Docstoc

Simultaneous quantitative detection of 12 pathogens using

Document Sample
Simultaneous quantitative detection of 12 pathogens using Powered By Docstoc
					Electrophoresis 2010, 31, 2405–2410                                                                                         2405

Gi Won Shin2Ã                         Research Article
Hee Sung Hwang2Ã
Mi-Hwa Oh3
Junsang Doh2,4ÃÃ                      Simultaneous quantitative detection of
Gyoo Yeol Jung1,2
                                      12 pathogens using high-resolution
1
 Department of Chemical
 Engineering, Pohang University
                                      CE-SSCP
 of Science and Technology,
 Pohang, Korea
2
 School of Interdisciplinary
 Bioscience and Bioengineering,       Several methods based on screening for a 16S ribosomal RNA gene marker have been
 Pohang University of Science         developed for rapid and sensitive detection of pathogenic microorganisms. One such
 and Technology, Pohang, Korea
3                                     method, CE-based SSCP (CE-SSCP), has enormous potential because the technique can
 National Institute of Animal
 Science, Rural Development           separate sequence variants using a simple procedure. However, conventional CE-SSCP
 Administration, Suwon, Korea         systems have limited resolution and cannot separate most 16S ribosomal RNA gene-
4
 Department of Mechanical             specific markers unless combined with additional modification steps. A high-resolution
 Engineering, Pohang University
 of Science and Technology,           CE-SSCP system that uses a poly(ethyleneoxide)-poly(propyleneoxide)-poly(ethylene-
 Pohang, Republic of Korea            oxide) triblock copolymer matrix was recently developed and shown to effectively
                                      separate highly similar PCR products. In this study, we developed a method based on a
                                      high-resolution CE-SSCP system using a poly(ethyleneoxide)-poly(propyleneoxide)-
Received February 17, 2010
                                      poly(ethyleneoxide) triblock copolymer that is capable of simultaneous and quantitative
Revised April 5, 2010
Accepted April 6, 2010                detection of 12 clinically important pathogens. Pathogen markers were amplified by PCR
                                      using universal primers and separated by CE-SSCP; each marker peak was well separated
                                      at baseline and showed a characteristic mobility, allowing easy identification of patho-
                                      gens. A series of experiments using different amounts of genomic pathogen DNA
                                      showed that the method had a limit of detection of 0.31–1.56 pg and a dynamic range of
                                      approximately 102. These results indicate that high-resolution CE-SSCP systems have
                                      considerable potential in the clinical diagnosis of bacteria-induced diseases.

                                      Keywords:
                                      16S rRNA gene / CE-SSCP / High-resolution / Pathogen detection / Polymer
                                      matrix                                         DOI 10.1002/elps.201000091




1 Introduction                                                   of spectrally distinct fluorescent tags typically prevents
                                                                 analysis by real-time PCR of more than four target bacterial
A rapid and sensitive detection method is required for the       species per sample in an analytic run [10]. Furthermore,
accurate diagnosis and effective treatment of patients with      multiplex PCR for simultaneous detection of multiple
bacteria-induced diseases [1–4]. One existing method, PCR-       targets requires optimization for any particular target set,
based pathogen detection, uses exponential amplification to       and multiplex analysis by real-time PCR typically suffers
rapidly detect marker genes from specific pathogens, and is       from reduced quantification power because of inaccuracies
the most sensitive of the existing molecular diagnostic          and the relatively narrow dynamic range. Another major
methods [5]. Real-time PCR, in particular, is one of the most    technology, DNA microarray analysis, is also used in
accurate DNA quantification methods currently available,          microbial diagnostics and offers advantages in multiplex
and is widely used in various applications, including            applications [11–13]. However, although the availability of
pathogen diagnosis [6–9]. However, the limited availability      commercial microarray platforms has made it possible to
                                                                 obtain reproducible results, DNA microarray technology has
                                                                 limitations in the accuracy of quantification [12, 14, 15].
Correspondence: Professor Gyoo Yeol Jung, Department of          DNA microarray analysis is based on the hybridization
Chemical Engineering, Pohang University of Science and           of probes to targets; thus, nonspecific hybridization is
Technology, Pohang, Gyeongbuk 790-784, Korea                     inevitable, and differences in the melting temperatures
E-mail: gyjung@postech.ac.kr                                     of probes and targets decrease the resolution of this
Fax: 182-54-279-5528
                                                                 technology. Therefore, it is difficult to obtain an accurate
                                                                  ÃThese authors are co-first authors.
Abbreviations: gDNA, genomic DNA; PEO-PPO-PEO, poly-
                                                                 ÃÃAdditional corresponding author: Professor Junsang Doh
(ethyleneoxide)-poly(propyleneoxide)-poly(ethyleneoxide);
rRNA, ribosomal RNA                                                E-mail: jsdoh@postech.ac.kr


& 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim                                                www.electrophoresis-journal.com
2406       G. W. Shin et al.                                                                Electrophoresis 2010, 31, 2405–2410


quantification of individual species from a microbial sample      polymer matrix for separation of variable sequences originating
because target signals are not completely separable.             from the 16S rRNA gene of the 12 pathogens, we tested PEO-
Considerable recent progress has been made in overcoming         PPO-PEO triblock copolymer solutions with different polymer
this problem, but inaccurate quantification remains one of        compositions and concentrations. Under optimized conditions,
the main disadvantages of microarrays [15].                      the 16S rRNA sequence for each bacterial pathogen was
     CE-SSCP analysis is a molecular diagnostic technique        detected as a single representative peak that was resolved from
that discriminates differences in the conformations of           the variable sequences of other targets. The resulting CE-SSCP
single-stranded DNA molecules [16]. In CE-SSCP analyses,         profile was validated by additional analyses of individual
only thermal denaturation and cooling of the PCR product         targets. The sensitivity of the assay system was assessed by
is needed to form single-stranded DNAs with distinctive          measuring the limit of detection for each pathogen genomic
tertiary conformations, enabling the separation of same-         DNA (gDNA). Additionally, a series of experiments using
sized targets with different sequences. Particularly when        different concentrations of pathogen gDNAs were performed
combined with 16S ribosomal RNA (rRNA) gene-specific              to demonstrate the ability of our improved system to provide
PCR, CE-SSCP separation can provide quantitative infor-          quantitative results in a multiplex format.
mation about a microbial population [17–23]. The 16S
rRNA gene, considered the ‘‘gold standard’’ for bacterial
identification, is an ideal choice of detection marker            2 Materials and methods
because the gene contains conserved sequences that flank
a highly variable region [24]. We previously demonstrated        2.1 Target pathogenic bacteria and gDNA isolation
that CE-SSCP technology could be used for simultaneous
assay of antimicrobial activity in a microbial community         Twelve pathogenic bacteria were used as a model set
[23], and for quantitative detection of multiple pathogens       (Table 1). Eight of these strains were provided by the Korea
[25, 26]. To detect microbial strains in earlier studies, we     Collection for Type Cultures (KCTC, Daejeon, Korea) and
used the 16S rRNA gene or rRNA per se as a target. Our           were grown according to the instructions provided using
initial efforts to simultaneously quantify four bacterial        culture media purchased from BD Diagnostics (Sparks, MD,
strains were based on amplification of the V2 region of the       USA). gDNA was isolated from these strains using a
16S rRNA gene using universal primers [23]. However,             DNeasy Blood & Tissue Kit (Qiagen, Valencia, CA, USA).
separation of a larger number of amplified products from a        gDNA from Brucella abortus was a kind gift from the
greater number of strains of interest was not possible           National Veterinary Research and Quarantine Service
because the resolution of conventional CE-SSCP techniques        (NVRQS, Anyang-si, Kyeonggi-do, Korea). gDNA of the
was insufficient. Subsequent studies showed that multiplex        remaining three strains was purchased from the American
pathogen quantification was possible with conventional            Type Culture Collection (ATCC, Manassas, VA, USA) or the
CE-SSCP if strain-specific primers were used instead of the       Korea Culture Center of Microorganisms (KCCM, Seoul,
universal primer set [25, 26]. In this case, however, accurate   Korea).
quantification of individual target pathogens required an
additional singleplex quantification step [25] or a template-
tagging procedure before amplification using a common             2.2 16S rRNA gene-specific PCR
primer set [26].
     Limitations in the resolution of CE-SSCP are caused         The primer set used to prepare DNA for separation analysis
primarily by the CE polymer matrix, which is the most critical   targeted the V2 region of the 16S rRNA gene. The primer
component in CE separation. Recently, a high-resolution CE-      sequences corresponding to the conserved flanking
SSCP system was developed using poly(ethyleneoxide)-poly         segments bordering the V2 region were 50 -GGC GAA
(propyleneoxide)-poly(ethyleneoxide) (PEO-PPO-PEO) solution      CGG GTG AGT AA-30 (V2Forward) and 50 -ACT GCT GCC
as a novel CE separation medium for SSCP analysis [27, 28].      TCC CGT AG-30 (V2Reverse). Forward primers were labeled
PEO-PPO-PEO is an amphiphilic triblock copolymer contain-        at the 50 end with 6-carboxyfluorescein. Pfu polymerase PCR
ing PEO as a hydrophilic block and PPO as a hydrophobic          premix and primers were obtained from Bioneer (Daejeon,
block. The resolution of SSCP separation in the triblock         Korea). PCR was performed in a 20 mL reaction volume
copolymer solution was significantly improved compared with       containing gDNA, 10 pmol of each primer (forward and
the resolution in poly(N,N-dimethylacrylamide) (commercial       reverse primers), 0.25 mM of each dNTP, 1 U of Pfu DNA
name: POP-Conformation Analysis Polymer) solution, a             polymerase in reaction buffer. For demonstration of multi-
commonly used commercial polymer for CE-SSCP analysis.           plex quantitative detection, serial dilutions of a pathogen
Furthermore, PEO-PPO-PEO triblock copolymers have poten-         gDNA were mixed with the other targets with constant
tial advantages attributable to enhanced dynamic coating and     amount (10 or 50 pg of each gDNA). The PCR cycles
sieving abilities that result from temperature-induced micelle   consisted of initial denaturation for 4 min at 951C, followed
formation [29, 30]. In this study, we report a novel technique   by 25 cycles of denaturation for 30 s at 951C, annealing for
for multiplex detection and quantification of 12 pathogens        30 s at 551C, and extension for 30 s at 721C, then a final
using the high-resolution CE-SSCP system. To optimize the        extension at 721C for 7 min. PCR products were diluted as


& 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim                                             www.electrophoresis-journal.com
Electrophoresis 2010, 31, 2405–2410                                                                       Nucleic acids        2407

Table 1. Target pathogen informationa)

Microbial name                           Source and strain IDb)           Typical disease                        Amplicon length (bp)

Proteus mirabilis                        KCTC 1117                        Urinary tract infection                255
Streptococcus agalactiae                 KCTC 11957                       Septicemia                             263
Enterococcus faecalis                    KCTC 3511                        Nosocomial infection                   265
Streptococcus pyogenes                   KCTC 3984                        Pyogenic infection                     263
Haemophilus ducreyi                      ATCC 700724                      Chancroid                              255
Brucella abortus                         NVRQS                            Zoonotic food poisoning                238
Enterococcus faecium                     KCTC 2022                        Meningitis                             265
Enterobacter aerogenes                   KCTC 22064                       Nosocomial infection                   255
Legionella pneumophila                   ATCC 33152                       Legionellosis                          255
Vibrio vulnificus                         KCTC 2959                        Cellulitis                             255
Vibrio cholerae                          KCTC 2715                        Cholera                                255
Neisseria meningitidis                   KCCM 41562                       Meningitis                             255

a) Strains are sorted and numbered in order of increasing signal elution time.
b) KCTC, Korea Collection for Type Cultures; ATCC; NVRQS, National Veterinary Research and Quarantine Service; KCCM, Korea Culture
   Center of Microorganisms.


appropriate in nuclease-free water and used for CE-SSCP             syringe pumping time necessary to achieve complete
analysis.                                                           replacement of polymer in the capillary varied with polymer
                                                                    type and concentration. A CCD camera fitted to the ABI 310
                                                                    Genetic Analyzer was used to detect fluorescence at
2.3 Polymer matrix preparation                                      wavelengths from 525 to 650 nm. Virtual filter sets were
                                                                    optimized for the ABI PRISM dye set (6-carboxyfluorescein),
A 7.2 wt% solution of POP-Conformation Analysis Polymer             and dyes were excited with a 10 mW argon ion laser at 488
(CAP; Applied Biosystems, Foster City, CA, USA) in 1 Â              and 514 nm.
EDTA buffer with 10 wt% glycerol was used as the
conventional nondenaturing polymer matrix. Pluronic
F127 and F108 PEO-PPO-PEO triblock copolymers,                      3 Results and discussion
purchased from Sigma-Aldrich (St. Louis, MO, USA), were
dissolved in 0.7 Â EDTA buffer (Applied Biosystems) at              3.1 Multiplex detection of 12 pathogens using
various concentrations. Bubbles in the solution were                    CE-SSCP
removed by centrifugation at 7000 Â g for 30 min at 41C.
                                                                    Previously, we showed that CE-SSCP combined with 16S
                                                                    rRNA gene-specific PCR could be used to identify four
2.4 CE-SSCP analysis                                                bacterial species [23]. All four targets were represented by
                                                                    single unique peaks that showed reproducible electrophore-
CE-SSCP analysis was performed as described in our                  tic mobility based on individual sequences. However, it was
previous reports [26]. A 1 mL sample of the amplified 16S            not possible to apply the same strategy to detect a larger
rRNA gene from each species was mixed with 13.5 mL                  number of microorganisms because of the low resolution of
deionized formamide (Applied Biosystems) and 0.5 mL ROX             conventional CE-SSCP [25, 26]. Here, employing a novel
500 size standards (Applied Biosystems). The sample                 PEO-PPO-PEO triblock copolymer matrix for CE-SSCP, we
mixtures were denatured at 941C for 4 min and immediately           have achieved a very large improvement in resolution that
cooled on ice for 3 min. CE-SSCP analyses were performed            makes it feasible to use the same approach to discriminate
on an ABI Prism 310 Genetic Analyzer (Applied Biosys-               more targets. As shown in Fig. 1A, a conventional polymer
tems), configured in accordance with the manufacturer’s              matrix under optimal conditions (7.2 wt% with 10 wt%
instructions, using noncoated capillaries (47 cm  50 mm;           glycerol) was unable to separate PCR products originating
Applied Biosystems). When triblock copolymer solutions              from the 12 pathogens, because of insufficient resolution.
were used, the capillary was filled with the polymer matrix          To improve resolution, we tested various concentrations of
and allowed to stand overnight to improve resolution.               the two types of triblock copolymers of different composi-
A similar phenomenon has been observed by Wu et al. [31],           tions (Pluronic F127 and F108) in the operable viscosity
but the reason for the improvement is not clear. After the          range. In a 15 wt% solution of Pluronic F108, 12 represen-
overnight waiting, over 200 running can be performed                tative peaks were obtained (Fig. 1B). Additional analyses
without loss of resolution. Samples were electrophoresed            of individual targets showed that each of the 12 peaks
using an injection voltage of 15.0 kV, an electrophoresis           represented a unique target. Therefore, the characteristic
voltage of 15.0 kV, and a running temperature of 351C. The          elution times obtained on CE-SSCP analysis (Table 2) could


& 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim                                                   www.electrophoresis-journal.com
2408        G. W. Shin et al.                                                                     Electrophoresis 2010, 31, 2405–2410




Figure 1. CE-SSCP separation of 16S rRNA gene-specific PCR
products using (A) a conventional polymer matrix (POP-CAP)
and (B) a PEO-PPO-PEO triblock copolymer (Pluronic F108).
Elution time (x-axis) and fluorescence intensity (y-axis) are
indicated in time units and relative fluorescence units, respec-
tively. For the electropherogram of Pluronic F108 (B), the
identities of peaks obtained by additional analyses are indicated
by numbers: (1) P. mirabilis, (2) S. agalactiae, (3) E. faecalis,
(4) S. pyogenes, (5) H. ducreyi, (6) B. abortus, (7) E. faecium,
(8) E. aerogenes, (9) L. pneumophila, (10) V. vulnificus, (11)       Figure 2. Serial dilutions of L. pneumophila mixed with the
V. cholerae, and (12) N. meningitidis.                              other targets. L. pneumophila gDNA was serially diluted
                                                                    from 5 ng to 50 pg; the amount of input target is indicated.
                                                                    The amounts of input gDNA for the other targets were
Table 2. Summary of quantitative multiplex detection of             10 pg (S. agalactiae, S. pyogenes, H. ducreyi, E. faecium,
          12 pathogens                                              En. aerogenes, V. vulnificus, V. cholerae, and N. meningitidis)
                                                                    or 50 pg (P. mirabilis, E. faecalis, and B. abortus). The signal from
                                                                    the dilution target decreased with increasing degree of dilution.
Microbe           Elution Limit of      Dynamic range Correlation
                  time    detection     (pg/reaction) coefficient
                  (min)   (pg/reaction)               (R 2)
                                                                    tens-to-hundreds of cells (Table 2). This level of sensitivity is
P. mirabilis      32.8    1.56         50–5050         0.99         sufficient to allow practical detection of pathogens in a
S. agalactiae     33.9    0.63         10–5010         0.98         variety of samples.
E. faecalis       34.3    1.56         50–5050         0.99
S. pyogenes       34.5    0.63         10–5010         0.98
H. ducreyi        34.9    0.31         10–5010         0.98         3.2 Quantification by relative peak area
B. abortus        35.2    1.56         50–5050         0.99
E. faecium        35.5    0.63         10–5010         1.00
                                                                    As peak area in CE-SSCP analysis reflects the initial amount
E. aerogenes      35.7    0.63         10–5010         0.99
                                                                    of RNA or DNA input [23, 25, 32], CE-SSCP separation
L. pneumophila    36.3    0.31         50–5000         1.00
V. vulnificus      36.9    0.31         10–5010         0.98
                                                                    combined with 16S rRNA gene-specific PCR can provide
V. cholerae       37.3    1.56         10–5010         0.98         quantitative information on microbial populations. The peak
N. meningitidis   40.4    0.31         10–5010         0.97         baselines were well separated (Fig. 1B); thus, peak area could
                                                                    be accurately obtained for each target, allowing a more
                                                                    precise quantification of individual targets than is achievable
                                                                    with the conventional system. For multiplex applications
be used to differentiate the 12 pathogenic strains. Compared        such as the present system, reproducible quantification can
with our previous efforts to enhance the resolution of CE-          be obtained using an internal standard [26]. Thus, instead of
SSCP separations, the multiplex detection approach used             using absolute peak areas, peak areas relative to the standard
here was much simpler, requiring no additional steps. We            are employed to minimize assay-to-assay variation in
also tested the lower limits of multiplex detection and found       quantification. To study the correlation between fold changes
that targets could be sensitively identified in samples              of the signal and input gDNA amounts, we tested serial
containing 0.31–1.56 pg of gDNA, which is equivalent to             dilutions of Legionella pneumophila gDNA mixed with other


& 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim                                                  www.electrophoresis-journal.com
Electrophoresis 2010, 31, 2405–2410                                                                        Nucleic acids          2409

                                                                      determine the presence of multiple pathogens. Moreover,
                                                                      the clear separation of peaks at baseline enabled accurate
                                                                      quantification of multiple targets in mixed samples. The
                                                                      broad dynamic range and the strong linear correlation
                                                                      between peak area and input gDNA amount indicate that
                                                                      this system can also provide highly accurate microbial
                                                                      community analysis if the members of the community are
                                                                      known. For the analysis of samples with unknown micro-
                                                                      organisms, better understanding on the microbial community
                                                                      in the sample may be achieved due to the high resolution of
                                                                      this system. Because of the improved resolution, changes of
                                                                      the community structure could be sensitively monitored.

                                                                          This work was supported by grants from the Ministry
                                                                      of Education, Science and Technology via MGAC, AEBRC
                                                                      (R11-2003-006-04003-0), and Converging Research Center
Figure 3. Correlation between relative peak area and amount of        Program.
input L. pneumophila gDNA. Serially diluted targets were
analyzed, and changes in CE-SSCP peak area were examined.                 The authors have declared no conflict of interest.
The linear correlation between the amount of diluted target gDNA
per reaction and relative peak area is represented by the R 2-value
(correlation coefficient). Relative peak areas of the other (nondi-
luted) targets are also shown. The x- and y-axes represent the log
                                                                      5 References
of gDNA amount (pg/reaction) and the relative peak area (target
peak area divided by P. mirabilis peak area), respectively.            [1] Schlundt, J., Int. J. Food Microbiol. 2002, 78, 3–17.
                                                                       [2] Hall, R. H., Microbes Infect. 2002, 4, 425–432.
targets. The results of 16S rRNA gene-specific PCR and CE-              [3] Hebart, H., Loffler, J., Meisner, C., Serey, F., Schmidt, D.,
SSCP analysis obtained using these samples showed that the                 Bohme, A., Martin, H., Engel, A., Bunje, D., Kem, W. V.,
peak area of the diluted target decreased with increasing                  Schumacher, U., Kanz, L., Einsele, H., J. Infect. Dis. 2000,
dilution (Fig. 2). Similar results were obtained using serial              181, 1713–1719.
dilutions of the other targets. The area of the diluted target         [4] Briese, T., Palacios, G., Kokoris, M., Jabado, O., Liu, Z.,
was linearly correlated with the amount of input gDNA,                     Renwick, N., Kapoor, V., Casas, I., Pozo, F., Limberger, R.,
whereas the relative peak areas of the other targets were not              Perez-Brena, P., Ju, J., Lipkin, W. I., Emerg. Infect. Dis.
                                                                           2005, 11, 310–313.
significantly changed (Fig. 3). The dynamic range, where the
amount of input gDNA is linearly correlated with peak area,            [5] Tang, Y. W., Procop, G. W., Persing, D. H., Clin. Chem.
                                                                           1997, 43, 2021–2038.
covered three orders of magnitude, and the correlation
coefficients (R2) were greater than 0.97, indicating strong             [6] Chun, J. A., Lee, W. H., Han, M. O., Lee, J. W., Yi, Y. B.,
                                                                           Park, G. Y., Chung, C. H., Biotechnol. Bioprocess Eng.
linear correlations (Table 2). These results demonstrate the
                                                                           2007, 12, 242–249.
potential of high-resolution CE-SSCP as a tool for analyzing
                                                                       [7] Espy, M. J., Uhl, J. R., Sloan, L. M., Buckwalter, S. P.,
microbial communities. Compared with conventional meth-
                                                                           Jones, M. F., Vetter, E. A., Yao, J. D., Wengenack, N. L.,
ods, such as microarrays, this improved method achieves                    Rosenblatt, J. E., Cockerill, F. R., 3rd, Smith, T. F., Clin.
excellent resolution and accurate quantification, and thus                  Microbiol. Rev. 2006, 19, 165–256.
provides high-quality results.                                         [8] Kaltenboeck, B., Wang, C., Adv. Clin. Chem. 2005, 40,
                                                                           219–259.
                                                                       [9] Monis, P. T., Giglio, S., Infect. Genet. Evol. 2006, 6,
4 Concluding remarks                                                       2–12.
                                                                      [10] Grace, M. B., McLeland, C. B., Gagliardi, S. J., Smith,
In this study, we developed and evaluated a multiplex                      J. M., Jackson, W. E., 3rd, Blakely, W. F., Clin. Chem.
pathogen detection system that separates and quantifies                     2003, 49, 1467–1475.
PCR products originating from a variable region of the 16S            [11] Lee, H. H., Yager, P., Biotechnol. Bioprocess Eng. 2007,
rRNA gene. Insufficient resolution in conventional CE-SSCP                  12, 634–639.
analysis has been overcome using a novel polymer matrix that          [12] Bodrossy, L., Sessitsch, A., Curr. Opin. Microbiol. 2004,
provides exceptionally improved separation of variable                     7, 245–254.
sequences. Sensitive multiplex detection was demonstrated             [13] Yang, J. R., Song, E., Kim, B. G., Kim, E. S., Sohng, J. K.,
for 12 important pathogens, the 16S rRNAs of which were                    Oh, M. K., Biotechnol. Bioprocess Eng. 2008, 13,
separated into 12 characteristic peaks by the improved CE-                 738–744.
SSCP procedure, on the basis of signature sequences. Now, in          [14] Zhou, J., Thompson, D. K., Curr. Opin. Biotechnol. 2002,
a single assay, it is possible to rapidly and sensitively                  13, 204–207.


& 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim                                                   www.electrophoresis-journal.com
2410       G. W. Shin et al.                                                                  Electrophoresis 2010, 31, 2405–2410


[15] Gentry, T. J., Wickham, G. S., Schadt, C. W., He, Z.,         [24] Pontes, D. S., Lima-Bittencourt, C. I., Chartone-Souza,
     Zhou, J., Microb. Ecol. 2006, 52, 159–175.                         E., Amaral Nascimento, A. M., J. Ind. Microbiol.
[16] Andersen, P. S., Jespersgaard, C., Vuust, J., Christiansen,        Biotechnol. 2007, 34, 463–473.
     M., Larsen, L. A., Hum. Mutat. 2003, 21, 455–465.             [25] Shin, G. W., Cho, Y. S., Hwang, H. S., Park, J. H., Jung,
[17] Bouallagui, H., Torrijos, M., Godon, J. J., Moletta, R.,           G. Y., Anal. Biochem. 2008, 383, 31–37.
     Cheikh, R. B., Touhami, Y., Delgenes, J. P., Hamdi, M.,       [26] Shin, G. W., Cho, Y. S., Hwang, H. S., Oh, M. H., Nam,
     Biotechnol. Lett. 2004, 26, 857–862.                               H. G., Park, J. H., Jung, G. Y., Electrophoresis 2009, 30,
[18] Chachkhiani, M., Dabert, P., Abzianidze, T., Partskha-             2728–2736.
     ladze, G., Tsiklauri, L., Dudauri, T., Godon, J. J., Bior-    [27] Kim, S. J., Shin, G. W., Choi, S. J., Hwang, H. S., Jung,
     esour. Technol. 2004, 93, 227–232.                                 G. Y., Seo, T. S., Electrophoresis 31, 1108–1115.
[19] Leclerc, M., Delgenes, J. P., Godon, J. J., Environ.          [28] Shin, G. W., Hwang, H. S., Seo, S. W., Oh, M. H., Ryu,
     Microbiol. 2004, 6, 809–819.                                       C. Y., Salvo, C. J., Feldman, S., Doh, J., Jung, G. Y.,
[20] Dabert, P., Delgenes, J. P., Godon, J. J., Appl. Microbiol.        J. Sep. Sci. 2010, 33, 1639–1643.
     Biotechnol. 2005, 66, 575–588.                                [29] Malmsten, M., Linse, P., Cosgrove, T., Macromolecules
[21] Khammar, N., Malhautier, L., Degrange, V., Lensi, R.,              1992, 25, 2474–2481.
     Godon, J. J., Fanlo, J. L., J. Appl. Microbiol. 2005, 98,     [30] Chu, B., Liang, D., J. Chromatogr. A 2002, 966, 1–13.
     476–490.                                                      [31] Wu, C., Liu, T., Chu, B., Electrophoresis 1998, 19,
[22] King, S., McCord, B. R., Riefler, R. G., J. Microbiol.              231–241.
     Methods 2005, 60, 83–92.                                      [32] Park, Y. S., Chu, H. S., Hwang, S. H., Seo, J. H.,
[23] Chung, J. H., Park, Y. S., Kim, J., Shin, G. W., Nam,              Choi, C. Y., Jung, G. Y., Electrophoresis 2006, 27,
     M. H., Oh, M. K., Kim, C. W., Jung, G. Y., Hyun Park, J.,          3836–3845.
     Electrophoresis 2007, 28, 2416–2423.




& 2010 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim                                               www.electrophoresis-journal.com

				
DOCUMENT INFO