FEMS Microbiology Letters, 364, 2017, fnx055
doi: 10.1093/femsle/fnx055
Advance Access Publication Date: 8 March 2017
Research Letter
R E S E A R C H L E T T E R – Physiology & Biochemistry
Quantifying ribosome dynamics in Escherichia coli
using fluorescence
Jurek Failmezger, Julian Ludwig, Alexander Nieß
and Martin Siemann-Herzberg∗
Institute of Biochemical Engineering, University of Stuttgart, 70569 Stuttgart, Germany
∗
Corresponding author: Institute of Biochemical Engineering, Allmandring 31, 70569 Stuttgart, Germany. Tel: +49711 68565161; E-mail:
[email protected]
One sentence summary: This study describes the evaluation of a fluorescent method to assess ribosome metabolism in Escherichia coli.
Editor: Richard Calendar
ABSTRACT
Ribosomes are a crucial component of the physiological state of a cell. Therefore, we aimed to monitor ribosome dynamics
using a fast and easy fluorescence readout. Using fluorescent-labeled ribosomal proteins, the dynamics of ribosomes during
batch cultivation and during nutritional shift conditions was investigated. The fluorescence readout was compared to the
cellular rRNA content determined by capillary gel electrophoresis with laser-induced fluorescence detection during
exponentially accelerating and decelerating growth. We found a linear correlation between the observed fluorescence and
the extracted rRNA content throughout cultivation, demonstrating the applicability of this method. Moreover, the results
show that ribosome dynamics, as a result of slowing growth, are accompanied by the passive effect of dilution of
preexisting ribosomes, de novo ribosome synthesis and ribosome degradation. In light of the challenging task of deciphering
ribosome regulatory mechanisms, our approach of using fluorescence to follow ribosome dynamics will allow more
comprehensive studies of biological systems.
Keywords: fluorescence; ribosome dynamics; rRNA; growth rate; ribosomal protein; capillary gel electrophoresis; ribosome
degradation
INTRODUCTION
Reliable analysis of ribosomes is indispensable to follow rRNA
degradation processes, compare rRNA turnover rates and investigate translation mechanisms and ribosomal function kinetics both in vivo and in vitro. Ribosome degradation in Escherichia
coli is a common reaction to some environmental conditions
(Deutscher 2003; Housley and Tollervey 2009) such as lack of carbon, nitrogen or phosphate. However, the principal mechanisms
of rRNA degradation in vivo and in vitro are still unclear (Housley
and Tollervey 2009). To date, the typical method of rRNA analysis is agarose or polyacrylamide gel electrophoresis where rRNA
molecules are evaluated by comparing the relative intensities
of electrophoretic bands (Zundel, Basturea and Deutscher 2009;
Basturea, Zundel and Deutscher 2011). More recent approaches
are based on capillary gel electrophoresis with laser-induced fluorescence detection (CGE-LIF), which allows the measurement
of 16S and 23S rRNA. We have previously shown that CGE-LIF
using MS2 RNA as an internal standard is a powerful method to
analyze rRNA in vivo (Hardiman et al. 2008) and to follow rRNA
degradation processes in vitro (Failmezger et al. 2016). However,
these methods are hampered by laborious RNA extraction procedures, many of which include potentially hazardous chemicals (Boedtker 1971; Reijnders, Sloof and Borst 1973). A simple
solution to this problem would be to monitor ribosome dynamics by fluorescence readout measurements, avoiding the purification of rRNA. The pool of ribosomal proteins offers a wide
selection of possible targets for fluorescent protein tagging. It
Received: 16 December 2016; Accepted: 7 March 2017
C FEMS 2017. All rights reserved. For permissions, please e-mail: [email protected]
1
2
FEMS Microbiology Letters, 2017, Vol. 364, No. 6
was previously demonstrated by Chai et al. (2014) that labeling of
ribosomal protein L9 and S6 with fluorescence proteins allowed
the examination of the intracellular ribosome distribution during cell division. In addition, Bakshi et al. (2014) monitored
diffusion of ribosomes in single cells during growth and drug
treatment. Nikolay et al. (2014) have shown that the fusion of
fluorescent proteins (enhanced green fluorescent protein [EGFP]
and mCherry) with ribosomal proteins L19 and S2 allows the
monitoring of ribosome assembly defects. Moreover, this modified E. coli strain MCrg shows a wild-type-like growth behavior, indicating that the translation apparatus is not affected
by the modified reporter proteins. While the mentioned reports demonstrate the tracking of ribosomes using fluorescence,
no growth rate-associated ribosome dynamic studies were
performed.
Here we show that the dynamic behavior of ribosomes during batch cultivation can be observed by monitoring the fluorescence of E. coli strain MCrg. Moreover, we demonstrate that the
fluorescence of tagged ribosomal proteins and the concentration
of 16S and 23S rRNA correlate during exponential and stationary
growth phases, thus leading to a quantitative understanding of
ribosome-related functions.
teins and DNA were precipitated by adding 250 μL of ice-cold
saturated NaCl solution. After incubation on ice, the precipitate
was collected by centrifugation (20 000 × g, 20 min, 4◦ C). A volume of 500 μL of the supernatant was transferred to fresh tubes.
rRNA precipitation was initiated by adding 1 mL ethanol, and the
samples were incubated at –70◦ C for 1 h to ensure complete precipitation. The rRNA was collected by centrifugation at 20 000 ×
g for 30 min at 4◦ C and washed with 1 mL ice-cold ethanol (70%).
The rRNA pellets were dried in a vacuum centrifuge and stored
at –70◦ C. Measurement of 16S and 23S rRNA was performed using an Agilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto,
USA). rRNA pellets were resuspended in RNase-free water and
heated to 65◦ C for 2 min. Analysis of rRNA was performed as
described by the manufacturer, and rRNA (sum of 16S and 23S)
was quantified by taking into account the known concentration
of the internal standard (see Supplementary Material, Fig. S1).
Estimation of ribosome dilution
Estimation of ribosome dilution as a function of the growth rate
was described by the following equation assuming a linear decrease of the growth rate,
MATERIAL AND METHODS
dcrRNA
= μ × crRNA = (a × t + b) crRNA
dt
Cell growth and preparation of cell lysates
Escherichia coli strain MCrg, a kind gift from R. Nikolay, was cultivated in 2 × TY media at 37◦ C in replicate shaking flasks. Cell
density was determined at 600 nm using a photometer. Samples
were taken throughout cultivation and the biomass was rapidly
chilled by incubation in an ice bath. Cells were harvested by centrifugation at 5000 × g for 20 min at 4◦ C, and the biomass was
flash frozen in liquid nitrogen and stored at –70◦ C until cell lysis. For cell lysis, the biomass was resuspended in lysis buffer
(10 mM magnesium acetate, 60 mM potassium acetate, 10 mM
Tris, pH 8.0) at identical biomass concentrations in each sample. Cells were lysed by sonication for 30 s while keeping the cell
suspension on ice. Cell debris was removed by centrifugation at
20 000 × g for 30 min at 4◦ C. For nutritional shift experiments, E.
coli MCrg was cultivated on mineral medium with the following
composition according to Wilms et al. (2001): 2.0 g L−1 Na2 SO4 ,
2.68 g L−1 NH4 SO4 , 0.5 g L−1 NH4 Cl, 14.6 g L−1 K2 HPO4 , 4.0 g L−1
NaH2 PO4 × 2H2 0, 0.5 g L−1 MgSO4 × 7 H2 O, 0.01 g L−1 thiamine,
3 mL L−1 trace element solution. Carbon source was acetate
(64 mM) and glucose (1 mM) or glucose (11 mM) and sorbitol (11
mM), respectively.
Fluorescence measurements
Fluorescence of L19-GFP and S2-mCherry was measured in a
Synergy 2 plate reader (Biotek Instruments, USA) using 250 μL
of cell lysates. EGFP was determined with excitation at 485 nm
and emission at 528 nm (gain set to 40), and mCherry with excitation at 528 nm and emission at 590 nm (gain set to 50).
rRNA extraction and analysis
rRNA was extracted from cell lysates as previously described by
Reddy (1987) and Hardiman et al. (2008) with some modifications:
20 μL lysate was added to tubes containing 5 μL diethyl phosphorocyanidate. Twenty microliters of internal standard (MS2
RNA, Roche Diagnostics, Mannheim, Germany) was added, followed by the addition of 500 μL of buffer (10 mM Tris-HCl [pH
8.0], 10 mM NaCl, 1 mM sodium citrate and 1.5% [w/v] SDS). Pro-
(1)
Here, crRNA (t) denotes the cellular rRNA content and μ the
specific growth rate. Parameters a and b were estimated by linear
regression of the growth rate between 2 and 7 h (R2 = 0.97), with
a = –0.25 1/h2 and b = 1.7 1/h. Regression analysis and function
R Excel.
evaluation were performed in Microsoft
RESULTS
Monitoring ribosomes by rRNA content during batch
cultivation
The intracellular ribosome concentration in Escherichia coli is
known to be highly dynamic, reflecting and being dependent on
the physiological state of the cell (for a review, see Deutscher
2003). This is emphasized by the fact that ribosome content
and growth rate are tightly coupled (Bremer and Dennis 1996).
Ribosomes are degraded under stress conditions such as nutrient deprivation (Basturea, Zundel and Deutscher 2011), and
therefore we decided to analyze ribosomes during bacterial
growth in batch culture. This cultivation strategy exhibits highly
dynamic growth patterns such as exponential acceleration followed by deceleration due to starvation. These rapid changes in
growth should be mirrored in the intracellular ribosome content.
Thus, we took samples throughout batch cultivation (Fig. 1a), extracted the RNA and analyzed the rRNA content by CGE-LIF (Fig
S1) during the different growth phases. Fig. 1b depicts the ratio of
rRNA per dry cell weight during cultivation. An initial increase in
the rRNA concentration up to 21% rRNA per cell dry weight was
clearly visible in the early exponential phase. This increase was
accompanied by an increase in the maximal specific growth rate
(μmax ) to 1.2 h−1 . Interestingly, after reaching this peak, the normalized rRNA content decreased slightly throughout exponential growth, which is consistent with the slowing of growth indicated by a lower specific growth rate. Eventually, growth became
irregular in the transition from exponential to stationary phase,
until a final growth arrest was observed during the stationary
phase. rRNA content dropped steadily, and reached a value of
around 2% rRNA per cell dry weight after 9 h of cultivation.
Failmezger et al.
3
Figure. 1 rRNA dynamics during batch cultivation in 2 × TY media at 37◦ C. (a) Escherichia coli strain MCrg was cultivated in replicate shaking flasks, and the cell dry
weight (open squares) and specific growth rate (filled triangles) were monitored over time. Growth periods: (1) adaptation phase; (2) exponential phase; (3) transition
phase; and (4) stationary phase. (b) rRNA content (sum of 16S and 23S rRNA) during cultivation. rRNA is depicted as a percentage of total cell dry weight. (c) Correlation
of rRNA content and growth rate (R2 = 0.9). Error bars show standard deviations.
Plotting rRNA content and specific growth rate reveals an expected linear correlation (Fig. 1c), demonstrating, as expected,
that ribosomes are growth limiting.
Monitoring ribosomes by fluorescence during batch
cultivation
An absolute prerequisite for following ribosome dynamics by
fluorescence is that the monitored fluorescent-labeled protein
reflects the overall dynamics of the ribosome. The E. coli ribosome consists of approximately 55 proteins and three rRNA
molecules. As rRNA possesses the major catalytic functions including polymer formation, decrease in rRNA content corresponds to a loss of ribosome function. Consequently, fluorescence readout and rRNA content have to be correlated to mirror
ribosome dynamics.
To test the feasibility of ribosome dynamic measurements
based on fluorescence, we measured the fluorescence of labeled
ribosomal proteins L19 and S2 and compared the fluorescence
readout to the determined rRNA content per biomass. Remarkably, the fluorescence of L19-GFP showed a similar dynamic as
rRNA content measured by CGE-LIF (Fig. 2a). In fact, we were able
to determine a linear correlation (R2 = 0.97) between fluorescence of L19-EGFP and the rRNA/dry cell weight ratio throughout
the cultivation of E. coli strain MCrg (Fig. 2b). Next, we analyzed
the appropriateness of the S2-mCherry fusion protein to reflect
rRNA dynamics (Fig. 2c). Again, both the increase and decrease
of rRNA content were in agreement with mCherry fluorescence.
This agreement was highlighted by a linear correlation of rRNA
and mCherry (R2 = 0.97) (Fig. 2d).
Ribosome decrease during slowing of growth is
accompanied by de novo synthesis
We detected a decrease in intracellular ribosome content effortlessly measured by fluorescence that likely resulted in a reduction of translation capacity as substrates became rate limiting. In
consequence, the growth rate dropped. Three scenarios may explain the perceived decline of cellular rRNA at slowing of growth:
(i) the decrease in ribosome content results from dilution of preexisting ribosomes by passive separation during cell division
(Birky 1982; Huh and Paulsson 2011); (ii) ribosomes are actively
degraded and diluted; and (iii) ribosomes are diluted and new ribosomes are synthesized. To characterize the decrease in rRNA
content in detail, we calculated the rRNA content per cell assuming only dilution of preexisting ribosomes according to the
measured cell growth rate using Equation 2.
cr R N A (t) = cr R N A0 × exp(0.5 × (t − 2)(a × (t − 2) − 2 × b)
(2)
Here, crRNA (t) denotes the simulated cellular rRNA content
and crRNA0 is the cellular rRNA content after 2 h. Parameters a and b were derived from a linear fit of the growth rate.
The results of this simulation, considering the decline of the
4
FEMS Microbiology Letters, 2017, Vol. 364, No. 6
Figure. 2 Ribosome dynamics followed by fluorescence. (a) Fluorescence of EGFP fused to ribosomal protein L19 (filled triangles), cellular rRNA content (open squares)
and ribosomes per cell (line). (b) Linear correlation of L19-EGFP and cellular rRNA. (c) Fluorescence of mCherry fused to ribosomal protein S2 (filled triangles), cellular
rRNA content (open squares) and ribosomes per cell (line). (d) Linear correlation of S2-mCherry and cellular rRNA. The number of ribosomes per cell was calculated
assuming a molecular mass of 1.51 MDa for rRNA and 2.7 MDa for ribosomes. Dry mass per cell values at different growth rates were taken from Bremer and Dennis
(1996) and interpolated. Error bars show standard deviations.
determined specific growth rates after 2 h of growth, were
not supported by the determined rRNA data. Instead, the decrease in cellular rRNA content was highly overestimated (Fig. 3).
Accordingly, taking into account only the active degradation
of existing ribosomes should result in an even stronger decrease. Therefore, it follows that the decline in rRNA content
observed represents a passive dilution of existing ribosomes
accompanied by de novo ribosome synthesis. Next, we calculated the fraction of de novo synthesized ribosomes for each
time point. Interestingly, following these calculations, we found
an almost constant de novo synthesis rate of ribosomes during slowing of growth under the conditions tested. Maintaining a constant ribosome synthesis rate most likely occurs at
the expense of reducing the metabolic protein fraction. Nevertheless, based on the identified scenario one can assume that
this rate is not high enough to keep a constant ribosome concentration in the cell. As the number of ribosomes per cell decreases, the translation rate, which is the product of elongation rate and amount of actively translating ribosomes, falls
concomitantly.
Monitoring of ribosomes by fluorescence during
nutritional shift
We aimed to depict ribosome dynamics as a result of certain induced growth patterns. Therefore, E. coli MCrg was cultivated in
the presence of acetate as the sole carbon source and the ribosome content was monitored. After 5 h of cultivation, the cells
were subjected to a nutritional upshift by glucose addition. The
resulting immediate upshift of the specific growth rate was associated with an increase of the cellular ribosome content, as
shown in Fig. 4a. We observed a sudden decrease of fluorescence
once glucose was depleted and growth stopped, indicating rapid
degradation of ribosomes when cells starved for carbon. Next,
ribosome dynamics in cells undergoing a diauxic transition by
switching from utilizing glucose as a carbon source to sorbitol
was studied (Fig. 4b). We found that after an initial drop of the ribosome content due to glucose depletion and growth arrest, the
ribosome level remains constant during the transition phase. After the transition phase, ribosomes are synthesized and growth
continues.
Failmezger et al.
Figure. 3 Decrease in ribosome content during slowing of growth assuming different scenarios. Measured cellular rRNA content (filled circles) and simulated
decrease of ribosomes as a result of passive dilution based on the growth rate
(line) were compared. Dilution of existing ribosomes by cell division was simulated based on the respective growth rate. The fraction of de novo synthesized
ribosomes (filled triangles) at each time point and the rRNA de novo synthesis
rate (open squares) were also calculated.
DISCUSSION
The rRNA content of a cell is an important characteristic to describe the physiological state during bacterial growth. In the
present work, we determined rRNA content using CGE-LIF in
combination with an internal standard that was added to the
cell lysate. However, similar to other rRNA analysis methods
using gel electrophoresis, rRNA extraction is obligatory for this
approach. Consequently, we asked whether it was possible to
mirror ribosome dynamics by simple and easy to monitor fluorescence measurements using fluorescent-labeled ribosomal
proteins. Indeed, we found a linear correlation between cellular
rRNA content and fluorescence signals from ribosomal proteins
S2 and L19. Both labeled proteins have previously been used to
identify subunit assembly defects upon treatment with translation inhibitors (Nikolay et al. 2014). Our results add to this work
5
and clearly demonstrate the feasibility of tracking ribosome dynamics during cultivation using fluorescence. Determining ribosome content by rRNA measurements revealed a highly dynamic
behavior in which content was directly coupled to changes in
the specific growth rate. In fact, the observed ribosome dynamics during batch cultivation, as presented here, shows a pattern similar to that previously described by Piir et al. (2011), who
monitored ribosomes by pulse-labeling experiments. Comparison of the observed ribosome dynamics with simulations corresponding to several possible scenarios, such as ribosome dilution, degradation or de novo synthesis, allowed us to link the
decrease in cellular rRNA content to dilution and concomitant de
novo synthesis of ribosomes. This finding is consistent with active ribosome concentration limiting the cell growth rate (Scott
et al. 2010). A well-known mechanism that characterizes the
transition from exponential to stationary phase is the ‘stringent
response’ (for a review, see Potrykus and Cashel 2008). Here, resources are shifted from expressing genes mainly required for
growth to operons necessary for amino acid biosynthesis. Consequently, the translation machinery is downregulated. In contrast, assuming a linear relationship between dilution of preexisting ribosomes and de novo synthesis, we calculated a constant
ribosome synthesis rate during slowing of growth. The use of
nutrient-rich media resulted in a very slow transition from growing cells to final growth arrest, as can be seen by the slowing
down of the growth rate over several hours. Whether the observed ribosome dynamics is related to a ppGpp-induced regulation (stringent response) will be addressed in future studies.
Furthermore, we were able to monitor ribosome dynamics at different induced growth pattern such as nutritional
upshift from acetate to glucose and diauxic transition from
glucose to sorbitol. Interestingly, in both cases an abrupt
growth arrest was always accompanied by a drastic decrease in the cellular ribosome content examined by fluorescence measurements. Accordingly, we conclude that degradation of ribosome complexes also includes fluorescent-labeled
ribosomal proteins, and therefore both ribosome synthesis
and degradation can be monitored by fluorescence, respectively. These observations clearly manifest the usability and
Figure. 4 Ribosome dynamics during nutritional shift conditions. (a) Escherichia coli MCrg was cultivated on acetate, and glucose was pulsed after 5 h of growth. (b)
Ribosome dynamics during diauxic growth on glucose and sorbitol. Illustrated are the cell dry weight (dashed line), specific growth rate (straight line) and ribosome
content (filled circles) (fluorescence of EGFP fused to ribosomal protein L19).
6
FEMS Microbiology Letters, 2017, Vol. 364, No. 6
the potential of the presented approach to profile ribosome
dynamics.
While the presented method takes advantage of fluorescentlabeled ribosomal proteins, one has to keep in mind that complex molecular mechanisms such as ribosomal subunit assembly are easily disturbed by fusion proteins. Similar to Nikolay
et al. (2014), we did not observe any growth defects of Escherichia
coli MCrg. Nevertheless, it was shown that a similar modification
of ribosomal protein S2 resulted in an incapability to form 100S
ribosomes (Shcherbakova, Hideki and Nobuo 2015).
In conclusion, our work shows the general suitability of
fluorescent-labeled ribosomal proteins to perform ribosome dynamics studies, opening the field of quantitative understanding
of translation.
SUPPLEMENTARY DATA
Supplementary data are available at FEMSLE online.
ACKNOWLEDGEMENTS
The authors thank the Institute of Cell Biology and Immunology,
University of Stuttgart, for the provision of the Agilent Bioanalyzer. The authors thank Dr R. Nikolay for providing Escherichia
coli MCrg.
FUNDING
This work was supported by the Bundesministerium für Bildung und Forschung, Berlin, Germany [BMBF grant number:
031A157D].
Conflict of interest: None declared.
REFERENCES
Bakshi S, Choi H, Mondal J et al. Time-dependent effects of
transcription-and translation-halting drugs on the spatial
distributions of the Escherichia coli chromosome and ribosomes. Mol Microbiol 2014;94:871–87.
Basturea GN, Zundel MA, Deutscher MP. Degradation of ribosomal RNA during starvation: comparison to quality control
during steady-state growth and a role for RNase PH. RNA
2011;17:338–45.
Birky CW. The partitioning of cytoplasmic organelles at cell
division. Int Rev Cytol 1982;15:49–89.
Boedtker H. Conformation independent molecular weight determinations of RNA by gel electrophoresis. Biochim Biophys Acta
1971;240:448–53.
Bremer H, Dennis PP. Modulation of chemical composition and
other parameters of the cell by growth rate. In: Neidhardt FC
(ed.). Escherichia coli and Salmonella. Washington: ASM Press,
1996, 1553–69.
Chai Q, Singh B, Peisker K et al. Organization of ribosomes and
nucleoids in Escherichia coli cells during growth and in quiescence. J Biol Chem 2014; 289:11342–52.
Deutscher MP. Degradation of stable RNA in bacteria. J Biol Chem
2003;278:45041–44.
Failmezger J, Nitschel R, Sanchez-Kopper A et al. Sitespecific cleavage of ribosomal RNA in Escherichia colibased cell-free protein synthesis systems. PLoS One 2016;
11:e0168764.
Hardiman T, Ewald JC, Lemuth K et al. Quantification of rRNA
in Escherichia coli using capillary gel electrophoresis with
laser-induced fluorescence detection. Anal Biochem 2008;374:
79–86.
Housley J, Tollervey D. The many pathways of RNA degradation.
Cell 2009;136:763–76.
Huh D, Paulsson J. Random partitioning of molecules at cell division. P Natl Acad Sci USA 2011;108:15004–9.
Nikolay R, Schloemer R, Schmidt S et al. Validation of a
fluorescence-based screening concept to identify ribosome
assembly defects in Escherichia coli. Nucleic Acids Res 2014,
DOI: 10.1093/nar/gku381.
Piir K, Paier A, Liiv A et al. Ribosome degradation in growing bacteria. EMBO Rep 2011;12:458–62.
Potrykus K, Cashel M. ppGpp: still magical? Microbiology
2008;62:35.
Reddy K, Gilman M. Preparation of bacterial RNA. In: Ausubel
FM, Brent R, Kingston RE et al. (eds). Current Protocols
in Molecular Biology. New York: John Wiley & Sons, 1987,
4.4.4–7.
Reijnders L, Sloof P, Borst P. The molecular weights of the
mitochondrial-ribosomal RNAs of Saccharomyces carlsbergensis. Eur J Biochem 1973;35:266–9.
Scott M, Gunderson CW, Mateescu EM et al. Interdependence of
cell growth and gene expression: origins and consequences.
Science 2010;330:1099–102.
Shcherbakova K, Hideki N, Nobuo S. Role of 100S ribosomes in bacterial decay period. Genes Cells 2015;20:
789–801.
Wilms B, Hauck A, Reuss M et al. R. High-cell-density fermentation for production of l-n-carbamoylase using an expression
system based on the Escherichia coli rhaBAD promoter. Biotechnol Bioeng 2001;73:95–103.
Zundel MA, Basturea GN, Deutscher MP. Initiation of ribosome degradation during starvation in Escherichia coli. RNA
2009;15:977–83.