Comparison of real-time PCR with SYBR Green I or 5'-nuclease assays and dot-blot hybridization with rDNA-targeted oligonucleotide probes in quantification of selected faecal bacteria

Erja Malinen, Anna Kassinen, Teemu Rinttilä and Airi Palva

University of Helsinki, Faculty of Veterinary Medicine, Department of Basic Veterinary Sciences, Section of Microbiology, PO Box 57, FIN-00014 Helsinki University, Finland

Correspondence
Airi Palva
Airi.Palva{at}Helsinki.fi


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
PCR primers and hybridization probes were designed for the 16S rRNA genes of six bacterial species or groups typically present in human faeces or used in the dairy industry. The primers and probes were applied for quantification of the target bacterial genomes added in artificial DNA mixtures or faecal DNA preparations, using dot-blot hybridization and real-time PCR with SYBR Green I and TaqMan chemistries. Dot-blot hybridization with 33P-labelled oligonucleotide probes was shown to detect a 10 % target DNA fraction present in mixed DNA samples. Applicability of the rDNA-targeted oligonucleotide probes without pre-enrichment of the 16S gene pool by PCR was thus limited to the detection of the predominant microbial groups. Real-time PCR was performed using a 96-well format and was therefore feasible for straightforward analysis of large sample amounts. Both chemistries tested could detect and quantify a subpopulation of 0·01 % from the estimated number of total bacterial genomes present in a population sample. The linear range of amplification varied between three and five orders of magnitude for the specific target genome while the efficiency of amplification for the individual PCR assays was between 88·3 and 104 %. Use of a thermally activated polymerase was required with the SYBR Green I chemistry to obtain a similar sensitivity level to the TaqMan chemistry. In comparison to dot-blot hybridization, real-time PCR was easier and faster to perform and also proved to have a superior sensitivity. The results suggest that real-time PCR has a great potential for analysis of the faecal microflora.

Abbreviations: GI, gastrointestinal


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
The microflora of the human gastrointestinal (GI) tract is mainly composed of uncultured organisms (Langendijk et al., 1995; Suau et al., 1999). Although the host–microbe interactions in this environment are poorly understood, the importance of the GI microbes to the well-being of the host is evident. The information provided by culture-based methods is biased, therefore analyses of nucleic acids are required to obtain a better understanding of the ecology and impact of the GI microflora. rRNA- or DNA-based applications have become popular due to the accumulation of sequence data in public databases, which are the primary requirement for designing specific oligonucleotide primers or probes. Recent rRNA- or DNA-based studies on faecal microbes have applied cloning and sequencing of rDNA (Suau et al., 1999), denaturing- or thermal-gradient gel electrophoresis (Zoetendal et al., 1998; Satokari et al., 2001a, b; Walter et al., 2001; Favier et al., 2002; Heilig et al., 2002), fluorescence in situ hybridization (Langendijk et al., 1995; Harmsen et al., 2000) or dot-blot hybridization with rRNA-targeted oligonucleotide probes (Doré et al., 1998; Sghir et al., 2000; Hopkins et al., 2001; Marteau et al., 2001), and species- or group-specific PCR (Wang et al., 1996; Matsuki et al., 1999).

Without a doubt, the available microbiological methods provide valuable information on the faecal microbial composition. However, they are not optimal for monitoring quantitative changes, especially those of subdominant microbial species or groups. Therefore, we used real-time PCR with TaqMan probes (5'-nuclease assay) or SYBR Green I for the specific detection of selected bacteria from faecal DNA. Real-time PCR is based on the continuous monitoring of changes in fluorescence during PCR and, in contrast to the conventional end-point detection PCR, quantification occurs during the exponential phase of amplification. Thus, the bias often observed in the PCR template-to-product ratios (Suzuki & Giovannoni, 1996) is avoided. The TaqMan assay is based on measuring the fluorescence released during PCR as the 5'-nuclease activity of Taq DNA polymerase cleaves a dual-labelled fluorescent hybridization probe, designed to bind inside the amplified region, during primer extension (Heid et al., 1996). SYBR Green I is an intercalating dye that gives a fluorescent signal when bound to double-stranded DNA, while being otherwise virtually non-fluorescent. SYBR Green I lacks the specificity of the TaqMan assay, where the fluorescent signal is derived from a specific probe. However, SYBR Green I can be applied to the fluorescent monitoring of any amplification reaction, which gives more flexibility to the real-time PCR.

In this study, we designed specific rDNA-targeted oligonucleotide primers or probes for six bacterial species or groups and applied them to the detection and quantification of target bacteria from faecal DNA with the two real-time PCR applications discussed above and rDNA-targeted dot-blot hybridization. Five faecal target bacteria were chosen for the assays. Bacteroides fragilis represented the Bacteroides group and Ruminococcus productus represented the Clostridium coccoides group; these two groups have been reported to be abundant in human faeces (Sghir et al., 2000). Bifidobacterium longum, one of the most common bifidobacterial species in human faecal samples (Mangin et al., 1994; Matsuki et al., 1999; Malinen et al., 2002), was also chosen for the assay. Although enterobacteria and lactobacilli form only a minor part of the GI microflora (Sghir et al., 2000), Escherichia coli and Lactobacillus acidophilus assays were also included. The remaining target species, Bifidobacterium lactis, does not belong to the human GI microflora but is increasingly used in the dairy industry as a probiotic bacterium.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Bacterial strains and culture conditions.
The strains used as positive and negative controls in this study were: Bacteroides fragilis DSM 2151T; Bifidobacterium lactis DSM 10140T; Bifidobacterium longum DSM 20219T; E. coli DSM 6897; L. acidophilus ATCC 4356T; R. productus DSM 2950T. E. coli was grown in Luria broth (Difco) at 37 °C with shaking (220 r.p.m.). Bacteroides fragilis and R. productus were grown in fastidious anaerobe broth (LabM), Bifidobacterium lactis and Bifidobacterium longum were grown in MRS broth or on agar supplemented with 0·05 % cysteine (Difco) and L. acidophilus was grown in MRS broth (Difco) at 37 °C in an anaerobic chamber (Concept Plus Anaerobe Work Station; Ruskinn Technology).

Extraction and purification of DNA from faeces and pure cultures.
Bacterial DNA was isolated from pure cultures of the reference strains and from faecal samples obtained from healthy volunteers. Removal of undigested particles from the faeces was performed by washing the samples with repeated low-speed centrifugations, followed by collection of the bacteria from combined wash supernatants with high-speed centrifugation (Apajalahti et al., 1998). Cell lysis and DNA extraction from faecal bacterial pellets and pure cultures were accomplished with a combination of physical, chemical and enzymic steps as described by Apajalahti et al. (1998). DNA concentrations were measured with a Versafluor fluorometer (Bio-Rad). This method is based on the quantification of the intensity of Hoechst 33258 dye (bisBenzimide H 33258), which is fluorescent when bound to double-stranded DNA, with an excitation wavelength of 360 nm and an emission wavelength of 460 nm. Due to the high sensitivity of Hoechst 33258, the amount of DNA required for concentration measurements is reduced and a linear quantification of double-stranded DNA can be obtained in the range of 20 ng to 10 mg.

Design of oligonucleotide primers and probes.
Hybridization probes and real-time PCR primers were designed for six bacterial species or groups, representing faecal microbes or bacteria used in the dairy industry. The on-line Internet tools CLUSTAL W (Thompson et al., 1994) and FASTA3 (Pearson & Lipman, 1988) provided by the European Bioinformatics Institute (http://www.ebi.ac.uk) as well as the PROBE MATCH and HIERARCHY BROWSER interfaces provided by the Ribosomal Database Project (Maidak et al., 2001; http://rdp.cme.msu.edu/html/) were utilized for identifying the primers and probes that were able to bind with the desired specificity to the rDNA of the selected target bacteria. The Bacteroides fragilis and Bifidobacterium lactis assays were designed to be species-specific, whereas the remaining four assays were set to detect closely related bacterial species. According to the sequence analyses, the Bifidobacterium longum assay detected Bifidobacterium infantis, Bifidobacterium longum, Bifidobacterium pseudolongum and Bifidobacterium suis. A recent study by Sakata et al. (2002) has suggested the unification of the species Bifidobacterium infantis, Bifidobacterium longum and Bifidobacterium suis into a single species. The E. coli assay targeted the Escherichia subgroup, composed of E. coli, Hafnia alvei and Shigella spp. Target bacteria for the L. acidophilus assay were L. acidophilus, Lactobacillus amylovorus, Lactobacillus amylolyticus, Lactobacillus crispatus, Lactobacillus gasseri and Lactobacillus johnsonii. The R. productus assay was specific for R. productus and C. coccoides. Base sequences and modifications of the oligonucleotides used in dot-blot hybridization and real-time PCR are shown in Tables 1 and 2, respectively. Whenever possible, dot-blot hybridization was performed with one of the PCR primers used in real-time PCR detection of the same target species or group. Specificity of the real-time PCR was obtained with one or two primers selective for the intended target(s) while the TaqMan probes used had broader ranges of target sequences and could therefore be used in conjunction with various primer sets.


View this table:
[in this window]
[in a new window]
 
Table 1. Dot-blot hybridization probes and washing temperatures used in this study

 
Dot-blot hybridization.
The oligonucleotide probes (Table 1) were 5'-end labelled with [{gamma}-33P]ATP (Amersham Pharmacia Biotech) using T4 polynucleotide kinase (New England Biolabs). The labelling reaction was incubated for 1 h at 37 °C in a total volume of 50 µl, consisting of 1xT4 polynucleotide kinase buffer (70 mM Tris/HCl, 10 mM MgCl2, 5 mM DTT, pH 7·6), 20 U T4 polynucleotide kinase, 1 µM [{gamma}-33P]ATP [specific activity >3000 Ci mmol-1 (>111 TBq mmol-1)] and 1 µM probe. Following the incubation, the reactions were extracted with phenol/chloroform and the unattached label was removed by using NAP 5-columns according to the manufacturer's instructions (Amersham Pharmacia Biotech). Efficiency of labelling was confirmed by liquid scintillation measurements.

Stringent washing temperatures (Ti) for the oligonucleotides were determined as the half-denaturation temperature by testing temperatures at 2·5 °C intervals ranging from calculated Tm -10 °C to Tm +10 °C, and 100 °C using the equation Tm=81·5 °C+16·6 °Cxlog10[Na+]+0·41x(mol% G+C)-600/N, where N is the number of nucleotides and the sodium ion concentration [Na+] is 0·39 M (Sambrook et al., 1989).

The dot-blot hybridization was performed in the following way. DNA samples were denatured using alkali and heat and were blotted onto a positively charged nylon membrane (Boehringer Mannheim). Pre-hybridization at 65 °C for 1 h in a buffer containing 5xDenhart's solution, 5xSSC, 0·5 % SDS and 100 µg denatured Herring Sperm DNA ml-1 (Sigma) was followed by the addition of the 5'-end labelled oligonucleotide probe to a final concentration of 7x10-12 mol ml-1. The hybridization was continued for 15 min at 65 °C, after which the hybridization oven was set to 30 °C and the hybridization was continued with declining temperature until the hybridization temperature had been under Ti for 2 h. The membranes were washed with a buffer containing 2xSSC and 0·5 % SDS, four times at room temperature (5 min) and once at Ti (15 min), followed by a rinse at room temperature. The washed membranes were wrapped in heat-sealed plastic bags and the dots were quantified using the Multi-Imager with IMAGING SCREEN BI and QUANTITY ONE software (Bio-Rad).

Real-time PCR.
Performance and optimal annealing temperatures of the PCR primers (Table 2) were first tested with gradient PCR (Peltier Thermal Cycler PTC-200; MJ Research). The standard reaction mixture consisted of 10 mM Tris/HCl (pH 8·8), 150 mM KCl, 0·1 % Triton X-100, 1·5 mM MgCl2, 200 µM each dNTP, 1 µM each primer and 0·03 U Dynazyme Polymerase II µl-1 (Finnzymes). The following PCR conditions were used. An initial DNA denaturation step at 95 °C for 5 min was followed by 30 cycles of denaturation at 95 °C for 30 s, primer annealing at 50–70 °C for 20 s and primer extension at 72 °C for 45 s, with a final extension step at 72 °C for 5 min. The specificity of the primers was confirmed against the non-target test bacteria; otherwise, results from the sequence searches against prokaryotic DNA databases were relied upon. PCR products were analysed on agarose gels by electrophoresis.


View this table:
[in this window]
[in a new window]
 
Table 2. PCR primers and 5'-nuclease assay probes used in this study

Details of the species detected by the different probes are given in the footnotes to Table 1. F and R indicate forward and reverse primers, respectively; the other sequences shown represent the probes.

 
Quantitative PCR was performed using an iCycler iQ apparatus (Bio-Rad) associated with the ICYCLER OPTICAL SYSTEM INTERFACE software (version 2.3; Bio-Rad). All PCRs were performed in triplicate in a volume of 25 µl, using 96-well optical-grade PCR plates and an optical sealing tape (Bio-Rad).

Optimal concentrations for various reaction components were tested for each primer set and chemistry with a dilution series of genomic DNA from the target test species. For the SYBR Green I chemistry, the effect of the polymerase type on PCR product formation was tested with a standard polymerase, Dynazyme II (Finnzymes), and the hot-start polymerases AmpliTaq Gold DNA Polymerase (Applied Biosystems), BlueTaq (Euroclone) and FastStart Taq DNA Polymerase (Roche).

The TaqMan assays were performed successfully with Dynazyme II; therefore, switching to a hot-start enzyme was not considered. Optimal reaction components and their concentrations for each PCR assay and chemistry (a total of 12 different reaction mixtures) are summarized in Table 3.


View this table:
[in this window]
[in a new window]
 
Table 3. Optimized reaction conditions for SYBR Green I (SG) and TaqMan (TM) assays with the six oligonucleotide sets used

 
Reaction mixtures for the optimized SYBR Green I-based assays consisted of a 1 : 75 000 dilution of SYBR Green I (Molecular Probes), 10 mM Tris/HCl (pH 8·3), 50 mM KCl, 0·01 % Tween 20, 2–3 mM MgCl2, 200 µM each dNTP, 0·5 µM each primer, 0·08 U BlueTaq µl-1 and either 10 µl of template or water (see Table 3 for detailed information on the concentrations of the reaction components). The thermal cycling conditions used were an initial DNA denaturation step at 95 °C for 10 min followed by 35 cycles of denaturation at 95 °C for 15 s, primer annealing at the optimal temperature (Table 3) for 20 s, extension at 72 °C for 30 s, and an additional incubation step at 85 °C for 30 s to measure the SYBR Green I fluorescence. Finally, melt-curve analysis was performed by slowly heating the PCRs to 95 °C (0·3 °C per cycle) with simultaneous measurement of the SYBR Green I signal intensity. When Dynazyme II was used the duration of the initial denaturation step was reduced to 4 min.

The reaction mixtures of the 5'-nuclease (TaqMan) assays consisted of 10 mM Tris/HCl (pH 8·8), 150 mM KCl, 0·1 % Triton X-100, 1·5–6·0 mM MgCl2, 200–300 µM each dNTP, 1 µM each primer, 80–400 nM fluorescent probe, 0·024–0·036 U Dynazyme II µl-1 and either 10 µl of template or water. The following thermal cycling parameters were used. An initial DNA denaturation at 95 °C for 4 min was followed by 30–40 cycles of denaturation at 95 °C for 15 s and a combined incubation step for primer annealing and extension at 58–60 °C, during which the fluorescent signal was measured.

Quantification of target bacterial DNA in mixed DNA samples.
Specificity and sensitivity of the dot-blot hybridization probes were studied by using 0, 10, 30, 90, 270 and 810 ng of the target strain genomic DNA for hybridization and by using the same amounts of target DNA mixed in 500 and 1000 ng of faecal DNA. Pooled DNA samples (1350 ng) from the genomic DNA of the non-target bacteria were used as negative controls. Faecal DNA samples (500 and 1000 ng) without added DNA were used for quantification of the target bacteria present in faeces.

In the real-time PCR assays, 20 ng of pooled DNA from non-target test bacteria or 20 ng of faecal DNA was applied in the PCRs together with a 10-fold dilution series of between 10 ng and 0·1 pg (approx. 2–5x106 to 10–50 target genomes) of genomic DNA from the target species. Faecal DNA samples (20 ng) without added DNA were used for quantification of the target bacteria present in faeces.

Additionally, real-time PCR was used for the quantification of pure cultures of bacterial cells that had been introduced into faecal samples. Bacterial densities in the pure cultures were estimated either by viable counts or by microscopy, after which replicate faecal samples were spiked with 109, 108, 107, 106, 105 or 0 pure cultured cells of each test species used. Collection of bacteria, lysis of the cells and isolation of DNA were preformed as described in ‘Extraction and purification of DNA from faeces and pure cultures’.


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Characterization of the 16S rDNA primers and probes for target bacteria
To compare real-time PCR and dot-blot hybridization for analysis of faecal bacteria, PCR primers and hybridization probes, targeted to the detection of 16S rDNA, were designed for six bacterial groups or species (Tables 1 and 2) using sequence databases and analysis tools available on the Internet. The specificity and sensitivity of the dot-blot hybridization probes and PCR primers were tested with Bacteroides fragilis, Bifidobacterium lactis, Bifidobacterium longum, E. coli, L. acidophilus and R. productus.

Results from dot-blot hybridizations with dilution series from the genomic DNA of test target species are shown in Table 4. The expected specificity was obtained but the sensitivity of the test was rather low (Table 4). With the Bifidobacterium lactis, Bifidobacterium longum, L. acidophilus and R. productus probes, 30 ng of genomic DNA extracted from the target species were required for the generation of a hybridization signal that was distinguishable from the background signal. This corresponds to the application of approximately 107 target genomes to the hybridization. With the E. coli and Bacteroides fragilis probes, 90 and 270 ng, respectively, of genomic DNA were required for the generation of the positive hybridization signal (Table 4).


View this table:
[in this window]
[in a new window]
 
Table 4. Detection of target bacteria using dot-blot hybridization with the six oligonucleotide probes designed

 
Specificity of the PCR assays (Table 2) was first confirmed by using conventional end-point PCR and gel electrophoresis. The assays were positive for the intended target species, resulting in the formation of PCR products of the expected sizes. No PCR products were observed with the non-target test species (data not shown).

Applicability of dot-blot hybridization for the detection of rDNA from faeces
To test whether target DNA could be reliably detected in DNA of faecal origin by dot-blot hybridization, reconstruction experiments were carried out. Analyses of artificial DNA mixtures were performed by adding 10, 30, 90 and 270 ng of target DNA to 500 and 1000 ng faecal DNA. Hybridization results from mixed DNA samples containing 1000 ng faecal DNA are shown in Fig. 1. The presence of additional faecal DNA was shown not to affect the sensitivity of the hybridization probes.



View larger version (18K):
[in this window]
[in a new window]
 
Fig. 1. Recovery of target DNA added to faecal DNA samples by dot-blot hybridization. Genomic DNA from the target bacterium was added in quantities of 10, 30, 90 and 270 ng to 1 µg of faecal DNA. {blacksquare}, Bacteroides fragilis assay; {triangleup}, Bifidobacterium lactis assay; {bullet}, Bifidobacterium longum assay; {blacktriangleup}, E. coli assay; {circ}, L. acidophilus assay; {square}, R. productus assay. Mean values ±SD values for three parallel samples are shown.

 
Applicability of real-time PCR for the detection of rDNA
SYBR Green I and TaqMan chemistries were compared in the real-time PCR, using six designed primer sets (Table 2). Each PCR assay was first optimized for the reaction conditions, by using the amplification efficiency, linearity and specificity as criteria for the selection of the final conditions (Table 3). Standard curves of the optimized assays with the SYBR Green I and TaqMan chemistries are shown in Fig. 2(a, b). The linear range of amplification varied between 0·1–1 pg and 1–10 ng of specific target genome, which corresponds to approximately 20–400 to 2x105–4x106 genomes, depending on the genome size of the target species. Standard curves had correlation coefficient values of between 0·979 and 0·998. The efficiencies of amplification for the individual assays, calculated from the formula Eff(n)=[10(-1/slope)-1], were between 88·3 and 100·1 % for the SYBR Green I assay, and between 91·0 and 104·0 % for the TaqMan assay. For the SYBR Green I chemistry, melt-curve analysis was used to detect the formation of the correct amplification product. While the PCR amplicons had Tm values of between 80 and 90 °C, primer dimers with lower Tm values were observed when small amounts of template DNA were used in PCRs with a basic polymerase (Fig. 3). To increase the sensitivity of the SYBR Green I assay, three hot-start polymerases provided by commercial suppliers were tested for their ability to prevent the formation of primer dimers. Although all of the tested hot-start polymerases reduced the amounts of primer dimers formed (data not shown), BlueTaq was shown to efficiently prevent the primer dimerization reactions occurring (Fig. 3) and was therefore used in the optimized SYBR Green I assay.



View larger version (16K):
[in this window]
[in a new window]
 
Fig. 2. Sensitivity and relationship of the observed threshold cycle to the target DNA starting quantity in the real-time PCR with the (a) SYBR Green I and (b) 5'-nuclease (TaqMan) chemistries. Purified genomic DNA from the test target species was used as the template in quantities of 10 000, 1000, 100, 10 and 1 pg. {blacksquare}, Bacteroides fragilis assay; {triangleup}, Bifidobacterium lactis assay; {bullet}, Bifidobacterium longum assay; {blacktriangleup}, E. coli assay; {circ}, L. acidophilus assay; {square}, R. productus assay.

 


View larger version (20K):
[in this window]
[in a new window]
 
Fig. 3. Effect of the polymerase type for the formation of PCR products and primer dimers with small amounts (1 pg) of specific template DNA in Bacteroides fragilis assay with SYBR Green I. A post-amplification melt-curve analysis, performed by plotting the negative value of the change in rate of fluorescence (-dRFU/dT) against temperature, is shown. {bullet}, Dynazyme II; {circ}, BlueTaq (hot-start polymerase). The mean values ±SD values from two replicates are displayed.

 
The detection of target DNA that had been introduced into mixed DNA from the non-target test bacteria by the SYBR Green I and TaqMan assays is shown in Fig. 4(a, b). A 1 pg addition of target DNA in 20 ng mixed DNA samples could be detected reliably with both methods. Fig. 5 demonstrates the quantification of target DNAs added in 20 ng faecal DNA preparations using the SYBR Green I (Fig. 5a) and TaqMan (Fig. 5b) assays. Generally, both real-time PCR assays gave consistent results for each target bacteria. The only exception was seen with the assays for R. productus, in which TaqMan gave almost a 100-fold higher estimate for the amount of target template originally present in the faeces than SYBR Green I (Fig. 5a, b). Repeating the assays did not change the results obtained; therefore, these contradicting results could have been caused by the less stringent conditions (higher Mg2+, dNTP and primer concentrations) used in the TaqMan assay compared to the SYBR Green I assay, resulting in suboptimal binding of the primers to templates with mismatched bases. As the TaqMan probe used was designed to bind to a broad range of target species, detection of falsely primed amplification products could have been possible.



View larger version (28K):
[in this window]
[in a new window]
 
Fig. 4. Detection of specific target DNAs mixed in non-target bacterial DNA from the other test species with the (a) SYBR Green I and (b) 5'-nuclease (TaqMan) assays. Purified genomic DNA from the test target species was added to 20 ng of non-target DNA pools in quantities of 10 000, 1000, 100, 10, 1 and 0·1 pg. {blacksquare}, Bacteroides fragilis assay; {triangleup}, Bifidobacterium lactis assay; {bullet}, Bifidobacterium longum assay; {blacktriangleup}, E. coli assay; {circ}, L. acidophilus assay; {square}, R. productus assay.

 


View larger version (31K):
[in this window]
[in a new window]
 
Fig. 5. Detection of specific target DNAs mixed in faecal DNA with the (a) SYBR Green I and (b) 5'-nuclease (TaqMan) assays. Purified genomic DNA from the test target species was added to 20 ng of faecal DNA in quantities of 10 000, 1000, 100, 10, 1 and 0·1 pg. {blacksquare}, Bacteroides fragilis assay; {triangleup}, Bifidobacterium lactis assay; {bullet}, Bifidobacterium longum assay; {blacktriangleup}, E. coli assay; {circ}, L. acidophilus assay; {square}, R. productus assay.

 
Reconstruction assays were performed by spiking faecal samples with a dilution series from each test bacteria; this was done before DNA isolation and purification. A preliminary test was performed for the detection and quantification of Bifidobacterium lactis using the TaqMan assay with the Bifidobacterium lactis PCR primer set (Table 2). Quantification of the target bacteria was performed successfully when 3·1x107 c.f.u. Bifidobacterium lactis cells were introduced into 1 g of faeces. According to real-time PCR results, parallel DNA preparations contained 1·87x107 (0·705±0·924 pg target DNA in 100 ng faecal DNA) and 8·94x107 (2·99±0·933 pg target DNA in 100 ng faecal DNA) target cells. The TaqMan and SYBR Green I assays were used to detect the presence of four other test bacteria in the samples, using the primers and probes listed in Table 2. Results from these assays are shown for E. coli and L. acidophilus in Fig. 6(a, b). For E. coli and L. acidophilus, the addition of 106 cells to a faecal sample could be demonstrated (Fig. 6a, b). However, Bacteroides fragilis and Bifidobacterium longum could also be detected in the faeces, due to the high numbers of these organisms in the unspiked control samples (approx. 108–109 cells); the addition of 105–109 target cells could not be demonstrated for these bacteria (data not shown). R. productus was omitted from the reconstruction assays, due to the contradictory results obtained for this organism with the TaqMan and SYBR Green I assays.



View larger version (14K):
[in this window]
[in a new window]
 
Fig. 6. Detection of specific target DNAs with the SYBR Green I ({blacktriangleup}) or 5'-nuclease (TaqMan; {square}) assays in a reconstruction assay with 109, 108, 107, 106 or 105 target bacterial cells introduced into replicate faecal samples prior to the collection of bacterial cells and isolation of DNA. Results are shown for (a) E. coli and (b) L. acidophilus. The mean values ±SD values from two replicates are displayed.

 

   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Direct nucleic-acid-level detection of bacteria is required for studying the GI microflora, which consists of a diverse population of bacterial species. Approximately 1011–1012 bacterial cells are present in 1 g of faeces, thus providing a challenging starting material for molecular analyses. In this study, we compared a dot-blot hybridization approach with two real-time PCR applications for the detection and quantification of target rDNAs in mixed DNA samples or faeces. For this purpose, oligonucleotide primers and probes were designed for selected representatives of the faecal microflora or for bacteria used by the dairy industry. Sequencing of 16S rDNA clone libraries has revealed an abundance of previously unknown microbial species to be present in the faecal microflora (Suau et al., 1999; Leser et al., 2002). Although the accumulation of sequence data from uncultured organisms has given a better insight into the GI microflora, an exact determination of primer or probe specificities, such as for the primers and probes used in this study, can be considered virtually impossible due to the diversity of the target ecosystem. Our aim was, however, to develop tools for studying the ecology of the GI microflora, with the focus of our assays, in practice, being on the detection and quantification of spatial, temporal or health-status-induced differences between or within populations.

Dot-blot hybridization with rRNA-targeted oligonucleotide probes is a well-established method for studying faecal microbes (Doré et al., 1998; Sghir et al., 2000; Hopkins et al., 2001; Marteau et al., 2001). rRNA is present in abundance in bacterial cells, thus increasing the sensitivity of the dot-blot hybridization to a reasonable level. In this study, dot-blot hybridization was performed with rDNA-targeted oligonucleotide probes to quantify the same starting material as was used for the real-time PCR applications tested here. The dot-blot hybridization could, at best, detect a 3 % DNA subpopulation within mixed DNA samples (Fig. 1). A similar sensitivity level has been reported by Muttray & Mohn (2000) for measurements of rDNA-to-rRNA ratios, thus the result was as expected, being approximately 10-fold less sensitive than what is obtained with the rRNA-targeted oligonucleotides (Sghir et al., 2000).

Real-time PCR applications have been described for the quantification of total bacteria (Suzuki et al., 2000; Bach et al., 2002; Nadkarni et al., 2002) and archaea (Suzuki et al., 2000) present in various environments, as well as for the sensitive and accurate detection and quantification of pathogenic bacteria (Nogva et al., 2000; Hein et al., 2001; Ge et al., 2001). To our knowledge, real-time PCR assays for studying members of the normal GI tract microflora have not been described. The sensitivity levels of the real-time PCR assays described in this study were between 200 and 400 target bacteria in pure cultures or mixed DNA samples, depending on the genome size of the target bacterium (Figs 4 and 5). Results obtained by spiking faeces with various amounts (109–105 cells g-1) of target bacteria (Fig. 6a, b) showed that real-time PCR is suitable for the analysis of real samples and is highly sensitive.

The SYBR Green I and TaqMan assays used here were both shown to be sensitive and specific methods for the quantification of target bacteria. Hein et al. (2001) used a hot-start enzyme to prevent primer dimer formation in SYBR Green I-based assays, which led to subsequent improvements in the sensitivity of these assays; however, an even better result was reported for TaqMan-based assays using the same hot-start polymerase. Here, it was found that the sensitivities of both chemistries were alike, but a hot-start polymerase was used in the SYBR Green I assays and a cost-effective standard polymerase was used in the TaqMan assays. Although for the TaqMan-based assays this probably led to a somewhat less sensitive result than would have been achieved using a hot-start enzyme, the acquired level of sensitivity, providing a means for the quantification of a 0·01 % subpopulation in a DNA sample, was considered sufficient for studying the GI tract ecology. Considering the inaccuracies that are likely to be introduced into results by the processes used to isolate DNA from population samples and by the subsequent measurement of the concentration of template DNA present in a given sample, the observed accuracy level of real-time PCR in the quantification of target genomes present in the in vitro DNA mixtures or faecal samples with added target cells was satisfactory (Figs 4, 5 and 6). Genome sizes and rRNA gene copy numbers are also likely to contribute to the amplification of different target bacteria.

When compared to rDNA-targeted dot-blot hybridization, real-time PCR had a superior sensitivity and also proved to be a more convenient and less expensive method for the quantification of selected bacterial populations. While a reasonable level of sensitivity was obtained here with the TaqMan chemistry using a standard, non-hot-start polymerase, the non-selective detection of any double-stranded DNA molecule by SYBR Green I-based assays can also be considered useful, especially when common primers are used for the quantification of diverse target DNA populations that have no suitable conserved sequence for the design of a detection probe.


   ACKNOWLEDGEMENTS
 
We thank Sinikka Ahonen for technical assistance and Dr Ilkka Palva for valuable discussions. This work was supported by the National Technology Agency of Finland (TEKES).


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Apajalahti, J. H. A., Särkilahti, L. K., Mäki, B. R. E., Heikkinen, J. P., Nurminen, P. H. & Holben, W. E. (1998). Effective recovery of bacterial DNA and percent-guanine-plus-cytosine-based analysis of community structure in the gastrointestinal tract of broiler chickens. Appl Environ Microbiol 64, 4084–4088.[Abstract/Free Full Text]

Bach, H.-J., Tomanova, J., Schloter, M. & Munch, J. C. (2002). Enumeration of total bacteria and bacteria with genes for proteolytic activity in pure cultures and in environmental samples by quantitative PCR mediated amplification. J Microbiol Methods 49, 235–245.[CrossRef][Medline]

Doré, J., Sghir, A., Hannequart-Gramet, G., Corthier, G. & Pochart, P. (1998). Design and evaluation of a 16S rRNA-targeted oligonucleotide probe for specific detection and quantitation of human faecal Bacteroides populations. Syst Appl Microbiol 21, 65–71.[Medline]

Favier, C. F., Vaughan, E. E., De Vos, W. M. & Akkermans, A. D. L. (2002). Molecular monitoring of succession of bacterial communities in human neonates. Appl Environ Microbiol 68, 219–226.[Abstract/Free Full Text]

Ge, Z., White, D. A., Whary, M. T. & Fox, J. G. (2001). Fluorogenic PCR-based quantitative detection of a murine pathogen, Helicobacter hepaticus. J Clin Microbiol 39, 2598–2602.[Abstract/Free Full Text]

Harmsen, H. J. M., Gibson, G. R., Elfferich, P., Raangs, G. C., Wildeboer-Veloo, A. C. M., Argaiz, A., Roberfroid, M. B. & Welling, G. W. (2000). Comparison of viable cell counts and fluorescence in situ hybridization using specific rRNA-based probes for the quantification of human fecal bacteria. FEMS Microbiol Lett 183, 125–129.[CrossRef][Medline]

Heid, C. A., Stevens, J., Livak, K. J. & Williams, P. M. (1996). Real time quantitative PCR. Genome Res 6, 986–994.[Abstract]

Heilig, H. G. H. J., Zoetendal, E. G., Vaughan, E. E., Marteau, P., Akkermans, A. D. L. & de Vos, W. M. (2002). Molecular diversity of Lactobacillus spp. and other lactic acid bacteria in the human intestine as determined by specific amplification of the 16S ribosomal DNA. Appl Environ Microbiol 68, 114–123.[Abstract/Free Full Text]

Hein, I., Lehner, A., Rieck, P., Klein, K., Brandl, E. & Wagner, M. (2001). Comparison of different approaches to quantify Staphylococcus aureus cells by real-time quantitative PCR and application of this technique for examination of cheese. Appl Environ Microbiol 67, 3122–3126.[Abstract/Free Full Text]

Hopkins, M. J., Sharp, R. & Macfarlane, G. T. (2001). Age and disease related changes in intestinal bacterial populations assessed by cell culture, 16S rRNA abundance, and community cellular fatty acid profiles. Gut 48, 198–205.[Abstract/Free Full Text]

Langendijk, P. S., Schut, F., Jansen, G. J., Raangs, G. J., Kamphuis, G. R., Wilkinson, M. H. & Welling, G. W. (1995). Quantitative fluorescence in situ hybridization of Bifidobacterium spp. with genus-specific 16S rRNA-targeted probes and its application in fecal samples. Appl Environ Microbiol 61, 3069–3075.[Abstract]

Leser, T. D., Amenuvor, J. Z., Jensen, T. K., Lindecrona, R. H., Boye, M. & Møller, K. (2002). Culture-independent analysis of gut bacteria: the pig gastrointestinal tract microbiota revisited. Appl Environ Microbiol 68, 673–690.[Abstract/Free Full Text]

Maidak, B. L., Cole, J. R., Lilburn, T. G. & 7 other authors (2001). The RDP-II (Ribosomal Database Project). Nucleic Acids Res 29, 173–174.[Abstract/Free Full Text]

Malinen, E., Mättö, J., Salmitie, M., Alander, M., Saarela, M. & Palva, A. (2002). PCR–ELISA. II: Analysis of Bifidobacterium populations in human faecal samples from a consumption trial with Bifidobacterium lactis Bb-12 and a galacto-oligosaccharide preparation. Syst Appl Microbiol 25, 249–258.[Medline]

Mangin, I., Bourget, N., Bouhnik, Y., Bisetti, N., Simonet, J.-M. & Decaris, B. (1994). Identification of Bifidobacterium strains by rRNA gene restriction patterns. Appl Environ Microbiol 60, 1451–1458.[Abstract]

Marteau, P., Pochart, P., Doré, J., Béra-Maillet, C., Bernalier, A. & Corthier, G. (2001). Comparative study of bacterial groups within the human cecal and fecal microbiota. Appl Environ Microbiol 67, 4939–4942.[Abstract/Free Full Text]

Matsuki, T., Watanabe, K., Tanaka, R., Fukuda, M. & Oyaizu, H. (1999). Distribution of bifidobacterial species in human intestinal microflora examined with 16S rRNA-gene-targeted species-specific primers. Appl Environ Microbiol 65, 4506–4512.[Abstract/Free Full Text]

Muttray, A. F. & Mohn, W. W. (2000). Quantitation of the population size and metabolic activity of a resin acid degrading bacterium in activated sludge using slot-blot hybridization to measure the rRNA : rDNA ratio. Microb Ecol 38, 348–357.[CrossRef]

Nadkarni, M. A., Martin, F. E. & Jacques, N. A. (2002). Determination of bacterial load by real-time PCR using a broad-range (universal) probe and primers set. Microbiology 148, 257–266.[Abstract/Free Full Text]

Nogva, H. K., Bergh, A., Holck, A. & Rudi, K. (2000). Application of the 5'-nuclease PCR assay in evaluation and development of methods for quantitative detection of Campylobacter jejuni. Appl Environ Microbiol 66, 4029–4036.[Abstract/Free Full Text]

Pearson, W. & Lipman, D. (1988). Improved tools for biological sequence comparison. Proc Natl Acad Sci U S A 85, 2444–2448.[Abstract]

Sakata, S., Kitahara, M., Sakamoto, M., Hayashi, H., Fukuyama, M. & Benno, Y. (2002). Unification of Bifidobacterium infantis and Bifidobacterium longum. Int J Syst Evol Microbiol Papers in Press, http://dx.doi.org/10.1099/ijs.0.02221-0

Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning: a Laboratory Manual, 2nd edn. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Satokari, R. M., Vaughan, E. L., Akkermans, A. D. L., Saarela, M. & De Vos, W. M. (2001a). Bifidobacterial diversity in human feces detected by genus-specific PCR and denaturing gradient gel electrophoresis. Appl Environ Microbiol 67, 504–513.[Abstract/Free Full Text]

Satokari, R. M., Vaughan, E. E., Akkermans, A. D. L., Saarela, M. & De Vos, W. M. (2001b). Polymerase chain reaction and denaturing gradient gel electrophoresis monitoring of fecal Bifidobacterium populations in a prebiotic and probiotic feeding trial. Syst Appl Microbiol 24, 227–231.[Medline]

Sghir, A., Gramet, G., Suau, A., Rochet, V., Pochart, P., & Doré, J. (2000). Quantification of bacterial groups within human fecal flora by oligonucleotide probe hybridization. Appl Environ Microbiol 66, 2263–2366.[Abstract/Free Full Text]

Suau, A., Bonnet, R., Sutren, M., Godon, J.-J., Gibson, G. R., Collins, M. D. & Doré, J. (1999). Direct analysis of genes encoding 16S rRNA from complex communities reveals many novel molecular species within the human gut. Appl Environ Microbiol 65, 4799–4807.[Abstract/Free Full Text]

Suzuki, M. T. & Giavannoni, S. J. (1996). Bias caused by template annealing in the amplification of mixtures of 16S rRNA genes by PCR. Appl Environ Microbiol 62, 625–630.[Abstract]

Suzuki, M. T., Taylor, L. T. & DeLong, E. F. (2000). Quantitative analysis of small-subunit rRNA genes in mixed microbial populations via 5'-nuclease assays. Appl Environ Microbiol 66, 4605–4616.[Abstract/Free Full Text]

Thompson, J. D., Higgins, D. G. & Gibson, T. J. (1994). CLUSTAL W: Improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res 22, 4673–4680.[Abstract]

Walter, J., Hertel, C., Tannock, G. W., Lis, C. M., Munro, K. & Hammes, W. P. (2001). Detection of Lactobacillus, Pediococcus, Leuconostoc, and Weissella species in human feces by using group-specific PCR primers and denaturing gradient gel electrophoresis. Appl Environ Microbiol 67, 2578–2585.[Abstract/Free Full Text]

Wang, R.-F., Cao, W.-W. & Cerniglia, C. E. (1996). PCR detection and quantitation of predominant anaerobic bacteria in human and animal fecal samples. Appl Environ Microbiol 62, 1242–1247.[Abstract]

Zoetendal, E. G., Akkermans, A. D. L. & De Vos, W. M. (1998). Temperature gradient gel electrophoresis of 16S rRNA from human fecal samples reveals stable and host-specific communities of active bacteria. Appl Environ Microbiol 64, 3854–3859.[Abstract/Free Full Text]

Received 30 August 2002; accepted 1 October 2002.