Identification and discrimination of bacteria using Fourier transform infrared spectroscopy

Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 116 (2013) 478–484

Contents lists available at ScienceDirect

Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy journal homepage: www.elsevier.com/locate/saa

Identification and discrimination of bacteria using Fourier transform infrared spectroscopy Jyoti Prakash Maity a,⇑, Sandeep Kar b, Chao-Ming Lin c, Chen-Yen Chen a, Young-Fo Chang a, Jiin-Shuh Jean b, Thomas R. Kulp d a

Department of Earth and Environmental Sciences, National Chung Cheng University, Ming-Shung, Chiayi County 62102, Taiwan Department of Earth Sciences, National Cheng Kung University, Tainan City 70101, Taiwan Department of Electronic Engineering, Hsiuping Institute of Technology, Dali City, Taichung 41280, Taiwan d Department of Geological Sciences and Environmental Studies, Binghamton University, 4400 Vestal Parkway E., Binghamton, NY 13902, United States b c

h i g h l i g h t s

g r a p h i c a l a b s t r a c t

 FTIR is applicable for bacterial

classification and identification.  Bacterial discrimination by proteins

specific bands at 1339, 1382 and 1096 cm 1.  Cell wall constituent at 2987, 2971 and 2900 cm 1 found in both B. flexus and OS1.  Amide bands (I/II/III) exhibits OS1 and B. flexus were closely related, except OS2.  Specific fingerprint at 1096 (OS2), 1339 (OS1) and 1382 cm 1 (B. flexusATCC49095).

a r t i c l e

i n f o

Article history: Received 21 December 2012 Received in revised form 17 June 2013 Accepted 22 July 2013 Available online 6 August 2013 Keywords: FTIR 16S rRNA B. flexus S. maltophilia Identification Discrimination

a b s t r a c t Bacterial spectra were obtained in the wavenumber range of 4000–600 cm 1 using FTIR spectroscopy. FTIR spectral patterns were analyzed and matched with 16S-rRNA signatures of bacterial strains OS1 and OS2, isolated from oil sludge. Specific spectral bands obtained from OS1 (FJ226761), reference strainBacillus flexus (ATCC 49095), OS2 (FJ215874) and reference strain Stenotrophomonas maltophilia (ATCC 19861) respectively, suggested that OS1 and ATCC 49095 were closely related whereas OS2 was different. The bands probably represent groups of proteins and lipids of specific bacteria. Separate peaks found in B. flexus were similar to those of OS1. The S. maltophilia (ATCC 19861) and OS2 exhibited a similar peak at 3272 cm 1. Amide bands (I, II and III) exhibited that OS1 and B. flexus were closely related, but were different from OS2. In the fingerprint region, peak at 1096 cm 1 and 1360 cm 1 exhibited the specific fingerprints of OS2 and reference strain S. maltophilia (ATCC 19861), respectively. The specific fingerprint signature was found at 1339 cm 1 for OS1 and at 1382 cm 1 for B. flexus ATCC 49095, allowing these two strains of B. flexus to be differentiated. This spectral signature originated from phospholipid and RNA components of the cell. Principle components analysis (PCA) of spectral regions exhibited with distinct sample clusters between Bacillus flexus (ATCC 49095), S. maltophilia (ATCC 19861), OS1 and OS2 in amide and fingerprint region. Ó 2013 Elsevier B.V. All rights reserved.

Introduction ⇑ Corresponding author. Tel.: +886 5 2720411x61217; fax: +886 5 2720807. E-mail address: [email protected] (J.P. Maity). 1386-1425/$ - see front matter Ó 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.saa.2013.07.062

Detection and characterization of microorganisms by Fourier transform infrared spectroscopy (FTIR) technique promises to be

J.P. Maity et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 116 (2013) 478–484

of great value because of the method’s inherent sensitivity, nondestructive nature, small sample size, rapidity, simplicity, and the potential for complete computerization [1–11]. Spectroscopic techniques provide a wealth of specific qualitative and quantitative information about a given organism [9,12,13]. The infrared spectrum of any compound, including those found in living cells, is known to give a unique ‘fingerprint’ [14]. Fourier Transformation Infrared (FTIR) spectroscopy provides a non-destructive, fast, easy to use, and highly sensitive method for microbial analysis [9,15]. This technique permits non-destructive chemical characterization of living cells producing unique and reproducible biochemical spectra for different cell types [16–19]. Five spectral windows have been defined and are used for the differentiation of bacteria species [7,20,21]. Previous authors reported that the spectral region of 3000–2800 cm 1 was commonly assumed to be dominated by fatty acid related compounds; the region of 1700–1500 cm 1 by carbonyl residual of proteins (Amide I and II); the carboxylic groups of peptide (AC@O), at about 1650 cm 1 and ACAOA at 1500 cm 1; free amino acids, and polysaccharides in the region of (1450–1400 cm 1), and the region of 1250–1200 cm 1 by RNA/ DNA (1245 cm 1 and 1080 cm 1 for mas PO2 and mas PO2 ) and phospholipids [22]. The region below 1500 cm 1 was referred as the fingerprint region (IUPAC), and contains information significant to strain-specific discrimination. Several reports have been published using FTIR spectroscopy as means of rapid identification of microorganisms [8,9,22–24], however there are insufficient reports on the identification of Bacillus flexus and S. maltophilia using the FTIR spectroscopy technique. In this study, FTIR spectroscopy was used for detection of unique spectral parameters representing biochemical differences among different bacteria. We compare the identification of bacteria using FTIR techniques to 16S-rRNA gene sequence analysis. Materials and methods Isolation and cultivation of bacteria Oil degrading bacteria were isolated from oil sludge of a sewage treatment plant (China Petroleum Refinery Company, Kaohsiung, Taiwan) using a dilution plate method in nutrient agar media (Sigma–Aldrich, Germany) under aerobic conditions at 30 °C [25]. The Isolated bacteria were grow in nutrient broth media (Peptone 5 g/L, beef extract 3 g/L, agar 15 g/L, and Milli-Q water, pH-7 at 25 °C). Cultivated bacteria was split into duplicate samples that were used for identification by FTIR spectroscopy and by 16S rRNA, respectively. FTIR spectroscopy: sample preparation, spectral data collection and measurement of bacteria spectra Samples (OS1, OS2, reference B. flexus (ATCC 49095) and S. maltophilia (ATCC 19861)) were prepared for the FTIR study as described by earlier workers [22,24]. Small amounts of bacterial colonies were suspended in 2 mL of saline solution and pelleted by centrifugation at 1000g for 2 min. The pellet was then re-suspended with 20 lL of saline solution and one drop of suspension was placed on a zinc selenide crystal (4 cm  1 cm) and air dried for 5 min in a laminar flow hood at room temperature before being examined by FTIR spectroscopy. The FTIR spectra was recorded in transmission mode from individual cells at 4 cm 1 resolution with Happ-Genzel apodization, using a FTIR spectrometer (ThermoFisher-Nicolet (TFN), Magna-IR 860) fitted with an IR microscope system (TFN, Continuum) and a liquid–nitrogen-cooled mercury– cadmium–tellurite (MCT) detector, including KBr beamsplitter and a 32 magnification IR objective. The spectra were obtained

479

in the wavenumber range of 4000–600 cm 1. For each bacterial sample, the spectrum was taken as an average of nine different measurements at various sites of the sample (n = 9). The peak position, baseline corrections and smoothing were automatically performed by peak resolve (Thermo-Fisher-Nicolet, OMNIC 7.1). The FTIR spectra (4000–600 cm 1) of bacteria were analyzed using Spectrum software (PerkinElmer V5.0) and KnowItAll software (Bio-Rad IR/NIR Edition) [26]. This software provided identification of the various functional groups and possible assignments of different bacteria. Identification by 16S rRNA gene sequence analysis: DNA isolation, PCR (Polymerase Chain Reaction) amplification, nucleotide sequencing and accession numbers Each strain of bacteria (OS1, OS2, reference B. flexus (ATCC 49095) and S. maltophilia (ATCC 19861)) was cultivated in nutrient broth for identification by 16S rRNA gene sequencing and the FTIR technique. Genomic DNA was extracted from bacteria using a DNA/ RNA extraction kit (Viogene, Taipei, Taiwan). The 16S rRNA gene region of DNA was amplified by PCR using a pair of forward and reverse primers [27,28]. GenBank accession numbers were assigned from nucleotide sequences (99% identity) (OS1: FJ226761; B. flexus: ATCC 49095; OS2: FJ215874, and S. maltophilia: ATCC 19861) (S. maltophilia (FJ009381) is the synonym of Pseudomonas beteli) (http://www.ncbi.nlm.nih.gov). Statistical analysis of FTIR spectral data A significant level of difference between experimental bacteria was determined by factor analysis (n = 9), which was performed with the help of the PCA using STATISTICA 5.1. Also, this software was used to perform Hierarchical Cluster Analysis (HCA) on the FTIR region of the different bacteria. Results and discussion Identification of bacteria using two different techniques: FTIR spectroscopy and 16S-rRNA gene sequence Different bacterial strains were investigated by FTIR spectroscopy to determine the specific spectroscopic biomarkers useful for identification and discrimination of selected bacterial species. The peak position, baseline corrections and smoothing were automatically performed by peak resolve for a profile of each peak from the same bacterial strain. FTIR spectra and possible assignments of each isolate (OS1 and OS2) were compared with B. flexus reference strain (ATCC 49095) and S. maltophilia (ATCC 19861) (Figs. 1–7; Table 1). The 16S-rRNA gene sequence of the OS1 bacterial isolate aligned with that of the reference strain of B. flexus with 99% nucleotide similarity, suggesting that the OS1 (FJ226761) isolate was closely related to B. flexus, whereas OS2 (FJ215874) was closely related to S. maltophilia (ATCC 19861) (Fig. 2). During FTIR analysis, B. flexus (ATCC 49095) delivered a well differentiated spectra from OS2 but provided an almost identical spectra to the OS1 strain, with minimal differences in intensity of the individual bands (Figs. 1 and 3), nucleic acid and cell wall constituents (Figs. 1, 4 and 5), proteinaceous structure, and the fingerprint region (Figs. 6 and 7) The bacterial spectra in the region of 1300–1500 cm 1 showed a unique absorption band at 1405 and 1313 cm 1. The bands at 1394 and 1313 cm 1 probably represent absorption that is specific to B. flexus (Fig. 3). On the other hand, the band at 1394 cm 1 was observed in both OS1 and B. flexus (ATCC 49095), indicating that OS1 is very similar to B. flexus. The bands at 1094 cm 1 were specific for OS2 and S. maltophilia (ATCC 19861). By contrast, the bands

480

J.P. Maity et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 116 (2013) 478–484

Fig. 1. FTIR spectra in the region of 3850–800 cm 1 from three different bacteria, reference strain B. flexus (ATCC 49095), strain OS1 (FJ226761), strain OS2 (FJ215874) and reference strain S. maltophilia (ATCC 19861). Results are nine different (n = 9) and separate experiments for each sample.

Fig. 2. Phylogenic relationship between the bacterial reference B. flexus (ATCC 49095), strain OS1 (FJ226761), strain OS2 (FJ215874) and reference strain S. maltophilia (ATCC 19861). The calculation of evolutionary distance and classification of phylogenetic relationship were determined using the Jukes-Cantor distance and neighbor-joining algorithm.

at 1382 cm 1 and 1360 cm 1 suggest that they are different strains, and probably represent groups of proteins and lipids specific to a particular bacterium. Identification by spectral band of cell wall constituent, lipids and proteins Many peaks were observed in CAH stretching region at 3000– 2800 cm 1 (Fig. 4) [29,30]. The spectral analysis of vibration modes in this range included msCH2 (2863–2843 cm 1), msCH3 (2882– 2862 cm 1), masCH2 (2936–2916 cm 1), and masCH3 (2972– 2952 cm 1) [10,23,24,29]. The methylene (ACH2) group and methyl (ACH3) group dominated in lipids and proteins, respec-

Fig. 3. Bacterial signature of the FTIR spectra in the region of 1500–1300 cm strain OS2 (FJ215874) and reference strain S. maltophilia (ATCC 19861).

1

tively [29]. In B. flexus ATCC 49095 and S. maltophilia (ATCC 19861), the bands of cell wall constituents found at 2987, 2971 and 2900 cm 1 were comparable to those of OS1 and OS2. However, a different band at 2882 cm 1 appeared for msCH3, which was specific to OS2. The presence of double bonds within the fatty acid carbon backbone can shift the unsaturated @CAH or @CH2 stretch absorption at 2987 cm 1. Cell proteins are typically indicated by a number of amide bands. The spectral region of the present study from 1665 to 1200 cm 1 (Figs. 1 and 6) was assumed to be dominated by proteins [24,31]. The bands in the range of 1680–1600 cm 1 were attributed to protein amide-I [22–24,32]. The bands marked as ms PO2 and mas PO2 , located at 1087 and 1241 cm 1, are the symmetric (ms) and the asymmetric (mas) PO2 stretching vibrations of DNA (amide-III contributes at 1241 cm 1), whereas lipids contributed at 1740 cm 1 originating from ester C@O group [22–24,31]. The intensity and shape of the band was notably dissimilar at 1450 cm 1 for both OS1 and B. flexus (ATCC 49095). However, a peak at 1313 cm 1 was demonstrated to correspond to amide-III for all of the bacteria (possibly representing a bacterial signature), whereas unique amide-III peaks at 1230, 1241 and 1250 cm 1 suggested that OS1 is a different strain from OS2 (Fig. 6). The peaks at 1338 cm 1 and 1360 cm 1 represented the difference between the strain of B. flexus and S. maltophilia (Figs. 6 and 7). According to the previous report, the amide bands dominate at 1658/1656 cm 1 (amide-I), 1535 cm 1 (amide-II) and 1240/1235 cm 1 (amide-III) in Pseudomonas spp. [23]. The NAH deformation of amides associated with proteins of amide II band at 1540 cm 1 and amide I band at 1650 cm 1 were observed from Pseudomonas aeruginosa [29]. The present results supports the findings of Filip et al. that the peak positions at 1660 cm 1, 1540 cm 1 and 1235 cm 1 were assigned to amide I, amide II and amide III in B. subtilis, respectively [24].

from three different bacteria, reference strain B. flexus (ATCC 49095), strain OS1 (FJ226761),

J.P. Maity et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 116 (2013) 478–484

Fig. 4. Cell wall constituents of bacteria in the FTIR spectral region of 3000–2800 cm strain OS2 (FJ215874) and reference strain S. maltophilia (ATCC 19861).

Fig. 5. Nucleic acid in the FTIR spectral region of 3360–3200 cm (FJ215874) and reference strain S. maltophilia (ATCC 19861).

1

1

481

from different bacteria, reference strain B. flexus (ATCC 49095), strain OS1 (FJ226761),

from three different bacterial reference strains; B. flexus (ATCC 49095), strain OS1 (FJ226761), strain OS2

Fig. 6. Proteinaceous structure of bacteria in the FTIR spectral region of 1665–1200 cm (FJ226761), strain OS2 (FJ215874) and reference strain S. maltophilia (ATCC 19861).

Similarly, Erukhimovitch et al. reported that the bands at 1655 and 1544 cm 1 were attributed to amide I and II in B. megaterium, Escherichia coli and Pseudomonas stutzeri [22]. Identification by spectral band of nucleic acid The absorption bands at 3307 cm 1, 3297 cm 1 and 3282 cm 1 were found to be similar among all bacteria in this study; however

1

from three different bacteria; reference strain B. flexus (ATCC 49095), strain OS1

these bands appeared more weakly in case of OS1 compared to the reference bacteria B. flexus (Fig. 5). The present results reveal that one specific absorption band at 3292 cm 1 occurred in OS2 and S. maltophilia (ATCC 19861). This band could be specific for OS2 and S. maltophilia (ATCC 19861), and indicates that the nucleic acid structure of reference strain B. flexus is different from OS2 (FJ215874). The band at 3315 cm 1 was observed to be specific for reference S. maltophilia (ATCC 19861). Parallel findings of

482

J.P. Maity et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 116 (2013) 478–484

Fig. 7. Fingerprint of bacteria in the FTIR spectral region of 1430–800 cm and reference strain S. maltophilia (ATCC 19861).

1

from three different strains; B. flexus (ATCC 49095), strain OS1 (FJ226761), strain OS2 (FJ215874)

Table 1 FT-IR absorption bands and their subsequent possible assignments of different bacteria. Band (cm

1

Possible assignments

Microorganism

References

3000 3240–3200 3440–3400 3240–3100 3400–3300 2960–2927

)

NH2 stretching in adenine, cytosine, quinine; H-bonded OH groups

Bacillus subtilis Strain OS1 (FJ226761) B. flexus (ATCC 49095) Strain OS2 (FJ215874) S. maltophilia (ATCC 19861) Bacillus subtilis

[12,23,24] This study

2999–2977 2999–2977

CAH stretching in aliphatics of cell wall (fatty acid and carbohydrate)

Strain OS2 (FJ215874) Strain OS1 (FJ226761)

[23] This study

3100–2900 2924–2875 3100–2900 2924–2875

Aliphatic CAH stretching (fatty acid)

B. flexus (ATCC 49095) Strain OS2 (FJ215874) S. maltophilia (ATCC 19861) Strain OS1 (FJ226761)

[23,24] This study

1660 1670–1645 1670–1618

NH2 bending, C@O, C@N stretching (amide-I band)

Bacillus subtilis S. maltophilia (ATCC 19861) B. flexus (ATCC 49095)

[12] This study

1680–1630 1680–1630 1680–1630

C@O stretching (amide-I band)

Strain OS2 (FJ215874) S. maltophilia (ATCC 19861) Strain OS1 (FJ226761)

[23,32] This study

1638–1618 1605–1590

NH2 stretching, C@O, C@N stretching (amide-II band)

Strain OS2 (FJ215874) Strain OS1 (FJ226761)

[23,32] This study

1540 1570–1515

Amide-II band

Bacillus subtilis S. maltophilia (ATCC 19861)

[12,23,24]

1305–1200 1305–1200 1305–1200

Amide III, combination of CAN stretching and NAH bending

B. flexus (ATCC 49095) Strain OS2 (FJ215874) Strain OS1 (FJ226761)

[23,32] This study

1235

Amide-III band

Bacillus subtilis

[12,24,32]

1265–1200 1265–1200 1265–1200

CAN and CAO stretching amide IV

Strain OS2 (FJ215874) Strain OS1 (FJ226761) B. flexus (ATCC 49095)

This study

1050–970

PO12 very strong stretching (glycopeptides, ribose)

Strain OS2 (FJ215874)

[23] This study

Strain OS1 (FJ226761)

1050–970 1025–870 1025–870

PAOAP stretching (phospholipids, ribose phosphate chain pyrophosphate)

earlier researchers [23,24,33] demonstrated that H-bonded OH groups and NH2 stretching of adenine guanine and cytosine were responsible for an absorption band at 3300 cm 1 in Pseudomonas spp. Identification by spectral band of fingerprint region In the fingerprint region, the cell wall carbohydrate peak at 1066 cm 1 was obtained from OS1 and B. flexus ATCC 49095 (Fig. 7). Our results correlate with the earlier report on B. subtilis, in which a phosphoric ester group (PAOAC) was found to absorb

Strain OS2 (FJ215874) Strain OS1 (FJ226761)

[23] This study

at 1030 cm 1 [24]. Also, similar peaks at 1075, 1056, 891, 879 and 871 cm 1 were found from OS1 and B. flexus. However, a peak at 1096 cm 1 exhibited a specific fingerprint for OS2 and S. maltophilia (ATCC 19861). The spectral region of 1100–1000 cm 1 was assigned as a PO2 stretching vibration of phosphate groups and a CAO stretching vibration in Pseudomonas spp. [23]. Tsuboi reported that PO2 stretching vibrations of phosphate groups and a CAO stretching vibrations are attributed to the nucleic acid and ribose-phosphate chain-backbone vibration between 967 cm 1 and 760 cm 1 [34]. The specific fingerprint signature by which two strains of B. flexus can be differentiated was found at 1339 cm 1

J.P. Maity et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 116 (2013) 478–484

for OS1 and at 1382 cm 1 for B. flexus ATCC 49095 [Fig. 7]. This peak originated from phospholipids and RNA in the cell. The present results indicate that FTIR is highly reproducible and specific at the strain level for bacteria. Thus, it allows accurate differentiation of closely related bacterial species. In comparison to conventional genotypic and phenotypic methods, FTIR was found to be accurate for a wider range of bacterial species and more economical for routine analysis [35]. Statistical discrimination of essential biomacromolecule using FTIR spectra The spectra were normalized before statistical discrimination analysis. The PCA was conducted on the second derivative FTIR spectra over the entire wavenumber range. Factor analysis was

483

performed using spectral regions of cell wall constituents (3000– 2800 cm 1), lipids (1500–1300 cm 1), amide-III (1300– 1200 cm 1), carbohydrates (1150–1000 cm 1) and the fingerprint region (1500–860 cm 1) for all strains (OS1, OS2, B. flexus (ATCC 49095) and S. maltophilia (ATCC 19861)) (Figs. 8a–e). All the strains possess their two representative factors of spectra together. The clustering of cell wall constituents (3000–2800 cm 1) (Fig. 8a) and lipid (1500–1300 cm 1) (Fig. 8b) of Bacillus flexus ATCC 49095) and OS1 partially overlapped, which confirms the high degree of similarity between these two strains, whereas clustering results for the strain OS2 and S. maltophilia (ATCC 19861) was noticeably different from others. Factor analysis of cell wall constituents, lipids, proteins, cell wall carbohydrates and the fingerprint region, and their dendrogram, all showed that characteristics of the spectra for OS1 (FJ226761) are quite distinct

Fig. 8. Factor analysis of B. flexus reference strain ATCC 49095, strain OS1 (FJ226761), strain OS2 (FJ215874) and reference strain S. maltophilia (ATCC 19861) in the FTIR spectral regions of (a) cell wall constituent (3000–2800 cm 1), (b) lipid of bacteria (1500–1300 cm 1), (c) bacterial protein amide-III (1300–1200 cm 1), (d) cell wall carbohydrate (1150–1000 cm 1), and (e) fingerprint region (910–860 cm 1). (f) Hierarchical classification of the FTIR region.

484

J.P. Maity et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 116 (2013) 478–484

from that of the OS2 (FJ215874) and S. maltophilia (ATCC 19861) but similar to B. flexus ATCC 49095. The dendrogram of lipids, amide and cell wall carbohydrates of B. flexus and OS1 suggests that they are different strains within the same species (Fig. 8f). Clear segregations with distinct sample clusters were observed between Bacillus flexus (ATCC 49095), S. maltophilia (ATCC 19861), OS1 and OS2 in amide (1300–1200 cm 1) (Fig. 8c) and fingerprint region (1500–860 cm 1) (Fig. 8e). The study shows the usefulness of identification of bacteria with spectral reference peaks of reference bacteria, and discriminates between two different bacteria of their genus at the species and strain level. This is needed to make a database for each bacterium. Conclusions Statistically and experimentally, FTIR spectroscopy can be used to identify and differentiate among different organisms with specific spectral signatures. The peak at 1313 cm 1 was revealed to correspond to amide-III for all of the bacteria representing possible bacterial signature. A peak at 1096 cm 1 represents a specific fingerprint for OS2. The peak at 1360 cm 1 exhibited the specific fingerprints of reference S. maltophilia (ATCC 19861). The specific fingerprint signature was found at 1339 cm 1 for OS1 and 1382 cm 1 for B. flexus ATCC 49095, allowing the two strains of B. flexus to be differentiated. This spectral signature originated from cell phospholipids and RNA. Moreover, the clear segregations (PCA analysis) with distinct sample clusters were revealed from Bacillus flexus (ATCC 49095), S. maltophilia (ATCC 19861), OS1 and OS2 in amide (1300–1200 cm 1) and fingerprint region (1500– 860 cm 1). The results exhibit that representative specific spectral biomarkers are applicable for identification and discrimination of different bacteria and that this technique should be widely applicable to other areas of microbial research. Acknowledgment The authors wish to thank the National Science Council of Taiwan for partial financial support. References [1] B. Dziuba, B. Nalepa, Food Technol. Biotechnol. 50 (2012) 399–405. [2] M. Wenning, N.R. Büchl, S. Scherer, J. Biophotonics. 8–9 (2010) 493–505. [3] A. Bosch, A. Minan, C. Vescina, J. Degrossi, B. Gatti, P. Montanaro, M. Messina, M. Franco, C. Vay, J. Schmitt, D. Naumann, O. Yantorno, J. Clinic. Microbiol. 46 (2008) 2535–2546. [4] A. Bombalska, M. Mularczyk-Oliwa, M. Kwas´ny, M. Włodarski, M. Kaliszewski, K. Kopczyn´ski, M. Szpakowska, E.A. Trafny, Spectrochim. Acta A 78 (2011) 1221–1226.

[5] F. Quilès, F. Humbert, A. Delille, Spectrochim. Acta A 75 (2010) 610–616. [6] E. San-Blas, N. Cubillán, M. Guerra, E. Portillo, I. Esteves, Spectrochim. Acta A 93 (2012) 58–62. [7] D. Naumann, D. Helm, H. Labischinski, P. Giesbrecht, in: W.H. Nelson (Ed.), The Characterization of Microorganisms by Fourier Transform Infrared Spectroscopy, VCH Publishers Inc., New York, 1991, pp. 43–96. [8] O. Preisner, J.A. Lopes, R. Guiomar, J. Machado, J.C. Menezes, Anal. Bioanal. Chem. 387 (2007) 1739–1748. [9] C.A. Rebuffo-Scheer, J. Dietrich, M. Wenning, S. Scherer, Anal. Bioanal. Chem. 390 (2008). 1629-1625. [10] P. Whittaker, M.M. Mossoba, S. Al-Khaldi, F.S. Fry, V.C. Dunkel, B.D. Tall, M.P. Yurawecz, J. Microbiol. Methods 55 (2003) 709–716. [11] K. Maquelin, C. Kirschner, L.-P. Choo-Smith, N.A. Ngo-Thi, D. Naumann, G.J. Puppels, in: P.R. Griffiths (Ed.), Vibrational Spectroscopic Studies of Microorganisms, Handbook of Vibrational Spectroscopy, John Wiley & Sons, Ltd., Chichester, United Kingdom, 2002. [12] N.A. Ngo Thi, D. Naumann, Anal. Bioanal. Chem. 387 (2007) 1769–1777. [13] K. Maquelin, C. Kirschner, L.P. Choo-Smith, N.A. Ngo-Thi, T. van Vreeswijk, M. Stämmler, H.P. Endtz, H.A. Bruining, D. Naumann, G.J. Puppels, J. Clinical. Microbiol. 41 (2003) 324–329. [14] M.A. Al-Holy, M. Lin, A.G. Cavinato, B.A. Rasco, Food Microbiol. 23 (2006) 162– 168. [15] B. Dziuba, A. Babuchowski, D. Nałe˛cz, M. Niklewicz, Int. Dairy J. 17 (2007) 183– 189. [16] D. McNaughton, Aust. Biochemist. 36 (2005) 55–58. [17] A.A. Kamnev, Spectroscopy 22 (2008) 83–95. [18] G. Fischer, S. Braun, R. Thissen, W. Dott, J. Microbiol. Methods 64 (2006) 63–77. [19] L. Mariey, J.P. Signolle, C. Amiel, J. Travert, Vib. Spectrosc. 26 (2001) 151–159. [20] M. Kümmerle, S. Scherer, H. Seiler, Appl. Environ. Microbiol. 64 (1998). 22072204. [21] J. Schmitt, H.-C. Flemming, in: E. Heitz, W. Sand, H.-C. Flemming (Eds.), FTIR Spectroscopy, Springer Verlag, Heidelberg, 1996, p. 143. [22] V. Erukhimovitch, V. Pavlov, M. Talyshinsky, Y. Souprun, M. Huleihel, J. Pharma, Biomed. Anal. 37 (2005) 1105–1108. [23] Z. Filip, S. Hermann, Eur. J. Soil Biol. 37 (2001) 137–143. [24] Z. Filip, S. Hermann, J. Kubat, Res. 159 (2004) 257. [25] R.C. Dubey, D.K. Maheshwari, Practical Microbiology, first ed., S Chand and Company Ltd, New Delhi, India, 2005. [26] C.-C. Liu, J.P. Maity, J.-S. Jean, Z. Li, S. Kar, O. Sracek, H.-J. Yang, C.-Y. Chen, A.H.M.S. Reza, J. Bundschuh, C.-Y. Lee, J. Geochem. Explor. 128 (2013) 62– 71. [27] H.P. Goorissen, H.T. Boschker, A.J. Stams, T.A. Hansen, Int. J. Syst. Evol. Microbiol. 53 (2003) 1223–1229. [28] J.P. Maity, S. Kar, J.-H. Liu, J.-S. Jean, C.-Y. Chen, J. Bundschuh, S.C. Santra, C.-C. Liu, J. Environ. Sci. Health A Tox. Hazard. Subst. Environ. Eng. 46 (2011) 1239– 1246. [29] H.M. Al-Qadiri, M.A. Al-Holy, M. Lin, N.I. Alami, A.G. Cavinato, B.A. Rasco, J. Agric. Food Chem. 54 (2006) 5749–5754. [30] N. Buensanteai, K. Thumanu, M. Sompong, D. Athinuwat, S. Prathuangwong, Afri. J. Microbiol. Res. 6 (2012) 603–610. [31] T. Udelhovent, D. Naumann, J. Schmitt., Appl. Spectrosc. 54 (2000) 1471–1479. [32] J. Kong, S. YU, Acta Biochim. Biophysic. Sinica 39 (2007) 549–559. [33] F.S. Parker, Applications of Infrared Spectroscopy in Biochemistry, Biology and Medicine, Adam Hilger, London, 1971. [34] M. Tsuboi, Appl. Spectrosc. Rev. 3 (1970) 45–90. [35] C. Kirschner, K. Maquelin, P. Pina, N.A. Ngo Thi, L.-P. Choo-Smith, G.D. Sockalingum, C. Sandt, D. Ami, F. Orsini, S.M. Doglia, P. Allouch, M. Mainfait, G.J. Puppels, D. Naumann, Classification and identification of Enterococci: a comparative phenotypic, genotypic, and vibrational spectroscopic study, J. Clinic. Microbiol. 39 (2001) 1763–1770.