Competitive polymerase chain reaction for quantification of nonculturable Enterococcus faecalis cells in lake water

FEMS Microbiology Ecology 30 (1999) 345^353

Competitive polymerase chain reaction for quanti¢cation of nonculturable Enterococcus faecalis cells in lake water Maria del Mar Lleo¨ *, Maria Carla Ta¢, Caterina Signoretto, Cecilia Dal Cero, Pietro Canepari Dipartimento di Patologia, Sezione di Microbiologia, Universita© di Verona, Strada Le Grazie, 8, 37134 Verona, Italy Received 20 May 1999 ; revised 31 August 1999; accepted 3 September 1999

Abstract Among the survival strategies developed by bacteria when faced with adverse environmental conditions, the viable but nonculturable (VNC) state has been described. In this state, bacteria are unable to form colonies but are still alive and capable of metabolic activity. The VNC state has been described in numerous Gram-negative species, but recently also in Enterococcus faecalis, a Gram-positive species which can be found in the environment. In this study we describe a competitive PCR (cPCR) protocol to detect and quantify a specific sequence of DNA from culturable and nonculturable E. faecalis cells present in water samples. The protocol was found to be specific and capable of detecting amounts of DNA up to 0.1 pg corresponding to approximately 2 cells ml31 . Moreover, it allows an internal standard to be used to quantify the amount of specific DNA present in samples from different environments. The application of this cPCR method to water samples from Lake Garda enabled us to demonstrate the presence of nonculturable forms of E. faecalis in lake water and to quantify their DNA and the corresponding concentration of nonculturable cells. ß 1999 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved. Keywords : Viable but nonculturable state; Enterococci ; Polymerase chain reaction; Molecular biology

1. Introduction It has been demonstrated that bacteria are capable of developing a survival strategy, i.e. the viable but nonculturable state (VNC), when faced with adverse environmental conditions [1,2]. Several physicochemical conditions such as temperature, nutrient concentration, aeration, etc., have been shown to

* Corresponding author. Tel.: +39 (45) 8098194; Fax: +39 (45) 584606; E-mail: [email protected]

induce entry into the VNC state. The in£uence of these parameters in stimulating cells to enter the VNC state is highly variable according to bacterial species; for example, whereas high temperatures (25³C) induce the VNC state in Campylobacter jejuni [3], temperatures below 10³C must be used to stimulate entry into the VNC state in Vibrio vulni¢cus [4]. Cells in the VNC state are unable to form colonies on plates but are still alive and capable of metabolic activity [1,5^7]. The VNC state has been demonstrated in numerous Gram-negative bacterial species, mostly of medical interest [2,5], but more recently we have shown that a Gram-positive species, namely

0168-6496 / 99 / $20.00 ß 1999 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved. PII: S 0 1 6 8 - 6 4 9 6 ( 9 9 ) 0 0 0 7 3 - 2

FEMSEC 1079 12-11-99

346

M.M. Lleo¨ et al. / FEMS Microbiology Ecology 30 (1999) 345^353

Enterococcus faecalis, is also able to develop this particular survival mechanism [8]. Moreover, VNC cells remain potentially pathogenic [1,6] and can resume active growth when optimal environmental conditions are restored [8,10^12]. Consequently, standard methods, based on plate counts after inoculation of water samples in selective and nonselective media, may no longer be adequate in that they are unable to detect bacteria in the nonculturable state and need to be replaced by other techniques with such capability. Some of these methods (polymerase chain reaction (PCR), £uorescent monoclonal antibodies, speci¢c DNA probes, etc.) have now been shown to be valid when applied to detection of VNC bacteria in di¡erent environments [13^16]. Among these methods, PCR has proved very useful because it allows a speci¢c segment of DNA to be ampli¢ed by a factor of 106 or more within hours, thus potentially permitting detection of cells which are present only in very limited amount [14,17]. Moreover, it has recently been demonstrated that the PCR technique can be employed to detect low numbers of a speci¢c bacterium against a large background of other prokaryotic and eukaryotic cells and of organic material which may be present in the sample [17^19]. These properties make PCR a particularly suitable method for analyzing environmental samples. However, molecular methods such as PCR need to be adapted to the di¡erent environments by varying the parameters in£uencing speci¢c ampli¢cation in order to obtain the highest degree of speci¢city and sensitivity. A more advanced method is competitive PCR which, using an internal standard consisting of a DNA fragment derived from the DNA target, is capable not only of detecting the presence of the DNA target but also of precisely quantifying it. In this study we describe a cPCR method capable of detecting low numbers (up to 2 cells ml31 ) of E. faecalis in water from Lake Garda, Verona, Italy. This method also enabled us to use an internal standard which is coampli¢ed with the target fragment to accurately quantify E. faecalis cells in a range of 2U102 ^2U105 cell ml31 . By comparing these data with those obtained by standard methods such as colony forming unit (CFU) counts, we were able to establish the order of magnitude of nonculturable E. faecalis forms present in Lake Garda.

2. Materials and methods 2.1. Bacterial strains and growth conditions The enterococcal strains used were E. faecalis SF97, a strain previously described [8], Enterococcus hirae ATCC 9790, one environmental strain of Enterococcus durans (ED5) and one of Enterococcus casseli£avius (ECS2) and three di¡erent clinical isolates of Enterococcus faecium (namely VR1, VR5, VR27). A clinical isolate from each of several Gram-negative species was also used: Escherichia coli, Aeromonas sobria, Enterobacter aerogenes, Pseudomonas £uorescens, Shigella dysenteriae, Serratia marcescens and Klebsiella pneumoniae. All the bacterial strains were identi¢ed using the respective speci¢c identi¢cation methods (ID 32 system, BioMerieux, France). Both Gram-positive and Gramnegative bacterial strains were grown on Tryptic Soy Broth (TSB, Difco, USA) at 37³C. Total count of culturable bacteria was performed by plating lake water samples on Tryptic Soy Agar (TSA, Difco), while the selective count of enterococci was performed by plating the same samples on a selective medium such as m-Enterococcus (20 g of tryptone, 5 g of yeast extract, 2 g of glucose, 4 g of K2 HPO4 , 0.4 g of NaN3 , 0.1 g of 2,3,5-triphenyl tetrazolium chloride, 10 g of agar, 1 l of water). For con¢rmation, colonies were tested on Bile EsculineAzide agar (BEA, Difco). The microcosm used to induce the VNC state in bacteria consisted of ¢ltered, sterilized lake water as described previously [8]. We consider that cells are in the VNC state when 10 ml of inoculated microcosm yield no colonies, while total cell number, evaluated with a Coulter Counter, remains unchanged [8]. 2.2. Collection and processing of water samples Two hundred ml water samples were taken from di¡erent sites in Lake Garda over two di¡erent periods, obtaining two di¡erent series: the ¢rst series (series a) was collected in July 1997 and comprised 29 samples and the second series (series b) consisted of 17 samples collected in June 1998. The sites for sample collection were chosen on account of their proximity to villages and public drains. In situ, 1 ml of each sample and di¡erent 10-fold

FEMSEC 1079 12-11-99

M.M. Lleo¨ et al. / FEMS Microbiology Ecology 30 (1999) 345^353

dilutions of it, were inoculated on TSA plates and incubated at 22³C and 35³C for 3 days. From each sample, an additional 10 ml were taken and ¢ltered on 0.22 Wm MF-Millipore ¢lters (Millipore Co., USA) and ¢lters were placed face up on m-Enterococcus plates and then incubated at 37³C and observed after 24 and 48 h [9]. As a con¢rmation procedure, colonies were transferred to a plate containing BEA medium and incubated at 37³C. The remaining volume was transported under refrigerated conditions to the laboratory and then particulate matter was concentrated approximately 200-fold by passing samples through 0.22 Wm Millipore ¢lters. Each ¢lter was then introduced into 20 ml of sterilized lake water and agitated to remove bacterial cells from the ¢lter. The bacterial suspension was then centrifuged at 20 000Ug for 15 min and the pellet suspended in 1.2 ml sterilized lake water. From each concentrated sample 100 Wl were inoculated on BEA plates and 10 Wl aliquots were stored to be used as templates in PCR reactions. The remaining 1 ml of concentrated sample was used for DNA extraction. 2.3. DNA extraction DNA was extracted from 1 ml of each concentrated sample. Brie£y, cell pellets obtained by centrifugation were resuspended on 1 ml SET bu¡er (20% sucrose, 50 mM EDTA, 50 mM Tris-HCl pH 7.6) and the suspensions were left on ice for 15 min. Subsequently, 35 Wl of lysozyme (5 mg ml31 ) were added, the suspension was mixed and left for an additional 15 min on ice. Then, 9 Wl of a 25% sodium dodecylsulfate (SDS) solution were added and incubated for 60 min at room temperature (RT). Proteinase K was added at a ¢nal concentration of 1 mg ml31 and incubated at RT for 60 min and then with 0.5 ml of 7.5 M ammonium acetate, mixed and incubated for an additional 15 min at RT. After centrifuging in an Eppendorf microfuge for 5 min, DNA remained in the supernatants from which it was precipitated in ethanol (2 volumes) and resuspended in 30 Wl TE bu¡er (10 mM Tris-HCl pH 8 and 1 mM EDTA). To evaluate the yield of DNA per cell, 20 ml of an E. faecalis SF97 culture at a concentration of 5U108 cells ml31 were pelleted. DNA was extracted as de-

347

scribed above and resuspended in 50 Wl of TE bu¡er. The DNA concentration was determined by co-migration on 0.8% agarose gel [20] of the E. faecalis chromosomal DNA with markers of standard concentrations prepared from a solution of lambda DNA 0.25 Wg Wl31 (Boehringer Mannheim, Germany). 2.4. Construction of internal standard The target sequence for PCR ampli¢cation was a 444 bp fragment located on the E. faecalis chromosome within a gene coding for penicillin binding protein 5 (PBP5) previously sequenced [21] and demonstrated to be species speci¢c [22]. The fragment was PCR ampli¢ed using two primers selected within the PBP5 gene: primer FWD: 5P CATGCGCAATTAATCGG 3P and primer IS: 5P CATAGCCTGTCGCAAAAC 3P. This fragment has two HindIII restriction sites (Fig. 1) that yielded three di¡erent fragments as visualized by 8% polyacrylamide gel electrophoresis: a 178 bp fragment containing the FWD priming region, a 194 bp fragment containing the IS priming region and a 72 bp internal fragment. Restriction fragments were ligated with T4 DNA ligase (Boehringer Mannheim) and then visualized in an 8% polyacrylamide gel [20]. One of the ligation products, a 372 bp fragment, was excised from the gel and left overnight in TE bu¡er as described by Leser [23]. The fragment obtained was ampli¢able by the FWD and IS primers, thus proving suitable for use as an internal standard. The 372 bp fragment was inserted on a plasmid suitable for linking PCR products, PCR ScriptTM plasmid (Stratagene, USA), and cloned on E. coli DH5K. The recombinant plasmid, pISM (Fig. 1), containing the 372 bp fragment, served as an internal standard reservoir. 2.5. DNA ampli¢cation Water samples or extracted DNA were used for PCR ampli¢cation on a Thermal Cycler (Perkin-Elmer, USA). Final concentrations of the di¡erent components in the PCR reaction were: 10 mM Tris-HCl pH 8.8, 50 mM KCl, 1.5 mM MgCl2 , 200 WM of each deoxynucleotide triphosphate, 100 pmol of each of the two primers described above and 2 U Taq DNA polymerase (Boehringer Mann-

FEMSEC 1079 12-11-99

348

M.M. Lleo¨ et al. / FEMS Microbiology Ecology 30 (1999) 345^353

heim). When cPCR was used, 2 Wl of appropriately diluted internal standard were added to the PCR mixture. PCR was conducted with 30 cycles consisting of 1.5 min denaturation at 94³C, 1.5 min annealing at 60³C and 2 min extension at 72³C with a ¢nal 5 min extension period at 72³C. 2.6. Gel electrophoresis and quanti¢cation Ten Wl of PCR products were separated by electrophoresis on 2.5% SeaKem ME (FMC Bio Products, USA) agarose gels in TAE bu¡er (0.04 M Tris-acetate and 1 mM EDTA pH 8) containing 0.5 Wg ml31 ethidium bromide. Bands were visualized with a UV transilluminator and photographed on Polaroid 667 ¢lm. To quantify PCR products, Polaroid ¢lms were digitized using a video camera (Image MasterTM, Pharmacia Biotech, Sweden) and analyzed with Image Master VDS software.

3. Results 3.1. Speci¢city and sensitivity of primers The FWD and IS primers produced the expected 444 bp fragment from exponentially growing E. faecalis strain SF97 genomic DNA. The same fragment was also ampli¢ed when a sample from a population of E. faecalis just reaching the VNC state was used, in a similar way to that previously described by us using a di¡erent pair of primers selected within the same E. faecalis gene [8] (Fig. 2). Speci¢city of the primers was tested using the DNA extracted from E. hirae ATCC 9790, E. durans, E. casseli£avius and from three di¡erent clinical isolates of E. faecium as a template. No PCR products were obtained in those cases (Fig. 2). Similarly, no products were obtained when primers were applied to a PCR reaction where the template consisted of a pool of DNA (2 ng of each) from di¡erent Gram-negative species, namely, E. coli, K. pneumoniae, E. aerogenes, A. sobria, S. marcescens, S. dysenteriae and P. £uorescens (Fig. 2). Using an annealing temperature of 60³C no background ampli¢cation was observed thus demonstrating that the DNA from eukaryotic and prokaryotic

Fig. 1. Diagram showing construction of an internal standard. The 444 bp fragment was digested with HindIII generating three fragments which were subsequently ligated with T4 ligase. The 377 bp fragment generated containing both priming regions was puri¢ed and cloned to be used as an internal standard.

cells other than E. faecalis which may be present in lake water did not interfere with the molecules contained in the PCR reaction. To test the sensitivity of the primers, 10-fold dilutions of the DNA extracted from a culture of E. faecalis were ampli¢ed by PCR. Ampli¢cation products visualizable on ethidium bromide-stained agarose gel were obtained when up to 0.1 pg of DNA (corresponding to approx. 2 cells ml31 ) was applied

FEMSEC 1079 12-11-99

M.M. Lleo¨ et al. / FEMS Microbiology Ecology 30 (1999) 345^353

Fig. 2. Ethidium bromide-stained gel with products from ampli¢cation by PCR of DNA from di¡erent bacterial species using FWD and IS primers. Reactions contained 2 ng of DNA from di¡erent bacterial species or serially diluted E. faecalis chromosomal DNA. Lane 1, E. faecalis SF97; lane 2, DNA pool from E. hirae ATCC 9790, E. durans ED5 and E. casseli£avius ECS2; lane 3, E. faecium DNA pool from strains VR1, VR5 and VR27; lane 4, E. faecalis SF97 in the VNC state; lane 5, DNA pool from di¡erent Gram-negative species; lanes 6 to 11, 10-fold dilutions (from 1 ng to 0.01 pg) of DNA extracted from E. faecalis SF97. M, molecular mass markers from Px-174-RF DNA HaeIII digest (Boehringer Mannheim).

to the PCR reaction (Fig. 2). The same experiment was done using 10-fold dilutions of an E. faecalis culture (108 cells ml31 ) instead of pure DNA: ampli¢cation products were obtained up to a cell concentration of 1 cell ml31 (data not shown). 3.2. Detection of E. faecalis speci¢c DNA sequences in water samples collected from Lake Garda In order to establish a protocol which proved as simple and sensitive as possible for samples to be submitted to PCR ampli¢cation, we used three different approaches which consisted in using: (i) lake water samples without any manipulation, (ii) concentrated (ca 200U) lake water samples, and (iii) DNA extracted, as speci¢ed in Section 2, from concentrated (ca 200U) lake water samples. The results obtained indicated that from 17 samples (those of the second series as indicated in Section 2) examined to detect speci¢c DNA by conducting PCR on the three kinds of materials, only two (12%) yielded positive signals when concentrated lake water samples were used as a template in PCR reactions. No positivity was found when unconcentrated water samples were used as a template, while eight samples (47%) tested positive when DNA was extracted and used as a template. These results indicate that extraction of DNA from samples was a necessary step to make the protocol su¤ciently sen-

349

sitive and capable of detecting even very small amounts of the E. faecalis speci¢c DNA sequence. We then examined the two di¡erent series of water samples collected from Lake Garda over two di¡erent periods; the results are shown in Table 1. The ¢rst series (a) was collected in July 1997 and consisted of 29 samples; of these, six samples (20%) presented colonies on m-Enterococcus plates and were positive for the presence of the E. faecalis PBP5 gene as revealed by PCR; 11 samples (38%) yielded no colonies on selective plates and were also negative for speci¢c DNA research by PCR; and 12 samples (41%) formed no colonies on nonselective and selective medium but tested positive for E. faecalis DNA by PCR (Table 1). The second series (b) consisted of 17 samples collected in June 1998. In this case, two samples (12%) belonged to the ¢rst class of samples described above and seven (41%) to the second class, while eight samples (47%) belonging to the third class, i.e. positive regarding PCR detection of speci¢c DNA, formed no colonies on BEA plates. The proportion of samples containing no culturable E. faecalis cells but testing positive at speci¢c DNA detection, as revealed by PCR, thus ranged from 40 to 47% in the two series of water samples from Lake Garda. 3.3. Quanti¢cation of E. faecalis DNA in the water from Lake Garda Prior to quantitative determination of E. faecalis speci¢c DNA by PCR in water samples from Lake Garda, we constructed a standard curve. To do this, we used the DNA extracted from a culture of E. faecalis strain SF97. Ten-fold dilutions were perTable 1 Detection of the presence of E. faecalis in 46 lake water samples by CFU count and PCR detection of speci¢c DNA Presence of

Number of samples (% of total)a

Colonies on m-Ent. plates

Speci¢c DNA by PCR

+ 3 3

+ 3 +

8 [6+2] (17) 18 [11+7] (39) 20 [12+8] (43)

a Numbers in brackets indicate the number of samples belonging to each of the two series.

FEMSEC 1079 12-11-99

350

M.M. Lleo¨ et al. / FEMS Microbiology Ecology 30 (1999) 345^353

Fig. 5. Ethidium bromide-stained gel with ampli¢cation products obtained by PCR on DNA samples extracted from E. faecalis cultures with 10-fold decreasing cell concentrations ranging from 106 to 1 cell ml31 (lanes 1 to 7). Fig. 3. Ethidium bromide-stained gel containing ampli¢cation products from competitive PCR. Reactions contained a constant amount of internal standard (approx. 150 pg) and serially diluted DNA extracted from an E. faecalis SF97 culture. Lane 1, 10 ng DNA; lane 2, 1 ng DNA; lane 3, 100 pg DNA; lane 4, 10 pg DNA; and lane 5, 1 pg DNA. M, molecular mass markers.

formed in order to test concentrations of DNA ranging from 100 ng to 1 pg and corresponding to cell concentrations between 2U106 and 20 cells ml31 . PCR reactions also contained a constant amount of internal standard calculated as described by Leser [23], i.e. an internal standard concentration that resulted in a nearly equimolar yield of the two fragments in cPCR with DNA from 103 to 104 bacterial cells per ml. Similar results were obtained when known numbers of E. faecalis cells, instead of pure DNA, were used as a template for cPCR. The PCR products were quanti¢ed from digitized images of

Fig. 4. Standard curve for cPCR for quantitative detection of E. faecalis DNA. The PCR products were quanti¢ed from digitized images of ethidium bromide-stained gels (see Fig. 3). Results are representative results of cPCR on three replicate DNA extractions from a culture inoculated with E. faecalis SF97 at a concentration of 108 cells ml31 .

ethidium bromide-stained 2.5% agarose gels (Fig. 3) and a standard curve was constructed (Fig. 4). As can be seen in Fig. 4, the standard curve was loglinear over three orders of magnitude of starting cell concentrations. This means that we could accurately calculate concentrations of DNA ranging from 10 pg to 10 ng, corresponding to cell concentrations between 2U102 and 2U105 ml31 . To test the accuracy of the DNA yield calculations, we conducted an experiment to measure the DNA yield when it was extracted from cultures with medium and low cell concentrations. In this experiment, PCR was conducted on di¡erent samples of DNA, each extracted from cultures with 10-fold decreasing cell concentrations. As is shown in Fig. 5, even when DNA was extracted from a culture with only theoretically 1 cell ml31 , the speci¢c 444 bp ampli¢cation band could be observed. Finally, it should be stressed here that to obtain accurate quanti¢cation of E. faecalis DNA, two dif-

Fig. 6. Ampli¢cation products from cPCR conducted on DNA extracted from 4 (1 to 4) water samples taken from Lake Garda. Two di¡erent volumes of the DNA solution (4 Wl (a) and 10 Wl (b)) were applied to the PCR reaction in order to accurately calculate the amount of DNA in each sample.

FEMSEC 1079 12-11-99

M.M. Lleo¨ et al. / FEMS Microbiology Ecology 30 (1999) 345^353

351

Table 2 Competitive PCR quanti¢cation of E. faecalis speci¢c DNA in lake water samples testing positive at PCR ampli¢cation Samplea

CFU ml31 on BEA plates

Amount of E. faecalis speci¢c DNA (ng) in the PCR mixture

Approx. cell conc. in the original water sample (cells ml31 )b

5a 6a 7a 8a 9a 10a 11a 12a 14a 17a 18a 21a 22a 27a 6b 8b 9b

0 0 0 1 0 5 0 0 0 0 0 0 0 0 0 0 0

1.4 1.5 6 1.7 0.1 0.5 1.8 1.3 0.1 9 0.5 1.2 1.6 1.5 1.5 1.6 0.6

3U102 3U102 1.4U103 4U102 24 1.2U102 4U102 3U102 24 2U103 1.2U102 3U102 3.8U102 3U102 3U102 3.8U102 1.4U102

a b

The number of the sample is followed by a letter indicating whether the sample belongs to the ¢rst (a) or to the second (b) series. A conversion factor of 1.5 was considered when number of nonculturable cells were estimated from DNA concentrations.

ferent volumes of TE bu¡er (4 and 10 Wl), in which DNA extracted from the water samples was suspended, were used as templates in cPCR reactions. Fig. 6 shows the ampli¢cation products obtained with cPCR from a number of lake water samples. Quanti¢cation of the DNA present in the samples is shown in Table 2. In this table the data are compared with those obtained using a standard method such as CFU ml31 counts. As can be observed in the various samples, E. faecalis speci¢c DNA amounts varied from 100 pg to 10 ng in the PCR mixture corresponding to 24 to 2U103 cells ml31 in the original sample volume which contained no cells capable of colony formation. Although it is important to point out that the cellular DNA content may vary considerably depending on the growth rate of the cells [24], when nonculturable cells were estimated from DNA concentrations an average DNA content of 1.5 molecules per cell was considered, on the basis of the data obtained by Shockman et al. [25] for enterococci.

4. Discussion In a previous study [8] we demonstrated that, as

described for Gram-negative species, a Gram-positive species, namely E. faecalis, is also able to enter the VNC state when it is maintained in the laboratory under low nutrient concentration conditions and at low temperatures. In this state, E. faecalis cells are viable and capable of resuming active growth, though only in some cases [8]. Because this bacterial species may be found in water [9], it would be interesting to analyze whether E. faecalis may also be present in viable but nonculturable forms in the natural environment. If this were true, it seems obvious that the classic methods used to monitor the microbiological quality of water may be insu¤cient in that they are able to reveal only culturable cells. Thus, new protocols based on molecular biology should be applied in order to reveal also nonculturable bacterial forms including those which are still viable. In this study we describe a cPCR protocol which is also capable of revealing the presence of DNA of nonculturable forms of E. faecalis in fresh water samples. In this case, environmental samples used to detect E. faecalis nonculturable cells were samples of water from Lake Garda. We chose this molecular method because PCR can amplify minimal amounts of a speci¢c DNA present in samples with complex

FEMSEC 1079 12-11-99

352

M.M. Lleo¨ et al. / FEMS Microbiology Ecology 30 (1999) 345^353

backgrounds consisting of many di¡erent prokaryotic and eukaryotic species. The PCR method designed to detect the presence of the E. faecalis species speci¢c gene coding for PBP5 proved highly speci¢c and sensitive enough to detect low amounts of DNA (up to 0.1 pg, corresponding to 2 cells ml31 ). The accuracy of the method is also assured by the fact that we were able to obtain ampli¢cation products when DNA was extracted from samples containing cell concentrations as low as 1 cell ml31 . Moreover, we encountered no di¡erences when amplifying DNA from nonculturable cells as compared to culturable ones, unlike the ¢ndings reported by Brauns et al. [14]. These authors failed to amplify DNA from nonculturable cells of V. vulni¢cus unless additional PCR cycles were employed. The reason for the decreased ampli¢cation of nonculturable forms could be a reduced DNA content per cell or also di¤culties in extracting DNA from nonculturable cells. However, the di¡erent approaches used may partly invalidate comparison of the results obtained here with those of Brauns. The use of an internal standard capable of competing with the E. faecalis DNA to bind to speci¢c primers enabled us to conduct a competitive PCR and to quantify E. faecalis cells in both culturable and nonculturable forms. This protocol was validated by detecting E. faecalis speci¢c DNA in water samples from Lake Garda. From the data obtained in this study it can be deduced that as many as 43% of the samples collected from di¡erent Lake Garda sites, which tested negative for the presence of E. faecalis cells when using a plating method, actually contained speci¢c E. faecalis DNA. It should be recalled here that the high percentage of positive samples recovered could be explained by the fact that water samples were obtained from `hot' sites where fecal contamination was more probable, i.e. in the vicinity of villages and public drains. We also considered the possibility that the absence of E. faecalis colonies on selective media was due to the presence in the sample of injured instead of nonculturable enterococci. This possibility was ruled out by examining selective and nonselective media: as is well known, injured bacteria can be recovered in nonselective media but do not grow on selective media [9]; on the contrary,

nonculturable bacteria do not form colonies on neither selective nor nonselective media [1]. This method thus reveals very useful and sensitive and speci¢c enough to detect the presence of enterococcal DNA contained in culturable and nonculturable cells in aquatic environments and in our opinion represents a good system to monitor microbiological quality of water. Furthermore, the application of cPCR to these samples enabled us to calculate that the amount of speci¢c DNA corresponds to E. faecalis cell concentrations ranging from 24 to 2U103 cells ml31 ; these values are much higher than the bacteriological recreational fresh water standards for enterococci established by the Canadian government (35 CFU per 100 ml) [9] or by the European Economic Community (100 CFU per 100 ml) [26]. The possibility of amplifying free E. faecalis DNA seems to us very unreasonable because in a complex environment, such as lake water, free DNA would be degraded very rapidly, i.e. within hours [14,27]. It is also important to stress that ampli¢cation products only signal that the appropriate target DNA sequences are present in the sample. The presence of the ampli¢cation products does not imply that the target organism is viable. In fact, it has recently been demonstrated that PCR can also amplify DNA from dead cells [28]. Although we have to take this eventuality into account, we also have to consider the possibility that some of these nonculturable cells may actually be viable and capable of resuming active growth, as already demonstrated in arti¢cial microcosms mimicking the physico-chemical conditions of the water of Lake Garda [8]. In our opinion, this potential hazard should not be underrated with a view to the microbiological quality of the water.

Acknowledgements This study was supported by grant 97.01061.PF49 from the Consiglio Nazionale delle Ricerche (CNR, Rome), Target Project on `Biotechnology' and by 1998 co¢nancing from the Ministero dell'Universita© e della Ricerca Scienti¢ca e Technologica (MURST, Rome).

FEMSEC 1079 12-11-99

M.M. Lleo¨ et al. / FEMS Microbiology Ecology 30 (1999) 345^353

References [1] Colwell, R.R. and Huq, H. (1994) Vibrios in the environment : viable but nonculturable Vibrio cholerae. In: Vibrio cholerae and Cholera : Molecular Global Perspectives (Kaye, T., Ed.), pp. 117^133. American Society for Microbiology, Washington, DC. [2] Oliver, J.D. (1993) Formation of viable but non culturable cells, In: Starvation in Bacteria (Kjelleberg, S., Ed.), pp. 239^272. Plenum Press, New York. [3] Rollins, D.M. and Colwell, R.R. (1986) Viable but nonculturable stage of Campylobacter jejuni and role in survival in the natural aquatic environment. Appl. Environ. Microbiol. 52, 531^538. [4] Wolf, P.W. and Oliver, J.D. (1992) Temperature e¡ects on the viable but non-culturable state of Vibrio vulni¢cus. FEMS Microbiol. Ecol. 101, 33^39. [5] Barer, M.R., Gribbon, L.T., Harwood, C.R. and Nwoguh, C.E. (1998) The viable but non-culturable hypothesis and medical bacteriology. Rev. Med. Microbiol. 4, 183^191. [6] Rahman, I., Shahamat, M., Kirchman, P.A., Rissek-Cohen, E. and Colwell, R.R. (1994) Methionine uptake and cythopathogenicity of viable but nonculturable Shigella dysenteriae type 1. Appl. Environ. Microbiol. 60, 3573^3578. [7] Shahamat, M., Mai, U., Paszko-Kolva, C., Kessel, M. and Colwell, R.R. (1993) Use of autoradiography to assess viability of Helicobacter pilori in water. Appl. Environ. Microbiol. 59, 1231^1235. [8] Lleo©, M.M., Ta¢, M.C. and Canepari, P. (1998) Nonculturable Enterococcus faecalis cells are metabolically active and capable of resuming active growth. Syst. Appl. Microbiol. 21, 333^339. [9] Toranzos, G.A. and Mc Feters, G.A. (1997) Detection of indicator microorganisms in environmental freshwaters and drinking waters. In: Manual of Environmental Microbiology (Hurst, C.J., Knudsen, G.R., McInerney, M.J., Stetzenbach, L.D. and Walter, M.V., Eds.), pp. 184^194. American Society for Microbiology, Washington, DC. [10] Dukan, S., Levi, Y. and Tonati, D. (1997) Recovery of culturability of an HOCl-stressed population of Escherichia coli after incubation in phosphate bu¡er: resuscitation or regrowth ? Appl. Environ. Microbiol. 63, 4204^4209. [11] Jians, X. and Chai, T. (1996) Survival of Vibrio parahaemolyticus at low temperatures under starvation conditions and subsequent resuscitation of viable, nonculturable cells. Appl. Environ. Microbiol. 62, 1300^1305. [12] Whitesides, M.D. and Oliver, J.D. (1997) Resuscitation of Vibrio vulni¢cus from the viable but nonculturable state. Appl. Environ. Microbiol. 63, 1002^1005. [13] Bej, A.K., Ste¡an, R.J., DiCesare, J., Ha¡, L. and Atlas, R.M. (1990) Detection of coliform bacteria in water by polymerase chain reaction and gene probes. Appl. Environ. Microbiol. 56, 307^314. [14] Brauns, L.A., Hudson, M.C. and Oliver, J.D. (1991) Use of polymerase chain reaction in detection of culturable and non

[15]

[16]

[17]

[18]

[19]

[20]

[21]

[22]

[23]

[24] [25]

[26]

[27]

[28]

353

culturable Vibrio vulni¢cus cells. Appl. Environ. Microbiol. 57, 2651^2655. Huq, A., Colwell, R.R., Rahman, R., Ali, A., Chowdhury, M.A.R., Parveen, S., Sack, D.A. and Russek-Cohen, E. (1990) Detection of Vibrio cholerae O1 in the aquatic environment by £uorescent-monoclonal antibody and culture methods. Appl. Environ. Microbiol. 56, 2370^2373. Islam, M.S., Hasan, M.K., Miah, M.A., Sur, G.C., Felsenstein, A., Venkatesan, M., Sack, R.B. and Albert, M.J. (1993) Use of polymerase chain reaction and £uorescent-antibody methods for detecting viable but nonculturable Shigella dysenteriae type 1 in laboratory microcosms. Appl. Environ. Microbiol. 59, 536^540. Tsai, Y.L. and Olson, B.H. (1992) Detection of low numbers of bacterial cells in soils and sediments by PCR. Appl. Environ. Microbiol. 58, 754^757. Leser, T.D., Boye, M. and Hendriksen, N.B. (1995) Survival and activity of Pseudomonas sp strain B13 (FR1) in a marine microcosm determined by quantitative PCR and an rRNAtargeting probe and its e¡ect on the indigenous bacterioplankton. Appl. Environ. Microbiol. 61, 1201^1207. Tiem, S.M., Krumme, M.L., Smith, R.L. and Tledje, J.M. (1994) Use of molecular techniques to evaluate the survival of a microorganism injected into an aquifer. Appl. Environ. Microbiol. 60, 1059^1067. Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) Molecular Cloning : A Laboratory Manual, 2nd Edn. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Signoretto, C., Boaretti, M. and Canepari, P. (1994) Cloning, sequencing and expression in Escherichia coli of the low-a¤nity penicillin binding protein of Enterococcus faecalis. FEMS Microbiol. Lett. 123, 99^106. Robbi, C., Signoretto, C., Boaretti, M. and Canepari, P. (1996) The gene encoding for penicillin binding protein 5 of Enterococcus faecalis is useful for development of a speciesspeci¢c DNA probe. Microb. Drug Resist. 2, 215^218. Leser, T.D. (1995) Quantitation of Pseudomonas sp strain B13 (FR1) in the marine environment by cPCR. J. Microbiol. Methods 22, 249^262. Slater, M. and Schaechter, M. (1974) Control of cell division in bacteria. Bact. Rev. 38, 199^221. Shockman, G.D., Daneo-Moore, L. and Higgins, M.L. (1974) Problems of cell wall and membrane growth, enlargement and division. Ann. N.Y. Acad. Sci. 235, 161^197. Gleick, P.H. (1993) Water quality and contamination. In: Water in Crisis, a Guide to the World's Fresh Water Resources (Gleick, P.H., Ed.), pp. 225^257. Oxford University Press, Oxford. Lorenz, M.G. and Wackernagel, W. (1994) Bacterial gene transfer by natural genetic transformation in the environment. Microbiol. Rev. 58, 563^602. Josephson, K.L., Gerba, C.P. and Pepper, I.L. (1993) Polymerase chain reaction detection of nonviable bacterial pathogens. Appl. Environ. Microbiol. 59, 3513^3515.

FEMSEC 1079 12-11-99