Simultaneous detection of enteric bacteria from surface waters by ...

Environ Monit Assess (2009) 158:535–544 DOI 10.1007/s10661-008-0602-1

Simultaneous detection of enteric bacteria from surface waters by QPCR in comparison with conventional bacterial indicators Y. J. Liu · C. M. Zhang · X. C. Wang

Received: 1 July 2008 / Accepted: 29 September 2008 / Published online: 18 October 2008 © Springer Science + Business Media B.V. 2008

Abstract A rapid quantitative polymerase chain reaction (QPCR) method was developed for simultaneous detection of enteric bacteria from surface waters by utilizing a pair of universal primers which targeted four bacteria strains, namely Shigella dysenteriae, Vibrio cholerae, Salmonella typhimurium, and Escherichia coli. It was estimated that the QPCR method had a 94% confidence, and a detection limit as 2.7 E. coli cells per sample in undiluted DNA extracts. The QPCR method was applied for the bacteriological examination of several surface waters in the urban area of Xi’an, China and comparison was made with the conventional bacteria indicators determined by conventional membrane filter (MF) method. As a result, the calibrator cell equivalents (CCE) determined by QPCR was 2.2 to five times of the total coliform CFU, and the characteristics of the bacterial quality of different waters could be well presented by the QPCR results with a higher

Y. J. Liu (B) · C. M. Zhang · X. C. Wang Key Laboratory of Northwest Water Resource, Ecology and Environment, Ministry of Education, Xi’an University of Architecture and Technology, Xi’an 710055, China e-mail: [email protected] C. M. Zhang e-mail: [email protected] X. C. Wang e-mail: [email protected]

sensitivity. The coefficient of variation (CV) of data obtained by QPCR was smaller than that by traditional MF method, indicating a more stable analysis result. The QPCR method could thus be used as a supplement of the conventional culture method for more sensitive detection of pathogenic enteric bacteria from water. Keywords Universal primer · Enteric pathogenic bacteria · Quantitative polymerase chain reaction (QPCR) · Membrane filtration (MF) · Surface water

Introduction Surface water pollution, mostly by receiving untreated sewage from populated area, is still the main reason for the great number of human morbidities and mortalities worldwide. In particular, waterborne infections such as typhoid fever, cholera, dysentery, and traveler’s diarrhea are caused by different types of bacterial pathogens and thus pose a major public health hazard (Hunter 1997). For these reasons, regular monitoring of waterborne pathogens becomes extremely important for the protection of public health. However, lack of accurate and costeffective diagnostic tests is often a major obstacle in the prevention and control of infections and outbreaks transmitted by waterborne pathogens.

536

Environ Monit Assess (2009) 158:535–544

Detection and enumeration of bacterial pathogens have been largely based on the use of selective culture and standard biochemical methods. These methods have a number of drawbacks (Kong et al. 2002; Fleisher 1990; Roszak and Colwell 1987; Rahman et al. 1996). Increasing interest is now being directed towards the possible use of molecular microbial analysis methods with shorter reporting times. One such technology is the quantitative polymerase chain reaction (QPCR). Primer sets and probes, associated with this technology, have now been developed for the specific detection of different kinds of waterborne pathogens (Ludwig and Schleifer 2000; Lyon 2001; Brinkman et al. 2003; Foulds et al. 2002; Blackstone et al. 2003; Frahm and Obst 2003; Guy et al. 2003; Noble et al. 2003; Alexandrino et al. 2004; Abu-Halaweh et al. 2005; Haugland et al. 2005). Portable instruments that can be operated at or near the site are available now (Depaola, personal communication), and rapid methods for processing water samples for QPCR analysis have also been developed (Brinkman et al. 2003; Haugland et al. 2005). It was reported that the potential overall time requirements of QPCR could be reduced to a few hours from taking samples to obtaining results (Brinkman et al. 2003). Another advantage of QPCR is the possibility of simultaneous detection of different pathogens as long as their homology of gene sequences can be identified (Guy et al. 2003). Considering this property of QPCR, in this study, the authors designed bacterial universal primers targeting four kinds of enteric pathogenic bacteria, namely Shigella dysenteriae, Vibrio cholerae, Salmonella typhimurium, and Escherichia coli which are most frequently accoutered in surface waters (Bej et al. 1991; Blackstone et al. 2003; Hunter 1997). The

Table 1 Character of surface water samples

QPCR protocol thus established was applied to several surface waters in the urban area of Xi’an city for simultaneous detection of four kinds of enteric bacteria. Comparison was also conducted between the results of QPCR and that of bacteria count by conventional culture method.

Materials and method Study site and water sampling In this study, surface water samples were collected from two rivers and two lakes in Xi’an urban area. For comparison, samples were also colleted from the secondary effluent from a waste-water treatment plant (WWTP). The characteristics of these surface water bodies are shown in Table 1. In a 4-month period from March to June 2007, water sampling was conducted regularly on the 5th, 15th, and 25th of each month except for the 5th of May as weather condition was unfavorable for taking representative samples. On-site measurements were conducted regarding pH, turbidity, air, and water temperature using potable meters. Two samples each with 1-l volume were collected at each exacting location on the water body in a way referring to the standard method (American Public Health Association 1998). One sample was used for microbiological and other chemical analysis and another was used for QPCR analysis. As sampling, a 1,000-ml sterilized polypropylene bottle was lowered to an appropriate sampling depth (1/2 water depth) and then the lid was removed to let water in. The collected samples were placed in a cooling box filled with ice to maintain a low temperature of 1–4◦ C in the

Number

Name of water body

Use of the water body

Remark

1

Chanhe River

2

Xingqinghu Lake

A water course receiving industrial discharge Landscaping water

3 4

Beihu Lake Heihe River

An aperiodically polluted river A polluted lake after water quality restoration A newly built artificial lake A well-protected river

5

Beishiqiao WWTP

Landscaping water Source water for drinking water supply Secondary effluent

Without disinfection


process of transport from the sampling site to the laboratory. The samples were treated for analyses within 6 h after collection. Microbiological analyses Total bacteria, total coliform and fecal coliform at each exacting locations on the water bodies were detected referring to the standard method (American Public Health Association 1998). Design of enteric bacterial universal primers The 16S rRNA gene sequences of S. dysenteriae, V. cholerae, S. typhimurium, and E. coli were searched from the GenBank. By analysis of the homology using DNA-STAR software (ver. 3.2), the universal primers were designed as 5 -aaggc gacgatccctagctggtctgagaggatga/c-3 (nt 246–280, E. coli. 16S rRNA numbering); 5 -gcttgccagta tcagatgcagttcccaggttgagc-3 (nt 556–521, E. coli. 16S rRNA numbering). The primers were then synthesized under the assistance of the Shanghai Bioengineering Co., China. Sample processing for QPCR analyses Reference bacteria strains The reference bacteria strains as shown in Table 2 were supplied by Shaanxi Research Institute of Microbiology, Xi’an, China. The medium used

Table 2 Reference bacteria strains Accession numbers

Strain

EF424586 EF420247 EF378646 EF422070 EF413067 EF394153 BD267944 NC009089 EF421208 DQ171719 DQ362495 DS179652 D12814 EF418614

Staphylococcus aureus Bacillus subtilis Pseudomonas aeruginosa Bacillus cereus Bacillus thuringiensis Bradyrhizobium japonicum Lactobacillu bulgaricus Clostridium Difficile Bifidobacterium longum Lactococcus lactis Shigella dysenteriae Vibrio cholerae Salmonella typhimurium Escherichia coli

537

was solid Luria–Bertani (LB) and LB broth (per liter: 10 g Bacto-tryptone, 5 g NaCl, 5 g yeast extract). Bacteria strains were cultivated in a LB broth culture medium at 37◦ C for 24 h, centrifuged at 7,000 rpm for 10 min, and then the bacteria were collected. DNA extraction A 1.5-ml reference bacterial culture liquid or 100 ml surface water sample was centrifuged at 7,000 rpm for 10 min and then the supernatant was removed. 567 μl broken buffer (40 mmol/l Tris– HCl, pH 8.0, 20 mmol/l CH3 COONa, 1 mmol/l EDTA, 1% SDS) was added to the sediment to re-suspend it at 37◦ C for 30 min. 66 μl 5 mol/l NaCl was added and the suspension was fully mixed at 65◦ C for 20 min, and then centrifuged at 10000 rpm for 10 min. The supernatant was moved to another new tube, and mixed with equal volume of mixture of isochoric phenol, chloroform, and isoamyl alcohol at a ratio of 25:24:1 for extraction. After centrifugation at 10,000 rpm for 5 min, the supernatant was collected and 0.6 time volume of isoamyl alcohol was added to obtain the DNA sediment. The DNA sediment was washed three times with 1 ml 70% ethanol and the ethanol was then removed after centrifugation. The DNA was dissolved in ddH2 O and stored for PCR amplification (Bej et al. 1991). QPCR standard curve E. coli bacterial liquid of known cell density was diluted with sterile distilled water at ten times gradient, and then the diluted liquid was centrifuged (7,000 rpm, 10 min) for reclaiming the bacterial cells. The bacterial cells were washed three times with sterile distilled water, and then the bacterial total DNA was extracted by the method mentioned above. The total DNA of each diluting concentration was taken as the DNA template for QPCR standard curve preparation. With each diluting concentration, the procedures were repeated three times and the average result was used for determining the bacterial cell density. The sensitivity and precision of the QPCR method could be evaluated by scatter plot and regression analysis of the relationship between log calibrator

538

cell equivalents (CCE) of E. coli and the QPCRmeasured cycle threshold (CT ) values for DNA extracts of serially diluted E. coli cells (Brinkman et al. 2003). QPCR analyses PCR analyses The total volume of the PCR amplification reaction system was 25 μl. It contained dNTP 0.2 mmol/l, Taq DNA polymerase 1.0 U, 1× PCR Buffer, 2.0 mmol/l MgCl2 , 0.1 mmol/l upstream and downstream primers, and 2 μl DNA template. The PCR amplification condition was as following: denaturalization at 94◦ C 5 min, 94◦ C 30 s, 55◦ C 30 s, 72◦ C 30 s, 35 cycles, 72◦ C extension 5 min. In negative comparison, the DNA template was replaced by sterile ddH2 O. The PCR product was analyzed by 2% agar sugar gelatin electrophoresis (including 0.5 mg/ml bromize pyrimidine), and after imaging with gelatin imaging system1000 (Bio-Rad, America), the PCR product was then reclaimed and purified using DNA reclaim reagent kit (Huamei Biological Engineering Company, China). DNA fragments sequencing was conduced under the assistance of Shanghai Biological Engineering Company, China. Homologous analysis was conducted using the DNA-STAR software (Perkin Elmer, Norwalk, CT, USA). QPCR analyses of water sample The QPCR system was constructed by adding fluorescence reagent Sybr Green I (Tiangen Biological and Chemical Company, China) to the PCR system mentioned above. Real time detection of the PCR process was performed by detecting the fluorescence signal. The total volume of QPCR system was 25 μl, containing 1× real MastrMix/1× SYBR solution, 0.1 mmol/l upstream and downstream primers, and 2μl DNA template. The QPCR mixing reaction liquid was added into an 8position tube (MJ Research TLS-0251) and closed tightly with a super clean lid (MJ Research TCS0803). The reaction tube was then put on to the QPCR instrument (American BIO-RAD MJ) for QPCR amplification under the following condi-


tions: pre-denaturalization 94◦ C 5 min, and denaturalization 94◦ C 30 s, 55◦ C 30 s, 72◦ C 30 s, 85◦ C 2 s, 35 cycles. Determinations of cycle threshold (CT ) were performed automatically by the instrument after manually adjusting the threshold fluorescence value to eight units (Haugland et al. 1999). At least three replicates DNA extracts of the pathogenic bacteria from each water sample were prepared. The cell density of the original water sample was obtained from the cell CCE determined by QPCR in diluted water sample multiplying the diluting coefficient ranging from 107 to 101 . The cell density of the pathogenic bacteria in a water sample was a relative quantity calculated from the QPCR standard curve of known cell density.

Results and discussion Universal primers and their specificities for target strains According to the high conservation of bacteria 16S rRNA gene, bacterial universal primers targeting S. dysenteriae, V. cholerae, S. typhimurium, and E. coli were designed. The primer sequences were compared with each other and the homology searches were performed based on the sequence similarity by using the BLAST program and the GenBank database. The BLAST results of correlative bacterial strains for the evaluation of specificity of PCR universal primers were not shown here. It was found by computer analysis that the universal primer pair had significant affinities for target genes of Salmonella spp., Shigella spp., Vibrio spp., and Escherichia spp. The universal primer pair above mentioned was tested by PCR on DNA templates prepared from S. dysenteriae, V. cholerae, S. typhimurium, E. coli, and the other ten control bacteria strains. As shown in Fig. 1, a limpid specific strap could be seen at 320 bp for all the four target strains on the electrophoregrams, but none for the control strains. Analysis results verified that the universal primer pair showed specificities only for their corresponding target genes. To confirm their identity, direct sequencing was conducted and all the

Environ Monit Assess (2009) 158:535–544 kb M 1.0

1

2

3

539 4

5

6

7

8

9

10

11

12

13

14

0.7 0.5 0.4 0.3

Fig. 1 PCR amplification results of the four target strains and ten control strains M DNA Marker; 2 S. dysenteriae; 6 V. cholerae; 8 S. typhimurium; 12 E. coli; 1, 3–5, 7, 8–11, 13–14 control strains

amplimers showed a high percentage of sequence similarity (>99%) with published 16S rRNA gene sequences from Shigella sp., Vibrio sp., Salmonella sp., and E. coli. in the GenBank database. This proved that the universal primer pair designed in this study was suitable for the specific detection of most genetic species of the four target strains. Effectiveness and reliability of the QPCR method Figure 2 shows the relationship of log CCE of E. coli and QPCR-measured cycle threshold (CT ) values for DNA extracts of serially diluted E. coli. Original cell density of the water sample could be calculated from the QPCR standard curve of E. coli from the CT value of the water sample. The slope of the equation can be used to examine the efficiency of PCR. For a 100% PCR efficiency, the slope was −3.32. A perfect QPCR standard curve was based on the PCR efficiency reaching to 90%–100% (100% PCR efficiency means that the quantity of DNA template can be doubled

Cycle Threshold

35 30

y = -3.203x + 32.221 R2 = 0.9946

25 20 15 1

2

3

4

5

log10 Cell Equivalents

Fig. 2 Scatter plot and regression analysis results of log10 calibrator cell equivalents (CCE) of E. coli on QPCRmeasured cycle threshold (CT ) values for DNA extracts of serially diluted E. coli. Analyses of at least three replicate DNA extracts of E. coli are shown. The regression is represented by the dashed line

after each cycle). Only when the linear regression analyses of standard curve had a high correlation coefficient (r2 ≥ 0.99), could the process and data of QPCR experiment be believable (Haugland et al. 1999, 2005; Brinkman et al. 2003). The original DNA template copy of E. coli examined in this study was from 6.8 × 101 to 6.8 × 105 CFU/ml. The log CCE of the E. coli original DNA template was positively proportional to its corresponding CT value. The regression coefficient was 0.994 indicating a strong linear relation. After second regression, the slope was −3.203, and the PCR efficiency was 99%. The standard deviation of CT value between the replicate DNA extracts of every diluted grade was less than 0.3. The CCE of S. dysenteriae, V. cholerae, S. typhimurium (for which cells density were known) were also examined by QPCR method with universal primers. The scatter plot and regression analysis results indicated that the slope did not change apparently. The results of QPCR method and bacterial counting for serially diluted E. coli cells were compared and the reliability of the QPCR examination result was 94%. Taking E. coli as representative bacteria strain, the sensibility of QPCR method with universal primers was examined. As a result, the minimum detection limit of the QPCR method (including the processes of bacteria cell reclamation, DNA extracting and QPCR) was 2.7 bacterial cell DNA extracts. QPCR and bacteria counting results for surface water samples Both QPCR and bacterial counting were conducted for surface water samples from the water bodies shown in Table 1 in the 4 months period from March to June 2007. Figure 3 shows the variations of the measured cell intensities in water

540


Geometric mean / 100mL

Heihe River 10000

coliform

total bacteria

f faecal coliform

QPCR

1000 100 10 1

3.5 3.15 3.25

4.5 4.15 4.25 5.8 5.15 5.25

6.5 6 6.15 6.25

Sampling date

Fig. 3 Geometric means of enteric bacteria densities per 100 ml of water from all sampling locations of Xi’an city: (filled circle) determined by QPCR method as calibrator cell equivalents; (filled square) total bacteria determined by MF method as CFU; (filled diamond) coliform determined by MF method as CFU; (filled upright triangle) fecal coliform determined by MF method as CFU. Unless otherwise specified, sampling date is from March 2007 to June 2007. Sampling visits at each location occurred three times at 8 am to 12 am of the 5th, 15th, and 25th of every month


Beihu Lake 100000 10000 1000 100 10 1

coliform

total bacteria

3.5 3.15 3.25

faecal coliform

4.5 4.15 4.25 5.8 5.15 5.25

QPCR

6.5 6 6.15 6.25

Sampling date


Xingqinghu Lake 1000000

coliform

total bacteria

faecal coliform

QPCR

100000 10000 1000 100 10 1

3.5 3.15 3.25

4.5 4.15 4.25 5.8 5.15 5.25

6.5 6 6.15 6.25

Sampling date


Chanhe River 100000 10000 1000 100 10 1

coliform

total bacteria

3.5 3.15 3.25

faecal coliform

4.5 4.15 4.25 5.8 5.15 5.25

QPCR

6.5 6 6.15 6.25

Sampling date


Beishiqiao Wastewater Treatment Plant 1000000 100000 10000 1000 100 10

coliform

total bacteria

faecal coliform

QPCR

1 3.5 3.15 3.25 4.5 4.15 4.25 5.8 5.15 5.25 6.5 6.15 6.25

Sampling date

at each sampling site regarding the enteric bacteria CCE determined by QPCR and the total bacteria, total coliform and fecal coliform CFU determined by conventional culture method. The results were also compared in Table 3 by the statistical parameters of the data at each site for each bacterial item. In general, the CCE determined by QPCR was always higher than the total coliform CFU, indicating that coliform bacteria are only part of the enteric bacteria encountered in surface waters. Because the universal primers designed for this study targeted four enteric bacteria strains as S. dysenteriae, V. cholerae, S. typhimurium, and E. coli, the QPCR result can be considered to represent the total density of four kinds of enteric bacteria. From Fig. 3 and Table 3, the enteric bacterial CCE, on average, was about 2.2 to five times of the total coliform CFU and 16 to 34 times of the fecal coliform CFU. It can thus be roughly estimated that total coliform may take about 20% to 45%, and fecal coliform may take about 3% to 6% of the total enteric bacteria. The pattern of variation in the cell density determined by QPCR with time, as shown in Fig. 3, was similar to that of total coliform, fecal coliform and total bacteria determined by culture method. This is most noticeable for Beihu Lake, Xingqinghu Lake and Chanhe River where the data on 8th of May appeared to be a peak for either QPCR or each of the other parameters. Because 8th of May was just after the 1-week May Day holidays when tourist activity was concentrated near these waters, these peaks reflected the strong effect of human activity on bacterial quality of surface water. The difference between these waters in bacterial quality was apparent. Heihe River, which


541

Table 3 Detection results analysis of enteric bacteria in surface water of Xi’an city (determined by QPCR method as calibrator cell equivalents; determined by MF method as CFU) Sample

Bacteria/100 ml

Geometric meana

Log10 meanb

SDc

Log10 SDd

C.V.e

Heihe River

Total bacteria E. coli Fecal coliforms QPCR Total bacteria E. coli Fecal coliforms QPCR Total bacteria E. coli Fecal coliforms QPCR Total bacteria E. coli Fecal coliforms QPCR Total bacteria

904.17 15.25 4.58 74.67 10455.00 181.17 21.25 591.00 38683.33 435.83 56.75 1998.33 49608.33 2509.17 170.42 5755.56 244408.33

2.96 1.18 0.66 1.87 4.02 2.26 1.33 2.77 4.59 2.64 1.75 3.30 4.70 3.40 2.23 3.76 5.39

407.4 8.87 4.29 26.51 6545.63 133.79 14.32 302.92 38502.42 348.78 52.81 853.83 18844.55 1031.92 59.48 1791.03 264382.92

2.61 0.95 0.63 1.42 3.82 2.13 1.16 2.48 4.59 2.54 1.72 2.93 4.28 3.01 1.77 3.25 5.42

0.45 0.58 0.94 0.36 0.63 0.74 0.67 0.51 1.00 0.80 0.93 0.43 0.38 0.41 0.35 0.31 1.08

E. coli Fecal coliforms QPCR

13200.00 1455.83 42300.00

4.12 3.16 4.63

12173.74 1037.54 25965.22

4.09 3.02 4.41

0.92 0.71 0.61

Beihu Lake

Xingqinghu Lake

Chanhe River

Beishiqiao Wastewater Treatment plant

a Geometric

mean of all sampling locations (N = 12 at every sampling location) visits

b Log mean of all sampling 10 c Standard deviation

d Log SD between sampling visits 10 e Coefficient of variation (=SD in original

units / mean)

is a well-protected river as source water for drinking water supply, was found to be of good quality according to Chinese environmental quality standards for surface water (State Environmental Protection Administration of China 2002) regarding the fecal coliform in a range between 0 and 10 CFU/100 ml (average as about 4 CFU/100 ml) and the total coliform between 3 and 34 CFU/100 ml (average as about 15 CFU/100 ml). The QPCR was measured in a range between 24 and 86 CCE/100 ml (average as about 75 CCE/100 ml) which was about 16 times of the fecal coliform and five times of the total coliform. Therefore, it can be considered that the possibility of existence of enteric pathogenic bacteria other than coliform is still high with such a well-protected river. The bacterial quality of the two urban lakes, namely Beihu Lake and Xingqinghu Lake, were acceptable for landscape (State Environmental Protection Administration of China 2002) regarding the fecal coliform on av-

erage as about 20 CFU/100 ml and 60 CFU/100 ml, respectively, and the total coliform on average as about 180 CFU/100 ml and 440 CFU/100 ml, respectively. However, the average QPCR measurements were as about 590 CCE/100 ml and 2000 CCE/100 ml, respectively, indicating that there were larger amount of enteric bacteria other than coliform bacteria in existence in the lake water. The condition of the Chanhe River was worse because it was an urban drainage receiving industrial discharge. The secondary effluent from the WWTP, as expected, contain high concentrations of fecal coliform and total coliform, and also showed a higher QPCR measurement. As summarized in Table 3, the variability of the QPCR results was somewhat higher than that of the MF method. This higher variability is reflected by the standard deviation values and may be related to a lower intrinsic precision of QPCR measurements that can result from analyses of small portions of the total sample extracts by

542 100000

y = 0.0813x + 1

CCE(QPCR)

10000

2 R = 0.8148

1000 100

(a)

10 1

1

10

100

1000

The relationships between the bacteria CCE determined by QPCR method and the CFU of total bacteria, total coliform, and fecal coliform determined by conventional culture method are shown in Fig. 4 by scatter plot on a log–log coordinates. Regression analysis was conducted between log(CCE) and log(CFU) of total bacteria, total coliform and fecal coliform. As a result, the correlativity was found to be the strongest between QPCR CCE and total coliform CFU (r2 = 0.93) and followed by total bacteria CFU (r2 = 0.90) and fecal coliform CFU (r2 = 0.83). In this study, the universal primers designed were targeting four enteric bacteria strains including E. coli, S. dysenteriae, V. cholerae, and S. typhimurium for QPCR analysis. Therefore, the strong correlativity between QPCR CCE and total coliform CFU may indicate that the QPCR method can be used as a supplement of the conventional culture

100000

1000000

100000

y = 1.6772x + 1

CCE(QPCR)

10000

R2 = 0.8607

1000 100

(b)

10 1

1

10

100

1000

10000

100000

CFU(coliform) 100000 10000 1000

(c)

100 y = 16.674x + 597.74

10

Scatter plot and regression analysis of the examination result

10000

CFU(total bacteria)

CCE(QPCR)

this method. The results obtained by the MF method, however, also differed greatly by sampling location for each of the visits. These results demonstrate the need for a comprehensive water sampling scheme, such as the one used in this study, to perform accurate detection of pathogenic bacteria from surface water. Table 3 also compared the coefficients of variation (CV) which is the ratio of the standard deviation (SD) over the mean value of the data regarding the measurements of total bacteria, total coliform, fecal coliform and QPCR. It is understood that the CV of QPCR measurement was the lowest in each case, indicating that the timely variation of data obtained by QPCR method was smaller than that by traditional culture method. There may be two reasons to explain this: one is that the QPCR method, based on DNA amplification, may be more sensitive and stable for bacteria detection, and another is that the QPCR method may not be able to distinguish live cells from the dead cells (Duprey et al. 1997; Kreader 1998; Ludwig and Schleifer 2000).


1

R2 = 0.6816

1

10

100

1000

10000

CFU(faecal coliform)

Fig. 4 Scatter plot and regression analysis results of geometric mean of total bacteria CFU densities a, coliform CFU densities, b and fecal coliform CFU densities, c determined by MF method vs. pathogenic bacteria CCE determined by QPCR method for all sampling visits to five different sampling location of Xi’an city

method for more sensitive detection of pathogenic enteric bacteria from water. However, because QPCR and conventional culture are based on different principles, the results by the two methods may not always correspond with each other (Haugland et al. 2005). In order to fully understand the common points and difference between the two methods, further studies are still needed for a detailed investigation of the distribution of the enteric bacteria belonging to each individual strain in surface water.


Conclusions From the former sections, the following conclusions could be drawn: 1. The universal primers designed in this study showed high specificities for the four target strains of S. dysenteriae, V. cholerae, S. typhimurium, and E. coli. This made it possible for simultaneous detection of enteric bacteria by rapid QPCR method. 2. Enteric bacteria detection by the QPCR method could be completed in less than 5 h. Its confidence level was 94%, and its detection limit was 2.7 E. coli cells per sample in undiluted DNA extracts. 3. By applying the QPCR method for the bacteriological examination of several surface waters in the urban area of Xi’an, China and comparing with the conventional bacteria indicators such as total bacteria, total coliform and fecal coliform determined by conventional membrane filter (MF) method, it was found that the CCE determined by QPCR was 2.2 to five times of the total coliform CFU, indicating that coliform bacteria were only part of the enteric bacteria encountered in surface waters. The pattern of variation of the CCE was similar to that of the CFU of the conventional bacteria indicators and the characteristics of the bacterial quality of different waters could be well presented by the QPCR results with a higher sensitivity. 4. The coefficient of variation (CV) of data obtained by QPCR was smaller than that by traditional MF method, indicating a more stable analysis result. By scatter plot and regression analysis of the data, it was found that the QPCR data correlated well with the total coliform data. The QPCR method could thus be used as a supplement of the conventional culture method for more sensitive detection of pathogenic enteric bacteria from water. In order to fully understand the common points and difference between the two methods, further studies are still needed for a detailed investigation

543

of the distribution of the enteric bacteria belonging to each individual strain in surface water. Acknowledgements The study is supported by the National Natural Science Foundation of China (NSFC, Grant No. 50478048) and the NSFC-JST (Japan Science and Technology Agency) Joint Research Program (Grant No. 50621140001).

References Alexandrino, M., Grohmann, E., & Szewzyk, U. (2004). Optimization of PCR-based methods for rapid detection of Campylobacter jejuni, Campylobacter coli and Yersinia enterocolitica serovar 0:3 in wastewater samples. Water Research, 38, 1340–1346. doi:10.1016/ j.watres.2003.10.036. Abu-Halaweh, M., Bates, J., Patel, B. K. (2005). Rapid detection and differentiation of pathogenic Campylobacter jejuni and Campylobacter coli by real-time PCR. Research in Microbiology, 156, 107–114, Jan– Feb. American Public Health Association, American Water Works Association and Water Environment Federation (1998). Standard methods for the examination of water and wastewater, 20th ed. Washington, DC: American Public Health Association. Bej, A. K., Steffan, R. J., DiCesare, J. L., Haf, L., & Atlas, R. M. (1991). Detection of Escherichia coli and Shigella spp. in water using the polymerase chain reaction and gene probes for uid. Applied and Environmental Microbiology, 57, 1013–1017. Blackstone, G. M., Nordstrom, J. L., Vickery, M. C., Bowen, M. D., Meyer, R. F., & DePaola, A. (2003). Detection of pathogenic Vibrio parahaemolyticus in oyster enrichments by real time PCR. Journal of Microbiological Methods, 53, 149–155. doi:10.1016/ S0167-7012(03)00020-4. Brinkman, N. E., Haugland, R. A., Wymer, L. J., Byappanahalli, M., Whitman, R. L., & Vesper, S. J. (2003). Evaluation of a rapid, quantitative real-time PCR method for cellular enumeration of pathogenic Candida species in water. Applied and Environmental Microbiology, 69, 1775–1782. doi:10.1128/ AEM.69.3.1775-1782.2003. Duprey, E., Caprais, M. P., Derrien, A., & Fach, P. (1997). Salmonella DNA persistence in natural seawaters using PCR analysis. Journal of Applied Microbiology, 82, 507–510. doi:10.1046/j.1365-2672.1997.00143.x. Fleisher, J. M. (1990). Conducting recreational water quality surveys: Some problems and suggested remedies. Marine Pollution Bulletin, 21, 562–567. doi:10.1016/0025-326X(90)90603-6. Foulds, I. V., Granacki, A., Xiao, C., Krull, U. J., Castle, A., & Horgen, P. A. (2002). Quantification of

544 microcystin-producing cyanobacteria and E. coli in water by 5 -nuclease PCR. Journal of Applied Microbiology, 93, 825–834. doi:10.1046/j.1365-2672. 2002.01772.x. Frahm, E., & Obst, U. (2003). Application of the fluorogenic probe technique (TaqMan PCR) to the detection of Enterococcus spp., and Escherichia coli in water samples. Journal of Microbiological Methods, 52, 123–131. doi:10.1016/S0167-7012(02)00150-1. Guy, R. A., Payment, P., Krull, U. J., & Horgen, P. A. (2003). Real-time PCR for quantification of Giardia and Crytosporidium in environmental water samples and sewage. Applied and Environmental Microbiology, 69, 5178–5185. doi:10.1128/AEM.69. 9.5178-5185.2003 . Haugland, R. A., Siefring, S. C., Wymer, L. J., Brenner, K. P., & Dufour, A. P. (2005). Comparison of Enterococcus measurements in freshwater at two recreational beaches by quantitative polymerase chain reaction and membrane filter culture analysis. Water Research, 39, 559–568. Haugland, R. A., Vesper, S. J., & Wymer, L. J. (1999). Quantitative measurement of Stachybotrys chartarum conidia using real time detection of PCR products with the TaqManTM fluorogenic probe system. Molecular and Cellular Probes, 13, 329–340. doi:10.1006/mcpr.1999.0258. Hunter, P. R. (1997). Drinking water and waterborne diseases. In P. R. Hunter (Ed.), Waterborne disease: Epidemiology and ecology (pp. 27–41). New York: Wiley. Kong, R. Y. C., Lee, S. K. Y., Law, T. W. F., Law, S. H. W., & Wu, R. S. S. (2002). Rapid detection of six types of bacterial pathogens in marine waters

Environ Monit Assess (2009) 158:535–544 by multiplex PCR. Water Research, 36, 2802–2812. doi:10.1016/S0043-1354(01)00503-6. Kreader, C. A. (1998). Persistence of PCR-detectable Bacteroides diastonis from human feces in river water. Applied and Environmental Microbiology, 64, 4103–4105. Ludwig, W., & Schleifer, K. H. (2000). How quantitative is quantitative PCR with respect to cell counts? Systematic and Applied Microbiology, 23, 556–562. Lyon, W. J. (2001). TaqMan PCR for detection of Vibrio cholerae O1, O139, non-O1, and non-O139 in pure cultures, raw oysters, and synthetic seawater. Applied and Environmental Microbiology, 67, 4685–4693. doi:10.1128/AEM.67.10.4685-4693.2001. Noble, R. T., Allen, S. M., Blackwood, A. D., Chu, W., Jiang, S. C., Lovelace, G. L., et al. (2003). Use of viral pathogens and indicators to differentiate between human and non-human fecal contamination in a microbial source tracking comparison study. Journal of Water and Health, 1, 195–207. Rahman, I., Shahamat, M., Chowdhury, M. A., & Colwell, R. R. (1996). Potential virulence of viable but nonculturable Shigella dysenteriae type 1. Applied and Environmental Microbiology, 62, 115–120. doi:10.1016/j.watres.2004.11.011. Roszak, D. B., & Colwell, R. R. (1987). Survival strategies of bacteria in the natural environment. Microbiological Reviews, 51, 365–379. State Environmental Protection Administration of China and General Administration of Quality Supervision, Inspection and Quarantine of China (2002). Environmental quality standards for surface water (GB38382002).