STUDYING OF THE BACTERIAL GROWTH PHASES USING
FOURIER TRANSFORM INFRARED SPECTROSCOPY AND
MULTIVARIATE ANALYSIS
HAMZAH M. AL-QADIRI1,7, NIVIN I. AL-ALAMI2, MENGSHI LIN3,
MURAD AL-HOLY4, ANNA G. CAVINATO5 and BARBARA A. RASCO6
1
Department of Nutrition and Food Technology
Faculty of Agriculture, The University of Jordan
Amman 11942, Jordan
2
Water and Environment Research and Study Center
The University of Jordan
Amman 11942, Jordan
3
Food Science Program
University of Missouri
Columbia, MO 65211-5160
4
Department of Clinical Nutrition and Dietetics
Faculty of Allied Health Sciences, Hashemite University
Zarqa, Jordan
5
Chemistry and Biochemistry Program
Eastern Oregon University
One University Blvd., La Grande, OR 97850-2899
6
Department of Food Science and Human Nutrition
Box 646376, Washington State University
Pullman, WA 99164-6376
Accepted for Publication December 3, 2007
ABSTRACT
This study examined the potential of Fourier transform infrared (FT-IR)
absorbance spectroscopy to detect biochemical changes in bacterial cells that
occur during bacterial growth phases in batch culture. Two bacterial strains,
Escherichia coli ATCC 25922 and Listeria innocua ATCC 51742 were cultured
in brain heart infusion (BHI) broth and incubated at 37C and cells recovered
7
Corresponding author. TEL: (+962-6)-5355000, ext. 22422; FAX: (+962-6) 5355577; EMAIL:
[email protected]
Journal of Rapid Methods & Automation in Microbiology 16 (2008) 73–89. All Rights Reserved.
© 2008, The Author(s)
73
Journal compilation © 2008, Blackwell Publishing
74
H.M. AL-QADIRI ET AL.
at: the lag phase (3 h), the log phase (7 h), the stationary phase (24 h), and the
death phase (10 days) from incubation in BHI (n = 2).
Major variations in the biochemical structure of bacterial cells during
growth were observed. Over the range of 1800–900 cm-1, loadings 1 (principal component [PC] 1) and 2 (PC2) accounted for 88% of the total variability
(76% and 12%, respectively) for E. coli cells, and 80% (72% and 8%, respectively) for L. innocua cells. Changes were attributable to cellular structure of
bacterial cells (the variation in nucleic acids, proteins, lipids and polysaccharides) that occur during the growth phases.
PRACTICAL APPLICATIONS
The focus of this study was to examine Fourier transform infrared (FTIR) spectral features of bacterial cells at different growth stages and to determine if variations in composition and distribution of the biochemical
components of cells during growth phases can be distinguished using this
spectroscopic technique. This could explain the primary functional principle
behind detection and discrimination of bacteria by FT-IR spectroscopy.
INTRODUCTION
The normal bacterial growth curve has four stages: lag phase, log (exponential) phase, stationary phase, and death phase (Atlas 1988). These phases
characterize typical bacterial growth in batch culture where a complex or a
defined liquid nutrient medium is used. In this closed environment, the
required nutrients become exhausted and the metabolic products start to accumulate due to the bacterial growth (Akerlund et al. 1995). Lag, log and
stationary phases are characterized by distinct biochemical reactions for synthesis of cellular components necessary for cell growth and division.
During the lag phase bacteria show no growth – although cells increase in
size – but begin to prepare for reproduction such synthesis of DNA and cellular
enzymes (Garbutt 1997). During the log phase the bacterial biomass increases
linearly with time, with the number of bacterial cells doubling with every unit
of time. This phase is defined as the balanced growth stage because the average
composition of the cells remains constant with bacterial culture properties
(i.e., protein and DNA) increasing at the same rate (Akerlund et al. 1995).
When the stationary phase is reached, there is no longer a net increase in viable
bacterial cell numbers and cellular metabolic activity is decreased meaning
that the growth rate is equal to the death rate (Akerlund et al. 1995). During
the transition between log and stationary phase, cellular components are syn-
DISCRIMINATION OF BACTERIAL GROWTH PHASES BY FT-IR
75
thesized at unequal rates (unbalanced growth); therefore, the biochemical
composition of cells in the stationary phase is different from that in the log
phase (Atlas 1988). Bacterial growth is inhibited during the stationary phase
since the essential nutrients required for bacterial growth are exhausted and
metabolic inhibitory byproducts accumulate. Eventually, death phase is
reached where bacterial cells are broken down (cell lysis) due to the additional
accumulation of inhibitory byproducts, depletion of cellular energy, and pH
changes (Garbutt 1997).
Studying the changes in biochemical composition, specifically the relative amounts and changes in the distribution of various cellular components
such as DNA, proteins, fatty acids and polysaccharides during bacterial
growth could improve our understanding of microbial physiology as well as
the behavior of different bacterial strains in various batch culture systems.
FT-IR spectroscopy is a new technique for studying the biochemical
structure of microorganisms. It is used to identify bacteria and yeasts (Helm
et al. 1991; Beattie et al. 1998; Rodriguez-Saona et al. 2001; Wenning et al.
2002) and to discriminate foodborne pathogens (Kummerle et al. 1998;
Choo-Smith et al. 2001; Mossoba et al. 2003; Al-Holy et al. 2006; Garip et al.
2007). Distinctive spectral “biochemical fingerprints” can be used to differentiate microbial strains and to ascertain how the cellular composition of microbial cells are different (Goodacre et al. 1996; Kansiz et al. 1999; Lin et al.
2004; Winder et al. 2004; Al-Qadiri et al. 2006; Garip et al. 2007). FT-IR
spectra between 4000–400 cm-1 include stretching vibrations and deformation
of C–H, C=O, N–H, P=O, and C–O–C and which reflect the differences in the
biochemical composition of microbial cells i.e., DNA, proteins, polysaccharides and fatty acids (Hong et al. 1999; Kansiz et al. 1999; Oberreuter et al.
2002; Winder et al. 2004). The main objective of this work was to study FT-IR
spectral features of bacterial cells at different growth stages and to determine
if variations in composition and distribution of the biochemical components of
cells during growth phases can be distinguished using this spectroscopic
technique.
MATERIALS AND METHODS
Bacterial Strains and Growth Conditions
Bacterial cells used in this study were obtained form the culture collection in the Department of Food Science and Human Nutrition, Washington
State University. Two bacterial strains were studied: Escherichia coli ATCC
25922 as Gram-negative, and Listeria innocua ATCC 51742 as Gram-positive.
Both strains were activated first by transferring bacterial inoculum from refrig-
76
H.M. AL-QADIRI ET AL.
erated slant to tryptic soy agar (TSA) (Difco, Sparks, MD) followed by
incubation of cultured plates at 37C for 24 h. A well-isolated single colony of
each strain was then inoculated into 100 mL of brain heart infusion (BHI)
broth (Bacto, Sparks, MD) and strains were grown aerobically at 37C.
Four treatments of both strains were studied in duplicate (n = 2). The first
treatment (lag phase) was examined after 3 h in BHI at 37C, the second
treatment (log phase) after 7 h, the third treatment (stationary phase) after 24 h,
and the fourth treatment (death phase) after 10 days. The bacterial count at each
treatment time was determined and the limiting factor for these experiments was
the cell numbers at lag phase treatment as the cell numbers were relatively low.
To eliminate the effect of the different bacterial counts on spectral measurements, the cell concentration present at lag phase was used for all treatments. To
determine the number of bacterial cells during the lag phase, 1 mL from the BHI
broth was taken after 3 h from incubation and appropriate serial dilutions were
made (using sterile 0.1% peptone; Bacto). Samples were enumerated (in
duplicate) by the standard spread plate technique on TSA, plates were then
incubated at 37C for 48 h and bacterial count was recorded as log viable
colony-forming unit (cfu)/mL. This procedure was performed after 7 h, 24 h
and 10 days from incubation to determine the bacterial count of the other phases.
Accordingly, appropriate dilutions were made for log, stationary and death
phases using sterile 0.9% saline solutions to give the same bacterial count as lag
phase before spectral measurements were obtained.
For the lag phase treatment, 50 mL broth of each strain was transferred
after 3 h from incubation to a 50-mL sterile centrifuge tube. The tubes were
then centrifuged at room temperature for 15 min at 4000 rpm (CENTRA CL-2
Model 120 Thermo IEC; Thermo IEC, San Jose, CA) to harvest bacterial cells.
To reduce the effect of media components and bacterial metabolites, the
resultant pellets were resuspended in 10 mL sterile 0.9% saline solutions and
centrifuged again; this step was repeated twice. The supernatants were
removed and the remaining wet pellets were resuspended in 500 mL sterile
0.9% saline solution and vigorously mixed to obtain a homogeneous cell
distribution. A 300 mL of each bacterial suspension was uniformly applied to
a 0.2-mm pore size aluminum oxide membrane filter (25 mm OD; Anodisc,
Whatman, Inc., Clifton, NJ). Under aseptic conditions, the filters were then
air-dried at room temperature (22C) for 30 min under a laminar flow hood.
This procedure yielded a filter with a relatively uniform coating of dried
bacterial cells. The Anodisc membrane filter was used because it does not
contribute significantly to the bacterial spectra and it gives a smooth and a flat
surface for bacterial cell adhesion (Rodriguez-Saona et al. 2001; Lin et al.
2004; Rodriguez-Saona et al. 2004). Samples from the log, the stationary and
the death phases received similar treatment except that an initial volume of
10 mL broth was used.
DISCRIMINATION OF BACTERIAL GROWTH PHASES BY FT-IR
77
FT-IR Spectroscopy and Measurement
FT-IR bacterial spectra were collected using Thermo Nicolet Avatar 360
FT-IR spectrometer (Thermo Electron, Inc., San Jose, CA). The dried Anodisc
membrane filters (covered with a homogeneous and dried smooth layer of
examined bacteria) were positioned in direct contact with an infrared attenuated total reflection zinc selenide crystal (Schmitt and Flemming 1998;
Rodriguez-Saona et al. 2001). FT-IR spectra were recorded from 4000–
600 cm-1 at resolution of 2 cm-1. Thirty spectra were acquired for each sample
treatment (growth phase) at room temperature (each spectrum composed of an
average of 128 separate scans). FT-IR spectra were mean centered and baseline
corrected.
Multivariate Statistical Analysis
The collected FT-IR bacterial spectra were analyzed using OMNIC
(Thermo Electron, Inc.) and Delight version 3.2.1 (Textron Systems, Wilmington, MA) software. To reduce the overlapping between relatively similar
bacterial spectra, data preprocessing algorithms were carried out before data
analysis. Data preprocessing included spectral binning by 2 cm-1 and smoothing with a Gaussian function of 15 cm-1, this followed by a second derivative
transformation, 15 cm-1 gap (Huang et al. 2001; Huang et al. 2002; Lin et al.
2003).
The multivariate statistical analysis techniques of principal component
analysis (PCA) and soft independent modeling of class analogy (SIMCA)
were used to statistically analyze the obtained spectral data (Kansiz et al.
1999; Lin et al. 2003; Oust et al. 2004; Rodriguez-Saona et al. 2004). PCA
was used to show if there is a natural clustering in the spectral data through
capturing the related variations among bacterial spectra (Goodacre et al. 1998;
Rodriguez-Saona et al. 2001; Nilsen et al. 2002). It was also performed to
obtain graphical representations of similarities and differences from infrared
spectral data (Martens and Naes 1989; Goodacre et al. 1998; Kansiz et al.
1999; Rodriguez-Saona et al. 2001). The first principal component (PC1)
characterizes the maximum percentage of bacterial spectral variation followed
by the second PC (PC2) that describes the second most important factor of the
remaining spectral variation, and so on (Rodriguez-Saona et al. 2004). SIMCA
was mainly used to classify samples according to their analogy to the training
sample treatments (Hampton et al. 2001–2002).
RESULTS AND DISCUSSION
Table 1 shows the average of viable bacterial counts (log10 cfu/ml) for
E. coli and L. innocua strains at the different growth phases after incubation
78
H.M. AL-QADIRI ET AL.
TABLE 1.
THE AVERAGE VIABLE BACTERIAL COUNTS (log10 cfu/ml)* of ESCHERICHIA COLI AND
LISTERIA INNOCUA STRAINS AT THE DIFFERENT GROWTH PHASES IN BATCH
CULTURE (BHI BROTH AT 37C)
Strain
E. coli ATCC 25922
L. innocua ATCC 51742
Growth phase treatment
Lag (3 h)
Log (7 h)
Stationary (24 h)
Death (10 days)
4.30
4.08
6.90
7.08
9.70
9.64
8.30
8.65
*Bacteria were counted by spread plating on tryptic soy agar.
BHI, brain heart infusion; cfu, colony-forming units.
on TSA at 37C for 48 h. As shown from Table 1, the bacterial count during
growth was different from phase to phase. Because there is a variation in
composition and distribution of bacterial biochemical structure including cell
wall and cell membrane structural lipids (phospholipids and lipopolysaccharides), cellular proteins and enzymes, nucleic acids, and polysaccharides this
should be reflected in spectral features at each growth stage.
Conventional methods to determine bacterial cell numbers during different growth phases (i.e., standard plate count and optical density) cannot
describe biochemical changes that occur throughout the growth process.
However, FT-IR spectroscopy was used to characterize these changes through
spectral features analysis of various functional chemical groups and polar
bonds that reflect the biochemical composition of the cellular constituents.
Figs. 1 and 2 show the representative FT-IR spectra (4000–600 cm-1) of
E. coli and L. innocua during the four growth phases. In region I (Figs. 1 and
2), major spectral variations were observed in amide I and amide II due
possibly to C=O stretching vibrations (~1650 cm-1) and N–H deformation or
bending (~1540 cm-1) of amides associated with proteins, respectively (Kansiz
et al. 1999; Choo-Smith et al. 2001; Filip and Hermann 2001; Lin et al. 2004).
Region II, commonly called the fingerprint region (1300–900 cm-1), contains vibrational features of cellular proteins (including enzymes), nucleic
acids, and cell membrane and cell wall components (Goodacre et al. 1996).
Variations between experimental treatments in this region could be associated
with P=O asymmetric and symmetric stretches of the phosphodiester backbone of nucleic acids at ~1242 and ~1080 cm-1 respectively, and C–O–C
stretching vibrations of polysaccharides (1200–900 cm-1) associated with the
bacterial cell wall and cell membrane peptidoglycan layer and lipopolysaccharide outer leaflet (Kansiz et al. 1999; Choo-Smith et al. 2001; Filip and
Hermann 2001; Lin et al. 2004).
The death phase for both strains had a distinctive spectral pattern. The
absorbance was relatively higher compared to the other growth phases and that
DISCRIMINATION OF BACTERIAL GROWTH PHASES BY FT-IR
79
A
B
Fatty acids
Polysaccharides
Amide I
C
Amide II DNA/RNA
D
I
II
FIG. 1. REPRESENTATIVE FT-IR SPECTRA OF ESCHERICHIA COLI ATCC 25922 DURING
GROWTH PHASES
(A) Lag phase. (B) Log phase. (C) Stationary phase. (D) Death phase.
A
B
C
D
I
II
FIG. 2. REPRESENTATIVE FT-IR SPECTRA OF LISTERIA INNOCUA ATCC 51742 DURING
GROWTH PHASES
(A) Lag phase. (B) Log phase. (C) Stationary phase. (D) Death phase.
80
H.M. AL-QADIRI ET AL.
might be due to: (1) cell lysis leading to leakage of the cellular materials
outside the cytoplasm causing a net increase in absorbance; or (2) the presence
of the dead cells on the membrane filter; these could not be enumerated on
TSA but at the same time would have contributed to the microbial cell density
on the membrane. However, in spite of the higher absorbance during the death
phase, the spectral pattern for death phase was also characterized by a relative
depletion of major cellular components (particularly in region II) including
proteins, nucleic acids and most predominantly polysaccharides (1100–
950 cm-1) due to the destruction of the cell wall and cell membrane. The
depletion of the cellular components might be associated to the reduction in
the metabolic activity during the death phase.
Spectral binning, smoothing and second derivative transformation were
performed before multivariate statistical analysis to: (1) detect more details
about bacterial biochemical structure by increasing the number of discriminative features associated with the bacterial spectra (Pedone et al. 2003; Sandt
et al. 2003); (2) reduce spectral pattern overlapping; and (3) characterize and
amplify the related variations among bacterial FT-IR spectra since unprocessed spectral patterns look similar (Kansiz et al. 1999; Lin et al. 2005).
Spectral normalization was conducted to balance the differences in absorbance
as a prerequisite to PCA and SIMCA analysis (Holt et al. 1995; Lin et al.
2003).
The composition and distribution of the biochemical components in bacterial cells are different during growth phases and FT-IR spectroscopy provides
a method and an indication of which specific biochemical changes occur
during different growth phases. The 15-point second order derivatization of
region I for both strains is shown in Figs. 3 and 4. After transformation, the
spectral variations (1800–1300 cm-1) among the four growth phases become
more distinctive for both strains. Major variations could be observed in the log,
the stationary and the death phases which makes these stages obviously distinguishable from each other and from the lag phase because of unique amide
I and amide II bands (proteins) in the range of ~1650 and ~1540 cm-1.
However, relatively minor variations could also be observed in the range of
~1455 and ~1398 cm-1 from CH3 and CH2 asymmetric and symmetric deformation of proteins.
Figs. 5 and 6 show the 15-point second derivative transformation of
bacterial spectra in the fingerprint region (region II). The major differences
might be seen in the P=O asymmetric stretch of the phosphodiester backbone
of nucleic acids (~1242 cm-1), P=O symmetric stretch of the nucleic acid
ribose or deoxyribose moieties (~1080 cm-1) (Kansiz et al. 1999; Filip and
Hermann 2001), and C–O–C stretching vibration of cell wall and cell membrane polysaccharides (Kansiz et al. 1999; Filip and Hermann 2001). There
were very clear differences in polysaccharide spectral features between the log
DISCRIMINATION OF BACTERIAL GROWTH PHASES BY FT-IR
81
1
2
3
4
FIG. 3. REPRESENTATIVE SECOND DERIVATIVE TRANSFORMATION SPECTRA
(1800–1300 cm-1) OF ESCHERICHIA COLI ATCC 25922 DURING GROWTH PHASES
), log phase (
), stationary phase (
), and death phase (
).
Lag phase (
1, 2: C=O stretching vibration and N–H deformation of amides associated with proteins (amide I and
amide II); 3, 4: CH3 AND CH2 asymmetric and symmetric deformation of proteins.
1
3
4
2
FIG. 4. REPRESENTATIVE SECOND DERIVATIVE TRANSFORMATION SPECTRA
(1800–1300 cm-1) OF LISTERIA INNOCUA ATCC 51742 DURING GROWTH PHASES
Lag phase (
), log phase (
), stationary phase (
), and death phase (
).
See caption of Fig. 3 for details.
82
H.M. AL-QADIRI ET AL.
1
2
FIG. 5. REPRESENTATIVE SECOND DERIVATIVE TRANSFORMATION SPECTRA
(1300–900 cm-1) OF ESCHERICHIA COLI ATCC 25922 DURING GROWTH PHASES
), log phase (
), stationary phase (
), and death phase (
).
Lag phase (
1: P=O asymmetric stretches of the phosphodiester backbone of nucleic acids; 2: P=O symmetric
stretches and C–O–C stretching vibrations of polysaccharides.
1
2
FIG. 6. REPRESENTATIVE SECOND DERIVATIVE TRANSFORMATION SPECTRA
(1300–900 cm-1) OF LISTERIA INNOCUA ATCC 51742 DURING GROWTH PHASES
Lag phase (
), log phase (
), stationary phase (
), and death phase (
).
See caption of Fig. 5 for details.
DISCRIMINATION OF BACTERIAL GROWTH PHASES BY FT-IR
83
B
D
A
C
FIG. 7. PRINCIPAL COMPONENTS ANALYSIS OF ESCHERICHIA COLI ATCC 25922 AT
DIFFERENT GROWTH PHASES
(A) Lag phase. (B) Log phase. (C) Stationary phase. (D) Death phase.
and the death phases. While there was a relative depletion in the polysaccharides peak during the death phase, possibly from cell lysis, the polysaccharides
content increased during the log phase showing an increase in cell biomass.
PCA and SIMCA statistical analysis techniques were employed to differentiate and classify bacteria by growth phases (Oust et al. 2004; RodriguezSaona et al. 2004). PCA captures the related variations among spectral data
and clusters groups with presumptively similar spectral properties.
A mean centered PCA was performed on the second derivative FT-IR
spectra. Figs. 7 and 8 show the two-dimensional PCA clustering results from
FT-IR spectral data for E. coli and L. innocua strains at different growth
phases. For both strains, clear segregation was observed between the different
growth phases. Bacterial cells during the lag, the log and the stationary phases
were clearly segregated with distinct clustering of these cells from those in the
death phase, which indicates that there is a significant difference in the biochemical composition of bacterial cells during the death phase. However,
bacterial cells during the lag and the stationary phases were closely clustered
together, but were distinct from bacterial cells in the log phase possibly
reflecting spectral differences associated with rapid cell growth during the log
phase.
To predict which spectral regions provide the most significant contributions to data variation during the growth phases, loadings plots are used to
highlight the contribution of each variable (wavenumber) to each principal
component (Kansiz et al. 1999; Lin et al. 2005). The first and the second
84
H.M. AL-QADIRI ET AL.
B
D
A
C
FIG. 8. PRINCIPAL COMPONENTS ANALYSIS OF LISTERIA INNOCUA ATCC 51742 AT
DIFFERENT GROWTH PHASES
(A) Lag phase. (B) Log phase. (C) Stationary phase. (D) Death phase.
loadings plot from PCA analysis over the range of 1800–900 cm-1 for E. coli
and L. innocua strains are shown in Figs. 9 and 10, respectively. Large positive
or negative loadings might be associated with spectral regions within the
original spectra that are responsible for bacterial differentiation (Holt et al.
1995). Over the range of 1800-900 cm-1, loadings 1 (PC1) and 2 (PC2)
accounted for 88% of the total variability (76% and 12%, respectively) in case
of E. coli cells, and 80% (72% and 8%, respectively) in case of L. innocua
cells. Accordingly, spectral variables in the range of 1800–900 cm-1 provided
the majority of the total variance in the FT-IR data compared to other variables.
The major contribution to the spectral variation were amide I and amide II
bands; plus the CH3 and CH2 asymmetric and symmetric deformation of
proteins, P=O asymmetric and symmetric stretches of the phosphodiester
nucleic acid backbone, and polysaccharide C–O–C stretching vibrations providing further indication that FT-IR could detect changes in the cellular structure of bacterial cells that occur during growth and death.
Sample classification was performed using SIMCA analysis by comparing spectral features of each growth phase with models constructed from
spectra from the remaining treatments. Test spectra were assigned to a class
based upon the degree of analogy. Table 2 shows SIMCA classification results
for each sample treatment (growth phase) for E. coli and L. innocua strains
compared to the other test treatments.
DISCRIMINATION OF BACTERIAL GROWTH PHASES BY FT-IR
85
PC1
PC1
PC2
PC1
PC1
FIG. 9. LOADINGS PLOT OF THE FIRST AND THE SECOND PRINCIPAL COMPONENTS
OBTAINED FROM PRINCIPAL COMPONENTS ANALYSIS OF ESCHERICHIA COLI ATCC
25922 FT-IR SPECTRA
PC1
PC2
PC1
PC2
PC1
FIG. 10. LOADINGS PLOT OF THE FIRST AND THE SECOND PRINCIPAL COMPONENTS
OBTAINED FROM PRINCIPAL COMPONENTS ANALYSIS OF LISTERIA INNOCUA ATCC
51742 FT-IR SPECTRA
86
H.M. AL-QADIRI ET AL.
TABLE 2.
SIMCA CLASSIFICATION RESULTS FOR EACH SAMPLE TREATMENT (GROWTH PHASE)
FOR ESCHERICHIA COLI AND LISTERIA INNOCUA STRAINS
E. coli ATCC 25922
L. innocua ATCC 51742
Growth phase
% of correctly classified
spectra
Growth phase
% of correctly classified
spectra
Lag
Log
Stationary
Death
90.0
96.7
93.3
83.3
Lag
Log
Stationary
Death
76.7
90.0
100.0
76.7
SIMCA, soft independent modeling of class analogy.
As shown from Table 2, bacterial samples during the log phase and the
stationary phase were correctly classified for both strains (90% or greater).
However, differentiation of cells in the lag and the death phases were less
reliable, particularly for L. innocua.
CONCLUSIONS
FT-IR spectroscopy can be used to determine biochemical changes that
occur during bacterial cell growth and to classify and discriminate bacterial
cells during the different phases. Differences in Gram-negative and Grampositive cells are clearly distinguishable. FT-IR spectral features for detection
and discrimination of different bacterial strains and between growth phases are
primarily in the amide I (~1650 cm-1) and amide II region (~1540 cm-1) that
associated with proteins, as well as nucleic acid and polysaccharide moieties
in the fingerprint region (1300–900 cm-1).
ACKNOWLEDGMENTS
This research was supported in part by the International Marketing
Program for Agriculture Commodities and Trade (IMPACT), Washington
State University; Department of Nutrition and Food Technology, The University of Jordan, Amman, Jordan; USDA NRI Competitive Grants Program
(NRICGP) Project No. 2000–01617 and Project No. 2000–01191. We are
grateful to Dr. Peter Griffiths, Dr. Gary D. Knerr and Dr. Lacey Averett from
University of Idaho, Dr. Dong-Hyun Kang, Dr. Joong-Han Shin and Mr. Peter
Gray from Washington State University for their technical support.
DISCRIMINATION OF BACTERIAL GROWTH PHASES BY FT-IR
87
REFERENCES
AKERLUND, T., NORDSTROM, K. and BERNANDER, R. 1995. Analysis
of cell size and DNA content in exponentially growing and stationaryphase batch cultures of Escherichia coli. J. Bacteriol. 177, 6791–
6797.
AL-HOLY, M.A., LIN, M., AL-QADIRI, H. and RASCO, B.A. 2006. Discrimination of foodborne pathogens by Fourier transform infrared spectroscopy and principal component analysis. J. Rapid Methods Auto.
Microbiol. 14, 189–200.
AL-QADIRI, H.M., LIN, M., CAVINATO, A.G. and RASCO, B.A. 2006.
Fourier transform infrared spectroscopy, detection and identification of
Escherichai coli O157:H7 and Alicyclobacillus strains in apple juice. Int.
J. Food Microbiol. 111, 73–80.
ATLAS, R.M. 1988. Microbiology Fundamentals and Applications, 2nd Ed.,
pp. 106–110, Macmillan Publishing Company, New York.
BEATTIE, S.H., HOLT, C., HIRST, D. and WILLIAMS, A.G. 1998. Discrimination among Bacillus cereus, B. mycoides and B. thuringiensis and
some other species of the genus Bacillus by Fourier transform infrared
spectroscopy. FEMS Microbiol. Lett. 164, 201–206.
CHOO-SMITH, L.P., MAQUELIN, K., VREESWIJK, T.V., BRUINING,
H.A., PUPPELS, G.J., NGO THI, N.A., KRISCHNER, C., NAUMANN,
D., AMI, D., VILLA, A.M. ET AL. 2001. Investigating microbial (micro)
colony heterogeneity by vibrational spectroscopy. Appl. Environ. Microbiol. 67, 1461–1469.
FILIP, Z. and HERMANN, S. 2001. An attempt to differentiate Pseudomonas
spp. and other soil bacteria by FT-IR spectroscopy. Eur. J. Soil Biol. 37,
137–143.
GARBUTT, J. 1997. Essentials of Food Microbiology, 1st Ed., pp. 29–37,
Arnold, London.
GARIP, S., BOZOGLU, F. and SEVERCAN, F. 2007. Differentiation of
mesophilic and thermophilic bacteria with Fourier transform infrared
spectroscopy. Appl. Spectrosc. 61, 186–192.
GOODACRE, R., TIMMINS, E.M., ROONEY, P.J., ROWLAND, J.J. and
KELL, D.B. 1996. Rapid identification of Streptococcus and Enterococcus species using diffuse reflectance-absorbance Fourier transform infrared spectroscopy and artificial neural networks. FEMS Microbiol. Lett.
140, 233–239.
GOODACRE, R., TIMMINS, E.M., BURTON, R., KADERBHAI, R.,
WOODWARD, A.M. and KELL, D.B. 1998. Rapid identification of
urinary tract infection bacteria using hyperspectral whole-organism fingerprinting and artificial neural networks. Microbiol. 144, 1157–1170.
88
H.M. AL-QADIRI ET AL.
HAMPTON, K.A., CAVINATO, A.G., MAYES, D.M., BOE, S.J. and
HOFFNAGLE, T.L. 2001–2002. Near infrared spectroscopic classification of gender and maturity in Chincook salmon (Oncorhynchus tshawytscha). East. Oregon Sci. J. XVII, 31–36.
HELM, D., LABISCHINSKI, H., SCHALLEHN, G. and NAUMANN, D.
1991. Classification and identification of bacteria by Fourier transform
infrared spectroscopy. J. Gen. Microbiol. 137, 69–79.
HOLT, C., HIRST, D., SUTHERLAND, A. and MACDONALD, F. 1995.
Discrimination of species in the genus Listeria by Fourier transform
infrared spectroscopy and canonical variate analysis. Appl. Environ.
Microbiol. 61, 377–378.
HONG, K., SUN, S., TIAN, W., CHEN, G.Q. and HUANG, W. 1999. A rapid
method for detecting bacterial polyhydroxyalkanoates in intact cells by
Fourier transform infrared spectroscopy. Appl. Microbiol. Biotechnol.
51, 523–526.
HUANG, Y., ROGERS, T.M., WENZ, M.A., CAVINATO, A.G., MAYES,
D.M., BLEDSOE, G.E. and RASCO, B.A. 2001. Detection of sodium
chloride in cured salmon roe by SW-IR spectroscopy. J. Agric. Food
Chem. 49, 4161–4167.
HUANG, Y., CAVINATO, A.G., MAYES, D.M., BLEDSOE, G.E. and
RASCO, B.A. 2002. Non-destructive prediction of moisture and sodium
chloride in cold smoked Atlantic salmon (Salmo salar). J. Food Sci. 67,
2543–2547.
KANSIZ, M., HERAUD, P., WOOD, B., BURDEN, F., BEARDALL, J. and
MCNAUGHTON, D. 1999. Fourier transform infrared microspectroscopy and chemometrics as a tool for the discrimination of cyanobacterial
strains. Phytochem. 52, 407–417.
KUMMERLE, M., SCHERER, S. and SEILER, H. 1998. Rapid and reliable
identification of food-borne yeasts by Fourier-transform infrared spectroscopy. Appl. Environ. Microbiol. 64, 2207–2214.
LIN, M., CAVINATO, A.G., HUANG, Y. and RASCO, B.A. 2003. Predicting
sodium chloride content in commercial king (Oncorhynchus tshawytscha)
and chum (O. keta) hot smoked salmon fillet portions by short-wavelength
near-infrared (SW-NIR) spectroscopy. Food Res. Int. 36, 761–766.
LIN, M., AL-HOLY, M., AL-QADIRI, H., KANG, D-H., CAVINATO, A.G.,
HUANG, Y. and RASCO, B.A. 2004. Discrimination of intact and injured
Listeria monocytogenes by Fourier transform infrared spectroscopy and
principal component analysis. J. Agric. Food Chem. 52, 5769–5772.
LIN, M., AL-HOLY, M., CHANG, S-S., HUANG, Y., CAVINATO, A.G.,
KANG, D-H. and RASCO, B.A. 2005. Rapid discrimination of Alicyclobacillus strains in apple juice by Fourier transform infrared spectroscopy.
Int. J. Food Microbiol. (in press).
DISCRIMINATION OF BACTERIAL GROWTH PHASES BY FT-IR
89
MARTENS, H. and NAES, T. 1989. Multivariate Calibration. John Wiley and
Sons, New York.
MOSSOBA, M., AL-KHALDI, S., JACOBSON, A., CROWE, L. and FRY, F.
2003. Application of a disposable transparent filtration membrane to
the infrared spectroscopic discrimination among bacterial species. J.
Microbiol. Methods 55, 311–314.
NILSEN, H., ESAIASSEN, M., HEIA, K. and SIGERNES, F. 2002. Visible/
near-infrared spectroscopy: A new tool for the evaluation of fish freshness. J. Food Sci. 67, 1821–1826.
OBERREUTER, H., SEILER, H. and SCHERER, S. 2002. Identification of
coryneform bacteria and related taxa by Fourier-transform infrared (FTIR) spectroscopy. Int. J. Syst.. Evol. Microbiol. 52, 91–100.
OUST, A., MORETRO, T., KIRSCHNER, C., NARVHUS, J.A. and
KOHLER, A. 2004. FT-IR spectroscopy for identification of closely
related lactobacilli. J. Microbiol. Methods 59, 149–162.
PEDONE, E., BARTOLUCCI, S., ROSSI, M., PIERFEDERIC, F.M., SCIRE,
A., CACCIAMAN, T. and TANFAN, F. 2003. Structural and thermal
stability analysis of Escherichia coli and Alicyclobacillus acidocaldarius
thioredoxin revealed a molten globule-like state in thermal denaturation
pathway of the proteins: An infrared spectroscopic study. Biochem. J.
373, 875–883.
RODRIGUEZ-SAONA, L.E., KHAMBATY, F.M., FRY, F.S. and CALVEY,
E.M. 2001. Rapid detection and identification of bacterial strains by
Fourier transform near-infrared spectroscopy. J. Agric. Food Chem. 49,
574–579.
RODRIGUEZ-SAONA, L.E., KHAMBATY, F.M., FRY, F.S., DUBOIS, J.
and CALVEY, E.M. 2004. Detection and identification of bacteria in a
juice matrix with Fourier transform-near infrared spectroscopy and multivariate analysis. J. Food Prot. 67, 2555–2559.
SANDT, C., SOCKALINGUM, G.D., AUBERT, D., LEPAN, H., LEPOUSE,
C., JAUSSAUD, M., LEON, A., PINON, J.M., MANFAIT, M. and
TOUBAS, D. 2003. Use of Fourier-transform infrared spectroscopy for
typing of Candida albicans strains isolated in intensive care units. J. Clin.
Microbiol. 41, 954–959.
SCHMITT, J. and FLEMMING, H-C. 1998. FTIR-spectroscopy in microbial
and material analysis. Int. Biodeterior. Biodegrad. 41, 1–11.
WENNING, M., SEILER, H. and SCHERER, S. 2002. Fourier-transform
infrared microspectroscopy, a novel and rapid tool for identification of
yeasts. Appl. Environ. Microbial. 68, 4717–4721.
WINDER, C.L., CARR, E., GOODACRE, R. and SEVIOUR, R. 2004. The
rapid identification of Acinetobacter species using Fourier transform
infrared spectroscopy. J. Appl. Microbiol. 96, 328–339.