Engineering Fatty Acid Overproduction in Escherichia coli for NextGeneration Biofuels
by
Rebecca M. Lennen
A dissertation submitted in partial fulfillment of
the requirements for the degree of
Doctor of Philosophy
(Chemical and Biological Engineering)
at the
UNIVERSITY OF WISCONSIN-MADISON
2012
Date of final oral examination: 8/3/2012
This dissertation is approved by the following members of the Final Oral Committee:
Brian F. Pfleger, Assistant Professor, Chemical and Biological Engineering
Jennifer L. Reed, Assistant Professor, Chemical and Biological Engineering
Sean P. Palecek, Professor, Chemical and Biological Engineering
Regina M. Murphy, Smith-Bascom Professor, Chemical and Biological Engineering
Douglas B. Weibel, Assistant Professor, Biochemistry
 Copyright by
Rebecca Lennen
2012
All Rights Reserved
i
Engineering Fatty Acid Overproduction in
Escherichia coli for Next-Generation Biofuels
Rebecca M. Lennen
Under the supervision of Professor Brian F. Pfleger
At the University of Wisconsin-Madison
Abstract
In order to replace dwindling petroleum resources and to curb emissions of CO2, it is
desirable to develop renewable substitutes for liquid transportation fuels and petrochemical
feedstocks. High energy density liquid fuels are currently required for the heavy transportation
sector, where it is not technologically feasible to store sufficient electrical power onboard.
Hydrolysis of lignocellulosic biomass to its component sugars for use as a carbon feedstock for
microbial conversion is among the most promising potential routes to producing renewable
liquid fuels at industrial scale. One biochemical pathway for producing medium- to long-chain
hydrocarbons with the potential to serve as replacements for petrodiesel is fatty acid biosynthesis,
which ordinarily produces saturated or monounsaturated acyl chains destined for incorporation in
phospholipids or lipid storage bodies as oils.
In this thesis, a metabolic engineering strategy is demonstrated in the versatile bacterium
Escherichia coli that hijacks the native fatty acid biosynthesis pathway to overproduce free fatty
acids of predominantly C12-C14 carbon chain length. Expression of an acyl-acyl carrier protein
(ACP) thioesterase to hydrolyze free fatty acids provides a product sink for feedback-inhibitory
acyl-acyl carrier protein (ACP) intermediates, resulting in deregulation of flux through fatty acid
biosynthesis. Up to 41 percent of the maximum theoretical yield of free fatty acids from glucose
ii
or glycerol has been achieved, which exceeds yields reported in most similar work and is the
highest reported yield of medium-chain length free fatty acids. It was also demonstrated that
free fatty acids can be readily catalytically decarboxylated to alkanes, providing a hybrid
biological and catalytic route to synthesize high-energy-density liquid fuel from renewable
carbon substrates.
To be economically viable, it is essential to metabolically engineer the microorganism and
optimize process conditions to produce the desired compound at close to the maximum
theoretical yield while also maintaining high productivities and titers. This is particularly true
for low-price, high-volume markets such as fuels. To achieve higher yields, productivities, and
titers, metabolic and process barriers must be iteratively discovered and overcome using
available molecular biology tools and traditional bioprocess engineering.
In this thesis, a
physiological characterization coupled with a functional genomics study was performed to
identify perturbations to cellular metabolism and physiology caused by free fatty acid production.
Cells in overproducing cultures exhibited greatly reduced viability, increased cell lysis, and a
large increase in unsaturated membrane lipids. Functional genomics analysis indicated the onset
of membrane stress responses, impaired proton motive force and aerobic respiration, and a strong
transcriptional response aimed toward but apparently unable to correct for increased unsaturated
acyl-ACP abundance.
Using the findings from the functional genomics analysis, new hypotheses were developed for
alleviating the toxic effects of endogenously produced free fatty acids.
One strategy
implemented a screen to identify and overexpress proteins involved in free fatty acid secretion,
with the goal of increasing efflux rates. A number of native efflux pumps were identified that
appear to confer increased resistance to medium-chain length free fatty acids, however
iii
overexpression of these pumps could not improve production or measures of viability over
strains with basal expression.
The second strategy was a proof-of-concept that involved
expressing an acyl-ACP thioesterase with a higher substrate specificity for unsaturated acylACPs, which resulted in a greatly reduced unsaturated membrane lipid content and correlated
with a reduced degree of cell lysis. Conversely, eliminating the native repressor protein for
genes required for unsaturated fatty acid biosynthesis in free fatty acid overproducing cultures
resulted in a greatly increased unsaturated membrane lipid content and a heightened degree of
cell lysis.
In summary, the work presented in this thesis demonstrates a viable strategy for producing
precursors for petrodiesel replacements and commodity chemicals, reveals severe physiological
stresses associated with this strategy, suggests two underlying causes of these stresses, and
reports on efforts to correct these underlying causes. While success was attained in restoring a
degree of fitness to free fatty acid producing cells, further work is needed to identify additional
barriers that limit yields and titers to current levels.
iv
Acknowledgments
First and foremost, I would like to thank my advisor, Professor Brian Pfleger, for his valuable
guidance through the years, and for teaching me just about everything I know about molecular
biology and synthetic biology, beginning with daily sessions in his office with Dan Agnew on
the basics of PCR and cloning. It was a good experience to be allowed the intellectual freedom
to pursue many of my own ideas, although some of those might have gotten inadvertently pushed
up in priority in the face of a list of tasks that would have taken another 5 years to finish, if you
get my drift. But thanks for not noticing. I also really appreciate allowing me to go off to
California for 3 months to complete an industrial internship at Genencor. That was definitely a
time that I wouldn't want to have missed, and I know it was a sacrifice to lose a summer's worth
of experiments.
Secondly, I'd like to thank all of the undergraduates who have worked with me (in order of
last to first): Max Kruziki, Kritika Kumar, Amy Claas, Gina Lewin, and Anna Mielke. And a
few other students who I had the opportunity to work with from back in the early days of the lab,
when it was a smaller place: Phil Angart, Scott Walhovd, and Asif Rahman. I'd like to believe it
wasn't too horrible an experience, as many of you stuck around for multiple semesters when you
could, and labored for free which was better of you than I ever would have done, and hope you
learned a lot and have all gone on to good things. And to Max, who has stuck it out with me for
nearly 2 years: have a great first semester at University of Minnesota - it'll be tough but I think
your success is guaranteed. I also had the luck to work with a few talented first year graduate
rotation students from other departments: Melanie Spero, Ted Mattes, and Anna Baker.
I would also like to express my thanks toward everyone else who I have worked with in the
lab over the years, in particular all of the other grad students and former and current postdocs
v
who make things interesting and sometimes assist in the fight against the forces of entropy: Dan
Agnew, Matt Begemann, Matt Copeland, Spencer Hoover, Andrew Markley, Daniel MendezPerez, Mark Politz, Jackie Rand, and Tyler Youngquist. I still haven't figured out who stole my
primers and phage lysates, but whoever it was, you are forgiven. It wasn't actually that big of a
deal. And special thanks to Victor Orler and Daniel Mendez-Perez for providing pBAD18-desC
in Chapter 6.
I also had the ability to work with a large number of incredibly helpful collaborators and to
discuss specific research topics and problems, and with others who have donated their own time
to train me on various techniques. These include in no particular order (and I apologize if I'm
forgetting anyone, and I undoubtedly am): Alan Higbee, Mary Lipton, Kristin Burnum, Bob
Zinkel, Ruwan Ranatunga, Tyler Wittkopp, Wes Marner, Syd Withers, Mike Donath, Jason
Peters, Dave Keating, Ryan West, Drew Braden, Jeff Gardner, Jay Lemke, Kyle Boldon, John
Ohlrogge, Mike Schwalbach, and Trey Sato.
Finally, thanks Tibi, for keeping me sane and caring for me and always being there, even if
mostly from a distance, during this very difficult time in my life. I'm really not sure what I
would have done without you. I hope we'll have a good life together wherever it may be.
vi
Table of Contents
Abstract…………………………………………………………………………………………………..... i
Acknowledgments…………………………………………………………….…………………………..iv
Table of Contents……………………………………………………………………………..................vi
List of Figures…………………………………………………………………………………………….. x
List of Tables………………………………………………………………………………………........ xiv
Chapter 1: Fatty acids and next generation biofuels…………………………………………. …1
1.1 Motivation and routes for biofuels…………………………………………………………….. 1
1.2 Fatty acid biosynthesis and degradation pathways………………………………………… 8
1.3 Regulation of fatty acid biosynthesis and degradation……………………………………. 13
1.4 Free fatty acids (FFAs)……………………………………………………………………….. 17
1.4.1 Survey of acyl-ACP thioesterases………………………………………...……….18
1.4.2 Metabolic engineering of FFA production…………………………………………21
1.5 Other useful products derived from fatty acid biosynthesis……………………............... 27
1.6 Conclusion……………………………………………………………………………………... 34
Chapter 2: Development of a medium-chain length fatty acid overproducing strain of
E. coli and a process for alkane production………………………….…………….……………. 35
2.1 Introduction…………………………………………………………………………………….. 36
2.2 Materials and methods……………………………………………………………………….. 38
2.2.1 Maximum theoretical yield determination……………………………….............. 38
2.2.2 Strain construction………………………………………………………………….. 38
2.2.3 Gene synthesis…………………………………………………………….............. 40
2.2.4 Plasmid construction………………………………………………………………...41
2.2.5 Cell transformation, media, and growth…………………………………………...43
2.2.6 Quantitative PCR for determination of copy number……………………………. 43
2.2.7 Fatty acid extraction and methylation…………………………………………….. 44
2.2.8 Gas chromatography/mass spectrometry of fatty acid methyl esters…………. 44
2.2.9 Decane extraction of fatty acids………………………………………….............. 45
2.2.10 Catalytic decarboxylation…………………………………………………………... 45
2.3 Results and discussion……………………………………………………………………….. 47
2.3.1 Determination of maximum theoretical yield of FFAs…………………………… 47
2.3.2 Initial strain construction and optimization of BTE copy number………………. 49
2.3.3 Co-overexpression of acetyl-CoA carboxylase………………………………….. 54
2.3.4 Extraction and conversion of fatty acids to alkanes…………………………….. 55
2.4 Conclusion……………………………………………………………………………………... 57
Chapter 3: Functional genomics study of FFA-overproducing E. coli………………………59
3.1 Introduction…………………………………………………………………………………….. 61
3.2 Materials and methods………………………………………………………………………...64
3.2.1 Strains, plasmids, enzymes, media, and bacterial cultivation…………………. 64
3.2.2 RNA sample preparation…………………………………………………………… 67
3.2.3 cDNA synthesis and hybridization………………………………………………....67
3.2.4 Microarray data analysis…………………………………………………...............68
3.2.5 Quantitative RT-PCR (qPCR)…………………………………………….............. 69
3.2.6 Protein sample preparation…………………………………………………………69
vii
3.2.7 Peptide analysis by capillary liquid chromatography/mass spectrometry…….. 70
3.2.8 Proteomics data analysis……………………………………………………..……. 70
3.2.9 Fatty acid extraction and analysis………………………………………………… 71
3.2.10 Glucose analysis……………………………………………………………………. 71
3.2.11 Acetate analysis…………………………………………………………………….. 72
3.2.12 Cell viability measurements from plate counts………………………………….. .72
3.2.13 Microscopy…………………………………………………………………………... 73
3.2.14 SYTOX flow cytometry assays……………………………………………………..73
3.2.15 Membrane polarization assays……………………………………………………. 74
3.2.16 Additional strain engineering………………………………………………………. 74
3.3 Results…………………………………………………………………………………………. 76
3.3.1 Cell viability and morphology……………………………………………………….76
3.3.2 Cultivation of E. coli for functional genomics study………………………………81
3.3.3 FFA overproduction increases the long-chain unsaturated fatty acid………….83
3.3.4 DNA microarray analysis……………………………………………………………85
3.3.5 Differential proteomics analysis…………………………………………………… 88
3.3.6 Induction of phage shock proteins………………………………………………… 90
3.3.7 Induction of MarA/Rob/SoxS regulon…………………………………………….. 91
3.3.8 Induction of other envelope stress responses and acid resistance response
in batch cultures…………………………………………………………………….. 94
3.3.9 Changes in energy metabolism…………………………………………………… 95
3.3.10 Decreased expression of unsaturated fatty acid biosynthesis pathway………. 97
3.3.11 Increased expression of β-oxidation genes……………………………………… 99
3.3.12 Differences in the acetate profile of fatty acid overproducing cultures………. 101
3.3.13 Increased expression of the AccB and AccC subunits of acetyl-CoA ………. 103
3.3.14 Carbohydrate transport and catabolism pathways…………………………….. 104
3.3.15 Decreased expression of PhoB regulon………………………………………....105
3.3.16 Initial attempts at functional genomics guided strain engineering…………….106
3.4 Discussion……………………………………………………………………………………..110
3.5 Conclusion……………………………………………………………………………………. 115
Chapter 4: Identification of transport proteins involved in FFA efflux in E. coli…………122
4.1 Introduction…………………………………………………………………………………… 122
4.2 Materials and methods……………………………………………………………………….126
4.2.1 Chemicals, reagents, enzymes, and oligonucleotide primers…………………..126
4.2.2 Strain construction…………………………………………………………………...127
4.2.3 Plasmid construction………………………………………………………………...128
4.2.4 Cell cultivation………………………………………………………………………..129
4.2.5 Cell viability measurements from plate counts……………………………………129
4.2.6 SYTOX flow cytometry assays……………………………………………………..130
4.2.7 Fatty acid extraction and analysis………………………………………………….130
4.2.8 Minimum inhibitory concentration (MIC) assays………………………………….131
4.3 Results…………………………………………………………………………………………132
4.3.1 Viability analysis of single gene/operon deletion strains……………………...…132
4.3.2 Fatty acid titers of single gene/operon deletion strains…………………………. 136
4.3.3 MIC of exogenous FFAs in single gene/operon deletion strains………………. 138
4.3.4 Viability analysis of double transporter gene/operon deletions…………………140
4.3.5 Fatty acid titers of double transporter gene/operon deletions…………………..141
4.3.6 Functional validation of drug efflux pump overexpression constructs………… 143
4.3.7 Effects of drug efflux pump overexpression in FFA overproducing strains……145
viii
4.3.8 MICs of exogenous free fatty acids toward strains overexpressing drug
efflux pumps…………………………………………………………………………. 149
4.4 Discussion……………………………………………………………………………………. 151
4.5 Conclusion……………………………………………………………………………………. 157
Chapter 5: Improved tolerance of E. coli toward FFA overproduction by modulation
of membrane acyl chain composition…………………………………………………………….158
5.1 Introduction…………………………………………………………………………………....158
5.2 Materials and methods……………………………………………………………………….162
5.2.1 Chemicals, reagents, enzymes, and oligonucleotide primers………………..... 162
5.2.2 Gene synthesis……………………………………………………………………… 163
5.2.3 Plasmid construction………………………………………………………………...163
5.2.4 Strain construction…………………………………………………………………...164
5.2.5 Cell cultivation………………………………………………………………………..165
5.2.6 SYTOX flow cytometry assays……………………………………………………..166
5.2.7 Fatty acid extraction and analysis………………………………………………… 166
5.2.8 RNA extraction and qPCR…………………………………………………………. 167
5.3 Results…………………………………………………………………………………………169
5.3.1 Effect of fabR deletion on unsaturated membrane-bound fatty acids………….169
5.3.2 Effect of fabR deletion on cell membrane integrity……………………………… 173
5.3.3 Effect of fabR deletion on FFA production……………………………………….. 174
5.3.4 Selection and expression of a predominantly medium-chain length
unsaturated acyl-ACP thioesterase………………………………………………..174
5.3.5 Unsaturated long-chain fatty acid biosynthesis in BTE and GeoTE
expressing cultures…………………………………………………………………. 178
5.3.6 Cell growth and viability analysis of GeoTE and BTE expressing cultures……181
5.3.7 Fatty acid production of GeoTE and BTE expressing cultures…………………184
5.4 Discussion…………………………………………………………………………………... 184
5.5 Conclusion…………………………………………………………………………………...188
Chapter 6: Conclusions and future directions………………………………………………… 190
6.1 Summary of findings………………………………………………………………………… 190
6.2 Recommendations for future work………………………………………………………….193
6.2.1 Optimization and engineering of efflux pump overexpression…………………. 193
6.2.2 Synthetic biology control of membrane lipid content……………………………. 197
6.2.2.1 Materials and methods…………………………………………………….. 200
6.2.2.1.1 Strain and plasmid construction…………………………………. 200
6.2.2.1.2 Bacterial cultivation conditions……………………………………201
6.2.2.1.3 Other methods…………………………………………………….. 202
6.2.2.2 Results………………………………………………………………………. 202
6.2.2.2.1 Functional expression of Cti in E. coli……………………………202
6.2.2.2.2 Effect of Cti on FFA overproduction…………………………….. 203
6.2.2.2.3 Functional expression of two phospholipid desaturases in
E. coli……………………………………………………………….. 205
6.2.2.2.4 Complementation of fabA with desaturases……………………. 206
6.2.2.2.5 Further progress in building a strain with rewired unsaturated
fatty acid biosynthesis…………………………………………….. 208
6.2.2.3 Discussion and suggested future experiments…………………………. 209
6.2.3 Further investigations in metabolism………………………………………………211
ix
References…………………………………………………………………………………………….. 213
Appendices…………………………………………………………………………………………….. 232
Appendix I…………………………………………………………………………………………. 232
Appendix II………………………………………………………………………………………….237
Appendix III…………………………………………………………………………………………239
Appendix IV………………………………………………………………………………………...244
x
List of figures
Figure 1.1
Recent historical and predicted range of crude oil prices to the year 2035………………. 2
Figure 1.2
Recent historical and predicted U.S. liquid fuel supply………………………………….. 2
Figure 1.3
Recent historical atmospheric CO2 concentrations and global surface temperature
anomalies…………………………………………………………………………………. 3
Figure 1.4
General schematic depicting biomass derived renewable fuel cycles……………………. 4
Figure 1.5
Schematic of the fatty acid biosynthesis pathway in E. coli…………………………….. 10
Figure 1.6
Schematic of fatty acid degradation (β-oxidation) pathway in E. coli………………….. 13
Figure 1.7
Allosteric regulation of fatty acid biosynthesis by acyl-ACPs…………………………. .15
Figure 1.8
Fatty-acid derived fuels and chemicals…………………………………………………. 28
Figure 2.1
Functional verification of fadD and araBAD gene deletions on agar plates containing
oleate and arabinose as sole carbon sources…………………………………………….. 40
Figure 2.2
Sequence of codon-optimized and custom-synthesized BTE gene……………………... 41
Figure 2.3
Representative mass spectra of fatty acid methyl esters present in methylated BTEexpressing culture extracts but not present in FAME standards………………………... 46
Figure 2.4
Mass spectrum of tridecane in a decarboxylated decane extraction…………………….. 48
Figure 2.5
Metabolic engineering strategy for overproduction of FFAs overlaid on the fatty acid
biosynthesis pathway……………………………………………………………………. 50
Figure 2.6
Growth curves and fatty acid titers of E. coli RL08 cultures harboring plasmids
expressing BTE and BTE-H204A………………………………………………………. 51
Figure 2.7
Representative GC/MS chromatograms of methylated fatty acid extracts from
strains expressing either BTE or BTE-H204A, and either empty vector or ACC……… 55
Figure 2.8
Growth and fatty acid production of strain RL08 harboring combinations of
plasmids pBAD33 or pBAD33-ACC, and pBAD35-BTE-H204A or pBAD35-BTE…...56
Figure 2.9
Chromatograms of decarboxylated fatty acid extracts in decane, and an envisioned
semi-continuous two-phase partioning bioreactor process for production of alkanes….. 58
Figure 3.1
Parallel strategies undertaken by our group to develop a second-generation
FFA-overproducing strain with improved production characteristics……………60
Figure 3.2
OD600-normalized abundances of BTE and BTE-H204A from shake flask cultures
determined by proteomics analysis……………………………………………………… 65
xi
Figure 3.3
Flow cytometry analysis (green fluorescence and forward scatter histograms) of
SYTOX stained cells expressing BTE and BTE-H204A……………………………….. 79
Figure 3.4
Phase contrast micrographs of cells expressing BTE and BTE-H204A………………… 80
Figure 3.5
Flow cytometry analysis (forward scatter and green fluorescence) of SYTOX stained
cells exposed to exogenously added lauric acid………………………………………… 82
Figure 3.6
Growth curves, sampling points, and glucose and fatty acid profiles of shake flask
cultures for the functional genomics study……………………………………………… 83
Figure 3.7
Growth curves and sampling times of fermentor cultures for the functional genomics
study……………………………………………………………………………………... 84
Figure 3.8
Analysis of C16-C18 fatty acids extracted from shake flask cultures…………………….. 85
Figure 3.9
C8-C14 fatty acid titers and percent unsaturated C16-C18 fatty acids from fermentor
cultures…………………………………………………………………………………... 86
Figure 3.10
Heat map of gene expression values at each shake flask culture sampling time with
hierarchical clustering analysis………………………………………………………….. 87
Figure 3.11
Cell polarization flow cytometry assay of FFA-overproducing and control cultures…... 92
Figure 3.12
Schematic of regulation of unsaturated biosynthesis by FabR………………………….. 98
Figure 3.13
Extracellular acetate concentrations from EZ glucose shake flask cultures expressing
BTE or BTE-H204A…………………………………………………………………….101
Fig. 3.14
Fatty acid titers of BTE and BTE-H204A expressing strains with deleted transcription
factors or selected overexpressed genes……………………………………………….. 109
Figure 4.1
Determination of threshold for non-intact SYTOX Green stained cells from green
fluorescence histograms……………………………………………………………….. 133
Figure 4.2
Normalized CFUs and percent intact cells for single gene deletions in strains TY06
and TY05………………………………………………………………………………. 135
Figure 4.3
Scatter plots of normalized CFUs versus percent intact cells by SYTOX Green
staining…………………………………………………………………………………. 136
Figure 4.4
Total fatty acid titers in TY05 deletion strains after 8 h, and in baseline and ∆tolC
strains after 24 h……………………………………………………………………….. 137
Figure 4.5
MIC assay for octanoate and decanoate against TY05 and selected single deletions in
TY05…………………………………………………………………………………… 139
Figure 4.6
Normalized CFUs and percent intact cells for double efflux pump deletions in strain
TY06 and TY05………………………………………………………………………... 142
xii
Figure 4.7
Plate reader growth curves of acrAB emrAB double deletion strains and negative
control strains………………………………………………………………………….. 143
Figure 4.8
Total fatty acid titers for double efflux pump deletions in TY05 and TY06………….. .144
Figure 4.9
MIC assay for SDS against TY05 ∆acrAB overexpressing selected efflux pump
system components…………………………………………………………………….. 146
Figure 4.10
Percent intact cells in TY05ara and TY05ara acrAB::kan overexpressing acrAB on a
plasmid…………………………………………………………………………………. 148
Figure 4.11
Percent polarized and percent intact cells in TY05ara overexpressing selected efflux
pumps…………………………………………………………………………………... 149
Figure 4.12
Fatty acid titers from TY05ara overexpressing selected efflux pumps………………... 150
Figure 4.13
MIC assay for octanoate and decanoate against TY05ara and TY05ara acrAB::kan
overexpressing selected efflux pumps…………………………………………………. 151
Figure 5.1
Alignment of amino acid sequence of BTE with bacterial acyl-ACP thioesterases
from Geobacillus sp. Y412MC10 (GeoTE) and Clostridium thermocellum (ClosTE)... 164
Figure 5.2
Comparative analysis of acid and base catalyzed FAME preparations from cultures
expressing BTE and GeoTE-H173A…………………………………………………... 170
Figure 5.3.
Normalized transcript levels of fabA and fabB in strains with deleted or overexpressed
fabR expressing BTE or BTE-H204A…………………………………………………. 171
Figure 5.4
Percent unsaturated C16-C18 fatty acids in strains with deleted or overexpressed fabR
expressing BTE or BTE-H204A……………………………………………………….. 172
Figure 5.5
Percent intact cells in fabR+ and ∆fabR strains expressing BTE or BTE-H204A…….. .174
Figure 5.6
Effect of fabR deletion on C8-C14 (predominantly free) fatty acid titer produced in
BTE-expressing cultures……………………………………………………………….. 175
Figure 5.7
Total fatty acid analysis of functional and mutagenized GeoTE and ClosTE…………. 177
Figure 5.8
Percentage unsaturated membrane lipids in cultures expressing combinations of BTE
and GeoTE……………………………………………………………………………... 179
Figure 5.9
Normalized transcript levels of fabA and fabB in strains expressing combinations of
BTE and GeoTE……………………………………………………………………….. 181
Figure 5.10
Percent intact cells in strains expressing combinations of BTE and GeoTE………….. 182
Figure 5.11
Plate reader growth curves of strains expressing combinations of BTE and GeoTE….. 182
xiii
Figure 5.12
Fatty acid titers and compositions in strains expressing combinations of BTE and
GeoTE………………………………………………………………………………….. 185
Figure 6.1
SecYEG and YidC interactions with nascent peptides to achieve insertion and
assembly in the inner membrane………………………………………………………. 195
Figure 6.2
Detail of erythromycin binding in the AcrB proximal pocket, and schematic of
substrate uptake and extrusion by AcrB……………………………………………….. 196
Figure 6.3
Three envisioned strategies for modulation of membrane fatty acids in BTEexpressing strains……………………………………………………………………… .199
Figure 6.4
Chromatograms of FAMEs extracted from cultures expressing Cti and an empty
vector control…………………………………………………………………………... 203
Figure 6.5
Percent composition by mass of saturated, cis-unsaturated, trans-unsaturated, and
cyclic fatty acids from cultures expressing Cti and an empty vector control………….. 204
Figure 6.6
Green fluorescence histograms from SYTOX Green staining and plate counts from
strains expressing Cti or an empty vector control, and BTE or BTE-H204A…………. 205
Figure 6.7
Percent composition by mass of saturated, cis-unsaturated, trans-unsaturated, and
cyclic C16-C18 fatty acids from cultures expressing combinations of Cti and BTE or
BTE-H204A……………………………………………………………………………. 206
Figure 6.8
Chromatograms of FAMEs extracted from cultures expressing DesA and an empty
vector control…………………………………………………………………………... 207
Figure 6.9
Chromatograms of FAMEs extracted from a culture expressing DesC and an empty
vector control…………………………………………………………………………... 207
Figure 6.10
Complementation of growth in a fabA deletion by DesA……………………………… 208
Figure 6.11
Map of template plasmid for integration of desA onto the chromosome……………… 209
xiv
List of tables
Table 1.1
Calculated maximum theoretical yields of selected FFAs on various carbon sources….. 22
Table 1.2
Literature summary for FFA titers, yields, and percent of maximum theoretical
yield……………………………………………………………………………………... 24
Table 2.1
Copy numbers determined for BTE and BTE-H204A harboring plasmids by qPCR….. .52
Table 3.1
Viable cell counts in BTE and BTE-H204A-expressing shake flask cultures………….. 77
Table 3.2
Cell viability analysis from exogenous addition of lauric acid…………………………. 81
Table 3.3
qPCR measurement of fold-changes in expression of selected genes in shake flask
cultures………………………………………………………………………………….. 88
Table 3.4
Selected protein fold-changes between BTE and BTE-H204A-expressing strains…….. 91
Table 3.5
Viable cell counts from cultures of transcription factor deletion strains expressing
BTE-H204A or BTE…………………………………………………………………… 108
Table 3.6
Viable cell counts from cultures co-expressing BTE-H204A or BTE and selected
genes cloned in pBAD18………………………………………………………………. 109
Table 3.7
Selected transcript fold-changes from shake flask and fermentor cultures……………. 117
Table 3.8
Proteins differentially expressed in EZ glucose shake flask cultures………………….. 119
Table 3.9
Proteins differentially expressed in EZ glycerol fermentor cultures…………………... 121
Table 4.1
Genes targeted with possible free fatty secretion or efflux roles………………………. 125
Table 5.1
Viability analysis of strains expressing combinations of BTE and GeoTE……………. 183
1
Chapter 1: Fatty acids for next-generation biofuels
1.1 Motivation and routes for biofuels
The production of transportation fuels from renewable sources is necessary to meet
continuing demand in the face of dwindling petroleum supplies while also curbing the release of
greenhouse gases. The Energy Information Administration (EIA) predicts a likely continuation
of increasing oil prices in the near future (Figure 1.1), with a decreasing U.S. supply of liquid
fuels originating from both domestic and imported sources (Figure 1.2). The supply gap is
anticipated to be filled by increased production of liquids from coal, natural gas, and a more than
doubling of the production of biofuels from 2010 levels. In addition to inherent supply issues,
atmospheric CO2 concentrations continue to increase and exceed any maximum prehistoric level
from the past 800,000 years (~300 ppm) as a result of the extraction and combustion of fossil
fuels. Both prehistoric cycles in CO2 concentration and the recent and unprecedented rise in CO2
concentrations in modern times trend with increasing temperatures, which are generally
recognized to be contributing to an increased frequency of violent weather patterns and local
climate extremes (Figure 1.3). To avert catastrophic environmental, economic, and humanitarian
consequences, it is essential to develop renewable energy sources to displace the use of fossil
fuels. In order to meet this challenge, it is likely that a combination of renewable and non-fossil
fuel sources, including solar, wind, geothermal, and nuclear energy will need to be expanded for
electricity generation, while biomass derived chemicals can provide portable transportation fuels
and feedstocks for the chemical industry.
2
History
2010
Projections
High Oil Price
AEO2012 Reference
Low Oil Price
Source: EIA, Annual Energy Outlook 2012
Figure 1.1 Recent historical and predicted range of crude oil prices to the year 2035. The US
Energy Information Administration predicts a continuing upward trend in in oil prices (y-axis in real 2010
dollars per barrel) from approximately $100/barrel in 2012 to $150/barrel. Prices of $200/barrel are the
predicted high end oil price in 2035. Source: U.S. Energy Information Administration Annual Energy
Outlook 2012 (www.eia.gov)
U.S. liquid fuels supply
million barrels per day
History
2010
Projections
5%
Biofuels including imports
10%
Natural gas plant liquids
15%
Petroleum production
36%
12%
36%
Liquids from coal
49%
Net petroleum imports
1%
36%
Source: EIA, Annual Energy Outlook 2012
Figure 1.2 Recent historical and predicted U.S. liquid fuel supply. The EIA predicts a decreasing
U.S. supply of fuels derived from domestic petroleum production and foreign petroleum imports between
2012 and 2035, with an increase in the proportion supplied by biofuels from 5% in 2010 to 12% by 2035.
Source: U.S. Energy Information Administration Annual Energy Outlook 2012 (www.eia.gov)
Renewable fuels are derived from carbon present in plant or algal biomass in the form of
either structural sugar polymers such as cellulose and hemicelluloses, energy storage sugar
polymers in the form of starch, aromatic polymers in the form of lignin, oils stored in plant seeds
3
Figure 1.3 Recent historical atmospheric CO2 concentrations and global surface temperature
anomalies. (Left) Atmospheric CO2 concentration (ppm) has been steadily increasing since 1960.
Source: National Oceanic and Atmospheric Administration, Earth System Research Laboratory
(www.esrl.noaa.gov/gmd/ccgg/trends/#mlo_full). (Right) Global surface temperature anomaly from 1880
to 2006 compared to the baseline period between 1951-1980. Temperatures have been on an upward
trend that correlates with increased atmospheric CO2. Source: NASA Goddard Institute for Space Studies
(data.giss.nasa.gov/gistemp/graphs_v3/)
or vegetative tissue, or algal lipid bodies.
Glycerol, a byproduct of transesterification of
triacylglycerols to produce biodiesel, is also a fermentable and inexpensive carbon source for
small scale renewable chemical production. A typical renewable fuel cycle is shown in Figure
1.4.
While the production of ethanol derived from sugarcane and corn starch, and the production of
biodiesel from soybean, rapeseed, and palm oils ("first generation biofuels") has seen significant
global implementation, the scale of production is intrinsically limited as only the plant seeds are
utilized from corn and biodiesel crops. Negative public perceptions are also pervasive toward
first generation biofuels, as growing these crops leads to a direct competition between food and
energy markets [Duffield 2007, Banerjee 2011, Graham-Rowe 2011]. Destruction of tropical
ecosystems has also been implicated in expanded biodiesel production from soybeans in Brazil
[Altieri 2009] and from palm oil in southeast Asia [Stone 2007].
4
Pre-treatment &
Saccharification
Biomass
CO2
Hexose
Pentose
Gasification
Fermentation
h
CO, H2
Pyrolysis
Bio-oil
Algae
Catalysis
Crude
Biofuels
Chemicals
Distribution
& Use
Combustion
Refining &
Purification
Figure 1.4 General schematic depicting biomass derived renewable fuel cycles. Plants and algae
fixate CO2 through photosynthetic processes to generate plant and algal biomass. Plant biomass can be
broken down by a number of routes described in the text and used either as a carbon source for microbial
fermentation, or in various catalytic processes to generate crude biofuels and chemicals. Alternatively,
algae can be engineered to directly produce biofuels and chemicals. Combustion of fuels completes the
cycle by releasing CO2 back into the atmosphere.
"Second generation" biofuels aim to overcome the economic and environmental drawbacks of
first generation biofuels and likely provide the only feasible strategies to enable a complete
replacement of petroleum-derived transportation fuels with renewable biofuels [Hill 2006,
Schmer 2008]. One alternative source of biofuels is the cultivation of lipid accumulating or
alkane secreting algae, which could be cultivated in photobioreactors or in large open seawater
ponds on non-arable land and would utilize CO2 directly via photosynthesis.
Various
technological challenges exist including the development of cost-effective photobioreactors,
contamination of open ponds, and cost-effective recovery of lipids or other desired compounds
from algal biomass. Several companies are currently pursuing a variety of routes toward algal
biofuels and their level of success should become more clear in the near future.
5
Another source of second generation biofuels is the utilization of plant structural polymers,
so-called cellulosic or lignocellulosic biofuels. A variety of chemical and biological process
technologies enable the generation of molecules of fuel value from these raw materials. Some
examples include pyrolysis to yield a deoxygenated bio-oil that can be catalytically upgraded to
molecules of fuel value [Mohan 2006, Alonso 2010]; gasification to yield syngas (CO and H2),
which can then be used in Fischer-Tropsch processes to obtain higher chain length carbon units
[Dry
2002];
ionic
liquid
treatment
with
metal
chloride
catalysts
to
obtain
5-
hydroxymethylfurfural, a versatile intermediate for production of further deoxygenated
compounds [Zhao 2007, Binder 2009, Alonso 2010]; and acid treatment of biomass at high
temperatures to obtain furfurals and levulinic acid, which can be dehydrated to γ-valerolactone,
another useful intermediate for production of deoxygenated building blocks such as butene
[Alonso 2010]. Another major route for biomass utilization, following various pre-processing
and pre-treatment steps, utilizes enzyme cocktails to hydrolyze cellulose and hemicellulose
polymers into their component sugar monomers or oligomers (primarily glucose, cellobiose,
xylose, and arabinose) [Chundawat 2011]. These enzyme cocktails contain carefully controlled
mixtures of endo- and exoglucanases, xylanases, β-glucosidases, cellobiohydrolase, and
numerous other enzymes (eg. Accellerase® 1500 product information sheet, Genencor, Palo Alto,
CA).
These released sugars can then serve as feedstocks for any variety of microbial
fermentation processes, through which native metabolism or metabolic engineering of
biochemical pathways enables the production of desired chemicals and fuels. The remainder of
this review and the subsequent work described in this dissertation is aimed toward the goal of
developing microbial conversion processes to produce high-energy density molecules of fuel
value from these lignocellulosic feedstocks.
6
While much focus in the research area of microbial conversion of hydrolyzed cellulosic
feedstocks has been placed on developing biomass-derived gasoline alternatives such as ethanol
and other short-chain alcohols such as n-butanol and isobutanol [Houghton 2006, Atsumi 2008,
Green 2011], higher energy density distillates such as diesel and jet fuel are needed by the heavy
transportation sector due to weight and range constraints. Unlike the case for personal vehicles,
these constraints also functionally prohibit the utilization of electricity as a power source in these
vehicles without significant advances in either battery or hydrogen storage technology for the
case of fuel cell vehicles. Diesel and jet fuel currently account for more than half of the world
refinery output destined for use in vehicles [Davis 2009]. Furthermore, diesel engines improve
fuel efficiency in small passenger vehicles over gasoline engines, and have already been widely
adopted in Europe. Medium- and long-chain hydrocarbons can potentially serve as replacements
for diesel, rendering them an attractive target for microbial production from lignocellulosic
feedstocks. Unlike ethanol, the low water solubility of longer carbon chain-length hydrocarbons
should result in reduced recovery costs and reduced toxicity in the fermentation broth due to
phase separation. These hydrocarbons are also more likely to be compatible with existing
transport and storage infrastructure and vehicle engines, and possess higher cloud points than
biodiesel blends [Knothe 2010, Röttig 2010], enabling year-round usage in all climates.
Two major biochemical pathways exist for production of highly reduced C8 or higher
molecules, both of which build up carbon-carbon bonds by an iterative series of condensation
reactions. These are the isoprenoid biosynthesis pathway, where precursor molecules from
central metabolism are used to generate isopentenyl diphoshate and its isomerized product,
dimethylallyl diphosphate. These in turn can be condensed as 5 carbon monomers to geranyl
diphosphate (C10), farnesyl diphosphate (C15), and geranylgeranyl diphosphate (C20) to ultimately
7
generate a range of unsaturated, branched, and highly reduced hydrocarbons such as isoprene,
myrcene, farnesene, geranylgeraniol, and phytoene [i.e. Kuzuyama 2002, Mijts 2004, Yoshikuni
2006, Yoon 2007, Wang 2011]. Cyclized products of useful fuel value such as limonene and
bisabolene can also be synthesized by heterologous expression of plant terpene synthases [Carter
2003, Reiling 2004, Peralta-Yahya 2011]. A number of excellent reviews describe metabolic
engineering of the isoprenoid pathway for biofuels applications [Fortman 2008, Rude 2009,
Peralta-Yahya 2010]. The second pathway is fatty acid biosynthesis, for which acetyl-CoA (or
rarely propionyl-CoA) serves as the precursor and for which long-chain, predominantly
unbranched aliphatic compounds are derived including common natural products such as
phospholipids and di- and triacylglycerols. A number of other natural products can also be
generated through this pathway including free fatty acids (FFAs), alkanes and olefins, fatty
alcohols, methyl ketones, and esters of fatty acids, which will be discussed in more detail in
Section 1.5. Compounds derived from the fatty acid biosynthesis pathway were selected as the
subject of investigation in the subsequent described work.
When developing a microbial conversion process for production of a target compound, it is
often possible to select a host organism which possesses the innate ability to synthesize the
compound of interest. For example, the bacterium Clostridium acetobutylicum naturally excretes
n-butanol as a fermentative byproduct, and the yeast Yarrowia lipolytica naturally accumulates
triacylglycerols as an energy storage medium. Metabolic engineering offers the opportunity to
genetically modify these organisms to optimize production of the naturally produced compound
via single gene or entire pathway manipulation. However, the ability to genetically engineer
these organisms is often hindered by a lack of synthetic biology tools (eg. inducible promoters
systems, expression and shuttle vectors, and chromosomal gene deletion tools), undeveloped
8
methodologies for efficient DNA transformation and recombination, and lack of knowledge
about the host organism's metabolism and regulation. Alternatively, common host organisms
such as Escherichia coli and Saccharomyces cerevisiae provide the advantage of decades of
well-developed genetic engineering and synthetic biology tools; understanding of their
metabolism, physiology, and gene regulation; and rapid and well-developed protocols for
transformation and recombination.
For these reasons, and also due to its diversity of carbon utilization (eg. ability to readily
metabolize pentose sugars) and rapid growth rate, Escherichia coli was selected as the
production organism in this study. The remainder of this chapter will focus on the biochemistry
and regulation of fatty acid biosynthesis and degradation in E. coli, and will review prior work
done in the area of engineering E. coli to produce diesel substitutes via fatty acid biosynthesis.
1.2 Fatty acid biosynthesis and degradation pathways
Fatty acid biosynthesis and degradation is best understood in Escherichia coli. The basic
pathways for fatty acid biosynthesis and fatty acid degradation in E. coli are shown in Figures
1.5 and 1.6, based on a decades-long accumulation of research that is well-summarized in
numerous reviews [Rawlings 1998, Rock 2002, Cronan 2008, Zhang 2008a, Zhang 2008b].
Enzymes of the fatty acid biosynthesis pathway act on acyl thioesters bound to a carrier protein.
Holo-acyl carrier protein (holo-ACP), which is synthesized from apo-ACP (AcpP) by addition of
a phosphopantetheine moiety from coenzyme A (CoA) by holo-ACP synthase (AcpS) [Polacco
1981]. The terminal thiol on the phosphopantetheine ligand is the reactive attachment point of
acyl thioesters. Fatty acid biosynthesis begins with condensation of acetyl-coenzyme A (CoA)
and CO2 to form malonyl-CoA, which is catalyzed by acetyl-CoA carboxylase (ACC) [Gucchait
9
1974]. ACC consists of 4 subunits, of which biotin carboxyl carrier protein (AccB), which is
post-translationally modified by addition of biotin by BirA, is carboxylated by addition of HCO3by biotin carboxylase (AccC). Acetyl-CoA carboxylase, consisting of a heterodimer of subunits
AccA and AccD, completes the second half-reaction by transferring the carboxyl group to acetylCoA to generate malonyl-CoA. All malonyl-CoA is directed toward fatty acid biosynthesis in E.
coli [Li 1992].
The malonyl group is transferred from CoA to ACP by malonyl-CoA-ACP transacylase
(FabD) [Harder 1974, Heath 1995a], and malonyl-ACP serves as a 2-carbon extender unit in
each cycle of fatty acid elongation. A number of other initiation steps have been shown to
generate the first β-ketoacyl-ACP in the elongation cycle, acetoacetyl-ACP in vitro. However
the primary pathway of physiological relevance involves condensation of malonyl-ACP with
acetyl-CoA by β-ketoacyl-ACP synthase III (FabH) to form acetoacetyl-ACP [Jackowski 1987,
Tsay 1992]. E. coli, as in most bacteria and plant chloroplasts, utilizes a Type II elongation
system in which the modifications to the growing fatty acid chain are made by individual
cytosolic proteins, as opposed to the Type I systems in animals and fungi which consist of a large
multi-subunit megasynthase complex [Cronan 2008]. First, the β-ketoacyl-ACP is reduced with
NADPH by β-ketoacyl-ACP reductase (FabG) to a β-hydroxyacyl-ACP [Toomey 1966], then
dehydrated by FabZ to the trans-2-enoyl-ACP [Birge 1972, Mohan 1994], and finally reduced by
enoyl-ACP reductase (FabI) to acyl-ACP [Weeks 1968, Bergler 1994, Bergler 1996].
Furthermore, a shunt for generating cis-enoyl-ACPs is catalyzed by FabA, in which βhydroxydecanoyl-ACP is converted to trans-2-decenoyl-ACP, which is also partially isomerized
to cis-3-decenoyl-ACP [Bloch 1971].
This cis unsaturation is retained through subsequent
rounds of elongation and incorporation into phospholipids.
Acyl-ACPs are then either
10
ATP, HCO3-
ADP, Pi
O
-
OOC N
O
HN
NH
NH
AccC
BCCP
S
biotin
S
carboxybiotin
BCCP
O
AccA, AccD
O
O
HO
S CoA
acetyl-CoA
O
FabD, FabB, FabH
O
S ACP
S CoA
FabD
malonyl-CoA
acetoacetyl-ACP
malonyl-ACP
Phospholipid biosynthesis
O
FabH
OH
CH3(CH2)x
Acyl-ACP
thioesterase
O
O
S ACP
CH3(CH2)x
S ACP
CH3(CH2)x
O
FabB or FabF
β-ketoacyl-ACP
acyl-ACP
ACP, CO2
NAD+
NADPH
cis-3-decenoyl-ACP
FabI
x=6
FabA
FabG
H2O
NADH
NADP+
O
CH3(CH2)x
OH
FabZ
S ACP
trans-2-enoyl-ACP
H 2O
CH3(CH2)x
O
S ACP
β-hydroxyacyl-ACP
Figure 1.5 Schematic of the fatty acid biosynthesis pathway in E. coli. Acetyl-CoA is converted to
malonyl-CoA by the four subunits of acetyl-CoA carboxylase (AccABCD; BCCP = AccB). MalonylCoA is converted to malonyl-ACP by malonyl-CoA:ACP transacylase (FabD), and to the acetoacetylACP, the first β-ketoacyl-ACP in the fatty acid elongation cycle, by multiple pathways catalyzed by FabD,
β-ketoacyl-ACP synthase III (FabH), and β-ketoacyl-ACP synthase I (FabB). The ketoacyl-ACP is
reduced twice and dehydrated once to yield an acyl-ACP in the elongation cycle by β-ketoacyl-ACP
reductase (FabG), enoyl-ACP reductase (FabI), and β-hydroxyacyl-ACP dehydratase (FabZ). The acylACP is then condensed with malonyl-ACP by FabB or β-ketoacyl-ACP synthase II (FabF). Cisunsaturated fatty acids are formed at the C10 chain length by 3-hydroxydecanoyl-ACP dehydrase (FabA)
which produces both cis and trans species, and FabB is essential for condensing cis-2-decenoyl-ACP.
Glycerol-3-phosphate acyltransferase (PlsB) and 1-acylglycerol-3-phosphate acyltransferase (PlsC) utilize
C16 to C18 acyl-ACPs as substrates for phospholipid biosynthesis. Acyl-ACP thioesterases hydrolyze
acyl-ACPs to yield FFAs.
11
condensed with another unit of malonyl-ACP by FabB or FabF to extend the acyl chain length
by two carbons [D'Agnolo 1975, Garwin 1980], or at the appropriate chain length (C14-C18)
become substrates for acyltransferases of phospholipid biosynthesis [Rock 1981, Kessels 1983].
FabB and FabF possess different substrate specificities, with FabB being essential due to its
capability of condensing cis-3-decenoyl-ACP [Feng 2009]. FabF is non-essential but greatly
reduced incorporation of cis-vaccenate (18:1∆11) into phospholipids results from its deletion,
rendering the cells unable to grow at a low temperature of 30°C due to a higher requirement for
unsaturated membrane lipids under these conditions [Garwin 1980a, Garwin 1980b]. This is due
to FabF having a much higher selectivity toward condensing cis-palmitoleoyl-ACP (16:1∆9)
than FabB [Edwards 1997].
Long-chain length acyl-ACPs (in addition to acyl-CoAs) are
substrates for two acyltransferases, PlsB and PlsC, as part of phospholipid biosynthesis [Zhang
2008a]. PlsB catalyzes the addition of the first acyl group to glycerol-3-phosphate, with the
favored acyl group substrate being a saturated C16 species [Rock 1981]. PlsC catalyzes the
addition of the second acyl group, favoring monounsaturated C16 and C18 species [Kessels 1983].
A more poorly characterized alternative pathway for phospholipid biosynthesis in E. coli
involves PlsX and YgiH [Zhang 2008a, Zhang 2008b]. PlsX appears to produce activated acylphosphates from acyl-ACPs, and YgiH may be able to utilize both acyl phosphates and acylACPs directly to generate the 1-acylglycerol-3-phosphate intermediate [Yoshimura 2007].
E. coli can also scavenge FFAs and grow on them as a sole carbon source through the βoxidation pathway [Clark 1996], which functions similarly to a fatty acid biosynthesis cycle in
reverse but using CoA thioester intermediates, and yields the versatile central metabolic
intermediate acetyl-CoA (Figure 1.6). In the context of target compounds derived from the fatty
acid biosynthesis pathway, β-oxidation primarily needs to be understood such that product
12
catabolism can be eliminated. However an alternative pathway for FFA production will be
discussed in section 1.4.2, wherein a hybrid β-oxidation cycle is reversed in direction to create an
anabolic process [Dellomonaco 2011].
Extracellular long-chain (C14 and greater) FFAs cannot diffuse across the outer membrane but
are efficiently imported to the periplasm by the outer membrane transporter FadL [Maloy 1981,
Black 1990] for which a structural mechanism was recently elucidated [Hearn 2009, Lepore
2011]. Conversely, FFAs shorter than C14 chain length may be able to diffuse through outer
membrane porins [Hearn 2009]. Transport of FFAs into and across the periplasm is poorly
understood, however the kinetics of intercalation into lipid bilayers and transmembrane flipping
is rapid [Black 2003, Kampf 2006]. FFAs in the outer leaflet of the inner membrane are
activated to acyl-CoAs by an acyl-CoA synthetase, FadD or FadK.
FadD has an in vitro
substrate chain length specificity that is highest for C10-C18 fatty acids [Kameda 1981], while
FadK has the highest specificity for C6-C8 fatty acids and primarily exhibits activity under
anaerobic conditions due to a trace level of expression under aerobic conditions [Campbell 2003,
Morgan-Kiss 2004]. Once FFAs are activated as CoA thioesters, acyl-CoA dehydrogenases
FadE, and putatively YdiO, catalyze the oxidation of acyl-CoA to 2-enoyl-CoA [Campbell 2002,
Campbell 2003]. Homologues FadB and FadJ possess multiple enzymatic activities including
independent active sites for enoyl-CoA hydratase and β-hydroxyacyl-CoA dehydrogenase
functions, catalyzing the formation of β-ketoacyl-CoA thioesters [O'Brien 1977, Campbell 2003].
FadA and FadI are homologous β-ketoacyl-CoA thiolases that complete one cycle of β-oxidation
[Yang 1983, Campbell 2003]. While FadB and FadA exhibit a preference for longer-chain
length species, FadJ and FadI exhibit a preference for short and medium-chain length species
[Campbell 2003].
13
O
OH
CH3(CH2)x
O
ATP
S CoA
acetyl-CoA
FadD
or
FadK
AMP + PPi
O
O
S CoA
CH3(CH2)x
acyl-CoA
S CoA
CH3(CH2)x
FadA or
FadI
O
β-ketoacyl-CoA
CoA
FAD
NADH
cis-3-enoyl-CoA
FadE or
YdiO
FadB
or FadJ
FadB or
FadJ
FADH2
NAD+
O
CH3(CH2)x
OH
FadB or FadJ
S CoA
trans-2-enoyl-CoA
H2O
CH3(CH2)x
O
S CoA
β-hydroxyacyl-CoA
Figure 1.6 Schematic of fatty acid degradation (β-oxidation) pathway in E. coli. FFAs are activated
for β-oxidation by an acyl-CoA synthetase (FadD or FadK under anaerobic conditions for shorter chain
lengths). A catabolic oxidation cycle consisting of dehydrogenation by FadE, hydration and
dehydrogenation by FadB or FadJ, and thiolation with CoA by FadA or FadI completes one turn of the
cycle and generates one molecule of acetyl-CoA.
1.3 Regulation of fatty acid biosynthesis and degradation
It was first found that cultures of E. coli starved of glycerol (to limit phospholipid
biosynthesis) exhibited a greatly decreased rate of production of acyl-ACPs [Jiang 1994]. Fatty
acid biosynthesis is also shut down by accumulation of the stationary phase alarmone, guanosine
tetraphosphate (ppGpp), and it was found that ppGpp accumulation inhibits PlsB, thereby
pointing to a product-regulated control point for fatty acid biosynthesis [Heath 1994].
Concomitantly, flux through fatty acid biosynthesis was found to be increased by overexpression
of a cytosolic form (lacking a periplasmic leader sequence) of E. coli thioesterase I (TesA')
[Jiang 1994], which hydrolyzes both acyl-CoAs and acyl-ACPs to generate FFAs [Cho 1993]. A
14
number of other heterologously expresssed acyl-ACP thioesterases from plants generate the
same effect, as will be discussed in more detail in section 1.5. It was postulated that acyl-ACPs
were feedback inhibiting unidentified enzymes in the fatty acid biosynthesis pathway. Later
work co-overexpressing cytosolic TesA (TesA') and the four subunits of ACC resulted in higher
levels of FFA production than overexpressing TesA' alone, indicating that ACC was the rate
limiting step and pointed toward ACC as being the enzyme in fatty acid biosynthesis that is
feedback-inhibited by acyl-ACPs [Davis 2000]. In vitro studies later confirmed that acyl-ACPs
directly inhibit ACC [Davis 2001], however the overall picture is not this simple, as acyl-ACPs
have also been found to inhibit FabH and FabI to lesser degrees [Heath 1996] (Figure 1.7). In
vitro reconstitution of a full cycle of fatty acid biosynthesis from acetyl-CoA and malonyl-CoA
to butyryl-ACP (using FabD, FabH, FabG, FabZ, and FabI) revealed FabI to catalyze the rate
limiting step [Heath 1995b]. A recent more thorough in vitro reconstitution that also included
FabB, FabZ, FabA, and TesA' to allow consecutive elongation cycles and a product sink for
acyl-ACPs, revealed that of the biosynthetic enzymes, only FabI and FabZ enhanced the rate of
fatty acid production in a dose-dependent manner [Yu 2011].
However, the full spectrum of
acyl-ACP mediated inhibition of all enzymes in fatty acid biosynthesis is difficult to obtain, and
virtually no studies have reliably established any kinetic parameters, due to there being dozens of
acyl-ACP intermediates (saturated acyl-ACPs, cis- and trans-enoyl-ACPs, β-ketoacyl-ACPs, and
β-hydroxyacyl-ACPs of between 4-18 carbon chain length) and their relative difficulty in
synthesizing and separating for detection.
Compared to allosteric regulation, relatively little is known about transcriptional and
translational level control of fatty acid biosynthesis. Levels of all four ACC subunit transcripts
have been found to correlate with growth rate [Li 1993], further supporting the role of ACC as
15
O
AccABCD
O
O
HO
S CoA
acetyl-CoA
O
FabD, FabB, FabH
O
S ACP
S CoA
malonyl-CoA
FabD
acetoacetyl-ACP
malonyl-ACP
Phospholipid biosynthesis
O
O
S ACP
CH3(CH2)x
S ACP
CH3(CH2)x
O
FabB or FabF
β-ketoacyl-ACP
acyl-ACP
FabG
FabI
O
CH3(CH2)x
OH
FabZ
S ACP
trans-2-enoyl-ACP
CH3(CH2)x
O
S ACP
β-hydroxyacyl-ACP
Figure 1.7 Allosteric regulation of fatty acid biosynthesis by acyl-ACPs. When acyl-ACPs
accumulate such as when phospholipid biosynthesis is inhibited, acyl-ACPs feedback inhibit
ACC, FabH, and FabI. FabI and FabZ are rate limiting steps of fatty acid biosynthesis when
malonyl-CoA is supplied in vitro.
the primary gatekeeper controlling flux through fatty acid biosynthesis. However the mechanism
of this transcriptional coordination remains unknown, as the subunits are located in three distal
operons (accA, accBC, and accD). AccB autoregulates transcription of accBC, most likely by
DNA binding within its promoter region [James 2004]. AccA and AccD form a heterodimer
which catalyzes the carboxyltransferase half-reaction of ACC, and it was recently shown that the
translation of the two subunits is coordinated by RNA binding by the AccAD complex, thereby
providing a feedback-inhibitory translational control mechanism [Meades 2009]. An increase in
malonyl-ACP levels during late-log to early stationary phase was observed when a plant acylACP thioesterase was heterologously expressed [Ohlrogge 1995], perhaps suggesting growth
rate control of transcription or translation of fabH and therefore entry to the elongation cycle,
despite the likely derepression of ACC under these conditions. Several genes involved in fatty
16
acid and phospholipid biosynthesis: plsX, fabH, fabD, fabG, acpP, and fabF, are co-localized
in a few experimentally verified overlapping transcription units, with stringent regulation from
one promoter [Podkovyrov 1996], therefore growth rate control of fabH would likely also imply
growth rate control of transcription or translation of many other fatty acid biosynthetic genes.
E. coli also has the capability of shifting the phospholipid-bound fatty acid composition in
response to environmental conditions. For example the membrane becomes more unsaturated at
reduced growth temperatures due to the cells needing to maintain their membrane fluidity [Marr
1962], possibly via a regulatory mechanism involving FabF. At the most basic level, unsaturated
fatty acid biosynthesis is controlled by a transcription factor FabR, which binds DNA upstream
of the promoter region of fabA and fabB [Zhang 2002]. While FabR has been shown to bind
both saturated and unsaturated acyl-ACPs, its DNA binding affinity is increased in the presence
of a higher ratio of unsaturated acyl-ACPs [Zhu 2009, Feng 2011]. Therefore when the balance
of unsaturated and saturated acyl-ACPs is disturbed, feedback inhibition restores the appropriate
level of unsaturated acyl-ACP biosynthesis. E. coli also possesses a cyclopropane fatty acid
synthase (Cfa) which adds a methyl group across cis-unsaturated C16 and C18 to form cyclic C17
and C19 species [Grogan 1997]. Expression of cfa is increased during stationary phase, and the
cyclic species appear to aid in acid exposure [Chang 1999], possibly by stabilizing the reactive
double bond against peroxidation.
Transcription of most genes and operons of the fatty acid β-oxidation pathway, including
fadE, fadD, fadIJ, fadL, and fadBA, are repressed by a transcription factor FadR. FadR in turn
releases its binding from DNA, thereby derepressing its regulon, when it binds acyl-CoAs that
would be expected to accumulate due to a basal level expression of fadD when FFAs are
available [DiRusso 1992]. Many genes of β-oxidation are also repressed by phosphorylated
17
ArcA, which is produced under microaerobic and anaerobic and conditions [Cho 2006], and
are putatively activated during the absence of rapidly metabolizable carbon sources, such as
glucose, by CRP-cAMP [Zheng 2004].
1.4 Free fatty acids (FFAs)
Free fatty acids (FFAs) are the most direct product that can be derived from fatty acid
biosynthesis in E. coli, as their production requires only heterologous expression of an acyl-ACP
thioesterase. As discussed in section 1.3, introduction of an efficiently expressed acyl-ACP
thioesterase by itself deregulates fatty acid biosynthesis and results in greatly increased carbon
flux through the pathway.
FFAs are themselves a useful commodity chemical for which
aliphatic medium- and long-chain lengths are currently most commonly derived from plant oils
such as coconut for C12-C14 FFAs and palm oil for C16-C18 FFAs [Johnson 1989, Gunstone 1997].
FFAs, and in particular medium-chain length FFAs, are commonly used for the production of
chemicals used in cleaning products, agrochemicals, biocidal agents, textile processing agents,
and polymer additives including compounds such as soaps (sodium laurate), anionic surfactants
(alkyl sulfates such as sodium dodecyl sulfate), and cationic surfactants (such as alkyl amines)
[Johnson 1989, Gunstone 1997]. FFA esterification by glycerol is one route for the synthesis of
monoglycerides, which are used as emulsifiers in food products [Gunstone 1997]. While not
currently large markets, medium-chain length FFAs have also been found to serve as effective
and non-toxic antimicrobial agents [Desbois 2010]. Active mosquito repelling fractions were
recently isolated from breadfruit inflorescences, and were found to contain a mixture of FFAs,
from which a blend of undecanoic acid, capric acid, and lauric acid was found to have a higher
18
efficacy than N,N-diethyl-m-toluamide (DEET) [Jones 2012]. Thus there is a potential for the
demand for FFAs to substantially increase.
Significant interest has also arisen in the catalytic community for developing an alternative
deoxygenation process for generating molecules of fuel value from triglycerides. This is due to
poor performance and storage properties of biodiesel (fatty acid methyl esters), the formation of
byproducts during the trans-esterification process including FFAs, and problems with catalyst
fouling and deactivation [Lestari 2009]. These processes involve FFA intermediates, so it is
feasible to consider sources of FFAs other than from triglyceride hydrolysis. Two possible
reaction pathways are hydrodeoxygenation and hydrodecarboxylation [Helwani 2009]. The first
published complete process that couples biological production of FFAs with an extraction and
decarboxylation to alkanes (following the hydrodecarboxylation route) will be discussed in
Chapter 2. Because of the numerous applications for medium-chain length FFAs, their ease of
production from fatty acid biosynthesis using acyl-ACP thioesterases (reviewed in the next
section), and the potential to easily extract and convert them to alkanes with a desirable range of
chain lengths, we chose to focus on developing a strain of E. coli that would produce mediumchain length FFAs.
1.4.1 Survey of acyl-ACP thioesterases
A fairly large number of acyl-ACP thioesterases from plant chloroplasts have been
characterized, as these enzymes are the major determinant of seed oil composition. In plants,
fatty acids biosynthesis takes place in the chloroplast, while lipid biosynthesis takes place in the
cytosol. This compartmentalization requires nascent acyl-chains to be hydrolyzed from ACP,
and the resulting FFAs are either transported across the chloroplast membrane or vectorially
19
acylated to CoA thioesters that are released to the cytosol. Acyl-ACP thioesterases catalyze
the critical step in this process by targeting acyl-ACPs of a certain chain length. Typically,
heterologous expression in E. coli is used to determine thioesterase chain length specificity.
Plant acyl-ACP thioesterases fall into two classes based on homology to enzymes from
Arabidopsis thaliana: FatA isoforms, which have a higher substrate specifity toward unsaturated
acyl chains, particularly oleoyl-ACP (18:1∆9); and FatB isoforms, which are more specific
toward saturated acyl chains [Jones 1995]. As would be expected, most thioesterases exhibit the
highest specificities in the C16-C18 range, including A. thaliana FatA (18:1∆9) [Salas 2002],
Madhuca longifolia FatB (16:0, 16:1, 18:0, 18:1) [Ghosh 2007], Coriandrum sativum FatA
(18:1∆9) [Salas 2002], A. thaliana FatB (16:0, 18:1, 18:0, 16:1) [Salas 2002], Helianthus annuus
FatA (18:1, 16:1) [Serrano-Vega 2005], and Brassica juncea FatB2 (16:0, 18:0) [Jha 2010],
among numerous others. Other plant seed oils exhibit enrichment of medium-chain length acyl
groups, and acyl-ACP thioesterases have been isolated with complementary specificities,
including: Cuphea palustris FatB1 and C. hookeriana FatB2 (8:0, 10:0), C. palustris FatB2 (14:0,
16:0) [Dehesh 1996a, Dehesh 1996b]; and Umbellularia californica FatB (12:0, 12:1, 14:0, 14:1)
[Voelker 1992, Voelker 1994]. Interestingly, A. thaliana FatB had a 2.5-fold higher in vitro
activity toward palmitoyl-ACP (16:0) with ACP derived from spinach, and a 3-fold lower
activity toward myristoyl-ACP (14:0) with ACP derived from spinach, than toward those acyl
groups on E. coli ACP [Salas 2002], suggesting a large contribution of the thioesterase-ACP
interaction toward both specificity and overall activity. Recent phylogenetic analysis across
sequence collections suggested two new subfamilies of plant and green algal acyl-ACP
thioesterases that have never been experimentally characterized [Jing 2011].
20
E. coli possesses two native thioesterases, thioesterase I (TesA) and II (TesB), which
predominantly cleave acyl-CoAs but also acyl-ACPs with much lower activity in vitro, and with
only TesA hydrolyzing >C12 thioesters [Spencer 1978]. TesA is a periplasm-directed enzyme
that likely does not have access to acyl-CoA or acyl-ACP species, and its physiological role
remains unknown [Cho 1993]. Deletion of a 5' leader peptide region traps TesA in the cytosol
(TesA'), where it hydrolyzes predominantly C14 and C16, and a smaller amount of C12 saturated
and unsaturated acyl-ACPs in β-oxidation deficient strains [Cho 1995]. TesA' has been used in a
number of metabolic engineering studies that will be reviewed in section 1.4.2.
Diverse bacteria also possess hundreds of annotated acyl-ACP thioesterases with essentially
no functional characterization prior to a large selection that were recently studied for their in vivo
activities in E. coli [Jing 2011]. While many appeared to express poorly, in general the bacterial
thioesterases hydrolyzed a wider range of acyl groups. Most commonly, 8:0, 12:0, 12:1, 14:0,
and 14:1 FFAs were detected, with some thioesterases producing significantly amounts of
butyric and hexanoic acids as well. Much of Chapter 5 will focus on functional investigations of
a codon-optimized acyl-ACP thioesterase from Geobacillus (now Paenibacillus) sp. Y412MC10.
In the work to be described in this dissertation, the acyl-ACP thioesterase from Umbellularia
californica, which primarily hydrolyzes lauroyl-ACP, was selected for two primary reasons.
First, FFA titers that were achieved in early heterologous expression studies of this enzyme in E.
coli appeared significantly higher than other acyl-ACP thioesterases, with titers of C12 to C14
species of approximately 180 mg/L [Voelker 1994].
Secondly, the product of catalytic
decarboxylation of saturated and unsaturated C12 FFAs would be undecane, and the products of
in vivo esterification (see section 1.5) would be lauric acid methyl or ethyl esters, both of which
should exhibit desirable properties as diesel fuel replacements or as components in diesel blends,
21
compared to longer chain length products. For example, biodiesel derived from coconut oil
(46-50% C12, 17-19% C14 chain length) has a significantly lower viscosity and cloud point and
higher cetane number than soy biodiesel (predominantly C16-C18 chain length, cetane number of
55) [Alleman 2006]. While fuel property data are difficult to obtain for undecane, the cetane
number has been estimated as between 79 to 83 and the viscosity at 25°C is significantly lower
than soy biodiesel [Wu 1999, Murphy 2004]. Experimental engine testing would be required to
fully characterize performance, as a number of other important properties of the fuel include
surface tension, enthalpy of vaporization, and vapor diffusivity under conditions relevant for
engine operation [Murphy 2004].
1.4.2 Metabolic engineering of FFA production
Prior to publication of the work to be discussed in Chapter 2, only one concerted metabolic
engineering effort had been conducted aiming to maximize production of FFAs in E. coli (or any
organism) [Lu 2008]. By contrast, all other reports of FFA overproduction were from efforts to
either characterize the substrate specificity of plant acyl-ACP thioesterases (Section 1.4.2) or to
further understand the regulation of fatty acid biosynthesis (Section 1.3), and involved minimal
optimization to achieve higher titers. Lu et al. used the combined body of this past work to
incorporate four modifications in E. coli:
elimination of β-oxidation by deletion of fadD;
overexpression of both TesA' and a plant acyl-ACP thioesterase from Cinnamomum camphorum;
and overexpression of the four subunits of E. coli ACC.
The combination of these four
modifications and optimization of induction time and growth temperature afforded the highest
titers of fatty acids in batch cultures in a rich undefined medium (LB medium), with a 19-fold
increase in molar quantities of total fatty acids from the original strain. The fatty acid profile
was predominantly saturated and unsaturated C14-C16
22
species on account of the substrate
specificity of TesA' and the thioesterase from C. camphorum [Yuan 1995].
A fed-batch
fermentation was conducted with the FFA-overproducing strain in a defined glycerolsupplemented medium, achieving a titer of 2.5 g/L and a yield of 0.048 g FA g-1 glycerol. It
should be noted that this was achieved by maintaining nearly continuous growth in a high
density culture with a continuous feed supplementing not only glycerol, but also ammonium
sulfate and magnesium sulfate. As a basis for comparison, maximum theoretical yields of
different chain length FFAs from different carbon sources are shown in Table 1.1. These were
calculated by constraint-based modeling as described in Materials and Methods in Chapter 2.
Table 1.1 Calculated maximum theoretical yields of selected FFAs on various carbon sources.
Values were calculated by constraint-based modeling using the iAF1260 metabolic network
reconstruction of E. coli as described in Chapter 2.
fatty acid
palmitic acid
myristic acid
lauric acid
glycerol
0.37
0.38
0.39
g FA/g carbon source
D-glucose D-xylose L-arabinose
0.34
0.28
0.28
0.34
0.29
0.28
0.35
0.29
0.29
Many additional studies have since been published, with most utilizing very similar sets of
genetic modifications including blocking of the β-oxidation pathway and expression of various
acyl-ACP thioesterases.
Table 1.2 summarizes the yields and percentage of maximum
theoretical yields obtained from a variety of batch and fed-batch studies including our work in
Chapter 2 and the best performing strain from our laboratory grown at 37°C in shake flasks
containing MOPS minimal medium supplemented with glucose (TY05, containing three
chromosomal integrations of BTE under control of IPTG-inducible trc promoters in the fadD,
fadE, and fadAB loci). In general, most overproduced FFAs are in the C14-C16 range (from TesA',
23
FatB-type thioesterase from C. camphorum, FatB-type thioesterase from Ricinus communis,
FatB-type thioesterase from Jatropha curcus). Some observations of note are that Steen et al.
[2010] obtained high titers in minimal medium using strain DH1, and achieved an increase in
titer from 0.7 g/L to 1.2 g/L from deletion of fadE rather than fadD. While no rationale was
posited, this could possibly be due to more efficient blockage of β-oxidation due to the existence
of other C14-C16 acyl-CoA synthetases in E. coli. However, we have not observed any increase
in titer in our baseline strain (K-12 MG1655) as a result of deletion of both fadD and fadE versus
fadD alone (data not shown). High titers of up to over 2 g/L have also been achieved by
expression of R. communis and J. curcas thioesterases in a ∆fadD strain [Zhang 2011a], however
the use of LB medium only allowed the calculation of a maximum of 41% of the theoretical
yield being achieved from glucose, as LB contains other utilizable carbon sources. By contrast,
our best performing strain achieves 41% of maximum theoretical in minimal medium but with
lower titers (~1 g/L). Fed-batch studies using a ∆fadL strain, which would theoretically prevent
re-import of extracellular C14 and higher FFAs [Black 1990], and expressing TesA' in minimal
media supplemented with glucose have achieved titers of up to 5.1 g/L [Liu 2012], however as in
the original study by Lu et al. [2008], these titers appear to come at the expense of yield, with up
to 12.9% of maximum theoretical achieved. A recent study has reported 3.8 g/L titer of fatty
acids from the same baseline strain and medium used by Steen at al. [2010] over the course of 72
h, with the only modification being the use of a different low copy vector (the vector used in the
original study was not specified) [Zhang 2012]. While maximizing FFA production was not the
objective of this study, it represents the highest percentage of maximum theoretical yield yet
achieved via fatty acid biosynthesis, of 56%.
24
Table 1.2 Literature summary for FFA titers, yields, and percent of maximum theoretical yield.
Base strain
Modifications
Thioesterase
Titer
(g/L)
BL21(DE3)
∆fadD, ACC+
TesA' + CcTE
0.38
BL21(DE3)
∆fadD, ACC
+
TesA' + CcTE
DH1
∆fadD
DH1
∆fadE
+
Media/carbon
Yield
% of
Time
(% w/w) theoretical (h)
Culture
type
Temp.
(°°C)
Reference
LB/none
N/A
N/A
> 18
batch
30
Lu 2008
2.5
M9/glycerol
4.8
12.8%
22
fed-batch 30
Lu 2008
TesA'
0.7
M9/2% glucose
3.5
10.3%
N/A
batch
37
Steen 2010
TesA'
1.2*
M9/2% glucose
6.0
17.6%
N/A
batch
37
Steen 2010
BTE
0.81
LB/0.4% glycerol
< 16.1
< 42.4%
29
batch
37
Lennen 2010
K-12 MG1655
∆fadD, ACC
K-12 MG1655
∆fadD
BTE
0.77
LB/0.4% glycerol
< 15.3
< 40.3%
24
batch
37
Lennen 2011
K-12 MG1655
∆fadD ∆fadE
∆fadAB
BTE
0.98
MOPS**/
0.7% glucose
14
41%
40
batch
37
Youngquist
(unpublished)
K-12 MG1655
ACC+, FabD+
S. pyogenes TE
0.16
LB/none
N/A
N/A
24
batch
37
Jeon 2011
K-12 MG1655
∆fadD
RcTE
2.1
LB/1.5% glucose
< 14
< 41%
36
batch
30
Zhang 2011a,
Li 2012
K-12 MG1655
∆fadD
JcTE
1.5
LB/1.5% glucose
< 10
< 29%
36
batch
30
Zhang 2011a
BL21(DE3)
∆fadD, ACC+
TesA' + CcTE
0.94
LB/none
N/A
N/A
> 20
batch
30
Liu 2010a
BL21(DE3)
∆fadE
TesA' + CcTE
0.45
LB?
N/A
N/A
> 18
batch
30
Yu 2011
+
BL21(DE3)
∆fadE, FabZ ,
FabG+, FabI+
TesA' + CcTE
0.65
LB?
N/A
N/A
> 18
batch
30
Yu 2011
K-12 MG1655
∆fadD, SaFabD+
RcTE
1.4
LB/1.5% glucose
< 9.3
< 27%
24
batch
30
Zhang 2011b
TesA'
5.1
M9/glucose
4.1
12.9%
38
fed-batch 37/34/30 Liu 2012
TesA'
4.8
M9/glucose
4.4
12.1%
38
fed-batch 37/34/30 Liu 2012
TesA'
~0.45* minimal/2% glucose 2.3/
7.4§
6.6%/
22%§
48
batch
37
Dellomonaco
2011
TesB
~0.70* minimal/2% glucose 3.5/
13.3§
10%/
39%§
48
batch
37
Dellomonaco
2011
BL21(DE3)
BL21(DE3)
∆fadL
+
K-12 MG1655
see text, FadAB
K-12 MG1655
see text, FadAB+
24
25
Table 1.2 (cont.) Literature summary for FFA titers, yields, and percent of maximum theoretical yield.
Base strain
Modifications
Thioesterase
Titer
(g/L)
Media/carbon
Yield
% of
Time
(% w/w) theoretical (h)
Culture
type
Temp.
(°°C)
Reference
K-12 MG1655
see text, FadAB+
FadM
~0.87* minimal/2% glucose 4.4/
28
13%/
85%§
48
batch
37
Dellomonaco
2011
K-12 MG1655
see text, FadAB+
FadM
~7
minimal/3% glucose 23/
28§
70%/
85%§
60
batch
37
(bioreactor)
Dellomonaco
2011
DH1
∆fadE
TesA'
3.8
minimal/2% glucose 19
56%
72
batch
Zhang 2012
37
* free fatty acids or extracellular fatty acids only
** modified MOPS minimal medium with reduced concentration of phosphate (0.37 mM)
§
authors' calculation per g carbon source consumed (other values are per g carbon source supplied)
Abbreviations: TesA', cytosolic form of E. coli thioesterase I; CcTE, acyl-ACP TE from Cinnamomum camphorum; BTE, acyl-ACP TE from Umbellularia
californica; S. pyogenes TE, oleoyl-ACP TE from Streptococcus pyogenes; RcTE, acyl-ACP TE from Ricinus communis; JcTE, acyl-ACP TE from Jatropha
curcus; SaFabD, FabD from Streptomyces avermitilis; FadM, E. coli acyl-CoA thioesterase; TesB, E. coli thioesterase II (acyl-CoA thioesterase); N/A either not
applicable because it cannot be calculated from information given, or not available because information was not provided. See references for further details.
25
26
A promising recently described alternative route to production of FFAs and other FFAderived products involves reversal of β-oxidation, such that β-oxidation operates anabolically
rather than in the usual catabolic direction [Dellomonaco 2011]. This is made possible by the
reversible nature of FadA and the two reactions in the β-oxidation cycle catalyzed by FadB. A
highly engineered strain was designed that facilitated high-level expression of enzymes of βoxidation (constitutive expression of fadR and atoC, deletion of arcA, replacement of crp with
crp*, a cAMP-independent mutation), deletion of alternative fermentative pathways (∆adhE
∆pta ∆frdA ∆fucO ∆yqhD), and deletion of fadD. Experimental results strongly suggested YdiO
catalyzed the reverse reaction to FadE, which catalyzes irreversible acyl-CoA dehydrogenation.
Curiously, YdiO was not overexpressed and fadE was not deleted in this strain. When FadA,
FadB, and various acyl-CoA thioesterases were expressed in this strain, high titers of FFAs
resulted (Table 1.2).
The best-performing acyl-CoA thioesterase, FadM, predominantly
hydrolyzed saturated C16-C18 species. In shake flask cultures, up to 13% of the theoretical yield
was attained on a basis of the glucose supplied to a FadM-expressing strain, however based on
the uptaken glucose, yields approached 85% of theoretical. In a bioreactor run, nearly all
glucose was utilized in 60 h and ~7 g/L of fatty acids were produced, representing 85% of the
maximum theoretical yield.
This high yield is likely approaching or exceeding the levels
required to begin process commercialization, and could be a result of pathway selection (see
Chapter 5 for further discussion) or optimal bioprocess conditions. It should be noted that the
bioreactor medium contained 1 mM betaine, an osmolyte that has previously been shown to
greatly improve microbial production of lactate [Zhou 2006].
27
1.5 Other useful products derived from fatty acid biosynthesis
While FFAs have utility as commodity chemicals, C10 and higher chain length FFAs are
solids at ambient temperature and are therefore are not biofuel candidates. Instead, FFAs must
be either catalytically deoxygenated or processed (from either FFAs or acyl-ACP precursors) by
recently elucidated metabolic pathways that generate fuel molecules directly in vivo, for which
progress to date will be discussed here. Most of the spectrum of products that can be derived in
vivo from fatty acid biosynthesis are shown in Figure 1.8. Fatty acid ethyl esters (FAEEs),
which are chemically similar to current biodiesel (fatty acid methyl esters), can be derived from
acyl-CoAs and endogenously produced ethanol by utilizing a promiscuous wax ester
synthase/acyl-CoA:diacylglycerol acyltransferase (WS/DGAT; AtfA) from Acinetobacter baylyi
[Kalscheuer 2003].
Significant potential exists for generating FAEEs with improved fuel
properties such as with shorter aliphatic chains and branched chains by enzyme engineering of
AtfA. Additional homologues of AtfA which exhibit different substrate specificities also exist in
other γ-proteobacteria and actinomycetes, including an isoprenoid wax ester synthase in
Marinobacter hydrocarbonoclasticus [Wältermann 2007, Holtzapple 2007, Barney 2012]. AtfA
has been utilized in several recent metabolic engineering efforts, together with heterologous
expression of pyruvate decarboxylase (Pdc) and alcohol dehydrogenase (AdhB) from
Zymomonas mobilis to enable efficient endogenous ethanol production [Ingram 1987].
A
preliminary report expressing atfA, pdc, and adhB on a plasmid in E. coli, with oleate supplied to
the fermentation broth, enabled production up to 1.28 g/L of the FAEE [Kalscheuer 2006].
Expression of atfA and endogenous overproduction of FFAs in E. coli (∆fadE, overexpression of
TesA') with exogenous addition of ethanol produced a titer of 0.4 g/L of FAEEs, while additional
optimization of complete de novo biosynthesis involving heterologous expression of pdc and
28
Figure 1.8 Fatty-acid derived fuels and chemicals. Products that can be derived from acyl-ACPs
include FFAs hydrolyzed from acyl-ACPs by a thioesterase; long-chain alkanes via an acyl-ACP
reductase and aldehyde decarbonylase from some cyanobacteria [Schirmer 2010]; long-chain terminal
alkenes by either a cytochrome P450 decarboxylase from a Jeotgalicoccus species, putatively from FFAs
[Rude 2011], or from acyl groups of unknown origin via an elongation decarboxylation mechanism
catalyzed by a multidomain protein homologous to type I polyketide synthases in Synechococcus sp. PCC
7002 [Mendez-Perez 2011]. FFAs can also be methylated directly with S-adenosylmethionine (SAM) to
form fatty acid methyl esters (FAMEs) via a fatty acid O-methyltransferase (FAMT) from
Mycobacterium marinum [Nawabi 2011]. Acyl-CoAs derived from FFAs by acyl-CoA synthetase
(FadD) or other acyl-CoA ligases can be ethyl esterified (FAEEs) by a wax ester synthase (AtfA) from
Acinetobacter baylyi [Kalscheuer 2003, Steen 2010]; reduced to fatty alcohols by acyl-CoA reductases
[Steen 2010]; condensed to form long-chain waxy internal alkenes and ketones by OleA from
Micrococcus luteus [Beller 2010]; and β-ketoacyl-CoAs can be hydrolyzed to 3-oxoacids by FadM and
putatively undergo spontaneous decarboxylation to methyl ketones [Goh 2012]. Adapted from [Liu
2010b].
adhB, overexpression of fadD, an increased copy number of atfA, and introduction of a dodecane
overlayer increased FAEEs to 0.674 g/L, representing approximately 7% of the maximum
29
theoretical yield [Steen 2010].
Fed batch optimization of another similar strain enabled
production up to 0.922 g/L but at severe expense of yield (reported as 0.025 g FAEE g-1 glucose)
[Duan 2011]. A dramatic improvement in titer (1.5 g/L) and yield (reported as 28% of maximum
theoretical) was recently reported using a dynamic sensor-regulator system based on the use of
two FadR-regulated promoters, to better balance fluxes of endogenously produced FFAs, acylCoAs, ethanol, and FAEEs, demonstrating the power of synthetic biology and a facile host in
metabolic engineering applications [Zhang 2012].
Metabolic engineering of E. coli for production of fatty acid methyl esters has also recently
been described by heterologous expression of a fatty acid O-methyltransferase (FAMT) from
Mycobacterium marinum [Nawabi 2011]. The enzyme was characterized as having a substrate
preference for C8-C12 saturated and 3-hydroxylated FFAs. Thus in situ production from glucose
was tested using a plant acyl-ACP thioesterase from Arabidopsis thaliana which provides a
predominantly saturated FFA profile, and an acyl-ACP thioesterase from Clostridium
acetobutylicum, which provides 3-hydroxylated FFAs.
As S-adenosylmethionine (SAM) is
required as the methyl donor by the FAMT, cells were further engineered to increase the supply
of SAM by deletion of metJ, encoding a global regulator of the methionine biosynthesis pathway
that feedback inhibits SAM biosynthesis in response to elevated concentrations of SAM. While
reported yields and titers were very low, the bottleneck was believed to be the lack of a
thioesterase that cleaves sufficient 3-hydroxydecanoic acid, the preferred substrate of the FAMT
from Mycobacterium marinum.
Long-chain length alkanes and olefins have been isolated from numerous cyanobacterial
species in very low quantities, generally falling into two classes for which most species can only
produce one: pentadecane and heptadecane; and nonadecene and nonadecadiene [Winters 1969,
30
Schirmer 2010]. A subtractive genome analysis between Synechocystis sp. PCC 6803 and
other strains that produce the former group, and Synechococcus sp. PCC 7002 which does not,
identified a small list of gene candidates from which it was determined that an acyl-ACP
reductase and an aldehyde decarbonylase were responsible for pentadecane and heptadecane
production from C16 and C18 acyl-ACPs. When heterologously expressed in E. coli, pentadecane,
pentadecene, and heptadecene were the major products, as 16:0, 16:1, and 18:1 are the
predominant intermediates (lacking in cyanobacteria due to introduction of acyl chain
unsaturations at the phospholipid level via acyl phospholipid desaturases [Murata 1995]). Up to
300 mg/L of alkanes could be produced in E. coli, however it is likely that higher titers are
currently being achieved under industrial development by LS9.
The gene responsible for nonadecene and nonadecadiene production in Synechococcus sp.
PCC 7002, ols, was also identified by a genetic study within our laboratory [Mendez-Perez 2011].
Ols is homologous to type I polyketide synthases and has a domain architecture suggesting
loading of an as-yet unidentified acyl substrate, one round of elongation with malonyl-CoA, and
a sulfotransferase domain suggesting sulfation, perhaps by the activated sulfate intermediate 3'phosphoadenosine 5'-phosphosulfate (PAPS). Sulfation may provide the activation necessary to
allow subsequent dehydration and decarboxylation to 1-nonadecene, while the product 1,14nonadecadiene is likely a product of an unidentified desaturase utilizing 1-nonadecene as a
substrate.
Feeding heptadecanoic acid to PCC 7002 cultures enabled the synthesis of 1-
octadecene. While it was possible to increase the titer produced in PCC 7002 by 2-fold and 5fold for 1-nonadecene and 1,14-nonadecadiene, respectively, by replacement of the native ols
promoter with a stronger promoter from Amaranthus hybridus, titers remain in the sub-µg/mL
31
range. While heterologous expression of the protein appears to have been achieved in E. coli,
no alkene products have been detected.
A second route for terminal alkene production has been identified in a Gram-positive
Jeotgalicoccus species, catalyzed by a cytochrome P450 enzyme named OleT [Rude 2011].
When heterologously expressed in E. coli, 1-pentadecene and 1,10-heptadecadiene are produced
at unreported titers, while 1-heptadecene appears when stearic acid (18:0) is fed. The proposed
mechanism of OleT involves oxygen radical attack at the β-carbon of fatty acids, which
facilitates decarboxylation. Of the three routes discussed for bacterial production of alkanes and
alkenes of potential value as biofuels, the most promising pathway for metabolic engineering
involves heterologous expression of the acyl-ACP reductase and decarbonylase, which are small,
soluble proteins (341 and 231 amino acids, respectively). While interesting biologically, large
polyketide synthases and cytochrome P450s are unwieldy and difficult to heterologously express
in E. coli [Khosla 1997, Chang 2007]. The host organisms from which Ols and OleT have been
isolated are also currently difficult to genetically engineer.
A third discovered microbial route to long-chain length alkenes comes from a gene cluster in
Micrococcus luteus, which when heterologously expressed in fatty acid overproducing E. coli
(DH1 ∆fadE with overexpression of TesA'), results in the production of small quantities of longchain length (C27 to C29) internal alkenes and monoketones [Beller 2010]. One of the proteins
expressed in the gene cluster, OleA, is hypothesized to catalyze a Claisen condensation between
β-ketoacyl-CoAs and fatty acyl-CoAs, which generates a diketone which is reduced and
dehydrated to a monoalkene. Waxy hydrocarbons of these chain lengths would not be useful as
fuels, but could serve other commodity and specialty chemical markets for waxes and lubricants.
32
Medium- to long-chain fatty alcohols, while not having suitable properties to serve as fuels,
are valuable commodity chemicals used in the production of nonionic and anionic surfactants,
and serve directly as additives in foods and cosmetics, plasticizers, and lubricants [Johnson 1989,
Gunstone 1997]. While industrially derived from plant oils, they are also natural products of a
variety of bacteria, plants, and animals, and are biosynthesized by acyl-CoA and acyl-ACP
reductases. Fatty aldehydes are detected as products of some enzymes [Steen 2010, Schirmer
2010, Teerawanichpan 2010], suggesting either spontaneous oxidation to fatty alcohols or the
existence of another enzymatic activity responsible for aldehyde oxidation.
An NADPH-
dependent fatty aldehyde reductase has been isolated from Marinobacter aquaeolei VT8
[Hofvander 2011, Willis 2011]. While a number of plant and algal [Doan 2009], protozoan
[Teerawanichpan 2010], and bacterial [Reiser 1997, Steen 2010, Schirmer 2010] acyl-CoA and
acyl-ACP reductases have been expressed in either E. coli or S. cerevisiae for the purpose of
functional investigations, only a handful of studies have reported titers higher than 10 mg/L.
These include the acyl-ACP reductase from Synechococcus sp. PCC 7942 described previously,
where expression in E. coli without the decarbonylase results in accumulation of over 100 mg/L
of a mixture of hexadecanol, hexadecanal, octadecenol, octadecenal, and tetradecanol [Schirmer
2010]. Utilizing acyl-CoA reductases in E. coli requires more pathway engineering as acylCoAs can only be overproduced by generating FFAs, and then activating them to acyl-CoAs
using an acyl-CoA ligase such as FadD. It is likely that all of these conversion steps need to be
carefully balanced to avoid the accumulation of intermediates, similar to the case for FAEEs.
Overexpression of TesA', overexpression of FadD, and heterologous expression of acr1, an acylCoA reductase from Acinetobacter calcoaceticus, in a E. coli ∆fadE strain resulted in production
of approximately 100 mg/L of C16 and C18 fatty alcohols [Steen 2010]. Swapping the acyl-ACP
33
thioesterase from TesA' to those from Umbellularia californica or Cuphea hookeriana shifted
the predominant chain length to C12 and C14, respectively, but with unspecified titers.
Methyl ketones are another class of FFA-derived products which could have useful value as
biofuels. Recent work demonstrated engineering of E. coli to overproduce C11 to C15 methyl
ketones, following up on the observation of methyl ketone production when Micrococcus luteus
FabH and FabF were heterologously expressed in E. coli [Goh 2012]. It was hypothesized that
overproduction of β-ketoacyl-CoAs would lead to production of methyl ketones, as FabH and
FabF overexpression would likely increase the accumulation of similar β-ketoacyl-ACPs.
Ultimately a strain with highly engineered β-oxidation including a deletion in fadE, replacement
of the FadE activity with a soluble acyl-CoA oxidase from Micrococcus luteus, overexpression
of E. coli FadB, and deletion of fadA, was found to greatly increase production of methyl ketones.
Additional overexpression of the poorly characterized acyl-CoA thioesterase FadM was found to
significantly increase the titer to approximately 50 mg/L, suggesting its role as a β-ketoacyl-CoA
thioesterase. Released β-ketoacids were hypothesized to spontaneously decarboxylate to yield
methyl ketones. Use of a decane overlay further improved titers to 380 mg/L.
A final category of FFA-derived products are bioplastics generated from polymerizing
hydroxylated fatty acids to form medium-chain length polyhydroxyalkanoates. This class of
bioplastics has been biosynthesized by feeding vegetable oils to E. coli cultures, which enter βoxidation to generate polymerizable 3-hydroxyacyl-CoA monomers [Gao 2011]. However, the
iterative nature of β-oxidation generates a spectrum of monomer chain lengths, yielding
undefined heteropolymers. An alternative strategy by our laboratory and others [Rehm 2001]
disrupts normal β-oxidation while taking advantage of an acyl-ACP thioesterase to generate
more uniform acyl chain lengths (mostly C12).
34
1.6 Conclusion
Microbially-derived FFAs and other fatty-acid derived chemicals are a promising source of
renewable, high energy density biofuels and commercial oleochemicals.
Decades of basic
biochemical and genetic research, and several very recent advances, have laid the foundation for
metabolic engineering of the versatile host microorganism E. coli to economically produce these
compounds from renewable carbon sources. The remainder of this thesis will discuss our own
efforts related to producing medium-chain length FFAs in E. coli. In Chapter 2, metabolic
engineering and optimization of a medium-chain length FFA overproducing strain of E. coli will
be presented, and a coupled catalytic process for converting these FFAs to a useful liquid fuel
product will be introduced. In Chapter 3, work aimed toward identifying the bottleneck to
production through physiological characterization and a functional genomics study of an FFAoverproducing strain will be presented.
Additional work stemming from the hypothesized
production bottlenecks and efforts to break these yield barriers will be presented in Chapter 4
and 5. Chapter 6 will discuss future directions for the body of work described in this thesis.
35
Chapter 2: Development of a medium-chain length fatty acid overproducing
strain of E. coli and a process for alkane production
This chapter will discuss our efforts to develop a platform strain of E. coli that overproduces
FFAs primarily of saturated C12 chain length, and some preliminary optimization of the strain.
The maximum theoretical yield of FFAs from various carbon sources was determined using
constraint-based modeling, in order to set a benchmark for all future metabolic engineering
efforts. Three genetic modifications to a laboratory strain of E. coli (K-12 MG1655) were
introduced. First, fatty acid β-oxidation was eliminated by deleting the gene encoding the first
enzyme in the pathway, FadD, which activates FFAs to acyl-CoAs. Second, a codon-optimized
FatB-type acyl-ACP thioesterase from Umbellularia californica (BTE) was codon-optimized and
heterologously expressed, and its expression level was optimized by correlating fatty acid titer
and growth properties with the copy number of the plasmid in which it was cloned. Third,
acetyl-CoA carboxylase (ACC) from E. coli was additionally overexpressed on a compatible
plasmid. The maximum titers of FFAs were achieved when BTE was expressed on low-copy
plasmids, and lower titers and reduced growth resulted from expression on higher copy plasmids.
Additional overexpression of ACC resulted in a negligible increase in FFA production,
indicating that ACC is not the rate-limiting step. While a precise percentage of the maximum
theoretical yield could not be calculated due to use of a rich undefined medium in this study, it
could be determined that no more than 42% of the maximum theoretical yield was being
achieved, leaving significant future work for further strain engineering. Finally, a complete
process will be introduced wherein FFAs can be extracted out of the culture medium and
* Portions of this chapter were published in Biotechnology and Bioengineering (Lennen 2010)
36
catalytically decarboxylated to medium-chain length alkanes, a potential drop-in replacement
for diesel.
2.1. Introduction
As discussed in depth in Chapter 1, microbially synthesized fatty acids are logical precursors
to diesel-like hydrocarbons, and offer the flexibility to exploit a variety of biomass-derived
carbon sources for their production. While fatty acid biosynthesis in E. coli (and likely in other
microorganisms) is heavily regulated, dramatic examples of decoupling fatty acid biosynthesis
from normal modes of regulation have been observed from heterologous expression of plant
acyl-ACP thioesterases. While several have been expressed in E. coli [Yuan 1995, Serrano-Vega
2005, Jha 2006, Ghosh 2007, and others reviewed in Chapter 1], the FatB-type medium-chain
acyl-ACP thioesterase from Californica umbellularia demonstrated the largest increase in total
fatty acid production, with a chain length distribution heavily skewed toward C12 and C14 species
[Voelker 1994]. One metabolic engineering strategy for overproducing FFAs in E. coli utilizing
the acyl-ACP thioesterase from Cinnamomum camphorum and a cytosolic form of thioesterase I
(TesA) from E. coli [Cho 1993] was published prior to this work [Lu 2008], and was analyzed in
more detail in Chapter 1.
In this study, we sought to further understand previously reported instabilities of plasmids
expressing the U. californica acyl-ACP thioesterase [Voelker 1994] by varying the copy number
of a codon-optimized version of the gene (BTE) on plasmids having different origins of
replication, in an effort to balance the level of active protein with possible metabolic burden and
product toxicity [Jones 2000]. Previous studies expressing acyl-ACP thioesterases in E. coli
have been conducted at 30°C due to this putative instability [Lu 2008, Feng 2009]. However, no
37
additional studies have been conducted to determine any rationale for this instability. With
this information, E. coli was engineered to overproduce FFAs via three modifications: (1)
expression of a codon-optimized acyl-ACP thioesterase from U. californica (BTE) on a suitable
plasmid; (2) deletion of the fadD gene encoding acyl-CoA synthetase, the first enzyme involved
in β-oxidation; and (3) overexpression of ACC. The latter modification was attempted due to
previous reports of ACC catalyzing the rate-limiting step in fatty acid biosynthesis when another
acyl-ACP thioesterase, a cytosol-directed version of E. coli thioesterase I (TesA'), was
overexpressed [Davis 2000].
Fatty acids can be converted to useful liquid fuels by chemical catalytic or enzymatic
reactions described in Chapter 1. Here, the overproduced FFAs were extracted from culture and
catalytically decarboxylated to alkanes.
Coupling the microbial fermentation to a scalable
existing chemical technology precludes the need to heterologously express a new in vivo
pathway to convert the fatty acids to a useful liquid product, which even if properly balanced
could reduce the overall yield of fatty acids due to increased cellular metabolic burdens.
Decarboxylation is a preferable conversion because alkanes have more desirable properties, such
as higher energy density and lower viscosity, than corresponding esters that would be produced
by catalytic esterification of fatty acids. Furthermore, the final medium-chain alkane product can
be recycled for reuse as an extractant of fatty acids from the culture medium and as the solvent
for the catalytic decarboxylation. E. coli has been selected because of its ease of genetic
manipulation, well-understood physiology (especially with regards to fatty acid biosynthesis),
and rapid growth rate. Because all prokaryotic and eukaryotic organisms possess the ability to
produce fatty acids as part of membrane lipid biosynthesis, small modifications to the methods
presented should be broadly applicable to other industrial microorganisms.
38
2.2 Materials and Methods
2.2.1 Maximum theoretical yield determination
Maximum theoretical yields of FFAs were determined for a range of carbon sources using the
iAF1260 genome-scale metabolic reconstruction for E. coli [Feist 2007], and solving the largescale linear programming problems using the COBRA Toolbox [Becker 2007] with the GNU
Linear Programming Kit (Free Software Foundation, Boston, MA) in MATLAB (MathWorks,
Natick, MA). The iAF1260 model was modified to allow reversible transport of FFAs between
the cytosol and periplasm (the original model only allowed for cytosolic import from the
periplasm) and by forcing the flux to zero through the reactions catalyzed by PlsB. This directed
production of FFAs through a fatty acid hydrolase route rather than by formation and cleavage of
diacylglycerols, which provided a small advantage from the standpoint of ATP utilization but is
not a physiologically relevant route for production of medium-chain length FFAs. Maximum
uptake fluxes were set to 10 mmol gDCW-1 (gram dry cell weight) h-1 for glycerol (within the
range that can be achieved by culture conditioning) [Fong 2005], 7 mmol gDCW-1 h-1 for Dglucose [Pramanik 1997], 2 mmol gDCW h-1 for xylose and L-arabinose, and 18.5 mmol gDCW1
h-1 for oxygen [Feist 2007, Edwards 2001]. The objective function was set to maximize FFA
production.
2.2.2 Strain construction
Bacterial strains, plasmids, and oligonucleotide primers used in this study are listed in Tables
I and II. Oligonucleotide primers were purchased from Integrated DNA Technologies, Inc.
(Coralville, IA). Chemicals and reagents were purchased from Fisher Scientific (Pittsburgh, PA)
unless otherwise specified. E. coli K-12 MG1655 was obtained from the E. coli Genetic Stock
39
Center (New Haven, CT) and was the background strain used in this work. Strain RL01 was
constructed by deleting the fadD gene from the MG1655 chromosome by λ Red-mediated
recombination [Datsenko 2000] using the λ Red recombinase encoded on plasmid pKD46.
Plasmid pKD13 was used as the template for amplification of the linear cassette using primers 1
and 2. The kan cassette was removed by expressing the FLP recombinase encoded on plasmid
pCP20. Strain NRD204 (K-12 MG1655 araBAD::cat) [De Lay 2007] was generously donated
by Dr. Cronan (University of Illinois at Urbana-Champaign). Strain RL08 was constructed by
P1 phage transduction [Thomason 2007b] of the araBAD::cat insertion into strain RL01 via a
modified protocol utilizing selective plates containing 1.25 mM sodium pyrophosphate as a
calcium chelator. The cat cassette was removed using pCP20. To generate strains RL08-BTE
and
RL08-BTE-H204A,
plasmids
pBAD34-BTE-kan
and
pBAD34-BTE-H204A-kan
(construction described in next section) were used as templates for amplification of linear
cassettes using primers 1 and 7. The linear cassettes include araC but do not include the rrnB
terminator 3' to BTE(-H204A). The cassettes were transformed into induced electrocompetent
strain DY330, which possesses a temperature-inducible λ Red recombinase on a prophage
integrated at the attB site, following previously published procedures [Thomason 2007a]. The
integrations in the fadD locus were P1 phage transduced into strain RL06 using a modified liquid
lysate generation procedure [Donath 2011] conducted at 30°C, followed by selection as
described above. The kan cassettes were removed using pCP20, generating the final strains. To
minimize the possible presence of multiple insertions by λ Red recombination, all recombinant
strains were used as donors to transduce back into the parent strain.
Gene insertions and
deletions were verified by colony PCR using primers 3 and 4 for fadD and 5 and 6 for araBAD,
and by the absence of growth after plating on M9 agar supplemented with either 0.1% (w/v)
40
sodium oleate (TCI America) as previously described [Overath 1969], or 0.4% (w/v) Larabinose (Calbiochem, San Diego, CA) as carbon sources (Figure 2.1).
MG1655
MG1655
K27
MG1655 ? fadD
A
g lycero l +
sodium o leate
K27
MG1655 ? fadD
MG1655 ? fadD ?araBAD
sodium oleate
B
glycerol +
L-arabinose
L -arabino se
Figure 2.1 Functional verification of fadD and araBAD gene deletions on agar plates containing
oleate and arabinose as sole carbon sources. (A) E. coli K-12 MG1655, K27 (a non-functional fadD
mutant (Overath 1969)), and MG1655 ∆fadD streaked on M9 minimal medium agar containing 0.4%
glycerol and 0.1% sodium oleate (left), or 0.1% sodium oleate (right). K27 and the ∆fadD strain cannot
grow on the oleate-only plate, indicating shut-down of β-oxidation. (B) K-12 MG1655 and MG1655
∆fadD ∆araBAD (RL08) streaked on M9 minimal medium agar containing 0.4% glycerol and 0.4% Larabinose (left), and 0.4% L-arabinose (right). The ∆araBAD strain cannot grow on the arabinose-only
plate.
2.2.3 Gene synthesis
The 897-bp portion of the DNA sequence of BTE lacking the 83 amino acids at the Nterminus that appear to be involved in thylakoid targeting (Voelker 1992) and containing the
XbaI site formerly used to clone a functional part of the gene in E. coli (Ohlrogge 1995) was
codon-optimized for expression in E. coli, with common restriction sites eliminated.
An
artificial ribosome binding site (AGGAGG), spacer sequence, start codon, and bases to create an
in-frame sequence were added upstream of the gene fragment. The full sequence (Figure 2.2)
was custom-synthesized (Integrated DNA Technologies, Inc.) and was received in plasmid
pUC57 (pUC57-BTE).
41
1
51
101
151
201
251
301
351
401
451
501
551
601
651
701
751
801
851
901
CCCGGGAGGA
CTGCCTCAAC
TCGTACTTTC
CCATCCTGGC
AAATCTGTTG
TAAACGTGAC
GCTACCCTAC
TCCGGTAACA
GGGCGAAATC
GCACTCGTCG
CCTGCTTTCA
GCAAAAACTG
CGCGCTGGAA
GTTGCTTGGG
CATTTCCTCT
TTCTGCGCAG
GTCTGCGACC
GCGTACGGAG
TAATTCCGGC
GGATTATAAA
TGCTGGATGA
GCAATTCGTT
CGTCATGAAC
GTATCCTGGG
CTGATGTGGG
TTGGGGTGAC
ATGGTATGCG
CTGACGCGTT
CCTGTCTACC
TCGATAACGT
AACGACTCCA
CGACCTGGAT
TCTTCGAGAC
TTTACTCTGG
CCTGACCACC
ATCTGCTGCA
TGGCGTCCAA
GGAACCTCGT
ATGACTCTAG
TCACTTCGGT
CTTATGAAGT
CACATGCAGG
TGATGGTTTC
TAGTGCGTCG
ACTGTGGAAG
TCGCGATTTT
GCACCTCCCT
ATCCCGGACG
GGCAGTTAAA
CCGCGGACTA
GTTAATCAGC
TGTGCCGGAC
AGTACCGTCG
GTAAGCGGCG
ACTGGAAGGC
AGCTGACGGA
GTTTAAGCTT
AGTGGAAACC
CTGCACGGTC
GGGTCCAGAT
AAGCCACCCT
GGCACTACTC
CACCCACGTA
TCGAGTGTTG
CTGGTCCGTG
GAGCGTTCTG
AAGTGCGCGG
GACGACGAAA
CATCCAGGGC
ATGTGAACAA
AGCATTTTCG
CGAATGTACT
GTTCTAGCGA
GGCTCCGAAG
TTCTTTCCGC
GAAACCAAAA
TGGTGTTTCG
CGTTCTACCT
GAATCACGCG
TGGAAATGTC
GCAGTAGAGC
GATTGGCGCG
ACTGTAAAAC
ATGAACACCC
TGAGATCGGT
TCAAGAAACT
GGTCTGACTC
CCTGAAATAC
AAAGCCATCA
CGCGACTCCG
GGCAGGTCTG
TCCTGCGTGC
GGCATCTCCG
Figure 2.2 Sequence of codon-optimized and custom-synthesized BTE gene. The full sequence of
synthesized DNA includes restriction sites at the 5' (XmaI) and 3' (HindIII) ends (italicized), an artificial
ribosome binding site (shown in blue), a spacer sequence (shown in red), an artificial start codon (bolded),
and two bases following the start codon and before the XbaI site to generate an in-frame sequence (shown
in green) is shown below. The codon corresponding to histidine-204 is underlined. The stop codon is
bolded.
2.2.4 Plasmid construction
All cloning was performed in E. coli DH10B cells (Invitrogen, Carlsbad, CA), and all
enzymes for cloning were purchased from New England Biolabs (Ipswich, MA). The accD gene
encoding the β subunit of acetyl-CoA carboxyltransferase, the accA gene encoding the α subunit
of acetyl-CoA carboxyltransferase, and the accBC operon encoding biotin carboxyl carrier
protein and biotin carboxylase were amplified by PCR from MG1655 genomic DNA with their
putative upstream ribosome binding sites (RBS) using primers 8 and 9, 10 and 11, and 12 and 13,
respectively. These PCR products were sequentially inserted into pBAD33 (Guzman et al.,
1995) between the SacI and XmaI sites (pRL1), SalI and XbaI sites (pRL2), and SphI and XhoI
sites to create the artificial operon accDABC in plasmid pBAD33-ACC.
The BTE fragment from plasmid pUC57-BTE between the XmaI and HindIII sites was
inserted into plasmid pBAD18 (Guzman 1995) to generate plasmid pBAD18-BTE. To create a
42
non-functional BTE for use as a negative control, a catalytic histidine at amino acid 204 was
identified based on prior alignments of BTE with other plant acyl-ACP thioesterases (Yuan
1996). A two-step megaprimer PCR procedure (Xu 2003) was used to mutagenize the first two
nucleotides of the histidine codon at position 204 to create an alanine codon. Primers 14 and 15
were used in the first reaction to generate a 3' megaprimer from pBAD18-BTE template. This
purified megaprimer and primer 16 were used to generate the complete BTE-H204A fragment,
which was inserted as described for BTE to form plasmid pBAD18-BTE-H204A.
To generate a low copy vector harboring the PBAD promoter system, the 1697-bp fragment
between the start of the araC gene and the end of the rrnB terminator was amplified from
pBAD18 (Guzman 1995) with primers 17 and 18. This insert was treated with T4 polynucleotide
kinase and blunt ligated into EcoRV-digested plasmid pBT-2 (Lynch 2006) to form plasmid
pBAD35.
A fragment containing the BTE gene was amplified from pBAD18-BTE using
primers 19 and 20 and inserted into pBAD35 between the XmaI and SphI sites to generate
plasmid pBAD35-BTE. The same procedure was performed using pBAD18-BTE-H204A as a
template to form plasmid pBAD35-BTE-H204A.
To generate a high copy vector harboring the PBAD promoter system, the 1693-bp fragment
between the start of the araC gene and the end of the rrnB terminator was amplified from
pBAD33 with primers 21 and 22. The 2284-bp fragment of pUC19 (New England Biolabs)
containing the origin and AmpR marker was amplified with primers 23 and 24. These two
fragments were digested with XhoI and BglII and ligated to form plasmid pBAD34. The
XmaI/HindIII fragments containing BTE or BTE-H204A from pBAD18-BTE and pBAD18BTE-H204A were inserted into pBAD34 to form plasmids pBAD34-BTE and pBAD34-BTEH204A. All plasmid constructs described above were verified by sequencing.
43
To create a template plasmid for generating a linear cassette for homologous recombination
of BTE into the chromosome, the kan cassette from pKD13 was amplified using primers 25 and
26, which introduced XhoI sites to the 5' and 3' ends of the PCR product. This product and two
vectors, pBAD34-BTE and pBAD34-BTE-H204A, were digested with XhoI and ligated to form
plasmids pBAD34-BTE-kan and pBAD34-BTE-H204A-kan. Clones were selected with the kan
resistance gene oriented in the same direction of transcription as araC on the plasmids.
2.2.5 Cell transformation, media, and growth
Plasmids were electroporated into strain RL08 and selected on Luria Bertani (LB) agar
(Becton Dickinson, Cockeysville, MD) containing 25 µg/mL kanamycin (pBAD35 constructs) or
50 µg/mL ampicillin (pBAD18 constructs) and 34 µg/mL chloramphenicol. All cultures were
grown in a 37°C shaker at 250 rpm. Overnight cultures inoculated from single colonies were
used to inoculate shake flasks containing LB medium (Becton Dickinson) supplemented with
0.4% glycerol and antibiotics. The cultures were induced at an OD600 of 0.2 with 0.2% (w/v) Larabinose. Samples of cell culture (2.5 mL) were taken for fatty acid analysis at indicated times.
2.2.6 Quantitative PCR for determination of copy number
Immediately prior to induction at OD 0.2, 1 mL of cell culture was collected and centrifuged
at 16000xg for 1 minute. The cell pellet was resuspended in 100 µL of deionized water. One
microliter of resuspended cell pellet was used directly as the template in a qPCR reaction with
MaximaTM SYBR green/fluorescein qPCR master mix (Fermentas, Glen Burnie, MD). Primers
27 and 28 were used for BTE amplification, and primers 29 and 30 for chromosomal ompA
amplification. SYBR Green fluorescence was monitored with a Bio-Rad CFX96 optical reaction
44
module (Bio-Rad, Hercules, CA). Threshold cycle (Ct) values were calculated by Bio-Rad
CFX Manager software.
2.2.7 Fatty acid extraction and methylation
To 2.5 mL samples of cell culture (three replicates for each culture at each sampling time), 5
µL of 10 mg/mL heptadecanoic acid (Fluka, Buchs, Switzerland) dissolved in ethanol and 50 µL
of 10 mg/mL pentadecanoic acid (Acros Organics, Geel, Belgium) dissolved in ethanol were
added as internal standards. Next, 100 µL of glacial acetic acid and 5.0 mL of a 1:1 (v/v)
chloroform/methanol mixture were added [Bligh 1959]. The samples were inverted several
times, vortexed vigorously, and centrifuged. The aqueous layer and cell debris were removed by
aspiration and the chloroform layer was stored at -80°C until further processing. To methylate
the fatty acids, the chloroform layer was thawed and evaporated under a nitrogen stream.
Residual water was removed by lyophilization for approximately 1 hour. To the dried residue,
0.5 mL of 1.25 M HCl in methanol (Fluka) was added, and the reaction was incubated overnight
(14 to 16 hours) at 50°C. The reaction mixtures were quenched by the addition of 5 mL of 100
mg/mL aqueous NaHCO3 (Sigma-Aldrich Corp., St. Louis, MO), and fatty acid methyl esters
were extracted twice into 0.5 mL hexane. The hexane layers were collected for analysis.
2.2.8 Gas chromatography/mass spectrometry of fatty acid methyl esters
GC/MS analysis was performed on a model 7890 Agilent GC (Agilent Technologies, Inc.,
Santa Clara, CA) with a 30 m x 0.25 mm DB-5 capillary column (Agilent Technologies) and a
model 5973N mass spectrometer. The oven temperature program was 100°C for 2 min, 150°C
for 4 min, and a ramp to 250°C at a rate of 4°C/min. One microliter of sample was injected with
45
a 1:10 split ratio. Peak identification was achieved by comparison to internal standards and to
the NIST Mass Spectral Database for species not present in the standard, including cis-3decenoic acid (10:1), cis-5-dodecenoic acid (12:1), 3-hydroxydodecanoic acid (12:0-OH), and
myristoleic acid (14:1) (Figure 2.3). Quantification was achieved by comparison of integrated
peaks to calibration curves of a fatty acid methyl ester standard (Supelco No. 18918) with added
methyl heptadecanoate (Fluka) and methyl pentadecanoate (Fluka).
Due to the high
concentration of dodecanoic acid in BTE-expressing cultures, 20-fold dilutions were injected to
accurately quantify this species. Concentrations of decenoic, dodecenoic, and tetradecenoic
methyl esters were estimated using the sensitivity ratio of hexadecenoic to hexadecanoic methyl
esters in the external standard.
Calculated sample concentrations were normalized to the
recovered concentrations of internal standards and averaged for all replicates.
2.2.9 Decane extraction of fatty acids
After 34 hours, 40 mL of decane (Acros Organics) was added to approximately 410 mL of
each culture, and the mixtures were placed in a shaker at 250 rpm for 30 minutes. The resulting
emulsions were acidified to facilitate phase separation by the addition of 10 mL of concentrated
hydrochloric acid, shaken for one minute, and centrifuged at 2500xg for 20 minutes or 16000xg
for 10 minutes. The decane layer was collected for catalytic conversion.
2.2.10 Catalytic decarboxylation
The extracted fatty acids dissolved in decane were decarboxylated to alkanes over a 1 weight
percent Pd/C catalyst (Sigma-Aldrich) in a plug flow reactor. The catalyst (1.1 g) was loaded
46
Figure 2.3 Representative mass spectra of fatty acid methyl esters present in methylated BTEexpressing culture extracts but not present in FAME standards. 10:1, top; 12:1, second from top;
12:0-OH, second from bottom; 14:1, bottom. Spectra are from a 23 hour extraction of a strain expressing
BTE on plasmid pBAD35-BTE. The molecular ions for 12:1 (m/z 212), 12:0-OH (m/z 229), and 14:1
(m/z 240) are all present.
into in a 0.25 inch tubular stainless steel reactor operating in an upflow configuration surrounded
by aluminum blocks heated externally by a well-insulated furnace. The catalyst bed was held in
place by plugs of quartz wool at the reactor entrance and exit. Prior to reaction, fresh catalyst
47
was reduced in flowing H2 (250 cm (STP) min at 300 C (5 C min ramp) for 4 h). The
3
.
-1
o
o .
-1
temperature was maintained at 300°C and the pressure was maintained at 12 bar for reaction
experiments. A liquid solution containing the fatty acids in the decane extraction was introduced
(0.05 mL/min) into the upflow reactor using an HPLC pump along with a 5 percent H2 co-feed
flow of 250 cm3(STP).min-1. The effluent liquid was collected at room temperature in a gasliquid separator (Penberthy, Prophetstown, IL) and drained for GC/MS analysis (Shimadzu GC2010 (Shimadzu, Kyoto, Japan) with a mass spectrometer and DB-5ms column from Alltech).
Alkane concentrations were quantified by comparison to external standards containing undecane,
dodecane, pentadecane, and hexadecane in a decane solvent. The tridecane peak was identified
by its mass spectrum (Figure 2.4) and its concentration was determined using an estimated
response factor. A small quantity of undecane (2 µmol/mL) was present as an impurity in the
decane received from the manufacturer. This concentration was subtracted from all undecane
concentrations of converted culture extractions to yield the reported concentrations.
2.3 Results and discussion
2.3.1 Determination of maximum theoretical yield of FFAs
Metabolic engineering efforts typically aim to maximize yield, productivity, and titer of the
desired target compound. The iAF1260 metabolic network reconstruction of E. coli [Feist 2007]
was the most comprehensive model available at the time of this study, and it was used with
minor modifications described in Materials and Methods to provide the metabolic network
connectivity and constraints for posing an optimization problem to maximize FFA production
from various carbon sources. This was particularly useful in the case of FFAs, for which
cofactor utilization is complex and for which the maximum theoretical yield cannot be readily
48
%
57
100.0
43
75.0
71
50.0
85
25.0
99
0.0
50.0
75.0
100.0
112
126
125.0
141
156
150.0
184
175.0
210 221
200.0
225.0
255
250.0
288
275.0
300.0
319
325.0
Figure 2.4 Mass spectrum of tridecane in a decarboxylated decane extraction. Extracted from the
34 hour culture of RL08/pBAD33-ACC/pBAD35-BTE. The molecular ion (m/z 184) is present.
deduced by examination of central metabolism. Table 1.1 (shown in Chapter 1) presents the
calculated maximum yields for palmitic (saturated C16), myristic (saturated C14), and lauric
(saturated C12) acids on a mass basis from glycerol, D-glucose, D-xylose, and L-arabinose,
which range between 0.3 to 0.4 g g-1. For comparison, the maximum theoretical yield of ethanol
from glucose is 0.51 g g-1. A more relevant measure of comparison is the energy yield, or kJ g-1,
where the energy is calculated from a combustion reaction as would be the case for utilizing
these molecules in a combustion engine. Using standard heats of combustion (∆Hcº, which
admittedly does not represent temperature and pressure within an engine) of 29.67 kJ g-1 for
ethanol and 36.83 kJ g-1 for lauric acid [Haynes 2012], the comparable energy yields have a
smaller difference than mass yields, at 15.1 kJ g-1 glucose for ethanol and 12.9 kJ g-1 glucose for
lauric acid. Conversion of lauric acid to undecane (∆Hcº = 47.54 kJ, [Haynes 2012]), accounting
for the loss of mass from decarboxylation (maximum theoretical yield of 0.27 g g-1 glucose),
results in an energy yield of 13.0 kJ g-1. Additional energy savings as a result of not needing to
distill the final product would likely recover this small loss of energy yield, if the fatty acid could
49
be produced at near the maximum theoretical yield, as has been approached with fermentation
products such as ethanol (88%), n-butanol (88%), and isobutanol (86%) [Ohta 1991, Shen 2011,
Atsumi 2008].
2.3.2 Initial strain construction and optimization of BTE copy number
The development of an initial fatty acid overproducing strain followed the metabolic
engineering strategy presented in Figure 2.5. First, to prevent catabolism of FFAs, the gene
encoding acyl-CoA synthetase (fadD) was deleted from the chromosome. Second, to prevent
consumption of the inducing agent L-arabinose, the araBAD operon was deleted. This operon
encodes three genes involved in the initial steps of L-arabinose degradation: L-ribulokinase, Larabinose isomerase, and L-ribulose-5-phosphate 4-epimerase.
The resulting strain, K-12
MG1655 ∆fadD ∆araBAD, is designated strain RL08. Finally, to hydrolyze FFAs from acylACP, a codon-optimized plant acyl-ACP thioesterase (BTE) from Umbellularia californica was
cloned into various arabinose-inducible plasmids and transformed into strain RL08. The four
selected plasmids were pBAD34 (pUC origin), pBAD18 (pBR322 origin without rop), pBAD33
(pACYC origin), and pBAD35 (pBBR1 origin). These plasmids range from very high reported
copy number with the pUC origin (pBAD34) [Yanisch-Perron 1985], to medium copy number in
pBAD18 [Guzman 1995, Cronan 2006] to low copy number in pBAD33 [Guzman 1995, Chang
1978]. The only reported copy number of the pBBR1 origin in E. coli is 30 to 40 copies per cell
[Antoine 1992]. An identical set of plasmids expressing a non-functional version of BTE with
histidine-204 mutagenized to an alanine were also transformed into strain RL08 to serve as
negative controls.
50
Figure 2.5 Metabolic engineering strategy for overproduction of FFAs overlaid on the fatty acid
biosynthesis pathway. FFAs are generated by heterologous expression of an acyl-ACP thioesterase from
Umbellularia californica (BTE), which hydrolyzes FFAs from ACP. Activation of FFAs for degradation
in the β-oxidation cycle is blocked by deletion of fadD (red X). The four subunits of E. coli ACC
(AccABCD) are overexpressed.
To determine the optimal plasmid copy number for expressing BTE, cultures of strain RL08
harboring each plasmid (with both functional and non-functional BTE) were grown in shake
flasks containing 50 mL of LB medium supplemented with 0.4% (v/v) glycerol. The OD600 was
monitored (Figure 2.6), and relative copy numbers of each plasmid were determined by
quantitative PCR from cell cultures immediately before induction of transcription at OD 0.2, and
during early stationary phase after an elapsed time of 7.7 hours from inoculation (Table 2.1).
Dramatically lower cell densities were observed after approximately 5 hours when BTE was
expressed on plasmids pBAD34 and pBAD18 (Figure 2.6). Non-functional BTE-H204A did not
exhibit similarly reduced cell densities, strongly suggesting that the decreased cell densities were
due to thioesterase activity.
51
A
4
600
fatty acids (mg/L)
6
OD600
B
pBAD34
pBAD18
pBAD33
pBAD35
chromosome
2
0
pBAD34
pBAD18
pBAD33
pBAD35
chromosome
OD 600
6
4
500
C12
non-C12
400
300
200
100
0
BTE – + – + – + – + – +
copy ~1
~3
~6 ~24 ~100
2
0
0
2
4
6
8
10
12
time (h)
Figure 2.6 Growth curves and fatty acid titers of E. coli RL08 cultures harboring plasmids
expressing BTE and BTE-H204A. (A) Growth curves of cultures harboring plasmids expressing BTE
(filled markers) or non-functional BTE-H204A (open markers) monitored by optical density (OD600).
Cells were grown at 37°C in shake flasks containing LB medium supplemented with 0.4% (v/v) glycerol
and appropriate antibiotics for each vector. Cells were induced at an OD600 of 0.2 by the addition of a
final concentration of 0.2% (w/v) L-arabinose. (B) Total fatty acid titers (mg/L culture medium, open
bars) and C12 chain length fatty acid titers (saturated and estimated unsaturated, filled bars) extracted from
cultures shown in (A) at approximately 23 hours post-inoculation. Error bars represent standard
deviations about the mean of two or three replicate samples for either C12 fatty acids (lower bars) or total
fatty acids including C12 (upper bars).
Relative copy numbers trended as expected (Table 2.1), with pBAD34 exhibiting the highest
number of copies per copy of the selected chromosomal internal standard, ompA, during preinduction exponential growth and early stationary phase. The poorly characterized pBBR1
origin (present in pBAD35) is shown here to be present at low copy numbers (~5), similar to that
of the pACYC origin during exponential growth. As expected, the pBAD18 construct was shown
to have an intermediate copy number. BTE copy numbers relative to ompA increased in early
stationary phase for all vectors except pBAD35, which showed no change. This is likely due to
52
Table 2.1 Copy numbers determined for BTE and BTE-H204A harboring plasmids by qPCR.
Copy numbers are shown as copies of BTE per copy of chromosomal gene ompA from cultures of strain
RL08 harboring BTE on the plasmids shown. Culture samples were harvested during early exponential
phase (OD600 of approximately 0.2) immediately prior to induction and during early stationary phase (7.7
hours after inoculation). Errors are propagated standard deviations about the mean of three replicate
samples.
Plasmid
pBAD34-BTE
pBAD34-BTE-H204A
pBAD18-BTE
pBAD18-BTE-H204A
pBAD33-BTE
pBAD33-BTE-H204A
pBAD35-BTE
pBAD35-BTE-H204A
Copy number (per ompA)
exponential
stationary
100 ± 9
96 ± 15
24 ± 3
24 ± 2
5 ±1
7 ±1
3 ±0
4 ±0
1200 ± 240
840 ± 170
110 ± 10
70 ± 8
19 ± 2
13 ± 3
5±0
2±0
the presence of multiple replication forks on the chromosome during rapid growth [Nordström
2006], or the onset of nutrient limitations in the medium that stop replication from oriC while
replication continues from many plasmid origins [Friehs 2004]. The plasmids expressing BTEH204A all displayed lower copy numbers during stationary phase than the corresponding
plasmids expressing BTE, possibly as a result of increased cell lysis in these induced cultures.
Variation in the lysis of whole cell templates has been previously implicated as a source of error
in determining plasmid copy numbers by quantitative PCR [Providenti 2006]. Consistent with
this hypothesis, no discrepancy in copy number was observed between plasmids harboring BTE
and BTE-H204A during exponential growth prior to induction.
To determine whether copy number correlated with fatty acid production, total fatty acids
were extracted and derivatized for analysis by GC/MS after 23 hours (Figure 2.6). In strains
expressing functional BTE, the predominant fatty acid chain length shifted dramatically from C16
to C12, with C12 fatty acids accounting for up to 75 percent of the total fatty acid composition.
For the culture expressing pBAD33-BTE, approximately 15 percent of C12 and 59 percent of C14
fatty acids were unsaturated.
53
Lower but significant levels of hydroxylated C12, an
intermediate in the fatty acid elongation cycle, were also observed but not quantified in
functional BTE expressing strains (Figure 2.3). Of the functional BTE expressing plasmids, the
lowest titer (0.25 ± 0.01 g/L) was observed from the highest copy number plasmid, pBAD34BTE. The two plasmids with the lowest overall copy number in both exponential and early
stationary phase, pBAD33-BTE and pBAD35-BTE, accumulated the highest titers of fatty acids
(0.54 ± 0.00 g/L and 0.48 ± 0.00 g/L, respectively), with medium-copy pBAD18-BTE
accumulating a slightly lower quantity (0.40 ± 0.03 g/L). Expression of a single copy of BTE
from the chromosome produces a sub-optimal titer of less than 0.2 g/L. Strains harboring a nonfunctional BTE gene on the various plasmids accumulated similar quantities of predominantly
C16 fatty acids, as expected.
It can therefore concluded that expression of BTE on a low copy number plasmid results in
optimal fatty acid production. One possible explanation is that higher quantities of functional
thioesterase are produced from the high copy number plasmids, resulting in an initially rapid rate
of accumulation of medium-chain FFAs.
As there is no known protein-mediated export
mechanism for FFAs, they likely pass across the inner membrane via a transmembrane flip
[Black 2003]. Dodecanoic acid, the dominant BTE product, has a much higher water solubility
than longer chain fatty acids [Vorum 1992] and can possibly cross the outer membrane through
porins [Hearn 2009]. Too rapid an accumulation of C12 and C14 FFAs in the inner membrane
may disrupt the membrane integrity and result in cell lysis. A second possible explanation is that
higher quantities of functional thioesterase significantly deplete C12 and C14 acyl-ACPs destined
for membrane phospholipid incorporation, resulting in a reduced number of viable cells.
54
2.3.3 Co-overexpression of acetyl-CoA carboxylase
The conversion of acetyl-CoA to malonyl-CoA by acetyl-CoA carboxylase (ACC) has been
identified as a rate-limiting step in fatty acid biosynthesis [Davis 2000]. To see if higher
production could be achieved over expression of BTE alone, the four subunits of acetyl-CoA
carboxylase (ACC) were cloned as an artificial operon (accDABC) on a low copy arabinoseinducible plasmid to yield pBAD33-ACC. To co-express BTE, pBAD35-BTE was selected due
to its compatibility with pBAD33-ACC and its high level of fatty acid overproduction. Four
cultures of E. coli strain RL08 harboring combinations of either pBAD33 or pBAD33-ACC, and
pBAD35-BTE or pBAD35-BTE-H204A, were grown in shake flasks in 500 mL of LB medium
supplemented with 0.4% (v/v) glycerol as a carbon source. Fatty acids were extracted and
derivatized for analysis by GC/MS at several times (Figures 2.7 and 2.8).
Elevated levels of fatty acids were detected at the onset of stationary phase in cultures
expressing BTE, with maximum accumulation observed at approximately 29 hours postinoculation. At this time, the strain expressing both BTE and overexpressing ACC exhibited a
small increase in in fatty acid titer (0.81 ± 0.02 g/L) relative to a strain expressing BTE alone
(0.70 ± 0.01 g/L). Microarray data from our lab (see Chapter 3) suggests that expression of BTE
alone dramatically increases gene expression of two subunits of ACC (accBC), possibly by an
uncharacterized transcriptional activation mechanism. This result is in agreement with a
previous observation by Ohlrogge et al. [1995] wherein expression of BTE in E. coli increases
the levels of biotin carboxyl carrier protein (AccB). The expression levels of other subunits were
not quantified in this prior work. Additionally, heterologous expression of BTE in the seeds of
the rapeseed plant, Brassica napus was shown to increase expression levels of its native biotin
carboxyl carrier protein and biotin carboxylase [Eccleston 1998]. Plants and bacteria both have
55
similar multi-subunit acetyl-CoA carboxylases [Cronan 2002]. Expression of BTE also likely
relieves allosteric inhibition of ACC by acyl-ACPs, which in combination with increased
transcription of at least 2 subunits, may partly explain the modest effect. The strain expressing
only ACC does not overproduce fatty acids, as previously observed due to feedback inhibition of
ACC by accumulated acyl-ACPs [Davis 2001].
Figure 2.7 Representative GC/MS chromatograms of methylated fatty acid extracts from strains
expressing either BTE or BTE-H204A, and either empty vector or ACC. Samples were taken at 34
hours; undiluted hexane extracts are shown. Very large increases in the levels of C12 and C14 species are
clearly observed in BTE-expressing cultures. Labelled peaks that are not present in the standard mixture
(10:1, 12:1, 12:0-OH, and 14:1) were identified by comparing their mass spectra to the NIST library.
2.3.4 Extraction and conversion of fatty acids to alkanes
To demonstrate a complete process for fuel production, fatty acids were extracted from
approximately 400 mL of the overproducing culture (the remaining volume from an original
56
10
A
OD600
1
ACC BTE
pBAD33
(–) / pBAD35-BTE(–)
H204A
pBAD33-ACC
(+) (–)/ pBAD35BTE-H204A
pBAD33
(–) / pBAD35-BTE
(+)
pBAD33-ACC
(+) (+)/ pBAD35-
0.1
BTE
0.01
B
[Fatty Acid] (µg/mL)
0
800
5
6 hr
10
15
20
time (h)
10 hr
18 hr
25
29 hr
30
35
34 hr
C12
600
Other
400
200
0
ACC (-) (+) (-) (+) (-) (+) (-) (+) (-) (+) (-) (+) (-) (+) (-) (+) (-) (+) (-) (+)
BTE (-) (-) (+) (+) (-) (-) (+) (+) (-) (-) (+) (+) (-) (-) (+) (+) (-) (-) (+) (+)
Figure 2.8 Growth and fatty acid production of strain RL08 harboring combinations of plasmids
pBAD33 or pBAD33-ACC, and pBAD35-BTE-H204A or pBAD35-BTE. Cells were grown at 37°C in
shake flasks containing LB medium supplemented with 0.4% (v/v) glycerol, 25 µg/mL kanamycin, and
34 µg/mL chloramphenicol. (A) OD600 as a function of time from inoculation, (B) total (filled bars) and
C12 chain length (saturated and estimated unsaturated, open bars) fatty acid titers (µg/mL culture medium)
for selected times during cell growth as indicated (6 h, 10 h, 18 h, 29 h, and 34 h). Error bars represent
standard deviations about the mean for three replicate samples for either C12 fatty acids (upper bars) or
total fatty acids including C12 (lower bars).
culture volume of 500 mL) overexpressing both ACC and BTE at 34 hours, with a total
measured saturated and unsaturated C12 fatty acid titer of 0.36 ± 0.02 g/L. Decane was selected
to facilitate analysis of the dominant undecane product, due to the presence of significant
undecane impurities in other commercially available alkanes of higher molecular weight, such as
tridecane. While decane is mildly toxic to microbes [Sardessai 2002], larger alkanes such as
dodecane are essentially non-toxic and their use to extract metabolites during cell growth has
57
been previously demonstrated [Janikowski 2002, Newman 2006]. The emulsion resulting
from decane addition was acidified and centrifuged to facilitate phase separation, and the decane
layer was collected. Approximately 60 percent of the decane added could be collected as a deemulsified liquid layer. Fatty acids in the decane extractions were decarboxylated at 100 percent
conversion in the presence of hydrogen over a 1 weight percent Pd/C catalyst in a plug flow
reactor operating at 300°C and 12 bar. The catalytic decarboxylation of fatty acids over Pd/C
catalysts has previously been demonstrated using a semi-batch reactor operating at 300°C [MäkiArvela 2007]. Under these conditions unsaturated fatty acids are fully hydrogenated, which is
desirable for stability of the product during storage. In the collected alkane product, 0.44 ± 0.03
g/L (culture volume) undecane was obtained (Figure 2.9), representing a complete recovery and
conversion of C12 fatty acids from the culture medium. Smaller amounts of tridecane and
pentadecane were also present, as expected from the fatty acid composition in which C14 and C16
species are most abundant after C12.
Ultimately, an industrial process can be envisioned in which a desired hydrocarbon product is
used to both extract fatty acids from a fermentor and act as the solvent for the decarboxylation
reaction (Figure 2.9). A product stream could be continuously or semi-continuously collected
that matches the fatty acid production rate, with the remainder of the alkane phase recycled as an
extractant in a two-phase partitioning bioreactor.
2.4 Conclusion
A strain of E. coli that exhibits an approximately seven-fold increase in fatty acid production
(predominantly C12 fatty acids) over the baseline strain (RL08) was metabolically engineered.
58
B
Hydrolyzed
cellulose
Alkane
+ FA
Recycle
A
Bioreactor
Alkane
Product
Catalyst
Bed
Figure 2.9 Chromatograms of decarboxylated fatty acid extracts in decane, and an envisioned
semi-continuous two-phase partioning bioreactor process for production of alkanes. (A)
Chromatograms of collected decane layers from a 34 hour fatty acid overproducing culture
(RL08/pBAD33-ACC/pBAD35-BTE, second from top trace) and a negative control culture
(RL08/pBAD33/pBAD35-BTE-H204A, second from bottom trace) following decarboxylation at 300°C
in a plug flow reactor containing 1% (w/w) Pd/C catalyst in the presence of hydrogen. A standard
containing undecane and dodecane in a decane solvent (top trace) and a blank decane sample (bottom
trace) are shown for comparison. (B) Envisioned semi-continuous process for alkane production from
FFA overproducing cultures. The FFA overproducing strain of E. coli is cultivated in a bioreactor and
hydrolyzed cellulose is fed as a carbon source. FFAs are extracted into an alkane layer (eg. dodecanoic
acid is extracted into undecane), and the extract is passed over a Pd/C catalyst bed to decarboxylate
dodecanoic acid to undecane. A portion of the undecane is recycled back to the bioreactor for continued
extraction.
One key aspect of the strategy was utilizing a low copy number vector for expression of BTE.
The successful conversion of overproduced fatty acids to a useful enriched liquid alkane stream
was demonstrated by a novel process that couples microbial fermentation of FFAs to a catalytic
reaction step. Further genetic and process improvements are currently underway to increase fatty
acid yields and alkane recovery.
59
Chapter 3: Functional genomics study of FFA-overproducing E. coli
In Chapter 2, a medium-chain length FFA overproducing strain of E. coli was engineered and
the expression level of BTE was optimized to maximize FFA titer. Due to an undefined medium
being used, it was not possible to determine a yield from glycerol. However if only glycerol
were utilized in the medium, it would correspond to approximately 42% of the maximum
theoretical yield. Later work in our group utilizing a minimal medium containing only glucose
as a carbon source, has enabled us to determine that up to 41% of the maximum theoretical yield
can be achieved in a batch culture. In this chapter, additional literature will be reviewed related
to overproduction of FFAs in E. coli subsequent to submission of the material in Chapter 2 for
publication, as well as efforts to determine the bottleneck to achieving higher yields through fatty
acid biosynthesis.
Our group has undertaken four parallel strategies to improve FFA yields that are shown in
Figure 3.1. None of the tested rational strain modifications resulted in increased fatty acid titers,
including heterologous expression of ACC from Corynebacterium glutamicum [Miyahisa 2005],
deletion of the ackA-pta operon to reduce the generation of alternative products from acetyl-CoA
pools [Zha 2009], and deletion of aas to prevent the possibility of FFAs futile cycling back into
the fatty acid elongation pathway.
To complement rational strain design efforts, a high-
throughput assay employing the lipophilic dye Nile red [Greenspan 1985] was used to screen a
transposon mutagenesis library to identify strains with higher Nile red fluorescence, with
secondary screens analyzing FFA titers [Hoover 2012]. This approach generated numerous hits,
most of which were identified as false positives in secondary rounds of screening. One hit
* Portions of this chapter were published in Applied and Environmental Microbiology (Lennen 2011)
60
Current fatty acid
overproducer
Rational strain
modifications
Random
approaches
Bioprocess
optimization
Functional
Genomics
Targets
for improved
production
Figure 3.1 Parallel strategies undertaken by our group to develop a second-generation FFAoverproducing strain with improved production characteristics. These strategies include individual
rational strain modifications, random approaches that utilize gene disruption or overexpression libraries
combined with high-throughput screening, bioprocess optimization to determine optimal nutrient
supplementation and feeding regimes, and functional genomics to attempt to rationalize an observed
phenotype with gene, protein, and metabolite profiles. This chapter will highlight a functional genomics
study conducted to determine physiological differences between a FFA-overproducing and a negative
control strain.
(∆fadB) increased titers by a marginal amount, and a strain with BTE chromosomally integrated
in fadD, fadE, and fadAB loci has been implemented as a new platform strain in our laboratory
(TY05).
Results are published elsewhere of a chemostat study aimed at establishing key
bioprocess parameters for optimizing FFA production [Youngquist 2012]. Work in this area is
still in progress to generate models that will guide the design of an optimal bioprocess for FFA
overproduction.
This chapter will discuss a fourth strategy, which involved a differential comparison of the
viability, morphology, transcript levels, and protein levels in strains of E. coli that overproduce
FFAs compared to control strains engineered to produce the same protein overexpression burden
using the non-functional thioesterase introduced in Chapter 2. By early stationary phase, an 85%
reduction in viable cell counts and exacerbated loss of inner membrane integrity were observed
61
in the FFA overproducing strain. These effects were enhanced in strains endogenously
producing FFAs compared to strains exposed to exogenously fed FFAs. Under two sets of
cultivation conditions, long-chain unsaturated fatty acid content greatly increased and the
expression of genes and proteins required for unsaturated fatty acid biosynthesis were
significantly decreased. Membrane stresses were further implicated by increased expression of
genes and proteins of the phage shock response, the MarA/Rob/SoxS regulon, and the nuo and
cyo operons of aerobic respiration. Gene deletion studies confirmed the importance of the phage
shock proteins and Rob for maintaining cell viability, however little to no change in FFA titers
was observed after 24 h cultivation. The results of this study served as a baseline for current
(Chapters 4 and 5) and future (Chapter 6) work to improve FFA yields and titers in E. coli.
3.1 Introduction
Microbially derived FFAs are attractive intermediates for producing a wide range of high
energy density biofuels from sustainable carbon sources such as biomass [Handke 2011]. As
discussed in Chapter 1, FFAs can be extracted from culture medium and catalytically converted
to esters or alkanes [Mäki-Arvela 2007, Lennen 2010]. Alternatively, enzymatic pathways exist
for intracellular conversion to esters [Kalscheuer 2006, Steen 2010], olefins [Beller 2010,
Mendez-Perez 2011, Rude 2011], alkanes [Schirmer 2010], or fatty aldehydes and fatty alcohols
[Doan 2009, Steen 2010, Teerawanichpan 2010], which are also described in detail in Chapter 1.
These pathways can either be exploited in their native host or heterologously expressed in a
genetically pliable microorganism [Alper 2009]. The physical and chemical properties of the
resulting products are dependent on chain length and hydrophobicity, however medium-chain
62
length (8 to 14 carbon) methyl esters, olefins, and alkanes exhibit many properties analogous
to those of diesel and jet fuel, and are therefore potential drop-in replacements [Knothe 2010,
Murphy 2004].
Several studies have demonstrated FFA overproduction in E. coli [Davis 2000, Lu 2008,
Lennen 2010, Steen 2010, Voelker 1994].
In each, the key strain modifications included
overexpression of one or more cytosolic acyl-acyl carrier protein (ACP) thioesterases and
deletion of fadD, or both fadD and fadE, encoding an acyl-coenzyme A (CoA) synthetase and
acyl-CoA dehydrogenase, respectively. Overexpression of an acyl-ACP thioesterase depletes the
level of acyl-ACP intermediates, which feedback inhibit enzymes of fatty acid biosynthesis
[Davis 2000, Heath 1996]. Deletion of fadD and/or fadE eliminates catabolism of fatty acids by
the aerobic β-oxidation pathway [Overath 1969, Klein 1971]. The additional overexpression of
the native E. coli acetyl-CoA carboxylase (ACC) has been shown to improve fatty acid yields in
some metabolically engineered strains [Davis 2000, Lu 2008] but to have little impact in others
(Chapter 2).
Reported yields (% w/w of FA from a supplied carbon source) in the literature from FFA
overproducing E. coli are 4.8% in a fed-batch fermentation with glycerol as the sole carbon
source [Lu 2008], and 6% in shake flasks with glucose as a sole carbon source [Steen 2010],
representing less than 20% of the maximum theoretical yield of FFAs from either glucose or
glycerol. Very recent studies have achieved up to 65% of the maximum theoretical yield via
fatty acid biosynthesis (Table 1.2), but no analysis has been performed to determine the rationale
for varying yields. In comparison, strains of E. coli have been engineered to produce ethanol at
over 88% of the maximum theoretical yield [Ohta 1991] and isobutanol at 86% of the maximum
theoretical yield [Atsumi 2008] from glucose. The limitation to achieving higher yields of FFAs
63
in engineered strains is currently unknown.
Potential metabolic bottlenecks have been
identified by in vitro studies using cell-free extracts supplemented with additional substrates,
cofactors, or enzymes involved in FFA biosynthesis [Liu 2010a]. It is not yet known whether the
observed findings, such as a potential limitation in malonyl-CoA levels, translates to whole cell
biocatalysts, as we have observed little or no improvement in FFA titers in acyl-ACP
thioesterase-expressing strains that simultaneously overexpress native E. coli ACC (Chapter 2)
or Corynebacterium glutamicum ACC (unpublished results).
In addition to metabolic
bottlenecks, limitations to production can stem from product inhibition, toxicity, or other indirect
effects [Nicolaou 2010].
In Chapter 2, we observed a copy number dependent variation in FFA titers and maximum
cell densities by expressing an acyl-ACP thioesterase from Umbellularia californica [Voelker
1992] on a series of plasmids with identical inducible promoters (Chapter 2). We postulated that
growth inhibition was due either to a depletion of acyl-ACPs that reduced the ability of cells to
synthesize phospholipids necessary for growth, or due to alteration of membrane integrity as a
result of accumulation of FFAs in the cell envelope. In this study, further characterization was
performed on the cellular impacts of endogenous FFA overproduction, including viable cell
counts, staining with the membrane-impermeable SYTOX Green nucleic acid dye, forward
scatter flow cytometry, and microscopic observation. Significant losses in cell viability were
observed beginning in the transition between log phase and stationary phase, accompanied by
increased permeability to SYTOX Green and drastic changes in cell morphology. The direct
cause of these observations was not obvious and prevented implementation of directed metabolic
engineering strategies to alleviate toxicity and increase FFA titer and yield. While prior studies
have examined the impact of hydrophobic compounds such as hexane [Hayashi 2003] and n-
64
butanol [Rutherford 2010] on E. coli gene expression, the only previous systems biology study
investigating exposure to FFAs was a differential proteomics analysis that identified several
proteins with increased expression in the presence of exogenously fed oleic acid (C18:1∆9) [Han
2008]. Exposure to endogenously produced FFAs is anticipated to result in a more severe
cellular impact than exogenously added FFAs, as excretion would require first traversing the
inner membrane, and subsequently the periplasm with its peptidoglycan network and the outer
membrane.
In order to understand the mechanisms underlying the observed phenotypes, we performed a
comparative systems biology study that identified key differences in transcript, protein, and lipid
profiles between a FFA overproducing strain and a control strain. This study, consisting of a
DNA microarray analysis coupled to a global LC/MS analysis of cellular peptides and
determination of the fatty acid profile by GC/MS as a function of time, represents the first
investigation of changes in gene expression in E. coli resulting from either exogenous or
endogenous exposure to medium-chain FFAs. Genes identified in two sets of parallel
experiments focused attention to transcriptional responses that were specific to FFA
overproduction. These filtered data sets demonstrated differential expression of stress regulons
and suggested targets for improving physiology and strain performance.
3.2 Materials and methods
3.2.1 Strains, plasmids, enzymes, media, and bacterial cultivation
Strains and plasmids are listed in Appendices I and II. All enzymes used for PCR and cloning
were purchased from New England Biolabs (Ipswich, MA) unless otherwise specified.
Oligonucleotides used for PCR, cloning, qPCR, and verification are listed in Appendix III. The
65
primary strain used in this work (RL08), is E. coli K-12 MG1655 ∆fadD ∆araBAD (Chapter
2).
These deletions eliminate aerobic β-oxidation (∆fadD) and prevent catabolism of L-
arabinose (∆araBAD), which was used as an inducing agent. FFA overproducing strains were
constructed by transforming RL08 with plasmids harboring a codon-optimized acyl-ACP
thioesterase from Umbellularia californica (BTE) [Voelker 1992] as described previously
(Chapter 2).
Plasmid pTrc99A-BTE places BTE under control of the IPTG-inducible Ptrc
promoter [Hoover 2011]. Plasmids pBAD35-BTE and pBAD33-BTE place BTE under control
of an arabinose-inducible PBAD promoter (Chapter 2). A set of control strains was constructed by
transforming RL08 with plasmids harboring a mutated gene encoding a non-functional
thioesterase (BTE-H204A) as previously described (Chapter 2). The non-functional thioesterase
is expressed at equal levels to BTE (Figure 3.2) providing protein overexpression stress to both
experimental and control strains. Plasmids were introduced by established electroporation
protocols or by a chemical transformation method [Chung 1989].
Figure 3.2 OD600-normalized abundances of BTE and BTE-H204A from shake flask cultures
determined by proteomics analysis. Further details are provided in sections 3.3.6-8 and 3.3.5.
Expression levels are shown for BTE and BTE-H204A at all protein sampling times in strain
RL08/pTrc99A-BTE (BTE) and pTrc99A-BTE-H204A (H204A) grown in EZ rich defined medium
supplemented with 0.2% glucose. Error bars represent the propagated standard deviations for the rolledup protein abundances normalized to the OD600 at that sampling point.
66
Two sets of functional genomics experiments were conducted under different cultivation
conditions.
First, biological triplicate cultures of strain RL08 harboring pTrc99A-BTE or
pTrc99A-BTE-H204A were grown at 37°C with 250 rpm shaking in 2 L shake flasks containing
500 mL of EZ Rich Defined medium (Teknova, Hollister, CA) supplemented with 0.2% glucose,
0.01 mM biotin, and 100 µg/mL ampicillin. Cultures were induced at an optical density at 600
nm (OD600) of 0.2 with a final concentration of 50 µM isopropyl β-D-thiogalactopyranoside
(IPTG).
Additional experiments were conducted in biological duplicate with strain RL08
harboring pBAD35-BTE and pBAD33, pBAD35-BTE and pBAD33-ACC, or pBAD35-BTEH204A and pBAD33, in a single wall baffled fermentor vessel (Applikon Biotechnology,
Schiedam, Netherlands) with a 12 L working volume containing 4 L of EZ Rich Defined
medium supplemented with 0.4% glycerol, 0.01 mM biotin, 34 µg/mL chloramphenicol, and 25
µg/mL kanamycin. A temperature of 37°C, pH 7.0 ± 0.2, and 700 rpm agitation were maintained
with an ez-Control unit (Applikon Biotechnology). A 5% solution of hydrochloric acid and 2 N
sodium hydroxide were utilized for maintaining pH, and a 1:1000 dilution of antifoam 204
(Sigma, St. Louis, MO) in water was added as necessary to minimize overflow due to foaming.
Air was sparged at a flow rate of 4 L/min. Cultures were induced with 0.2% L-arabinose at an
OD600 of 0.2.
Cultures sampled for qPCR analysis, flow cytometry, CFU/mL counts, and microscopy were
grown in biological duplicate or triplicate in baffled shake flasks with a 4X headspace, as
described for microarray experiments with RL08/pTrc99A-BTE. Exogenous addition of fatty
acids was performed with strains RL08 and RL08/pTrc99A, with 0.5 g/L of lauric acid (from a
50 g/L stock in ethanol), or an equivalent volume of ethanol as a negative control, added at an
OD600 of 0.1 to 0.2.
67
3.2.2 RNA sample preparation
Samples for RNA isolation for microarray experiments were collected by adding 10
OD600·mL of culture to 1/8th volume of a chilled solution of 5% saturated phenol in absolute
ethanol. Deionized water was added such that the ratio of aqueous solution to phenol/ethanol
was constant. After centrifugation (5000 × g, 4°C, 12 min), each supernatant was removed, and
the remaining cell pellets were flash frozen in a dry ice/ethanol bath and stored at -80°C. Total
RNA was purified from frozen cell pellets using the Qiagen (Valencia, CA) RNAprotect Bacteria
Reagent kit with lysozyme treatment and proteinase K digestion of cell lysates, and purification
of total RNA with the Qiagen RNeasy Mini Kit as described in the product handbook. For
quantitative RT-PCR, 0.8 OD600·mL of each culture was centrifuged (16000 × g, 1 min),
supernatants were removed, and cell pellets were stored at -80°C. RNA was extracted using the
Qiagen RNeasy Mini Kit according to the manufacturer's instructions. Residual DNA was
removed by DNase digestion (DNA-free Kit, Ambion, Austin, TX).
RNA samples were
quantified using a NanoDrop spectrophotometer (Thermo Scientific, Wilmington, DE), and RNA
quality was assessed using an Agilent BioAnalyzer (Agilent Technologies, Santa Clara, CA)
using the Prokaryote Total RNA Nano Series II chip set.
3.2.3 cDNA synthesis and hybridization
For microarray hybridization, 10 µg of RNA was used to synthesize cDNA, and 1 µg was
labeled with Cy3 according to the NimbleGen Arrays User's Guide (version 5.0). Cy3-labeled
cDNA samples were hybridized to Roche NimbleGen (Madison, WI) Escherichia coli K-12
(accession number NC_000913) microarray slides containing 4 arrays per slide and 72000
probes per array. The arrays were scanned on a GenePix 4000B scanner (Molecular Devices,
-5
68
Sunnyvale, CA) with a photomultiplier tube gain to achieve 1 × 10 normalized counts at
saturated intensity level. The images were processed and .pair files were generated based on
non-normalized intensities.
For quantitative RT-PCR, cDNA was synthesized using the GoScript Reverse Transcription
System (Promega, Madison, WI) with 1 µg of template RNA, 2.5 mM MgCl2, and random
hexamer primers. Extension was carried out for 1 h at 42°C.
3.2.4 Microarray data analysis
Scanned probe intensities were imported into ArrayStar software (DNASTAR, Madison, WI),
and processed with Robust Microarray Analysis (RMA) background correction, quantile
normalization to adjust the probe intensity distributions to be the same across all cross-compared
arrays, and RMA median polish summarization to generate gene level signal values from the
normalized probe intensities [Bolstad 2003]. Mean intensities were determined across all array
and sample technical replicates, and then across biological replicates to yield the final expression
values. Linear fold-changes in expression values are reported, with P-values calculated across
biological replicates by a Moderated t-Test [Smyth 2004] using ArrayStar software. Ratios with
a fold-change greater than or equal to |2| and with P-values less than or equal to 0.05 were
considered significant. To generate a filtered list of significantly regulated genes found in both
the glucose-supplemented and glycerol-supplemented cultures, significance criteria were relaxed
such that the fold change was greater than or equal to |1.8| and the P-value was less than or equal
to 0.1 for the glycerol-supplemented cultures, or 0.05 for the glucose-supplemented cultures.
Complete data are available as a NCBI GEO DataSet (Series GSE29424).
69
3.2.5 Quantitative RT-PCR (qPCR)
The cDNA was diluted 10-fold in water and 5.0 µL was used amplified using the Maxima
SYBR green/fluorescein qPCR Master Mix (Fermentas, Glen Burnie, MD). Primers are listed in
Appendix III. For the selected housekeeping gene rrs (encoding 16S rRNA), 5.0 µL of 100-fold
diluted cDNA was used as a template, due to its high abundance. SYBR Green fluorescence was
monitored with a Bio-Rad CFX96 optical reaction module (Bio-Rad, Hercules, CA). Threshold
cycle (Ct) values were calculated by Bio-Rad CFX Manager software, and fold changes were
calculated as 2-∆∆Ct with inner normalization to the rrs housekeeping gene [Livak 2001].
Standard deviations of ∆Ct values for three biological triplicates were propagated to obtain
standard errors for each fold-change value.
3.2.6 Protein sample preparation
For protein isolation, approximately 20 OD600·mL of culture was centrifuged as described for
RNA isolation. The resulting cell pellets were flash frozen in a dry ice/ethanol bath and stored at
-80°C. Cells were resuspended in 100 mM ammonium bicarbonate, pH 8.0 buffer and lysed
with a Barocycler NEP3229 using FT500 PULSE Tubes (Pressure BioSciences Inc., South
Easton, MA) (10 cycles, 35,000 psi for 20 s/cycle). Protein concentrations of the lysates were
determined using a Bradford assay, and 100 µg protein was denatured by the addition of 7 M
urea and 5 mM dithiothreitol, followed by incubation at 60 °C for 30 min. The samples were
diluted 10-fold with 100 mM ammonium carbonate and calcium chloride was added to a final
concentration of 1 mM. Samples were digested with trypsin (1:50 trypsin:protein ratio) for 3 h at
37 °C and desalted using 1 mL, 50 mg Discovery DSC-18 solid phase extraction (SPE) columns
(Supelco, St. Louis, MO). Each SPE column was conditioned using methanol and rinsed with
70
0.1% trifluoroacetic acid (TFA) in water. Samples were introduced to the columns and washed
with 95:5 water/acetonitrile containing 0.1% TFA. Excess liquid was removed from the columns
under vacuum, and samples were eluted using 80:20 acetonitrile/water and 0.1% TFA and
concentrated in a Savant SpeedVac (Thermo Scientific) to a final volume of 50-100 µL. Final
peptide concentrations were determined using a BCA protein assay.
3.2.7 Peptide analysis by capillary liquid chromatography/mass spectrometry
Identification and quantification of individual peptides was performed by LC-MS/MS. The
HPLC system and method used for capillary LC have been described in detail elsewhere [Lipton
2006]. MS/MS data acquisition on an LTQ Velos orbitrap mass spectrometer (Thermo
Scientific) began 20 min after sample injection and continued for 100 min over a mass (m/z)
range of 400 to 2000. For each cycle using a dynamic exclusion time of 60 s, the six most
abundant ions from MS were selected for MS/MS using a collision energy setting of 45%.
3.2.8 Proteomics data analysis
Quantitative analysis utilized the PNNL-developed Accurate Mass and Time (AMT) Tag
approach [Zimmer 2006] using a previously established AMT tag peptide database for E. coli
containing the characteristic accurate mass and LC separation elution time. The end result was a
usable list of quantitative protein identifications for each sample. Tools used for quantitative
analysis were a visualization program, VIPER [Monroe 2007], to correlate LC-MS features to
the peptide identifications in the AMT tag database, and DAnTE software for peptide peak
intensity value determination, normalization, and protein roll-up using Rrollup parameters
[Polpitiya 2008]. Only those proteins that were detected in two out of two instrumental technical
71
replicates, in two out of three sample technical replicates (fermentor samples only), and in all
biological replicates were considered for further analysis. The rolled up protein abundances or
abundances of single peptides, when they represented the only peptide identified from a protein,
were averaged and fold-changes in abundance between the BTE-expressing cultures and BTEH204A expressing cultures were calculated at each sampling point. A two-tailed t-test assuming
equal sample variance was performed using Microsoft Excel (Redmond, WA) on the averaged
biological replicate level data. Only those proteins that had a greater than |2|-fold change in
abundance and decrease in abundance and P ≤ 0.05 were considered to be significantly
differentially expressed.
3.2.9 Fatty acid extraction and analysis
Fatty acids were extracted and methylated from cell cultures as described in Chapter 2. Gas
chromatography/mass spectrometry (GC/MS) analysis and peak identification and quantification
was performed on a model 7890 Agilent GC with a model 5975 mass spectrometer as described
in Chapter 2, with the exception of also running BTE-expressing samples at higher split ratios to
measure highly abundant fatty acids, in place of using sample dilutions. Technical duplicates
from each independent culture were averaged, and the average fatty acid concentrations from
two (fermentors) or three (shake flasks) independent cultures were determined.
3.2.10 Glucose analysis
Culture samples (0.5 mL) were incubated in centrifuge tubes at 100°C for 15 min and filtered
using 0.22 µm centrifugal filters (16000 × g, 10 to 15 min). The flow-through was stored at 20°C prior to analysis. Samples were diluted five-fold in deionized water prior to injection of 2
72
µL onto a Waters ACQUITY UPLC system (Milford, MA) with an evaporative light
scattering detector. Samples were separated on a Waters ACQUITY UPLC BEH Glycan 1.7 µm
column with 0.2 mL/min of mobile phase (74.9% acetonitrile, 24.9% water, and 0.2%
triethylamine, adjusted to pH 9.1 with glacial acetic acid). Quantification was achieved by
comparison of peak areas with a glucose standard curve.
3.2.11 Acetate analysis
Approximately 10 mL of culture was centrifuged (5000 × g, 12 min, 4°C) and the supernatant
was stored at -80°C. Supernatants were thawed and centrifuged (16000 × g, 1 min) to remove
cell debris. To preserve samples during analysis, 0.1 mL of 85% phosphoric acid was added to 1
mL of culture supernatant and 1 µL was injected at 250°C using a 5:1 split ratio onto a Shimadzu
GC-2010 gas chromatograph (Columbia, MD) with a 30 m × 0.25 mm Supelco PTE-5 column
(manufacturer, city) and helium carrier gas at a column flow rate of 2.7 mL/min. The column
was held at 60°C for 1 minute and ramped at 15°C min-1 to a final temperature of 210°C.
Detection was achieved with a flame ionization detector held at 250°C with flow rates of 40 mL/
min hydrogen, 400 mL/min air, and 30 mL/min helium makeup gas. Acetate concentrations
were quantified by comparison of peak heights (with a retention time between 4.9 to 5.0
minutes) determined by Shimadzu GCsolution software to a sodium acetate standard curve.
3.2.12 Cell viability measurements from plate counts
Volumes of cell culture were serially diluted in phosphate buffered saline (PBS) (137 mM
NaCl, 27 mM KCl, 10 mM Na2HPO4, 2 mM KH2PO4, pH 7.4) and spread onto LB agar plates
(containing no antibiotics) at indicated times. Individual colonies were counted after overnight
73
incubation at 37°C and an additional overnight incubation at room temperature, due to the
wide range of colony sizes observed in BTE-expressing strains after overnight incubation.
3.2.13 Microscopy
Ten µL of culture was spotted directly onto glass slides beneath a glass coverslip, and cells
were imaged on an Olympus IX70 fluorescence microscope (Olympus America, Center Valley,
PA) with Hoffman modulation contrast (HMC 20) condenser and Olympus Plan 100x/1.25 oil
immersion objective. Images were acquired using a SPOT RT monochrome digital camera
(Diagnostic Instruments, Inc., Sterling Heights, MI) on Metavue version 7.1.0.0 software
(Molecular Devices).
3.2.14 SYTOX flow cytometry assays
To assess cell permeability, cell pellets collected by centrifugation of 0.5 to 1 mL culture
samples were resuspended in 1 mL of PBS, diluted 1000 to 2000-fold in 1 mL PBS, and stained
by addition of 1 µL of 5 mM SYTOX Green in DMSO (Invitrogen). Staining proceeded for 10
to 30 min prior to flow cytometric analysis using a Guava EasyCyte Plus flow cytometer
(Millipore, Billerica, MA) with 488 nm excitation and simultaneous measurement of forward
scatter and 525 nm (green) emission on logarithmic scale photodetectors. Forward scatter with
no minimum threshold was selected as the trigger for events and 5000 events were collected per
sample. Green fluorescence histograms were constructed by binning logarithmic scale green
fluorescence values between 0 and 1000 in increments of 10, and averaging the number of events
per bin between three biological replicates.
74
3.2.15 Membrane polarization assays
3,3'-Diethyloxacarbocyanine iodide (DiOC2) and carbonyl cyanide 3-chlorophenylhydrazone
(CCCP) (Sigma) stocks were prepared in dimethyl sulfoxide (DMSO) at concentrations of 3 mM
and 0.5 mM, respectively. This method was adapted from Novo et al. [1999]. Approximately
0.5-1 OD-mL of culture was centrifuged (16000 × g, 1 min) at specified collection times and
resuspended in 1 mL of NaPBS (10 mM sodium phosphate, 145 mM sodium chloride, pH 7.4)
containing 1 mM ethylenediaminetetraacetic acid (EDTA) diluted from an 0.5 M, pH 8.0 stock.
EDTA was added to assist in permeabilization of the outer membrane to allow DiOC2 uptake.
Resuspended cells were diluted 1:1000 in NaPBS + 1 mM EDTA, and 10 µL of 0.5 mM CCCP
was added to positive control samples to induce depolarization. To test samples, 10 µL DMSO
was added. Finally, 10 µL of 3 mM DiOC2 was added and samples were incubated at room
temperature for at least 30 minutes prior to flow cytometric analysis as described in section
3.3.14. Both green (525 nm) and red (680 nm) emission were simultaneously detected, and the
ratio of red to green logarithmic scale fluorescence values was calculated.
Red to green
fluorescence ratio histograms were constructed by binning fluorescence ratios between 0 to 2.5
in increments of 0.01, and averaging the number of events per bin between three biological
replicates.
3.2.16 Additional strain engineering
Strains containing marA::kan, rob::kan, and soxS::kan, and pspF::kan loci (Appendix I)
were obtained from the Keio collection [Baba 2006]. Gene deletions were transduced into strain
RL08 by P1 phage as previously described (Chapter 2) to generate strains RL09, RL10, RL11,
and RL12, respectively, which were verified by colony PCR using the primers 40-44, 47, and 48
75
(Appendix III) and GoTaq Green Master Mix (Promega, Madison, WI). The kan cassette was
removed from strain RL12 by transforming plasmid pCP20 [Cherepanov 1995], which encodes a
FLP recombinase. The strain was cured of pCP20 by repeated incubations at 43°C to generate
strain RL13.
The marA, rob, and soxS genes were amplified by PCR using E. coli K-12 MG1655 genomic
DNA as template (primers 45-50, Appendix III). The forward primers introduced an artificial
ribosome binding site (AGGAGG) and spacer sequence (ATTATAAA) prior to the start codon
of each gene. PCR products were cloned into pBAD18 [Guzman 1995] between the XbaI and
HindIII sites to generate plasmids pBAD18-MarA, pBAD18-Rob, and pBAD18-SoxS. The
pspABCDE operon and the putative native ribosome binding site upstream of pspA was
amplified by PCR (primers 51-52, Appendix III). The PCR product and cloned into pBAD18
between the SacI and XmaI sites to generate plasmid pBAD18-Psp.
To test the impact of these additional gene deletions or overexpressed genes on cell viability
and fatty acid production, pBAD33-BTE or pBAD33-BTE-H204A, and pBAD18, pBAD18MarA, pBAD18-Rob, pBAD18-SoxS, or pBAD18-Psp were transformed into strain RL08 or
strain RL08 with additional gene deletions by a chemical method [Chung 1989].
Fresh
transformants were grown in biological triplicate overnight in LB medium supplemented with
the appropriate antibiotics (34 µg/mL chloramphenicol and 50 µg/mL ampicillin). Cultures were
inoculated to an initial OD600 of 0.01 in 50 mL LB medium supplemented with 0.4% glycerol
and appropriate antibiotics in 250 mL shake flasks. Cultures were induced with 0.2% arabinose
at an OD600 of 0.2. After 8 h, cells were dilution plated as described above for measuring
CFU/mL. After 24 h, 200 µL of 1:10 diluted antifoam 204 (Sigma, St. Louis, MO) in ethanol
was added to each culture, and samples with significant foam (all BTE-expressing cultures) were
76
heated with gently swirling in an 85°C water bath for 5-10 min to assist with foam collapse.
Samples were taken in technical duplicate from each culture for fatty acid analysis as described
above. Technical duplicate fatty acid concentrations were averaged prior to analyzing the mean
and standard deviation of biological triplicate cultures.
3.3 Results
3.3.1 Cell viability and morphology
In prior studies, the OD600 of RL08 expressing BTE from higher copy number plasmids was
significantly reduced following induction relative to RL08 expressing non-functional BTEH204A from the same plasmids (Chapter 2).
Furthermore, an optimum of FFA titer was
achieved when BTE was expressed on low copy plasmids, strongly supporting the presence of a
toxic effect resulting from FFA overproduction that is independent of effects resulting from
protein overexpression (Chapter 2).
Additionally, induced shake flask cultures of
RL08/pTrc99A-BTE exhibit reductions in OD600 during stationary phase [Hoover 2011] that
may be indicative of cell lysis.
To test the hypothesis that FFA overproduction is toxic,
measurement of CFU/mL was performed at four sampling times (mid-log phase at OD600
between 0.8 and 1.0, transition period after log-phase growth has ceased but prior to plateauing
of OD600 at 4.5 h post-inoculation, early stationary phase at 5.8 h, and mid-stationary phase at 10
h) on LB agar plates for cultures of RL08/pTrc99A-BTE-H204A and RL08/pTrc99A-BTE
(Table 3.1). Cultures expressing BTE contained nearly 50% fewer CFU/mL in mid-log phase
and approximately 85% fewer CFU/mL in early stationary phase than cultures expressing BTEH204A at the same sampling times. In addition, the colonies grown from stationary phase
cultures demonstrated strain-dependent differences in size after 16 to 24 h incubation (data not
77
shown). Colonies from BTE-H204A-expressing cultures plated all sampling times and BTEexpressing cultures during mid-log phase exhibited homogeneous colony size distributions. In
contrast, colonies from BTE-expressing cultures plated during the transition between log phase
and stationary phase and throughout stationary phase displayed a wide distribution of colony
sizes, with many small colonies that were only visible after additional incubation. The number
of small colonies was largest from mid-stationary phase cultures, and lower for early stationary
phase and mid-log phase cultures.
These observations support our hypothesis that BTE
expression has a negative impact on cell physiology beginning prior to the onset of stationary
phase.
Table 3.1 Viable cell counts in BTE and BTE-H204A-expressing shake flask cultures. CFU/mL
measurements were taken from induced cultures of RL08/pTrc99A-BTE-H204A and RL08/pTrc99ABTE grown in shake flasks containing EZ rich defined medium supplemented with 0.2% glucose at
specified times.
Viable cell counts (CFU mL-1)
Sampling time
mid-log
transition
early stationary
mid-stationary
BTE-H204A
(2.44 ± 0.29) ×108
9
(2.44 ± 0.28) ×10
(3.6 ± 0.6 ) ×109
(5.3 ± 1.9 ) ×109
BTE
(9.9 ± 2.3) ×107
8
(4.1 ± 0.1) ×10
(5.3 ± 1.9) ×108
(9.7 ± 1.3) ×108
An alternative measure of bacterial cell viability can be obtained by staining cells with
SYTOX Green nucleic acid dye, which is impermeable to cells when the inner membrane is
intact. Staining of intact cells produces a weak green fluorescence associated with staining of the
cell surface [Roth 1997]. In contrast, stained cells with compromised inner membranes exhibit
an intense green fluorescence associated with nucleic acid binding by the dye [Roth 1997].
Samples of strain RL08/pTrc99A-BTE-H204A and RL08/pTrc99A-BTE
were taken
78
immediately prior to induction at OD600 0.2 (pre-induction), and after 4.3 h (transition), 6.3 h
(early stationary phase), and 8.3 h (mid-stationary phase) and stained with SYTOX Green to
obtain a time course of measurements.
Log-scale forward scatter and green fluorescence
measurements were collected by flow cytometry.
Forward scatter histograms (Figure 3.3)
exhibit nearly identical distributions between BTE-H204A and BTE-expressing cells before
induction, with larger forward scatter values typical of the relatively elongated cells of log phase
growth [Åkerlund 1995].
During the transition from log phase growth and continuing
throughout stationary phase, histograms for BTE-H204A-expressing cells shift toward smaller
forward scatter values. This is consistent with the contraction in size of wild type E. coli cells as
they deplete nutrients but continue dividing during early stationary phase [Åkerlund 1995]. In
contrast, BTE-expressing cells exhibit a broader distribution of forward scatter values, indicative
of increased heterogeneity in cell size. Phase contrast microscopy of cells sampled immediately
after the times described above (Figure 3.4) revealed similar morphologies between BTE-H204A
and BTE-expressing strains during mid-log phase, with relatively elongated cells typical of logphase growth [Åkerlund 1995]. During stationary phase, BTE-H204A-expressing cells contract
in size, which is typical of wild-type E. coli cells as they deplete nutrients but continue dividing
during early stationary phase [Åkerlund 1995]. BTE-expressing cells, however, appear to take
on a wide range of different cell lengths.
Beginning in mid-stationary phase, many cell
aggregates are observed. It should be stressed that the morphologies observed may be dependent
on the growth medium and antibiotics present, as we have previously microscopically observed
cells expressing BTE under PBAD control (in pBAD33) in LB medium supplemented with 0.4%
v/v glycerol and 34 µg/mL chloramphenicol to have less variation in cell length but more
variation in diameter. Aggregates that stain strongly with Nile red, a lipophilic dye, are observed
79
A
B
Figure 3.3 Flow cytometry analysis (green fluorescence and forward scatter histograms) of SYTOX
stained cells expressing BTE and BTE-H204A. (A) Green fluorescence (485 nm excitation, 525 nm
emission) histograms of SYTOX Green stained cells taken from cultures of RL08/pTrc99A-BTE (gray)
and RL08/pTrc99A-BTE-H204A (black) at sampling times described in the text. Cells expressing BTEH204A yield a homogeneous population exhibiting weak green fluorescence. In contrast, cells expressing
BTE exhibit a binary distribution with an increased population of cells exhibiting bright green
fluorescence beginning in early stationary phase. Increased fluorescence is indicative of compromised
inner membrane integrity. (B) Averaged log-scale forward scatter of cells taken from cultures of
RL08/pTrc99A-BTE (gray) and RL08/pTrc99A-BTE-H204A (black) grown in EZ rich defined medium
supplemented with 0.2% glucose at sampling times defined in the text. Control cells expressing BTEH204A demonstrated a shift from larger forward scatter values during log phase growth (OD600 0.2) to
much smaller forward scatter values during stationary phase. In contrast, cells expressing BTE exhibit a
broadened forward scatter distribution corresponding to a larger average but more heterogeneous range of
cell sizes.
by mid-stationary phase under all culturing conditions in BTE-expressing strains (data not
shown). These aggregates are likely composed of excreted fatty acids, cell debris, and entire
80
exponential phase
BTE-H204A
early stationary phase
BTE
BTE-H204A
BTE
Figure 3.4 Phase contrast micrographs of cells expressing BTE and BTE-H204A. During mid-log
phase (left), both BTE and BTE-H204A expressing cells are relatively elongated and appear similar in
both cultures. In early stationary phase (right), BTE-H204A-expressing cells have contracted in size and
BTE-expressing cells have a wide range of cell sizes.
cells.
Green fluorescence histograms of the BTE-expressing strain (Figure 3.3) reveal an increasing
population of cells having intense green fluorescence from mid-log phase into stationary phase,
indicating compromised inner membrane integrity. In contrast, the BTE-H204A-expressing
strain exhibits a single population of cells possessing weak green fluorescence.
To determine whether the observed phenotypes were specific to endogenous FFA
overproduction or exacerbated by the presence of ampicillin, cell viability and SYTOX staining
was performed on RL08 and RL08/pTrc99A exposed to exogenously added lauric acid (0.5 g/L)
at an OD600 mimicking the induction time of fatty acid overproducing strains in order to present a
stringent comparison to endogenous production. Little to no drop in viable cell counts were
observed at 4.5 h and 10 h post-inoculation (Table 3.2). The heterogeneous distribution of
colony sizes observed in BTE-expressing cells was absent in both RL08 and RL08/pTrc99A
exposed to exogenous FFAs (Figure 3.5). The colony size distribution of cells exposed to lauric
acid was also indistinguishable from control cells.
Furthermore, SYTOX Green staining
indicated a significantly lower proportion of cells exhibiting bright green fluorescence from
81
cultures with exogenously added lauric acid compared to cells that endogenously overproduce
FFAs (Figure 3.5).
Table 3.2 Cell viability analysis from exogenous addition of lauric acid. Viable cell counts (CFU
/mL) from cultures of RL08 and RL08/pTrc99A grown in shake flasks containing EZ rich defined
medium supplemented with 0.2% glucose at 4.5 hours (late exponential phase) and 10 hours growth (midstationary phase). At an OD600 of approximately 0.2, lauric acid was added to a final concentration of 0.5
g/L from an ethanol stock. An equivalent volume of ethanol (corresponding to 1% (v/v)) was added to
negative control cultures.
-1
Viable cell counts (CFU mL )
Strain
Condition
transition
RL08
RL08/pTrc99A
RL08
RL08/pTrc99A
Ethanol
Ethanol
Lauric acid
Lauric acid
(3.76 ± 0.96) ×10
(3.22 ± 0.08) ×109
(2.99 ± 0.09) ×109
9
(2.24 ± 0.31) ×10
mid-stationary
9
9
(5.23 ± 1.36) ×10
(6.03 ± 0.64) ×109
(5.6 ± 0.5 ) ×109
9
(4.2 ± 0.2 ) ×10
3.3.2 Cultivation of E. coli for functional genomics study
To determine a functional basis for the observed losses in viability and membrane integrity,
global mRNA and protein levels were measured from triplicate shake flask cultures of
RL08/pTrc99A-BTE-H204A and RL08/pTrc99A-BTE in a rich defined medium supplemented
with glucose. RNA and protein samples were isolated during mid-log phase (OD600 of 0.8, two
doublings after induction), during early stationary phase (defined as 1 h after the instantaneous
growth rate dropped to one-fourth of the log phase growth rate), and during mid-stationary phase
(defined as 4 h after the early stationary phase sampling point). In addition, culture samples
were taken periodically to monitor OD600, glucose levels, and fatty acid content from the same
cultures (Figure 3.6). As expected, expression of BTE resulted in production of elevated levels
of fatty acids, primarily with a carbon chain length between 8 and 14 (and predominantly 12
carbons) relative to cultures expressing BTE-H204A. In a second set of experiments, RL08
82
Figure 3.5 Flow cytometry analysis (forward scatter and green fluorescence) of SYTOX stained
cells exposed to exogenously added lauric acid. RL08 and RL08/pTrc99A were exposed to either 0.5
g/L lauric acid (from a stock solution in ethanol), or an equivalent volume of ethanol (corresponding to
1%) at an OD600 of approximately 0.2. (a) averaged forward scatter histograms at 4.5 h (transition
between log and stationary phase). (b) averaged forward scatter histograms at 10 h (mid-stationary phase).
(c) averaged green fluorescence histograms at 4.5 h. (d) averaged green fluorescence histograms at 10 h.
RL08 and RL08/pTrc99A exhibited similar forward scatter distributions indicative of similar cell size
distributions with and without exogenous addition of lauric acid. Only a small percentage of RL08 and
RL08/pTrc99A exposed to lauric acid at both sampling times exhibited bright green fluorescence
indicative of compromised inner membrane integrity by SYTOX Green staining.
harboring a combination of plasmids expressing either BTE or BTE-H204A (in pBAD35), and
the four subunits of ACC or an empty vector (pBAD33), were cultivated in fermentors in a rich
defined medium supplemented with glycerol. Samples for mRNA and protein quantification
were isolated during late log phase (OD600 of 2.0), early stationary phase (defined as for the
shake flask cultures, mRNA only), and mid-stationary phase (12 h).
Growth curves and
sampling times are depicted in Figure 3.7. The shake flask cultures and fermentor cultures were
grown in the presence of different antibiotics (ampicillin and chloramphenicol/kanamycin),
carbon sources (glucose and glycerol), and utilize different inducing agents (IPTG and
83
arabinose). Genes and proteins that were differentially expressed in both the shake flask and
fermentor experiments would be more specifically representative of the perturbations resulting
from fatty acid overproduction.
Figure 3.6 Growth curves, sampling points, and glucose and fatty acid profiles of shake flask
cultures for the functional genomics study. (a) Optical density (OD600) of RL08/pTrc99A-BTE-H204A
(solid diamonds) and RL08/pTrc99A-BTE (empty diamonds) expressing cultures grown at 37°C in EZ
rich defined medium supplemented with 0.2% glucose in shake flasks. A decrease in OD600 is observed
beginning in early stationary phase for BTE-expressing cultures relative to BTE-H204A-expressing
cultures. Glucose utilization is also indicated for BTE-H204A (solid squares) and BTE (empty squares),
and proceeds at the same rate in all cultures. Sampling times for RNA and protein are indicated with
dashed vertical lines. (b) Medium (8 to 14 carbon) chain length fatty acid titers in BTE-H204A (solid
circles) and BTE (empty circles) expressing cultures.
3.3.3 FFA overproduction increases the long-chain unsaturated fatty acid content
A GC/MS analysis of the fatty acid composition of FFA overproducing and control strains
was conducted. In both functional genomics experiments, the percentage of unsaturated long-
84
Figure 3.7 Growth curves and sampling times of fermentor cultures for the functional genomics
study. Individual growth curves of strain RL08 harboring pBAD33-ACC and pBAD35-BTE (ACC+
BTE+), pBAD33 and pBAD33-BTE-H204A (ACC- BTE-), and pBAD33 and pBAD35-BTE (ACCBTE+) grown in fermentors in EZ rich medium supplemented with 0.4% glycerol as described in
Materials and Methods. Each strain was grown in biological duplicate and is indicated above by a
separate curve. The sampling times for RNA and protein are indicated by the enlarged and circled points
(late log phase, early stationary phase, and mid-stationary phase). Proteins were only measured at OD600
2.0 and at mid-stationary phase. The increase in OD600 observed between 10 and 15 hours for one
replicate of the ACC- BTE- strain (black circles) corresponded to a pressure increase due to a clogged
outlet filter.
chain (16 to 18 carbon chain length) fatty acids was significantly increased in FFA after 24 h
(Figure 3.8). This increase in unsaturated fatty acid content was due almost entirely to a
decrease in palmitic acid (C16:0) content (Fig. 3.8) in the BTE-expressing strain rather than from
an increase in palmitoleic (C16:1) and vaccenic acid (C18:1) content. A significant decrease in
stearic acid (C18:0) content from approximately 2.0% to 1.5% is also detectable. A similar trend
was previously observed when BTE was expressed in E. coli K27, a fadD mutant strain, with
large increases in C16:1 and C18:1 observed in extracted phospholipids [Voelker 1994]. It has
been postulated that this shift in phospholipid composition is due to the depletion of the pool of
saturated acyl-ACPs by BTE [Voelker 1994]. BTE-expressing cultures grown in fermentors
85
exhibited similar increases in unsaturated 16 to 18 carbon chain length fatty acids (48% versus
22% in BTE-H204A expressing cultures at 20 h, Figure 3.9).
A
B
BTE-H204A
% unsaturated C16-C18
60
BTE
50
16:0
18:0
40
30
BTE-H204A
BTE
16:1
20
0
5
10
15
time (h)
20
25
18:1
Figure 3.8 Analysis of C16-C18 fatty acids extracted from shake flask cultures. (A) Percentage of
unsaturated C16-C18 fatty acids out of total C16-C18 fatty acids (note that C17∆ was not analyzed in these
samples). (B) Titers of individual long-chain fatty acids in BTE-H204A (solid) and BTE (empty)
expressing cultures as a function of growth time, including, from top to bottom, palmitic acid (C16:0),
stearic acid (C18:0), palmitoleic acid (C16:1), and vaccenic acid (C18:1). Decreases in concentration of
the saturated species (16:0 and 18:0) was observed, with little change in the concentration of unsaturated
species (16:1 and 18:1).
3.3.4 DNA microarray analysis
Transcript profiles of BTE-H204A and BTE-expressing strains were compared to identify
stress responses and transcriptional cascades linked to FFA overproduction. During mid-log
86
A
B
Figure 3.9 C8-C14 fatty acid titers and percent unsaturated C16-C18 fatty acids from fermentor
cultures. (A) Combined titer of medium-chain fatty acids (C8:0, C9:0, C10:0, C10:1, C11:0, C12:0,
C12:1, C14:0, C14:1) over the course of the bioreactor runs. Note that fatty acid titer increases around the
time of transition to stationary phase. The drop in fatty acid titer observed after 15 hrs is due to culture
foaming (which was limited by addition of antifoam 204) which deposits fatty acids above the liquid
height (sampling point) of the vessel. (B) Percentage of unsaturated fatty acids (C16:1 and C18:1)
relative to the total C16 and C18 fatty acid pool. These fatty acids are the common constituents of lipids
found in the inner membrane. Note the difference in stationary phase (after 10 hours) between the BTE+
(white and grey symbols) and BTE- (black symbols) cultures.
phase, only a small set of genes were identified as being significantly differentially regulated
(fold change ≥ |2| and P ≤ 0.05) in BTE-expressing shake flask cultures relative to the
corresponding BTE-H204A control. Out of the 4240 genes analyzed on DNA microarrays, 20
genes displayed increased expression levels and 3 genes displayed decreased expression levels
87
compared to the control strain. More dramatic changes were observed during stationary phase,
with 283 genes having increased expression levels and 350 genes having decreased expression
levels in early stationary phase. In mid-stationary phase, 477 genes increased and 564 genes
decreased in expression. The extent of changes in mean gene expression values for each strain at
the three sampling times is presented as a hierarchical clustering analysis (Figure 3.10).
Additional transcript profiles were obtained from strain RL08 grown in fermentors. Fewer
statistically significant, differentially expressed genes were identified in the fermentor cultures
Figure 3.10 Heat map of gene expression values at each shake flask culture sampling time with
hierarchical clustering analysis. Gene expression values were measured at three sampling times (midlog phase; early stat, early stationary phase; mid stat, mid-stationary phase) for strain RL08/pTrc99ABTE-H204A and RL08/pTrc99a-BTE grown in EZ rich defined medium supplemented with 0.2%
glucose in shake flasks. Hierarchical clustering was performed in ArrayStar software, with the closest
pair clusters of gene expression profiles for each strain and sampling time indicated at top. Mid-log phase
and early stationary phase profiles between the two strains were the most similar to each other. The midstationary phase profile for BTE was more similar to the early stationary phase profiles of both strains
than to the mid-stationary phase profile of BTE-H204A. All stationary phase profiles were more similar
to each other than to the mid-log phase profiles.
88
which were grown in biological duplicate. A cross-comparison of significant differentially
expressed genes present in both the shake flask and fermentor cultures across all sampling times
(defined here as a fold change ≥ |1.8| and P ≤ 0.05 for shake flask cultures, P ≤ 0.10 for
fermentor cultures) identified a key set of 150 up-regulated genes and 112 down-regulated genes
(Appendix IV). To validate results from the microarray data, three genes were selected for
qPCR analysis of cDNA derived from RNA extracted from independent cultures of
RL08/pTrc99A-BTE or
RL08/pTrc99A-BTE-H204A
grown
in rich defined medium
supplemented with glucose. Fold-changes in expression calculated from qPCR cycle threshold
values showed similar trends to microarray fold-changes (Table 3.3).
Table 3.3 qPCR measurement of fold-changes in expression of selected genes in shake flask
cultures. Fold-changes are for BTE-expressing strains versus BTE-H204A expressing strains. Sampling
times are as defined in Materials and Methods. Standard errors are provided between biological
triplicates. Values in parentheses below each qPCR fold-change correspond to the measured fold-changes
observed in the equivalent DNA microarray experiment, with the P-value provided in italics.
locus
gene name
b0954
fabA
b1307
pspD
b2237
inaA
mid-log
-1.12 ± 0.58
(-1.28, 0.087)
1.14 ± 0.62
(-1.03, 0.82)
1.08 ± 0.71
(1.58, 0.001)
early stationary
mid-stationary
-4.05 ± 1.43
(-2.54, 0.004)
38.8 ± 11.9
(4.99, 0.002)
1.80 ± 0.61
(7.72, 0.000)
-8.15 ± 2.92
(-4.80, 0.000)
57.0 ± 24.5
(9.60, 0.002)
2.76 ± 0.89
(5.72, 0.000)
3.3.5 Differential proteomics analysis
In shake flask experiments, a total of 2214 proteins were detected in at least one technical
replicate from at least one sampling time (mid-log phase, early stationary phase, or midstationary phase) from at least one peptide, and of these 1424 proteins were identified from at
least two peptides. In fermentor experiments for three strains (BTE-H204A-expressing, BTEexpressing, and BTE-expressing and ACC-overexpressing), a total of 3270 proteins were
89
detected in at least one technical replicate from at least one sampling time (late log phase,
mid-stationary phase) from at least one peptide, and of these 2498 proteins were identified from
at least two peptides. Filtering, averaging, and statistical tests were applied between technical
and biological replicates as described in Materials and Methods. This analysis resulted in a total
of 42 proteins (37 from more than one peptide) in mid-log phase, 127 proteins (122 from more
than one peptide) in early stationary phase, and 110 proteins (104 from more than one peptide) in
mid-stationary phase shake flask cultures identified as differentially expressed with P ≤ 0.05,
with a much more select number of proteins exhibiting fold-changes greater than or equal to 2
with P ≤ 0.05 (Table 3.7 at the end of this chapter). Of the proteins with P ≤ 0.05 and a greater
than 2-fold change in expression, 44% were identified as being differentially expressed in the
same direction at the transcript level from at least one sampling time. Comparing differential
expression between the BTE-expressing strain and the BTE-H204A-expressing strains grown in
fermentors, 80 proteins (79 from more than one peptide) in late log phase, and 99 proteins (98
from more than one peptide) in mid-stationary phase were differentially regulated with P ≤ 0.05,
with only a handful of proteins exhibiting both P≤ 0.05 and a greater than 2-fold change in
expression (Table 3.8 at the end of this chapter). In each proteomics data set, BTE and BTEH204A were identified in the proteomics data sets from dozens of individual peptides, with a
single peptide containing either the original (WNDLDVNQHVNNLK) or mutated histidine
residue.
As expected, the histidine-containing peptide was present in all BTE-expressing
samples, and the alanine-containing peptide was present in all BTE-H204A-expressing samples.
Because of the low genome coverage and relatively small number of proteins identified as
significantly differentially expressed, the proteomics analysis was used primarily for validation
of trends in the microarray data sets in this study, and a complete analysis was not performed.
90
3.3.6 Induction of phage shock proteins and loss of proton motive force
Among the most significantly up-regulated genes beginning in stationary phase in both the
fermentor and shake flask cultures were those encoding the phage shock proteins (Table 3.7 at
the end of this chapter). PspA was also identified in proteomics data as being significantly
differentially expressed in mid-stationary phase shake flask cultures expressing BTE, and in
BTE-expressing, ACC-overexpressing cultures grown in fermentors (Table 3.4). Phage shock
proteins are encoded in two operons (pspABCDE and pspG) and are induced under a variety of
conditions that induce cell envelope stress and affect energy generation [Joly 2010], including
treatment with various organic solvents such as hexane [Kobayashi 1998] and ethanol [Brissette
1990], and following exposure to octanoic acid [Weiner 1994] but not isobutanol [Brynildsen
2009]. Transcription from the promoters of both operons is σ54-dependent and activated by the
transcriptional dual regulator PspF [Joly 2010]. PspA is believed to play a physical role in
stabilizing membranes [Engl 2009, Standar 2008] and has been shown to suppress proton
leakage of damaged liposomes composed of E. coli phospholipids [Kobayashi 2007].
To determine whether BTE-expressing cells were actually becoming depolarized as a result
of loss of proton motive force, a ratiometric flow cytometry assay using 2,2-diethyloxacarbocyanine iodide (DiOC2) [Novo 1999] was performed at various culture sampling times for RL08
harboring pTrc99A-BTE-H204A and pTrc99A-BTE cultivated as previously described.
A
difficulty in using a ratio of red to green fluorescence emission values is that the ratio values
appeared to depend on cell size, possibly because red fluorescence emits from the full volume
ofthe cell while green fluorescence emits from only the cell membrane. At each sampling time,
positive control samples of both cultures were run which were treated with CCCP, a proton
ionophore, to induce depolarization and determine the location of the red/green fluorescence
91
Table 3.4 Selected protein fold-changes between BTE and BTE-H204A-expressing strains. Cells
were cultured in EZ Rich Defined Medium supplemented with 0.2% glucose in shake flasks, and in EZ
Rich Defined Medium supplemented with 0.4% glycerol in fermentors at sampling times specified in the
text. P-values for all fold-changes shown were less than 0.05. No significant changes in expression
levels were measured at the mid-log phase sampling point for EZ glucose cultures.
EZ glucose
locus
protein
name
b1014
b1304
b1493
b1611
b1743
b2323
b3506
PutA
PspA
GadB
FumC
Spy
FabB
Slp
EZ glycerol
BTE+ ACC-
early stationary
-4.20
-
mid-stationary
2.10
5.52
4.63
3.53
5.24
-3.46
2.10
late log
2.20
-2.36
-
BTE+ ACC+
mid-stationary
-5.85
-2.11
late log
-2.46
-
mid-stationary
4.32
2.28
-7.00
-
ratio peak for this state. While BTE-H204A-expressing cultures exhibited a larger depolarized
percentage of the cell population during exponential growth, BTE-expressing cultures did exhibit
a larger population of depolarized cells during two early stationary phase sampling points
(approximately 6 and 8 hours post-inoculation) (Figure 3.11). This population of depolarized
cells may be responsible for the greatly increased expression of genes encoding phage shock
proteins in early and mid-stationary phase in the cultures sampled for microarray analysis.
3.3.7 Induction of MarA/Rob/SoxS regulon
Under all tested conditions, BTE-expressing strains exhibited increased expression of many
genes activated by the MarA/Rob/SoxS transcription factors (Table 3.7). MarA, Rob, and SoxS
bind to a similar consensus sequence of DNA upstream of the promoters of regulon members in
a degenerate manner, with one, two, or all three of the regulators activating transcription from a
given promoter [Martin 1999]. The activity of the transcription factors is modulated by various
mechanisms dependent on environmental conditions. In the presence of salicylate and other
92
CCCPtreated
BTE
BTEH204A
Figure 3.11 Cell polarization flow cytometry assay of FFA-overproducing and control cultures.
Shake flask cultures of RL08 harboring pTrc99A-BTE-H204A or pTrc99A-BTE were grown in EZ Rich
Defined Medium supplemented with 0.2% glucose and induced as previously described. Culture samples
were harvested approximately 7.5 hours post-inoculation (early stationary phase) and treated as described
in 3.3.15. Histograms were generated depicting depolarized (low red/green ratio) and polarized (high
red/green ratio) populations. BTE-expressing cultures exhibited a small population of depolarized cells
(colored in red), whereas BTE-H204A-expressing cultures appeared to be a uniform population of
polarized cells (dark blue).
aromatic compounds, MarA activation occurs via inactivation of the DNA-binding repressor
MarR [Alekshun 1999].
Similarly, SoxS activation occurs via inactivation of the cognate
repressor SoxR in the presence of superoxide [Demple 1996]. Rob has no known cognate
repressor, and its activity is directly modulated by the presence of decanoate, bile salts, and
dipyridyl [Rosner 2002, Rosenberg 2003]. As early as during mid-log phase growth in BTEexpressing shake flask cultures and at the late log phase sampling point in fermentors,
upregulation of marR was observed. Early stationary phase, and further increased mid-stationary
phase induction of many members of the regulon was observed in both shake flasks and
fermentors, including fumC (encoding an aerobically active fumarase), marA, pqiA (encoding a
paraquat-inducible protein with no known function), inaA (encoding a protein with unknown
function), marR, ybjC (encoding a protein with unknown function), putA (encoding a FAD+-
93
dependent proline dehydrogenase), nfsA (encoding an NADPH-dependent nitroreductase), and
nfo (encoding a DNA endonuclease involved in repairing oxidative damage). Expression of
mdtG, a recently ascribed member of the MarA/Rob/SoxS regulon which putatively encodes an
efflux pump, and for which overexpression increases resistance to fosfomycin and deoxycholate
[Fàbrega 2010] was also increased. MarA, Rob, and SoxS additionally activate transcription of
MicF RNA, which acts as an antisense repressor of ompF (encoding an outer membrane porin)
transcription [Andersen 1989].
Strongly decreased expression of ompF was observed in
stationary phase in both shake flasks and fermentors (Table 3.7). PutA was identified as having
increased mid-stationary phase expression in BTE-expressing shake flask cultures, in
exponential phase fermentor cultures expressing only BTE, and in mid-stationary phase
fermentor cultures expressing BTE and overexpressing ACC (Table 3.4).
FumC was also
observed to have increased expression in shake flask proteomics data in BTE-expressing cultures
during mid-stationary phase (Table 3.4).
Notably, no significant increase in expression of acrAB was observed in BTE-expressing
cultures (Table 3.7), which encodes an inner membrane multidrug efflux pump (AcrB) and a
periplasmic linker protein (AcrA) to an outer membrane channel (TolC). It has been previously
shown that acrAB transcription can be induced by high concentrations (5 mM) of exogenously
supplied decanoate via activation of Rob [Ma 1995, Rosenberg 2003]. Decanoate is similar to
the predominant overproduced fatty acid, laurate, however the maximum titer of endogenously
produced free fatty acids (chain length C12 to C14) in this study was approximately 1.5 mM,
which may be too low to observe significant induction of acrAB. Transcription of the acrAB
operon is also modulated by two divergently transcribed repressors, AcrR and MprA, which
appear to serve as a secondary attenuators of acrAB expression [Grkovic 2002]. Expression of
94
acrR was moderately increased in BTE-expressing cultures in early stationary phase in shake
flasks and in mid-stationary phase in fermentors (Table 3.7). Expression of mprA is moderately
increased in early stationary phase and strongly increased in mid-stationary phase in both shake
flasks and fermentors (Table 3.7). It has been previously noted that transcription of acrR is
increased approximately 4-fold more than transcription of acrAB under general stress conditions,
including exposure to 4% ethanol, 0.5 M NaCl, and during stationary phase [Ma 1996], which is
consistent with the pattern of expression observed in this study.
3.3.8 Induction of other envelope stress responses and acid resistance response
in batch cultures
Other regulons implicated in cell envelope stress include those of the alternative sigma factor
σE, and the two-component systems CpxAR and BaeSR [Rowley 2006]. Several genes in the σE
regulon, including opgG, opgH, plsB, and lpxP, were strongly up-regulated in early to midstationary phase in BTE-expressing shake flask cultures but not in fermentors (see GEO Series
GSE29424 or Supplementary Material in [Lennen 2011]). Significantly increased expression of
members of the CpxR regulon was also observed in mid-stationary phase in shake flask cultures
but not fermentors, including cpxP, degP, mdtD, and spy (mdtD and spy are also in the BaeR
regulon). Spy was also identified as being significantly differentially expressed in the shake
flask proteomics data (Table 3.4), but the other proteins were either non-significantly changed or
not detected. Other members of the σE and CpxR regulon were significantly down-regulated in
mid-stationary phase in BTE-expressing shake flask cultures, including ompF, csgDEFG, and
hlpA-lpxD-fabZ-lpxA, an operon transcribed from a promoter upstream of hlpA.
95
A number of genes on a region of the genome referred to as the acid fitness island
[Hommais 2004], primarily involving glutamate-dependent acid resistance, were significantly
increased in expression in stationary phase shake flask cultures. These genes include slp and
hdeAB-yhiD, which are repressed by MarA during the transition to stationary phase [Price 2000,
Ruiz 2008], but are also affected by numerous other regulators. Increased expression of these
genes was also accompanied by dramatic increases in expression of gadA, gadB, gadC, gadE,
and mdtEF, with more moderate increases in expression (< 5-fold) of gadX and gadW
(Supplementary Material), strongly suggesting the relative lack of MarA involvement. Slp and
GadB were also identified as having increased expression in shake flask proteomic data (Table
3.4). There was no activation of genes on the acid fitness island in fermentor cultures, but rather
a reduction in expression of hdeAB in the BTE-expressing strain in mid-stationary phase (see
GEO Series GSE29424 or Supplementary Material in [Lennen 2011]).
3.3.9 Changes in energy metabolism
Genes involved in energy metabolism were up-regulated in BTE-expressing strains in both
shake flask and fermentor cultures (Table 3.7). These include many members of the nuo operon
(nuoH, nuoI, nuoJ, nuoN) encoding subunits of NADH:ubiquinone oxidoreductase I, and the cyo
operon (cyoC, cyoD, cyoE) encoding subunits of the cytochrome bo terminal oxidase complex
and heme O synthase. Other members of the operon were up-regulated in the mid-stationary
phase shake flask data set with either P-values or fold-changes not meeting the selected
significance criteria or not also being identified with the significance criteria in the fermentor
data set, including nuoG, nuoL, nuoK, cyoA, and cyoB. Similarly for fermentor data, additional
genes identified in late log-phase include nuoM, nuoK, nuoG, nuoL, cyoA, and cyoB. These gene
96
products form core components of the electron transport chain, with complex regulation of
transcription by a number of different transcription factors involved in sensing oxygen levels
(ArcA, FNR, Fur) and nitrate levels (NarL) among others. Furthermore, expression of the nuo
operon is also dependent on the proportion of NADH present relative to total reducing
equivalents, as it is activated by growth on C4 dicarboxylates, which have a reduced proportion
of NADH [Bongaerts 1995]. Upregulation of the nuo and cyo operons has also been observed on
n-butanol exposure, which was suggested to either be indicative of an increased energy
requirement or impairment of respiration [Rutherford 2010]. While an increased population of
depolarized cells was observed in BTE-expressing cultures (Figure 3.11), it is not clear how this
event causes elevated transcription of the nuo and cyo operons.
Additional genes involved in cellular respiration were also identified as differentially
regulated in BTE-expressing strains in both data sets (Table 3.7), including increased stationary
phase expression of fdoG and fdoI encoding subunits of the aerobic formate dehydrogenase-O
(fdoH was also increased but with low statistical confidence in fermentors), ubiC and ubiA
involved in ubiquinone biosynthesis, hybA encoding a subunit of hydrogenase-2 (other subunits
also exhibited increased differential expression but with lower statistical confidence), frdA
encoding a subunit of fumarate reductase, and yfeH encoding an AI-2 induced putative
cytochrome oxidase. Highly increased expression of hyaABCDEF, encoding hydrogenase-1 and
associated proteins, was observed only in shake flask cultures. Decreased expression of appCBA,
encoding cytochrome bd-II terminal oxidase and an acid phosphatase, was observed transiently
in shake flasks in early stationary phase, and in mid-stationary phase in fermentors. Strongly
reduced expression of yqfA, encoding a subunit of a predicted oxidoreductase, and reduced
expression of yodB encoding a predicted cytochrome, was also transient in shake flasks in early
97
stationary phase, and during late log phase to early stationary phase in fermentors. The
soluble transhydrogenase encoded by sthA, whose primary activity appears to be regeneration of
NADH from NADPH and for which transcription is increased under conditions known to
increase NADPH availability and decreased during growth on glycerol [Sauer 2004], was nearly
5-fold more abundant in BTE-expressing mid-stationary phase shake flask cultures, and nondifferentially regulated in fermentor cultures.
3.3.10 Decreased expression of unsaturated fatty acid biosynthesis pathway
In both functional genomics experiments, the percentage of unsaturated long-chain (16 to 18
carbon) fatty acids was significantly increased due to a decrease in C16:0 levels in FFA
overproducing strains. Fatty acid unsaturation in E. coli is introduced by 3-hydroxydecanoylACP dehydrase (FabA) by dehydration of the hydroxylated C10 acyl-ACP and isomerization to
both cis-3- and trans-2-enoyl-ACP [Cronan 2008]. The cis species are not reduced by FabI and
are elongated as unsaturated acyl-ACPs. Additionally, an alternative β-ketoacyl-ACP synthase,
FabB, is required for elongation of cis-3-decenoyl-ACP [Feng 2009]. Transcription of fabA and
fabB are activated by FadR and repressed by FabR, which bind within the promoter region
(Figure 3.12). FadR activation is abolished upon its binding to acyl-CoAs [DiRusso 1992],
whereas FabR repression is abolished upon its binding to either unsaturated acyl-CoAs or
unsaturated acyl-ACPs [Zhu 2009]. As such, FabR is considered a controller of the level of
membrane lipid unsaturation. In BTE-expressing strains in both the shake flask and fermentor
data sets, fabB was strongly down-regulated in early stationary phase, with decreased expression
evident as early as the mid-log phase and late log phase in shake flask and fermentor cultures,
respectively (Table 3.7). FabB was also identified as being significantly decreased in abundance
98
in BTE-expressing cultures under all conditions (Table 3.4). Additionally, fabA was reduced
in expression but to a lesser extent than fabB. Deletion studies of fadR and fabR and in vitro
studies have identified FadR as having a greater influence on fabA expression, and FabR as
having a greater influence on fabB expression [Feng 2011]. Therefore it is suggestive that FabR
is playing a role in controlling fabB and fabA transcription in BTE-expressing strains by sensing
a high ratio of unsaturated to saturated acyl-ACPs, and that this high ratio results from depletion
of saturated acyl-ACPs by BTE. Despite this high level of transcriptional regulation, cultures
appear to be unable to produce sufficient palmitoyl-ACP to maintain a typical degree of longchain fatty acid saturation.
malonyl-ACP
FabR
O
O
S ACP
CH3(CH2)x
S ACP
CH3(CH2)x
O
FabB or FabF
β-ketoacyl-ACP
acyl-ACP
O
FabR
S ACP
FabI
FabG
cis-3-decenoyl-ACP
CH3(CH2)5
FabA
O
CH3(CH2)x
OH
FabZ
S ACP
trans-2-enoyl-ACP
CH3(CH2)x
O
S ACP
β-hydroxyacyl-ACP
Figure 3.12 Schematic of regulation of unsaturated biosynthesis by FabR. The transcription factor
FabR transcriptionally represses transcription of fabA and fabB. FabR binds DNA more strongly when
bound to unsaturated acyl-ACPs than when bound to saturated acyl-ACPs.
99
3.3.11 Increased expression of β-oxidation genes
Increased expression of several genes in the β-oxidation pathway was observed primarily in
mid-stationary phase in BTE-expressing cultures grown in both shake flasks and fermentors.
Genes increased in expression under both conditions include fadA, fadE, fadH, fadJ, and fadI
(Table 3.7). In the shake flask proteomics data set, FadE, FadA, and FadL were identified as
having significantly increased expression during mid-stationary phase (Table 3.8). These genes
are all repressed by FadR, which binds to DNA in the promoter region. Acyl-CoAs bind to FadR
and cause a conformational shift that releases FadR from its DNA binding site, thereby
activating expression of the β-oxidation pathway [DiRusso 1992, Cronan 1997]. In a prior study
investigating the effect of exogenous addition of oleic acid to protein expression in E. coli K-12
W3110, increased expression of FadA, FadB, and FadL was observed [Han 2008], consistent
with fatty acid activation to acyl-CoAs and subsequent de-repression of FadR-regulated genes.
However, the relief of repression of the β-oxidation genes in strain RL08 is curious, as it
possesses a deletion in the gene encoding the aerobically active acyl-CoA synthetase, fadD. E.
coli also possesses an anaerobically active short-chain acyl-CoA synthetase, FadK, which has the
highest in vitro activity toward 6 and 8 carbon chain length fatty acids [Morgan-Kiss 2004]. No
significant change in expression of either fadK transcript or FadK protein (which was only
detected from a single peptide in only one instrumental replicate) was observed in BTEexpressing cultures at any time point in any of the microarray data sets, consistent with prior
observations that fadK is not expressed at either the transcript or protein level under aerobic
conditions [Campbell 2003, Morgan-Kiss 2004].
Endogenously produced free fatty acids
produced by overexpressing E. coli cytosolic TesA in a fadD mutant strain, which exhibited no
detectable in vitro acyl-CoA synthetase activity, produced no detectable β-galactosidase activity
100
in a fadBA::lacZ chromosomal insertion [Cronan 1997].
This was suggestive that
endogenously produced free fatty acids cannot induce FadR derepression, however other studies
have indicated both weak in vitro activation of FadR by FFAs [DiRusso 1992] and in vivo
activation of an engineered regulatory system utilizing FadR where TesA' is overexpressed
[Zhang 2012]. It is hypothesized that FFAs are directly inducing FadR derepression of the fad
regulon in this work, and it should be noted that the FadR binding affinities of all FFAs produced
by BTE have not been characterized. It is also possible that an as-yet unidentified protein
exhibits a previously undetected acyl-CoA synthetase activity. For example, while unlikely to be
strictly an acyl-CoA synthetase, YccU encodes a predicted CoA and NAD(P)-binding protein
that appears on exogeneous addition of oleic acid [Han 2008], and was also identified as being
upregulated in mid-stationary phase in BTE-expressing strains in mid-stationary phase in both
shake flask and fermentor cultures (+2.2-fold, P = 0.026; +2.0-fold, P = 0.096).
Additional evidence for the involvement of FadR can be derived from the gene expression
profile of another member of the FadR regulon, iclR, encoding isocitrate lyase regulator. The βoxidation pathway (which generates acetyl-CoA) is linked to acetate metabolism via activation
of iclR expression by FadR in the absence of acyl-CoA. IclR in turn represses expression of the
aceBAK operon, encoding genes involved in the glyoxylate bypass [Cozzone 2005]. In BTEexpressing shake flask cultures, expression of iclR is down-regulated in early to mid-stationary
phase (-2.0-fold, P = 7.3 × 10-4; -2.0-fold, P = 0.058), consistent with antagonization of FadR
binding to DNA. However expression of the aceBAK operon is not increased in BTE-expressing
cultures as might be expected. Rather, a significant decrease in expression of aceB is observed
in early to mid-stationary phase (-2.2-fold, P = 0.0024; -2.5-fold, P = 3.1 × 10-4). IclR activity is
modulated by other metabolic intermediates such as pyruvate and glyoxylate [Lorca 2007] and
101
other transcriptional regulators are implicated in aceBAK expression [Gama-Castro 2011],
either of which may account for the observed pattern of expression.
3.3.12 Differences in the acetate profile of fatty acid overproducing cultures
In BTE-expressing cultures, acetate levels were found to be reduced in cell-free supernatants
beginning before the onset of stationary phase and continuing for approximately 5 hours during
stationary phase until all acetate was re-assimilated (Figure 3.13). Similar reduced levels of
acetate have recently been observed in other FFA-overproducing strains [Li 2012]. Acetogenesis
occurs during anaerobic growth on glucose to balance reducing equivalents produced during
glycolysis, and aerobically during growth on excess glucose, possibly as a means to either
balance carbon flux or increase CoA availability in the TCA cycle during rapid growth [Wolfe
2005]. While the precise mechanism by which acetate levels are reduced in BTE-expressing
cultures is not yet understood, some hypotheses can be made based on differential transcript and
25
BTEBTE+
acetate (mM)
20
15
10
5
0
0
2
4
6
8
time (h)
10
12
14
16
Figure 3.13 Extracellular acetate concentrations from EZ glucose shake flask cultures expressing
BTE or BTE-H204A. A decreased level of log-phase excretion and/or increased stationary phase uptake
of acetate was observed in BTE-expressing cultures.
102
protein abundances. Notably, the levels of acetyl-CoA synthetase (Acs) are significantly
increased in early to mid-stationary phase in BTE-expressing shake flask cultures (Table 3.8),
despite non-significant regulation at the transcript level except during mid-stationary phase,
when a decrease in acs expression is observed (-1.9-fold, P = 0.017). Overexpression of Acs
results in more efficient acetate assimilation [Lin 2006] and increased production of malonylCoA [Zha 2009].
Induction of acs, as measured by β-galactosidase or chloramphenicol
acetyltransferase transcriptional fusions, increases upon depletion of glucose due to repression of
acs transcription by catabolite repression protein (CRP-cAMP) [Browning 2004].
This
explanation does not appear to explain the current results, as glucose levels appear to be depleted
synchronously in the BTE-H204A- and BTE- expressing cultures (Figure 3.6). Inactivation of
iclR and fadR, which cause constitutive activation of the glyoxylate shunt, have also been shown
to reduce expression of acs [Shin 1997, Kumari 2000] and thereby to reduce the rate of acetate
assimilation. This is also an inconsistent explanation with the current results, as iclR levels are
reduced in BTE-expressing cultures while acetate levels are also reduced relative to BTEH204A-expressing cultures.
One additional observation is that the transcript levels of many genes whose enzyme products
either utilize acetyl-CoA as a substrate, or generate acetyl-CoA as a product, are differentially
regulated in early to mid-stationary phase in BTE-expressing cultures. These include pta, atoDA,
cysE, accBC, glcB, aceB, prpC, leuA, adhE, atoB, aceF, fadA, fadI, pflB, rimI, speG, lacA, maa,
nhoA, rimL, tmcA, and ydbK (see GEO Series GSE29424 or Supplementary Material in [Lennen
2011]) representing the majority of genes whose encoded products are involved in acetyl-CoA
metabolism as annotated in the EcoCyc database [Keseler 2011]. Other acetyl-CoA utilizing or
generating enzymes were identified as being differentially increased at the protein level,
103
including GltA, AdhE, GlcB, and Acs as described above (Table 3.8). While not further
explored here, this may be indicative of an uncharacterized global transcriptional response of
genes encoding acetylated proteins to the level of acetyl-CoA flux.
3.3.13 Increased expression of the AccB and AccC subunits of acetyl-CoA
carboxylase
As expected, in BTE-expressing, ACC-overexpressing cultures in fermentors, the four genes
encoding ACC were all highly increased in expression, with proteomics data also indicating
statistically significant increases for the four subunits during exponential phase and midstationary phase (Table 3.9). Upregulation of genes in the accBC operon, encoding biotin
carboxyl carrier protein and biotin carboxylase, was also observed in cultures only expressing
BTE, both at OD600 of 2.0 in fermentors, and during early to mid-stationary phase in shake flasks
(Table 3.7). AccB and AccC were present at increased levels only at the mid-stationary phase
sampling point in shake flask (+1.7-fold, P = 0.187; +1.7-fold, P = 0.065) or fermentor cultures
(+1.7-fold, P = 0.140; +1.6-fold, P = 0.048) but with less than 2-fold increases and with P > 0.05
for most comparisons.
Expression of accA was decreased in BTE-expressing shake flask
cultures beginning in early stationary phase (-1.6-fold, P = 0.003) into mid-stationary phase (3.6-fold, P = 0.009), and only to a small extent in mid-stationary phase in fermentors (-1.7-fold,
P = 0.046). No significant change in expression of accD was observed. Transcription of the
genes encoding all four subunits of ACC is increased or decreased in response to nutrient
upshifts or downshifts, respectively [Li 1993], and transcription of accBC is autoregulated by
AccB [James 2004]. However, no prior knowledge of ACC transcriptional regulation can
explain the pattern of expression observed in BTE-expressing strains, with only accBC
104
exhibiting increased transcription. Expression of BTE has previously been reported to
increase production of AccB (other subunits were not analyzed) [Ohlrogge 1995], in line with
the results presented here, and to decrease levels of AccB (other subunits were not identified) on
addition of exogeneous oleic acid [Han 2008].
Other conditions that result in increased
expression of accB or accBC include rpoE (encoding σE) overexpression [Kabir 2005] and
heterologous expression of a mevalonate production pathway [Kizer 2008], respectively.
3.3.14 Carbohydrate transport and catabolism pathways
Several genes involved in maltose transport and metabolism are differentially induced in
BTE-expressing strains under all conditions in mid-log phase cultures, with subsequent
decreased expression relative to BTE-H204A-expressing strains in stationary phase.
These
include malE and malF which are located within an operon, and the malK-lamB-malM operon
(Appendix IV). The maltose ABC transporter is composed of MalKFGE. These operons are
activated by MalT, which oligomerizes in the presence of maltotriose to bind upstream of the
promoters, and CRP-cAMP, which is present at elevated levels in the absence of
phosphoenolpyruvate:carbohydrate phosphotransferase system (PTS) carbon sources [reviewed
in Boos 2000]. The three genes encoding the D-xylose ABC transporter, xylFGH, as well as
xylA, encoding xylose isomerase, are also upregulated in the BTE-expressing strains in stationary
phase.
Similarly, the araFGH operon, encoding the arabinose ABC transporter, manYZ,
encoding two subunits of the mannose PTS permease, and fucI, encoding L-fucose isomerase,
and rbsA, encoding one subunit of the ribose ABC transporter, are also upregulated. The genes
encoding the remaining subunits of the ribose ABC transporter, rbsB and rbsC, and the fucose
MFS transporter fucP, are also upregulated in stationary phase in the shake flask data set. Aside
105
from transcription factors that sense the specific carbohydrates and relieve repression of
transcription, expression of all of these genes is activated by CRP-cAMP.
Interestingly,
increased expression of malE and manXYZ and at early times has also been observed following
n-butanol exposure in E. coli [Rutherford 2010], and overexpression of manXYZ has been
observed to confer tolerance to hexane [Okochi 2007], implicating alternative roles for sugar
transporters. All of the maltose transport genes identified in this study were also shown to be
upregulated by overexpression of baeR [Nishino 2005], the response regulator of the BaeSR twocomponent system implicated in sensing envelope stress.
3.3.15 Decreased expression of PhoB regulon
PhoB is the response regulator of the PhoR/PhoB two-component system and is activated
(phosphorylated) under limiting phosphate conditions by PhoR [reviewed in Wanner 1996 and
Hsieh 2010]. Phosphorylated PhoB subsequently activates transcription of a number of genes
involved in phosphate transport and metabolism. A large number of phosphorylated PhoBactivated genes are strongly down-regulated in fatty acid overproducing strains at various
sampling times in both shake flasks and fermentors, including phoA, the pstSCAB-phoU operon,
yibD, phoE and phoB. PstSCAB encodes an inorganic phosphate ABC transporter, PhoA an
alkaline phosphatase with broad specificity toward phosphate monoesters, PhoU an auxiliary
protein involved in phosphate transport, and PhoE an outer membrane porin with preference for
inorganic phosphate and organophosphates. In BTE-expressing shake flask cultures, there is
additional statistically significant down-regulation of amn and phnCD during early stationary
phase, and argP, psiE, and eda during mid-stationary phase. PsiF, Amn, UgpB, and PstB were
all identified as having reduced protein expression in stationary phase BTE-expressing shake
106
flask cultures, while PhoA and PstB had reduced expression in BTE and ACCoverexpressing fermentor cultures (Tables 3.7 and 3.8). Transcription of eda is repressed by
phosphorylated PhoB at one promoter, but binding of other transcription factors at farther
upstream promoters can also repress transcription [Keseler 2011], therefore the observed
repression may not be due to the binding of phosphorylated PhoB. Increased expression of the
ugpBAECQ and oppABCDF operons are also observed in stationary phase in shake flasks. The
former operon, encoding a glycerol-3-phosphate ABC transporter, exhibits dual regulation by
phosphorylated PhoB. Despite increased transcript levels of ugpB, there was reduced protein
expression of UgpB, particularly in early stationary phase (Table 3.8).
The latter operon,
encoding an oligopeptide ABC transporter, has been formerly observed to be inhibited during
phosphate limitation [Smith 1992], consistent with phosphorylated PhoB-mediated repression. It
is not known why the cells appear to be sensing an excess of phosphate, however it may be a
result of the release of organophosphate esters from lysed cells.
Recent results from our
laboratory have indicated higher fatty acid yields from cultures in a MOPS minimal medium
with reduced phosphate levels (data not shown).
3.3.16 Initial attempts at functional genomics guided strain engineering
Stress responses that were induced as a result of fatty acid overproduction (phage shock
proteins and MarA/Rob/SoxS regulon) were selected as targets for additional strain engineering.
RL08 harboring pBAD33-BTE was selected as the base strain because it produces higher titers
of fatty acids than strains harboring other BTE-containing plasmids when grown in LB medium
supplemented with 0.4% glycerol (Chapter 2) [Hoover 2009].
The phage shock proteins
(encoded by pspABCDE and pspG), of which some members may function to repair or stabilize
107
damaged cell membranes [Joly 2010] were eliminated by deletion of the necessary
transcriptional activator, PspF (strain RL13), in order to gauge its importance in maintaining
viability in BTE-expressing cells (Table 3.5). After 8 h growth, there was no significant change
in CFU/mL when pspF was deleted in BTE-H204A-expressing cells, however there was a 4-fold
decrease in viability in BTE-expressing cells. This observation suggested that the phage shock
system played a role in reducing the loss of viability resulting from FFA overproduction.
Despite the loss in viability after 8 h, there was no significant difference in total fatty acid titers
after 24 h between RL08/pBAD33-BTE and RL13/pBAD33-BTE (Figure 3.14). Next, the phage
shock system was overexpressed by cloning the pspABCDE operon into pBAD18 (pBAD18-Psp).
Overexpression of pspABCDE has previously been shown to improve tolerance of E. coli to
hexane exposure, but to a lesser degree than culturing the cells in the presence of 10 mM MgSO4
(44). A 2-fold decrease in CFU/mL at 8 h was found in RL08/pBAD18-Psp expressing BTEH204A, and an over 7-fold decrease in CFU/mL was found when expressing BTE (Table 3.5).
Total fatty acid titers were similarly negatively affected in the strain overexpressing pspABCDE
and BTE over the strain expressing BTE alone (Figure 3.14). Interestingly, the decrease in titer
was almost entirely in medium-chain length (C8 to C14) rather than long-chain length fatty acids
(C16 to C18).
To gauge the physiological importance of the MarA, Rob, and SoxS regulons on maintenance
of viability during FFA overproduction, deletions of marA, rob, and soxS were constructed by P1
phage transduction of kan cassettes from the corresponding Keio collection mutants [Baba 2006]
into strain RL08. No decreases in CFU/mL after 8 h growth were observed from deletions in
RL08 expressing BTE-H204A, however deletion of soxS resulted in a small statistically
significant increase in CFU/mL (Table 3.5). In BTE-expressing cultures, a small but significant
108
Table 3.5 Viable cell counts from cultures of transcription factor deletion strains expressing
BTE-H204A or BTE. Cultures were sampled 8 h post-inoculation from induced cultures of RL08
harboring plasmids pBAD33-BTE-H204A or pBAD33-BTE with additional specified gene deletions
grown in LB medium supplemented with 0.4% glycerol.
Viable cell counts (CFU mL-1)
Strain
RL08
RL08 ∆pspF
RL08
RL08 marA::kan
RL08 rob::kan
RL08 soxS::kan
BTE-H204A
BTE
(2.02 ± 0.18) ×109
(2.49 ± 0.49) ×109
(7.3 ± 0.2) ×108
(1.8 ± 0.1) ×108
(3.2 ± 0.9) ×109
(3.0 ± 0.6) ×109
9
(2.7 ± 0.1) ×10
(5.3 ± 0.2) ×109
(5.9 ± 1.8) ×108
(4.6 ± 0.8) ×108
8
(3.3 ± 0.2) ×10
(4.4 ± 2.6) ×108
loss of viability was observed in the rob deletion strain. While only Rob is known to be
activated by FFAs [Rosenberg 2003], there were no trends in the microarray data to suggest that
it had a more important role than MarA or SoxS, and genes annotated as being exclusively
activated by Rob (ybiS, mltF, and aslB) [Gama-Castro 2011] did not have highly increased
expression. In addition to having reduced CFU/mL after 8 h, the rob deletion strain expressing
BTE also produced a lower C8 to C14 fatty acid titer relative to RL08 (Figure 3.14).
Overexpression of marA, rob, and soxS on pBAD18 did not result in any statistically significant
changes to CFU/mL (Table 3.6) or fatty acid titers (Figure 3.14). While overexpression of either
marA, rob, or soxS has previously been observed to increase tolerance toward hexane [Hayashi
2003], cyclohexane [Hayashi 2003, White 1997], triclosan [McMurry 1998], antibiotics [Asako
1997, Warner 2010], and cationic antimicrobial peptides [Warner 2010], it did not increase
tolerance to endogenously overproduced FFAs.
109
Table 3.6 Viable cell counts from cultures co-expressing BTE-H204A or BTE and selected
genes cloned in pBAD18. Cultures were sampled 8 h post-inoculation from induced cultures of RL08
harboring plasmids pBAD33-BTE-H204A or pBAD33-BTE and additional specified plasmids (see
Appendix II) in LB medium supplemented with 0.4% glycerol.
Viable cell counts (CFU mL-1)
Strain
BTE-H204A
BTE
RL08/pBAD18
RL08/pBAD18-Psp
(3.24 ± 0.11) ×109
9
(1.59 ± 0.23) ×10
(5.3 ± 1.6) ×108
7
(6.8 ± 3.5) ×10
RL08/pBAD18
RL08/pBAD18-MarA
RL08/pBAD18-Rob
RL08/pBAD18-SoxS
(2.34 ± 0.23) ×109
(2.74 ± 0.55) ×109
(2.31 ± 0.18) ×109
(2.97 ± 0.69) ×109
(2.94 ± 0.78) ×108
(4.5 ± 2.6) ×108
(2.0 ± 1.0) ×108
(4.9 ± 2.3) ×108
Fig. 3.14 Fatty acid titers of BTE and BTE-H204A expressing strains with deleted transcription
factors or selected overexpressed genes. Titers of 16 to 18 carbon chain length fatty acids (white),
which are primarily associated with membrane lipids, and 8 to 14 carbon chain length fatty acids (gray),
which are primarily either FFAs or minorly associated with lipid A of (a) strain RL13 (∆pspF) expressing
BTE-H204A (BTE-) or BTE (BTE+), and (b) strains RL09 (marA::kan), RL10 (rob::kan), and RL11
(soxS::kan) expressing BTE-H204A or BTE. Strain RL08 was used as a negative control against gene
deletion strains. All fatty acid samples were taken after 24 h of shake flask growth in LB medium
supplemented with 0.4% glycerol as described in Materials and Methods.
110
3.4 Discussion
Overproduction of medium-chain length (C8 to C14) FFAs in E. coli by heterologous
expression of the Umbellularia californica acyl-ACP thioesterase (BTE) results in severe
phenotypic consequences beginning as early as the transition out of log phase growth, with
concomitant large reductions in CFU/mL and compromised inner membrane integrity as
indicated by staining with SYTOX Green. Additionally, heterogeneous cell size is observed by
flow cytometry analysis in BTE-expressing cells, indicating possible defects in cell division on
entry into stationary phase. The observed degree of toxicity and morphological changes are
unique to endogenous production and cannot be replicated by exogenous addition of lauric acid.
Previous studies investigating the exogenous addition of 0.4% decanoic, methyldecanoic acid, or
undecanoic acid have similarly observed little growth inhibition [Fay 1975, Fay 1977], however
much lower concentrations (0.01%) of methyldecanoic acid induced lysis of E. coli spheroplasts
[Fay 1977]. The outer membrane of E. coli is decorated with hydrophilic lipopolysaccharides,
providing a protective barrier against diffusive entry of hydrophobic compounds and completely
excluding free fatty acids with 14 or more carbon chain lengths without the presence of the outer
membrane fatty acid transporter FadL [Black 1990]. In contrast, endogenous production exposes
the unprotected inner membrane to FFAs. While FFAs are able to passively diffuse across the
inner membrane, the rate of flipping across the bilayer or exiting the membrane into the
periplasm may be slower than the rate of intracellular production, which would lead to
accumulation in either the cell membranes or the periplasm.
In this study, we found that overproduction of medium-chain length FFAs in E. coli resulted
in significantly perturbed gene expression profiles that correlated temporally with elevated rates
of fatty acid production. Cross-comparison of differentially expressed genes under two sets of
111
growth conditions and with BTE and BTE-H204A expressed on plasmids under the control
of different inducible promoters allowed the identification of a core set of stress responses that
may be generally induced as a result of endogenous FFA overproduction. Two stress response
regulons were identified as having all or many members up-regulated in both sets of microarray
data, with additional proteins identified as differentially expressed in the more limited set of
statistically significant proteomic data. All members of the phage shock system were strongly
induced in BTE-expressing cultures (Table 3.7), with PspA identified as having an increased
abundance at the protein level (Tables 3.7 and 3.8). Activation of the phage shock system
appears to occur in response to dissipation of the proton motive force, however the exact
mechanism of signal transduction remains elusive [Joly 2010]. It is appears that at least one of
the phage shock proteins, PspA, plays a role in restoring membrane polarization [Kobayashi
2007]. Many of the key genes of aerobic respiration were upregulated, including the nuo and cyo
operons, possibly as a result of the degradation of membrane properties or proton motive force.
The products of these genes are necessary to generate a proton gradient across the inner
membrane for aerobic energy generation. Exogenously added fatty acids act as proton motive
force uncoupling agents in mitochondria by various mechanisms including diffusive flipping of
proton-bound free acids into the cytosol, and direct interaction of membrane bound fatty acid
anions with protons in membrane proteins such as ATP synthase [Wojtczak 1999]. Export of
endogenously overproduced FFAs should result in a net efflux of protons by diffusive
mechanisms, but they may still behave as uncoupling agents by causing futile cycling of protons
that were originally bound for the cytosol.
Alternatively, FFAs may accumulate in the
periplasmic space and cause uncoupling by diffusing back into the cytosol at later times during
112
culture growth, essentially mimicking exogenously added FFAs. The observation that the
phage shock system was only induced in stationary phase supports this hypothesis.
Numerous members of the MarA/Rob/SoxS regulon, which are implicated in multiple
antibiotic resistance, resistance to solvents and other small toxic molecules, and oxidative stress,
were also strongly induced (Table 3.7) as early as during mid-log phase, although no clear trend
could be ascertained as to whether one or more of these transcription factors was responding to a
particular environmental stimulus. Rob activity is known to be increased by medium-chain
FFAs such as decanoate, and it has been suggested that E. coli possesses natural defense
mechanisms for dealing with high concentrations of fatty acids and bile salts in the enteric
environment [Rosenberg 2003].
While members of other cell envelope stress responses
including the σE, CpxR, and BaeR regulons were identified in shake flask cultures, induction of
these regulons was notably absent in fermentor cultures.
Thus it seems likely that other
environmental factors in combination with FFA production were the cause of this response. For
example, increased intercalation of FFAs in cell membranes of BTE-expressing cells could
render the cells more sensitive to ampicillin (present in the shake flask cultures), but not to
chloramphenicol or kanamycin (present in the fermentors), and this secondary impact could be
responsible for the induced stress responses. In light of the mode of action of ampicillin as a
peptidoglycan transpeptidase inhibitor, it is also tempting to speculate that the wide range of cell
lengths present in BTE-expressing cells during stationary phase is a result of inhibited cell
division due to effectively increased ampicillin exposure. Another possibility is that the less
controlled conditions present in shake flask cultures with respect to temperature and pH could
have imposed additional stresses on the cells, leading to further cell envelope stress cascades.
113
Increased acid stress or other indirect transcriptional cascades could be responsible for the
very large induction of genes present on the acid fitness island in shake flask cultures.
The induction of stress responses could be a direct result of a change in membrane properties
due to FFA intercalation in the inner membrane or because of other indirect effects on the
composition of membrane lipids. Notably, large decreases in expression of fabB and fabA were
observed in BTE-expressing strains, with the onset of decreased expression of fabB occurring as
early as mid-log phase. FabB was identified as having reduced expression in both proteomics
data sets. This early onset of differential regulation points to a rapid and severe perturbation of
unsaturated membrane lipid homeostasis that precedes most stress responses, as both fabB and
fabA are required for long-chain unsaturated fatty acid biosynthesis.
They are negatively
regulated by FabR, which effectively senses the ratio of unsaturated to saturated acyl-ACPs and
tunes the transcription of genes required for unsaturated fatty acid biosynthesis accordingly [Zhu
2009]. Despite the high degree of transcriptional repression, the fatty acid profile revealed a
sharply increased degree of C16 to C18 fatty acid unsaturation due primarily to a decrease in
C16:0 content in BTE-expressing strains (Figure 3.8). These antagonizing observations seem to
point to an inability of the cells to compensate for an extreme enrichment of unsaturated acylACPs as a result of the saturated acyl-ACP substrate preference of BTE, rather than an attempt
by the cells to purposefully alter their phospholipid unsaturated fatty acid content. An inability
to effectively control the degree of membrane lipid unsaturation could lead to severe
consequences on maintenance of membrane integrity and could also exacerbate or be the primary
factor responsible for proton motive force dissipation.
Indeed, Bacillus subtilis and B.
megaterium mutants defective in phospholipid desaturase activity exhibit an increased degree of
membrane saturation while attaining increased resistance to proton motive force uncouplers such
114
as carbonyl cyanide 3-chlorophenylhydrazone (CCCP) [Dunkley 1991]. Regardless of the
mechanism, any detrimental impact of unsaturated membrane lipid content to cell viability
presents metabolic engineering opportunities in the form of correcting the degree of membrane
lipid unsaturation.
Alternatively, if the primary membrane stress arises from FFA intercalation in the inner or
outer membranes or accumulation in the periplasm, then a possible metabolic engineering
strategy is to overexpress either a native or heterologous fatty acid exporter. While speculated to
exist in plant chloroplast membranes [Koo 2004], no specific fatty acid export protein has been
convincingly identified to date. However, less specific exporter proteins may exist in the form of
drug efflux pumps, which can confer resistance to a wide range of toxic small molecules [Ramos
2002, Pos 2009]. The AcrAB-TolC efflux pump, a member of the MarA and Rob regulons, is
not only activated in response to fatty acids via Rob [Ma 1995, Rosenberg 2003], but
overexpression has been shown to confer greatly increased resistance to chemicals such as
sodium dodecyl sulfate and deoxycholate [Nishino 2001]. A number of other drug efflux pumps
that are not in the MarA/Rob/SoxS regulon confer similar resistances [Nishino 2001, Fàbrega
2010].
Gene deletions of pspF and rob negatively affected cell viability (and in the case of rob,
reduce C8 to C14 fatty acid titers) in BTE-expressing cells, underscoring the physiological
importance of the phage shock response and Rob regulon under conditions of endogenous FFA
overproduction. Despite their importance, higher cell viabilities or fatty acid titers were not
observed when overexpressing the genes encoding the phage shock proteins, or marA, rob, or
soxS on multicopy plasmids. Further optimization of expression may be necessary. In the case
of overexpression of marA, rob, and soxS, additional deletions of attenuating transcriptional
115
repressors such as marR, acrR, and mprA may be necessary to achieve higher fatty acid
yields. Strategies for specifically repairing membrane damage may need to be carried out in
tandem with modifications that reduce the degree of membrane lipid unsaturation. Alterations in
growth conditions can also be explored, such as the addition of 10 mM Mg2+ or Ca2+, which has
previously been shown to reduce the lethality of hexane toward E. coli [Aono 1994].
3.5 Conclusion
In summary, several lines of evidence for the toxicity of intracellularly produced mediumchain length FFAs have been obtained.
In fatty acid overproducing strains, this includes
compromised inner membrane integrity, altered cellular morphology, reduced viability, and a
large increase in percent composition of unsaturated 16 to 18 carbon chain length fatty acids. A
functional basis for many of these observations was determined by an analysis of differential
gene and protein expression between a fatty acid overproducing and a non-overproducing strain.
Genes encoding the phage shock proteins, and many members of the MarA/Rob/SoxS regulon
exhibited increased expression in the FFA overproducing strain, indicating transcriptional
cascades in response to unfavorable conditions such as proton motive force dissipation and the
presence of toxic small molecules and oxidative stress. The increase in unsaturated long-chain
fatty acid content could be functionally linked to changes in gene (fabB and fabA) and protein
expression (FabB).
The importance of the phage shock proteins and the Rob regulon to
maintaining cell viability under conditions of FFA overproduction were validated by testing of
gene deletion strains. This study opens up the possibility for directed rational strain engineering
to achieve improved yields of free fatty acids. Chapters 4 and 5 will discuss work toward two
different areas of rational strain engineering: identifying native FFA export proteins in E. coli,
116
and further investigation into the effect of unsaturated membrane lipid content on cell
viability and FFA production.
117
Table 3.7 Selected transcript fold-changes from shake flask and fermentor cultures. Foldchanges are shown for BTE-expressing strains versus BTE-H204A-expressing strains cultured in EZ Rich
Defined Medium supplemented with 0.2% glucose (EZ glucose) in shake flasks, and in EZ Rich Defined
Medium supplemented with 0.4% glycerol (EZ glycerol) in fermentors at sampling times specified in the
text. Only those fold-changes with P-values greater than 0.05 (EZ glucose) or 0.10 (EZ glycerol) are
shown.
EZ glucose
locus
gene name
EZ glycerol
mid-log
early stationary
mid-stationary
late log
early stationary
mid-stationary
Phage shock proteins
b1304
pspA
b1305
pspB
pspC
b1306
b1307
pspD
b1308
pspE
pspG
b4050
-1.02 (0.714)
1.03 (0.635)
-1.05 (0.704)
-1.03 (0.818)
-1.15 (0.323)
9.45 (0.000)
5.18 (0.001)
4.11 (0.002)
4.99 (0.002)
2.74 (0.000)
13.85 (0.000)
11.04 (0.001)
7.32 (0.001)
5.07 (0.000)
9.60 (0.002)
11.40 (0.008)
13.23 (0.011)
-2.02 (0.187)
-2.34 (0.263)
-2.46 (0.265)
-2.68 (0.215)
2.11 (0.228)
10.47 (0.000)
7.09 (0.000)
5.34 (0.000)
9.78 (0.004)
8.35 (0.005)
13.59 (0.001)
12.63 (0.000)
12.07 (0.000)
8.76 (0.001)
19.10 (0.000)
19.47 (0.001)
38.14 (0.003)
MarA/Rob/SoxS
regulon
b0447
ybaO
b0463
acrA
b0578
nfsB
b0683
fur
b0684
fldA
b0819
ybiS
b0850
ybjC
b0851
nfsA
rimK
b0852
b0853
ybjN
b0871
poxB
b0950
pqiA
b0951
pqiB
b1014
putA
b1053
mdtG
ptsG
b1101
b1277
ribA
b1530
marR
b1531
marA
b1532
marB
b1611
fumC
zwf
b1852
b1973
zinT
b2159
nfo
b2237
inaA
b3035
tolC
b3037
ygiB
b3038
ygiC
b3506
slp
b3507
dctR
b3508
yhiD
b3509
hdeB
b3510
hdeA
b3624
rfaZ
b3625
rfaY
b3800
aslB
b3908
sodA
b3924
fpr
b4025
pgi
b4062
soxS
b4177
purA
b4396
rob
b4637
uof
1.09 (0.619)
-1.10 (0.776)
1.20 (0.305)
1.02 (0.755)
-1.06 (0.396)
-1.11 (0.397)
1.11 (0.741)
1.36 (0.004)
1.18 (0.055)
1.21 (0.530)
1.06 (0.264)
1.09 (0.444)
1.21 (0.233)
1.11 (0.414)
1.36 (0.153)
1.00 (0.955)
1.02 (0.664)
-1.01 (0.942)
2.50 (0.037)
1.73 (0.018)
1.20 (0.320)
1.11 (0.212)
1.12 (0.265)
2.75 (0.094)
1.04 (0.859)
1.58 (0.001)
-1.01 (0.975)
-1.04 (0.512)
-1.03 (0.923)
-1.08 (0.735)
1.02 (0.836)
1.06 (0.817)
-1.02 (0.707)
-1.01 (0.895)
1.10 (0.112)
-1.01 (0.942)
-1.01 (0.954)
1.05 (0.366)
-1.24 (0.028)
-1.07 (0.366)
1.09 (0.301)
-1.03 (0.604)
-1.19 (0.386)
1.73 (0.018)
-1.45 (0.005)
1.37 (0.043)
1.22 (0.311)
1.20 (0.062)
-2.28 (0.000)
1.62 (0.103)
2.06 (0.020)
1.66 (0.121)
1.85 (0.002)
1.14 (0.183)
2.52 (0.003)
1.06 (0.839)
-10.18 (0.063)
1.83 (0.013)
-2.61 (0.000)
-1.42 (0.007)
2.72 (0.004)
5.21 (0.000)
-1.36 (0.037)
4.39 (0.000)
-1.77 (0.001)
1.75 (0.001)
2.00 (0.032)
7.72 (0.000)
-1.07 (0.536)
1.78 (0.006)
1.61 (0.002)
3.95 (0.000)
1.38 (0.006)
2.05 (0.025)
3.94 (0.002)
3.11 (0.002)
-1.06 (0.561)
1.09 (0.395)
1.47 (0.034)
1.47 (0.002)
1.51 (0.006)
-1.09 (0.405)
2.56 (0.008)
-1.78 (0.007)
1.35 (0.203)
-11.32 (0.042)
4.64 (0.000)
-1.33 (0.175)
1.79 (0.088)
1.69 (0.031)
2.80 (0.001)
-2.05 (0.324)
2.65 (0.000)
4.29 (0.000)
2.54 (0.005)
1.25 (0.174)
1.76 (0.014)
5.92 (0.000)
2.94 (0.003)
2.53 (0.018)
2.20 (0.007)
-2.86 (0.014)
1.03 (0.912)
5.20 (0.000)
6.25 (0.000)
-1.43 (0.065)
8.51 (0.000)
-1.31 (0.084)
1.30 (0.463)
2.46 (0.004)
5.72 (0.000)
-1.34 (0.112)
2.77 (0.003)
5.53 (0.000)
22.72 (0.000)
1.69 (0.006)
6.85 (0.020)
67.71 (0.000)
36.27 (0.000)
-1.08 (0.569)
-1.11 (0.409)
1.65 (0.028)
2.29 (0.001)
2.22 (0.014)
2.47 (0.020)
1.22 (0.748)
-5.92 (0.000)
2.85 (0.002)
1.10 (0.587)
-1.13 (0.567)
1.46 (0.472)
1.29 (0.298)
1.42 (0.298)
-1.24 (0.357)
1.12 (0.701)
1.02 (0.904)
1.42 (0.292)
2.06 (0.079)
-3.48 (0.119)
-1.04 (0.733)
1.21 (0.566)
6.06 (0.235)
-1.19 (0.284)
1.25 (0.408)
1.33 (0.452)
1.96 (0.031)
2.24 (0.191)
-1.34 (0.111)
1.35 (0.257)
-1.43 (0.271)
1.12 (0.691)
-1.23 (0.292)
1.34 (0.073)
-1.23 (0.135)
1.29 (0.451)
2.14 (0.104)
-1.66 (0.049)
1.05 (0.707)
1.05 (0.751)
-1.16 (0.433)
-1.49 (0.166)
1.43 (0.218)
1.12 (0.760)
1.55 (0.131)
1.22 (0.677)
-1.46 (0.280)
-1.38 (0.211)
-5.94 (0.118)
1.46 (0.087)
-1.19 (0.499)
1.22 (0.391)
1.06 (0.618)
1.18 (0.153)
1.53 (0.043)
1.21 (0.464)
1.16 (0.144)
1.23 (0.276)
1.73 (0.003)
1.21 (0.415)
1.14 (0.484)
1.32 (0.198)
-1.20 (0.266)
-1.08 (0.618)
-1.05 (0.660)
1.84 (0.085)
-1.01 (0.961)
1.16 (0.146)
1.37 (0.052)
2.79 (0.013)
2.91 (0.001)
1.17 (0.461)
1.04 (0.873)
1.10 (0.442)
-1.02 (0.852)
1.47 (0.228)
2.10 (0.001)
-1.03 (0.081)
1.33 (0.058)
1.34 (0.110)
-1.06 (0.828)
1.05 (0.558)
-1.13 (0.270)
-1.16 (0.503)
-1.21 (0.239)
1.31 (0.043)
1.35 (0.038)
-1.03 (0.763)
1.49 (0.065)
-1.08 (0.558)
-1.06 (0.842)
-1.46 (0.396)
1.21 (0.112)
-1.17 (0.272)
3.28 (0.125)
1.28 (0.461)
-1.09 (0.844)
1.74 (0.164)
1.25 (0.204)
-1.07 (0.635)
5.32 (0.002)
2.32 (0.007)
1.30 (0.182)
-1.65 (0.157)
1.12 (0.659)
2.14 (0.100)
1.32 (0.310)
2.27 (0.046)
2.09 (0.033)
1.96 (0.060)
1.41 (0.113)
5.96 (0.007)
6.10 (0.010)
1.30 (0.157)
2.15 (0.013)
1.01 (0.984)
-1.77 (0.177)
1.89 (0.044)
7.52 (0.005)
1.00 (0.978)
1.10 (0.777)
-1.50 (0.248)
-1.04 (0.949)
-1.05 (0.773)
-2.05 (0.209)
-3.67 (0.018)
-3.27 (0.017)
-1.05 (0.757)
1.02 (0.921)
1.16 (0.425)
1.05 (0.863)
-1.43 (0.152)
-1.58 (0.089)
4.52 (0.057)
-1.21 (0.479)
-1.07 (0.856)
1.12 (0.396)
1.26 (0.323)
118
Table 3.7 (cont.)
EZ glucose
locus
gene name
Energy metabolism
b0428
cyoE
b0429
cyoD
b0430
cyoC
cyoB
b0431
b0432
cyoA
b0972
hyaA
b0973
hyaB
b0974
hyaC
b0975
hyaD
b0976
hyaE
b0977
hyaF
b0978
appC
b0979
appB
b0980
appA
b1974
yodB
b2276
nuoN
b2277
nuoM
b2278
nuoL
b2279
nuoK
b2280
nuoJ
b2281
nuoI
b2282
nuoH
b2410
yfeH
yqfA
b2899
b2996
hybA
b3892
fdoI
b3894
fdoG
b4039
ubiC
b4040
ubiA
b4154
frdA
sthA
b3962
Unsaturated fatty
acid biosynthesis
b0954
fabA
b2323
fabB
β-oxidation
b3845
fadA
b0221
fadE
b3081
fadH
b2342
fadI
b2341
fadJ
Other genes
b0464
acrR
b0929
ompF
b2684
mprA
b3255
accB
b3256
accC
EZ glycerol
mid-log
early stationary
mid-stationary
late log
early stationary
mid-stationary
1.09 (0.241)
1.03 (0.796)
1.07 (0.256)
1.01 (0.747)
1.13 (0.339)
-1.05 (0.584)
1.02 (0.922)
-1.01 (0.803)
-1.02 (0.915)
-1.00 (0.988)
1.00 (0.957)
-1.07 (0.515)
1.13 (0.514)
1.07 (0.584)
1.67 (0.056)
1.08 (0.591)
1.09 (0.672)
-1.03 (0.538)
-1.00 (0.986)
1.02 (0.726)
1.02 (0.751)
1.04 (0.702)
-1.07 (0.757)
-2.12 (0.136)
-1.31 (0.105)
-1.02 (0.869)
-1.21 (0.022)
-1.14 (0.311)
-1.02 (0.811)
-1.67 (0.564)
1.47 (0.010)
1.24 (0.095)
-1.14 (0.287)
-1.10 (0.518)
1.28 (0.037)
-1.42 (0.071)
-1.95 (0.016)
-1.24 (0.057)
-1.18 (0.337)
1.04 (0.659)
-1.04 (0.694)
-3.07 (0.001)
-2.26 (0.000)
-2.06 (0.001)
-2.57 (0.000)
-1.21 (0.218)
-1.81 (0.000)
-1.70 (0.001)
-1.70 (0.083)
-1.25 (0.143)
-1.06 (0.584)
-1.02 (0.865)
1.36 (0.083)
-8.77 (0.006)
-1.16 (0.280)
1.55 (0.042)
2.48 (0.017)
1.95 (0.007)
1.75 (0.001)
1.16 (0.211)
1.46 (0.028)
3.65 (0.003)
2.47 (0.007)
2.00 (0.015)
2.15 (0.001)
2.24 (0.004)
15.29 (0.000)
14.59 (0.000)
19.18 (0.000)
12.76 (0.000)
5.47 (0.001)
5.06 (0.005)
-1.59 (0.175)
1.12 (0.427)
-1.28 (0.260)
-1.20 (0.499)
3.06 (0.103)
1.64 (0.448)
1.91 (0.104)
1.80 (0.119)
2.03 (0.081)
2.91 (0.008)
3.27 (0.008)
2.53 (0.038)
-1.89 (0.111)
2.84 (0.001)
2.91 (0.000)
12.49 (0.000)
2.60 (0.005)
2.07 (0.014)
2.22 (0.001)
4.71 (0.000)
1.81 (0.096)
1.78 (0.057)
1.83 (0.129)
1.90 (0.205)
2.27 (0.159)
-1.21 (0.352)
-1.13 (0.417)
-1.33 (0.219)
-1.36 (0.165)
-1.11 (0.473)
-1.20 (0.371)
-1.52 (0.097)
-1.17 (0.466)
-1.67 (0.074)
-1.30 (0.380)
2.42 (0.015)
2.93 (0.006)
1.68 (0.019)
1.76 (0.012)
1.82 (0.030)
1.92 (0.054)
1.83 (0.071)
1.20 (0.338)
-14.66 (0.020)
2.25 (0.015)
1.05 (0.903)
-1.39 (0.300)
-1.04 (0.890)
-1.29 (0.202)
1.83 (0.094)
1.65 (0.271)
1.66 (0.168)
1.51 (0.224)
1.50 (0.279)
1.67 (0.075)
-1.07 (0.688)
-1.16 (0.219)
-1.07 (0.680)
-1.14 (0.242)
-1.05 (0.607)
-1.13 (0.407)
-1.74 (0.014)
-1.82 (0.016)
-1.11 (0.453)
-2.40 (0.016)
-1.24 (0.353)
-1.16 (0.578)
-1.26 (0.290)
-1.25 (0.268)
-1.22 (0.273)
-1.17 (0.366)
1.01 (0.968)
-1.03 (0.769)
-32.77 (0.000)
-1.26 (0.073)
1.14 (0.489)
1.13 (0.438)
1.26 (0.313)
1.21 (0.192)
-1.09 (0.586)
1.26 (0.087)
1.98 (0.042)
1.95 (0.046)
2.05 (0.060)
1.56 (0.331)
1.48 (0.228)
1.27 (0.682)
1.16 (0.665)
1.15 (0.595)
1.35 (0.211)
1.29 (0.280)
-1.02 (0.946)
-2.95 (0.092)
-2.76 (0.087)
-1.98 (0.017)
1.04 (0.793)
-1.75 (0.378)
-1.62 (0.285)
-1.24 (0.745)
-1.29 (0.673)
1.06 (0.899)
-1.14 (0.777)
-1.23 (0.712)
1.81 (0.053)
-1.15 (0.487)
-1.07 (0.689)
2.33 (0.070)
2.32 (0.072)
2.80 (0.018)
4.51 (0.001)
1.36 (0.024)
-1.28 (0.087)
-1.87 (0.000)
-2.54 (0.004)
-12.95 (0.001)
-4.80 (0.000)
-1.57 (0.334)
-2.94 (0.019)
-10.28 (0.001)
-3.58 (0.004)
-27.33 (0.000)
-1.24 (0.271)
-2.35 (0.099)
1.68 (0.001)
3.96 (0.002)
3.30 (0.001)
3.07 (0.006)
3.32 (0.001)
2.15 (0.009)
3.14 (0.053)
1.29 (0.040)
1.76 (0.017)
2.09 (0.041)
3.64 (0.020)
2.93 (0.032)
1.13 (0.545)
-10.43 (0.002)
6.11 (0.000)
3.82 (0.002)
2.69 (0.001)
1.30 (0.251)
2.75 (0.299)
2.12 (0.082)
7.55 (0.060)
7.78 (0.063)
1.93 (0.003)
1.60 (0.019)
1.15 (0.048)
1.49 (0.100)
-1.24 (0.064)
1.37 (0.023)
1.23 (0.026)
1.96 (0.000)
-7.72 (0.000)
2.10 (0.000)
2.17 (0.048)
2.09 (0.002)
1.37 (0.035)
1.32 (0.078)
1.66 (0.018)
1.66 (0.108)
3.44 (0.000)
1.58 (0.024)
3.46 (0.006)
1.87 (0.027)
2.11 (0.029)
1.67 (0.072)
2.60 (0.017)
-6.93 (0.037)
10.39 (0.001)
119
Table 3.8 Proteins differentially expressed in EZ glucose shake flask cultures. Differentially
expressed proteins were defined as having a minimum 2-fold change and P<0.05 in the fatty acid
overproducing strain grown in EZ rich defined medium supplemented with 0.2% glucose.
Protein name and
gene locus
Increased expression
FadE
b0221
GltA
b0720
SucA
b0726
MoaB b0782
OmpX b0814
ClpA
b0882
Agp
PutA
b1002
b1014
AdhE
b1241
PspA
DbpA
YncB
Sra
GadB
LsrF
FumC
Spy
YeaG
FliC
FliD
HchA
GatA
GatZ
FadL
TalA
YfiD
RaiA
GabD
GlcB
GlcG
ExbB
DkgA
YhcB
AaeA
OmpR
b1304
b1343
b1449
b1480
b1493
b1517
b1611
b1743
b1783
b1923
b1924
b1967
b2094
b2095
b2344
b2464
b2579
b2597
b2661
b2976
b2977
b3006
b3012
b3233
b3241
b3405
Slp
YhhA
LivJ
XylF
AldB
MtlA
MtlD
LldD
EnvC
TnaA
CorA
Fre
FadA
GlpK
Acs
DcuB
AspA
FrdB
YtfQ
b3506
b3448
b3460
b3566
b3588
b3599
b3600
b3605
b3613
b3708
b3816
b3844
b3845
b3926
b4069
b4123
b4139
b4153
b4227
YjiY
b4354
Decreased expression
CarA
b0032
PsiF
b0384
YaiE
b0391
IspE
b1208
LsrB
b1516
Gene product/function
mid-log ratio
BTE/BTE-H204A
early stat. ratio
mid-stat. ratio
acyl coenzyme A dehydrogenase
citrate synthase
2-oxoglutarate decarboxylase, thiamin-requiring
molybdopterin biosynthesis protein B
outer membrane protein X
ATPase and specificity subunit of ClpA-ClpP ATP-dependent serine
protease, chaperone activity
glucose-1-phosphatase / inositol phosphatase
fused DNA-binding transcriptional regulator / proline dehydrogenase /
pyrroline-5-carboxylate dehydrogenase
fused acetaldehyde-CoA dehydrogenase / iron-dependent alcohol
dehydrogenase / pyruvate-formate lyase deactivase
regulatory protein for the phage-shock-protein operon
ATP-dependent RNA helicase, specific for 23S rRNA
predicted NADP-dependent, Zn-dependent oxidoreductase
stationary-phase-induced ribosome-associated protein
glutamate decarboxylase B, PLP-dependent
putative autoinducer-2 (AI-2) aldolase
fumarate hydratase (fumarase C), aerobic Class II
envelope stress induced periplasmic protein
protein kinase, function unknown; autokinase
flagellar filament structural protein (flagellin)
flagellar filament capping protein
Hsp31 molecular chaperone
galactitol-specific enzyme IIA component of PTS
subunit of tagatose-1,6-bisphosphate aldolase 2
long-chain fatty acid outer membrane transporter
transaldolase A
autonomous glycyl radical cofactor
cold shock protein associated with 30S ribosomal subunit
succinate-semialdehyde dehydrogenase I, NADP-dependent
malate synthase G
conserved protein
membrane spanning protein in TonB-ExbB-ExbD complex
2,5-diketo-D-gluconate reductase A
conserved protein
p-hydroxybenzoic acid efflux system component
DNA-binding response regulator in two-component regulatory system
with EnvZ
outer membrane lipoprotein
conserved protein, DUF2756 family
leucine/isoleucine/valine transporter subunit
D-xylose transporter subunit
aldehyde dehydrogenase B
subunit of mannitol PTS permease
mannitol-1-phosphate dehydrogenase, NAD(P)-binding
L-lactate dehydrogenase, FMN-linked
activator of AmiB,C murein hydrolases, septal ring factor
tryptophanase / L-cysteine desulfhydrase, PLP-dependent
magnesium/nickel/copper transporter
NAD(P)H-flavin reductase
3-ketoacyl-CoA thiolase (thiolase I)
glycerol kinase
acetyl-CoA synthetase
C4-dicarboxylate antiporter
aspartate ammonia-lyase
fumarate reductase iron-sulfur protein
galactofuranose binding protein: periplasmic-binding component of
ABC superfamily
predicted inner membrane protein
2.23
-
∞
2.05
2.43
2.03
3.10
3.33
-
2.71
-
2.62
2.10
2.04
-
2.14
carbamoyl phosphate synthetase small subunit, glutamine amidotransferase
conserved protein, PsiF family, pho regulon
conserved protein, UPF0345 family
4-diphosphocytidyl-2-C-methylerythritol kinase
autoinducer-2 binding protein
-
2.28
2.23
2.73
-
2.66
2.29a
∞
2.32
2.82
2.62
-
5.52
2.38
4.63
3.53
5.24
2.12
2.48
∞
4.15
5.57
2.02
3.45
3.63
2.22
3.54
3.84
2.02
2.61
2.56
2.05
-
2.45
2.33
3.01
-
2.10
4.02
3.15
2.16
4.61
2.96
2.14
4.42
2.26
4.42
2.19
4.18
2.39
2.60
2.04
2.29
4.18
2.73
-
-
-2.04
-5.20
-2.08
-∞
-2.32
-5.14
-
120
Table 3.8 (cont.)
Protein name and
gene locus
Gene product/function
Decreased expression (cont.)
HdhA b1619
YebV b1836
Amn
b1982
BglX
b2132
PurF
b2312
FabB
b2323
YfeR
b2409
PurM
b2499
StpA
b2669
MltA
b2813
SspA
b3229
UgpB b3453
BcsA
b3533
PstB
b3725
IlvD
b3771
GlnA
b3870
MetB
b3939
MetL
b3940
MetF
b3941
PyrI
b4244
YjiH
b4330
7-α-hydroxysteroid dehydrogenase, NAD dependent
predicted protein
AMP nucleosidase
β-D-glucoside glucohydrolase, periplasmic
amidophosphoribosyltransferase
3-oxoacyl-[acyl carrier protein] synthase I
predicted DNA-binding transcriptional regulator
phosphoribosylaminoimidazole synthetase
DNA binding protein, nucleoid-associated
membrane-bound lytic murein transglycosylase A
stringent starvation protein A
glycerol-3-phosphate transporter subunit
cellulose synthase, catalytic subunit
phosphate transporter subunit
dihydroxyacid dehydratase
glutamine synthetase
cystathionine γ-synthase, PLP-dependent
fused aspartokinase II / homoserine dehydrogenase II
5,10-methylenetetrahydrofolate reductase
aspartate carbamoyltransferase, regulatory subunit
predicted inner membrane protein
mid-log ratio
-∞
-∞
BTE/BTE-H204A
early stat. ratio
mid-stat. ratio
-2.02
-3.06
-2.06
-2.03
-4.20
-2.54
-2.26
-8.18
-∞
-3.78
-3.97
-2.98
-2.35
-
-2.76
-3.46
-2.49
-2.43
-2.52
-∞
-3.15
-2.49
-2.19
-3.16
-2.39
-2.39
-
Bolded protein names also were differentially expressed greater than 2.0-fold (P<0.05) as transcripts at any of the three sampling points.
a
No transcript data has been obtained for this gene.
121
Table 3.9 Proteins differentially expressed in EZ glycerol fermentor cultures. Differentially
expressed proteins were defined as having a minimum 2-fold change and P<0.05 in the fatty acid
overproducing strain grown in EZ rich defined medium supplemented with 0.4% glycerol. Both fatty
acid overproducing strains (BTE+ ACC+, BTE+ ACC-) are shown compared against the non-fatty acid
overproducing strain (BTE- ACC-)
Protein name and
gene locus
Gene product/function
(BTE+)/(BTE- ACC-)
late log ratio
mid-stat. ratio
Increased expression (BTE+ ACC-)
PutA
b1014
fused DNA-binding transcriptional regulator / proline dehydrogenase /
pyrroline-5-carboxylate dehydrogenase
YddK b1471
predicted protein
YdhC
b1660
predicted transporter
2.20
-
2.53
-
2.43
Decreased expression (BTE+ ACC-)
OppD b1246
oligopeptide tranporter subunit
GatZ
b2095
D-tagatose 1,6-bisphosphate aldolase 2, subunit
ArnC b2254
undecaprenyl phosphate-L-Ara4FN transferase
FabB
b2323
3-oxoacyl-[acyl-carrier-protein] synthase I
Slp
b3506
outer membrane lipoprotein
-2.36
-
-2.05
-2.56
-∞
∞
-5.85
-2.11
4.18
-
7.39
2.17
4.32
-
2.28
∞
2.04
Increased expression (BTE+ ACC+)
AccA
b0185
acetyl-CoA carboxylase, carboxytransferase, α subunit
MiaB
b0661
tRNA-i(6)A37 methylthiotransferase
PutA
b1014
fused DNA-binding transcriptional regulator / proline dehydrogenase /
pyrroline-5-carboxylate dehydrogenase
PspA
b1304
regulatory protein for phage-shock-protein system
YmgA b1165
connector protein for RcsB regulation of biofilm
TrpD
b1263
fused glutamine amidotransferase (component II) of anthranilate synthase /
anthranilate phosphoribosyl transferase
Trg
b1421
methyl-accepting chemotaxis protein III, ribose and galactose sensor receptor
YddK b1471
predicted protein
CheB
b1883
fused chemotaxis regulator: protein-glutamate methylesterase in two-component
regulatory system with CheA
AccB
b3255
acetyl-CoA carboxylase, BCCP subunit
AccC
b3256
acetyl-CoA carboxylase, biotin carboxylase subunit
FrwD
b3953
predicted enzyme IIB component of PTS
-
2.13
3.90
2.05
3.76
-
30.6
38.8
Decreased expression (BTE+ ACC+)
LeuD
b0071
3-isopropylmalate dehydratase small subunit
LeuC
b0072
3-isopropylmalate dehydratase large subunit
PhoA
b0383
bacterial alkaline phosphatase
Rnk
b0610
regulator of nucleoside diphosphate kinase
LipA
b0628
lipoate synthase
OppB b1244
oligopeptide transporter subunit
OppD b1246
oligopeptide transporter subunit
OppF
b1247
oligopeptide transporter subunit
YciF
b1258
predicted ruberythrin/ferritin-like metal-binding protein
PptA
b1461
4-oxalocrotonate tautomerase
GatZ
b2095
D-tagatose 1,6-bisphosphate aldolase 2, subunit
FbaB
b2097
fructose-bisphosphate aldolase class I
ElaB
b2266
conserved protein
FabB
b2323
3-oxoacyl-[acyl-carrier-protein] synthase I
TalA
b2464
transaldolase A
CysH
b2762
3'-phosphoadenosine 5'-phosphosulfate reductase
YgiM
b3055
SH3 domain protein
DeaD
b3162
ATP-dependent RNA helicase
PstB
b3725
phosphate transporter subunit
IlvC
b3774
ketol-acid reductoisomerase, NAD(P)-binding
GlnA
b3870
glutamine synthetase
Ppc
b3956
phosphoenolpyruvate carboxylase
-2.46
-2.90
-
-2.27
-2.65
-28.3
-2.15
-2.25
-2.91
-2.56
-3.23
-2.25
-∞
-3.91
-2.58
-2.44
-7.00
-2.48
-3.32
-2.32
-4.22
-3.25
-6.03
-2.65
∞
Bolded protein names also were differentially expressed greater than 2.0-fold (P<0.05) as transcripts at any of the three sampling points.
122
Chapter 4: Identification of transport proteins involved in FFA efflux in E. coli
In Chapter 3, it was found that BTE-expressing, FFA overproducing E. coli cultures exhibited
greatly reduced viabilities, increased degrees of cell lysis, an increased population of depolarized
cells, and altered cell morphologies compared to BTE-H204A-expressing, non-FFA
overproducing control cultures. These phenotypes were linked to genotypes via a transcriptomic,
proteomic, and lipidomic analysis, which revealed an elevated unsaturated membrane lipid
content, induction of membrane stress responses, and increased expression of genes involved in
aerobic respiration. It was postulated that improving the rate of secretion of FFAs may alleviate
their intracellular toxicity.
In this chapter, 15 native E. coli genes and operons that were
speculated to be involved in FFA secretion were screened for their impacts on cell viability, lysis,
and FFA titer in a FFA overproducing strain. Inner membrane and periplasmic subunits of
multidrug efflux pumps (AcrAB, EmrAB, MdtEF, MdtABC) were identified as reducing
viability in single and double deletion screens, while the outer membrane component of all the
identified efflux pumps, TolC, appears to be essential for FFA overproduction. While it was not
possible to achieve higher FFA titers when overexpressing the efflux pumps, overexpression of
AcrAB, MdtEF, and MdtEF-TolC increased the minimum inhibitory concentration of decanoate
in an acrAB deletion strain.
4.1 Introduction
Efforts to metabolically engineer microorganisms to overproduce fuels and chemicals can
often induce toxicity and stresses as a direct result of the target compound that lead to reductions
in product yields, titers, and productivities [Nicolaou 2010, Dunlop 2011b]. This constitutes a
123
major challenge toward achieving economical production of these compounds for uses such
as fuels and bulk plastics. Hydrophobic compounds, including those of interest as high-energy
density biofuels such as n-butanol, and straight and branched hydrocarbons, intercalate in the
cytoplasmic membrane [Sikkema 1995]. This can result in altered membrane fluidity, membrane
protein function, and aerobic respiration [Ingram 1976, Ingram 1980, Sikkema 1995, Sardessai
2002]. Recent work has focused on metabolic engineering of E. coli and other organisms for the
production of free fatty acids (reviewed in Chapter 1), which can be catalytically converted to
alkanes (Chapter 2) [Mäki-Arvela 2007], and fatty acid derived products such as ethyl esters and
fatty alcohols (reviewed in Chapter 1). Production of free fatty acids and fatty acid ethyl esters
via fatty acid biosynthesis has thus far been limited to between 12% and 65% of the maximum
theoretical yield (see Table 1.1), which may be a result of toxicity of products and intermediates.
In Chapter 3, we identified physiological perturbations resulting from free fatty acid
overproduction in Escherichia coli that included greatly reduced cell viability, membrane
integrity, large increases in membrane unsaturated fatty acid content, induction of membrane
stress responses, and a loss of proton motive force coupled with increased expression of genes
involved in aerobic respiration (Chapter 3). As a result of these findings, we suggested two
parallel strategies for continued strain engineering for improved yields of free fatty acids. These
included improving the efflux of free fatty acids from the cell, and restoring the membrane
unsaturated fatty acid content to lower levels.
In this work, a list of native E. coli genes that could potentially be involved in fatty acid
export was assembled based on four criteria. First, genes previously observed to be involved in
free fatty acid import were targeted, as a dual role in export could also be possible. This
encompassed three genes: fadL, tolC, and prc; for which a transposon mutagenesis screen of
124
membrane-bound proteins identified defective or absent growth on oleate [Azizan 1994].
Second, a number of drug efflux pumps, all of which associate with the tolC outer membrane
channel, have previously been observed to confer resistance toward both SDS and bile salts
[Sulavik 2001, Nishino 2001], and we postulate that they may also be involved in free fatty acid
export. These include emrAB, acrAB, mdtABCD, acrD, mdtEF, and acrEF. Third, several
annotated multidrug efflux pumps were identified as having increased expression in fatty acid
overproducing strains at any sampling point in one or both experimental data sets (fermentors in
glycerol-supplemented media, or shake flasks with glucose-supplemented media). These include
cmr, mdtD, mdtG, mdtK, emrAB, and mdtEF (Table 4.1). Fourth, the transcriptional activators
marA, rob, and soxS, for which many members of their regulons were upregulated in fatty acid
overproducing strains, were targeted due to their role in regulating drug efflux pumps such as
acrAB [Rosenberg 2003] and mdtG [Fàbrega 2010], the tolC outer membrane channel [Aono
1998b], and the outer membrane porin ompF via antisense repression with micF sRNA
[Andersen 1989]. Specifically, rob was observed in our prior study as having a negative impact
on viable cell counts when deleted (Chapter 3), which is interesting given previous studies
detailing its activation by free fatty acids [Rosenberg 2003] and subsequent induction of acrAB
at sufficiently high fatty acid concentrations [Ma 1995]. As many genes in the MarA/Rob/SoxS
regulons have unknown functions, it may be possible to identify a new protein potentially
involved in fatty acid export by preventing the activation of an entire regulon.
After gene targets were identified, gene and operon deletions were constructed in a free fatty
acid overproducing strain (TY05) containing three copies of a codon-optimized acyl-ACP
thioesterase from Umbellularia californica (BTE) under control of an arabinose-inducible
promoter integrated into chromosomal loci of genes involved in β-oxidation on the chromosome
125
Table 4.1 Genes targeted with possible free fatty secretion or efflux roles.
Locus
Gene/operon
name
Reason for selection
Source or
reference
b3035
tolC
b0462-3
acrAB
Prior role in alleviating FFA toxicity; outer membrane
component for most MFA-type efflux pumps in E. coli;
member of Rob regulon
member of Rob regulon; confers SDS resistance
b2344
fadL
b1830
b2470
b3265-6
b2074-7
prc
acrD
acrEF
mdtABCD
b0842
cmr
Azizan 1994
Aono 1998b
(see text)
Rosenberg 2003,
Nishino 2001
Black 1990
Chapter 3
Azizan 1994
Nishino 2001
Nishino 2001
Nishino 2001
Chapter 3
Chapter 3
b1053
mdtG
b1663
mdtK
b2685-6
emrAB
b3513-4
mdtEF
b1531
marA
b4062
soxS
b4396
rob
b0929
ompF
Necessary role in outer membrane import of long-chain
FFAs; increased expression in microarray data set
Identified role in fatty acid import
Confers SDS resistance
Confers SDS resistance
Confers SDS resistance; increased expression of mdtD
in microarray data set
Efflux pump; increased expression in both microarray
data sets
Efflux pump; member of Rob regulon; increased
expression in both microarray data sets
Efflux pump; decreased expression in one microarray
data set and increased expression in another
Confers SDS resistance; increased expression in one
microarray data set
Confers SDS resistance; increased expression in one
microarray data set
Overlapping regulon with Rob; increased expression in
both microarray data sets
Overlapping regulon with Rob; increased expression in
both microarray data sets
Activation by FFAs; increased expression in one
microarray data set
Outer membrane protein; indirectly in Rob regulon;
strongly decreased expression in both microarray data
sets
Fàbrega 2010
Chapter 3
Chapter 3
Nishino 2001
Chapter 3
Nishino 2001
Chapter 3
Chapter 3
Chapter 3
Rosenberg 2003
Chapter 3
Andersen 1989
Chapter 3
[Youngquist 2012]. The same gene and operon deletions were also constructed in a negative
control strain (TY06) containing three copies of BTE with an active site mutation (BTE-H204A)
that render the protein non-functional (Chapter 2) [Yuan 1996]. The gene deletion strains were
screened for decreases in viability relative to TY05, as well as a lack of decrease in viability
relative to TY06, by plate counts and a flow cytometry assay employing SYTOX green, an inner
membrane impermeable nucleic acid dye [Roth 1997]. This strategy of screening for deleterious
mutations in endogenous FFA overproducing strains was necessitated by previous findings that
126
exogenous addition of lauric acid, the major species that is overproduced, does not elicit the
same toxicity as endogenous production (Chapter 3). However, gene and operon deletions that
resulted in reduced viabilities in BTE-expressing, FFA-overproducing cultures were tested by a
minimum inhibitory concentration (MIC) assay for their impact on viability in the presence of
more toxic, exogenously added FFAs such as octanoic and decanoic acid.
Genes and operons that appeared to be important in maintaining viability under fatty acid
overproducing conditions were cloned into expression vectors and analyzed for their ability to
improve fatty acid production. While all the overexpressed drug efflux pump components
improved tolerance toward SDS in an acrAB deletion strain, none improved tolerance over
strains with intact chromosomal acrAB. Furthermore, none of the drug efflux pumps, when
overexpressed in a strain with intact chromosomal acrAB improved fatty acid production or
viability when BTE was expressed, or improved the MIC toward exogenous octanoic or decanoic
acid.
However, when acrAB was disrupted, the MIC of octanoic and decanoic acid was
increased in all selected drug efflux pumps except mdtABCD.
4.2 Materials and Methods
4.2.1 Chemicals, reagents, enzymes, and oligonucleotide primers
All chemicals were purchased from Fisher Scientific (Pittsburgh, PA) unless otherwise
specified.
Restriction enzymes, PCR enzymes for cloning, and associated reagents were
purchased from New England Biolabs (Ipswich, MA).
Shrimp alkaline phosphatase was
purchased from Fermentas (Glen Burnie, MD). GoTaq Green Master Mix was purchased from
Promega (Madison, WI) and used for colony PCR. All oligonucleotide primers were purchased
127
from Integrated DNA Technologies (Coralville, IA). Plasmid DNA and gel extracted DNA
was purified using the QIAPrep Spin Miniprep and QIAEX II kits (Qiagen, Valencia, CA).
4.2.2 Strain construction
Bacterial strains, plasmids, and oligonucleotide primers used in this study are listed in
Appendices I, II, and III. The primary background strains used in deletion studies in this work
are TY05 and TY06, which are E. coli K-12 MG1655 harboring three integrations in aerobic βoxidation loci of a codon-optimized gene encoding an acyl-ACP thioesterase from Umbellularia
californica or a mutated gene encoding a non-functional thioesterase (BTE-H204A), respectively,
under control of IPTG-inducible trc promoters, as previously described [Youngquist 2012].
These strains are therefore deficient in aerobic β-oxidation and can be maintained without the
addition of antibiotics.
Additional single gene deletions in strains TY05 and TY06 were
constructed by P1 phage transduction of kan cassettes from the Keio collection of gene knockout
mutants [Baba 2006], as previously described (Chapters 2 and 3) [Thomason 2007b]. Deletions
of multiple gene operons (acrAB, mdtEF, emrAB, acrEF, and mdtABCD) were constructed by
amplifying kan cassettes from template plasmid pKD13 [Datsenko 2000] with primers harboring
40-bp 5'-extensions homologous to regions upstream and downstream of the targeted operons
(primers 61-70, Appendix III). These linear cassettes were electroporated into strain DY330, for
which recombinase function was induced by a temperature shift from 30°C to 42°C for 15
minutes prior to being made electrocompetent. Phage P1 lysates were prepared using a modified
liquid procedure [Donath 2011], with cells grown and infected at 30°C. These lysates were then
used to P1 transduce the kan containing loci into strains TY05 and TY06. The kan cassette was
removed using pCP20 [Cherepanov 1995], and the presence of the desired deleted locus and the
128
presence of the BTE or BTE-H204A containing integrations in remaining FRT-site
containing loci (fadD, fadE, fadAB) was confirmed by colony PCR (primers 3-6, 39-42, 53-60,
73-102, Appendix III).
Strain TY05ara was constructed by sequential P1 phage transductions using lysates harboring
Φ(∆araEp kan Pcp8-araE), araFGH::kan, and araBAD::cat loci from strains BW27271,
BW27269, and NRD204, respectively [Khlebnikov 2001]. These chromosomal modifications
allow homogeneous induction with L-arabinose. Antibiotic resistance genes were removed after
each transduction using pCP20, and the presence of all desired FRT-site containing loci were
confirmed by colony PCR using the listed primers. TY05ara acrAB::kan was also constructed by
P1 phage transduction using a lysate derived from TY05 acrAB::kan. The kan locus was not
removed by pCP20 as deleterious recombination events were always observed if greater than 6
FRT sites existed on the chromosome. All loci were re-confirmed by colony PCR of the
transductant.
4.2.3 Plasmid construction
Plasmid pBAD33* is pBAD33 [Guzman 1995] harboring araC with cysteine-280 converted
to a premature stop codon (AraC-C280*). This mutation has been previously observed to reduce
inhibition of gene expression from the PBAD promoter in the presence of IPTG [Lee 2007],
allowing the use of both inducing agents simultaneously. The plasmid was generated by aroundthe-world PCR using primers 103 and 104 (Appendix III) with template pBAD33, which
introduced the C280* mutation and an XhoI restriction site at the 5' and 3' ends. The PCR
product was then digested with XhoI and ligated to form plasmid pBAD33*.
129
Selected genes and operons were amplified by PCR from MG1655 genomic DNA with
their putative upstream ribosome binding sites and added 5' and 3' restriction sites using primers
listed in Appendix III (primers 121-130). These PCR products were subsequently inserted into
pBAD33* between the added restriction sites. The selected operons were also inserted into
pBAD33*-tolC to generate artificial operons with tolC.
4.2.4 Cell cultivation
All strains tested for viability and fatty acid production were grown aerobically in a shaker
(New Brunswick Scientific, Enfield, CT) at 37°C and 250 rpm in 250 mL shake flasks with a 5X
headspace in Difco LB medium (BD, Franklin Lakes, NJ) supplemented with 0.4% v/v glycerol.
Cultures were induced with 1 mM IPTG at OD 0.2 to induce expression of BTE or BTE-H204A.
Strains were grown in biological triplicate from overnight cultures inoculated with independent
colonies on streak plates or fresh plasmid transformation plates.
For strains harboring
transporters cloned in pBAD33-C280* plasmids, chloramphenicol was added to a concentration
of 34 µg/mL and cultures were induced at OD 0.2 with 0.2% w/v L-arabinose in addition to
IPTG as described above.
4.2.5 Cell viability measurements from plate counts
Volumes of cell culture were serially diluted in phosphate buffered saline (PBS) (137 mM
NaCl, 27 mM KCl, 10 mM Na2HPO4, 2 mM KH2PO4, pH 7.4) and spread onto LB agar plates
(containing no antibiotics) at indicated times. Individual colonies were counted after overnight
incubation at 37°C and an additional overnight incubation at room temperature, due to the wide
range of colony sizes observed in BTE-expressing strains after overnight incubation.
130
4.2.6 SYTOX flow cytometry assays
To assess cell permeability, cell pellets were collected 8 h post-inoculation by centrifugation
of 0.1 mL culture samples, and were resuspended in 1 mL of 0.22 µm filter-sterilized PBS,
diluted 1000-fold in 1 mL PBS, and stained by addition of 1 µL of 5 mM SYTOX Green in
DMSO (Invitrogen). Staining proceeded for 10 to 30 min prior to flow cytometric analysis using
a Guava EasyCyte Plus flow cytometer (Millipore, Billerica, MA) with 488 nm excitation and
simultaneous measurement of forward scatter and 525 nm (green) emission on logarithmic scale
photodetectors. Forward scatter with no minimum threshold was selected as the trigger for
events and 5000 events were collected per sample.
Green fluorescence histograms were
constructed by binning logarithmic scale green fluorescence values between 0 and 1000 in
increments of 10, and averaging the number of events per bin between three biological replicates.
Two distinct populations were evident from the green fluorescence histograms, allowing a
logarithmic-scale green fluorescence intensity of 420 to serve as the cut-off between cells
counted as intact (less than or equal to 420) and non-intact (greater than 420).
4.2.7 Fatty acid extraction and analysis
Fatty acids were extracted and methylated from cell cultures as previously described in
Chapters 2. Gas chromatography/mass spectrometry (GC/MS) analysis and peak identification
and quantification was performed on a model 7890 Agilent GC with a model 5975 mass
spectrometer as previously described in Chapters 2 and 3. Average fatty acid concentrations
from biological triplicate cultures were determined.
131
4.2.8 Minimum inhibitory concentration (MIC) assays
Agar plates containing varying concentrations of sodium dodecyl sulfate (SDS) were prepared
by mixing equal volumes of autoclaved 2X YT agar (16 g/L Bacto tryptone, 10 g/L Bacto yeast
extract, 5 g/L sodium chloride, 30 g/L agar, adjusted to pH 7.0 with NaOH) and varying
concentrations of SDS (0, 0.1, 0.5, 1, 5, 10, 15, 25, 50 mg/mL) in sterile water plus final
concentrations of 34 µg/mL chloramphenicol and 0.2% w/v L-arabinose for plasmid
maintenance and induction. Overnight 5 mL cultures of TY05 ∆acrAB harboring pBAD33* or
transporter genes cloned into pBAD33* were grown in LB medium containing 34 µg/mL
chloramphenicol. These cultures were diluted 1:100 in 5 mL LB medium containing 34 µg/mL
chloramphenicol and 0.2% w/v arabinose and grown in a shaker for 4 hours at 37°C and 250 rpm.
Serial 10-fold dilutions were prepared in PBS, and 3.0 µL of 10-4, 10-5, and 10-6 dilutions were
spotted onto plates containing SDS.
Agar plates containing varying concentrations of FFAs were prepared by mixing equal
volumes of 2X LB agar with sterile water containing 10% w/v stocks of free fatty acids in
ethanol, for free fatty acids that are solids at room temperature (decanoic acid and higher chain
lengths). The FFA stock was pre-mixed with 5 mL of 10% Brij 35 per 50 mL of final LB agar
mixture, prior to addition of extra water or 2X LB agar, to assist in dispersion. An equivalent
volume of ethanol was added to 0 g/L FFA plates. For plates containing varying concentrations
of octanoic acid (0, 1, 2, 3, 4, 5, 7.5, and 10 g/L), pure octanoic acid was mixed with 5 mL of
10% Brij 35 per 50 m of final LB agar mixture, without the use of an ethanol stock. Where
applicable, final concentrations of 34 µg/mL chloramphenicol and 0.2% w/v L-arabinose, or only
0.2% w/v L-arabinose, were also added to the plates. Cultures were grown as described above
132
except TY05 and gene deletion strains were grown in unsupplemented LB medium. Strains
harboring plasmids were grown as described for individual experiments in the Results section.
4.3 Results
4.3.1 Viability analysis of single gene/operon deletion strains
The targeted genes and operons listed in Table 4.1 were deleted in E. coli TY05 and TY06,
which are strains that harbor three chromosomal copies of BTE or BTE-H204A, respectively,
under control of trc promoters. The probability of non-specific effects that could result from
deleting proteins involved in antibiotic efflux, or from altered membrane properties as a result of
FFA overproduction, was reduced due to these strains not requiring antibiotics for maintenance
of thioesterase expression. Shake flask cultures of each strain were grown in biological triplicate
in LB medium supplemented with 0.4% glycerol.
At 8 h post-inoculation, viability was
measured by plating serial dilutions of culture and calculating colony forming units (CFUs) per
mL after overnight incubation of the plates, and by a flow cytometry assay employing staining
with SYTOX Green nucleic acid dye, which is impermeable to intact inner membranes. The
portion of total cells in the population (log-scale green fluorescence of 50 or higher) that
exhibited bright green fluorescence (log-scale value of 440 or higher) after staining with SYTOX
Green was considered non-intact (Figure 4.1). While CFU/mL is dependent both on the absolute
number of cells and the percentage of viable and culturable cells, the flow cytometry assay is
independent of the absolute number of cells but also provides separate information on the
number of cells per mL in each culture.
The data collected was used to identify gene deletions that decreased viability by one or
ideally both measures in FFA-overproducing TY05, but that had no impact on viability in non-
133
number of events
600
TY06
TY05
+ipp
non-intact
intact
500
400
300
200
100
0
0
200
400
600
green fluorescence bin
800
1000
Figure 4.1 Determination of threshold for non-intact SYTOX Green stained cells from green
fluorescence histograms. Induced cultures of TY06 and TY05 were stained after 8 h growth, and a
TY05 strain (TY05/pBAD33*) was additionally treated with 25% v/v isopropanol (ipp, green curve) for
10 minutes, to demonstrate justification of a green fluorescence value of 440 as the threshold between
intact and non-intact cells. Error bars represent standard deviations in each histogram bin from three
biological replicate cultures.
FFA-overproducing TY06. The results of the CFU/mL screen were normalized to the number of
events measured in the flow cytometry assay per mL of culture to provide a measure of viable
cells comparable to the SYTOX Green assay (Figure 4.2). For gene deletions in strain TY06, in
general the ratio of CFU to flow cytometer events was close to 1.0, indicating close to 100%
viability. In contrast, TY05 exhibited a ratio of 0.43 ± 0.16, with a few deletions exhibiting
statistically significant (P < 0.05) reduced ratios. These were TY05 ∆acrAB (ratio of 0.11 ± 0.03,
P < 0.01), TY05 ∆rob (ratio of 0.13 ± 0.10, P < 0.01), TY05 ∆mdtABCD (ratio of 0.20 ± 0.07, P
< 0.05), and TY05 ∆emrAB (ratio of 0.16 ± 0.02, P < 0.05). Notably, these same deletions in
TY06 exhibited no statistically significant deviation from the CFU to number of events ratio
compared to strain TY06, indicating that the observed reductions in viability were specific to the
condition of fatty acid overproduction. A few additional gene deletions in TY05 exhibited
reduced average viabilities compared to TY05 but with 0.15 > P > 0.10. These were TY05
134
∆mdtEF (ratio of 0.27 ± 0.09, P = 0.13) and TY05 ∆prc (ratio of 0.27 ± 0.03, P = 0.11). The
latter strain had approximately equivalent plate counts to TY05, but smaller cell size and higher
event/mL counts in both TY06 ∆prc and TY05 ∆prc. Furthermore, it should be noted that TY05
∆tolC exhibited greatly reduced CFU/mL counts compared with TY05, but also highly reduced
flow cytometry counts and OD600 at 8 hours. No reduction in CFU/mL counts, flow cytometry
counts, or OD600 was observed for TY06 ∆tolC compared to TY06.
The results of the SYTOX Green flow cytometry screen were qualitatively very similar to
the normalized CFUs (Figure 4.2). All TY06 background strains were nearly 100% intact.
Conversely, TY05 background strains were less than 50% intact and showed variability that
mirrored the data collected in the CFU/mL screen. The correlation between normalized CFUs
and percent intact by SYTOX Green is further supported by scatter plots (Figure 4.3), where a
linear fit to the points produces a R2 value of 0.0016 for TY06 strains, and 0.6984 for TY05
strains (excluding TY05 ∆tolC for reasons described below). Compared to TY05 (33.8 ± 11.4
percent), TY05 ∆rob (15.0 ± 6.4 percent) and TY05 ∆acrAB (10.9 ± 1.6 percent) exhibited
statistically significant (P < 0.05) reductions in percentage of intact cells. A few additional gene
deletions in TY05 exhibited reduced intact percentages compared to TY05 but with 0.15 > P >
0.10, including TY05 ∆mdtABCD (22.6 ± 5.1 percent, P = 0.12) and TY05 ∆emrAB (22.4 ± 4.5
percent, P = 0.11). Percent intact values were not calculated for tolC deletion strains due to
shifted green fluorescence histograms compared to all other strains. Analysis of forward scatter
values by flow cytometry indicated a larger average cell size for TY05 ∆tolC than other TY05
strains, with TY06 ∆tolC exhibiting similar forward scatter histograms as other TY06 strains
(data not shown). As the outer membrane is permeable to SYTOX Green, a combination of
135
changes in cell size and/or defective efflux of SYTOX Green from the periplasmic space may
be responsible for the shifted green fluorescence histograms.
In all, a total of 4 gene deletions in TY05 resulted in both reduced CFU/flow cytometry event
ratios and reduced percent intact cells compared to TY05. These were TY05 ∆rob, TY05
∆acrAB, TY05 ∆emrAB, and TY05 ∆mdtABCD. Rob is a transcription factor that is known to
activate its regulon upon activation by free fatty acids [Ma 1995, Rosenberg 2003], and deletion
was previously observed to cause a reduction in CFU/mL in a strain of E. coli (K-12 MG1655
1.2
P=
0.09
**
1.0
P=
0.07
*
0.8
0.6
100
P=
0.12
80
percent intact
CFU/flow cytometry events
1.4
60
40
0.4
20
0.2
0
70
0.6
60
0.5
0.4
P=
0.11
*
0.3
0.2
0.1
TY
05
0.0
P=
0.13
**
**
*
percent intact
0.7
50
P=0.12
P=0.11
40
30
20
*
*
10
0
TY
05
CFU/flow cytometer events
TY
06
TY
06
0.0
Figure 4.2 Normalized CFUs and percent intact cells for single gene deletions in strains TY06 and
TY05. (top) values for strain TY06 (BTE-H204A-expressing), in black; (bottom) values for strain TY05
(BTE-expressing), in red; (left) normalized CFUs (CFU/mL from plate counts divided by cell/mL counts
by flow cytometry); (right) percent intact cells measured by SYTOX Green flow cytometry assay. Error
bars are propagated standard errors using cell counts (for percent intact, the number of cells with green
fluorescence >50 for total cells, and the number of cells with green fluorescence >440 for non-intact cells).
* = P-value < 0.05 compared against TY06 or TY05. ** = P-value < 0.01 compared against TY06 or
TY05.
136
∆fadD ∆araBAD ∆rob) expressing BTE on a plasmid (Chapter 3) compared to the same
strain without deletion of rob. Similarly, deletion of marA and soxS, which encode transcription
factors with regulons overlapping that of Rob but that are activated by different mechanisms,
previously showed no significant reduction in CFU/mL upon deletion (Chapter 3), with the same
effect observed here. Members of the Rob regulon include acrAB and tolC. AcrAB, EmrAB,
and MdtABC all encode inner membrane and periplasmic linker protein components of drug
efflux pumps that interact with an outer membrane component, TolC [Fralick 1996, BorgesWalmsley 2003, Nagakubo 2002]. MdtD is an uncharacterized putative inner membrane protein
that may function as a separate drug efflux system [Nagakubo 2002].
0.7
1.5
2
R = 0.6984
2
R = 0.0016
0.6
CFU/flow cytometer events
CFU/flow cytometer events
1.4
1.3
1.2
1.1
1
0.9
0.8
0.7
0.5
0.4
0.3
0.2
0.1
0.6
0.5
0
86
88
90
92
94
96
98
100
102
104
106
0
10
20
percent intact
30
40
50
60
percent intact
Figure 4.3 Scatter plots of normalized CFUs versus percent intact cells by SYTOX Green staining.
(Left) TY06 background strain data points with linear fit having R2 value of 0.0016. (Right) TY05
background strain data points with linear fit having R2 value of 0.6984.
4.3.2 Fatty acid titers of single gene/operon deletion strains
Fatty acid titers were analyzed at 8 h and 24 h post-inoculation from separate sets of
biological triplicate cultures, due to the destructive conditions required to collapse the foam in
order to obtain accurate volumetric titers. While it would have been anticipated that an increased
percentage of non-intact cells would result in lower fatty acid titers, statistically equivalent titers
137
were observed between strain TY05 and all gene deletions for which reduced viability was
observed (Figure 4.4).
Interestingly, free fatty acid production at both 8 h and 24 h in TY05 ∆tolC was nearly
abolished (Figure 4.4). After 8 h, TY05 produced 428 ± 29 mg/L of C8-C14 fatty acids, while
TY05 ∆tolC produced 12 ± 1 mg/L. At 24 h, TY05 produced a total of 740 ± 70 mg/L of C8-C14
fatty acids while TY05 ∆tolC produced 11 ± 2 mg/L. At 24 h, the composition of fatty acids in
TY06 ∆tolC was very similar to that of TY06, with total titers of predominantly C16-C18 fatty
acids of 163 ± 7 and 149 ± 4 mg/L in TY06 and TY06 ∆tolC, respectively. It is evident that the
drastic change in phenotype upon tolC deletion is specific to conditions of free fatty acid
overproduction, and that tolC appears to be required for FFA overproduction. While TolC is
known for its role in facilitating efflux of a large range of molecules through the outer membrane,
it was also previously observed to be essential for supporting growth of E. coli when oleate is
1000
700
total fatty acids (mg/L)
total fatty acids (mg/L)
800
600
500
400
300
200
TY06
TY05
800
600
400
200
100
0
TY
05
TY
05
0
TY06/TY05
∆tolC
strain
Figure 4.4 Total fatty acid titers in TY05 deletion strains after 8 h, and in baseline and ∆tolC strains
after 24 h. (Left) Fatty acids were extracted from TY05 cultures grown for 8 h in LB + 0.4% glycerol in
two separate experiments (the results of TY05 are shown for each experiment). In general no significant
differences were observed except for TY05 ∆tolC. (Right) Fatty acids were extracted from cultures grown
24 h in LB + 0.4% glycerol at 37°C. While deletion of tolC in TY06 (non-FFA overproducing) had
virtually no effect, deletion of tolC in TY05 (FFA overproducing) nearly completely abolished fatty acid
production, particularly of C8-C14 species.
138
supplemented as a sole carbon source [Azizan 1994].
Its role in free fatty acid
overproduction cannot be limited in the present study to that of FFA efflux, as TolC has been
implicated in other roles such as altered NAD+/NADH ratios, activation of the phage shock
regulon [Dhamdhere 2010], export of siderophores, efflux of cysteine, and acid tolerance
[reviewed in Zgurskaya 2011].
4.3.3 MIC of exogenous FFAs in single gene/operon deletion strains
To further confirm the role of the identified genes in conferring resistance to free fatty acids,
deletion strains in TY05 were plated under non-inducing conditions (no added IPTG) on LB agar
containing varying concentrations of octanoic and decanoic acid. The pH was adjusted to 7 in all
plates by addition of equimolar amounts of NaOH, and it was confirmed that the maximum
concentration of Na+ present was not growth inhibitory toward TY05 or TY05 ∆acrAB in a plate
containing NaCl (data not shown). Dodecanoic and tetradecanoic acids were non-inhibitory at
plate concentrations of 2 g/L, which was also above their solubility limits. We have previously
observed minimal toxicity of 0.5 g/L dodecanoic acid added to cultures and have postulated that
endogenously produced dodecanoic acid, or the mixture of fatty acids resulting from expression
of BTE, exhibits a higher degree of toxicity (Chapter 3). In contrast, sodium octanoate and
decanoate appear soluble at concentrations of up to 10 g/L and 5 g/L, respectively, and both elicit
growth inhibition in strain TY05 at concentrations below these apparent solubilities. Saturated
overnight cultures of TY05, TY05 ∆acrAB, TY05 ∆emrAB, TY05 ∆mdtEF, TY05 ∆mdtABCD,
TY05 ∆tolC, and TY05 ∆rob were diluted 1:100 in 5 mL of LB agar in glass tubes and incubated
at 37°C for 4 h with 250 rpm shaking. At this time, all cultures had an OD600 between 2.5 to 2.9.
Samples from each culture were serially diluted in PBS and 3 µL of 104, 105, and 106-fold
139
dilutions were spotted on plates containing varying concentrations of octanoate and
decanoate (Figure 4.5). After overnight incubation at 37°C, growth of TY05 was observed up to
5 g/L octanoate and 4 g/L decanoate. Growth was observed for TY05 ∆acrAB only for 0 g/L
decanoate (no growth at 0.5 g/L) and up to 3 g/L octanoate, while TY05 ∆rob grew on up to 3
g/L decanoate and 4 g/L octanoate. Growth was observed of TY05 ∆mdtABCD up to 4 g/L
decanoate (same as TY05) but only up to 4 g/L octanoate. TY05 ∆tolC was the most inhibited,
Figure 4.5 MIC assay for octanoate and decanoate against TY05 and selected single deletions in
TY05. (top) TY05 exhibits visible growth up to 5 g/L octanoate. Deletions in acrAB, rob, mdtABCD,
and tolC resulted in reduced MICs of 3, 4, 4, and 0 g/L octanoate, respectively. (bottom) TY05 exhibits
visible growth up to 4 g/L decanoate. Deletions in acrAB, rob, and tolC resulted in reduced MICs of 0.5,
3, and 0 g/L decanoate, respectively.
140
showing greatly reduced growth on plates containing 0 g/L octanoate and no growth on 0 g/L
decanoate, despite the similar OD600 to all other strains grown in liquid LB medium. Presumably,
the presence of either Brij-35 or ethanol was responsible for the inhibition of growth on plates.
However, while some growth was observed for TY05 ∆tolC on 0 g/L octanoate, no growth was
observed for 1 g/L or any higher concentration, indicative of a fatty acid-specific effect. All
other strains tested had similar MICs toward octanoate and decanoate as background strain TY05.
4.3.4 Viability analysis of double transporter gene/operon deletions
The activities of many efflux pumps are masked by the basal activity of the AcrAB-TolC
complex, which confers resistance to the widest characterized range of compounds [Sulavik
2001, Nishino 2001]. AcrAB also is responsible for conferring resistance to sodium dodecyl
sulfate (SDS) at concentrations exceeding 50 mg/mL in strain TY05, while the minimum
inhibitory concentration (MIC) for TY05 ∆acrAB is less than 0.1 mg/mL (see Figure 4.8 in
section 4.3.6). The structural similarity of SDS and lauric acid, and the loss of viability observed
under conditions of fatty acid overproduction in TY05 ∆acrAB implicate AcrAB as a transporter
of FFAs while also suggesting that AcrAB may be masking the full effects of other identified
efflux pumps, all of which also confer varying degrees of resistance against SDS [Sulavik 2001,
Nishino 2001] and possibly FFAs. To determine if additional efflux pumps were involved,
deletions of additional genes were constructed in TY05 ∆acrAB. Deletions in emrAB, mdtEF,
and mdtABCD were selected based on reduced viability compared to TY05 but with lower
statistical confidence than TY05 ∆acrAB. Additional deletions in acrEF and acrD were selected
for their known capability of conferring SDS resistance when overexpressed in an acrAB
141
deletion strain [Nishino 2001], despite that they were not significant hits in the single
deletion screen.
The strains were tested as described above by determining CFU/mL and percent intact cells
by flow cytometry after staining with SYTOX Green. Normalized CFUs and percent intact cells
by SYTOX Green assay are shown in Figure 4.6. Biological triplicates of TY05 ∆acrAB were
independently run as negative controls at the same time as the double gene/operon deletion
strains. None of the double deletion strains exhibited reduced normalized CFUs or percent intact
cells relative to TY05 ∆acrAB with P < 0.05, however TY05 ∆acrAB ∆mdtEF was lower by both
measures with a P-value of between 0.14 to 0.16.
Interestingly, TY05 ∆acrAB ∆emrAB exhibited a very different phenotype from either single
deletion strain, exhibiting higher CFU/mL, higher CFUs normalized to flow cytometer events,
and higher percent intact cells.
The same two deletions in TY06 produced no detectable
differences in any measure from TY06 or either single deletion in TY06. Growth of TY05
∆acrAB ∆emrAB stalled at a lower final OD600, both in shake flasks at the 8 h sampling point
(7.7 ± 0.4 for TY05 ∆acrAB, 5.6 ± 0.5 for TY05 ∆acrAB ∆emrAB) and in a plate reader growth
curve compared to other negative control strains (Figure 4.7).
4.3.5 Fatty acid titers of double transporter gene/operon deletions
Fatty acid titers were analyzed at 8 h (in TY05 strains only) and 24 h post-inoculation for all
double deletions (Figure 4.8). Double deletions in strain TY06 all exhibited similar fatty acid
titers and profiles as in TY06 and single deletions in TY06, ranging between 145 and 170 mg/L
at 24 h. At 8 h, similar titers were observed for TY05, TY05 ∆acrAB, and all double deletions
(~500 mg/L) except TY05 ∆acrAB ∆emrAB, where the titer was significantly reduced (177 ± 26
142
1.0
CFU/flow cytometry events
A
BTE-H204A
BTE
0.8
0.6
0.4
0.2
TY
0
5
0.0
100
B
BTE-H204A
BTE
90
80
percent intact
70
60
50
40
30
20
10
TY
05
0
Figure 4.6 Normalized CFUs and percent intact cells for double efflux pump deletions in strain
TY06 and TY05. Cultures were sampled 8 h post-inoculation. (A) Normalized CFUs (CFU/mL from
plate counts divided by cells/mL flow cytometry counts). (B) Percent intact cells by SYTOX Green flow
cytometry assay. Only double deletions were tested in TY06. TY05 and TY05 ∆acrAB were tested again
as negative controls.
mg/L). The majority of the reduction in fatty acid titer was in C8-C14 fatty acids. After 24 h, the
titer of TY05 ∆acrAB ∆emrAB increased significantly (490 ± 16 mg/L) but remained less than
the titers of TY05 and TY05 ∆acrAB (~700 mg/L).
decreased titer at 24 h of 638 ± 21 mg/L.
TY05 ∆mdtEF exhibited a slightly
143
1.2
TY05
TY05 ∆acrAB
TY06 ∆acrAB ∆emrAB
TY05 ∆acrAB ∆emrAB
1
OD600
0.8
0.6
0.4
0.2
0
0
2
4
6
8
10
time (h)
Figure 4.7 Plate reader growth curves of acrAB emrAB double deletion strains and negative control
strains. Biological triplicate cultures were grown in 96-well plates in LB + 0.4% glycerol with shaking at
37°C and induced after 2 h with 1 mM IPTG. TY05 ∆acrAB ∆emrAB exhibited a reduced OD600 at 8 h
(marked with vertical line), the sampling time at which CFU/mL and SYTOX Green staining was
performed from shake flask cultures.
4.3.6 Functional validation of drug efflux pump overexpression constructs
All identified gene/operon deletions that exhibited reduced viabilities encode for TolCassociated multidrug efflux pumps that have previously been shown to confer resistance to SDS
[Sulavik 2001, Nishino 2001]. While wild-type E. coli exhibits an MIC toward SDS of greater
than 12.8 mg/mL, deletions in acrAB reduce the MIC to between 0.05 and 0.1 mg/mL [Sulavik
2001, Nishino 2001]. Overexpression of acrAB, emrAB, mdtEF, and mdtABCD on high copy
plasmids in an acrAB deletion strain increases the MIC toward SDS from 0.05 mg/mL to greater
than 0.4 mg/mL, 0.1 mg/mL, 0.2 mg/mL, and 0.2 mg/mL, respectively [Nishino 2001].
Therefore an increase in MIC of SDS was used to validate the functional expression of multidrug
fflux pumps cloned into pBAD33*. As previously observed, strain TY05 (with intact acrAB)
harboring empty vector pBAD33* exhibited no inhibition of growth on plates containing up to
50 mg/mL of SDS, beyond the aqueous solubility limit (data not shown). Also in accordance
144
TY05 8 h
TY06 24 h
TY05 24 h
800
total fatty acids (mg/L)
700
600
500
400
300
200
100
0
TY05
∆acrAB
∆acrAB
∆emrAB
∆acrAB ∆acrAB ∆acrAB
∆mdtEF ∆mdtABCD ∆acrEF
∆acrAB
∆acrD
Figure 4.8 Total fatty acid titers for double efflux pump deletions in TY05 and TY06. Fatty acids
were sampled 8 h and 24 h post-inoculation. TY05 ∆acrAB ∆emrAB exhibits greatly reduced fatty acid
production (primarily reduced C8-C14) relative to other TY05 strains after 8 h. Low titers relative to other
TY05 strains persist in TY05 ∆acrAB ∆emrAB after 24 h.
with prior literature, strain TY05 ∆acrAB harboring pBAD33* exhibited an MIC of less than 0.1
mg/mL. Complementation of TY05 ∆acrAB with pBAD33*-acrAB fully restored the MIC to
greater than 50 mg/mL SDS (Figure 4.9). Complementation of TY05 ∆acrAB with pBAD33*emrAB, pBAD33*-mdtEF, and pBAD33*-mdtABCD restored the MIC to less than 0.1 mg/mL,
0.1 mg/mL, and 0.1 mg/mL after one night incubation at 37°C, respectively (Figure 4.9). After
two nights incubation at 37°C, the MICs were 0.1 mg/mL, between 0.1 to 0.5 mg/mL, and 0.1
mg/mL, respectively. Functional overexpression was therefore validated in all four constructs.
While resistance can be conferred without overexpression of tolC, encoding the outer
membrane component of each drug efflux pump, it is not known whether additional
overexpression of tolC can improve observed MICs and function of overexpressed inner
145
membrane and periplasmic efflux pump components. Thus each drug efflux pump was also
cloned in an artificial operon with tolC harboring its native ribosome binding site. Interestingly,
TY05 ∆acrAB harboring pBAD33*-acrAB-tolC completely lost the resistance to SDS observed
with pBAD33*-acrAB, with an MIC of less than 0.1 mg/mL despite robust growth in LB
containing chloramphenicol and L-arabinose (Figure 4.9). However, pBAD33*-emrAB-tolC and
pBAD33*-mdtABCD-tolC conferred equivalent MICs of SDS as the non-tolC containing
plasmids.
Only pBAD33*-mdtEF-tolC exhibited an improved MIC over pBAD33*-mdtEF
alone, with MIC increasing from between 0.1 to 0.5 mg/mL to 1.0 mg/mL after two nights
incubation at 37°C (Figure 4.9).
As a result, pBAD33*-mdtEF-tolC was the only tolC
expressing construct selected to go forward with FFA plate MIC assays and overexpression in
endogenous FFA overproducing strains.
4.3.7 Effects of drug efflux pump overexpression in FFA overproducing strains
To determine if increased levels of expression of genes identified in deletion screens could
improve endogenous production of FFAs, the efflux pump constructs selected above were tested
in FFA-overproducing strain TY05ara. In preliminary tests, TY05ara and TY05ara acrAB::kan
harboring either pBAD33* as a negative control or pBAD33*-acrAB were grown in shake flasks
and induced at an OD600 of 0.2 with 1 mM IPTG and 0.2% w/v L-arabinose. At 8 h postinoculation, samples were taken for SYTOX Green staining and subsequent flow cytometry
analysis. Samples of cultures were taken for fatty acid analysis at 8 h and 24 h post-inoculation.
Reduced growth was observed in TY05ara acrAB::kan harboring pBAD33*, presumably due to
enhanced chloramphenicol sensitivity despite expression of a chloramphenicol acetyltransferase.
This has previously been observed in acrAB deletion strains [Potrykus 2003], and could be due
146
A
B
Figure 4.9 MIC assay for SDS against TY05 ∆acrAB overexpressing selected efflux pump system
components. TY05 ∆acrAB harbored empty vector as a negative control, or overexpressed inner
membrane/periplasmic linker components encoded in operons by themselves or in artificial operons with
tolC. (A) Overexpression of acrAB restores growth to TY05 ∆acrAB up to 50 mg/mL SDS, but
overexpression of acrAB-tolC confers no resistance. (B) After 24 h, overexpression of mdtEF, mdtEFtolC, mdtABCD, and mdtABCD-tolC confer resistance to 0.1 mg/mL SDS in TY05 ∆acrAB. After 48 h,
growth is observed at 0.1 mg/mL for overexpression of emrAB, emrAB-tolC, mdtEF, mdtABCD, and
mdtABCD-tolC, and up to 1 mg/mL for mdtEF-tolC.
to inefficient efflux of the acetylated chloramphenicol product, which may also possess some
toxicity. Another possibility is decreased efflux of chloramphenicol relative to the activity of
chloramphenicol acetyltransferase, rendering the cells sensitive to a normally non-toxic
147
concentration. At 8 h post-inoculation, TY05ara acrAB::kan/pBAD33* had a larger overall
cell size as determined by forward scatter flow cytometry analysis, and a shifted green
fluorescence histogram from SYTOX Green staining (excluded from the analysis but most likely
representing a high proportion of intact cells shifted toward higher green fluorescence values due
to an increased degree of cell surface staining). TY05ara/pBAD33* and TY05ara/pBAD33*acrAB exhibited nearly identical green fluorescence and forward scatter histograms, with 51 ± 10
and 46 ± 14 percent intact cells, respectively (Figure 4.10). The phenotype of the TY05ara
strains was recovered in TY05ara acrAB::kan/pBAD33*-acrAB, indicating successful
complementation of acrAB but with no overall reduction in cell lysis (29 ± 9 percent intact).
Fatty acid titers were also not improved at either 8 h or 24 h in either acrAB overexpressing
strain relative to TY05ara/pBAD33* (TY05ara acrAB::kan/pBAD33* had a greatly reduced fatty
acid titer at 8 h which was recovered to the same level as TY05ara/pBAD33* after 24 h, data not
shown), and were in fact slightly reduced.
We have observed nearly complete inner membrane depolarization in LB medium containing
either 0.4% or 0.8% glycerol at 3 h post-inoculation (mid-log phase) using a flow cytometry
assay with the membrane polarization probe diethyloxacarbocyanine (DiOC2), and full
depolarization at 8 h post-inoculation (data not shown). As all of the tested multi-drug efflux
pumps in this study are proton antiporters for which export of FFAs would likely be dependent
on a polarized inner membrane, further experiments were conducted with supplementation of 0.5
mM Ca2+ and 0.5 mM Mg2+ into the medium. It has previously been noted that LB medium is
likely divalent cation limited [Nikaido 2009], and these species are necessary to stabilize outer
membrane lipopolysaccharides [Nikaido 1996]. TY05ara harboring pBAD33*, -acrAB, -emrAB,
-mdtEF, -mdtABCD, and -mdtEF-tolC were grown in LB supplemented with 0.4% glycerol, 34
148
70
60
percent intact
50
40
30
20
10
0
TY05ara/
pBAD33*
TY05ara/
pBAD33*acrAB
TY05ara
acrAB::kan /
pBAD33*-acrAB
Figure 4.10 Percent intact cells in TY05ara and TY05ara acrAB::kan overexpressing acrAB on a
plasmid. Cultures of TY05ara and TY05ara acrAB::kan harboring either empty vector pBAD33* or
pBAD33*-acrAB were grown in LB + 0.4% glycerol + 34 µg/mL chloramphenicol and sampled for
SYTOX Green staining after 8 h. Percent intact cells could not be calculated due to shifting of green
fluorescence histograms from apparent growth defects in TY05ara acrAB::kan/pBAD33* (see text).
µg/mL chloramphenicol, 0.5 mM CaCl2, and 0.5 mM MgSO4. After 3 h, over 70% of cells from
all cultures except for those overexpressing mdtABCD (approximately 50%) appeared polarized,
and polarized populations persisted up to at least 8 h (Figure 4.11). A much greater percentage
of cells appeared intact by SYTOX Green staining after 8 h in all cultures than in the absence of
added divalent cations, except those overexpressing mdtABCD (Figure 4.11). Reductions in
percent intact cells were also observed in cultures overexpressing mdtEF-tolC. No improvement
in percent intact cells was observed as a result of overexpressing any of the tested drug efflux
pumps. Furthermore, no improvement in total or C8-C14 fatty acid titers were observed after 8 or
24 h in TY05ara overexpressing drug efflux pumps relative to TY05ara harboring empty vector
(Figure 4.12). A reduction in C8-C14 fatty acid titer was observed as a result of overexpressing
mdtABCD.
149
100
A
3h
90
8h
120
B
100
70
60
percent intact
percent polarized
80
**
50
40
30
20
**
80
60
**
40
20
10
lC
EF
-to
CD
m
dt
m
dt
AB
dt
EF
m
em
rA
B
AB
ac
r
pB
AD
33
*
lC
D
dt
EF
- to
m
dt
AB
C
m
m
dt
EF
em
rA
B
AB
0
ac
r
pB
AD
33
*
0
Figure 4.11 Percent polarized and percent intact cells in TY05ara overexpressing selected efflux
pumps. Cultures of TY05ara harboring either empty vector pBAD33* or selected overexpressed efflux
pumps in pBAD33* were grown in LB + 0.4% glycerol + 34 µg/mL chloramphenicol + 0.5 mM each
MgSO4 and CaCl2. (A) Percent polarized cells determined by a DiOC2 flow cytometry assay for cells
sampled 3 and 8 h post-inoculation. (B) Percent intact cells determined by SYTOX Green staining.
4.3.8 MICs of exogenous free fatty acids toward strains overexpressing drug
efflux pumps
To determine if overexpression of identified drug efflux pumps conferred resistance toward
exogenously added medium-chain length FFAs, we determined the MIC of octanoate and
decanoate against TY05ara and TY05ara acrAB::kan harboring pBAD33*, -acrAB, -emrAB, mdtEF, -mdtABCD, and -mdtEF-tolC in the absence of BTE induction (Figure 4.13)
Strains
overexpressing efflux pumps were not able to increase the MIC of octanoate or decanoate above
that of TY05ara/pBAD33* (which grew on up to 5 g/L octanoate and 5 g/L decanoate),
consistent with prior endogenous FFA production results. In contrast, TY05ara acrAB::kan/
pBAD33* exhibited growth only up to 3 g/L octanoate and 0.5 g/L decanoate. Two efflux
pumps (AcrAB and MdtEF-TolC) were able to complement the acrAB deletion and increase the
150
A
B
C8-C14
C16-C18
600
500
total fatty acids (mg/L)
500
total fatty acids (mg/L)
C8-C14
C16-C18
600
400
300
200
400
300
200
100
100
0
pBAD33*
pBAD33*acrAB
pBAD33*- pBAD33*- pBAD33*- pBAD33*emrAB
mdtEF
mdtABCD mdtEF-tolC
0
pBAD33*
pBAD33*acrAB
pBAD33*- pBAD33*- pBAD33*- pBAD33*emrAB
mdtEF
mdtABCD mdtEF-tolC
Figure 4.12 Fatty acid titers from TY05ara overexpressing selected efflux pumps. (A) C8-C14 and
C16-C18 fatty acid titers 8 h post-inoculation. (B) Fatty acid titers 24 h post-inoculation. No improvements
were observed over the negative control strain TY05ara/pBAD33*. Reduced C8-C14 titers were evident
from overexpression of MdtABCD after 8 h.
MIC of octanoate to 5 g/L and 4 g/L, respectively. When challenged with decanoate, no efflux
pumps restored the baseline MIC against TY05ara/pBAD33*, but most of the selected
overexpressed
efflux
pumps
enabled
growth
above
the
MIC
against
TY05ara
acrAB::kan/pBAD33*. Overexpression of AcrAB, EmrAB, MdtEF, and MdtEF-TolC conferred
resistance toward up to 3 g/L, 2 g/L, 1 g/L, and 1 g/L decanoate, respectively. In summary, the
native level of AcrAB activity is sufficient to confer resistance to exogenous medium-chain
length FFAs. Overexpression of AcrAB in acrAB deletion strains only restores the wild-type
tolerance, at best, but does not improve it further. A number of drug efflux pumps (specifically
EmrAB, MdtEF, and MdtEF-TolC), when overexpressed can partially complement the reduced
tolerance resulting from deletion of acrAB, but also do not improve tolerance in strains either
possessing or lacking chromosomal acrAB.
151
Figure 4.13 MIC assay for octanoate and decanoate against TY05ara and TY05ara acrAB::kan
overexpressing selected efflux pumps. (Left) TY05ara (top) and TY05ara acrAB::kan (bottom)
harboring drug efflux pumps and plated on varying concentrations of octanoate. No overexpressed efflux
pump increases the MIC of octanoate in TY05ara. In TY05ara acrAB::kan, overexpression of AcrAB
and MdtEF-TolC increase the MIC (see text). (Right) The same strains plated on varying concentrations
of decanoate. No overexpressed efflux pump increases the MIC of decanoate in TY05ara. In TY05ara
acrAB::kan, overexpression of AcrAB, EmrAB, MdtEF, and MdtEF-TolC increase the MIC (see text).
4.4 Discussion
Many renewable biochemicals and biofuels are toxic to the host organism at the high
concentrations required by economically feasible bioprocesses [Nicolaou 2010, Shen 2011,
Dunlop 2011b]. In particular, endogenous production of biofuels and chemicals may result in
toxic effects at concentrations lower than observed from exogenous addition of the same
compound.
When produced inside the cell, molecules must traverse the inner and outer
membranes. Therefore the cell cannot effectively use the same strategies, such as reduced
permeability of the outer membrane through decreased porin expression, lipopolysaccharide
modifications, or other modifications to the lipid bilayer that are effective in combating exposure
to exogeneous agents. This phenomenon has been observed for lauric acid, as endogenous
152
production via BTE overexpression results in greatly reduced measures of viability and
increased cell lysis relative to both exogenous addition of lauric acid to the growth medium
(Chapter 3) and addition of lauric acid to plates at concentrations at or above the titers present in
overproducing strains (however it should be noted that the precise mixture of FFAs generated by
BTE expression cannot be replicated by exogenous addition due to the presence of commercially
unavailable and unidentified species).
It has been suggested that microbial efflux pumps, which have been shown to confer
resistance to a wide range of antibiotics, solvents, and cationic or lipophilic compounds [eg.
Sikkema 1995, Brown 2001, Nishino 2001, Pos 2009], could be utilized in production hosts to
improve secretion of the endogenously produced compound, while also conferring resistance to
high levels of the compound as it accumulates in the fermentation broth [Ramos 2002, Nicolaou
2010, Dunlop 2011b].
Furthermore, it has been suggested that keeping intracellular
concentrations of the final product at low concentrations could serve to reduce product inhibition
and improve the flux through reversible reactions [Dunlop 2011b]. A number of efflux pumps
have been isolated from E. coli, and numerous Pseudomonads which confer increased resistance
to toxic solvents such as toluene and hexane [eg. White 1997, Ramos 1998, Aono 1998a].
Competition assays have also identified drug efflux pumps from libraries that are enriched
following a challenge from a toxic level of exogenously added hydrophobic compound [Dunlop
2011a].
These strategies are difficult to employ when exogenous addition of the target
compound does not elicit significant growth inhibition, as is the case for saturated C12 and higher
chain length FFAs.
Additionally, the aforementioned competition assay and other studies
performed in E. coli have necessitated deletion of acrAB to render the cells sensitive to most
compounds of interest [Nishino 2001, Dunlop 2011a].
In one of these studies, a plasmid
153
encoding AcrB was found to be highly enriched in competition assays against five advanced
biofuels [Dunlop 2011a].
We surmised that E. coli already possesses the ability to secrete FFAs via native drug efflux
pumps or other transport machinery, and sought to identify the responsible genes in this study
such that they could be tested as targets for overexpression to allow for improved production of
FFAs. The lines of evidence in support of this assumption were (1) that the native enteric
environment of E. coli is rich in fatty acids [Rosenberg 2003]; (2) that a number of drug efflux
pumps, when overexpressed in an acrAB deletion strain, confer an increased MIC of SDS
[Nishino 2001], which is structurally very similar to FFAs, and that the sensitivity toward SDS is
striking in an acrAB deletion strain (Results; MIC increased from less than 0.1 mg/mL to greater
than 50 mg/mL); and (3) that the transcription factor Rob is activated by 5 mM decanoate, which
leads to increased transcription of acrAB [Ma 1995, Rosenberg 2003], and that we had also
observed upregulation of the overlapping MarA/Rob/SoxS regulons in BTE-expressing cultures
but only observed decreased measures of viability when rob was deleted (Chapter 3). To date
only two drug efflux pump systems, EmhABC in Pseudomonas fluorescens cLP6a [Adebusuyi
2011] and FarAB in Neisseria gonorrhoeae [Lee 1999] have been implicated in FFA export
[Adebusuyi 2011]. Interestingly, the FarAB system was first identified based on homology to E.
coli EmrAB [Lee 1999]. Multidrug efflux systems have previously been proposed to have
alternative physiological roles such as membrane lipid turnover [Poole 2008]. This may be
another rationale for the observed activity in E. coli, and could be related to the induction of
many efflux systems under cell envelope stress conditions, as these conditions may necessitate
active membrane lipid remodeling.
154
Since selections or screens could not be performed using exogenously added lauric acid, a
screen was devised based on the observation that FFA overproduction causes a significant
increase in non-intact cells as indicated by SYTOX Green staining during early stationary phase
(Chapter 3). This impact is particularly severe in LB medium without supplementation of
divalent cations (Mg2+, Ca2+), the medium used in this study. Genes encoding suspected FFA
export proteins were deleted from a plasmid-free, FFA-overproducing strain of E. coli
[Youngquist 2012]. This allowed for cultivation without the addition of antibiotics, which could
cause non-FFA related toxicity in efflux pump mutants. Three out of fifteen single gene or
operon deletions either resulted in statistically significant reductions of both normalized CFUs
and percent intact cells, or nearly abolished FFA production in the BTE-expressing cultures.
Conversely, no deletion significantly altered viability or fatty acid content in control (non-FFAoverproducing) cultures.
The three hits were rob, acrAB, and tolC.
We had previously
identified Rob as being important to maintaining viability in another fatty acid overproducing
strain, likely due to its role in activating its regulon, which includes acrAB but also a number of
genes encoding proteins with ill-defined functions, in response to intracellular FFAs (Chapter 3).
AcrAB appears to be most important drug efflux pump characterized in E. coli to date for
conferring SDS resistance [Nishino 2001], therefore it is perhaps unsurprising that it also was the
only single efflux pump deletion to render cells more sensitive toward endogenous FFA
production. As further validation of their roles, deletion strains of rob and acrAB also exhibited
growth at lower maximum concentrations of exogenously added octanoate and decanoate than in
strains where these genes were intact.
TolC, which serves as the outer membrane component of the AcrB and many other inner
membrane drug efflux pump subunits, had previously been identified in a transposon
155
mutagenesis screen for membrane-associated proteins that were essential for growth on
decanoate or oleate as a sole carbon sources [Azizan 1994]. However, its role was not associated
with import of FFAs, but rather toward remediating an undefined toxic effect of FFAs on growth.
One possibility is that TolC was necessary for effluxing excess toxic quantities of FFAs beyond
the levels that could be processed by acyl-CoA synthetase (FadD). These excess FFAs could
have been either imported into the periplasm by FadL, the dominant importer of FFAs across the
outer membrane [Black 1990] or could have entered the periplasm and cytosol by diffusive
processes. TolC could also play a number of secondary roles that are unrelated to direct FFA
efflux, as it has been implicated in extrusion of intracellular metabolites including signaling
molecules such as cAMP [Hantke 2011], siderophores for iron acquisition [Bleuel 2005], and
excess cysteine [Wiriyathanawudhiwong 2009].
Furthermore, disruption of tolC triggers
induction of phage shock proteins and reduced NADH oxidase activity, suggesting a role for
TolC in maintaining inner membrane integrity or in direct interaction with enzymes of aerobic
respiration [Dhamdhere 2010]. Thus while tolC is critical for achieving overproduction of FFAs,
the precise role of TolC in efflux of FFAs remains unresolved in this study.
A number of additional efflux pumps were identified in the single deletion screens as
reducing normalized CFUs and percent intact cells, but with P-values between 0.05 and 0.15 in
one or both measures. These included deletions in emrAB, mdtEF, and mdtABCD. These three
efflux pumps are also known to confer a small increase in SDS resistance when overexpressed in
an acrAB deletion strain, in addition to acrD and acrEF [Nishino 2001] which did not have any
impacts on viability in this study as single deletions. To determine if the effects of deleting these
genes and operons was masked by the presence of chromosomal acrAB, these deletions were
tested in strain TY05 ∆acrAB for their effect on viability and FFA production. No differential
156
effects versus the deletion in acrAB alone were observed except for in TY05 ∆acrAB
∆emrAB, which exhibited a phenotype approaching that observed for TY05 ∆tolC, with reduced
levels of growth and delayed onset of stationary phase, coupled with a dramatic reduction in FFA
production after 8 h (but largely recovered at 24 h). Thus both acrAB and emrAB appear to be
important for FFA efflux in BTE-expressing cultures, although deletion of only one of these
operons only has an impact on viability and has no impact on FFA titers. We hypothesize that
multiple RND-type efflux pumps may fill compensatory roles, albeit with varying activities, to
achieve the same net level of excretion of FFAs until multiple inner membrane components are
deleted.
Hits identified in the deletion screens were individually overexpressed by themselves, or as an
artificial operon together with tolC. While all of these constructs conferred increased resistance
toward SDS in an acrAB deletion strain, the only tolC containing construct that improved
resistance over the non-tolC containing construct was pBAD33*-mdtEF-tolC. These constructs
were overexpressed in TY05ara (acrAB+) under conditions conducive to maintaining the proton
motive force (addition of excess divalent cations), however no overexpressed efflux pump
improved either viability or fatty acid titers over negative control cultures. Concomitantly, none
of the overexpressed efflux pumps improved resistance to exogenous addition of octanoate or
decanoate in an acrAB+ strain. However, many of the efflux pumps did improve resistance
against octanoate and decanoate in an acrAB- strain. This included overexpression of acrAB,
demonstrating rescue of the deleted genes, only mdtEF-tolC for both octanoate and decanoate,
and emrAB and mdtEF for only decanoate.
Thus it can be concluded that of the genes investigated in this study, acrAB appears to be
providing a native level of FFA efflux that cannot be exceeded by either overexpression of
157
acrAB or other targeted efflux pumps. This may be due to an already high basal level of
expression or saturation of membrane protein insertion machinery.
Other efflux pumps,
particularly mdtEF and emrAB, can complement the activity of acrAB albeit at lower degrees of
efficacy. New insights were gained, including expanding the range of substrates for these efflux
pumps to include C8-C12 FFAs, providing a newly recognized mechanism for which FFAs appear
to be actively excreted from cells, and possibly providing future metabolic engineering targets
for increasing FFA titer and yield once critical barriers are revealed. These could include
saturation of membrane protein insertion pathways, proton motive force dissipation due to
insufficiently active aerobic respiration or other mechanisms, and negative consequences of FFA
overproduction that remained uncorrected in this study (eg. elevated unsaturated membrane lipid
content) (Chapter 3).
4.5 Conclusion
E. coli possesses an innate ability to secrete medium-chain length FFAs, and we have
demonstrated that this ability is conferred by a selection of RND-type multidrug efflux pump
systems (primarily AcrAB-TolC, EmrAB-TolC, and MdtEF-TolC). As overexpression of these
systems failed to confer increased endogenous FFA production, and only improved tolerance in a
∆acrAB strain, exceeding the native level of secretion has proven to be a challenge. Additional
strategies for utilizing the identified efflux pumps to improve FFA production will be discussed
in Chapter 6.
Chapter 5:
158
Improved tolerance of E. coli toward FFA overproduction by
modulation of membrane acyl chain composition
A key hypothesis that was generated in the functional genomics study detailed in Chapter 3
was that endogenous FFA production negatively impacted membrane integrity, and that these
effects could constitute a physiological barrier to achieving higher yields of FFAs. Chapter 4
addressed increasing FFA export as one strategy for alleviating the negative impact of
endogenously produced FFAs. While initial efforts to increase efflux pump expression did not
increase FFA titer or tolerance to exogenously added FFAs above the baseline strain, efflux
pumps capable of secreting FFAs were identified that will enable further optimization and
protein engineering studies (Chapter 6). In Chapter 5, a second negative physiological effect of
FFA production, increased unsaturated fatty acid content of the cell membrane, will be further
explored. A successful strategy for increasing tolerance towards and transient productivity of
FFAs was achieved by redirecting acyl-ACP pools to restore proper membrane composition.
5.1 Introduction
One of the most promising routes toward the production of renewable substitutes for
petrodiesel in E. coli makes use of intermediates derived from fatty acid biosynthesis. By
introducing a cytosolic acyl-acyl carrier protein (ACP) thioesterase, feedback inhibition of
enzymes of fatty acid biosynthesis (acetyl-CoA carboxylase (AccABCD), β-ketoacyl-ACP
synthase III (FabH), enoyl-ACP reductase (FabI)) by acyl-ACP intermediates is released,
increasing flux through fatty acid biosynthesis [Cho 1995, Davis 2000]. The released free fatty
acids (FFAs) can either be separated from the culture medium and catalytically decarboxylated
159
to alkanes (Chapter 2) [Mäki-Arvela 2007], or they can be directed into heterologous
pathways that utilize acyl-CoAs by overexpression of the native E. coli acyl-CoA synthetase
(FadD) or heterologous expression of other acyl-CoA ligases [Steen 2010, Zhang 2012]. AcylCoA derived products include fatty acid ethyl esters, fatty alcohols, and methyl ketones
(reviewed in section 1.5), with the product profiles dependent on selection of the acyl-ACP
thioesterase (reviewed in section 1.4.1).
Past reports have indicated that heterologous expression of the acyl-ACP thioesterase from
Umbellularia californica (BTE) in E. coli results in greatly elevated levels of unsaturated and
cyclic phospholipids (Chapter 3) [Voelker 1994], which are derived from unsaturated acyl
groups. BTE belongs to the FatB family of plant acyl-ACP thioesterases, and has a specificity
predominantly for saturated C12 acyl-ACPs (~60-70% of FFAs), while also hydrolyzing
unsaturated C12 (~10%), saturated C14 (~10%), and unsaturated C14 (~10%) [Voelker 1994]. It
was postulated that the heightened unsaturated membrane lipid content was a result of altered
long-chain acyl-ACP pools due to BTE expression.
Altered membrane content has been
observed in cells expressing other thioesterases with the changes directly correlated to
thioesterase substrate specificity. For instance, expression of the FatA type thioesterase from
Helianthus annuus in E. coli, which predominantly cleaves unsaturated C16 and C18 acyl-ACPs
(but was likely non-optimally expressed with only a 23% increase in total fatty acids reported),
increased the saturated phospholipid acyl content by approximately 5% [Serrano-Vega 2005].
Overexpression of a cytosolic form of E. coli thioesterase I (TesA'), which was reported to
generate a FFA distribution of approximately 54% unsaturated and 46% saturated, also resulted
in a 7.5% reduction of unsaturated and cyclopropanated phospholipid acyl group content [Cho
1995].
160
In previous work, we observed strong decreases in expression of genes involved in
unsaturated fatty acid biosynthesis (fabA and fabB) in cultures expressing BTE compared to
cultures expressing a non-functional thioesterase with a catalytic histidine mutagenized to an
alanine, BTE-H204A (Chapter 3). This was concomitant with greatly increased unsaturated C16C18 fatty acid content associated with membrane phospholipids (Chapter 3) [Voelker 1994].
FabA, which catalyzes formation of the cis double bond in elongating acyl chains at the C10
chain length, and FabB, which is essential for condensing cis-3-decenoyl-ACP with malonylACP, are both regulated at the transcriptional level by DNA binding of FabR to their promoter
region [Zhang 2002, Feng 2009]. FabR binds DNA and both saturated and unsaturated acylACPs, with a weakened affinity for DNA when bound to unsaturated acyl-ACPs [Zhu 2009,
Feng 2011]. FabR thus plays a key role in modulating unsaturated membrane lipid content [Zhu
2009]. Due to this mechanism of repression (Figure 3.12), we hypothesized that expression of
BTE, which primarily cleaves saturated acyl-ACPs, was enriching the acyl-ACP pool in
unsaturated acyl-ACPs. The high proportion of unsaturated acyl-ACPs increased FabR-mediated
repression, reducing transcription of fabA and fabB, but this repression was insufficient to
prevent a highly elevated unsaturated membrane lipid content which would be anticipated to
greatly affect membrane properties.
E. coli and other bacteria adjust both the phospholipid head group and acyl chain
compositions of their cell membrane in response to environmental conditions such as
temperature and organic solvent exposure, to maintain homeostasis in physical properties such as
fluidity [eg. Marr 1962, Ingram 1976, Sikkema 1995, Ramos 2002, Huffer 2011]. Long-chain
alcohols and aromatic compounds generally induce increases in saturated fatty acid content
[Ingram 1976, Huffer 2011]. This has been interpreted as an attempt by the cell to rigidify the
161
cell membrane and become more impermeable to the hydrophobic intercalating compound
[Sikkema 1995]
Alternatively, biophysical studies on the interaction of n-alkanes with
phospholipid bilayers suggest that alkane intercalation disturbs interactions between acyl chains,
resulting in further disordering [McIntosh 1980] that could potentially be offset by a shift toward
higher fatty acid saturation. Treatment of cells with n-hexanol similarly results in both an
increase in membrane fluidity and an increase in saturated fatty acids [Ingram 1980]. While no
studies have been reported on the effect of exogenous addition of FFAs or other amphiphilic
compounds on membrane fatty acid composition, a variety of antibacterial effects have been
reported in many bacteria including disruption of the electron transport chain, uncoupling of
oxidative phosphorylation, and cell lysis [Desbois 2010]. These were similar to effects that we
observed in cells that endogenously produced FFAs: increased membrane depolarization,
induction of phage shock proteins, increased expression of genes involved in aerobic respiration,
increased percentages of non-intact cells as measured using SYTOX Green, decreased viability,
and induction of other membrane stress responses (Chapter 3).
An inability of cells to
effectively regulate their membrane lipid unsaturation, such as appeared to be imparted by BTE
expression, could be at least partly responsible for many of the observed negative physiological
effects of endogenously produced FFAs.
In this study, we begin by further investigating the effect of BTE expression on unsaturated
membrane lipid content, cell lysis, FFA titers, and fabA and fabB expression levels when fabR is
deleted, eliminating the mechanism for feedback repression of unsaturated fatty acid biosynthesis.
Higher levels of fabA and fabB expression were observed, correlated with a greatly elevated
unsaturated membrane lipid content, a highly exacerbated degree of cell lysis, and depressed
FFA titers, underscoring the importance of FabR-mediated control of unsaturated fatty acid
162
biosynthesis toward tolerance of endogenous FFA production. As a demonstration of the
ability to modulate membrane lipid composition by acyl-ACP thioesterase selection, a
thioesterase from Geobacillus sp. Y412MC10 (GeoTE) which was reported to hydrolyze a high
percentage of unsaturated medium-chain length FFAs [Jing 2011], was expressed by itself and in
combination with BTE. Expression of GeoTE both alone and in tandem with BTE reduces the
membrane unsaturated fatty acid content, relieves transcriptional repression of fabA and fabB,
and decreases the population of lysed cells while maintaining total FFA titers. A 2.6-fold
increase in FFA titer on a per-cell basis was achieved by co-expressing BTE and GeoTE. While
overall increased FFA titers and yields were not attained in this study, possibly due to the
existence of other unaddressed physiological or metabolic bottlenecks, insights were gained on
the interplay between cell physiology and metabolic engineering of FFA overproduction.
5.2 Materials and Methods
5.2.1 Chemicals, reagents, enzymes, and oligonucleotide primers
All chemicals were purchased from Fisher Scientific (Pittsburgh, PA) unless otherwise
specified.
Restriction enzymes, PCR enzymes for cloning, and associated reagents were
purchased from New England Biolabs (Ipswich, MA).
Shrimp alkaline phosphatase was
purchased from Fermentas (Glen Burnie, MD). GoTaq Green Master Mix was purchased from
Promega (Madison, WI) and used for colony PCR. All oligonucleotide primers were purchased
from Integrated DNA Technologies (Coralville, IA). Plasmid DNA and gel extracted DNA was
purified using the QIAPrep Spin Miniprep and QIAEX II kits (Qiagen, Valencia, CA).
163
5.2.2 Gene synthesis
The Geobacillus sp. Y412MC10 (presently renamed Paenibacillus sp. Y412MC10, GenBank
genome accession number CP001793.1) acyl-ACP thioesterase (GeoTE) formerly annotated in
GenBank under accession number EDV77528, and Clostridium thermocellum acyl-ACP
thioesterase (ClosTE) annotated in GenBank under accession number ABN54268 were codonoptimized and synthesized by GeneArt (Life Technologies, Regensburg, Germany) with
elimination of common restriction sites. Prior to synthesis, the codon-optimized ORF was
analyzed using the Ribosome Binding Site Calculator (https://salis.psu.edu/software/) [Salis
2009] to generate a ribosome binding site (RBS) and spacer sequence that resulted in a high
predicted rate of translation (~25000 au) from the designated start codon, with orders-ofmagnitude lower predicted rates of translation from other potential start codons. The RBS,
spacer sequence, and flanking restriction sites at the 3' (XmaI) and 5' (HindIII) termini were
added to the final sequence to be synthesized. A site-directed mutant was also ordered from
GeneArt, following alignment of GeoTE and ClosTE with BTE (Figure 5.1), which identified a
conserved catalytic histidine at positions 173 and 171, respectively.
The full synthesized
sequences were amplified by PCR from the provided GeneArt standard ampicillin-resistant
bacterial cloning vectors using primers (see Appendix III) 117 and 118 for GeoTE, and 119 and
120 for ClosTE, digested with XmaI and HindIII, and ligated into plasmid pBAD18 to generate
plasmids pBAD18-GeoTE, pBAD18-GeoTE-H173A, pBAD18-ClosTE, and pBAD18-ClosTEH171A.
5.2.3 Plasmid construction
Construction of plasmid pBAD33* is described in Chapter 4. The fabR gene was amplified
164
by PCR from MG1655 genomic DNA with flanking 5' (XmaI) and 3' (HindIII) restriction
sites and an artificial RBS and spacer sequence generated from the RBS Calculator forward
design tool (primers 107-108, Appendix III). The artificial RBS and spacer sequence results in a
predicted translation rate of 32100 au from the desired start codon of fabR. The PCR product
was subsequently digested with XmaI/HindIII, and ligated into pBAD33*
BTE
Clostridium
Geobacillus
Consensus
_clipseq
_clipseq#2
_clipseq#3
Consensus
_clipseq
_clipseq#2
_clipseq#3
Consensus
_clipseq
_clipseq#2
_clipseq#3
Consensus
_clipseq
_clipseq#2
_clipseq#3
Consensus
_clipseq
_clipseq#2
_clipseq#3
Consensus
_clipseq
_clipseq#2
_clipseq#3
(1) MTLEWKPKPKLPQLLDDHFGLHGLVFRRTFAIRSYEVGPDRSTSILAVMN
(1) --------------------MQKKRFSKKYEVHYYEINSMQEATLLSLLN
(1) ------------------MELMIDKWTEEYTIQSVDADFKGDCRWSSLLS
(1)
L
KFSK Y I SYEI
D SILSLLN
51
100
(51) HMQEATLNHAKSVGILGDGFGTTLEMSKRDLMWVVRRTHVAVERYPTWGD
(31) YMEDCAISHSTSAGYGVN------ELLAADAGWVLYRWLIKIDRLPKLGE
(33) ILQRAADRHIEALGISRE------EMIERGMGWMLITLELEMRRMPRDME
(51) HMQDAAI HA SLGI D
EMI RDLGWVL R I IDRLPK GE
101
150
(101) TVEVECWIGASGNNGMRRDFLVRDCKTGEILTRCTSLSVLMNTRTRRLST
(75) TITVQTWASSFERFYGNREFIVLDGRDNPIV-KASSVWIYFNIKKRKPMR
(77) NVYVDTWSRGSKGALWHRDYRIKNGDGELLG-EGRSVWALVDIHKRKILR
(101) TV VDTWA AS
RDFIVKDGK
IL KASSVWILMNIKKRKILR
151
200
(151) IPDEVRGEIGPAFIDNVAVKDDEIKKLQKLNDSTADYIQGGLTPRWNDLD
(124) IPLEMGDAYG------IDETRALEEPFTDFDFDFEPKVIEEFTVKRSDID
(126) PSMFPYEVPI------GQETVGELPSKAVLPEGVQLDDAYTYSVRYSGID
(151) IPLEM D G
I ET AEI
L D
I
FTVRWSDID
201
250
(201) VNQHVNNLKYVAWVFETVPDSIFESHHISSFTLEYRRECTRDSVLRSLTT
(168) TNSHVNNKKYVDWIMETVPQQIYDNYKVTSLQIIYKKESSLGSGIKAGCV
(170) TNGHLNNARYADLCFDVLDEQELREGLVTGFKITYLNEARLKDTMLIKRS
(201) TN HVNN KYVDWIFETVPDQIFD H VTSF I YKKEASL S IKA S
251
300
(251) VSGGSSEAGLVCDHLLQLEGGSEVLRARTEWRPKLTDSFRGISVIPAEPR
(218) IDEQNTDNPRLLHKIWDKNTGLELVSAETIWQKIQS-------------(220) AEENNR-----VYVQGTSPDGTNFFEAAIVRES----------------(251) IDENNSD
LLH I
GSELL A TIW
S
301
(301) V(254) -(248) --
Figure 5.1 Alignment of amino acid sequence of BTE with bacterial acyl-ACP thioesterases from
Geobacillus Y412MC10 (GeoTE) and Clostridium thermocellum (ClosTE). The catalytic histidine at
position 204 in BTE and HVNN motif aligns with histidine-173 in GeoTE and histidine-171 in ClosTE.
Valine-205 in BTE aligns with a similar leucine residue in GeoTE.
5.2.4 Strain construction
Bacterial strains, plasmids, and oligonucleotide primers used in this study are listed in
Appendices I, II, and III, respectively. All enzymes used for PCR and cloning were purchased
from New England Biolabs (Ipswich, MA) unless otherwise specified. The background strain
165
used in this work is RL08ara, which is the previously described strain E. coli RL08 (K-12
MG1655 ∆fadD ∆araBAD) with the additional deletion of araFGH encoding a high-affinity
arabinose transporter, and replacement of the native promoter of araE, a low-affinity arabinose
transporter, with a constitutive promoter.
These modifications were previously shown to
generate homogeneous induction from the PBAD promoter across a population of cells, and to
enable to titration of arabinose to modulate gene expression from a PBAD promoter [Khlebnikov
2001]. Strain RL08ara was constructed by sequential P1 phage transductions using lysates
harboring Φ(∆araEp kan Pcp8-araE) and araFGH::kan loci from strains BW27271 and
BW27269, respectively [Khlebnikov 2001]. Antibiotic resistance genes were removed after each
transduction using pCP20 [Cherepanov 1995].
Plasmid pCP20 was removed by repeated
elevated temperature cures at 43°C, and the presence of all desired FRT-site containing loci were
confirmed by colony PCR (primers 3-6, 53-56, Appendix III). Promoter replacement of araE
was verified sequentially following both transduction and elevated temperature cure of pCP20.
Strain RL08ara ∆fabR was constructed by P1 phage transduction of the fabR::kan cassette
from strain JW3935-4 [Baba 2006]. The kanamycin resistance gene was removed using pCP20
as described above, which was subsequently removed, and all FRT-site containing loci were
confirmed by colony PCR (primers 3-6, 53-56, 105-106, Appendix III)
5.2.5 Cell cultivation
All strains tested for viability, RNA extraction, and fatty acid production were grown
aerobically in a shaker (New Brunswick, Enfield, CT) at 37°C and 250 rpm in 250 mL shake
flasks with a 5X headspace in Difco LB medium (BD, Franklin Lakes, NJ) supplemented with
0.4% v/v glycerol. Chloramphenicol was added to a concentration of 34 µg/mL in strains
166
harboring pBAD33* derived plasmids. Ampicillin was added to a concentration of 50
µg/mL in strains harboring pBAD18 derived plasmids, and 100 µg/mL in strains harboring
pTrc99A derived plasmids. Cultures were induced at OD600 0.2 with 0.2% arabinose (except
where otherwise noted) for expression of genes on pBAD33, pBAD33*, and pBAD18, and 50
µM IPTG for expression of genes on pTrc99A. Strains were grown in biological triplicate from
overnight cultures inoculated with independent colonies on streak plates or fresh plasmid
transformation plates.
5.2.6 SYTOX flow cytometry assays
To assess cell permeability, cell pellets were collected and stained by addition of 1 µL of 5
mM SYTOX Green in DMSO (Invitrogen) and measured by flow cytometry as described in
Chapter 4. Data analysis was also performed as described in Chapter 4.
5.2.7 Fatty acid extraction and analysis
Total fatty acids were extracted from cell cultures and methylated by acid catalysis as
described in Chapter 2. Bound fatty acids were extracted and methylated by base catalysis where
described in the text. Samples were extracted from 2.5 mL of cell culture spiked with 5 µL of 10
mg/mL heptadecanoic acid (Fluka) in ethanol and 10 µL of 5 mg/mL 1,2-dipentadecanoyl-snglycero-3-phosphoethanolamine (Avanti Polar Lipids, Alabaster, AL) in chloroform with 5 mL
of 1:1 chloroform:methanol, vortexed thoroughly, and centrifuged at 1000 x g for 10 min. The
aqueous upper layer and interfacial cell debris was removed, and the bottom organic layer was
evaporated to dryness under a nitrogen stream. To the dried residue, 0.5 mL of 0.5 M sodium
methoxide in methanol (Sigma) was added, and the reactions were allowed to proceed at 50°C
167
for 10 minutes [Christie 2003]. To quench the reaction, 0.1 mL of glacial acetic acid was
added, followed by 5 mL of deionized water. Fatty acid methyl esters were extracted twice into
0.5 mL of hexane, and the collected hexane layers were quantified as for total fatty acids. Gas
chromatography/ mass spectrometry (GC/MS) analysis and peak identification and quantification
was performed on a model 7890 Agilent GC with a model 5975 mass spectrometer as described
in Chapters 2 and 3. Average fatty acid concentrations from biological triplicate cultures were
determined by normalizing to recovered FFA internal standards (pentadecanoic and
heptadecanoic acid) from acid-catalyzed methylations, and from the recovered fatty acids
derived from the phospholipid internal standard in base-catalyzed methylations (heptadecanoic
acid was added only for verification that FFAs were not being methylated).
5.2.8 RNA extraction and qPCR
From shake flask cultures expressing combinations of pBAD33-BTE-H204A, pBAD33-BTE,
pBAD18-GeoTE, and pBAD18-GeoTE-H173A, cells were collected in mid-log phase (OD ~0.8,
3.25 hours post-inoculation) and in early stationary phase (OD varied between 0.6 to 4.6, 5.25
hours post-inoculation) by spinning down 0.8 mL and 0.333 mL of culture, respectively,
centrifuging at 16000 x g for 1 min, aspirating off the supernatant, and flash freezing the cell
pellet in a dry ice/ethanol bath. Cell pellets were stored at -80°C. To extract RNA, cell pellets
were resuspended in 100 µL of TE buffer, pH 8.0, containing 400 µg/mL lysozyme and
incubated for 5 minutes. From this point forward, the Qiagen RNeasy Plus Kit was used
according to the manufacturer's instructions, beginning with the addition of 350 µL of Buffer
RLT Plus to each sample and applying each sample to gDNA removal columns. For shake flask
cultures expressing combinations of pTrc99A-BTE, pTrc99A-BTE-H204A, pBAD33*, and
168
pBAD33*-fabR in strains RL08ara and RL08ara ∆fabR, approximately 0.7 OD600-mL of
cells were collected in early stationary phase (4.58 h post-inoculation) and centrifuged at 16000
x g for 1 min. The supernatant was removed by aspiration and cell pellets were flash frozen in a
dry ice/ethanol bath prior to storage at -80°C. RNA was extracted and purified as described
above, with the exception of using the Qiagen RNeasy Kit.
Following RNA clean-up, any contaminating DNA was removed using the DNA-freeTM Kit
(Applied Biosystems, city, state) according to the manufacturer's instructions. The absence of
DNA contamination was confirmed by using 0.5 µL of each RNA sample as template in a PCR
reaction with GoTaq Green Master Mix (Promega, Madison, WI) and primers specific for acpP
(primers 131-132, Appendix III) First-strand cDNA synthesis was performed for the first set of
shake flask cultures using the Promega GoScriptTM Reverse Transcription System (Madison, WI)
according to the manufacturer's instructions, with 0.5 µg of RNA, random primers, 2.5 mM
MgCl2, and a 1 hour extension time at 42°C. First-strand cDNA synthesis for the second set of
shake flask cultures was performed using the Bio-Rad iScriptTM Reverse Transcription Supermix
(Hercules, CA) for RT-qPCR with 1 µg RNA template. The presence of cDNA was confirmed
in each reaction using 0.5 µL of each sample as template as described above for confirming the
lack of DNA contamination in RNA samples. To quantify the relative expression of fabA, fabB,
and fabR transcript in each sample, qPCR reactions were set up using Fermentas Maxima SYBR
Green master mix was used according to the manufacturer's directions, with 2.0 µL of 5-fold
diluted cDNA in water was used as template, and primers listed in Appendix III (35-36, 121-122,
133-134). SYBR green fluorescence was monitored with a Bio-Rad CFX96 optical reaction
module. Cycle quantification (Cq) values were calculated by Bio-Rad CFX Manager software,
and relative expression values to the double negative controls were calculated as 2-∆Cq.
169
5.3 Results
5.3.1 Effect of fabR deletion on unsaturated membrane-bound fatty acids
In Chapter 3, a sharply elevated percentage of long-chain (C16 to C18) unsaturated fatty acids
was observed in a fatty-acid-overproducing, BTE-expressing strain of E. coli (RL08) compared
to a non-fatty-acid-overproducing, BTE-H204A-expressing strain. Due to the medium-chain
length acyl-ACP substrate specificity of BTE [Voelker 1992, Voelker 1994], these unsaturated
long-chain fatty acids are predominantly associated with inner and outer membrane
phospholipids. This was verified in strain RL08ara harboring pBAD33-BTE and pBAD18GeoTE-H173A, encoding a non-functional Geobacillus sp. TE (GeoTE) with a catalytic histidine
converted to an alanine, by comparative analysis of fatty acid methyl esters obtained via acidand base-catalyzed methylations by GC/MS (Figure 5.2). Acid catalysis methylates both free
fatty acids and bound fatty acids in phospholipids and other species, whereas base catalysis
methylates only bound fatty acids. The difference between the total and bound fatty acid profiles
provides the free fatty acid profile, which definitively demonstrates within experimental error
that BTE does not significantly cleave C16 to C18 chain length fatty acids.
Despite the increased membrane unsaturated content in BTE-expressing strains, a strong
repression of fabA and fabB was observed beginning as early as the mid-log phase of growth and
continuing into stationary phase (Chapter 3), with the relative repression ratio consistent with in
vitro studies of FabR-mediated repression of these genes [Feng 2011]. We postulated that
expression of BTE results in a depletion of saturated acyl-ACPs relative to unsaturated acylACPs, resulting in an increased level of FabR mediated repression of fabA and fabB. However
the repression appears to be insufficient to prevent an enrichment of unsaturated fatty acids in the
cell membranes. If this hypothesis is correct, deletion of fabR should result in both increased
170
400
fatty acid titer (mg/L)
350
300
8h acid
8h base
24h acid
24h base
250
200
150
100
50
:0
:1
18
C
18
C
C
8:
0
C
10
:1
C
10
:0
C
11
:0
C
12
:1
C
12
:0
C
13
:0
C
14
:1
C
14
:0
C
16
:1
C
16
:0
0
Figure 5.2 Comparative analysis of acid and base catalyzed FAME preparations from cultures
expressing BTE and GeoTE-H173A. Cultures were sampled 8 h and 24 h post-inoculation. Free fatty
acids can be calculated by subtracting the base titers from acid titers. BTE hydrolyzes predominantly
12:1, 12:0, 14:1, and 14:0 fatty acids, and does not appreciably hydrolyze C16-C18 fatty acids.
levels of fabA and fabB, as well as a further increase in unsaturated content that is amplified by
expression of BTE. To test this hypothesis, shake flask cultures were cultivated of RL08ara (see
Materials and Methods) and RL08ara ∆fabR harboring combinations of plasmids pTrc99A-BTEH204A and pTrc99A-BTE, and pBAD33* and pBAD33*-fabR to test the effect of FabR
overexpression. Cultures were induced with 50 µM IPTG and 0.2% L-arabinose at an OD600 of
between 0.05 to 0.12. Levels of fabA and fabB transcript were quantified by qPCR in early
stationary phase (approximately 4.5 hours after inoculation) in strains RL08ara and RL08ara
∆fabR harboring pBAD33* and pTrc99A-BTE-H204A or pTrc99A-BTE, as well as strain
RL08ara ∆fabR harboring pBAD33*-fabR and pTrc99A-BTE-H204A or pTrc99A-BTE in order
to test the effect of complementation of fabR (Figure 5.3).
Strain RL08ara/pTrc99A-BTE
exhibited statistically significant (P < 0.05) decreased levels of fabA and fabB relative to strain
RL08ara/pTrc99A-BTE-H204A, consistent with previous results comparing BTE- and BTE-
171
H204A-expressing strains under different growth conditions (Chapter 3).
Strains with
deletions in fabR exhibited heightened levels of fabA and fabB expression in both BTE-H204A
(fabA, P = 0.069; fabB, P = 0.033) and BTE-expressing strains (fabA, P = 0.034; fabB, P = 0.031)
relative to RL08ara/pTrc99A-BTE-H20A, with a higher fold-change in fabA and fabB expression
between the BTE-expressing strains than between the BTE-H204A-expressing strains, as
postulated due to the increased degree of FabR-mediated repression that is present when BTE is
expressed.
inoculation.
Total fatty acids were also analyzed following extraction 8 h and 24 h postAs predicted by the increased levels of fabA and fabB, strain RL08ara
∆fabR/pTrc99A-BTE-H204A also exhibited a higher percentage of unsaturated C16-C18 fatty
acids at 8 h (48.0 ± 0.8%) than RL08ara/pTrc99A-BTE-H204A (34.8 ± 1.7%), and strain
RL08ara ∆fabR/pTrc99A-BTE exhibited an even larger increase (81.1 ± 0.3%) than RL08ara/
4.0
relative expression level
3.5
* = P<0.05 for C
3.0
q
values
*
*
fabA
fabB
*
2.5
2.0
1.5
1.0
0.5
0.0
**
*
+fabR
∆fabR
+fabR
RL08ara RL08ara ∆fabR
BTE-H204A BTE BTE-H204A BTE BTE-H204A BTE
Figure 5.3. Normalized transcript levels of fabA and fabB in strains with deleted or overexpressed
fabR expressing BTE or BTE-H204A. qPCR was performed using cDNA synthesized from RNA
extracted from cell pellets harvested 4 hours, 35 minutes after inoculation. In the strain with intact
chromosomal fabR (RL08ara), expression of BTE reduces both fabA and fabB levels. In ∆fabR strains
expressing both BTE-H204A and BTE, levels of fabB are increased with P<0.05. In ∆fabR strains
overexpressing FabR on a plasmid (+fabR), levels of fabA and fabB are generally restored to lower levels
than those observed in strains lacking fabR. P-values are calculated by a t-test against the expression
levels of fabA or fabB in strain RL08ara expressing BTE-H204A.
172
% unsaturated C16-C18
80
70
60
50
40
30
20
10
0
RL08ara
BTE-H204A
RL08ara
∆fabR
BTE
BTE-H204A
∆fabR
BTE
∆fabR
+fabR
BTE-H204A
∆fabR
+fabR
BTE
Figure 5.4 Percent unsaturated C16-C18 fatty acids in strains with deleted or overexpressed fabR
expressing BTE or BTE-H204A. Fatty acids were extracted from cultures 8 hours post-inoculation, and
do not include C17∆. Expression of BTE dramatically increases unsaturated content in strains harboring
chromosomal fabR (RL08ara), and further increases unsaturated content in the ∆fabR strain.
Overexpression of fabR on a plasmid in the ∆fabR strain restores unsaturated content to a lower level than
present in RL08ara.
pTrc99A-BTE (65.4 ± 1.2%) (Figure 5.4). Fatty acid compositions followed the same trend
when extracted 24 h post-inoculation.
Strains overexpressing FabR on pBAD33*-fabR exhibited growth defects (data not shown),
likely as a result of the expression of fabR being at too high a level, despite use of a low copy
plasmid. Measurement of relative fabR transcript levels by qPCR indicated a 106 ± 65 (P <
0.001) and 82 ± 51 (P < 0.001) fold increase in strain RL08ara ∆fabR/pTrc99A-BTE-H204A and
RL08ara ∆fabR/pTrc99A-BTE overexpressing fabR, respectively, over the native levels of fabR
expression in strain RL08ara. Expression levels of overexpressed transcription factors requires
careful control due to their DNA-binding nature.
Despite the observed growth defects,
overexpression of fabR restored decreased levels of fabA and fabB in RL08ara ∆fabR (Figure
5.3), and also restored unsaturated C16-C18 fatty acid levels in RL08ara ∆fabR expressing either
173
BTE-H204A or BTE to similar or lower levels as those present RL08ara (Figure 5.4). While
not further considered in this study, fabR overexpression at lower levels may be a viable strategy
for reversing the increased unsaturated fatty acid membrane composition resulting from
expression of acyl-ACP thioesterases with predominantly saturated acyl-ACP substrate
specificity.
5.3.2 Effect of fabR deletion on cell membrane integrity
Cells from the same cultures described above were collected 8 h post-inoculation and stained
with SYTOX Green. The green fluorescence of individual cells in the population was measured
by flow cytometry as described in Materials and Methods.
SYTOX Green is ordinarily
impermeable to intact inner cell membranes, but results in a bright green fluorescence upon
nucleic acid binding in cells with non-intact inner membranes [Roth 1997]. RL08ara/pTrc99ABTE-H204A is 98.4 ± 0.2 percent intact while RL08ara ∆fabR/pTrc99A-BTE-H204A exhibits
only a slight reduction to 94.0 ± 0.6 percent intact (Figure 5.5). However, RL08ara/pTrc99ABTE is 75.5 ± 4.0 percent intact while RL08ara ∆fabR/pTrc99A-BTE has a dramatic reduction to
14.4 ± 5.2 percent intact (Figure 5.5). Therefore deletion of fabR has a specifically deleterious
effect on cell membrane integrity under conditions of fatty acid overproduction. This effect
could be a direct result of the severely elevated unsaturated fatty acid membrane content in
RL08ara ∆fabR, or due to the combination of free fatty acid overproduction and elevated
unsaturated content.
174
5.3.3 Effect of fabR deletion on FFA production
The negative physiological consequences of fabR deletion manifested in FFA production of
BTE-expressing strains. RL08ara/pTrc99A-BTE exhibited higher titers of C8-C14 fatty acids
(predominantly FFAs) at both 8 h and 24 h post-inoculation compared to RL08ara
∆fabR/pTrc99A-BTE (Figure 5.6). At 8 h, titers for these two strains were 333 ± 21 mg/L and
259 ± 11 mg/L, respectively.
At 24 h, titers were 629 ± 48 mg/L and 479 ± 30 mg/L,
respectively.
100
BTE-H204A
BTE
percent intact
80
60
40
20
0
RL08ara
RL08ara
∆fabR
Figure 5.5 Percent intact cells in fabR+ and ∆fabR strains expressing BTE or BTE-H204A. Cells
were harvested 8 h post-inoculation and stained with SYTOX Green before analysis by flow cytometry.
In BTE-H204A-expressing cultures, deletion of fabR has little effect on percent intact cells. In BTEexpressing cultures, a dramatic increase in cell lysis is observed as a result of deletion of fabR.
5.3.4 Selection and expression of a predominantly medium-chain length
unsaturated acyl-ACP thioesterase
BTE is in the FatB family of plant acyl-ACP thioesterases, all of which exhibit predominantly
saturated acyl-ACP substrate specificities [Jones 1995]. The FatA family of plant acyl-ACP
thioesterases primarily hydrolyze cis-18:1∆9-ACP [Jones 1995], and none has been
characterized in the literature that can hydrolyze medium-chain length unsaturated species as a
C8-C14 fatty acids (mg/L)
175
8h
24 h
600
400
200
0
RL08ara
RL08ara
∆fabR
Figure 5.6 Effect of fabR deletion on C8-C14 (predominantly free) fatty acid titer produced in BTEexpressing cultures. Reduced titers were observed at both 8 h and 24 h growth in the ∆fabR strain.
significant portion of its product profile. Recently, a number of bacterial acyl-ACP thioesterases
were further characterized by heterologous expression in a fadD deficient strain of E. coli [Jing
2011]. Thioesterases from Bacteroides fragilis, Bacteroides thetaiotaomicron, Streptococcus
dysgalactiae, Lactobacillus brevis, Lactobacillus plantarum, Bdellovibrio bacteriovorus,
Clostridium thermocellum, a Geobacillus sp., Desulfovibrio vulgaris, and Bryantella
formatexigens all produced greater quantities of unsaturated than saturated C12-C14 species.
However, only C. thermocellum and Geobacillus sp. produced more unsaturated C12-C14 fatty
acids than octanoic acid, therefore these acyl-ACP thioesterases (hereafter referred to as ClosTE
and GeoTE) were selected for codon-optimization and chemical synthesis. Catalytic histidines
were identified that aligned with His-204 and shared a NXHVNN motif previously identified by
alignment of plant FatA and FatB thioesterases [Yuan 1996], with the exception of GeoTE
having a leucine residue in place of valine (Figure 5.1). Non-functional mutations of ClosTE
(ClosTE-H171A) and GeoTE (GeoTE-H173A) were generated in order to provide negative
176
controls that account for protein production effects. ClosTE, GeoTE, ClosTE-H171A, and
GeoTE-H173A were cloned into pBAD18 and tested in strain RL08ara. Both non-functional
bacterial thioesterases produced fatty acid profiles similar to BTE-H204A-expressing cells
(Figure 5.7). Expression of ClosTE produced elevated levels of 8:0, 10:1, 10:0, 12:1, 12:0, 14:1,
and 14:0, as well as detectable levels of nonanoic, 3-hydroxyoctanoic, and 3-hydroxydecanoic
acids (data not shown). However, titers of C8-C14 fatty acids were only 36 mg/L compared with
12 mg/L (mostly 14:0 likely associated with phospholipids) in ClosTE-H171A-expressing
cultures. These low titers were likely a result of a low level of functional thioesterase expression,
which may not be surprising given that C. thermocellum is a thermophilic organism that grows
optimally at temperatures in excess of 55°C.
Expression of GeoTE produced highly elevated levels of 8:0, 10:1, 10:0, 12:1, 12:0, 14:1, and
14:0 fatty acids and detectable levels of odd-chain 9:0, 11:0, and 13:0 fatty acids (Figure 5.7).
Furthermore, β-hydroxylated octanoic, decanoic, dodecanoic, tetradecanoic, and hexadecanoic
species were detected according to their elution time patterns and primary m/z ion of 103 (and
often their molecular ion) (data not shown). Smaller tridecanone and pentadecanone peaks were
also identified from their mass spectra (data not shown), which may be spontaneous
decarboxylation
products
following
hydrolysis
of
3-oxotetradecanoyl-ACP
and
3-
oxohexadecanoyl-ACP, similar to observations by Goh and coworkers [2012] by overexpression
of E. coli FadM, an acyl-CoA thioesterase. Elevated levels of 16:1 but decreased levels of 16:0,
18:1, and 18:0 were observed in cells expressing GeoTE compared with cells expressing GeoTEH173A. Titers of C8-C14 fatty acids were 248 mg/L compared with 12 mg/L in GeoTE-H173Aexpressing cultures, and the proportion of C8-C14 species that were unsaturated was 55 percent,
compared with a typical value in BTE-expressing cultures of 23 percent. As a result of this high
177
level of medium-chain length fatty acid production with a higher percentage of unsaturated
species than BTE, GeoTE was selected for moving forward for further characterization of its
impact on membrane fatty acid content.
100
fatty acid titer (mg/L)
A
GeoTE-H173A
GeoTE
80
60
40
20
C
8:
0
C
10
:1
C
10
:0
C
11
:0
C
12
:1
C
12
:0
C
13
:0
C
14
:1
C
14
:0
C
16
:1
C
16
:0
C
18
:1
C
18
:0
0
80
fatty acid titer (mg/L)
B
ClosTE-H171A
ClosTE
60
40
20
:1
:0
:1
:0
C
18
C
18
C
16
:1
:0
:0
C
16
C
14
C
14
:1
:0
:0
C
13
C
12
C
12
:1
:0
C
11
C
10
C
10
C
8:
0
0
Figure 5.7 Total fatty acid analysis of functional and mutagenized GeoTE and ClosTE. (A) GeoTE.
(B) ClosTE. Cultures were grown for 24 hours in LB plus 0.4% glycerol at 37°C. Expression of GeoTE
resulted in high titers of overexpressed FFAs with a diverse profile and 12:1 as the highest concentration
product. Expression of ClosTE produced low FFA titers.
5.3.5
178
Unsaturated long-chain fatty acid biosynthesis in BTE and GeoTE
expressing cultures
Voelker and Davies [1994] observed large decreases in 16:0 and increases in 16:1 and 18:1 as
a percentage of fatty acids derived from phospholipids in BTE-expressing cultures of E. coli.
They postulated that this was a result of BTE acting predominantly on 12:0 and 14:0 acyl-ACPs
to generate free 12:0 and 14:0 fatty acids. Similarly, in Chapter 3 we found that the unsaturated
C16-C18 fatty acid content of BTE-expressing cultures was 56 percent, compared to 30 percent in
BTE-H204A-expressing cultures.
To determine the effect of GeoTE expression and
GeoTE/BTE coexpression on membrane lipid composition and to compare this with that of BTE
expression, shake flask cultures of RL08ara harboring plasmid combinations of pBAD33-BTE or
pBAD33-BTE-H204A, and pBAD18-GeoTE or pBAD18-GeoTE-H173A were grown in shake
flasks and sampled for fatty acids 8 h and 24 h post-inoculation. Fatty acid methylation was
performed with both acid-catalyzed and base-catalyzed methods as described in section 5.2.7, in
order to differentiate between free (total minus bound) and bound fatty acids (primarily from
membrane phospholipids). Analysis of only bound fatty acids by base-catalyzed methylation
revealed that cultures coexpressing GeoTE and BTE-H204A, as well as cultures co-expressing
GeoTE and BTE, maintained a lower percentage of C16-C18 unsaturated fatty acids than cultures
coexpressing BTE and GeoTE-H173A (45.3 ± 0.8, 39.6 ± 7.4, and 65.1 ± 1.7 percent,
respectively), and a similar percent unsaturation to negative control cultures coexpressing BTEH204A and GeoTE-H173A (37.4 ± 3.1 percent) (Figure 5.8). When C17 cyclopropane fatty acids,
which are derived from methylation across the double bond of 16:1 fatty acids by cyclopropane
fatty acid synthase (Cfa), are included in a combined C16-C18 unsaturated plus cyclic fatty acid
179
percentage, this results in an increased value for the GeoTE and BTE expressing cultures
(60.4 ± 5.0 percent) versus the GeoTE and BTE-H204A expressing cultures (57.5 ± 0.2 percent)
and GeoTE-H173A and BTE-H204A expressing cultures (50.2 ± 0.6 percent), with GeoTEH173A and BTE-expressing cultures maintaining the highest percentage of cyclic plus
unsaturated fatty acids (71.4 ± 1.2 percent). These trends are consistent with the anticipated
highest depletion of saturated acyl-ACPs by BTE, followed by BTE and GeoTE co-expression,
followed by GeoTE alone. GeoTE and BTE co-expressing cultures exhibit the highest
percentage of bound fatty acids as cyclic C17. Increased expression of cfa is known to occur
under stationary phase and acidic conditions [Wang 1994, Chang 1999], with implicated
transcriptional regulators including RpoS, Crp, Fur, and FNR [Keseler 2011], however no further
analysis was conducted in this study to determine the cause of increased cyclic C17 in these
cultures.
% unsaturated C16-C18 (base)
80
70
60
50
40
30
20
10
0
BTE
GeoTE
(-)
(-)
(-)
(+)
(+)
(-)
(+)
(+)
Figure 5.8 Percentage unsaturated membrane lipids in cultures expressing combinations of BTE
and GeoTE. Membrane-bound lipids were analyzed by base-catalyzed methylation of fatty acid extracts
from samples taken from cultures 24 h post-inoculation. C17∆ was not included in the unsaturated
percentage. GeoTE-expressing cells have a reduced unsaturated content relative to BTE-expressing cells.
GeoTE and BTE co-expressing cells also have reduced unsaturated content.
180
Microarray analysis and qPCR to quantify levels of fabA transcript previously indicated
strongly repressed levels of fabA and fabB in BTE-expressing cultures compared to BTEH204A-expressing cultures, which we postulated was due to depletion of saturated acyl-ACPs in
BTE-expressing cultures, thereby causing an increase in FabR-mediated repression of these
genes (Chapter 3). If expression of GeoTE, or co-expression of GeoTE and BTE is resulting in a
shift toward a larger ratio of saturated to unsaturated acyl-ACPs, it would therefore be expected
that fabA and fabB levels would be restored to higher levels than in cultures only expressing BTE.
Shake flask cultures of RL08ara were grown with the plasmid combinations described previously,
and RNA was harvested from flash frozen cell pellets collected 3.25 hours post-inoculation
(approximately 1 hour post-induction). The levels of fabA and fabB transcript were quantified
by qPCR from biological triplicate cultures of each strain (Figure 5.9). Indeed, in both the
GeoTE/BTE-H204A and GeoTE/BTE coexpressing cultures, levels of fabA and fabB were
restored to similar levels as present in the control culture (BTE-H204A and GeoTE-H173A coexpressed), and were higher than levels present in the BTE/GeoTE-H173A coexpressing cultures.
Expression of fabB was higher with statistical significance (P < 0.05) in the GeoTE/BTE-H204A
expressing cultures than in the GeoTE-H173A/BTE-H204A expressing cultures, indicating that
expression of a thioesterase with this degree of unsaturated acyl-ACP substrate specificity may
be resulting in a lower amount of FabR repression than is present in wild-type cells. That
expression of fabA was not higher with statistical significance is also consistent within this
framework, as fabA is repressed less strongly, and by analogy also de-repressed less strongly, by
FabR than fabB due to weaker binding of FabR with the promoter region upstream of fabA
relative to the promoter region upstream of fabB [Feng 2011].
181
*
2.0
fabA
relative expression level
fabB
1.5
1.0
*
0.5
*
0.0
BTE
GeoTE
(-)
(-)
(-)
(+)
(+)
(-)
(+)
(+)
Figure 5.9 Normalized transcript levels of fabA and fabB in strains expressing combinations of
BTE and GeoTE. qPCR was performed using cDNA synthesized from RNA extracted from cell pellets
harvested approximately 1 hour post-induction. Levels of fabA and fabB were only decreased from cells
expressing only non-functional thioesterases in BTE-expressing cells. Levels were statistically the same
or higher in cells expressing GeoTE or GeoTE and BTE. * = P-value < 0.05 for Cq values compared
against fabA or fabB in cultures expressing only non-functional thioesterases (BTE- GeoTE-).
5.3.6 Cell growth and viability analysis of GeoTE and BTE expressing cultures
SYTOX Green staining and flow cytometry analysis of cell samples collected 8 h postinoculation revealed differences in the percentage of intact cells between strains expressing the
two thioesterases (Figure 5.10). BTE-H204A and GeoTE-H173A expressing cells exhibited a
typical non-functional thioesterase expressing value of 94.8 ± 1.3 percent intact, while cells
expressing only functional BTE were 34.6 ± 2.1 percent intact. Expression of only functional
GeoTE, however, increased the percentage of intact cells to 77.8 ± 5.2 percent, while coexpression of functional GeoTE and BTE resulted in 67.7 ± 4.5 percent intact cells. This latter
result may reflect that the cells have not entered an equivalent stationary phase to the other
cultures at the selected sampling time, as there are fewer cells as measured by flow cytometry
182
(Table 5.1), larger cell size as measured by forward scatter (data not shown), and altered
growth (OD600) as monitored in a 96-well plate in a plate reader (Figure 5.11).
100
BTE-H204A
BTE
90
percent intact
80
70
60
50
40
30
20
10
0
GeoTE-H173A
GeoTE
Figure 5.10 Percent intact cells in strains expressing combinations of BTE and GeoTE. Cells were
harvested 8 h post-inoculation and stained with SYTOX Green before analysis by flow cytometry.
Cultures only expressing BTE were the least intact. Cultures only expressing GeoTE, and both GeoTE
and BTE appeared primarily intact.
0.6
BTE- GeoTEBTE- GeoTE+
BTE+ GeoTEBTE+ GeoTE+
0.5
OD600
0.4
0.3
0.2
0.1
0
0
2
4
6
8
10
time (h)
Figure 5.11 Plate reader growth curves of strains expressing combinations of BTE and GeoTE.
Cells were grown in a 96-well plate at 37°C with shaking and otherwise as described for shake flask
cultures. Reduced OD600 was observed 8 h post-inoculation in GeoTE and BTE co-expressing cultures,
correlating with reduced cell counts by flow cytometry in shake flask cultures.
183
A percent viable cells analysis obtained by dividing CFU/mL by the number of flow
cytometry events measured per mL of original cell culture generally followed trends observed by
the SYTOX Green assay for percent intact cells (Table 5.1). While non-FFA overproducing
BTE-H204A and GeoTE-H173A expressing cells were 90 ± 14 percent viable, BTE and GeoTEH173A expressing cells were only 5 ± 1 percent viable, and BTE-H204A and GeoTE expressing
cells were 45 ± 12 percent viable. In contrast to the SYTOX Green assay, BTE and GeoTE
expressing cells were only 3 ± 2 percent viable. As we have observed previously (Chapter 3),
plated cells from BTE-expressing cultures exhibited a wide variety of different colony sizes after
1 night incubation at 37°C, and 1 additional night incubation at room temperature was required
for all colonies to become visible. Plated cells from GeoTE-expressing cultures exhibited an
even larger distribution of colony sizes, with new colonies appearing after 1 night at 37°C and 4
nights at room temperature. BTE plus GeoTE expressing cultures similarly required a total of 5
days before new colonies stopped appearing. Therefore it appears that large percentages of
strains expressing one or both thioesterases are non-lysed but also non-culturable on LB agar.
Table 5.1 Viability analysis of strains expressing combinations of BTE and GeoTE. Cultures of
RL08ara harboring combinations of pBAD33-BTE-H204A (BTE-) or pBAD33-BTE (BTE+), and
pBAD18-GeoTE-H173A (GeoTE-) or pBAD18-GeoTE (GeoTE+) grown in LB + 0.4% glycerol and
antibiotics at 37°C for 8 hours. Reported values are forward scatter triggered flow cytometry events per
mL original culture volume of SYTOX Green stained cells, plate counts (CFU/mL) after 5 days
incubation, and plate counts normalized to flow cytometry events (CFU event-1), as another estimate of
percentage live cells.
BTE
GeoTE
Flow cytometer
(events mL-1 )
Plate counts
(CFU mL-1)
Normalized CFU
(CFU event-1)
(-)
(-)
(+)
(+)
(-)
(+)
(-)
(+)
(5.75 ± 0.43) ×109
(4.93 ± 0.15) ×109
(3.61 ± 0.50) ×109
(1.37 ± 0.76) ×109
(5.2 ± 0.7 ) ×109
(2.20 ± 0.60) ×109
(1.9 ± 0.4 ) ×108
(4.0 ± 1.6 ) ×107
0.90 ± 0.14
0.45 ± 0.12
0.05 ± 0.01
0.03 ± 0.02
184
5.3.7 Fatty acid production of GeoTE and BTE expressing cultures
After 8 h growth, cultures expressing only functional GeoTE exhibit a higher total fatty acid
titer (513 ± 12 mg/L) than cultures expressing only functional BTE (364 ± 42 mg/L; Figure 5.12).
This higher productivity could represent a faster rate of production per cell, or could be due to a
larger percentage of viable, intact cells in these cultures. Cultures expressing both functional
thioesterases produced 255 ± 117 mg/L after 8 hours, possibly reflecting the lower cell counts
and reduced growth in these cultures. After 24 hours, titers for GeoTE, BTE, and both GeoTE
and BTE expressing cultures reached 569 ± 8, 651 ± 57, and 653 ± 60 mg/L respectively (Figure
5.12). The fatty acid composition of cultures expressing both thioesterases was intermediate
between cultures expressing only functional GeoTE or BTE.
5.4 Discussion
The complex physiological and metabolic perturbations that result from free fatty acid
production render it challenging to determine the barrier preventing higher yields. In previous
work, we observed cell lysis, decreased viability, increased membrane unsaturated fatty acid
content, and transcriptional evidence of membrane stresses, cell depolarization, and impaired
aerobic respiration (Chapter 3). Furthermore, greatly varying titers (and correspondingly, yields
in batch cultures) have been reported by a number of groups working toward FFA
overproduction (see Table 1.2 and section 1.4.2), with no clear rationale to explain the
differences observed or to guide the development of improved strains with yields exceeding 30%
of maximum theoretical in minimal medium. In vitro experiments in which E. coli cell lysates
were supplemented with additional enzymes and cofactors [Liu 2010a], and full reconstitution of
FFA biosynthesis [Yu 2011] have yielded significant insights into potential
185
A
600
C18:0
C18:1
C17delta
C16:0
C16:1
C14:0
C14:1
C13:0
C12:0
C12:1
C11:0
C10:0
C10:1
C9:0
C8:0
fatty acid titer (mg/L)
500
400
300
200
100
0
BTE
GeoTE
B
(-)
(-)
(-)
(+)
(+)
(-)
(+)
(+)
fatty acid titer (mg/L)
700
C18:0
C18:1
C17delta
C16:0
C16:1
C14:0
C14:1
C13:0
C12:0
C12:1
C11:0
C10:0
C10:1
C9:0
C8:0
600
500
400
300
200
100
0
BTE
GeoTE
(-)
(-)
(-)
(+)
(+)
(-)
(+)
(+)
Figure 5.12 Fatty acid titers and compositions in strains expressing combinations of BTE and
GeoTE. (A) 8 h post-inoculation; (B) 24 h post-inoculation. Cultures expressing only functional GeoTE
exhibited the highest titer at 8 h, with little further increase observed after 24 h. Nearly equivalent titers
were reached after 24 h in cultures expressing only BTE, or co-expressing BTE and GeoTE.
metabolic limitations and basic biochemistry. However, they lack the dynamic context of
transcriptional and translational regulation and the connection between fatty acid biosynthesis
and the remainder of the cellular metabolism. For instance, achieving the maximum theoretical
186
yield of FFAs requires aerobic respiration and the use of the membrane-bound
transhydrogenase PntAB (as determined by constraint-based modeling, see Chapter 2), and
therefore maintenance of inner cell membrane integrity and properties.
We previously identified the substantially increased membrane unsaturated fatty acid content
of BTE-expressing, fatty acid overproducing cells as being a potential cause of the increase in
non-intact and non-viable cells as measured by SYTOX Green staining and plate counting
(Chapter 3). This was largely due to the observation of a greatly reduced abundance of fabA and
fabB transcripts, which implicated a cellular response to attempt to mitigate the elevated ratio of
unsaturated to saturated acyl-ACP intermediates via FabR. Despite FabR-mediated repression,
unsaturated membrane fatty acid content was still highly elevated.
To further probe the
importance of the native level of FabR-mediated repression of fabA and fabB transcription, fabR
was deleted in both FFA-overproducing (BTE-expressing), and non-FFA-overproducing (BTEH204A-expressing cultures). This resulted in a lack of repression of fabA and fabB, which were
increased in expression in both BTE- and BTE-H204A-expressing cultures.
As a result,
unsaturated C16-C18 content was elevated, with a particularly drastic increase evident in BTEexpressing cells. This increase in unsaturated C16-C18 fatty acids was also accompanied by a
much lower percentage of intact cells than present in the fabR+ BTE-expressing strain, and a
decrease in FFA production. This criticality of FabR strongly suggests that its level of repression
is insufficient in fabR+ BTE-expressing cultures, and that the cells cannot down-regulate fabA
and fabB to a sufficient degree to counter the effects on the acyl-ACP pool resulting from BTE
expression.
Overexpression of FabR is one potential strategy to increase the level of repression if
insufficient levels of FabR are present. When overexpressed on a low-copy plasmid (pBAD33*),
187
-
unsaturated content was restored in a fabR , BTE-expressing strain to slightly lower levels
than in a fabR+, BTE-expressing strain, but still significantly higher than in a fabR+ BTE-H204Aexpressing strain. However, severe impacts were observed on growth, perhaps due to nonspecific DNA binding by very high levels of FabR. A more promising strategy may entail
promoter engineering of PfabA and PfabB to achieve increased repression by FabR.
A second strategy pursued in this study involved modulating the unsaturated acyl-ACP pool
by co-expression of BTE with an unsaturated acyl-ACP thioesterase.
While it may not
necessarily be ideal to alter the FFA product profile, unsaturated species are hydrogenated and
decarboxylated to alkanes when a downstream ex-vivo catalytic process is implemented (Chapter
2). A thioesterase with a more unsaturated FFA product profile (unsaturated C12 and C14) from
Geobacillus sp. Y412MC10 (GeoTE) [Jing 2011] was codon-optimized and heterologously
expressed in E. coli. GeoTE-expressing cultures, and cultures co-expressing BTE on a low-copy
plasmid and GeoTE on a medium-copy plasmid, exhibited reduced percentages of membrane
unsaturated fatty acids and increased levels of fabA and fabB transcripts compared to cultures
only expressing BTE. Furthermore, cell lysis was greatly reduced in early stationary phase in
GeoTE-expressing cultures and cultures expressing both thioesterases, although the latter
resulted in lagged growth and lower viable cell counts. That higher 24 h titers were not achieved
in cultures exhibiting restored unsaturated membrane content relative to BTE-expressing cultures,
despite a higher productivity for the GeoTE-expressing culture during the first 8 hours, likely
indicates the existence of another barrier preventing further FFA overproduction. Thus reducing
cell lysis by restoring the unsaturated fatty acid content is likely only one of multiple genetic
modifications that will need to be engineered in order to increase FFA yields and titers.
188
Varying ratios of unsaturated to saturated acyl-ACP substrate specificities of different
thioesterases may be one source of the variable titers reported in different engineered FFAoverproducing strains, in which the primary distinguishing feature is choice of thioesterase. It
may explain results such as those observed by Khosla et al. [2008], wherein co-expression of a
cytosolic form of E. coli TesA (thioesterase I) and the plant FatB-type thioesterase from
Cinnamomum camphorum (CcTE) produced higher titers than expression of CcTE alone. The
substrate specificity for CcTE is most enriched in saturated C14 [Yuan 1995], whereas TesA has
a broad C12-C18 specificity that is highest toward unsaturated C16 and C18 fatty acids [Cho 1995].
It is perhaps supportive that the increases in titer from co-expressing CcTE and TesA were not
simply additive, and may have instead been the result of reduced cell lysis and increased viability.
High titers also reported from a strain designed to undergo reversal of β-oxidation and
overexpressing an acyl-CoA thioesterase (E. coli FadM) [Dellomonaco 2011], may be due to this
pathway being independent from fatty acid biosynthesis and therefore lacking impacts to
membrane-bound fatty acids such as those imposed by acyl-ACP thioesterase expression.
Membrane fatty acid content and other physiological characteristics such as those measured here
are not typically reported in FFA overproduction studies, however we suggest their increased use
to gain a unifying understanding of FFA production bottlenecks in the context of different
engineered strains. This level of understanding is necessary to further advance the field of
advanced biofuels.
5.5 Conclusion
The unsaturated fatty acid content of the membrane is a key parameter that influences cell
lysis under conditions of endogenous FFA production. While FabR is a major regulator, its level
189
of repression appears to be insufficient to prevent elevated unsaturated acyl-ACP pools as a
result of expression of BTE, which hydrolyzes primarily saturated acyl-ACPs. Work described
in this chapter further solidified the relationship between acyl-ACP thioesterase expression and
membrane unsaturated fatty acid content.
It was also shown that acyl-ACP pools can be
modulated by thioesterase selection. Expression of a thioesterase with primarily unsaturated
acyl-ACP substrate specificity (GeoTE) reduced membrane unsaturated fatty acids, increased
cell viability, reduced lysis, and increased FFA productivity during the first 8 hours of growth.
Further strategies geared toward modulation of membrane lipid content for improved cellular
fitness will be discussed as part of Chapter 6.
190
Chapter 6: Conclusions and future directions
6.1 Summary of findings
In this thesis, we followed a typical iterative metabolic engineering process of first
introducing genetic modifications to develop a strain that produced a desired target compound
(free fatty acids) (Chapter 2). Second, the strain was characterized to determine the impact of the
engineered pathway on cell physiology (Chapter 3). This data was used to generate hypotheses
describing potential barriers to higher production (Chapter 3).
Next, experiments were
performed to test hypotheses and gain missing fundamental information (Chapters 4 and 5).
Finally, genetic modifications were introduced using the information obtained to engineer a
second generation strain with improved production characteristics (Chapters 3, 4, and 5).
It is widely acknowledged within the metabolic engineering community that the simplest step
is development of an initial strain that can produce some small quantity of a target compound,
while more difficult steps involve determination of production bottlenecks and engineering the
cells to circumvent these barriers and ultimately approach the maximum theoretical yield. This
view was borne out in the body of work discussed here, as development of a second generation
strain with improved overall production characteristics has not yet been achieved. However,
significant advances have been made in understanding the physiology and metabolism of FFA
overproduction, with findings likely to prove useful for engineering microorganisms for
production of other high energy density biofuels and commodity chemicals such as those
discussed in Chapter 1.
In Chapter 2, a platform strain was developed that overproduces medium-chain length FFAs
(primarily 12:0) with total fatty acid titers of approximately 0.8 g/L from a rich undefined
191
medium supplemented with glycerol. The expression level of the acyl-ACP thioesterase
required for FFA overproduction (BTE) was optimized, and similar low levels of expression
were used in the remainder of the described work. That a higher level of expression of BTE
resulted in impaired growth and reduced FFA titers provided a first hint of possible physiological
challenges. Additional overexpression of acetyl-CoA carboxylase (ACC) resulted in marginal
increases in fatty acid titer (mostly in longer-chain species), indicating that generation of
malonyl-CoA was not a bottleneck. A process to convert FFAs to a potential biofuel was
demonstrated by coupling a biological process with heterogeneous chemical catalysis. This
process featured extraction of FFAs from the culture medium into a separate alkane phase, which
could be passed over a heterogeneous catalyst bed to fully decarboxylate FFAs to their
corresponding alkanes. Such a continuous or semi-continuous process could be designed to
recycle a portion of the alkanes produced by decarboxylation as the solvent for subsequent
extraction. Such a process (or any process for biofuel applications) would require cells to
produce FFAs at yields approaching the maximum theoretical limit to be economically viable.
Therefore subsequent work focused on identifying and circumventing barriers to increasing FFA
yield.
In Chapter 3, a differential physiological characterization was performed between BTE (FFAoverproducing) and BTE-H204A-expressing (non-overproducing) strains, revealing drastic
differences in cell viability, lysis, and morphology that could not be mimicked by exogenous
addition of FFAs. To determine a functional basis for these observations, two large sampling
experiments conducted under different sets of environmental conditions were performed to
determine differential transcript, protein, lipid, and other selected metabolite profiles between
BTE-expressing and BTE-H204A-expressing cultures. Membrane stress responses including
192
induction of the MarA/Rob/SoxS regulons and phage shock proteins were observed, in
addition to increased expression of genes involved in aerobic respiration.
When certain
regulators of these processes were disrupted, cells displayed markedly decreased fitness,
confirming their role in maintenance of cell viability under conditions of FFA overproduction.
Unfortunately, second generation strains that overexpressed transcription factors or phage shock
proteins failed to restore viability or lead to increased FFA production.
These findings led to the hypothesis that endogenous FFAs generate toxic intracellular effects,
perhaps as a result of their accumulation in the membranes more rapidly than their export. This
led to the work described in Chapter 4. Native genes that were hypothesized to be involved in
FFA export or secretion were deleted and the resulting strains were assayed for further reduced
viability, increased lysis, or decreased FFA production compared to the baseline FFAoverproducing strain. This strategy successfully identified a number of genes that appear to be
involved in FFA excretion and tolerance toward exogenously added FFAs that exhibit higher
toxicities than exogenously added laurate. Other than the transcription factor Rob (identified in
Chapter 3), these were all components of RND-type drug efflux pump systems that have
previously been shown to confer resistance to SDS, which is structurally similar to the
predominant FFA product, lauric acid. These efflux pump systems are AcrAB-TolC, EmrABTolC, MdtEF-TolC, and MdtABC-TolC. Ongoing work to overexpress these efflux pumps to
achieve rates of FFA secretion above that conferred by native activity of AcrAB has thus far
proven challenging. Future suggested work in this area is discussed here in Chapter 6.
A second major hypothesis developed in Chapter 3 is that an elevated unsaturated membrane
lipid content is altering membrane properties and affecting cell physiology in BTE-expressing
cells, which was counter to the highly repressed expression levels of genes of unsaturated fatty
193
acid biosynthesis. This suggested a metabolic incapability of the cells to counter an increase
in unsaturated acyl-ACP pools that result from BTE expression. We hypothesized that the
elevated unsaturated content could be responsible for exacerbation of membrane stresses, cell
lysis, and further impairment of aerobic respiration. This led to the work in Chapter 5 which
sought to conduct further experiments to demonstrate the correlation between unsaturated
membrane lipid content and impaired physiological characteristics. In particular, deletion of the
gene encoding the transcriptional repressor of unsaturated fatty acid biosynthetic genes (fabA and
fabB) led to unsaturated membrane lipid content in BTE-expressing cells that was over 80%, and
these cultures also had an extreme degree of cell lysis.
These observations led to the
implementation of two strategies: overexpression of FabR, and co-expression of an acyl-ACP
thioesterase with a predominantly unsaturated acyl-ACP substrate specificity. While preliminary
work involving overexpression of FabR was unsuccessful, the unsaturated membrane lipid
content was greatly reduced by the second strategy. The strain only expressing this alternative
thioesterase (GeoTE) exhibited higher viability, reduced cell lysis, and higher fatty acid titers
after 8 h, while a strain co-expressing both BTE and GeoTE exhibits a nearly 2.5-fold increase in
per-cell fatty acid production. However, overall improved titers and yields have not yet been
achieved. Alternative strategies for synthetic biology control of membrane lipid properties and
suggestions for future experiments are discussed in this chapter.
6.2 Recommendations for future work
6.2.1 Optimization and engineering of efflux pump overexpression
In Chapter 4, native proteins responsible for conferring tolerance toward endogenous FFA
194
production were identified by investigating the impact of gene and operon deletions on
viability, cell lysis, and FFA production. All of the genes that conferred increased tolerance to
FFA-overproducing cells encoded inner membrane efflux pumps of the resistance-nodulationdivision family (RND): AcrAB, MdtEF, MdtABC (mdtD was included in the deletion), and
EmrAB. Additionally, deletion of the gene encoding the outer membrane pore TolC resulted in
essentially abolished FFA production. TolC acts as the outer membrane component of all of the
aforementioned inner membrane and periplasmic linker subunits of each efflux pump.
Overexpression of each efflux pump, with and without additional overexpression of TolC, failed
to either increase FFA production in a BTE-expressing host strain, or to confer increased
tolerance toward exogenously added FFAs in cells harboring a chromosomal copy of acrAB.
A number of possible explanations could explain the lack of improved production or lack of
increased minimum inhibitory concentration, despite demonstration of functional overexpression
of the efflux pumps. One possibility is that the required pathway for membrane protein insertion
is saturated. A number of translocase and insertase systems in E. coli are responsible for
insertion of nascent peptides into the inner membrane, with only the SecYEG translocase and
YidC insertase characterized in detail. A typical initial step is the binding of signal recognition
particle (SRP) to hydrophobic segments of the nascent peptide during translation, which directs
the complex to the membrane where SRP binds with FtsY [Dalbey 2011]. Once directed to the
inner membrane, translocation and insertion is accomplished by SecYEG and YidC, which can
function together in a variety of ways to achieve insertion and assembly of a peptide chain in the
inner membrane (Figure 6.1). Overexpression of membrane or secreted proteins in E. coli is a
recognized challenge, and various systems have been designed or selected for membrane protein
production by overexpression of insertases or through the use of isolated strains harboring better
195
properties [eg. Baneyx 2004, Wagner 2007, Wagner 2008]. Additionally, chaperones are
often required for efficient membrane protein folding or assembly of subunits [Baneyx 2004,
Nannenga 2011], and it would need to be determined whether any chaperones were required for
these purposes for the identified efflux pumps.
Figure 6.1 SecYEG and YidC interactions with nascent peptides to achieve insertion and assembly
in the inner membrane. (a) SecYEG mediating insertion of a nascent peptide and translocation of
membrane domains. (b) YidC and SecYEG acting to properly fold and insert trans-membrane segments to
form protein subunits. (c) YidC acts independently to insert some membrane proteins. (d) Sequential
functioning of SecYEG and YidC for separate N- and C-terminal trans-membrane domains. Adapted
from [Dalbey 2011].
One interesting observation was that native level expression of AcrAB can confer resistance
to up to 50 mg/mL SDS whereas a strain lacking acrAB can tolerate less than 0.1 mg/mL, more
than a 500-fold decrease. The dynamic range of resistance conferred by AcrAB toward FFAs is
lower, with native expression conferring resistance to 5 g/L octanoate and the acrAB deletion
strain exhibiting resistance to 3 g/L octanoate, a 1.67-fold decrease. A 10-fold decrease in
resistance is observed for decanoate with loss of acrAB.
While this may be due to
complementation of FFA efflux activity in the acrAB- strain by other pumps such as EmrAB and
MdtEF, it may also be due to a lower affinity of AcrAB toward FFAs than toward SDS. Crystal
structures of AcrB bound to different antibiotic substrates have revealed three substrate channels
196
that deliver substrates to a proximal distal pocket, followed by transport to a distal pocket
[Murakami 2002, Nakashima 2011] (Figure 6.2).
With knowledge of the locations of the
substrate binding pockets and the residues that may be interacting with hydrophobic substrates, it
may be possible to rationally engineer a small library of mutagenized residues and test them
individually, or to conduct rounds of error prone PCR in tandem with gene shuffling to generate
libraries for selection on 5 g/L octanoate or 2 g/L decanoate in an acrAB- strain to identify
beneficial mutations that increase the affinity of AcrB toward medium-chain length FFAs.
Figure 6.2 Detail of erythromycin binding in the AcrB proximal pocket, and schematic of substrate
uptake and extrusion by AcrB. (a) Erythromycin is shown in yellow. Key residues important for
binding were identified by site-directed mutagenesis. (b) Peristaltic mechanism of AcrB driven by
conformational changes on binding of substrate in the proximal and distal pockets. Adapted from
[Nakashima 2011].
197
6.2.2 Synthetic biology control of membrane lipid content
In Chapter 5, it was demonstrated that unsaturated fatty acid incorporation into membrane
lipids, while elevated in BTE-expressing cultures relative to BTE-H204A-expressing cultures, is
further dramatically increased in BTE-expressing cultures with fabR deleted from the
chromosome. This further increase in unsaturated membrane fatty acids also resulted in a
dramatically decreased percentage of intact cells as measured by SYTOX Green staining. An
acyl-ACP thioesterase with higher selectivity toward unsaturated acyl-ACPs (GeoTE) was
heterologously expressed as a demonstration that both unsaturated membrane fatty acid content
and the percentage of non-intact cells could be decreased.
Expression or co-expression (with BTE) of an acyl-ACP thioesterase with a higher specificity
toward hydrolysis of unsaturated acyl-ACPs was only one of the envisioned strategies for
achieving higher tolerance to FFA overproduction via modulation of membrane unsaturated fatty
acid content (Figure 6.3A). Another envisioned strategy is the heterologous expression of a fatty
acyl cis-trans isomerase (Cti) from Pseudomonas putida, which isomerizes cis-unsaturated acyl
groups (16:1∆9 and 18:1∆11) on phospholipids acyl groups to the trans form [Holtwick 1997]
(Figure 6.3B). Cti activity has been implicated in tolerance of P. putida to toluene and ethanol
[Heipieper 1994, Junker 1999]. It is believed that this tolerance is primarily conferred by a
decreased membrane fluidity, as trans unsaturated fatty acids would exhibit properties
intermediate between cis-unsaturated and saturated species [Ramos 2002]. E. coli does not
natively synthesize trans-unsaturated fatty acids, but it was hypothesized that importing this
capability could reduce membrane fluidity and therefore reduce cell lysis as a result of BTEexpression. Preliminary results (shown below) demonstrate successful heterologous expression
of Cti in E. coli. From extracts of these cultures, trans-unsaturated C16 and C18 fatty acids were
198
detected. Despite the presence of these species, no decrease in cell lysis or increase in FFA
production is observed. However suggestions will be made for further optimization for use in
FFA overproducing strains and other potential applications of this approach will be discussed.
A third envisioned strategy involves a complete rewiring of unsaturated fatty acid
biosynthesis in E. coli. This would involve first deleting fabA, as FabA is responsible for
introducing the cis-unsaturation during fatty acid elongation. Because unsaturated fatty acids are
required for a functional cell membrane, an alternative route to generating them can be
introduced in the form of a phospholipid acyl desaturase (Figure 6.3C), which catalyzes the
formation of a cis-unsaturation on phospholipids already inserted into the membrane. This route
for unsaturated fatty acid biosynthesis is utilized in cyanobacteria, Bacillus subtilis, and most
higher organisms, the majority of which lack FabA homologues [Aguilar 2006]. One wellcharacterized phospholipid acyl desaturase which has been successfully heterologously
expressed in E. coli is DesA from Bacillus subtilis, however it is a ∆5 desaturase that produces
unsaturated species not ordinarily present in E. coli including 16:1∆5 and 18:2∆5∆11 [Bonamore
2006]. An alternative ∆9 desaturase is DesC from Synechococcus sp. PCC 7002 [Sakamoto
1997], which would be expected to generate one of the major native unsaturated fatty acids in E.
coli, 16:1∆9, but not the other predominant species, 18:1∆11. The expression level of the
desaturase is anticipated to need careful control to maintain proper physiology. The optimal
expression level can be achieved by utilizing existing synthetic biology tools for controlling
transcription. Preliminary work demonstrating successful heterologous expression of B. subtilis
DesA and Synechococcus sp. PCC 7002 DesC will be presented, as well as efforts toward
building a strain where the level of membrane fatty acid unsaturation can be controlled by the
199
A
B
cis-trans
isomerase
C
Figure 6.3 Three envisioned strategies for modulation of membrane fatty acids in BTE-expressing
strains. (A) Co-expression of a predominantly unsaturated acyl-ACP thioesterase would be expected to
restore the ratio of saturated (s) and unsaturated (u) acyl groups that are incorporated into phospholipids.
This strategy was undertaken in Chapter 5 with GeoTE. (B) Heterologous expression of a phospholipid
acyl cis-trans isomerase will convert many of the cis-unsaturated fatty acids enriched in the membrane to
trans-unsaturated fatty acids, which have physical properties intermediate between cis-unsaturated and
saturated species. (C) Design a strain that lacks FabA and is therefore unable to synthesize unsaturated
acyl-ACPs, and instead synthesize unsaturated phospholipid acyl groups using a desaturase expressed at a
low level.
addition of inducing agents. Future experiments and other possible strategies will also be
discussed.
200
6.2.2.1 Materials and methods
6.2.2.1.1 Strain and plasmid construction
Strain DY330 fabA::kan was constructed by amplifying the kan cassette flanked by FRT sites
from pKD13 template plasmid using primers 109 and 110 (Appendix III), which contain 40-bp
homologous regions on the 5' overhangs to regions immediately upstream and downstream
(including the stop codon) of fabA. The resulting PCR product was DpnI treated, gel extracted,
and electroporated into strain DY330. Strains that had successfully integrated the kan cassette in
place of fabA were selected on LB plates containing 25 µg/mL kanamycin and 0.1% w/v sodium
oleate (TCI America, city, state) pre-mixed with a final concentration of 1% v/v Brij 35
(manufacturer, city, state). Deletion of fabA was confirmed by colony PCR using primers 111
and 112, and by lack of ability to grow on LB plates not supplemented with sodium oleate.
The cti gene encoding fatty acyl cis-trans isomerase from Pseudomonas putida was amplified
from P. putida KT2440 genomic DNA using primers 123 and 124, which introduce XbaI and
HindIII sites on the 5' and 3' ends, as well as an artificial ribosome binding site and spacer
sequence on the 5' end prior to the cti start codon. The PCR product and pBAD33* vector were
digested with XbaI and HindIII and ligated to form pBAD33*-cti.
Plasmid pBAD33*-desA-His was generated by first PCR amplifying desA from genomic
DNA isolated from Bacillus subtilis 3610 using primers 113 and 114. The 5' overhang of 113
contains an XmaI site, artificial ribosome binding site and spacer sequence, and a T to G
substitution at position 6 in the desA reading frame to eliminate an out-of-frame CTG start codon
with a high predicted translation initiation rate from the Ribosome Binding Site Calculator [Salis
2009].
Primer 114 contains a 5' overhang that introduces an XbaI site and C-terminal
hexahistidine tag immediately prior to the stop codon.
201
The desC gene encoding a phospholipid acyl desaturase was amplified from genomic
DNA isolated from Synechococcus sp. PCC7002 using primers 115 and 116, which introduce an
XbaI site and artificial ribosome binding site and spacer sequence, and HindIII site respectively.
The PCR product and vector were digested with XbaI and HindIII and ligated to form pBAD18desC.
Template plasmids for integration of phospholipid desaturases (pBAD33*-kan and
pBAD33*-desA-His-kan) were generated by digesting the kan cassette flanked by FRT sites
from pBAD34-BTE-kan using XhoI and gel extracting the appropriate size fragment. Plasmid
pBAD34-BTE-kan was originally generated by amplifying the kan cassette from pKD13
[Datsenko 2000] using primers 25 and 26, which contain XhoI sites on 5' overhangs, and ligating
into plasmid pBAD34-BTE (Chapter 2).
The kan cassette was ligated into the XhoI site
downstream of araC-C280* in pBAD33* and pBAD33*-desA-His, and clones were selected for
which kan is transcribed in the opposite direction as araC-C280*. Plasmid pBAD33*-desC-kan
was generated by digesting and gel extracting the fragment containing desC and part of araC
from pBAD18-desC using AflII and HindIII, and ligated into the corresponding sites in plasmid
pBAD33*-kan.
6.2.2.1.2 Bacterial cultivation conditions
Cultures were grown in 250 mL baffled shake flasks containing 50 mL LB medium
supplemented with 34 µg/mL chloramphenicol at 30°C or 37°C, as specified in the text, when
not testing FFA overproduction. Cultures were induced at an OD600 of 0.2 with 0.2% w/v Larabinose. When FFA overproduction was also being tested, cultures were supplemented with
202
0.4% v/v glycerol and 100 µg/mL ampicillin for maintenance of pTrc99A-BTE or pTrc99ABTE-H204A, and also induced with 50 µM IPTG.
To test complementation of desaturases in DY330 fabA::kan, transformants were maintained
on LB agar plates supplemented with 34 µg/mL chloramphenicol and 0.1% w/v sodium oleate,
streaked onto plates additionally supplemented with 0.2% w/v L-arabinose to induce expression
of the desaturase, and subsequently streaked onto plates containing only chloramphenicol and Larabinose to test the ability of desaturases to complement deletion of fabA.
6.2.2.1.3 Other methods
Fatty acid extraction and methylation, GC/MS analysis, flow cytometry analysis, and plate
counting were performed as previously described in Chapters 2, 3, 4, and 5. The trans-18:1∆11
FAME was identified and quantified using a series of external standards containing both cis and
trans-vaccenic acid methyl esters (Sigma). The trans-16:1∆9 fatty acid was identified as a new
peak that eluted just after the cis species, similar to trans-18:1∆11. As a standard was not
commercially available, its concentration was estimated by applying the sensitivity ratio between
the cis- and trans-18:1∆11 species and assuming the same sensitivity ratio between 16:1∆9
species.
6.2.2.2 Results
6.2.2.2.1 Functional expression of Cti in E. coli
Functional expression of Cti was validated by fatty acid extraction, methylation, and GC/MS
analysis to identify the appearance of trans-unsaturated fatty acids in strain RL08ara harboring
plasmids pBAD33*-cti, and pBAD33* as an empty vector control.
These species are not
203
naturally present in E. coli. After 8 h and 24 h cultivation in LB medium at both 30°C and
37°C, the appearance of new fatty acid peaks was evident (Figure 6.4). The trans-18:1 species
eluted at the same retention time as a neat standard of trans-vaccenic acid (18:1∆11). A species
here identified as trans-palmitoleic acid (16:1∆9) also appeared with a slightly higher retention
time than cis-palmitoleic acid. Higher percentages of trans-unsaturated fatty acids (tUFAs) were
detected at 37°C than 30°C, in contrast to past reports where tUFAs could not be detected when
Cti was expressed in E. coli at 37°C [Holtwick 1997]. In Cti-expressing cultures grown at 37°C
for 8 h, approximately 30% of unsaturated fatty acids are trans (Figure 6.5). By 24 h, 44% are
converted to tUFAs. Cyclic C17 fatty acid concentrations appear nearly unchanged as a result of
Cti expression.
TIC intensity
pBAD33
pBAD33-cti
cis 16:1
trans 16:1
cis 18:1
trans 18:1
17
19
21
23
time (min)
Figure 6.4 Chromatograms of FAMEs extracted from cultures expressing Cti and an empty vector
control. Fatty acids were extracted from cultures grown at 37°C for 24 hours. Only shown are C16-C18
peaks eluting between 17-24 minutes. Unlabeled peaks are 16:0 (approx 18.5 min), 17∆ (approx 20.2
min), 17:0 internal standard (approx. 20.8 min), an unidentified peak that is not a FAME (immediately
before cis-18:1), and 18:0 (approx. 23.0 min).
6.2.2.2.2 Effect of Cti on FFA overproduction
Shake flask cultures of RL08ara harboring combinations of plasmids pBAD33*-cti or
204
70
percent composition
60
saturated
cis-unsaturated
trans-unsaturated
cyclic
50
40
30
20
10
0
pBAD33* 8 h
cti 8 h
pBAD33* 24 h
cti 24 h
Figure 6.5 Percent composition by mass of saturated, cis-unsaturated, trans-unsaturated, and
cyclic fatty acids from cultures expressing Cti and an empty vector control. Fatty acids were
extracted from cultures grown at 37°C for 8 and 24 h as indicated. Cti-expressing cultures exhibit 44%
conversion of cis to trans-unsaturated fatty acids at 24 h.
pBAD33*, and pTrc99A-BTE-H204A or pTrc99A-BTE, were grown at 37°C in LB medium
supplemented with 0.4% glycerol, 34 µg/mL chloramphenicol, and 100 µg/mL ampicillin, and
induced at an OD600 of approximately 0.2 with 0.2% w/v L-arabinose and 50 µM IPTG to induce
expression of both Cti and BTE or BTE-H204A. Cultures were sampled for plate counting and
SYTOX Green staining 8 h post-inoculation, and for fatty acid extractions 8 h and 24 h postinoculation. Surprisingly, there was no change in percent intact cells (population represented by
green fluorescence > 440) as measured by SYTOX Green staining and flow cytometry analysis
(Figure 6.6) as a result of co-expressing Cti and BTE or BTE-H204A versus the empty vector
control (pBAD33*). Once again, 44% of cis-unsaturated fatty acids were converted to tUFAs in
BTE and Cti co-expressing cultures, despite overall elevated unsaturated fatty acid levels in
typically observed as a result of BTE expression (Figure 6.7). However, despite significant
anticipated effects on on C8-C14 fatty acid (predominantly FFAs) production, despite lower CFU/
205
500
BTE-H204A/pBAD33*
BTE-H204A/cti
BTE/pBAD33*
BTE/cti
450
350
6.E+09
5.E+09
300
CFU/mL
number of events
400
BTE-H204A
BTE
250
200
150
4.E+09
3.E+09
2.E+09
100
1.E+09
50
*
0.E+00
0
0
200
400
600
green fluorescence bin
800
1000
pBAD33*
pBAD33*-cti
strain
Figure 6.6 Green fluorescence histograms from SYTOX Green staining and plate counts from
strains expressing Cti or an empty vector control, and BTE or BTE-H204A. (A) Equivalent
populations of intact (low green fluorescence) and lysed (high green fluorescence) cells are observed after
8 h growth between cultures expressing empty vector and BTE-H204A (98.2 ± 0.6% intact) and Cti and
BTE-H204A (96.9 ± 0.5% intact), and empty vector and BTE (70.0 ± 8.6% intact) and Cti and BTE (61.0
± 8.5% intact). (B) Viable cell counts sampled at 8 h. Cti has no effect on viable cell counts in BTEH204A expressing cultures. Cti expression reduces viable cell counts (from (3.4±1.2)×108 to
(8.7±2.0)×107 CFU/mL) in BTE-expressing cultures with P = 0.021.
mL in the BTE and Cti co-expressing cultures compared to the cultures expressing only BTE
(Figure 6.6).
6.2.2.2.3 Functional expression of two phospholipid desaturases in E. coli
Functional expression of DesA and DesC was validated by fatty acid extraction, methylation,
and GC/MS analysis to identify the appearance of new unsaturated fatty acids or increased levels
of native unsaturated species in strain RL08ara harboring plasmids pBAD33*-desA-His, and
pBAD18-desC as an empty vector control. After 8 h and 24 h cultivation in LB medium at 37°C,
the appearance and enhancement of new fatty acid peaks was evident in cultures expressing
DesA (Figure 6.8). The enhancement in the 16:1 peak is believed to be the formation of 16:1∆5
due to the known activity of DesA [Bonamore 2006]. Another peak eluting immediately before
206
70
percent composition
60
saturated
cis-unsaturated
trans-unsaturated
cyclic
50
40
30
20
10
0
99H/33*
99H/cti
99B/33*
99B/cti
Figure 6.7 Percent composition by mass of saturated, cis-unsaturated, trans-unsaturated, and
cyclic C16-C18 fatty acids from cultures expressing combinations of Cti and BTE or BTE-H204A.
Plasmids are indicated as empty vector negative control (33*), BTE (99B), and BTE-H204A (99H).
Shown are results after 24 h growth at 37°C. Elevated levels of cis-unsaturated fatty acids resulting from
BTE expression led to similarly elevated levels of tUFAs, with 44% of cis species converted to trans.
16:1 is speculated to be 16:2∆5∆9, a result of desaturation of 16:1∆9. While this species has not
been observed by heterologous expression in E. coli previously [Bonamore 2006], it may only be
detectable under environmental conditions when significant amounts of 16:1∆9 is present in the
membrane. It is also possible that this peak is 16:1∆5, and that the enhanced 16:1 peak is only
an increased level of 16:1∆9. An enhancement of the 18:1 peak is also observed, which could
possibly be formation of 18:1∆5, or 18:2∆5∆11 as has previously been observed. Expression of
DesC, a ∆9 desaturase, only results in an increased level of 16:1∆9 at the expense of a decreased
level of its substrate 16:0 (Figure 6.9).
6.2.2.2.4 Complementation of fabA with desaturases
DY330 fabA::kan was constructed as described in Materials and Methods, and was verified
by colony PCR of the fabA locus and by its ability to grow on plates supplemented with 0.1%
207
pBAD33
pBAD33-desA
16:1
TIC intensity
16:2?
18:1
17
19
21
23
time (min)
Figure 6.8 Chromatograms of FAMEs extracted from cultures expressing DesA and an empty
vector control. Fatty acids were extracted from cultures grown at 37°C for 24 hours. Only shown are
C16-C18 peaks eluting between 17-24 minutes. In DesA-expressing cultures, a new unsaturated FAME
appears at an elution time immediately prior to 16:1∆9, which could be evolution of 16:1∆5 or 16:2∆5∆9.
The 16:1∆9 peak area increases, which could also be a contribution from 16:1∆5. The peak assigned to
18:1∆11 also increases, which could potentially be unresolved 18:2∆5∆11. Other unlabeled peaks are as
indicated in Figure 6.6.
TIC
pBAD18
pBAD18-desC
16
18
20
time (min)
22
24
Figure 6.9 Chromatograms of FAMEs extracted from a culture expressing DesC and an empty
vector control. Fatty acids were extracted from a culture grown at 37°C for 24 hours. Only shown are
C16-C18 peaks eluting between 16-24 minutes. In DesC-expressing cultures, an increased level of 16:1∆9
is observed, with a decreased level of 16:0.
w/v oleate, but inability to grow on plates without added oleate.
DY330 fabA::kan was
transformed with pBAD33-C280*-desA-His and pBAD18-desC and transformants were tested
208
for their ability to complement the lack of both fabA and oleate supplementation under
inducing conditions. Unfortunately, DesC appeared to not express at 30°C (which was also
verified in a shake flask experiment), therefore it could not be tested in DY330 fabA::kan.
DY330 can only be maintained at 30°C due to the presence of a temperature-inducible λ Red
prophage in its chromosome. Complementation will only be able to be tested after fabA::kan is
transduced to another strain. Expression of DesA successfully complements the deletion of fabA
in strain DY330 fabA::kan. Transformants grow when streaked on LB agar supplemented with
only chloramphenicol and arabinose, and the same transformant colonies do not grow when
struck on LB agar plates containing only chloramphenicol (Figure 6.10).
+ inducer
- inducer
Figure 6.10 Complementation of growth in a fabA deletion by DesA. Transformant colonies of
DY330 fabA::kan/pBAD33-C280*-desA-His struck on LB agar plates supplemented with arabinose to
induce expression of DesA (left) and lacking arabinose (right). Cells only grow in the presence of
arabinose.
6.2.2.2.5 Further progress in building a strain with rewired unsaturated fatty acid
biosynthesis
Template plasmids were designed and constructed to enable the integration of a cassette
containing desA and desC under control of arabinose-inducible promoters into any desired
chromosomal locus by PCR amplification of the assembled cassette with primers containing 5'
overhangs that are homologous to regions flanking the targeted locus on the chromosome (Figure
209
6.11). The cassette also includes the IPTG-insensitive araC-C280* repressor, and allows for
selection of successful recombinants on kanamycin.
The KanR gene can subsequently be
eliminated using yeast flippase recombinase (encoded on pCP20) due to flanking FRT sites.
FRT site 1
kan
araC C280*
PBAD promoter
FRT site 2
desA-His
pBAD33*-desA-kan
7719 bp
p15A ori
CM(R)
M13 Ori
Figure 6.11 Map of template plasmid for integration of desA onto the chromosome. The cassette to
be integrated includes the region from FRT site 2 to the 3' terminus of desA-His.
6.2.2.3 Discussion and suggested future experiments
Preliminary experiments heterologously expressing Cti did not result in any changes in
percent intact cells or fatty acid titers. This seemed rather remarkable in light of the significant
change in composition of membrane lipids, but one possibility is that non-optimal temporal
control of Cti expression was utilized, and that remodeling of the membrane largely occurred
after most FFAs were produced. Thus a suggested experiment is to induce Cti expression earlier
than BTE expression, and to also express Cti from a higher copy plasmid to attempt to achieve a
higher rate of conversion of cis to trans. Cti expression may also confer resistance to other
solvents and chemicals of interest such as ethanol, n-butanol, isobutanol, toluene, and C10 or
lower n-alkanes and hydrocarbons.
210
Significant progress has been made toward constructing a strain with rewired unsaturated
fatty acid biosynthesis. This strain will have a deletion in fabA and will instead synthesize
controlled levels of unsaturated fatty acids by inducible expression of a chromosomally
integrated cassette harboring desA or desC.
A necessary initial experiment will entail
determining the level of induction of desaturases to support optimal growth, and to correlate the
membrane unsaturated content with level of growth (where enough cells are present to obtain
detectable FAME profiles). Once this range of inducer concentration is determined, the strains
can be transformed with BTE and a range of inducer concentrations can again be tested that
support optimal growth. As this strain will no longer be able to synthesize cis-unsaturated acylACPs, BTE expression should have no impact on membrane lipid composition, and many
experiments can be envisioned that explore physiological and production characteristics and
compare them to the strain with native unsaturated fatty acid biosynthesis.
However, as
preliminary complementation studies have shown, the re-engineered strain may have poor
growth and fitness characteristics due to it being unable to synthesize both of the unsaturated
species ordinarily present in the E. coli membrane (16:1∆9, 18:1∆11).
Additional experiments are envisioned that, rather than introducing an alternative route to
synthesizing unsaturated membrane lipids, would instead involve tuning expression levels of
fabA and fabB.
First, it should be determined whether FabA or FabB is rate-limiting in
unsaturated fatty acid biosynthesis. This could be achieved by overexpressing FabA and FabB
on plasmids, and determining which enzyme increases the ratio of unsaturated to saturated FFAs
produced by an acyl-ACP thioesterase. Next, rationally designed promoters, random promoter
libraries, or selected libraries of FabR binding sites with mutagenized bases can be tested for
their effect on expression of the rate limiting enzyme and on unsaturated fatty acid biosynthesis
211
and FFA production in a BTE-expressing strain. A simple preliminary experiment would
involve replacing the PfabA region with that of PfabB, as FabR more strongly represses fabB than
fabA.
Ultimately, it is sought to engineer a production host where high-level acyl-ACP
thioesterase expression does not perturb membrane lipid homeostasis.
6.2.3 Further investigations in metabolism
Much additional understanding of FFA production within the context of cellular metabolism
is required to be able to identify other possible bottlenecks. Fatty acid biosynthesis requires
utilization of both NADH and NADPH as reducing equivalents. While prior work has suggested
that NADPH is not limiting in cell-free lysates of a FFA-overproducing strain [Liu 2010a], it is
suggested that the levels of NADPH, NADH, NADP+, and NAD+ be measured and compared in
an FFA-overproducing strain and a negative control. Our own results suggest impairment of
aerobic respiration and cell depolarization as a result of FFA production (Chapter 3), and
exogenously added FFAs are believed to uncouple oxidative phosphorylation either by acting as
proton carriers by flipping across the inner membrane or increasing the permeability of the inner
membrane to protons [Desbois 2010]. The primary transhydrogenase in E. coli that catalyzes the
interconversion of NADH to NADPH, PntAB, is an inner membrane enzyme driven directly by
proton translocation.
Intracellular ATP levels should also be monitored, as FFAs have
previously been shown to reduce or prevent ATP synthesis [Desbois 2010]. This could occur
either indirectly via proton motive force dissipation, or by direct inhibition of ATP synthase by
FFAs [Wojtczak 1999, Desbois 2010].
A second area that warrants further investigation is determining the fate of excreted acetate.
As shown in Chapter 3, BTE-expressing cells accumulate lower maximum levels of acetate in
the extracellular medium than BTE-H204A-expressing cells.
212
Despite lower levels, the
maximum concentration measured is non-negligible at 15 mM, which would correspond to 2.5
mM lauric acid on a carbon basis, or 0.50 g/L. Secreted acetate is uptaken in stationary phase to
form acetyl-CoA, which can either be directed toward fatty acid biosynthesis, the TCA cycle to
generate ATP and 2 molecules of CO2 per acetyl-CoA, or through the glyoxylate shunt where
carbon is conserved. We have observed no changes in fatty acid titer in strains with deletions in
ackA-pta (data not shown) encoding acetate kinase and phosphate acetyltransferase, similar to
observations recently reported by another group [Li 2012] and suggesting either that uptaken
acetate is primarily directed to fatty acid biosynthesis, or that metabolic inefficiencies as a result
of interfering with exponential phase acetate production compensate for any benefits in
conserving carbon by preventing acetate formation. In order to determine whether uptaken
acetate is being incorporated into FFAs, a suggested experiment is to cultivate BTE-expressing
cultures until the stationary phase uptake period, resuspend the cells in depleted medium
supplemented with radiolabeled acetate, and measure the production of radiolabeled FFAs.
Another suggested experiment is investigation of perturbation of FFA production by deletion and
overexpression of acetyl-CoA synthetase (Acs).
213
References
Adebusuyi, A. A. and J. M. Foght. 2011. An alternative physiological role for the EmhABC
efflux pump in Pseudomonas fluorescens cLP6a. BMC Microbiol. 11:252.
Aguilar, P. S. and D. de Mendoza. 2006. Control of fatty acid desaturation: a mechanism
conserved from bacteria to humans. Molec. Microbiol. 62:1507-1514.
Åkerlund, T., K. Nordström, and R. Bernander. 1995. Analysis of cell size and DNA content
in exponentially growing and stationary-phase batch cultures of Escherichia coli. J. Bacteriol.
177:6791-6797.
Alekshun, M. N., and S. B. Levy. 1999. Alteration of the repressor activity of MarR, the
negative regulator of the Escherichia coli marRAB locus, by multiple chemicals in vitro. J.
Bacteriol. 181:4669-4672.
Alleman, T. L. and R. L. McCormick. 2006. Analysis of coconut-derived biodiesel and
conventional diesel fuel samples from the Philippines. NREL/MP-540-38643.
Alonso, D. M., J. Q. Bond, and J. A. Dumesic. 2010. Catalytic conversion of biomass to
biofuels. Green Chem. 12:1493-1513.
Alper, H., and G. Stephanopoulos. 2009. Engineering for biofuels: exploiting innate microbial
capacity or importing biosynthetic potential? Nat. Rev. Microbiol. 7:715-723.
Altieri, M. A. 2009. The ecological impacts of large-scale agrofuel monoculture production
systems in the Americas. Bull. Sci. Technol. Soc. 29:236-244.
Amann, E., B. Ochs, and K.-J. Abel. 1988. Tightly regulated tac promoter vectors useful for
the expression of unfused and fused proteins in Escherichia coli. Gene 69:301-315.
Andersen, J., S. A. Forst, K. Zhao, M. Inouye, N. Delihas. 1989. The function of micF RNA.
micF RNA is a major factor in the thermal regulation of OmpF protein in Escherichia coli. J.
Biol. Chem. 264:17961-17970.
Antoine, R. and C. Locht. 1992. Isolation and molecular characterization of a novel broad-hostrange plasmid from Bordetella bronchiseptica with sequence similarities to plasmids from
Gram-positive organisms. Mol. Microbiol. 6:1785-1799.
Aono, R., H. Kobayashi, K. N. Joblin, and K. Horikoshi. 1994. Effects of organic solvents on
growth of Escherichia coli K-12. Biosci. Biotechnol. Biochem. 58:2009-2014.
Aono, R. 1998. Improvement of organic solvent tolerance level of Escherichia coli by
overexpression of stress-responsive genes. Extremophiles 2:239-248.
Aono, R., N. Tsukagoshi, and M. Yamamoto. 1998b. Involvement of outer membrane protein
TolC, a possible member of the mar-sox regulon, in maintenance and improvement of
organic solvent tolerance of Escherichia coli K-12. J. Bacteriol. 180:938-944.
Asako, H., H. Nakajima, K. Kobayashi, M. Kobayashi, and R. Aono. 1997. Organic solvent
tolerance and antibiotic resistance increased by overexpression of marA in Escherichia coli.
Appl. Environ. Microbiol. 63:1428-1433.
Atsumi, S., T. Hanai, and J. C. Liao. 2008. Non-fermentative pathways for synthesis of
branched-chain higher alcohols as biofuels. Nature 451:86-89.
Azizan, A. and P. N. Black. 1994. Use of transposon TnphoA to identify genes for cell envelope
proteins of Escherichia coli required for long-chain fatty acid transport: the periplasmic
protein Tsp potentiates long-chain fatty acid transport. J. Bacteriol. 176:6653-6662.
214
Baba, T., A. Takeshi, M. Hasegawa, Y. Takai, Y. Okumura, M. Baba, K. A. Datsenko, M.
Tomita, B. L. Wanner, H. Mori. 2006. Construction of Escherichia coli K-12 in-frame,
single-gene knockout mutants: the Keio collection. Mol. Syst. Biol. 2:2006.0008.
Banerjee, A. 2011. Food, feed, fuel: transforming the competition for grains. Dev. Change
42:529-557.
Baneyx, F. and M. Mujacic. 2004. Recombinant protein folding and misfolding in Escherichia
coli. Nat. Biotechnol. 22:1399-1408/
Barney, B. M., B. D. Wahlen, E. Garner, J. Wei, and L. C. Seefeldt. 2012. Appl. Environ.
Microbiol. in press, June 8.
Becker, S. A., Feist, A. M., Hannum, G., Palsson, B. Ø., and M. J. Herrgard. 2007.
Quantitative prediction of cellular metabolism with constraint-based models: the COBRA
Toolbox. Nat. Protoc. 2:727-738.
Beller, H. R., E.-B. Goh, and J. D. Keasling. 2010. Genes involved in long-chain alkene
biosynthesis in Micrococcus luteus. Appl. Environ. Microbiol. 76:1212-1223.
Bergler, H., P. Wallner, A. Ebeling, B. Leitinger, S. Fuchsbichler, H. Aschauer, G. Kollenz,
G. Högenauer, and F. Turnowsky. 1994. Protein EnvM is the NADH-dependent enoylACP reductase (FabI) of Escherichia coli. J. Biol. Chem., 269:5493-5496.
Bergler, H., S. Fuchsbichler, G. Högenauer, and F. Turnowsky. 1996. The enoyl-[acylcarrier-protein] reductase (FabI) of Escherichia coli, which catalyzes a key regulatory step in
fatty acid biosynthesis, accepts NADH and NADPH as cofactors and is inhibited by
palmitoyl-CoA. Eur. J. Biochem. 242:689-694.
Binder, J. B. and R. T. Raines. 2009. Simple chemical transformation of lignocellulosic
biomass into furans for fuels and chemicals. J. Am. Chem. Soc. 131:1979-1985.
Birge, C. H. and P. R. Vagelos. 1972. Acyl carrier protein XVII: purification and properties of
β-hydroxyacyl carrier protein dehydrase. J. Biol. Chem. 247:4930-4938.
Black, P. N. 1990. Characterization of FadL-specific fatty acid binding in Escherichia coli.
Biochim. Biophys. Acta 1046:97-105.
Black, P. N. and C. C. DiRusso. 2003. Transmembrane movement of exogenous long-chain
fatty acids: proteins, enzymes, and vectorial esterification. Microbiol. Mol. Biol. Rev. 67:454472.
Bleuel, C., C. Grosse, N. Taudte, J. Scherer, D. Wesenberg, G. J. Krauss, D. H. Nies, and G.
Grass. 2005. TolC is involved in enterobactin efflux across the outer membrane of
Escherichia coli. J. Bacteriol. 187:6701-6707.
Bligh, E. G. and W. J. Dyer. 1959. A rapid method of total lipid extraction and purification.
Can. J. Biochem. Physiol. 37:911-917.
Bloch, K. 1971. "β-hydroxydecanoyl thioester dehydrase," in The Enzymes, ed. P. D. Boyer.
Academic Press, New York.
Bolstad, B. M., R. A. Irizarry, M. Åstrand, and T. P. Speed. 2003. A comparison of
normalization methods for high density oligonucleotide array data based on variance and bias.
Bioinformatics 19:185-193.
Bonamore, A., A. Macone, G. Colotti, R. M. Matarese, and A. Boffi. 2006. The desaturase
from Bacillus subtilis, a promising tool for the selective olefination of phospholipids. J.
Biotechnol. 121:49-53.
215
Bongaerts, J., S. Zoske, U. Weidner, and G. Unden. 1995. Transcriptional regulation of
the proton translocating NADH dehydrogenase genes (nuoA-N) of Escherichia coli by
electron acceptors, electron donors, and gene regulators. Mol. Microbiol. 16:521-534.
Boos, W. and A. Bohm. 2000. Learning new tricks from an old dog: MalT of the Escherichia
coli maltose system is part of a complex regulatory network. Trends Genet. 16:404-409.
Borges-Walmsley, M. I., J. Beauchamp, S. M. Kelly, K. Jumel, D. Candlish, S. E. Harding,
N. C. Price, and A. R. Walmsley. 2003. Identification of oligomerization and drug-binding
domains of the membrane fusion protein EmrA. J. Biol. Chem. 278:12903-12912.
Brissette, J. L., M. Russel, L. Weiner, and P. Model. 1990. Phage shock protein, a stress
protein of Escherichia coli. Proc. Natl. Acad. Sci. USA 87:862-866.
Brown, M. H. and R. A. Skurray. 2001. Staphylococcal multidrug efflux protein QacA. J. Mol.
Microbiol. Biotechnol. 3:163-170.
Browning, D. F., C. M. Beatty, E. A. Sanstad, K. E. Gunn, S. J. Busby, and A. J. Wolfe.
2004. Modulation of CRP-dependent transcription at the Escherichia coli acsP2 promoter by
nucleoprotein complexes: anti-activation by the nucleoid proteins FIS and IHF. Mol.
Microbiol. 51:241-254.
Brynildsen, M. P., and J. C. Liao. 2009. An integrated network approach identifies the
isobutanol response network of Escherichia coli. Mol. Syst. Biol. 5:277.
Campbell, J. W. and J. E. Cronan, Jr. 2002. The enigmatic Escherichia coli fadE gene is yafH.
J. Bacteriol. 184:3759-3764.
Campbell, J. W., R. M. Morgan-Kiss, and J. E. Cronan, Jr. 2003. A new Escherichia coli
metabolic competency: growth on fatty acids by a novel anaerobic β-oxidation pathway. Mol.
Microbiol. 47:793-805.
Carter, O. A., R. J. Peters, and R. Croteau. 2003. Monoterpene biosynthesis pathway
construction in Escherichia coli. Phytochemistry 64:425-433.
Chang, A. C. Y. and S. N. Cohen. 1978. Construction and characterization of amplifiable
multicopy DNA cloning vehicles derived from the P15A cryptic miniplasmid. J. Bacteriol.
134:1411-1156.
Chang, Y.-Y. and J. E. Cronan, Jr. 1999. Membrane cyclopropane fatty acid content is a major
factor in acid resistance of Escherichia coli. Mol. Microbiol. 33:249-259.
Chang, M. C. Y., R. A. Eachus, W. Trieu, D.-K. Ro, and J. D. Keasling. 2007. Engineering
Escherichia coli for production of functionalized terpenoids using plant P450s. Nat. Chem.
Biol. 3:274-277.
Cherepanov, P. P., and W. Wackernagel. 1995. Gene disruption in Escherichia coli: TcR and
KmR cassettes wth the option of Flp-catalyzed excision of the antibiotic-resistance
determinant. Gene 158:9-14.
Cho, H. and J. E. Cronan, Jr. 1993. Escherichia coli thioesterase I. Molecular cloning and
sequencing of the structural gene and identification as a periplasmic enzyme. J. Biol. Chem.
268:9238-9245.
Cho, H. and J. E. Cronan, Jr. 1995. Defective export of a periplasmic enzyme disrupts
regulation of fatty acid synthesis. J. Biol. Chem. 270:4216-4219.
Cho, B. K., E. M. Knight, and B. Ø. Palsson. 2006. Transcriptional regulation of the fad
regulon genes of Escherichia coli by ArcA. Microbiology 152:2207-19
Choi, J. H. and S. Y. Lee. 2004. Secretory and extracellular protein production of recombinant
proteins using Escherichia coli. Appl. Microbiol. Biotechnol. 64:625-635.
216
Christie, W. W. 2003. Lipid Analysis, 3rd edition. Oily Press, Bridgewater, UK.
Chundawat, S. P. S., G. T. Beckham, M. E. Himmel, and B. E. Dale. 2011. Deconstruction of
lignocellulosic biomass to fuels and chemicals. Annu. Rev. Chem. Biomol. Eng. 2:121-145.
Chung, C. T., S. L. Niemela, and R. H. Miller. 1989. One-step preparation of competent
Escherichia coli: transformation and storage of bacterial cells in the same solution. Proc.
Natl. Acad. Sci. USA 86:2172-2175.
Clark, D. P. and J. E. Cronan, Jr. 1996. "Two-carbon compounds and fatty acids as carbon
sources," in F. C. Neihardt (ed.), Escherichia coli and Salmonella: Cellular and Molecular
Biology, 2nd edition. ASM Press, Washington, DC.
Cozzone, A. J. and M. El-Mansi. 2005. Control of isocitrate dehydrogenase catalytic activity
by protein phosphorylation in Escherichia coli. J. Mol. Microbiol. Biotechnol. 9:132-146.
Cronan, J. E., Jr. 1997. In vivo evidence that acyl coenzyme A regulates DNA binding by the
Escherichia coli FadR global transcription factor. J. Bacteriol. 179:1819-1822.
Cronan, J. E. 2006. A family of arabinose-inducible Escherichia coli expression vectors having
pBR322 copy control. Plasmid 55:152-157.
Cronan, J. E., Jr., and C. O. Rock. October 2008, posting date. Chapter 3.6.4, Biosynthesis of
membrane lipids. In R. Curtiss III et al. (ed.), EcoSal - Escherichia coli and Salmonella:
cellular and molecular biology. ASM Press, Washington, DC. http://www.ecosal.org.
Cronan, J. E., Jr. and G. L. Waldrop. 2002. Multi-subunit acetyl-CoA carboxylases. Prog.
Lipid Res. 41:407-435.
Dalbey, R. E., P. Wang, and A. Kuhn. 2011. Assembly of bacterial inner membrane proteins.
Annu. Rev. Biochem. 80:161-187.
Datsenko, K. A. and B. L. Wanner. 2000. One-step inactivation of chromosomal genes in
Escherichia coli K-12 using PCR products. Proc. Natl. Acad. Sci. USA 97:6640-6645.
Davis, M. S., J. Solbiati, and J. E. Cronan, Jr. 2000. Overproduction of acetyl-CoA
carboxylase activity increases the rate of fatty acid biosynthesis in Escherichia coli. J. Biol.
Chem. 275:28593-29598.
Davis, M. S. and J. E. Cronan, Jr. 2001. Inhibition of Escherichia coli acetyl coenzyme A
carboxylase by acyl-acyl carrier protein. J. Biol. Chem. 282:20319-20328.
Davis, S. C., S. W. Diegel, and R. G. Boundy. 2009. Transportation energy data book: edition
28. U.S. Department of Energy. ORNL-6984.
Dehesh, K., A. Jones. D. S. Knutzon, and T. A. Voelker. 1996a. Production of high levels of
8:0 and 10:0 fatty acids in transgenic canola by overexpression of Ch FatB2, a thioesterase
cDNA from Cuphea hookeriana. Plant J. 9:167-172.
Dehesh, K., P. Edwards, T. Hayes, A. M. Cranmer, and J. Fillatti. 1996b. Two novel
thioesterases are key determinants of the bimodal distribution of acyl chain length of Cuphea
palustris seed oil. Plant Physiol. 110:203-210.
De Lay, N. R. and J. E. Cronan. 2007. In vivo functional analysis of the type II acyl carrier
proteins of fatty acid biosynthesis. J. Biol. Chem. 282:20319-20328.
Dellomonaco, C., J. M. Clomburg, E. N. Miller, and R. Gonzalez. 2011. Engineered reversal
of the β-oxidation cycle for the synthesis of fuels and chemicals. Nature 476:355-359.
Demple, B. 1996. Redox signaling and gene control in the Escherichia coli soxRS oxidative
stress regulon - a review. Gene 179:53-57.
217
Desbois, A. P. and V. J. Smith. 2010. Antibacterial free fatty acids: activities, mechanisms
of action and biotechnological potential. Appl. Microbiol. Biotechnol. 85:1629-1642.
Dhamdhere, G. and H. I. Zgurskaya. 2010. Metabolic shutdown in Escherichia coli cells
lacking the outer membrane channel TolC. Molec. Microbiol. 77:743-754.
DiRusso, C. C., T. L. Heimert, and A. K. Metzger. 1992. Characterization of FadR, a global
transcriptional regulator of fatty acid metabolism in Escherichia coli. J. Biol. Chem.
267:8685-8691.
Doan, T. T. P., A. S. Carlsson, M. Hamberg, L. Bülow, S. Stymne, and P. Olsson. 2009.
Functional expression of five Arabidopsis fatty acyl-CoA reductase genes in Escherichia coli.
J. Plant Physiol. 166:787-796.
Donath, M. J., II, M. A. Dominguez, and S. T. Withers, III. 2011. Development of an
automated platform for high-throughput P1-phage transduction of Escherichia coli. J. Lab
Autom. 16:141-147.
Dry, M. E. 2002. High quality diesel via the Fischer-Tropsch process - a review. J. Chem.
Technol. Biotechnol. 77:43-50.
Duan, Y., Z. Zhu, K. Cai, X. Tan, and X. Lu. 2011. De novo biosynthesis of biodiesel by
Escherichia coli in optimized fed-batch cultivation. PLoS One 6:e20265.
Duffield, J. A. 2007. Biodiesel: production and economic issues. Inhal. Toxicol. 19:1029-1031
Dunkley, E. A., Jr., S. Clejan, and T. A. Krulwich. 1991. Mutants of Bacillus species isolated
on the basis of protonophore resistance are deficient in fatty acid desaturase activity. J.
Bacteriol. 173:7750-7755.
Dunlop, M. J., Z. Y. Dossani, H. L. Szmidt, H. C. Chu, T. S. Lee, J. D. Keasling, M. Z. Hadi,
and A. Mukhopadhyay. 2011. Engineering microbial biofuel tolerance and export using
efflux pumps. Mol. Syst. Biol. 7:487.
Dunlop, M. J. 2011b. Engineering microbes for tolerance to next-generation biofuels.
Biotechnol. Biofuels 4:32.
Eccleston, V. S. and J. B. Ohlrogge. 1998. Expression of lauroyl-acyl carrier protein
thioesterase in Brassica napus seeds induces pathways for both fatty acid oxidation and
biosynthesis and implies a set point for triacylglycerol accumulation. Plant Cell 10:613-621.
Edwards, P., J. Sabo Nelsen, J. G. Metz, and K. Dehesh. 1997. Cloning of the fabF gene in an
expression vector and in vitro characterization of recombinant fabF and fabB encoded
enzymes from Escherichia coli. FEBS Lett. 402:62-66.
Edwards, J. S., Ibarra, R. U., and B. Ø Palsson. 2001. In silico predictions of Escherichia coli
metabolic capabilities are consistent with experimental data. Nat. Biotechnol. 19:125-130).
Engl, C., G. Jovanovic, L. J. Lloyd, H. Murray, M. Spitaler, L. Ying, J. Errington, and M.
Buck. 2009. In vivo localizations of membrane stress controllers PspA and PspG in
Escherichia coli. Mol. Microbiol. 73:382-396.
Fàbrega, A., R. G. Martin, J. L. Rosner, M. Mar Tavio, and J. Vila. 2010. Constitutive SoxS
expression in a fluoroquinolone-resistant strain with a truncated SoxR protein and
identification of a new member of the marA-soxS-rob regulon, mdtG. Antimicrob. Agents
Chemother. 54:1218-1225.
Fay, J. P., and R. N. Farias. 1975. The inhibitory action of fatty acids on the growth of
Escherichia coli. J. Gen. Microbiol. 91:233-240.
218
Fay, J. P., and R. N. Farias. 1977. Inhibitory action of a non-metabolizable fatty acid on the
growth of Escherichia coli: role of metabolism and outer membrane integrity. J. Bacteriol.
132:790-795.
Feist, A. M., Henry, C. S., Reed, J. L., Krummenacker, M., Joyce, A. R., Karp, P. D.,
Broadbelt, L. J., Hatzimanikatis, V., and B. Ø. Palsson. 2007. A genome-scale metabolic
reconstruction for Escherichia coli K-12 MG1655 that accounts for 1260 ORFs and
thermodynamic information. Molec. Sys. Biol. 3:121.
Feng, Y., and Cronan, J. E. 2009. Escherichia coli unsaturated fatty acid synthesis: complex
transcription of the fabA gene and in vivo identification of the essential reaction catalyzed by
FabB. J. Biol. Chem. 284:29526-29535.
Feng, Y., and Cronan, J. E. 2011. Complex binding of the FabR repressor of bacterial
unsaturated fatty acid biosynthesis to its cognate promoters. Mol. Microbiol. 80:195-218.
Fong, S. S., Joyce, A. R., and B. Ø. Palsson. 2005. Parallel adaptive evolution cultures of
Escherichia coli lead to convergent growth phenotypes with different gene expression states.
Genome Res. 15:1365-1372.
Fortman, J. L., S. Chhabra, A. Mukhopadhyay, H. Chou, T. S. Lee, E. Steen, and J. D.
Keasling. 2008. Biofuel alternatives to ethanol: pumping the microbial well. Trends
Biotechnol. 26:375-381.
Fralick, J. A. 1996. Evidence that TolC is required for functioning of the Mar/AcrAB efflux
pump of Escherichia coli. J. Bacteriol. 178:5803-5805.
Friehs, K. 2004. Plasmid copy number and plasmid stability. Adv. Biochem. Eng. Biotechnol.
86:47-82.
Gama-Castro, S., H. Salgado, M. Peralta-Gil, A. Santos-Zavaleta, L. Muñiz- Rascado, H.
Solano-Lira, V. Jimenez-Jacinto, V. Weiss, J. S. García-Sotelo, A. López-Fuentes,
L.
Porrón-Sotelo, S. Alquicira-Hernández, A. Medina-Rivera, I.
Martínez-Flores, K.
Alquicira-Hernández, R. Martínez-Adame, C. Bonavides-Martínez, J. Miranda-Ríos, A.
M. Huerta, A. Mendoza-Vargas, L. Collado- Torres, B. Taboada, L. Vega-Alvarado, M.
Olvera, L. Olvera, R. Grande, E. Morett, and J. Collado-Vides. 2011. RegulonDB
version 7.0: transcriptional regulation of Escherichia coli K-12 integrated within genetic
sensory response units (Gensor Units). Nucleic Acids Res. 39:D98-D105.
Gao, X., J.-C. Chen, Q. Wu, and G.-Q. Chen. 2011. Polyhydroxyalkanoates as a source of
chemicals, polymers, and biofuels. Curr. Opin. Biotechnol. 22:768-774.
Garwin, J. L., A. L. Klages, and J. E. Cronan, Jr. 1980a. Structural, enzymatic, and genetic
studies of β-ketoacy-acyl carrier protein synthases I and II of Escherichia coli. J. Biol. Chem.
255:11949-11956.
Garwin, J. L., A. L. Klages, and J. E. Cronan, Jr. 1980b. β-Ketoacyl-acyl carrier protein
synthase II of Escherichia coli. Evidence for funtion in the thermal regulation of fatty acid
synthesis. J. Biol. Chem. 255:3263-3265.
Ghosh, S. K., A. Bhattacharjee, J. K. Jha, A. K. Mondal, M. K. Maiti, A. Basu, D. Ghosh, S.
Ghosh, and S. K. Sen. 2007. Characterization and cloning of a stearoyl/oleoyl specific fatty
acyl-acyl carrier protein thioesterase from the seeds of Madhuca longifolia (latifolia). Plant
Physiol. Biochem. 45:887-897.
Goh, E.-B., E. E. K. Baidoo, J. D. Keasling, and H. R. Beller. 2012. Engineering of bacterial
methyl ketone synthesis for biofuels. Appl. Environ. Microbiol. 78:70-80.
219
Graham-Rowe, D. 2011. Agriculture: beyond food versus fuel. 464:S6-S8.
Green, E. M. 2011. Fermentative production of butanol - the industrial perspective. Curr. Opin.
Biotechnol. 22:337-343.
Greenspan, P., E. P. Mayer, and S. D. Fowler. 1985. Nile red: a selective fluorescent stain for
intracellular lipid droplets. J. Cell Biol. 100:965-973.
Grkovic, S., M. H. Brown, and R. A. Skurray. 2002. Regulation of bacterial drug export
systems. Microbiol. Mol. Biol. Rev. 66:671-701.
Grogan, D. W. and J. E. Cronan, Jr. 1997. Cyclopropane ring formation in membrane lipids of
bacteria. Microbiol. Mol. Biol. Rev. 61:429-441.
Gucchait, R. B., S. E. Polakis, P. Dimroth, E. Stoll, J. Moss, and M. D. Lane. 1974. Acetyl
coenzyme A carboxylase system of Escherichia coli. J. Biol. Chem. 249:6633-6645.
Gunstone, F. D. and F. B. Padley, eds. 1997. Lipid Technologies and Applications. Marcel
Dekker, New York.
Guzman, L.-M., D. Belin, M. J. Carson, and J. Beckwith. 1995. Tight regulation, modulation,
and high-level expression by vectors containing the arabinose PBAD promoter. J. Bacteriol.
177:4121-4130.
Han, M.-J., J. W. Lee, S. Y. Lee, and J. S. Yoo. 2008. Proteome-level responses of Escherichia
coli to long-chain fatty acids and use of fatty acid inducible promoter in protein production. J.
Biomed. Biotechnol. 2008:735101.
Handke, P., S. A. Lynch, and R. T. Gill. 2011. Application and engineering of fatty acid
biosynthesis in Escherichia coli for advanced fuels and chemicals. Metab. Eng. 13:28-37.
Hantke, K., K. Winkler, and J. E. Schultz. 2011. Escherichia coli exports cyclic AMP via
TolC. J. Bacteriol. 193:1086-1089,
Harder, M. E., R. C. Ladenson, S. D. Schimmel, and D. F. Silbert. 1974. Mutants of
Escherichia coli with temperature-sensitive malonyl coenzyme A-acyl carrier protein
transacylase. J. Biol. Chem. 249:7468-7475.
Hayashi, S., R. Aono, T. Hanai, H. Mori, T. Kobayashi, and H. Honda. 2003. Analysis
of
organic solvent tolerance in Escherichia coli using gene expression profiles from DNA
microarrays. J. Biosci. Bioeng. 95:379-383.
Haynes, W. M., ed. 2012. "Heats of Combustion, Interactive Table," in CRC Handbook of
Chemistry and Physics , 92nd edition (Internet Version 2012). CRC Press/Taylor and Francis,
Boca Raton, FL.
Hearn, E. M., D. R. Patel, B. W. Lepore, M. Indic, and B. van den Berg. 2009.
Transmembrane passage of hydrophobic compounds through a protein channel wall. Nature
458:367-370.
Heath, R. J., S. Jackowski, and C. O. Rock. 1994. Guanosine tetraphosphate inhibition of fatty
acid and phospholipid synthesis in Escherichia coli is relieved by overexpression of glycerol3-phosphate acyltransferase (plsB). J. Biol. Chem. 269:26584-26590.
Heath, R. J. and C. O. Rock. 1995a. Regulation of malonyl-CoA metabolism by acyl-acyl
carrier protein and β-ketoacyl-acyl carrier protein synthases in Escherichia coli. J. Biol.
Chem. 270:15531-15538.
Heath, R. J. and C. O. Rock. 1995b. Enoyl-acyl carrier protein reductase (FabI) plays a
determinant role in competing cycles of fatty acid elongation in Escherichia coli. J. Biol.
Chem. 270:26538-26542.
220
Heath, R. J., and C. O. Rock. 1996. Regulation of fatty acid elongation and initiation by acylacyl carrier protein in Escherichia coli. J. Biol. Chem. 271:1833-1836.
Heipieper, H. J. and J. A. de Bont. 1994. Adaptation of Pseudomonas putida S12 to ethanol
and toluene at the level of fatty acid composition of membranes. Appl. Environ. Microbiol.
60: 4440-4444.
Helwani, Z., M. R. Othman, N. Aziz, W. J. N. Fernando, and J. Kim. 2009. Technologies for
production of biodiesel focusing on green catalytic techniques: a review. Fuel Proc. Technol.
90:1502-1514.
Hill, J., E. Nelson, D. Tilman, S. Polansky, and D. Tiffany. 2006. Environmental, economic,
and energetic costs and benefits of biodiesel and ethanol biofuels. Proc. Natl. Acad. Sci. USA
103:11206-11210.
Hofvander, P., T. T. Doan, and M. Hamberg. 2011. A prokaryotic acyl-CoA reductase
performing reduction of fatty acyl-CoA to fatty alcohol. FEBS Lett. 585:3538-3543.
Holtwick, R., F. Meinhardt, and H. Keweloh. 1997. cis-trans Isomerization of unsaturated
fatty acids: cloning and sequencing of the cti gene from Pseudomonas putida P8. Appl.
Environ. Microbiol. 63:4292-4297.
Holtzapple, E. and C. Schmidt-Dannert. 2007. Biosynthesis of isoprenoid wax ester in
Marinobacter hydrocarbonoclasticus DSM 8798: identification and characterization of
isoprenoid coenzyme A synthetase and wax ester synthases. J. Bacteriol. 189:3804-3812.
Hommais, F., E. Krin, J.-Y. Coppée, C. Lacroix, E. Yeramian, A. Danchin, and P. Bertin.
2004. GadE (YhiE): a novel activator involved in the response to acid environment
in
Escherichia coli. Microbiology 150:61-72.
Hoover, S. W., W. D. Marner II, A. K. Brownson, R. M. Lennen, T. M. Wittkopp, J.
Yoshitani, S. Zulkifly, L. E. Graham, S. Chaston, K. D. McMahon, and B. F. Pfleger.
2011. Bacterial production of free fatty acids from freshwater, macroalgal cellulose. Appl.
Microbiol. Biotechnol. 91:435-446.
Hoover, S. W., J. T. Youngquist, P. A. Angart, S. T. Withers, R. M. Lennen, and B. F.
Pfleger. 2012. Isolation of improved free fatty acid overproducing strains of Escherichia coli
via Nile red based high-throughput screening. Environ. Prog. Sustainable Energy 31:17-23.
Houghton, J., S. Weatherwax, and J. Ferrell. 2006. Breaking the biological barriers to
cellulosic ethanol: a joint research agenda. DOE/SC-0095.
Huffer, S., M. E. Clark, J. C. Ning, H. W. Blanch, and D. S. Clark. 2011. Role of alcohols in
growth, lipid composition, and membrane fluidity of yeasts, bacteria, and archaea. Appl.
Environ. Microbiol. 77:6400-6408.
Ingram, L. O. 1976. Adaptation of membrane lipids to alcohols. J. Bacteriol. 125:670-678.
Ingram, L. O. and N. S. Vreeland. 1980. Differential effects of ethanol and hexanol on the
Escherichia coli cell envelope. J. Bacteriol. 144:481-488.
Ingram, L. O., T. Conway, D. P. Clark, G. W. Sewell, and J. F. Preston. 1987. Genetic
engineering of ethanol production in Escherichia coli. Appl. Environ. Microbiol. 53:24202425.
Jackowski, S. and C. O. Rock. 1987. Acetoacetyl-acyl carrier protein synthase, a potential
regulator of fatty acid biosynthesis in bacteria. J. Biol. Chem. 262:7927-7931.
James, E. S. and J. E. Cronan. 2004. Expression of two Escherichia coli acetyl-CoA
carboxylase subunits is autoregulated. J. Biol. Chem. 279:2520-2527.
221
Janikowski, T. B., D. Velicogna, M. Punt, and A. Daugulis. 2002. Use of a two-phase
partitioning bioreactor for degrading polycyclic aromatic hydrocarbons by a Sphingomonas
sp. Appl. Microbiol. Biotechnol. 59:368-376.
Jeon, E., S. Lee, J.-I. Won, S. O. Han, J. Kim, and J. Lee. 2011. Development of Escherichia
coli MG1655 strains to produce long chain fatty acids by engineering fatty acid synthesis
(FAS) metabolism. Enzyme Microb. Technol. 49:44-51.
Jha, J. K., M. K. Maiti, A. Bhattacharjee, A. Basu, P. C. Sen, and S. K. Sen. 2006. Cloning
and functional expression of an acyl-ACP thioesterase FatB type from Diploknema
(Madhuca) butyracea seeds in Escherichia coli. Plant Physiol. Biochem. 44:645-655.
Jha, S. S., J. K. Jha, B. Chattopadhyaya, A. Basu, S. K. Sen, and M. K. Maiti. 2010. Cloning
and characterization of cDNAs encoding for long-chain saturated acyl-ACP thioesterases
from the developing seeds of Brassica juncea. Plant Physiol. Biochem. 48:476-480.
Jiang, P. and J. E. Cronan, Jr. 1994. Inhibition of fatty acid synthesis in Escherichia coli in the
absence of phospholipid synthesis and release of inhibition by thioesterase action. J.
Bacteriol. 176:2814-2821.
Jing, F., D. C. Cantu, J. Tvaruzkova, J. P. Chipman, B. J. Nikolau, M. D. Yandeau-Nelson,
and P. J. Reilly. 2011. Phylogenetic and experimental characterization of an acyl-ACP
thioesterase family reveals significant diversity in enzymatic specificity and activity. BMC
Biochem. 12:44.
Johnson, R. W. and E. Fritz, eds. 1989. Fatty Acids in Industry. .Marcel Dekker, New York.
Joly, N., C. Engl, G. Jovanovic, M. Huvet, T. Toni, X. Sheng, M. P. H. Stumpf, and M.
Buck. 2010. Managing membrane stress: the phage shock protein (Psp) response, from
molecular mechanisms to physiology. FEMS Microbiol. Rev. 34:797-827.
Jones, A., H. M. Davies, and T. A. Voelker. 1995. Palmitoyl-acyl carrier protein (ACP)
thioesterases and the evolutionary origin of plant acyl-ACP thioesterases. Plant Cell 7:359371.
Jones, K. L., S. W. Kim, and J. D. Keasling. 2000. Low-copy plasmids can perform as well or
better than high-copy plasmids for metabolic engineering of bacteria. Metab. Eng. 2:328-338.
Jones, A. M. P., J. A. Klun, C. L. Cantrell, D. Ragone, K. R. Chauhan, P. N. Brown, and S.
J. Murch. 2012. Isolation and identification of mosquito (Aedes aegypti) biting deterrent
fatty acids from male inflorescences of breadfruit (Artocarpus altilis (Parkinson) Fosberg). J.
Agric. Food Chem. 60:3867-3873.
Jovanovic, G., L. J. Lloyd, M. P. H. Stumpf, A. J. Mayhew, and M. Buck. 2006. Induction
and function of the phage shock protein extracytoplasmic stress response in Escherichia coli.
J. Biol. Chem. 281:21147-21161.
Junker, F. and J. L. Ramos. 1999. Involvement of the cis/trans isomerase Cti in solvent
resistance of Pseudomonas putida DOT-T1E. J. Bacteriol., 181:5693-5700.
Kabir, M. S., D. Yamashita, S. Koyama, T. Oshima, K. Kurokawa, M. Maeda, R.
Tsunedomi, M. Murata, C. Wada, H. Mori, and M. Yamada. 2005. Cell lysis directed by
σE in early stationary phase and effect of induction of the rpoE gene on global gene
expression in Escherichia coli. 151:2721-2735.
Kalscheuer, R. and A. Steinbüchel. 2003. A novel bifunctional wax ester synthase/acylCoA:diacylglycerol acyltransferase mediates wax ester and triacylglycerol biosynthesis in
Acinetobacter calcoaceticus ADP1. J. Biol. Chem. 278:8075-8082.
222
Kalscheuer, R., T. Stölting, and A. Steinbüchel. 2006. Microdiesel: Escherichia coli
engineered for fuel production. Microbiology 152:2529-2536.
Kameda, K. and W. D. Nunn. 1981. Purification and characterization of acyl coenzyme A
synthetase from Escherichia coli. J. Biol. Chem. 256:5702-5707.
Kampf, J. P., D. Cupp, and A. M. Kleinfeld. 2006. Different mechanisms of free fatty acid
flip-flop and dissociation revealed by temperature and molecular species dependence of
transport across lipid vesicles. J. Biol. Chem. 281:21566-21574.
Keseler, I. M., J. Collado-Vides, A. Santos-Zavaleta, M. Peralta-Gil, S. Gama-Castro, L.
Muñiz-Rascado, C. Bonavidez-Martinez, S. Paley, M. Krummenacker, T. Altman, P.
Kaipa, A. Spaulding, J. Pacheco, M. Latendresse, C. Fulcher, M. Sarker, A. G. Shearer,
A. Mackie, I. Paulsen, R. P. Gunsalus, P. D. Karp. 2011. EcoCyc: a comprehensive
database of Escherichia coli biology. Nucleic Acids Res. 39:D583-590.
Kessels, J. M., H. Ousen, and H. Van den Bosch. 1983. Facilitated utilization of endogenously
synthesized lysophosphatidic acid by 1-acylglycerophosphate acyltransferase from
Escherichia coli. Biochim. Biophys. Acta 753:227-235.
Khlebnikov, A., K. A. Datsenko, T. Skaug, B. L. Wanner, and J. D. Keasling. 2001.
Homogeneous expression of the PBAD promoter in Escherichia coli by constitutive expression
of the low-affinity high-capacity AraE transporter. Microbiology 147:3241-3247.
Khosla, C. 1997. Harnessing the biosynthetic potential of modular polyketide synthases. Chem.
Rev. 97:2577-2590.
Kizer, L., D. J. Pitera, B. F. Pfleger, and J. D. Keasling. 2008. Application of functional
genomics to pathway optimization for increased isoprenoid production. Appl. Environ.
Microbiol. 74:3229-3241.
Klein, K. R., R. Steinberg, B. Fiethen, and P. Overath. 1971. Fatty acid degradation in
Escherichia coli. An inducible system for the uptake of fatty acids and further
characterization of old mutants. Eur. J. Biochem. 19:442-450.
Knothe, G. 2010. Biodiesel and renewable diesel: a comparison. Prog. Energy Combustion Sci.
36:364-373.
Kobayashi, H., M. Yamamoto, and R. Aono. 1998. Appearance of a stress-response protein,
phage-shock protein A, in Escherichia coli exposed to hydrophobic organic solvents.
Microbiology 144:353-359.
Kobayashi, R., T. Suzuki, and M. Yoshida. 2007. Escherichia coli phage-shock protein A
(PspA) binds to membrane phospholipids and repairs proton leakage of the damaged
membranes. Mol. Microbiol. 66:100-109.
Koo, A. J. K., J. B. Ohlrogge, and M. Pollard. 2004. On the export of fatty acids from the
chloroplast. J. Biol. Chem. 279:16101-16110.
Kumari, S., C. M. Beatty, D. F. Browning, S. J. Busby, E. J. Simel, G. Hovel-Miner, and A.
J. Wolfe. 2000. Regulation of acetyl coenzyme A synthetase in Escherichia coli. 182:41734179.
Kuzuyama, T. 2002. Mevalonate and nonmevalonate pathways for the biosynthesis of isoprene
units. Biosci. Biotechnol. Biochem. 66:1619-1627.
Lee, E. H. and W. M. Shafer. 1999. The farAB-encoded efflux pump mediates resistance of
gonococci to long-chained antibacterial fatty acids. Mol. Microbiol. 33:839-845.
223
Lee, S. K., H. H. Chou, B. F. Pfleger, J. D. Newman, Y. Yoshikuni, and J. D. Keasling.
2007. Directed evolution of AraC for improved compatibility of arabinose- and lactoseinducible promoters. Appl. Environ. Microbiol. 73:5711-5715.
Lennen, R. M., D. J. Braden, R. M. West, J. A. Dumesic, and B. F. Pfleger. 2010. A process
for microbial hydrocarbon synthesis: overproduction of fatty acids in Escherichia coli and
catalytic conversion to alkanes. Biotechnol. Bioeng. 106:193-202.
Lennen, R. M., M. A. Kruziki, K. Kumar, R. A. Zinkel, K. E. Burnum, M. S. Lipton, S. W.
Hoover, D. R. Ranatunga, T. M. Wittkopp, W. D. Marner, II, and B. F. Pfleger. 2011.
Membrane stresses induced by overproduction of free fatty acids in Escherichia coli. Appl.
Environ. Microbiol. 77:8114-8128.
Lepore, B. W., M. Indic, H. Pham, E. M. Hearn, D. R. Patel, and B. van den Berg. 2011.
Ligand-gated diffusion across the bacterial outer membrane. Proc. Natl. Acad. Sci. USA
108:10121-10126.
Lestari, S., P. Mäki-Arvela, J. Beltramini, G. Q. M. Lu, and D. Yu. Murzin. 2009.
Transforming triglycerides and fatty acids into biofuels. ChemSusChem 2:1109-1119.
Li, S.-J. and J. E. Cronan, Jr. 1992. The gene encoding the biotin carboxylase subunit of
Escherichia coli acetyl-CoA carboxylase. J. Biol. Chem. 267:855-863.
Li, S.-J. and J. E. Cronan, Jr. 1993. Growth rate regulation of Escherichia coli acetyl
coenzyme A carboxylase, which catalyzes the first committed step of lipid biosynthesis. J.
Bacteriol. 175:332-340.
Li, M., X. Zhang, A. Agrawal, and K.-Y. San. 2012. Effect of acetate formation pathway and
long chain fatty acid CoA-ligase on the free fatty acid production in E. coli expressing acylACP thioesterase from Ricinus communis. Metab. Eng. 14:380-387.
Lin, H., N. M. Castro, G. N. Bennett, and K.-Y. San. 2006. Acetyl-CoA synthetase
overexpression in Escherichia coli demonstrates more efficient acetate assimilation and
lower acetate accumulation: a potential tool in metabolic engineering. Appl. Microbiol.
Biotechnol. 71:870-874.
Lipton, M. S., M. F. Romine, M. E. Monroe, D. A. Elias, L. Pasa-Tolic, G. A. Anderson, D.
J. Anderson, J. Fredrickson, K. K. Hixson, C. Masselon, H. Mottaz, N. Tolic, and R. D.
Smith. 2006. AMT tag approach to proteomic characterization of Deinococcus radiodurans
and Shewanella oneidensis. Methods Biochem. Anal. 49:113-134.
Liu, T., H. Vora, and C. Khosla. 2010a. Quantitative analysis and engineering of fatty acid
biosynthesis in E. coli. Metab. Eng. 12:378-386.
Liu, T. and C. Khosla. 2010b. Genetic engineering of Escherichia coli for biofuel production.
Annu. Rev. Genet. 44:53-69.
Liu, H., C. Yu, D. Feng, T. Cheng, X. Meng, W. Liu, H. Zou, and M. Xian. 2012. Production
of extracellular fatty acid using engineered Escherichia coli. Microb. Cell Fact. 11:41.
Livak, K. J., and T. D. Schmittgen. 2001. Analysis of relative gene expression data using realtime quantitative PCR and the 2-∆∆Ct method. Methods 25:402-408.
Lorca, G. L., A. Ezersky, V. V. Lunin, J. R. Walker, S. Altamentova, E. Evdokimova, M.
Vedadi, A. Bochkarev, and A. Savchenko. 2007. Glyoxylate and pyruvate are antagonistic
effectors of the Escherichia coli IclR transcriptional regulator. J. Biol. Chem. 282:1647616491.
Lu, X., H. Vora, and C. Khosla. 2008. Overproduction of free fatty acids in E. coli:
implications for biodiesel production. Metab. Eng. 10:333-339.
224
Lynch, M. D. and R. T. Gill. 2006. Broad host range vectors for stable genomic library
construction. Biotechnol. Bioeng. 94:151-158.
Ma, D., M. Alberti, C. Lynch, H. Nikaido, J. E. Hearst. 1996. The local repressor AcrR plays
a modulating role in the regulation of acrAB genes of Escherichia coli by global stress
signals. Mol. Microbiol. 19:101-112.
Ma, D., D. N. Cook, M. Alberti, N. G. Pon, H. Nikaido, J. E. Hearst. 1995. Genes acrA and
acrB encode a stress-induced efflux system of Escherichia coli. Mol. Microbiol. 16:45-55.
Mäki-Arvela, P., I. Kubickova, M. Snåre, K. Eränen, and D. Yu. Murzin. 2007. Catalytic
deoxygenation of fatty acids and their derivatives. Energy Fuels 21:30-41.
Maloy, S. R., C. L. Ginsburgh, R. W. Simons, and W. D. Nunn. 1981. Transport of long and
medium chain fatty acids by Escherichia coli K12. J. Biol. Chem. 256:3735-3742.
Marr, A. G. and J. L. Ingraham. 1962. Effect of temperature on the compositions of fatty acids
in Escherichia coli. J. Bacteriol. 84:1260-1267.
Martin, R. G., W. K. Gillette, S. Rhee, and J. L. Rosner. 1999. Structural requirements for
marbox function in transcriptional activation of mar/sox/rob regulon promoters in
Escherichia coli: sequence, orientation, and spatial relationship to the core promoter. Mol.
Microbiol. 34:431-441.
McIntosh, T. J., S. A. Simon, and R. C. MacDonald. 1980. The organization of n-alkanes in
lipid bilayers. Biochim. Biophys. Acta 24:445-463.
McMurry, L. M., M. Oethinger, and S. B. Levy. 1998. Overexpression of marA, soxS, or
acrAB produces resistance to triclosan in laboratory and clinical strains of Escherichia coli.
FEMS Microbiol. Lett. 166:305-309.
Meades, G., Jr., B. K. Benson, A. Grove, and G. L. Waldrop. 2009. A tale of two functions:
enzymatic activity and translational repression by carboxyltransferase. Nucleic Acids Res.
38:1217-1227.
Mendez-Perez, D., M. B. Begemann, and B. F. Pfleger. 2011. Modular synthase-encoding
gene involved in α-olefin biosynthesis in Synechococcus sp. strain PCC 7002. Appl. Environ.
Microbiol. 77:4264-4267.
Mijts, B. N., P. C. Lee, C. Schmidt-Dannert. 2004. Engineering carotenoid biosynthetic
pathways. Methods Enzymol. 388:315-329.
Miyahisa, I., M. Kaneko, N. Funa, H. Kawasaki, H. Kojima, Y. Ohnishi, and S. Horinouchi.
2005. Efficient production of (2S)-flavonones by Escherichia coli containing an artificial
biosynthetic gene cluster. Appl. Microbiol. Biotechnol. 68:498-504.
Mohan, S., T. M. Kelly, S. S. Eveland, C. R. H. Raetz, and M. S. Anderson. 1994. An
Escherichia coli gene (fabZ) encoding (3R)-hydroxymyristoyl acyl carrier protein dehydrase.
J. Biol. Chem. 269:32896-32903.
Mohan, D., C. U. Pittman, Jr., and P. H. Steele. 2006. Pyrolysis of wood/biomass for bio-oil: a
critical review. Energy Fuels 20:848-889.
Monroe, M. E., N. Tolić, N. Jaitly, J. L. Shaw, J. N. Adkins, and R. D. Smith. 2007. VIPER:
an advanced software package to support high-throughput LC-MS peptide identification.
Bioinformatics 23:2021-2023.
Morgan-Kiss, R. M. and J. E. Cronan. 2004. The Escherichia coli fadK (ydiD) gene encodes
an anaerobically regulated short chain acyl-CoA synthetase. J. Biol. Chem. 279:37324-37333.
225
Murakami, S., R. Nakashima, E. Yamashita, and A. Yamaguchi. 2002. Crystal structure
of bacterial multidrug efflux transporter AcrB. Nature 419:587-593.
Murata, N. and H. Wada. 1995. Acyl-lipid desaturases and their importance in the tolerance
and acclimatization to cold of cyanobacteria. Biochem. J. 308:1-8.
Murphy, M. J., J. D. Taylor, and R. L. McCormick. 2004. Compendium of experimental
cetane number data. NREL/SR-540-36805.
Nagakubo, S., K. Nishino, T. Hirata, A. Yamaguchi. 2002. The putative response regulator
BaeR stimulates multidrug resistance of Escherichia coli via a novel multidrug exporter
system, MdtABC. J. Bacteriol. 184:4161-4167.
Nakashima, R., K. Sakurai, S. Yamasaki, K. Nishino, and A. Yamaguchi. 2011. Structures of
the multidrug exporter AcrB reveal a proximal multisite drug-binding pocket. Nature
480:565-569.
Nannenga, B. L. and F. Baneyx. 2011. Reprogramming chaperone pathways to improve
membrane protein expression in Escherichia coli. Protein Sci. 20:1411-1420.
Nawabi, P., S. Bauer, N. Kyrpides, and A. Lykidis. 2011. Engineering Escherichia coli for
biodiesel production utilizing a bacterial fatty acid methyltransferase. Appl. Environ.
Microbiol. 77:8052-8061.
Newman, J. D., J. Marshall, M. Chang, F. Nowroozi, E. Paradise, D. Pitera, K. L. Newman,
and J. D. Keasling. 2006. High-level production of amorpha-4,11-diene in a two-phase
partitioning bioreactor of metabolically engineered Escherichia coli. Biotechnol. Bioeng.
95:684-691.
Nicolaou, S. A., S. M. Gaida, and E. T. Papoutsakis. 2010. A comparative view of metabolite
and substrate stress and tolerance in microbial bioprocessing: from biofuels and chemicals, to
biocatalysis and bioremediation. Metab. Eng. 12:307-331.
Nikaido, H. 1996. "Outer membrane," in F. C. Neihardt (ed.), Escherichia coli and Salmonella:
Cellular and Molecular Biology, 2nd edition. ASM Press, Washington, DC.
Nikaido, H. 2009. The limitations of LB medium. Small Things Considered: The Microbe Blog.
November 9. http://schaechter.asmblog.org/schaechter/ 2009/11/the-limitations-of-lbmedium.html.
Nishino, K., and A. Yamaguchi. 2001. Analysis of a complete library of putative drug
transporter genes in Escherichia coli. J. Bacteriol. 183:5803-5812.
Nishino, K., T. Honda, and A. Yamaguchi. 2005. Genome-wide analyses of Escherichia coli
gene expression responsive to the BaeSR two-component regulatory system. J. Bacteriol.
187:1763-1772.
Nordström, K. and S. Dasgupta. 2006. Copy-number control of the Escherichia coli
chromosome: a plasmidologist's view. EMBO Rep. 7:484-489.
Novo, D., N. G. Perlmutter, R. H. Hunt, and H. M. Shapiro. 1999. Accurate flow cytometric
membrane potential measurement in bacteria using diethyloxacarbocyanine and a ratiometric
technique. Cytometry 35:55-63.
O'Brien, W. J. and F. E. Frerman. 1977. Evidence for a complex of three β-oxidation enzymes
in Escherichia coli: induction and localization. J. Bacteriol. 132:532-540.
Ohlrogge, J., L. Savage, J. Jaworski, T. Voelker, and D. Post-Beittenmiller. 1995. Alteration
of acyl-acyl carrier protein pools and acetyl-CoA carboxylase expression in Escherichia coli
by plant medium chain acyl-acyl carrier protein thioesterase. Arch. Biochem. Biophys.
317:185-190.
226
Ohta, K., D. S. Beall, J. P. Mejia, K. T. Shanmugam, and L. O. Ingram. 1991. Genetic
improvement of Escherichia coli for ethanol production: chromosomal integration of
Zymomonas mobilis genes encoding pyruvate decarboxylase and alcohol dehydrogenase II.
Appl. Environ. Microbiol. 57:893-900.
Okochi, M., M. Kurimoto, K. Shimizu, and H. Honda. 2007. Increase of organic solvent
tolerance by overexpression of manXYZ in Escherichia coli. Appl. Microbiol. Biotechnol.
73:1394-1399.
Overath, P., G. Pauli, and H. U. Schairer. 1969. Fatty acid degradation in Escherichia coli. An
inducible acyl-CoA synthetase, the mapping of old-mutations, and the isolation of regulatory
mutants. Eur. J. Biochem. 7:559-574.
Peralta-Yahya, P. P. and J. D. Keasling. 2010. Advanced biofuel production in microbes.
Biotechnol. J. 5:147-162.
Peralta-Yahya, P. P., M. Ouellet, R. Chan, A. Mukhopadhyay, J. D. Keasling, and T. S. Lee.
2011. Identification and microbial production of a terpene-based advanced biofuel. Nat.
Commun. 2:483.
Podkovyrov, S. M. and T. J. Larson. 1996. Identification of promoter and stringent regulation
of transcription of the fabH, fabD, and fabG genes encoding fatty acid biosynthetic enzymes
of Escherichia coli. Nucleic Acids Res. 24:1747-1752.
Polacco, M. L. and Cronan, J. E., Jr. 1981. A mutant of Escherichia coli conditionally
defective in the synthesis of holo-[acyl carrier protein]. J. Biol. Chem. 256:5750-5754.
Polpitiya, A. D., W. J. Qian, N. Jaitly, V. A. Petyuk, J. N. Adkins, D. G. Camp, II, G. A.
Anderson, and R. D. Smith. 2008. DAnTE: a statistical tool for quantitative analysis of omics data. Bioinformatics. 24:1556-1558.
Poole, K. 2008. Bacterial multidrug efflux pumps serve other functions. Microbe 3:179-185.
Pos, K. M. 2009. Drug transport mechanism of the AcrB efflux pump. Biochim. Biophys. Acta
1794:782-793.
Potrykus, J. and G. Wegrzyn. 2003. The acrAB locus is involved in modulating intracellular
acetyl coenzyme A levels in a strain of Escherichia coli CM2555 expressing the
chloramphenicol acetyltransferase (cat) gene. Arch. Microbiol. 180:362-366.
Pramanik, J. and J. D. Keasling. 1997. Stoichiometric model of Escherichia coli metabolism:
incorporation of growth-rate dependent biomass composition and mechanistic energy
requirements. Biotechnol. Bioeng. 56:398-421.
Price, G. P., and A. C. St. John. 2000. Purification and analysis of expression of the stationary
phase-inducible Slp lipoprotein in Escherichia coli: role of the Mar system.FEMS Microbiol.
Lett. 193:51-56.
Providenti, M. A., J. M. O'Brien, R. J. Ewing, E. S. Paterson, and M. L. Smith. 2006. The
copy-number of plasmids and other genetic elements can be determined by SYBR-Greenbased quantitative real-time PCR. J. Microbiol. Methods 65:476-487.
Ramos, J. L., E. Duque, P. Godoy, and A. Segura. 1998. Efflux pumps involved in toluene
tolerance in Pseudomonas putida DOT-T1E. J. Bacteriol. 180:3323-3329.
Ramos, J. L., E. Duque, M.-T. Gallegos, P. Godoy, M. I. Ramos-González, A. Rojas, W.
Terán, and A. Segura. 2002. Mechanisms of solvent tolerance in Gram-negative bacteria.
Annu. Rev. Microbiol. 56:743-768 (2002).
227
Rawlings, B. J. 1998. Biosynthesis of fatty acids and related metabolites. Nat. Prod. Rep.
15:275-308.
Rehm, B. H. and A. Steinbüchel. 2001. Heterologous expression of the acyl-acyl carrier protein
thioesterase gene from the plant Umbellularia californica mediates polyhydroxyalkanoate
biosynthesis in recombinant Escherichia coli. Appl. Microbiol. Biotechnol. 55:205-209.
Reiling, K. K., Y. Yoshikuni, V. J. Martin, J. Newman, J. Bohlmann, and J. D. Keasling.
2004. Mono and diterpene production in Escherichia coli. Biotechnol. Bioeng. 87:200-212.
Reiser, S. and C. Somerville. 1997. Isolation of mutants of Acinetobacter calcoaceticus
deficient in wax ester synthesis and complementation of one mutation with a gene encoding a
fatty acyl coenzyme A reductase. J. Bacteriol. 179:2969-2975.
Rock, C. O., S. E. Goelz, and J. E. Cronan, Jr. 1981. Phospholipid synthesis in Escherichia
coli. Characteristics of fatty acid transfer from acyl-acyl carrier protein to sn-glycerol-3phosphate. J. Biol. Chem. 256:736-742.
Rosenberg, E. Y., D. Bertenthal, M. L. Nilles, K. P. Bertrand, and H. Nikaido. 2003. Bile
salts and fatty acids induce the expression of Escherichia coli AcrAB multidrug efflux pump
through their interaction with Rob regulatory protein. Mol. Microbiol. 48:1609-1619.
Rosner, J. L., B. Dangi, A. M. Gronenborn, and R. G. Martin. 2002. Posttranscriptional
activation of the transcriptional activator Rob by dipyridyl in Escherichia coli. J. Bacteriol.
184:1407-1416.
Roth, B. L., M. Poot, S. T. Yue, and P. J. Millard. 1997. Bacterial viability and antibiotic
susceptibility testing with SYTOX Green nucleic acid stain. Appl. Environ. Microbiol.
63:2421-2431.
Röttig, A., L. Wenning, D. Bröker, and A. Steinbüchel. 2010. Fatty acid alkyl esters:
perspectives for production of alternative biofuels. Appl. Microbiol. Biotechnol. 85:17131733.
Rowley, G., M. Spector, J. Kormanec, and M. Roberts. 2006. Pushing the envelope:
extracytoplasmic stress responses in bacterial pathogens. Nat. Rev. Microbiol. 4:383-394.
Rude, M. A. and A. Schirmer. 2009. New microbial fuels: a biotech perspective. Curr. Opin.
Microbiol. 12:274-281.
Rude, M. A., T. S. Baron, S. Brubaker, M. Alibhai, S. B. del Cardayre, and A. Schirmer.
2011. Terminal olefin (1-alkene) biosynthesis by a novel p450 fatty acid decarboxylase from
Jeotgalicoccus sp. Appl. Environ. Microbiol. 77:1718-1727.
Ruiz, C., L. M. McMurry, and S. B. Levy. 2008. Role of the multidrug resistance regulator
MarA in global regulation of the hdeAB acid resistance operon in Escherichia coli. J.
Bacteriol. 190:1290-1297.
Rutherford, B. J., R. H. Dahl, R. E. Price, H. L. Szmidt, P. I. Benke, A. Mukhopadhyay,
and J. D. Keasling. 2010. Functional genomic study of exogenous n-butanol stress in
Escherichia coli. Appl. Environ. Microbiol. 76:1935-1945.
Sakamoto, T. and D. A. Bryant. 1997. Temperature-regulated mRNA accumulation and
stabilization for fatty acid desaturase genes in the cyanobacterium Synechococcus sp. strain
PCC 7002. Mol. Microbiol. 23:1281-1292.
Salas, J. J. and J. B. Ohlrogge. 2002. Characterization of substrate specificity of plant FatA and
FatB acyl-ACP thioesterases. Arch. Biochem. Biophys. 403:25-34.
Salis, H. M., E. A. Mirsky, and C. A. Voigt. 2009. Automated design of synthetic ribosome
binding sites to control protein expression. Nat. Biotechnol. 27:946-950.
228
Sardessai, Y. and S. Bhosle. 2002. Tolerance of bacteria to organic solvents. Res. Microbiol.
153:263-268.
Sauer, U., F. Canonaco, S. Heri, A. Perrenoud, and E. Fischer. 2004. The soluble and
membrane-bound transhydrogenases UdhA and PntAB have divergent functions in NADPH
metabolism of Escherichia coli. J. Biol. Chem. 279:6613-6619.
Schirmer, A., M. A. Rude, X. Li, E. Popova, and S. B. del Cardayre. 2010. Microbial
biosynthesis of alkanes. Science 329:559-562.
Schmer, M. R., K. P. Vogel, R. B. Mitchell, and R. K. Perrin. 2008. Net energy of cellulosic
ethanol from switchgrass. Proc. Natl. Acad. Sci. USA 105:464-469.
Serrano-Vega, M. J., R. Garcés, and E. Martínez-Force. 2005. Cloning, characterization and
structural model of a FatA-type thioesterase from sunflower seeds (Helianthus annuus L.)
Planta 221:868-880.
Shen, C. R., E. I. Lan, Y. Dekishima, A. Baez, K. M. Cho, and J. C. Liao. 2011. Driving
forces enable high-titer anaerobic 1-butanol synthesis in Echerichia coli. Appl. Environ.
Microbiol. 77:2905-2915.
Shin, S., S. G. Song, D. S. Lee, J. G. Pan, and C. Park. 1997. Involvement of iclR and rpoS in
the induction of acs, the gene for acetyl coenzyme A synthetase of Escherichia coli K12.
FEMS Microbiol. Lett. 146:103-108.
Sikkema, J., J. A. M. De Bont, and B. Poolman. 1995. Mechanisms of membrane toxicity of
hydrocarbons. Microbiol. Rev. 59:201-222.
Smith, M. W. and J. W. Payne. 1992. Expression of periplasmic binding proteins for peptide
transport is subject to negative regulation by phosphate limitation in Escherichia coli. FEMS
Microbiol. Lett. 79:183-190.
Smyth, G. K. 2004. Linear models and empirical Bayes methods for assessing differential
expression in microarray experiments. Stat. Appl. Genet. Mol. Biol. 3:3.
Spencer, A. K., A. D. Greenspan, J. E. Cronan, Jr. 1978. Thioesterases I and II of Escherichia
coli. Hydrolysis of native acyl-acyl carrier protein thioesters. J. Biol. Chem. 253:5922-5926.
Standar, K., D. Mehner, H. Osadnik, F. Berthelmann, G. Hause, H. Lünsdorf, and T.
Brüser. 2008. PspA can form large scaffolds in Escherichia coli. FEBS Lett. 582:3585-3589.
Steen, E. J., Y. Kang, G. Bokinsky, Z. Hu, A. Schirmer, A. McClure, S. B. del Cardayre,
and J. D. Keasling. 2010. Microbial production of fatty-acid-derived fuels and chemicals
from plant biomass. Nature 463:559-562.
Stone, R. 2007. Can palm oil plantations come clean? Science, 317:1491.
Teerawanichpan, P., and X. Qiu. 2010. Fatty acyl-CoA reductase and wax synthase from
Euglena gracilis in the biosynthesis of medium-chain wax esters. Lipids 45:263-273.
Thomason, L., D. L. Court, M. Bubunenko, N. Costantino, H. Wilson, S. Datta, and A.
Oppenheim. 2007a. Recombineering: genetic engineering in bacteria using homologous
recombination. Curr. Protoc. Mol. Biol. 1:16.
Thomason, L. C., N. Costantino, and D. L. Court. 2007b. E. coli genome manipulation by P1
transduction. Curr. Protoc. Mol. Biol. 1:17.
Toomey, R. E. and S. J. Wakil. 1966. Studies on the mechanism of fatty acid synthesis. XV.
Preparation and general properties of β-ketoacyl acyl carrier protein reductase from
Escherichia coli. Biochim. Biophys. Acta 116:189-197.
229
Tsay, J.-T., W. Oh, T. J. Larson, S. Jackowski, and C. O. Rock. 1992. Isolation and
characterization of the β-ketoacyl-acyl carrier protein synthase III gene (fabH) from
Escherichia coli K-12. J. Biol. Chem. 267:6807-6814.
Voelker, T. A., and H. M. Davies. 1994. Alteration of the specificity and regulation of fatty
acid synthesis of Escherichia coli by expression of a plant medium-chain acyl-acyl carrier
protein thioesterase. J. Bacteriol. 176:7320-7327.
Voelker, T. A., A. C. Worrell, L. Anderson, J. Bleibaum, C. Fan, D. J. Hawkins, S. E.
Radke, and H. M. Davies. 1992. Fatty acid biosynthesis redirected to medium chains in
transgenic oilseed plants. Science 257:72-74.
Vorum, H., R. Brodersen, U. Kragh-Hansen, and A. O. Pedersen. 1992. Solubility of longchain fatty acids in phosphate buffer at pH 7.4. Biochim. Biophys. Acta 1126:135-142.
Wagner, S., L. Baars, A. J. Ytterberg, A. Klussmeier, C. S. Wagner, O. Nord, P.-Å. Nygren,
K. J. van Wijk, and J.-W. de Gier. 2007. Consequences of membrane protein
overexpression in Escherichia coli. Mol. Cell Proteomics 6:1527-1550.
Wagner, S., M. M. Klepsch, S. Schlegel, A. Appel, R. Draheim, M. Tarry, M. Högborn, K. J.
van Wijk, D. J. Slotboom, J. O. Persson, and J.-W. de Gier. 2008. Tuning Escherichia
coli for membrane protein overexpression. Proc. Natl. Acad. Sci. USA 105:14371-14376.
Wältermann, M., T. Stöveken, and A. Steinbüchel. 2007. Key enzymes for biosynthesis of
neutral lipid storage compounds in prokaryotes: properties, function, and occurrence of wax
ester synthases/acyl-CoA:diacylglycerol acyltransferases. Biochimie 89:230-242.
Wang, A. Y., Cronan, J. E., Jr. 1994. The growth phase-dependent synthesis of cyclopropane
fatty acids in Escherichia coli is the result of an RpoS(KatF)-dependent promoter plus
enzyme instability. Mol. Microbiol. 11:1009-1017.
Wang, C., S. H. Yoon, H. J. Jang, Y. R. Chung, J. Y. Kim, E. S. Choi, and S. W. Kim. 2011.
Metabolic engineering of Escherichia coli for α-farnesene production. Metab. Eng. 13:648655.
Wanner, B. L. 1996. "Phosphorus assimilation and control of the phosphate regulon," in F. C.
Neihardt (ed.), Escherichia coli and Salmonella: Cellular and Molecular Biology, 2nd edition.
ASM Press, Washington, DC.
Warner, D. M., and S. B. Levy. 2010. Different effects of transcriptional regulators MarA,
SoxS and Rob on susceptibility of Escherichia coli to cationic antimicrobial peptides
(CAMPs): Rob-dependent CAMP induction of the marRAB operon. Microbiology 156:570578.
Weeks, G. and S. J. Wakil. 1968. Studies on the mechanism of fatty acid synthesis. XVIII.
Preparation and general properties of the enoyl acyl carrier protein reductases from
Escherichia coli. J. Biol. Chem. 243:1180-1189.
Weiner, L., and P. Model. 1994. Role of an Escherichia coli stress-response operon in
stationary-phase survival. Proc. Natl. Acad. Sci. USA 91:2191-2195.
White, D. G., J. D. Goldman, B. Demple, and S. B. Levy. 1997. Role of the acrAB locus in
organic solvent tolerance mediated by expression of marA, soxS, or robA in Escherichia coli.
J. Bacteriol. 179:6122-6126.
Willis, R. M., B. D. Wahlen, L. C. Seefeldt, and B. M. Barney. 2011. Characterization of a
fatty acyl-CoA reductase from Marinobacter aquaeolei VT8: a bacterial enzyme catalyzing
the reduction of fatty acyl-CoA to fatty alcohol. Biochemistry 50:10550-10558.
230
Winters, K., P. L. Parker, and C. Van Baalen. 1969. Hydrocarbons of blue-green algae:
geochemical significance. Science 163:467-468.
Wiriyathanawudhiwong, N., I. Ohtsu, Z. D. Li, H. Mori, and H. Takagi. 2009. The outer
membrane TolC is involved in cysteine tolerance and overproduction in Escherichia coli.
Appl. Microbiol. Biotechnol. 81:903-913.
Wojtczak, L., and M. R. Więckowski. 1999. The mechanisms of fatty acid-induced proton
permeability of the inner mitochondrial membrane. J. Bioenerg. Biomembr. 31:447-455.
Wolfe, A. J. 2005. The acetate switch. Microbiol. Molec. Biol. Rev. 69:12-50.
Wu, J., A. H. Nhaesi, and A.-F. A. Asfour. 1999. Viscosities of eight binary liquid n-alkane
systems at 293.15 K and 298.15 K. J. Chem. Eng. Data 44:990-993.
Xu, Z., A. Colosimo, and D. C. Gruenert. 2003. Site-directed mutagenesis using the
megaprimer method. Methods Mol. Biol. 235:203-207.
Yang, S. Y. and H. Schulz. 1983. The large subunit of the fatty acid oxidation complex from
Escherichia coli is a multifunctional polypeptide. Evidence for the existence of a fatty acid
oxidation operon (fadAB) in Escherichia coli. J. Biol. Chem. 258:9780-9785.
Yanisch-Perron, C., J. Vieira, and J. Messing. 1985. Improved M13 phage cloning vectors
and host strains: nucleotide sequences of the M13mp18 and pUC19 vectors. Gene 33:103119.
Yoon, S. H., H. M. Park, J. E. Kim, S. H. Lee, M. S. Choi, J. Y. Kim, D. K. Oh, J. D.
Keasling, and S. W. Kim. 2007. Increased β-carotene production in recombinant
Escherichia coli harboring an engineered isoprenoid precursor pathway with mevalonate
addition. Biotechnol. Prog. 23:599-605.
Yoshikuni, Y., T. E. Ferrin, and J. D. Keasling. 2006. Designed divergent evolution of
enzyme function. Nature 440:1078-1082.
Yoshimura, M., T. Oshima, and N. Ogasawara. 2007. Involvement of the YneS/YgiH and
PlsX proteins in phospholipid biosynthesis in both Bacillus subtilis and Escherichia coli.
BMC Microbiol. 7:69.
Youngquist, J. T., R. M. Lennen, D. R. Ranatunga, W. H. Bothfeld, W. D. Marner II, and B.
F. Pfleger. 2012. Kinetic modeling of free fatty acid production in Escherichia coli based on
continuous cultivation of a plasmid free strain. Biotechnol. Bioeng. 109:1518-1527.
Yu, X., T. Liu, F. Zhu, and C. Khosla. 2011. In vitro reconstitution and steady-state analysis of
the fatty acid synthase from Escherichia coli. Proc. Natl. Acad. Sci. USA 108:18643-18648.
Yuan, L., T. A. Voelker, and D. J. Hawkins. 1995. Modification of the substrate specificity of
an acyl-acyl carrier protein thioesterase by protein engineering. Proc. Natl. Acad. Sci. USA
92:10639-10643.
Yuan, L., B. A. Nelson, and G. Caryl. 1996. The catalytic cysteine and histidine in the plant
acyl-acyl carrier protein thioesterases. J. Biol. Chem. 271:3417-3419.
Zgurskaya, H. I., G. Krishnamoorthy, A. Ntreh, and S. Lu. 2011. Mechanism and function of
the outer membrane channel TolC in multidrug resistance and physiology of enterobacteria.
Front. Microbiol. 2:189.
Zha, W., S. B. Ruben-Pitel, Z. Shao, and H. Zhao. 2009. Improving cellular malonyl-CoA
level in Escherichia coli via metabolic engineering. Metab. Eng. 11:192-198.
Zhang, Y.-M., H. Marrakchi, and C. O. Rock. 2002. The FabR (YijC) transcription factor
regulates unsaturated fatty acid biosynthesis in Escherichia coli.
231
Zhang, Y.-M. and C. O. Rock. 2008a. Acyltransferases in bacterial glycerophospholipid
synthesis. J. Lipid Res. 49:1867-1874.
Zhang, Y.-M. and C. O. Rock. 2008b. Membrane lipid homeostasis in bacteria. Nat. Rev.
Microbiol. 6:222-233.
Zhang, X., M. Li, A. Agrawal, and K.-Y. San. 2011a. Efficient free fatty acid production in
Escherichia coli using plant acyl-ACP thioesterases. 13:713-722.
Zhang, X., A. Agrawal, and K.-Y. San. 2011b. Improving fatty acid production in Escherichia
coli through the overexpression of malonyl CoA-acyl carrier protein transacylase. Biotechnol.
Prog. 28:60-65.
Zhang, F., J. M. Carothers, and J. D. Keasling. 2012. Design of a dynamic sensor-regulator
system for production of chemicals and fuels derived from fatty acids. Nat. Biotechnol.
30:354-359.
Zhao, H., J. E. Holladay, H. Brown, and Z. C. Zhang. 2007. Metal chlorides in ionic liquid
solvents convert sugars to hydroxymethylfurfural. Science 316:1597-1600.
Zheng, D., C. Constantinidou, J. L. Hobman, and S. D. Minchin. 2004. Identification of the
CRP regulon using in vitro and in vivo transcriptional profiling. Nucleic Acids Res. 32:58745893.
Zhou, S., T. B. Grabar, K. T. Shanmugam, and L. O. Ingram. 2006. Betaine tripled the
volumetric productivity of D(-)-lactate by Escherichia coli strain SZ132 in mineral salts
medium. Biotechnol. Lett. 28:671-676.
Zhu, K., Y.-M. Zhang, and C. O. Rock. 2009. Transcriptional regulation of membrane lipid
homeostasis in Escherichia coli. J. Biol. Chem. 284:34880-34888.
Zimmer, J. S., M. E. Monroe, W. J. Qian, and R. D. Smith. 2006. Advances in proteomics
data analysis and display using an accurate mass and time tag approach. Mass Spectrom. Rev.
25:450-482.
232
Appendices
Appendix I: Table of bacterial strains.
Strain
Relevant genotype/property a
F- λ- ilvG- rfb-50 rph-1
K-12 MG1655 fadD::kan
K-12 MG1655 ∆fadD
K-12 MG1655 araBAD::cat
K-12 MG1655 ∆araBAD
K-12 MG1655 ∆fadD araBAD::cat
K-12 MG1655 ∆fadD ∆araBAD
F- mcrA ∆(mrr-hsdRMS-mcrBC) φ80lacZ∆M15
∆lacX74 recA1 endA1 araD139 ∆(ara, leu)7697
galU galK λ- rpsL nupG
DY330
K-12 W3110 ∆lacU169 gal490 pgl∆8 λcI857 ∆(cro-bioA)
(TetR)
RL08-BTE
K-12 MG1655 ∆araBAD fadD::araC-PBAD-BTE
RL08-BTE-H204A K-12 MG1655 ∆araBAD fadD::araC-PBAD-BTE-H204A
BW25113
lacIq rrnB3 F- ∆(araD-araB)567 ∆lacZ4787(::rrnB-3) λrph-1 ∆(rhaD-rhaB)568 hsdR514
JW5249-1
BW25113 marA752::kan
JW4359-1
BW25113 rob-721::kan
JW4023-5
BW25113 soxS756::kan
JW1296-5
BW25113 pspF739::kan
RL09
K-12 MG1655 ∆fadD ∆araBAD marA::kan
RL10
K-12 MG1655 ∆fadD ∆araBAD rob::kan
RL11
K-12 MG1655 ∆fadD ∆araBAD soxS::kan
RL12
K-12 MG1655 ∆fadD ∆araBAD pspF::kan
RL13
K-12 MG1655 ∆fadD ∆araBAD ∆pspF
TY05
K-12 MG1655 fadD::Ptrc-BTE fadE::Ptrc-BTE
fadAB::Ptrc-BTE
TY06
K-12 MG1655 fadD::Ptrc-BTE-H204A fadE::Ptrc-BTEH204A fadAB::Ptrc-BTE-H204A
BW27269
BW25113 araFGH::kan903
K-12 MG1655
RL01
RL02
NRD204
RL06
RL07
RL08
DH10B
BW27270
BW25113 Φ(∆araEp kan PCP18-araE)
RL14
RL15
RL16
K-12 MG1655 ∆fadD ∆araBAD araFGH::kan
K-12 MG1655 ∆fadD ∆araBAD ∆araFGH
K-12 MG1655 ∆fadD ∆araBAD ∆araFGH Φ(∆araEp
kan PCP18-araE)
Source or
reference
CGSC
Chapter 2
Chapter 2
DeLay 2007
Chapter 2
Chapter 2
Chapter 2
Invitrogen
Thomason
2007a
Chapter 2
Chapter 2
Baba 2006
Baba 2006
Baba 2006
Baba 2006
Baba 2006
Chapter 3
Chapter 3
Chapter 3
Chapter 3
Chapter 3
Youngquist
2012
Youngquist
2012
Khlebnikov
2010
Khlebnikov
2010
Chapter 5
Chapter 5
Chapter 5
233
Appendix I (cont.): Table of bacterial strains.
Strain
Relevant genotype/property a
Source or
reference
RL08ara
K-12 MG1655 ∆fadD ∆araBAD ∆araFGH Φ(∆araEp
PCP18-araE)
BW25113 ∆fabR751::kan
K-12 MG1655 ∆fadD ∆araBAD ∆araFGH Φ(∆araEp
PCP18-araE) fabR::kan
K-12 MG1655 ∆fadD ∆araBAD ∆araFGH Φ(∆araEp
PCP18-araE) ∆fabR
DY330 fabA::kan
K-12 MG1655 fadD::Ptrc-BTE fadE::Ptrc-BTE
fadAB::Ptrc-BTE araFGH::kan
K-12 MG1655 fadD::Ptrc-BTE fadE::Ptrc-BTE
fadAB::Ptrc-BTE ∆araFGH
K-12 MG1655 fadD::Ptrc-BTE fadE::Ptrc-BTE
fadAB::Ptrc-BTE ∆araFGH Φ(∆araEp kan PCP18-araE)
K-12 MG1655 fadD::Ptrc-BTE fadE::Ptrc-BTE
fadAB::Ptrc-BTE ∆araFGH Φ(∆araEp PCP18-araE)
K-12 MG1655 fadD::Ptrc-BTE fadE::Ptrc-BTE
fadAB::Ptrc-BTE ∆araFGH Φ(∆araEp PCP18-araE)
araBAD::cat
K-12 MG1655 fadD::Ptrc-BTE fadE::Ptrc-BTE
fadAB::Ptrc-BTE ∆araFGH Φ(∆araEp PCP18-araE)
∆araBAD
BW25113 tolC732::kan
BW25113 ∆fadL752::kan
BW25113 ∆prc-755::kan
BW25113 ∆acrD790::kan
BW25113 ∆mdtG723::kan
BW25113 ∆mdtK740::kan
BW25113 ∆cmr-742::kan
BW25113 ∆ompF746::kan
DY330 acrAB::kan
DY330 mdtEF::kan
DY330 acrEF::kan
DY330 emrAB::kan
DY330 mdtABCD::kan
TY05 marA::kan
TY06 marA::kan
TY05 rob::kan
TY06 rob::kan
Chapter 5
JW3935-4
RL17
RL18
RL31
RL36
RL37
RL38
RL39
RL40
TY05ara
JW5503-1
JW2341-1
JW1819-1
JW2454-1
JW1040-1
JW1655-1
JW0826-1
JW0912-1
RL46
RL47
RL48
RL49
RL50
RL51
RL52
RL53
RL54
Baba 2006
Chapter 5
Chapter 5
Chapter 6
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Baba 2006
Baba 2006
Baba 2006
Baba 2006
Baba 2006
Baba 2006
Baba 2006
Baba 2006
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
234
Appendix I (cont.): Table of bacterial strains.
Strain
Relevant genotype/property a
Source or
reference
RL55
RL56
RL57
RL58
RL59
RL60
RL61
RL62
RL63
RL64
RL65
RL66
RL67
RL68
RL69
RL70
RL71
RL72
RL73
RL74
RL75
RL76
RL77
RL78
RL79
RL80
RL81
RL82
RL83
RL84
RL85
RL86
RL87
RL88
RL89
RL90
RL91
RL92
RL93
RL94
TY05 soxS::kan
TY06 soxS::kan
TY05 tolC::kan
TY06 tolC::kan
TY05 fadL::kan
TY06 fadL::kan
TY05 prc::kan
TY06 prc::kan
TY05 acrD::kan
TY06 acrD::kan
TY05 mdtG::kan
TY06 mdtG::kan
TY05 mdtK::kan
TY06 mdtK::kan
TY05 cmr::kan
TY06 cmr::kan
TY05 ompF::kan
TY06 ompF::kan
TY05 acrAB::kan
TY06 acrAB::kan
TY05 mdtEF::kan
TY06 mdtEF::kan
TY05 acrEF::kan
TY06 acrEF::kan
TY05 emrAB::kan
TY06 emrAB::kan
TY05 mdtABCD::kan
TY06 mdtABCD::kan
TY05 ∆marA
TY06 ∆marA
TY05 ∆rob
TY06 ∆rob
TY05 ∆soxS
TY06 ∆soxS
TY05 ∆tolC
TY06 ∆tolC
TY05∆fadL
TY06 ∆fadL
TY05 ∆prc
TY06 ∆prc
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
235
Appendix I (cont.): Table of bacterial strains.
Strain
Relevant genotype/property a
Source or
reference
RL95
RL96
RL97
RL98
RL99
RL100
RL101
RL102
RL103
RL104
RL105
RL106
RL107
RL108
RL109
RL110
RL111
RL112
RL113
RL114
RL115
RL116
RL117
RL118
RL119
RL120
RL121
RL122
RL123
RL124
RL125
RL126
RL127
RL128
RL129
RL130
RL131
RL132
RL133
RL134
TY05 ∆acrD
TY06 ∆acrD
TY05 ∆mdtG
TY06 ∆mdtG
TY05 ∆mdtK
TY06 ∆mdtK
TY05 ∆cmr
TY06 ∆cmr
TY05 ∆ompF
TY06 ∆ompF
TY05 ∆acrAB
TY06 ∆acrAB
TY05 ∆mdtEF
TY06 ∆mdtEF
TY05 ∆acrEF
TY06 ∆acrEF
TY05 ∆emrAB
TY06 ∆emrAB
TY05 ∆mdtABCD
TY06 ∆mdtABCD
TY05 ∆acrAB acrD::kan
TY06 ∆acrAB acrD::kan
TY05 ∆acrAB mdtEF::kan
TY06 ∆acrAB mdtEF::kan
TY05 ∆acrAB acrEF::kan
TY06 ∆acrAB acrEF::kan
TY05 ∆acrAB emrAB::kan
TY06 ∆acrAB emrAB::kan
TY05 ∆acrAB mdtABCD::kan
TY06 ∆acrAB mdtABCD::kan
TY05 ∆acrAB ∆acrD
TY06 ∆acrAB ∆acrD
TY05 ∆acrAB ∆mdtEF
TY06 ∆acrAB ∆mdtEF
TY05 ∆acrAB ∆acrEF
TY06 ∆acrAB ∆acrEF
TY05 ∆acrAB ∆emrAB
TY06 ∆acrAB ∆emrAB
TY05 ∆acrAB ∆mdtABCD
TY06 ∆acrAB ∆mdtABCD
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
236
Appendix I (cont.): Table of bacterial strains.
Strain
Relevant genotype/property a
Source or
reference
RL135
DH5α
TY05ara acrAB::kan
fhuA2 ∆(argF-lacZ)U169 phoA glnV44 Φ80 ∆(lacZ)M15
gyrA96 recA1 relA1 endA1 thi-1 hsdR17
Chapter 4
Invitrogen
237
Appendix II: Table of plasmids.
Plasmid
Relevant genotype/property a
pKD13
template plasmid, R6K gamma origin, AmpR, KanR
pKD46
pCP20
pBAD33
pBAD18
pBAD24
pRL1
pRL2
pBAD33-ACC
pBAD18-BTE
pBAD18-BTEH204A
pBT-2
pUC19
pBAD34
pBAD34-BTE
pBAD34-BTEH204A
pBAD35
Source or
reference
Datsenko
2000
carries Red recombinase under ParaB control, R101 origin, Datsenko
repA101(ts), AmpR
2000
carries yeast FLP recombinase under constitutive
Cherepanov
promoter, pSC101 origin, λ cI857+, λ pR Repts, AmpR,
1995
CmR
PBAD promoter, pACYC origin, CmR
Guzman 1995
R
PBAD promoter, pBR322 origin, Amp
Guzman 1995
PBAD promoter, pBR322 origin, AmpR
Guzman 1995
pBAD33 carrying accD under PBAD control, CmR
Chapter 2
Chapter 2
pRL1 carrying accDA under PBAD control, CmR
pRL2 carrying accDABC under PBAD control, CmR
Chapter 2
R
pBAD18 carrying BTE under PBAD control, Amp
Chapter 2
pBAD18 carrying BTE with H204A under PBAD control, Chapter 2
AmpR
Lynch 2006
pBBR1-MCS origin, KanR
R
pUC origin, lacZα, Amp
New England
Biolabs
PBAD promoter, pUC origin, AmpR
Chapter 2
R
pBAD34 carrying BTE under PBAD control, Amp
Chapter 2
Chapter 2
pBAD34 carrying BTE-H204A under PBAD control,
R
Amp
pBT-2 containing araC-PBAD-MCS-rrnB fragment from
Chapter 2
R
pBAD18, Kan
pBAD35 carrying BTE under PBAD control, KanR
Chapter 2
R
pBAD35 carrying BTE-H204A under PBAD control, Kan Chapter 2
pBAD35-BTE
pBAD35-BTEH204A
pBAD34-BTE-kan pBAD34-BTE with FRT-kan-FRT cassette 3' to araC,
KanR, AmpR
pBAD34-BTEpBAD34-BTE-H204A with FRT-kan-FRT cassette 3' to
H204A-kan
araC, KanR, AmpR
pBAD33-BTE
pBAD33 carrying BTE under PBAD control, CmR
pBAD33-BTEpBAD33 carring BTE-H204A under PBAD control, CmR
H204A
pTrc99A
Ptrc promoter, pBR322 origin, AmpR
pTrc99A-BTE
pTrc99a carrying BTE under Ptrc control, AmpR
pTrc99A-BTEpTrc99a carrying BTE-H204A under Ptrc control, AmpR
H204A
pBAD18-MarA
pBAD18 carrying marA under PBAD control, AmpR
Chapter 2
Chapter 2
Chapter 2
Chapter 2
Amann 1988
Hoover 2011
Hoover 2011
Chapter 3
238
Appendix II (cont.): Table of plasmids.
Plasmid
Relevant genotype/property a
pBAD18 carrying rob under PBAD control, AmpR
pBAD18 carrying soxS under PBAD control, AmpR
pBAD18 carrying pspABCDE under PBAD control, AmpR
pBAD33 with araC-C280* mutation
pBAD33* carrying tolC under PBAD control, CmR
pBAD33* carrying acrAB under PBAD control, CmR
pBAD33* carrying acrAB-tolC artificial operon
under PBAD control, CmR
pBAD33* carrying mdtEF under PBAD control, CmR
pBAD33* carrying mdtEF-tolC artificial operon
under PBAD control, CmR
pBAD33* carrying emrAB under PBAD control, CmR
pBAD33* carrying emrAB-tolC artificial operon
under PBAD control, CmR
pBAD33* carrying mdtABCD under PBAD control,
CmR
pBAD33* carrying mdtABCD-tolC artificial operon
under PBAD control, CmR
pBAD18 carrying Geobacillus sp. TE under PBAD control,
AmpR
pBAD18-GeoTE- pBAD18 carrying Geobacillus sp. TE with H173A
H173A
mutation under PBAD control, AmpR
pBAD18-ClosTE pBAD18 carrying Clostridium thermocellum TE under
PBAD control, AmpR
pBAD18-ClosTE- pBAD18 carrying Clostridium thermocellum TE with
H171A
H171A mutation under PBAD control, AmpR
pBAD33*-fabR
pBAD33* carrying fabR under PBAD control, CmR
pBAD33*-desA- pBAD33* carrying desA from Bacillus subtilis 3610
His
with C-terminal His tag under PBAD control, CmR
pBAD18-desC
pBAD18 carrying desC from Synechococcus sp. PCC7002
under PBAD control, AmpR
pBAD33*-kan
Template plasmid with kan cassette from pKD13 inserted
behind araC-C280* in pBAD33*, CmR, KanR
pBAD33-C280*- Template plasmid with kan cassette from pKD13 inserted
desA-His-kan
behind araC-C280* in pBAD33*-desA-His, CmR, KanR
KanR
pBAD33-C280*- Template plasmid with kan cassette from pKD13 inserted
desC-kan
behind araC-C280* in pBAD33*-desC, CmR, KanR
pBAD18-Rob
pBAD18-SoxS
pBAD18-Psp
pBAD33*
pBAD33*-tolC
pBAD33*-acrAB
pBAD33*-acrABtolC
pBAD33*-mdtEF
pBAD33*-mdtEFtolC
pBAD33*-emrAB
pBAD33*-emrABtolC
pBAD33*mdtABCD
pBAD33*mdtABCD-tolC
pBAD18-GeoTE
a
Source or
reference
Chapter 3
Chapter 3
Chapter 3
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 4
Chapter 5
Chapter 5
Chapter 5
Chapter 5
Chapter 5
Chapter 6
Chapter 6
Chapter 6
Chapter 6
Chapter 6
Abbreviations: Amp, ampicillin; Cm, chloramphenicol; Kan, kanamycin; R, resistance; ts, temperature sensitive.
Appendix III: Table of oligonucleotide primers.
Primer name
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
fadDKO_fwd
239
Sequence (5' to 3')a
GGTTGCGATGACGACGAACACGCATTTTAGAGGTGAAGAAGTGT
AGGCTGGAGCTGCTTC
fadDKO_rev
CGCCGGATTAACCGGCGTCTGACGACTGACTTAACGCTCAATTC
CGGGGATCCGTCGACC
fadDKO_colPCR_fwd ACGGCATGTATATCATTTGGG
fadDKO_colPCR_rev
CTTTAGTGGGCGTCAAAAAAAAC
araBADKO_colPCR_fwd AAGCGGGACCAAAGCCATGAC
araBADKO_colPCR_rev AGGAGACTTCTGTCCCTTGCG
BTE-int_rev
CGTTAGATTTCGGTAACTCATCACGAAACTCCACCAGTTACATC
CGCCAAAACAGCCAAG
accD_fwd
CCCGAGCTCAGGTCCCTAATGAGCTGGATTGAAC
accD_rev
CCCCCCGGGTCAGGCCTCAGGTTCCTGATCC
accA_fwd
GGGCCCGGGAGGAATACTATGAGTCTGAATTTCCTTG
accA_rev
GGGGTCGACCTCGAGTTTACGCGTAACCGTAGCTCATC
accBC_fwd
CCCCTCGAGACGGAACCCACTCATGGATATTC
accBC_rev
CCCGCATGCTTATTTTTCCTGAAGACCGAGTTTTTTC
BTE-H204A_mega_rev TCTCATCCGCCAAAAC
BTE-H204A_mega_mut TGTTAATCAGGCTGTGAACAACCTGAAATACG
BTE-H204A_mega_fwd TTGGGCTAGCGAATTC
pBAD18-to-pBT_fwd
TTATGACAACTTGACGGCTACATC
pBAD18-to-pBT_rev
AGAGTTTGTAGAAACGCAAAAAGGC
BTE-to-pBAD35_fwd
ACGCTTTTTATCGCAACTCTC
BTE-to-pBAD35_rev
GGGGCATGCTTAAACACGAGGTTCGCGC
pBAD33ara_fwd
GGGCTCGAGTTATGACAACTTGACGGCTACATC
pBAD33ara_rev
GGGAGATCTAGAGTTTGTAGAAACGCAAAAAGGC
pUC19ara_fwd
GGGCTCGAGGTGCCTAATGAGTGAGCTAACTC
pUC19ara_rev
GGGAGATCTTAGTTAAGCCAGCCCCGACAC
kan-to-pUC19_fwd
GGGCTCGAGGTGTAGGCTGGAGCTGCTTC
kan-to-pUC19_rev
CCCCTCGAGATTCCGGGGATCCGTCGACC
qPCR_BTE_fwd
CTGTCTACCATCCCGGAC
qPCR_BTE_rev
TCAGTTTTTGCAGTTTCTTGATTTCG
qPCR_ompA_fwd
TGTTGAGTACGCGATCACTC
qPCR_ompA_rev
GTTGTCCGGACGAGTGC
rrs_qPCR_fwd
AATGTTGGGTTAAGTCCCGCAACG
rrs_qPCR_rev
GACTTGACGTCATCCCCACC
pspD_qPCR_fwd
CTGGGCGATAAAATCAGTTGCTCG
pspD_qPCR_rev
ACGCTGTGCCAGTTTATTAGCAGC
fabA_qPCR_fwd
CACTTTATTGGCGATCCGGTTATG
fabA_qPCR_rev
CTTTACCTTCGCCGCCCAGC
Appendix III (cont.): Table of oligonucleotide primers.
240
Primer namea
Sequence (5' to 3')b
37
38
39
40
41
42
43
44
45
CAGTATCGCCTTATTCTGACGAAG
AGCCATGCTGACGATTAATGCTATG
CGGCGCGGCAATATGTGAAC
CAACTGGCTGCGTGGTTTGTTC
CGTTTCGCCACTTCGCCG
GCGTCGAAACTGAGGAGCAG
CATGCCAGGATGAGTTAGCG
CCGCATCCGGCAAGTTGTA
GGGTCTAGAAGGAGGATTATAAAATGTCCAGACGCAATACT
GACG
GGCAAGCTTCTAGCTGTTGTAATGATTTAATGGATG
GGGTCTAGAAGGAGGATTATAAAATGGATCAGGCCGGCATT
ATTCG
CGGAAGCTTAACGACGGATCGGAATCAGCAG
GGGTCTAGAAGGAGGATTATAAAATGTCCCATCAGAAAATT
ATTCAGGATC
GGGAAGCTTACAGGCGGTGGCGATAATC
GGGGAGCTCAGGACAACATTATGGGTATTTTTTCTCG
GGGCCCGGGTTAACCTTTGACCTTCGGCATTGCG
GGTACCAAAGACAACAAGGATTTCC
CTATACTTACATGTCTGTAAAGCGCG
CATGGCGACCAACAATACTC
TTCCGCCTCAATATGACG
CGATTGATGGTAAAACGGTGTTGTT
CTGAAGTGCGGATAAAAACAGCAA
GGAGTGAATAAGTAACGCATCC
GCTGTCGCGTCTTATCGTGC
ACCATTGACCAATTTGAAATCGGACACTCGAGGTTTACATGTGT
AGGCTGGAGCTGCTTC
CCGCTTACGCGGCCTTAGTGATTACACGTTGTATCAATTCCGGG
GATCCGTCGACC
TTAAAGAACCGTTATTTCTCAAGAATTTTCAGGGACTAAAGTGT
AGGCTGGAGCTGCTTC
CTGAACCTTCATGTTCAACCTTACTCTCATTTACACGTTAATTC
CGGGGATCCGTCGACC
TTGGGTAAATAACGCGCTTTTGGTTTTTTGAGGAATAGTAGTGT
AGGCTGGAGCTGCTTC
ATATAAAGGCACCCGAAAGCGCCTTTATGTTTCTGATTTAATTC
CGGGGATCCGTCGACC
inaA_qPCR_fwd
inaA_qPCR_rev
marA_colPCR_fwd
marA_colPCR_rev
soxS_colPCR_fwd
soxS_colPCR_rev
pspF_colPCR_fwd
pspF_colPCR_rev
marA_fwd
46 marA_rev
47 rob_fwdc
48 rob_revc
49 soxS_fwd
50
51
52
53
54
55
56
57
58
59
60
61
soxS_rev
pspABCDE_fwd
pspABCDE_rev
araFGH_colPCR_fwd
araFGH_colPCR_rev
ParaE_colPCR_fwd
ParaE_colPCR_rev
fadE_colPCR_fwd
fadE_colPCR_rev
fadAB_colPCR_fwd
fadAB_colPCR_rev
acrAB_KO_fwd
62 acrAB_KO_rev
63 mdtEF_KO_fwd
64 mdtEF_KO_rev
65 acrEF_KO_fwd
66 acrEF_KO_rev
Appendix III: Table of oligonucleotide primers.
Primer namea
67 emrAB_KO_fwd
241
Sequence (5' to 3')b
TCGGCTCAGCCGATGAGTTAAGAAGATCGTGGAGAACAATGTGT
AGGCTGGAGCTGCTTC
68 emrAB_KO_rev
TGAACTGGCTTAGTTGTACTTAGTGCGCACCGCCTCCGCCATTC
CGGGGATCCGTCGACC
69 mdtABCD_KO_fwd
ATTCCGCGAAACGTTTCAGGAAGAGAAACTCTTAACGATGGTGT
AGGCTGGAGCTGCTTC
70 mdtABCD_KO_rev
GTAATACCGGGTCGCCAGAACTTCATTGCGCGCTCCTTTTATTC
CGGGGATCCGTCGACC
71 rob_colPCR_fwd
GTCAAGCCCTAAAACATACTCTAC
72 rob_colPCR_rev
GATGCCTGGTGAGTACGATTC
73 tolC_colPCR_fwd
ACGCCAACCTTTTGCGGTAG
74 tolC_colPCR_rev
GAAGAATGCGGCAGATAACC
75 fadL_colPCR_fwd
GCCACTGGTCTGATTTCTAAG
76 fadL_colPCR_rev
CAAAACTCAGGAGGTAAGCAATG
77 prc_colPCR_fwd
AGTCTGGATAGTGCGTAAGTC
78 prc_colPCR_rev
GCTGAATTCGGGTATGTCTTTG
79 acrD_colPCR_fwd
TCGTACCTTGCCGCTACAGTG
80 acrD_colPCR_rev
CAAAGTACAAACAGCAAGAACCCG
81 mdtG_colPCR_fwd
GATAAAAGCTCTCTGGATTGCG
82 mdtG_colPCR_rev
GCAGCGTCAATGGCTTCTTC
83 mdtK_colPCR_fwd
GAATGGCTATTTTTTCACTGGAG
84 mdtK_colPCR_rev
GATAACAGATGCCAGTCGG
85 cmr_colPCR_fwd
GTAGCTATACTCGTAATAATGTAAG
86 cmr_colPCR_rev
CCTTATGTTCGCCATCTTGC
87 ompF_colPCR_fwd
AGACACCAAACTCTCATCAATAGTTC
88 ompF_colPCR_rev
AACGCAGGCTGTTTTTGCAAGAC
89 acrAB_colPCR_fwd
CAGGCGTTAGATTTACATACATTTG
90 acrAB_colPCR_rev
CCGTGGTTAATACTGGTTTTCG
91 mdtEF_colPCR_fwd
GTGCCTGTATCCCACCTTAC
92 mdtEF_colPCR_rev
GATGACGAATGGCTGGAGTG
93 acrEF_colPCR_fwd
CCCGCGTCAAAATAAAACAGTAG
94 acrEF_colPCR_rev
CGGAGGTTATAAATCTTGCGG
95 emrAB_colPCR_fwd
CTCCCGTCTCGACCAGATGG
96 emrAB_colPCR_rev
CTCGTTGCAGGAAGCGCAGG
97 mdtABCD_colPCR_fwd CATTCAGACGATTCCAGACA
98 mdtABCD_colPCR_rev AGCCACGCTCAAAACTGATAC
99 pBAD33-C280*_fwd
GGGCTCGAGTTAACCGGCACGGAACTCGCTCG
100 pBAD33-C280*_rev
GGGCTCGAGTTGGTAACGAATCAGACAATTGACGGC
101 fabR_colPCR_fwd
GTACGTAAAAGAACCGGCCAAAG
Appendix III: Table of oligonucleotide primers.
242
Primer namea
Sequence (5' to 3')b
102 fabR_colPCR_rev
103 fabR_RBSeng_fwd
GCTGCTGCTGGCGTTAGTTG
GCGCCCGGGAATCATACCCCATCGGTACCCATAGGTAGGTTCAC
ATGTTCATTCTCTGGTATAGTGCC
GGGTCTAGATTACTCGTCCTTCACATTTCCCG
CTATCCTGCGTGCTTCAATAAAATAAGGCTTACAGAGAACGTGT
AGGCTGGAGCTGCTTC
CGGAGGTTTCGCCTTTTGATACTCTGTCTGATTATAATCAATTC
CGGGGATCCGTCGACC
TACGTTGGCTGAACTGGTTTATTCC
CATGGCGGACAGGGTAGTAATG
GCGCCCGGGAGGAGGATTATAAAATGACGGAACAAACCATTGCA
CATAAACAAAAACAG
GGGTCTAGATCAGTGGTGATGGTGATGATGGGCATTCTTCCGCA
GCTTCTG
CCCTCTAGAAGGAGGAAAAAAAAATGACTGTTGCCACTTCCCAA
AAACTCC
CCCAAGCTTTTACTTGGCCGGAGCGAGACGAATATTTTTGG
GCGCCCGGGAAGGAGGTATATAAAATGG
GGGAAGCTTAGTGGTGATGGTGATGATGGCTTTCACGAACAATT
GCTGCTTCAAA
GCGCCCGGGAGGAGGTAAATTAAATGC
GGGAAGCTTAGTGGTGATGGTGATGATGGCTCTGAATTTTCTGC
CAAATGGTTTC
ATGCATGGCGTTGATACCCCAATC
AACACTTCACGGATAGCTGCCAG
GGGTCTAGAAGGAGGTAAATAAAATGGTGCATCGTATCCTTGCC
G
GGGAAGCTTCAGAGGTTCTCGTAGCGGTTC
TTTTCCCGGGCACTCGAGGTTTACATATGAAC
CCCTCTAGATCAATGATGATCGACAGTATGG
GGGTCTAGATTCACCACAAGGAATGCAAATGAAGAAATTGCTCC
CCATTCTTATC
GGGGCATGCTCAGTTACGGAAAGGGTTATGACC
GCCCCCGGGTCGTGGAGAACAATATGAGCGCAAATG
GGGTCTAGATTAGTGCGCACCGCCTCCG
GGGGAGCTCAAGAATTTTCAGGGACTAAAA
AAACCCGGGCATTTACACGTTACGCTTTTT
AAAGAGCTCCAGGAAGAGAAACTCTTAACGATG
AAACCCGGGACTTCATTGCGCGCTCCTTTT
104 fabR_rev
105 fabA_KO_fwd
106 fabA_KO_rev
107 fabA_colPCR_fwd
108 fabA_colPCR_rev
109 desA_fwd
110 desA-His_rev
111 desC_fwd
112 desC_rev
113 GeoTE_fwd
114 GeoTE-His_rev
115 ClosTE_fwd
116 ClosTE-His_rev
117 fabB_qPCR_fwd
118 fabB_qPCR_rev
119 cti_fwd
120 cti_rev `
121 acrAB_fwd
122 acrAB_rev
123 tolC_fwd
124 tolC_rev
125 emrAB_fwd
126 emrAB_rev
127 mdtEF_fwd
128 mdtEF_rev
129 mdtABCD_fwd
130 mdtABCD_rev
Appendix III: Table of oligonucleotide primers.
Primer namea
Sequence (5' to 3')b
131 acpP_qPCR_fwd
132 acpP_qPCR_rev
133 fabR_qPCR_fwd
134 fabR_qPCR_rev
GCGGATTCTCTTGACACCGTTG
TGATGTAATCAATGGCAGCCTGAAC
ATTGCGGAACTTGCGGACTATCTG
TGCCGACGTTGTTCGACGCC
a
Primers containing 'qPCR' in the name were used for amplification of cDNA in quantitative PCR reactions.
Primers containing 'colPCR' were used colony PCR verification of chromosomal gene insertions and deletions.
Primers containing restriction sites were used for amplification of insertions for cloning.
b
Restriction sites are underlined
c
These primers were also used for colony PCR verification of the rob::kan insertion in Chapter 3
243
244
Appendix IV
Table of genes with fold-changes between BTE-expressing and BTE-H204A expressing cultures
greater than 1.8 and with P-values less than 0.05 (EZ glucose shake flask cultures) or less than
0.10 (EZ glycerol fermentor cultures) in both shake flask and fermentor cultures for at least one
sampling time. Sampling times are defined in Chapter 3. Fold-changes are reported as linear
values.
locus
b0030
b0048
b0052
b0080
b0090
b0096
b0097
b0117
b0120
b0135
b0221
b0312
b0428
b0429
b0430
b0464
b0475
b0729
b0842
b0846
b0850
b0851
b0853
b0861
b0864
b0950
b0965
b1002
b1014
b1036
b1053
b1165
b1166
b1167
b1256
b1304
b1305
b1306
b1307
b1308
b1384
b1423
b1444
b1463
b1498
b1514
b1515
b1530
b1531
b1610
b1611
b1655
b1682
b1744
b1745
b1773
b1818
b1819
b1833
b1900
b1901
b1959
b1976
b2105
b2126
b2157
b2159
b2188
b2215
b2237
b2276
b2280
gene name
rihC
folA
pdxA
fruR
murG
lpxC
secM
yacH
speD
yadC
fadE
betB
cyoE
cyoD
cyoC
acrR
hemH
sucD
cmr
ybjK
ybjC
nfsA
ybjN
artM
artP
pqiA
yccU
agp
putA
ycdZ
mdtG
ymgA
ymgB
ymgC
ompW
pspA
pspB
pspC
pspD
pspE
feaR
ydcJ
ydcW
nhoA
ydeN
lsrC
lsrD
marR
marA
tus
fumC
ydhO
sufC
astE
astB
ydjI
manY
manZ
yebS
araG
araF
yedA
yeeI
yohL
yehU
yeiE
nfo
yejM
ompC
inaA
nuoN
nuoJ
mid-log glu
early stat glu
mid stat glu
late log gly
early stat gly
mid stat gly
fold-change P-value
fold-change P-value fold-change P-value fold-change P-value fold-change P-value fold-change P-value
1.04
0.746
1.92
0.009
1.65
0.077
2.13
0.036
-1.04
0.775
1.08
0.732
1.12
0.731
2.08
0.006
1.75
0.068
1.94
0.022
1.10
0.549
1.16
0.730
1.11
0.370
1.50
0.061
2.94
0.024
2.17
0.066
-1.06
0.600
-1.29
0.253
-1.09
0.241
1.19
0.501
2.68
0.000
2.46
0.009
1.01
0.934
-1.49
0.325
1.05
0.875
3.57
0.000
8.01
0.000
2.01
0.058
-1.02
0.905
-1.82
0.059
1.04
0.456
1.32
0.016
1.81
0.070
-1.54
0.203
1.23
0.201
4.56
0.003
-1.05
0.566
1.20
0.086
2.73
0.001
1.35
0.198
1.07
0.546
2.10
0.059
1.09
0.356
2.55
0.001
11.79
0.000
-1.13
0.450
1.04
0.647
3.03
0.010
1.05
0.491
2.56
0.000
5.53
0.000
2.97
0.095
1.46
0.369
-1.72
0.114
-1.06
0.596
1.95
0.005
1.54
0.010
-1.01
0.900
-1.06
0.677
2.39
0.031
1.07
0.569
-1.07
0.546
3.30
0.001
1.60
0.166
1.76
0.017
3.46
0.006
1.20
0.523
1.01
0.940
2.58
0.001
-1.34
0.197
1.90
0.013
3.63
0.001
1.09
0.241
1.47
0.010
3.65
0.003
1.81
0.096
1.65
0.271
1.98
0.042
1.03
0.796
1.24
0.095
2.47
0.007
1.78
0.057
1.66
0.168
1.95
0.046
1.07
0.256
-1.14
0.287
2.00
0.015
1.83
0.129
1.51
0.224
2.05
0.060
1.49
0.100
1.96
0.000
1.13
0.545
1.30
0.251
1.66
0.018
2.60
0.017
-1.10
0.146
1.48
0.009
4.27
0.001
1.03
0.883
-1.12
0.394
2.41
0.046
1.40
0.011
-1.28
0.201
3.44
0.024
3.84
0.071
1.09
0.488
-1.24
0.702
1.24
0.591
1.61
0.041
2.48
0.001
-1.02
0.974
-1.14
0.292
3.04
0.074
-1.03
0.503
2.06
0.006
2.01
0.225
1.16
0.460
1.12
0.450
2.25
0.026
1.36
0.004
1.62
0.103
2.65
0.000
1.12
0.701
1.73
0.003
5.32
0.002
1.18
0.055
2.06
0.020
4.29
0.000
1.02
0.904
1.21
0.415
2.32
0.007
1.06
0.264
1.85
0.002
1.25
0.174
2.06
0.079
1.32
0.198
-1.65
0.157
-1.22
0.311
3.06
0.003
15.19
0.000
1.87
0.089
-1.10
0.413
1.98
0.057
-1.02
0.759
1.95
0.001
3.94
0.000
1.10
0.473
-1.01
0.943
1.99
0.052
1.21
0.233
2.52
0.003
5.92
0.000
-1.04
0.733
-1.08
0.618
2.14
0.100
-1.11
0.206
1.66
0.007
2.19
0.026
-1.86
0.021
-1.11
0.392
1.95
0.096
-1.01
0.810
3.08
0.000
4.52
0.000
3.19
0.036
-1.05
0.626
2.24
0.055
1.36
0.153
-10.18
0.063
2.53
0.018
6.06
0.235
1.84
0.085
2.27
0.046
1.13
0.042
1.25
0.094
2.58
0.010
1.57
0.196
-1.01
0.883
1.88
0.032
1.00
0.955
1.83
0.013
2.20
0.007
-1.19
0.284
-1.01
0.961
2.09
0.033
1.06
0.510
3.56
0.001
-4.02
0.003
-4.67
0.174
1.46
0.604
25.95
0.033
1.09
0.202
5.60
0.001
-3.60
0.005
-7.53
0.165
1.19
0.745
20.66
0.031
-1.05
0.299
2.33
0.028
-5.49
0.000
-3.19
0.287
1.28
0.709
26.34
0.020
-1.48
0.595
2.34
0.027
-2.96
0.087
3.88
0.026
1.04
0.908
-2.03
0.127
-1.02
0.714
9.45
0.000
11.04
0.001
-2.02
0.187
10.47
0.000
12.63
0.000
1.03
0.635
5.18
0.001
7.32
0.001
-2.34
0.263
7.09
0.000
12.07
0.000
-1.05
0.704
4.11
0.002
5.07
0.000
-2.46
0.265
5.34
0.000
8.76
0.001
-1.03
0.818
4.99
0.002
9.60
0.002
-2.68
0.215
9.78
0.000
19.10
0.000
-1.15
0.323
2.74
0.000
11.40
0.008
2.11
0.228
8.35
0.005
19.47
0.001
1.00
1.000
1.86
0.002
2.09
0.006
2.93
0.037
-1.13
0.217
-1.14
0.433
-1.00
0.997
2.91
0.000
2.69
0.005
1.97
0.072
-1.49
0.041
-1.31
0.224
-1.02
0.632
1.18
0.215
2.07
0.110
2.29
0.064
1.19
0.466
-2.74
0.103
-1.03
0.646
5.04
0.000
12.69
0.000
-1.47
0.070
-1.26
0.068
4.26
0.059
1.07
0.385
4.28
0.000
2.34
0.002
3.41
0.045
-1.15
0.362
-1.42
0.634
1.03
0.669
2.40
0.002
-5.56
0.001
1.15
0.614
-1.09
0.707
3.68
0.024
1.03
0.769
2.35
0.001
-2.17
0.073
1.14
0.379
-1.14
0.225
3.65
0.010
2.50
0.037
2.72
0.004
5.20
0.000
1.96
0.031
2.79
0.013
5.96
0.007
1.73
0.018
5.21
0.000
6.25
0.000
2.24
0.191
2.91
0.001
6.10
0.010
1.03
0.907
1.55
0.001
1.83
0.006
-1.71
0.036
-1.01
0.916
2.16
0.009
1.11
0.212
4.39
0.000
8.51
0.000
1.35
0.257
1.04
0.873
2.15
0.013
1.02
0.626
2.40
0.000
-1.02
0.944
2.32
0.004
-1.02
0.906
-1.20
0.627
1.83
0.012
1.29
0.149
2.67
0.010
-1.60
0.223
-1.13
0.330
1.99
0.055
1.11
0.075
3.19
0.000
5.62
0.000
2.39
0.110
-1.00
0.996
-1.50
0.112
1.06
0.352
2.88
0.000
6.74
0.000
2.24
0.070
1.37
0.120
-1.45
0.115
1.08
0.188
1.49
0.103
2.16
0.002
1.97
0.032
1.16
0.494
1.13
0.815
1.09
0.108
-1.83
0.000
1.90
0.008
2.80
0.043
1.14
0.371
1.27
0.179
1.17
0.110
-1.87
0.005
1.81
0.014
3.06
0.041
1.12
0.412
1.58
0.051
1.14
0.109
4.07
0.007
11.91
0.000
-1.08
0.583
-1.02
0.916
2.52
0.045
1.04
0.642
1.12
0.319
8.35
0.012
4.12
0.087
-1.32
0.540
-1.00
0.989
1.33
0.069
1.26
0.167
6.18
0.009
3.63
0.047
-1.96
0.183
1.05
0.768
1.55
0.160
1.81
0.000
1.55
0.077
2.52
0.021
2.78
0.009
1.42
0.130
-1.10
0.275
1.99
0.001
4.22
0.000
3.64
0.063
1.11
0.646
-1.54
0.133
1.11
0.761
-1.24
0.121
2.02
0.043
-2.00
0.040
1.03
0.863
2.69
0.027
-1.03
0.651
1.49
0.010
2.74
0.000
2.27
0.020
1.10
0.314
-1.58
0.106
1.13
0.107
2.45
0.001
1.21
0.149
1.02
0.957
1.03
0.845
3.11
0.058
1.04
0.859
2.00
0.032
2.46
0.004
-1.23
0.292
1.47
0.228
1.89
0.044
1.25
0.552
1.46
0.034
3.97
0.001
1.81
0.070
1.00
0.999
-1.08
0.635
1.19
0.749
1.80
0.002
1.23
0.440
-1.10
0.657
2.48
0.019
-1.61
0.197
1.58
0.001
7.72
0.000
5.72
0.000
1.34
0.073
2.10
0.001
7.52
0.005
1.08
0.591
-1.21
0.218
3.06
0.103
2.42
0.015
-1.24
0.353
-1.75
0.378
1.02
0.726
-1.25
0.143
2.03
0.081
1.82
0.030
-1.22
0.273
1.06
0.899
245
Appendix IV (cont.)
Table of genes with fold-changes between BTE-expressing and BTE-H204A expressing cultures
greater than 1.8 (cont).
locus
b2281
b2282
b2327
b2341
b2342
b2377
b2410
b2584
b2666
b2684
b2703
b2705
b2710
b2802
b2809
b2869
b2875
b2877
b2889
b2996
b3009
b3014
b3038
b3050
b3074
b3081
b3126
b3127
b3161
b3194
b3207
b3224
b3225
b3238
b3255
b3256
b3261
b3337
b3383
b3404
b3528
b3565
b3566
b3567
b3568
b3608
b3665
b3749
b3784
b3806
b3845
b3892
b3894
b3925
b3936
b4033
b4034
b4035
b4036
b4037
b4039
b4040
b4045
b4050
b4062
b4139
b4154
b4175
b4242
b4346
b4359
b4380
b4381
b4382
mid-log glu
early stat glu
mid stat glu
late log gly
early stat gly
mid stat gly
gene name fold-change P-value fold-change P-value fold-change P-value fold-change P-value fold-change P-value fold-change P-value
nuoI
1.02
0.751
-1.06
0.584
2.91
0.008
1.92
0.054
-1.17
0.366
-1.14
0.777
nuoH
1.04
0.702
-1.02
0.865
3.27
0.008
1.83
0.071
1.01
0.968
-1.23
0.712
yfcA
1.11
0.631
1.03
0.744
1.81
0.004
2.32
0.093
-1.08
0.633
1.44
0.107
yfcX
1.15
0.048
1.28
0.160
2.15
0.009
2.93
0.032
1.32
0.078
1.67
0.072
yfcY
1.09
0.825
1.60
0.019
3.32
0.001
3.64
0.020
1.37
0.035
2.11
0.029
yfdY
-1.01
0.956
2.01
0.001
-1.15
0.585
-4.54
0.031
1.24
0.175
4.09
0.082
yfeH
-1.07
0.757
1.36
0.083
2.53
0.038
1.20
0.338
-1.03
0.769
1.81
0.053
yfiQ
-1.08
0.634
2.01
0.003
3.65
0.001
2.17
0.016
-1.03
0.806
-1.16
0.388
yqaE
-1.10
0.733
1.87
0.011
1.15
0.674
-5.19
0.054
1.04
0.778
2.19
0.094
mprA
1.37
0.023
2.10
0.000
6.11
0.000
2.12
0.082
3.44
0.000
10.39
0.001
srlE
1.16
0.396
-1.10
0.601
3.81
0.000
7.86
0.005
-1.20
0.129
-1.12
0.711
srlD
-1.01
0.920
1.28
0.089
2.55
0.007
1.99
0.063
-1.09
0.348
1.10
0.749
norV
-1.03
0.753
1.19
0.268
2.94
0.153
-1.31
0.153
1.43
0.015
4.87
0.006
fucI
1.06
0.555
2.76
0.000
3.20
0.022
1.83
0.081
-1.00
0.984
-1.48
0.337
ygdI
1.02
0.753
2.81
0.002
3.25
0.003
-1.45
0.335
1.36
0.025
2.81
0.011
ygeV
-1.14
0.359
2.39
0.000
2.90
0.001
5.52
0.039
1.01
0.952
-1.00
0.989
yqeB
-1.04
0.826
1.09
0.463
4.44
0.005
4.49
0.046
1.03
0.836
-1.12
0.600
ygfJ
-1.37
0.153
4.15
0.000
9.53
0.000
3.19
0.023
-1.32
0.147
1.65
0.225
idi
1.08
0.354
3.29
0.001
2.96
0.008
-1.54
0.275
1.02
0.873
2.64
0.099
hybA
-1.31
0.105
-1.16
0.280
2.84
0.001
2.25
0.015
-1.26
0.073
-1.07
0.689
yghB
-1.24
0.058
1.85
0.016
2.50
0.003
1.43
0.052
-1.31
0.243
2.94
0.065
yqhH
1.00
0.968
1.06
0.649
2.18
0.019
-1.17
0.306
-1.10
0.390
1.83
0.062
ygiC
-1.03
0.923
1.61
0.002
5.53
0.000
2.14
0.104
1.34
0.110
-1.50
0.248
yqiJ
1.53
0.316
12.75
0.000
9.46
0.000
1.49
0.112
1.55
0.048
7.15
0.000
ygjH
1.03
0.627
1.10
0.508
2.06
0.028
-2.96
0.287
-1.14
0.604
2.50
0.091
fadH
1.02
0.909
1.93
0.003
3.07
0.006
2.09
0.041
1.17
0.155
1.87
0.027
garL
1.09
0.218
3.03
0.005
-1.21
0.225
3.20
0.103
-1.10
0.490
1.93
0.125
garP
1.01
0.923
2.44
0.009
1.35
0.354
6.46
0.007
-1.33
0.047
2.92
0.143
mtr
1.12
0.649
12.16
0.000
21.56
0.001
-2.39
0.210
1.50
0.167
6.02
0.010
yrbE
1.14
0.303
1.79
0.003
2.06
0.006
1.20
0.294
1.19
0.307
2.27
0.011
yrbL
1.26
0.352
7.67
0.000
1.41
0.022
-1.00
0.993
13.58
0.000
13.59
0.013
nanT
1.07
0.658
1.52
0.052
4.27
0.002
3.90
0.033
1.08
0.617
3.22
0.029
nanA
1.07
0.439
1.83
0.007
2.33
0.024
5.34
0.140
1.22
0.344
2.93
0.021
yhcN
1.93
0.130
4.69
0.000
4.90
0.000
-3.68
0.015
1.67
0.096
3.05
0.068
accB
1.23
0.026
2.17
0.048
3.82
0.002
7.55
0.060
1.58
0.024
-1.42
0.662
accC
1.27
0.278
2.09
0.002
2.69
0.001
7.78
0.063
1.26
0.119
-1.25
0.657
fis
-1.04
0.852
1.84
0.002
-1.95
0.006
8.05
0.082
1.27
0.356
-2.41
0.114
bfd
1.14
0.111
-1.35
0.023
2.00
0.030
-2.25
0.299
1.35
0.324
5.39
0.083
yhfZ
-1.05
0.381
1.79
0.027
1.86
0.074
2.42
0.005
-1.19
0.188
-1.23
0.682
envZ
1.18
0.577
2.56
0.000
6.47
0.000
1.50
0.171
1.02
0.878
-1.24
0.285
dctA
1.59
0.016
-1.67
0.053
2.27
0.006
8.55
0.031
1.32
0.064
-1.22
0.336
xylA
1.00
0.971
3.15
0.000
1.30
0.302
21.85
0.041
1.24
0.287
-2.07
0.015
xylF
1.07
0.387
2.11
0.002
2.27
0.016
36.39
0.020
-1.07
0.590
-3.22
0.004
xylG
1.03
0.628
4.57
0.000
1.62
0.042
13.69
0.006
-1.15
0.215
-5.02
0.002
xylH
1.09
0.270
5.48
0.000
2.24
0.015
8.01
0.009
-1.05
0.750
-3.68
0.003
gpsA
1.00
0.936
1.53
0.009
2.57
0.001
1.87
0.068
-1.15
0.251
1.18
0.320
ade
1.00
0.984
1.58
0.054
2.53
0.002
1.98
0.034
-1.09
0.377
1.09
0.728
rbsA
-1.18
0.498
5.16
0.000
16.73
0.000
3.40
0.048
-1.01
0.929
-1.25
0.268
rfe
-1.05
0.650
1.31
0.109
2.28
0.002
2.74
0.087
1.16
0.418
1.27
0.217
cyaA
-1.10
0.080
1.49
0.002
2.82
0.002
-1.90
0.112
-1.05
0.745
2.32
0.058
fadA
1.09
0.545
1.68
0.001
3.96
0.002
3.14
0.053
1.29
0.040
1.32
0.250
fdoI
-1.02
0.869
1.55
0.042
2.91
0.000
1.05
0.903
1.14
0.489
2.33
0.070
fdoG
-1.21
0.022
2.48
0.017
12.49
0.000
-1.39
0.300
1.13
0.438
2.32
0.072
glpX
-1.04
0.834
3.03
0.000
9.71
0.002
4.20
0.030
-1.08
0.406
-1.40
0.477
rpmE
-1.16
0.069
8.62
0.000
3.93
0.001
3.22
0.052
1.22
0.649
2.79
0.304
malF
1.89
0.001
-2.15
0.057
1.32
0.192
3.31
0.050
-1.21
0.101
-38.29
0.001
malE
2.25
0.000
-2.83
0.014
-1.12
0.436
7.58
0.102
-1.30
0.312
-31.57
0.001
malK
2.62
0.000
-2.99
0.002
1.11
0.385
6.74
0.029
-1.12
0.473
-35.25
0.002
lamB
3.75
0.000
-1.77
0.038
-1.16
0.236
10.78
0.033
-1.71
0.264
-20.72
0.007
malM
3.06
0.000
-1.56
0.022
-1.56
0.022
5.85
0.030
-1.32
0.304
-14.74
0.005
ubiC
-1.14
0.311
1.95
0.007
2.60
0.005
-1.04
0.890
1.26
0.313
2.80
0.018
ubiA
-1.02
0.811
1.75
0.001
2.07
0.014
-1.29
0.202
1.21
0.192
4.51
0.001
yjbJ
-1.12
0.376
1.90
0.006
-2.20
0.001
-7.40
0.194
-1.28
0.199
4.31
0.072
yjbO
1.05
0.779
13.85
0.000
13.23
0.011
-3.22
0.301
13.59
0.001
38.14
0.003
soxS
1.09
0.301
2.56
0.008
1.22
0.748
-5.94
0.118
-1.46
0.396
4.52
0.057
aspA
-1.38
0.680
2.43
0.001
19.48
0.000
9.53
0.115
-1.45
0.243
2.00
0.028
frdA
-1.67
0.564
1.16
0.211
2.22
0.001
1.83
0.094
-1.09
0.586
1.36
0.243
hflC
1.03
0.642
-1.10
0.315
2.52
0.017
1.00
0.999
-1.16
0.160
1.89
0.058
mgtA
1.94
0.006
-1.02
0.878
-1.21
0.243
3.66
0.084
1.32
0.050
-1.12
0.547
mcrB
-1.29
0.057
1.83
0.008
1.37
0.046
2.96
0.095
1.08
0.433
1.10
0.720
mdoB
1.24
0.508
2.53
0.001
2.37
0.081
1.80
0.103
1.08
0.616
1.55
0.121
yjjI
-1.71
0.579
4.80
0.000
18.14
0.000
-1.66
0.120
-1.31
0.114
5.44
0.051
deoC
-1.16
0.022
1.27
0.022
6.77
0.000
-3.12
0.034
1.01
0.944
4.89
0.015
deoA
-1.04
0.572
-1.17
0.223
7.80
0.000
-2.87
0.068
-1.15
0.572
4.27
0.004
246
Appendix IV (cont.)
Table of genes with fold-changes between BTE-expressing and BTE-H204A expressing cultures
greater than 1.8 (cont).
locus
b4460
b4485
b4485
b4554
gene name
araH
ytfR
ytfR
yibT
mid-log glu
early stat glu
mid stat glu
late log gly
early stat gly
mid stat gly
fold-change P-value
fold-change P-value fold-change P-value fold-change P-value fold-change P-value fold-change P-value
1.12
0.464
-1.05
0.755
4.07
0.032
2.24
0.047
-1.29
0.473
1.07
0.674
-1.03
0.627
2.24
0.001
7.19
0.001
3.64
0.084
-1.85
0.120
1.03
0.912
-1.03
0.627
2.24
0.001
7.19
0.001
3.64
0.084
-1.85
0.120
1.03
0.912
1.14
0.466
6.63
0.000
5.95
0.000
-4.25
0.003
2.18
0.003
5.15
0.050
247
Appendix IV (cont.)
Table of genes with fold-changes between BTE-expressing and BTE-H204A expressing cultures
less than 1.8 and with P-values less than 0.05 (EZ glucose shake flask cultures) or less than 0.10
(EZ glycerol fermentor cultures) in both shake flask and fermentor cultures for at least one
sampling time. Sampling times are defined in Chapter 3. Fold-changes are reported as linear
values.
locus
b0001
b0197
b0241
b0287
b0325
b0368
b0383
b0399
b0440
b0553
b0624
b0710
b0800
b0814
b0828
b0838
b0912
b0929
b0954
b0978
b0979
b0980
b1037
b1038
b1039
b1040
b1059
b1087
b1111
b1130
b1178
b1180
b1446
b1525
b1539
b1541
b1606
b1607
b1619
b1637
b1740
b1787
b1811
b1841
b1892
b1908
b1919
b1920
b1929
b1974
b2007
b2235
b2252
b2253
b2254
b2255
b2294
b2323
b2414
b2422
b2424
b2425
b2576
b2622
b2670
b2673
b2751
b2752
b2762
b2763
b2764
b2804
mid-log glu
early stat glu
mid stat glu
late log gly
early stat gly
mid stat gly
gene name fold-change P-value fold-change P-value fold-change P-value fold-change P-value fold-change P-value fold-change P-value
thrL
1.04
0.776
-1.30
0.206
-1.99
0.064
-1.98
0.065
-1.15
0.641
-1.40
0.460
metQ
-1.16
0.459
-1.66
0.015
-2.85
0.054
-1.12
0.415
1.11
0.526
-2.49
0.072
phoE
-1.10
0.084
-4.88
0.000
-1.04
0.762
-25.17
0.048
2.38
0.126
-10.18
0.236
yagU
1.04
0.518
-6.52
0.001
-6.38
0.001
-2.06
0.053
-1.18
0.384
-1.39
0.598
yahK
1.02
0.836
-1.24
0.188
-1.81
0.057
-2.71
0.183
-1.20
0.143
-2.01
0.030
tauD
-1.02
0.821
-1.18
0.204
-1.94
0.004
-3.54
0.082
-1.27
0.183
-15.51
0.251
phoA
1.20
0.183
-49.27
0.000
-1.05
0.859
-14.51
0.130
1.17
0.641
-20.43
0.077
phoB
1.06
0.801
-7.75
0.000
-1.48
0.048
-9.76
0.159
1.05
0.657
-3.24
0.059
hupB
-1.06
0.401
1.20
0.078
-4.22
0.028
2.13
0.158
1.62
0.005
-2.70
0.075
nmpC
-1.92
0.000
-5.07
0.000
-27.64
0.001
3.98
0.234
-1.27
0.674
-6.31
0.066
crcB
1.05
0.672
1.41
0.022
-1.99
0.008
-1.94
0.009
-1.18
0.347
2.03
0.065
ybgI
-1.02
0.805
-1.65
0.001
-1.89
0.007
-1.92
0.020
1.02
0.879
-1.32
0.161
ybiB
-1.10
0.167
-1.28
0.101
-2.05
0.004
-3.04
0.072
1.04
0.781
1.17
0.399
ompX
-1.08
0.290
1.08
0.665
-2.88
0.002
-7.21
0.018
1.10
0.475
1.94
0.031
iaaA
-1.12
0.568
-2.13
0.002
1.08
0.573
-2.53
0.053
-1.46
0.164
-2.14
0.131
yliJ
1.03
0.524
-1.41
0.102
-2.99
0.004
-2.50
0.003
1.27
0.059
-1.86
0.287
ihfB
-1.04
0.540
-1.57
0.003
-3.36
0.000
-2.02
0.084
1.02
0.898
1.10
0.585
ompF
-1.24
0.064
-7.72
0.000
-10.43
0.002
2.75
0.299
1.66
0.108
-6.93
0.037
fabA
-1.28
0.087
-2.54
0.004
-4.80
0.000
-2.94
0.019
-3.58
0.004
-1.24
0.271
appC
-1.07
0.515
-3.07
0.001
-1.59
0.175
-1.52
0.097
-1.74
0.014
-2.95
0.092
appB
1.13
0.514
-2.26
0.000
1.12
0.427
-1.17
0.466
-1.82
0.016
-2.76
0.087
appA
1.07
0.584
-2.06
0.001
-1.28
0.260
-1.67
0.074
-1.11
0.453
-1.98
0.017
csgG
1.19
0.048
-3.17
0.003
-2.05
0.001
-2.36
0.212
1.42
0.110
-3.71
0.025
csgF
-1.01
0.891
-6.82
0.000
-2.34
0.009
-2.58
0.259
1.54
0.076
-10.26
0.012
csgE
1.11
0.330
-5.93
0.000
-2.28
0.041
-2.63
0.387
1.41
0.259
-8.10
0.018
csgD
-1.03
0.646
-4.76
0.000
-5.73
0.005
-3.72
0.247
1.29
0.202
-7.34
0.014
solA
1.09
0.554
-2.42
0.003
-1.93
0.088
-1.11
0.653
1.14
0.497
-2.22
0.100
yceF
-1.01
0.894
-1.09
0.523
-2.38
0.003
-2.20
0.091
1.25
0.111
-1.08
0.718
ycfQ
-1.26
0.486
-1.29
0.034
-1.89
0.036
-2.16
0.087
-1.19
0.118
-1.20
0.522
phoP
-1.02
0.864
-1.65
0.002
-2.86
0.001
-1.98
0.075
-1.02
0.826
1.14
0.418
ycgK
1.36
0.406
-1.22
0.169
-2.29
0.003
-2.26
0.017
1.24
0.101
-2.02
0.049
ycgM
-1.16
0.513
-1.67
0.002
-2.65
0.003
-2.83
0.091
-1.00
0.986
1.29
0.367
ydcY
-1.11
0.511
-1.50
0.101
-2.37
0.001
-2.29
0.065
1.53
0.135
1.42
0.294
yneI
1.03
0.773
-1.23
0.073
-2.76
0.000
-2.40
0.021
1.27
0.250
-1.48
0.480
ydfG
-1.11
0.050
-1.55
0.223
-2.25
0.002
1.01
0.955
1.03
0.887
-2.04
0.075
ydfZ
-1.53
0.428
-2.61
0.000
-1.91
0.002
-1.87
0.013
1.21
0.345
-1.53
0.098
folM
-1.24
0.272
-2.97
0.000
-3.01
0.003
-1.95
0.018
-1.38
0.040
-1.16
0.445
ydgC
1.16
0.454
-2.77
0.000
-2.89
0.000
-2.58
0.072
-1.04
0.918
-1.10
0.571
hdhA
-1.24
0.371
-1.73
0.001
-2.04
0.027
-2.65
0.039
-1.30
0.147
1.34
0.400
tyrS
1.07
0.190
-2.35
0.001
-3.36
0.000
1.30
0.329
1.14
0.226
-2.58
0.057
nadE
-1.07
0.218
-1.23
0.192
-1.87
0.013
-1.98
0.034
1.16
0.387
-1.75
0.210
yeaK
-1.07
0.262
-3.73
0.000
-2.45
0.002
-1.83
0.078
1.22
0.174
-1.79
0.323
yoaH
-1.04
0.669
-1.72
0.001
-2.59
0.004
-1.90
0.073
1.04
0.884
1.66
0.160
yobA
1.00
0.988
-1.68
0.005
-2.80
0.003
-1.95
0.009
-1.20
0.466
1.09
0.674
flhD
1.32
0.017
-3.25
0.001
-4.43
0.001
4.35
0.183
1.24
0.568
-3.59
0.093
yecA
-1.13
0.628
-1.51
0.021
-2.54
0.001
-2.30
0.053
1.00
0.999
1.10
0.761
yedO
1.02
0.825
-1.30
0.020
-2.06
0.008
-3.18
0.008
-1.64
0.086
-1.22
0.343
fliY
-1.01
0.948
-2.05
0.001
-3.45
0.001
-3.45
0.029
-1.77
0.079
-1.82
0.254
yedE
1.00
0.998
-2.98
0.001
-4.96
0.000
6.71
0.031
1.24
0.147
-2.28
0.343
yodB
1.67
0.056
-2.57
0.000
-1.20
0.499
-1.30
0.380
-2.40
0.016
1.04
0.793
yeeX
-1.14
0.473
-1.48
0.014
-3.45
0.000
-2.13
0.078
1.40
0.051
-1.35
0.224
nrdB
-1.19
0.454
-1.86
0.035
-4.20
0.002
-1.30
0.450
-1.09
0.461
-1.85
0.029
ais
1.01
0.895
-2.90
0.000
-1.19
0.237
-1.69
0.365
-5.32
0.000
1.43
0.131
yfbE
1.08
0.392
-3.96
0.000
-1.44
0.096
-3.06
0.424
-7.61
0.000
1.07
0.853
yfbF
1.06
0.703
-2.17
0.003
3.64
0.002
-2.40
0.461
-5.41
0.006
1.02
0.973
yfbG
-1.05
0.359
-2.10
0.001
1.35
0.161
-1.44
0.388
-2.21
0.021
1.18
0.380
yfbU
-1.16
0.361
-1.60
0.001
-1.87
0.002
-1.89
0.031
1.11
0.625
-1.21
0.404
fabB
-1.87
0.000
-12.95
0.001
-1.57
0.334
-10.28
0.001
-27.33
0.000
-2.35
0.099
cysK
-1.15
0.371
-7.01
0.000
-1.84
0.081
-1.97
0.073
-1.29
0.211
-4.38
0.090
cysA
-1.00
0.976
-4.17
0.001
1.45
0.044
-5.90
0.058
-2.20
0.101
-6.17
0.134
cysU
1.04
0.798
-2.20
0.001
1.82
0.007
-6.37
0.062
-3.07
0.159
-8.65
0.081
cysP
-1.01
0.903
-3.45
0.001
1.02
0.891
-8.33
0.063
-2.50
0.200
-6.66
0.133
srmB
1.15
0.598
-2.16
0.013
-2.71
0.001
1.12
0.601
1.24
0.167
-2.02
0.101
intA
-1.26
0.456
1.00
0.971
-2.25
0.001
-1.89
0.012
1.01
0.964
-1.39
0.143
ygaW
1.01
0.804
-3.56
0.113
-3.85
0.010
-3.49
0.011
1.38
0.023
1.22
0.229
nrdH
1.33
0.233
-1.10
0.413
-3.38
0.003
-2.88
0.039
1.02
0.866
1.64
0.469
cysN
-1.08
0.734
-8.92
0.000
-1.58
0.023
-5.77
0.054
-2.38
0.079
-10.58
0.132
cysD
-1.21
0.192
-5.30
0.001
-1.29
0.178
-5.90
0.076
-2.47
0.124
-10.01
0.102
cysH
-1.17
0.574
-3.75
0.004
-1.43
0.069
-4.09
0.171
-2.60
0.151
-10.88
0.098
cysI
-1.12
0.673
-8.61
0.000
-1.21
0.325
-4.42
0.139
-2.99
0.091
-17.14
0.071
cysJ
1.03
0.615
-5.08
0.000
1.12
0.510
-4.13
0.237
-2.97
0.130
-14.87
0.101
fucU
-1.08
0.535
-1.32
0.045
-2.76
0.001
-2.50
0.013
1.25
0.173
-1.52
0.162
248
Appendix IV (cont.)
Table of genes with fold-changes between BTE-expressing and BTE-H204A expressing cultures
less than 1.8 (cont.)
locus
b2899
b2913
b2924
b2988
b3028
b3041
b3103
b3185
b3234
b3291
b3321
b3356
b3417
b3461
b3470
b3615
b3688
b3713
b3724
b3725
b3726
b3728
b3766
b3774
b3842
b3850
b3858
b4032
b4033
b4034
b4035
b4077
b4114
b4219
b4238
b4294
b4340
b4505
b4543
b4592
mid-log glu
early stat glu
mid stat glu
late log gly
early stat gly
mid stat gly
gene name fold-change P-value fold-change P-value fold-change P-value fold-change P-value fold-change P-value fold-change P-value
yqfA
-2.12
0.136
-8.77
0.006
-1.89
0.111
-14.66
0.020
-32.77
0.000
-1.15
0.487
serA
-1.18
0.353
-4.81
0.000
-3.67
0.004
-1.92
0.034
-1.05
0.746
-2.20
0.204
mscS
-1.03
0.688
-7.20
0.000
-11.76
0.000
-8.02
0.086
-1.04
0.847
-1.29
0.471
gss
-1.17
0.576
-2.47
0.001
-1.11
0.563
-2.34
0.022
-1.10
0.674
-1.59
0.103
mdaB
1.07
0.166
-1.80
0.002
-2.08
0.030
-1.25
0.276
1.18
0.230
-4.73
0.089
ribB
-1.20
0.225
-1.85
0.091
-3.59
0.030
-1.53
0.182
-1.01
0.976
-3.94
0.014
yhaH
-1.12
0.489
-1.05
0.634
-2.35
0.039
-2.87
0.064
1.14
0.230
1.71
0.312
rpmA
-1.12
0.466
-1.19
0.101
-14.07
0.000
3.22
0.084
1.52
0.148
-2.92
0.028
degQ
-1.09
0.465
-1.40
0.072
-3.01
0.003
-1.92
0.008
1.17
0.241
-1.55
0.086
mscL
-1.14
0.577
1.28
0.039
-2.05
0.003
-3.20
0.083
1.17
0.145
1.10
0.744
rpsJ
-1.09
0.494
1.35
0.165
-5.20
0.001
4.79
0.189
1.76
0.389
-1.89
0.028
yhfA
-1.23
0.488
-2.68
0.002
-3.53
0.001
-1.83
0.058
1.04
0.678
1.18
0.332
malP
1.70
0.001
-1.50
0.006
-2.52
0.015
1.44
0.056
-1.06
0.765
-7.82
0.009
rpoH
-1.13
0.209
-1.32
0.025
-3.38
0.001
-2.33
0.053
1.04
0.664
-1.49
0.270
yhhP
-1.06
0.795
2.53
0.001
-2.03
0.090
-1.82
0.246
-1.68
0.265
-1.92
0.052
yibD
-1.04
0.453
-8.53
0.000
1.09
0.670
-8.68
0.315
-1.10
0.854
-40.76
0.077
yidQ
-1.21
0.090
-1.26
0.426
-2.44
0.002
1.62
0.315
-1.01
0.898
-1.87
0.082
yieF
-1.07
0.520
1.01
0.960
-2.55
0.004
-1.31
0.370
1.19
0.499
-2.26
0.007
phoU
-1.01
0.805
-6.60
0.000
-5.13
0.000
-7.18
0.092
-1.03
0.799
-4.62
0.056
pstB
1.02
0.794
-7.74
0.000
-4.78
0.000
-4.90
0.129
-1.04
0.704
-4.50
0.071
pstA
1.20
0.673
-18.25
0.000
-2.64
0.001
-5.44
0.219
-1.18
0.116
-4.74
0.047
pstS
1.33
0.330
-40.50
0.000
-6.05
0.000
-7.11
0.160
-1.09
0.386
-8.03
0.062
ilvL
-1.19
0.105
-1.95
0.015
-4.28
0.002
-6.16
0.055
-1.04
0.896
-1.07
0.816
ilvC
-1.11
0.316
-5.01
0.003
-1.77
0.054
-3.35
0.029
-1.22
0.092
-6.26
0.147
rfaH
-1.17
0.700
-2.41
0.001
-4.65
0.000
-1.84
0.047
-1.14
0.429
1.10
0.834
hemG
-1.16
0.186
-1.83
0.000
-1.48
0.329
1.17
0.509
-1.09
0.575
-2.57
0.029
yihD
-1.31
0.171
-1.83
0.005
-3.03
0.007
-2.35
0.088
1.32
0.035
-1.75
0.354
malG
1.39
0.005
-3.16
0.023
-1.09
0.586
1.09
0.796
-1.02
0.816
-15.09
0.009
malF
1.89
0.001
-2.15
0.057
1.32
0.192
3.31
0.050
-1.21
0.101
-38.29
0.001
malE
2.25
0.000
-2.83
0.014
-1.12
0.436
7.58
0.102
-1.30
0.312
-31.57
0.001
malK
2.62
0.000
-2.99
0.002
1.11
0.385
6.74
0.029
-1.12
0.473
-35.25
0.002
gltP
1.01
0.934
-5.40
0.001
-3.26
0.001
-3.43
0.049
1.15
0.665
1.06
0.710
eptA
1.05
0.805
-2.39
0.000
1.09
0.658
-2.08
0.432
-4.76
0.000
1.20
0.448
msrA
1.13
0.274
1.03
0.714
-1.98
0.005
-2.27
0.045
1.01
0.948
-1.03
0.879
nrdD
-1.76
0.320
1.58
0.008
-3.65
0.000
-3.78
0.060
-1.70
0.045
-1.05
0.919
insA
-1.00
0.995
1.29
0.068
-2.23
0.008
-2.50
0.052
-1.20
0.619
-1.08
0.856
yjiR
1.03
0.493
-1.99
0.003
-4.53
0.000
-1.83
0.053
1.21
0.308
1.27
0.175
ykgN
-1.18
0.089
-1.29
0.114
-1.96
0.030
-1.85
0.087
-1.36
0.375
-1.36
0.245
ypaA
1.10
0.177
-1.23
0.048
-2.31
0.007
-1.34
0.164
1.13
0.426
-1.87
0.052
yccB
1.14
0.737
-2.13
0.000
-1.08
0.503
-2.01
0.082
-1.44
0.067
-1.68
0.055