Culture Independent Raman Spectroscopic Identification of Urinary

Article pubs.acs.org/ac

Culture Independent Raman Spectroscopic Identification of Urinary Tract Infection Pathogens: A Proof of Principle Study Sandra Kloß,† Bernd Kampe,† Svea Sachse,‡ Petra Rösch,† Eberhard Straube,‡ Wolfgang Pfister,‡ Michael Kiehntopf,§ and Jürgen Popp*,†,∥ †

Institute Institute § Institute ∥ Institute ‡

of of of of

Physical Chemistry and Abbe Center of Photonics, University of Jena, Helmholtzweg 4, D-07743 Jena, Germany Medical Microbiology, Jena University Hospital, Erlanger Allee 101, D-07747 Jena, Germany Clinical Chemistry and Laboratory Diagnostics, Jena University Hospital, Erlanger Allee 101, D-07747 Jena, Germany Photonic Technology, Albert-Einstein-Straße 9, D-07745 Jena, Germany

S Supporting Information *

ABSTRACT: Urinary tract infection (UTI) is a very common infection. Up to every second woman will experience at least one UTI episode during her lifetime. The gold standard for identifying the infectious microorganisms is the urine culture. However, culture methods are time-consuming and need at least 24 h until the results are available. Here, we report about a culture independent identification procedure by using Raman microspectroscopy in combination with innovative chemometrics. We investigated, for the first time directly, urine samples by Raman microspectroscopy on a single-cell level. In a first step, a database of eleven important UTI bacterial species, which were grown in sterile filtered urine, was built up. A support vector machine (SVM) was used to generate a statistical model, which allows a classification of this data set with an accuracy of 92% on a species level. This model was afterward used to identify infected urine samples of ten patients directly without a preceding culture step. Thereby, we were able to determine the predominant bacterial species (seven Escherichia coli and three Enterococcus faecalis) for all ten patient samples. These results demonstrate that Raman microspectroscopy in combination with support vector machines allow an identification of important UTI bacteria within two hours without the need of a culture step.

U

MS)9−11 have also been explored and have found their way into clinical analysis. Recently, bacterial identification approaches using 1H nuclear magnetic resonance (1H NMR) spectroscopy12 and microcalorimetry13 were also reported. Within the last few years vibrational spectroscopic methods especially, which are label-free and nondestructive, like infrared (IR)14−20 and Raman spectroscopy in combination with powerful statistical data evaluation procedures, have shown their great potential to rapidly identify bacteria.21−26 Also, epidemiological studies using Raman spectroscopy are possible.27−30 Raman microspectroscopy (i.e., the combination of Raman spectroscopy with conventional light microscopy) in combination with chemometrical methods has been successfully applied to classify bacteria on the species level with high accuracy.21,22,31 For the identification of a certain species, a reference database is required.32,33 In particular, Raman microspectroscopy is able to investigate single bacterial cells,

rinary tract infection (UTI) is considered to be among the most common bacterial infections. Statistically, one out of three women has an UTI episode already at the age of 24 years, making an antimicrobial treatment necessary.1 Fourty to fifty percent of women will experience at least one UTI episode during her lifetime.1,2 UTIs account for more than 40% of all nosocomial infections in Germany.3 While the majority of these infections proceed without any complications, in some cases, serious progress of UTIs like bacteremia, sepsis, and even death are observed.2,4 In particular, for these severely progressing infections, a fast and reliable identification of the causing pathogens is of utmost importance. Currently, besides clinical examination, urine cultures including resistogram of isolated bacteria, which needs at least 24 h, are the gold standard to diagnose UTI.5 Until the result of the urine culture is available, an initial antimicrobial therapy against common UTI-causing microorganisms is normally started. However, this procedure could lead to rising antibiotic resistances due to selection pressure.6 To shorten the time until results are available, alternative bacterial classification methods like polymerase chain reaction (PCR)-based techniques7,8 and matrix-assisted laser desorption time-of-flight mass spectrometry (MALDI-TOF © 2013 American Chemical Society

Received: June 17, 2013 Accepted: September 6, 2013 Published: September 6, 2013 9610

dx.doi.org/10.1021/ac401806f | Anal. Chem. 2013, 85, 9610−9616

Analytical Chemistry

Article

occurrence of an inhibitory zone was determined (Table 3, column 4). For the Raman spectroscopic measurements, 1 mL of the urine was directly centrifuged at 10000g and the supernatant was discarded. Afterward, the pellet was washed twice with PBS similar to the cultured samples. The resulting pellet was suspended in 1 mL of autoclaved deionized water, and 4 μL of the suspension was spread on the nickel foil and allowed to dry at room temperature before the Raman microspectroscopic investigation took place. Raman spectra were only taken of particles which were in the size of single bacterial cells. An example of the image processing is given in Figure S1 of the Supporting Information. Spectroscopic Instrumentation. All Raman spectroscopic measurements were performed with the Raman microscope BioParticleExplorer (MicrobioID 0.5, RapID, Berlin, Germany). A frequency-doubled solid-state Nd:YAG diode pumped laser (LCM-S-111, Laser-Export Company Ltd., Moscow, Russia) at 532 nm was used for excitation. The laser beam was focused with a 100× magnification objective (MPLFLN 100×, NA: 0.9, Olympus Corporation, Tokyo, Japan) on the sample with a laser power of approximately 7 mW. The backscattered light was focused on a single stage monochromator (HE 532, Horiba Jobin Yvon, Munich, Germany) which is equipped with a 920 lines/ mm grating and collected with a thermoelectrically cooled CCD camera (DV401A-BV, Andor Technology, Belfast, Northern Ireland). The spectral resolution was about 10 cm−1. For each bacterial cell, two consecutive Raman spectra were measured at the same position, which were afterward combined for spike removal (see also Figure S1 of the Supporting Information). Integration times between 6 and 30 s for a single bacterial cell, within a wavenumber range of 3319−70 cm−1, were chosen. Data Preprocessing and Chemometrical Analysis. All preprocessing was done using the “R” software package43 with inhouse developed scripts.44 The first step consisted in removing the spectral background, using a method based on the SNIP clipping algorithm.45 For this, we used a fourth-order clipping filter. Since with the BioParticleExplorer, two consecutive spectra of one cell are always measured, spikes were eliminated afterward by a robust variant of the upper-bound spectrum algorithm.46 The single-cell Raman spectra were then wavenumber-calibrated by using an acetaminophen Raman spectrum measured at the same day as the reference.47,48 The wavenumber regions between 3100 and 2650 cm−1 and 1750−450 cm−1 were used for the chemometrical analysis. Finally, all spectra were vectornormalized. As has been shown previously, support vector machines (SVMs) are especially suited to classify and identify different bacterial species.33,49 SVMs belong to the group of maximum margin classifiers and will efficiently find the optimal solution for given parameters. A classification model was built using the all-pairs approach for SVMs with a linear kernel and a cost factor of 2.50 This model was afterward used to predict the independent test data set and the patient urine samples. Mean Raman spectra were calculated using the preprocessed, vector-normalized single-cell spectra of all classification batches of one species.

which could be isolated from different matrices, making culture steps unnecessary.33−38 Up to now, Raman spectroscopy has only been employed to identify UTI pathogens for cultured samples, including E. coli, K. pneumoniae, and Proteus spp.39,40 or additional Enterococcus spp.41 The preceding culture step of these Raman studies marginalizes the time benefit of Raman spectroscopy as compared to classical culture methods. The aim of the study presented in the following is to build up a reference database of single-cell Raman spectra of in urine-cultured UTI-causing bacteria, which allows a subsequent identification of bacteria directly in patient urine samples without a culture step.

■

MATERIAL AND METHODS The study was approved by the ethics committee of the Friedrich-Schiller University, Jena, Germany. Bacterial Culture. For the construction of a reference database, the following bacterial strains were used: Enterococcus faecalis (E. faecalis) DSM 20478, Enterococcus faecium (E. faecium) DSM 20477, Staphylococcus epidermidis (S. epidermidis) DSM 20044, Staphylococcus haemolyticus (S. haemolyticus) DSM 20263, Staphylococcus hominis (S. hominis) DSM 20328, Staphylococcus saprophyticus (S. saprophyticus) DSM 20229, Staphylococcus aureus (S. aureus) ATCC 43300, Escherichia coli (E. coli) DSM 10806, E. coli ATCC 35218, Klebsiella pneumoniae (K. pneumoniae) ATCC 700603, Pseudomonas aeruginosa (P. aeruginosa) ATCC 27853, and Proteus mirabilis (P. mirabilis) DSM 4479. All strains were provided by the Institute of Medical Microbiology, University of Jena, and were originally purchased from the German Collection of Microorganisms and Cell Cultures (DSMZ) and the American Type Culture Collection (ATCC). The bacteria were cultured on Columbia blood agar (CBA, Oxoid) plates at 37 °C for 24 h. Five milliliter sterile filtered urine (syringe filter 0.8/0.2 μm pore size, Pall Corporation, New York) was inoculated with one colony of the CBA culture and incubated at 37 °C for 24 h. One milliliter of the urine culture was centrifuged for 5 min at 10000g (Eppendorf MiniSpin plus, Eppendorf AG, Hamburg, Germany), and the supernatant was discarded. The resulting pellet was resuspended in 1 mL of phosphate-buffered saline [PBS, prepared in-house, composition per 1000 mL: 1.44 g Na2HPO4 × 2 H2O (Roth), 8 g NaCl, 0.2 g KH2PO4, and 0.2 g KCl (all purchased from Merck)] and centrifuged at 10000g for 5 min. This washing step was done twice. Afterward the pellet was suspended in 1 mL autoclaved deionized water, and 4 μL of this suspension was spread on a nickel foil and allowed to dry at room temperature. For each strain, at least five independent batches were prepared. Direct Patient Urine Preprocessing. Ten urine samples from patients with a UTI were included in this study; the samples and their origins are listed in Table 3 (columns 1−2). Only samples with at least 104 colony forming unit (CFU)/mL were used for the study because these cell counts are significant for UTI.42 Also, samples with less bacterial load (down to 103 CFU/ sample) are possible, if they are of interest for some special cases.34 The containing bacterial species in the patient urine samples were also typed according to their biochemical properties by using Vitek 2 (Biomerieux, Marcy l’Etoile, France) (Table 3, column 3). To determine possible growth inhibitors within the urine samples, an agar diffusion test on Columbia agar (Oxoid) inoculated with 106 CFU/mL of Bacillus subtilis ATCC 6633 has been performed. The test plates were inoculated with one droplet of the urine sample and after 24 h incubation at 37 °C, the

■

RESULTS AND DISCUSSION Raman spectra of single bacterial cells can be seen as a characteristic spectroscopic fingerprint of the investigated cell. The diameter of the used laser spot is in the same dimension as the diameter of a single bacterial cell (around 1 μm). Therefore, 9611



Article

Table 1. Observed Raman Bands and Their Tentative Assignment wavenumber (cm−1) 3059 2935 1665

1606 1573 1451 1334

1310

1241 1127 1099 1004 781

748 723 a

Figure 1. Mean spectra of the investigated species: (a) E. faecalis (429 spectra), (b) E. faecium (256 spectra), (c) S. epidermidis (227 spectra), (d) S. haemolyticus (225 spectra), (e) S. hominis (207 spectra), (f) S. saprophyticus (237 spectra), (g) S. aureus (285 spectra), (h) E. coli (360 spectra), (i) K. pneumoniae (233 spectra), (j) P. aeruginosa (249 spectra), and (k) P. mirabilis (244 spectra).

tentative band assignmenta ν(CH) olefinic ν(CH2) asymmetric ν(CH3) symmetric amide I ν(CC) nucleic acids (CC) ring vibrations of phenylalanine, tyrosine ring vibrations of guanine and adenine cytochrome c δ(CH2/CH3) δ(CH2) (in proteins) ring vibrations of guanine and adenine tryptophan CH2/CH3 twisting, wagging, bending modes of lipids cytochrome c amide III ν(PO2−) asymmetric (DNA bases) cytochrome c phenylalanine ν(C−N) ring breathing modes of phenylalanine ring breathing modes of cytosine, uracil and thymine O−P−O backbone of DNA DNA cytochrome c ρ(CH2)

reference (cm−1) 306022 293522 293555 166638 166755 166355 1603/160555 157522 158356 1440−146022 145038 133755 133755 1339/133755 1313−130755 131156 124555 124355 112856 110455 109955 100455 786−78055 78555 74855 75056 72222

ν: stretching vibration, δ: deformation vibration, ρ: rocking vibration.

unknown bacterial composition and culture conditions. The variability of the data set in the form of the standard deviations of the mean spectra is available in Figures S2 and S3 of the Supporting Information. In Figure 1, the mean Raman spectra of all bacterial species in the CH-stretching (3100−2650 cm−1) and the fingerprint wavenumber region (1750−450 cm−1) are shown. Each mean spectrum was constructed from at least 200 single-cell Raman spectra. The most prominent band at around 2935 cm−1 can be assigned to symmetric and asymmetric CH2 and CH3 stretching vibrations.22,55 Other CH vibration bands can be found at 3059, 1451, 1334, and 723 cm−1.38 Such CH structures are common for proteins, lipids, and carbohydrates. Table 1 provides an overview of the observed Raman bands and their tentative band assignments. Besides the CH-vibrations, additional protein bands can be found at 1665, 1606, 1334, 1241, 1099, and 1004 cm−1. Typical bands for DNA structures are at 1665, 1573, 1334, 1241, 781, and 748 cm−1. The enhanced bands at 1573, 1127, 1310, and 748 cm−1, which can be found for some species like for example, Staphylococci (c−g), are due to a higher cytochrome c content which exhibits an electronic absorption in the range of the Raman excitation wavelength and therefore experiences a resonance Raman enhancement.56 Figure 1 shows that the differences between the Raman spectra of the single bacterial cells of different species in spiked urine

the Raman spectrum represents a superposition of the molecular vibrations of all cellular compounds.51 The most important step in bacterial identification by means of Raman microspectroscopy is the buildup of a database of reference Raman spectra. In order to establish such a database, we used eleven different UTI relevant bacterial species, namely, E. coli (Ecol), K. pneumoniae (Kpne), P. aeruginosa (Paer), P. mirabilis (Pmir), E. faecalis (Efca), E. faecium (Efci), S. aureus (Saur), S. epidermidis (Sepi), S. hominis (Shom), S. haemolyticus (Shae), and S. saprophyticus (Ssap). These bacterial species were cultured in sterile filtered urine samples (at least four different batches per species, each in different urine) followed by recording Raman spectra of single bacterial cells. Different batches for each species were used to increase the variability of the data set. This variability is important because the Raman spectra of single bacterial cells, even of the same bacterial species, differ according to the growth stage22,52 and the nutrition situation (composition of culture medium).22,53,54 Since during single-cell measurements no standardization process can take place, possible variation has to be included in the database, in order to be able to identify bacteria grown in urine samples with 9612



Article

Table 2. (A) SVM Results for Classification Model and (B) Identification of an Independent Test Data Set of Samples from Spiked Urine (A) true classified asa

Efca

Efci

Sepi

Shae

Shom

Efca Efci Sepi Shae Shom Ssap Saur Ecol Kpne Paer Pmir

422 5 0 0 0 0 0 2 0 0 0

14 242 0 0 0 0 0 0 0 0 0

1 0 208 2 0 9 5 1 0 0 1

0 0 2 208 6 2 7 0 0 0 0

0 0 5 4 177 15 4 0 0 0 2

Ssap

Saur

Ecol

Kpne

Paer

Pmir

Sensb

Specb

2 0 10 4 19 194 3 2 0 0 3 (B)

1 0 7 3 7 3 264 0 0 0 0

0 0 2 0 0 1 0 323 26 1 7

1 0 0 0 0 0 0 26 203 2 1

0 0 0 0 0 0 0 2 3 244 0

0 0 0 0 0 0 0 8 2 1 233

98.4 94.5 91.6 92.4 85.5 81.9 92.6 89.7 87.1 98.0 95.5

99.2 99.8 99.0 99.5 98.8 98.8 99.2 98.3 98.8 99.8 99.4

true a

identified as

Efca

Efci

Sepi

Shae

Shom

Ssap

Saur

Ecol

Kpne

Paer

Pmir

Sensb

Specb

Efca Efci Sepi Shea Shom Ssap Saur Ecol Kpne Paer Pmir

53 0 0 0 0 0 0 0 0 0 0

1 49 0 0 0 0 0 0 0 0 0

0 0 41 0 1 1 1 3 0 0 0

1 0 0 50 0 0 1 0 0 0 0

0 0 0 0 44 5 0 0 0 0 0

0 0 0 0 0 29 1 0 0 0 0

0 0 0 0 0 0 50 0 0 0 0

0 0 1 0 0 0 0 30 4 0 2

0 0 0 0 0 0 0 0 40 0 0

0 0 0 0 0 0 0 0 0 39 0

0 0 0 0 0 0 0 3 0 0 64

100 98.0 87.2 96.2 89.8 96.7 100 81.1 100 100 95.5

99.5 100 99.8 100 99.8 98.7 99.3 98.7 99.1 100 99.5

a Efca = E. faecalis, Efci = E. faecium, Sepi = S. epidermidis, Shae = S. haemolyticus, Shom = S. hominis, Ssap = S. saprophyticus, Saur = S. aureus, Ecol = E. coli, Kpne = K. pneumoniae, Paer = P. aeruginosa, Pmir = P. mirabilis. bSens = sensitivity (%). Spec = specificity (%).

Figure 2. Evaluation of patient urine specimens: assignment of measured single-cell Raman spectra based on the before built SVM model.

samples are rather subtle and often not visible by the naked eye, making the application of chemometrical methods to distinguish between the bacterial species necessary. As described in the previous section, we used a SVM with a linear kernel and a cost factor of 2 to build a statistical model which is able to classify/

identify the bacterial species. The performance of the SVM was assessed by a 10-fold cross validation. The resulting confusion table is shown in Table 2A. In total, 2718 out of 2952 spectra were correctly classified, resulting in an accuracy of 92.1%. The best results could be achieved for E. 9613



Article

Table 3. Result for the Evaluation of Direct Patient Urine Samples without a Preceding Culture Step sample

origina

real speciesb

inhibitor test

predicted asc

spectrad

abundance (%)

EC1 EC2 EC3 EC4 EC5 EC6 EC7 EF1 EF2 EF3

urine ur. cath. urine urine urine ur. cath. urine urine blad. punct. ur. cath.

E. coli E. coli E. coli E. coli E. coli E. coli E. coli E. faecalis E. faecalis E. faecalis

positive negative positive negative negative positive negative positive negative positive

E. coli E. coli E. coli E. coli E. coli E. coli E. coli E. faecalis E. faecalis E. faecalis

40/60 37/51 35/51 41/50 25/31 40/49 43/49 46/50 45/46 50/54

66.7 72.5 68.6 82.0 80.6 81.6 87.8 92.0 97.8 92.6

a

Ur. cath. = permanent urinary catheder. blad. punct. = bladder punction. bDetermined by Vitek 2. cBy Raman microspectroscopy. dNumber of correct identified spectra/number of measured spectra.

faecalis, where 422 out of 429 single-cell spectra were classified correctly. This yields a sensitivity of 98.4% together with a specificity of 99.2%. The specificities for the whole data set range between 98.8% in cases of S. hominis, S. saprophyticus, and K. pneumoniae and 99.8% for E. faecium and P. aeruginosa. Sensitivities vary between 81.9% for S. saprophyticus and 98.4% for E. faecalis. Most of the misclassified S. saprophyticus spectra were classified as other coagulase-negative Staphylococci (33 out of 43). Incorrect assignments also occurred between E. coli and K. pneumoniae: 26 E. coli spectra out of 350 were classified false positive as K. pneumoniae and 26 K. pneumoniae spectra out of 233 were false positive as E. coli. This could be explained by their close genetic relationship within Enterobacteriaceae. To confirm the predictive capacity of the classification model, an independent batch for each species was cultured in sterile, filtered urine, and single-cell Raman spectra were measured. The before-built SVM model was used to predict this independent test data set to check for overfitting. In Table 2B, the correspondent results are summarized: 95.1% of the spectra were correctly assigned (489 out of 514 spectra) while the sensitivities range between 81.1% for E. coli and 100% for E. faecalis, S. aureus, K. pneumoniae, and P. aeruginosa. Specificities between 98.7% for S. saprophyticus as well as E. coli and 100% for E. faecium, S. haemolyticus, and P. aeruginosa were reached. These results suggest that the built SVM model can be used to identify independent test data. Evaluation of Direct Urine Specimens. Finally, the abovedescribed SVM model was tested on ten real patient urine samples to provide a proof of principle that Raman microspectroscopy can be applied to directly identify single bacterial cells out of infected urine samples without the need for timeconsuming culture steps. Thereby, the challenge consists in identifying direct patient samples where the following three conditions are unknown as compared to the reference database: (1) the composition of the patients’ urine (matrix), (2) the dwell time in the urine (“culture” time), and (3) the bacterial strains might be different as compared to the ones used to build up the database. To cope with these challenges, it was necessary to implement as much as possible, variations into the classification reference data set. The bacteria were isolated by using the same centrifugation and washing steps as for the bacteria used to build up the reference database described in the preceding section. Figure 2 displays the SVM identification results of the real urine patient samples by showing the assignments of the singlecell spectra for each urine sample. For each of the urine samples,

one predominant germ could be determined. Seven samples were predicted to contain E. coli and three E. faecalis. This agrees with the microbiological analysis of the samples (see Table 3). At least 66.7% of the spectra of each specimen were identified correctly. The abundance ranges from 66.7 to 97.8%, whereby it was higher for E. faecalis isolates (92.0 to 97.8%) than for E. coli samples (66.7 to 87.8%). In five out of the ten samples, the bacterial growth was suppressed as proven by the inhibitor test (Table 3); this means that the patients were pretreated with antibiotics. In accordance with our results, substances causing suppression of bacterial growth in these samples seem to have only a little influence on the identification abundance. The two E. faecalis isolates with a positive inhibitor test were identified with an abundance of 92.0% (EF1) and 92.6% (EF3), respectively. The E. faecalis without an inhibitor in the urine (EF2) was identified with an abundance of 97.8%. The chemometrical evaluation for E. coli provides for the three samples with a positive inhibitor test, an abundance between 66.7% (EC1) and 81.6% (EC6). For samples without growth inhibitor, the abundance ranges from 72.5% (EC2) to 87.8% (EC7). We set the threshold for a correct identification of a whole sample to an abundance of 65% to account for false positive classifications, which are included in the reference database. In addition, possible contaminations of the urine samples due to, for example, skin flora are not unusual.

■

CONCLUSIONS

Here, we report to the best of our knowledge about the first single-cell typing method for a direct identification of UTI pathogens. We accomplished a Raman microspectroscopic identification of bacterial species in real patient urine samples without any time-consuming culture steps. Prior to this identification, a reference database needs to be established. For that purpose, eleven UTI relevant species were used to build up a reference database. The applicability of the reference model to evaluate infected real urine samples was tested for ten patients’ urine samples without a preceding culture step. For all ten samples, the correct species could be determined by Raman microspectroscopy. It is very promising for further studies, that it was also possible to identify the containing species in urine samples from antibiotic-pretreated patients. Since Raman spectroscopy can be applied on single bacterial cells, it is possible also to investigate samples with mixed bacterial species. In addition, the investigation of patient samples is performed without prior culture, therefore, even minor sample contaminations due to sampling might be identified aside the main pathogens present in the urine. In further studies, the amount 9614



Article

(15) Goodacre, R.; Timmins, Ã . a. M.; Burton, R.; Kaderbhai, N.; Woodward, A. M.; Kell, D. B.; Rooney, P. J. Microbiology 1998, 144, 1157−1170. (16) Sahu, R. K.; Mordechai, S.; Pesakhov, S.; Dagan, R.; Porat, N. Biopolymers 2006, 83, 434−442. (17) Bosch, A.; Miñań , A.; Vescina, C.; Degrossi, J.; Gatti, B.; Montanaro, P.; Messina, M.; Franco, M.; Vay, C.; Schmitt, J.; Naumann, D.; Yantorno, O. J. Clin. Microbiol. 2008, 46, 2535−2546. (18) Maquelin, K.; Kirschner, C.; Choo-Smith, L. P.; van den Braak, N.; Endtz, H. P.; Naumann, D.; Puppels, G. J. J. Microbiol. Methods 2002, 51, 255−271. (19) Kirschner, C.; Maquelin, K.; Pina, P.; Ngo Thi, N. A.; ChooSmith, L.-P.; Sockalingum, G. D.; Sandt, C.; Ami, D.; Orsini, F.; Doglia, S. M.; Allouch, P.; Mainfait, M.; Puppels, G. J.; Naumann, D. J. Clin. Microbiol. 2001, 39, 1763−1770. (20) Saraiva, R. G.; Lopes, J. A.; Machado, J.; Gameiro, P.; Feio, M. J. J. Biophotonics DOI 10.1002/jbio.201200131. (21) Rösch, P.; Harz, M.; Schmitt, M.; Peschke, K.-D.; Ronneberger, O.; Burkhardt, H.; Motzkus, H.-W.; Lankers, M.; Hofer, S.; Thiele, H.; Popp, J. Appl. Environ. Microbiol. 2005, 71, 1626−1637. (22) Harz, M.; Rösch, P.; Peschke, K. D.; Ronneberger, O.; Burkhardt, H.; Popp, J. Analyst 2005, 130, 1543−1550. (23) Lu, X.; Huang, Q.; Miller, W. G.; Aston, D. E.; Xu, J.; Xue, F.; Zhang, H.; Rasco, B. A.; Wang, S.; Konkel, M. E. J. Clin. Microbiol. 2012, 50, 2932−2946. (24) Walter, A.; März, A.; Schumacher, W.; Rösch, P.; Popp, J. Lab Chip 2011, 11, 1013−1021. (25) Geßner, R.; Winter, C.; Rösch, P.; Schmitt, M.; Petry, R.; Kiefer, W.; Lankers, M.; Popp, J. ChemPhysChem 2004, 5, 1159−1170. (26) Chan, J. W. J. Biophotonics 2013, 6, 36−48. (27) Willemse-Erix, D. F. M.; Scholtes-Timmerman, M. J.; Jachtenberg, J.-W.; van Leeuwen, W. B.; Horst-Kreft, D.; Bakker Schut, T. C.; Deurenberg, R. H.; Puppels, G. J.; van Belkum, A.; Vos, M. C.; Maquelin, K. J. Clin. Microbiol. 2009, 47, 652−659. (28) Willemse-Erix, H. F. M.; Jachtenberg, J.; Barutci, H.; Puppels, G. J.; van Belkum, A.; Vos, M. C.; Maquelin, K. J. Clin. Microbiol. 2010, 48, 736−740. (29) Willemse-Erix, D.; Bakker-Schut, T.; Slagboom-Bax, F.; Jachtenberg, J.-w.; Lemmens-den Toom, N.; Papagiannitsis, C. C.; Kuntaman, K.; Puppels, G.; van Belkum, A.; Severin, J. t. A.; Goessens, W.; Maquelin, K. J. Clin. Microbiol. 2012, 50, 1370−1375. (30) te Witt, R.; Vaessen, N.; Melles, D. C.; Lekkerkerk, W. S. N.; van der Zwaan, E. A. E.; Zandijk, W. H. A.; Severin, J. A.; Vos, M. C. J. Clin. Microbiol. 2013, 51, 1434−1438. (31) Schmid, U.; Rösch, P.; Krause, M.; Harz, M.; Popp, J.; Baumann, K. Chemom. Intell. Lab. Syst. 2009, 96, 159−171. (32) Stöckel, S.; Schumacher, W.; Meisel, S.; Elschner, M.; Rösch, P.; Popp, J. Appl. Environ. Microbiol. 2010, 76, 2895−2907. (33) Stöckel, S.; Meisel, S.; Elschner, M.; Rösch, P.; Popp, J. Anal. Chem. 2012, 84, 9873−9880. (34) Meisel, S.; Stöckel, S.; Elschner, M.; Rösch, P.; Popp, J. Analyst 2011, 136, 4997−5005. (35) Stöckel, S.; Meisel, S.; Elschner, M.; Rösch, P.; Popp, J. Angew. Chem., Int. Ed. 2012, 51, 5339−5342. (36) Krause, M.; Radt, B.; Rösch, P.; Popp, J. J. Raman Spectrosc. 2007, 38, 369−372. (37) Krause, M.; Rösch, P.; Radt, B.; Popp, J. Anal. Chem. 2008, 80, 8568−8575. (38) Harz, M.; Kiehntopf, M.; Stöckel, S.; Rösch, P.; Straube, E.; Deufel, T.; Popp, J. J. Biophotonics 2009, 2, 70−80. (39) Kastanos, E. K.; Kyriakides, A.; Hadjigeorgiou, K.; Pitris, C. J. Raman Spectrosc. 2009, 41, 958−963. (40) Kastanos, E.; Kyriakides, A.; Hadjigeorgiou, K.; Pitris, C. Int. J. Spectrosc. 2012, 2012, 195317. (41) Jarvis, R. M.; Goodacre, R. FEMS Microbiol. Lett. 2004, 232, 127− 132. (42) Kass, E. H. J. Urol. (N.Y., NY, U.S.) 2002, 167, 1016−1020. (43) R Development Core Team. R Foundation for Statistical Computing: Vienna, Austria, 2011.

and species of contaminating bacterial cells has to be investigated both by Raman spectroscopy and other single-cell typing methods, additional to the gold-standard urine culture. The results of this study are very encouraging for a fast and cultureindependent identification of UTI pathogens directly out of urine specimen by using Raman microspectroscopy. Restrictively, it has to be mentioned that the herein presented proof-ofprinciple was only done with the most important UTI species E. coli and E. faecalis, since these are the most common UTI pathogens. For other species, the applicability has to be shown in further studies. Nevertheless, the whole identification can be done within approximately two hours after the specimen collection, which is an enormous time benefit in comparison to the gold standard urine culture.

■

ASSOCIATED CONTENT

* Supporting Information S

Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Tel: +49-3641-948320 and +49-3641-206300. Fax: +49-3641-948302. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS Funding of the research project FastDiagnosis (13N11350) from the Federal Ministry of Education and Research, Germany (BMBF) is gratefully acknowledged.

■

REFERENCES

(1) Foxman, B.; Barlow, R.; D’Arcy, H.; Gillespie, B.; Sobel, J. D. Annals of Epidemiology 2000, 10, 509−515. (2) Foxman, B. Am. J. Med. 2002, 113, 5−13. (3) Gastmeier, P.; Kampf, G.; Wischnewski, N.; Hauer, T.; Schulgen, G.; Schumacher, M.; Daschner, F.; Rüden, H. Journal of Hospital Infection 1998, 38, 37−49. (4) Al-Hasan, M. N.; Eckel-Passow, J. E.; Baddour, L. M. J. Infect. 2010, 60, 278−285. (5) Schmiemann, G.; Kniehl, E.; Gebhardt, K.; Matejczyk, M. M.; Hummers-Pradier, E. Deutsches Ä rzteblatt International 2010, 107, 361− 367. (6) Tenover, F. C. Am. J. Med. 2006, 119, S3−10 and discussion S62− 70. (7) Lu, J.; Yu, R.; Yan, Y.; Zhang, J.; Ren, X. J. Microbiol. Meth. 2011, 86, 78−81. (8) Lehmann, L. E.; Hauser, S.; Malinka, T.; Klaschik, S.; Weber, S. U.; Schewe, J.-C.; Stüber, F.; Book, M. PLoS ONE 2011, 6, e17146. (9) Ferreira, L.; Sánchez-Juanes, F.; Muñoz-Bellido, J. L.; GonzálezBuitrago, J. M. Clin. Microbiol. Infec. 2011, 17, 1007−1012. (10) Ferreira, L.; Sánchez-Juanes, F.; González-Á vila, M.; CembreroFuciños, D.; Herrero-Hernández, A.; González-Buitrago, J. M.; MuñozBellido, J. L. J. Clin. Microbiol. 2010, 48, 2110−2115. (11) Wang, X. H.; Zhang, G.; Fan, Y. Y.; Yang, X.; Sui, W. J.; Lu, X. X. J. Microbiol. Methods 2013, 92, 231−235. (12) Gupta, A.; Dwivedi, M.; Mahdi, A. A.; Khetrapal, C. L.; Bhandari, M. J. Proteome Res. 2012, 11, 1844−1854. (13) Bonkat, G.; Braissant, O.; Widmer, A. F.; Frei, R.; Rieken, M.; Wyler, S.; Gasser, T. C.; Wirz, D.; Daniels, A. U.; Bachmann, A. BJU Int. 2012, 110, 892−7. (14) Sousa, C.; Grosso, F.; Meirinhos-Soares, L.; Peixe, L.; Lopes, J. J. Biophotonics 10.1002/jbio.201200075. 9615



Article

(44) Bocklitz, T.; Walter, A.; Hartmann, K.; Rösch, P.; Popp, J. Anal. Chim. Acta 2011, 704, 47−56. (45) Morhác, M.; Kliman, J.; Matoušek, V.; Veselský, M.; Turzo, I. Nucl. Instrum. Methods Phys. Res., Sect. A 1997, 401, 113−132. (46) Zhang, D.; Jallad, K. N.; Ben-Amotz, D. Appl. Spectrosc. 2001, 55, 1523−1531. (47) Dörfer, T.; Bocklitz, T.; Tarcea, N.; Schmitt, M.; Popp, J. Z. Phys. Chem. 2011, 225, 753−764. (48) Bocklitz, T.; Putsche, M.; Stüber, C.; Käs, J.; Niendorf, A.; Rösch, P.; Popp, J. J. Raman Spectrosc. 2009, 40, 1759−1765. (49) Meisel, S.; Stöckel, S.; Elschner, M.; Melzer, F.; Rösch, P.; Popp, J. Appl. Environ. Microbiol. 2012, 78, 5575−5583. (50) Chang, C.-C.; Lin, C.-J. ACM Transactions on Intelligent Systems and Technology 2011, 2, 1−27. (51) Harz, M.; Rösch, P.; Popp, J. Cytometry, Part A 2009, 75A, 104− 113. (52) Neugebauer, U.; Schmid, U.; Baumann, K.; Ziebuhr, W.; Kozitskaya, S.; Deckert, V.; Schmitt, M.; Popp, J. ChemPhysChem 2007, 8, 124−137. (53) Walter, A.; Kuhri, S.; Reinicke, M.; Bocklitz, T.; Schumacher, W.; Rösch, P.; Merten, D.; Büchel, G.; Kothe, E.; Popp, J. J. Raman Spectrosc. 2012, 43, 1058−1064. (54) Huang, W.; Bailey, M.; Thompson, I.; Whiteley, A.; Spiers, A. Microb. Ecol. 2007, 53, 414−425. (55) Movasaghi, Z.; Rehman, S.; Rehman, I. U. Appl. Spectrosc. Rev. 2007, 42, 493−541. (56) Pätzold, R.; Keuntje, M.; Theophile, K.; Müller, J.; Mielcarek, E.; Ngezahayo, A.; Ahlften, A. A. V. J. Microbiol. Methods 2008, 72, 241− 248.

9616


Culture Independent Raman Spectroscopic Identification of Urinary

Recommend Documents