Establishment of an Analytical System for the Human Fecal Microbiota, Based on Reverse Transcription-Quantitative PCR Targeting of Multicopy rRNA Molecules

Abstract

An analytical system based on rRNA-targeted reverse transcription-quantitative PCR (RT-qPCR) was established for the precise evaluation of human intestinal microbiota. Group- and species-specific primer sets for Clostridium perfringens, Lactobacillus spp. (six subgroups and three species), Enterococcus spp., and Staphylococcus spp. targeting 16S rRNA gene sequences were newly developed for the quantitative analysis of such subdominant populations in human intestines. They were used together with previously reported group-specific primer sets for Enterobacteriaceae, Pseudomonas spp., and six predominant bacterial groups (the Clostridium coccoides group, the Clostridium leptum subgroup, the Bacteroides fragilis group, Bifidobacterium spp., the Atopobium cluster, and Prevotella spp.) for the examination of fecal samples from 40 healthy adults by RT-qPCR with lower detection limits of 10² to 10⁴ cells per g of feces. The RT-qPCR method gave data equivalent to those yielded by qPCR for predominant populations of more than 10⁸ cells per g of feces and could quantify bacterial populations that were not detectable (Staphylococcus and Pseudomonas) or those only detected at lower incidences (Prevotella, C. perfringens, Lactobacillus, and Enterococcus) by qPCR or the culture method. The RT-qPCR analysis of Lactobacillus spp. at the subgroup level revealed that a subject has a mean of 4.6 subgroups, with an average count of log₁₀(6.3 ± 1.5) cells per g of feces. These results suggest that RT-qPCR is effective for the accurate enumeration of human intestinal microbiota, especially the entire analysis of both predominant and subdominant populations.

The human intestinal tract harbors a large, active, and complex microbial community, which consists of several hundred bacterial species (8) with concentrations of 10¹¹ viable microorganisms per g of content (47). Human intestinal microbiota have been reported to have numerous important functions, including the digestion of food, the metabolism of endogenous and exogenous compounds, immunopotentiation, protection against pathogens, and the regulation of host fat storage (22, 26, 51). For almost a century, human intestinal microbiota have been investigated in great detail using anaerobic culture techniques (35). Conventional culture techniques for the enumeration of different populations involve the use of selective microbiological media, followed by the isolation of pure cultures and the application of confirmatory biochemical tests. Despite the effectiveness and advantage of sensitivity, these procedures are extremely labor-intensive and time-consuming and therefore might be undesirable. Moreover, classification and identification based on phenotypic traits do not always provide clear-cut results and sometimes are unreliable.

Recently, in place of cultivation-based techniques, a number of molecular methods based on the use of rRNA gene sequence information have been applied to the analyses of intestinal microbiota (46, 59, 60). Techniques such as the clone library method (8, 11, 26, 47, 58), denaturing gradient gel electrophoresis (3, 10, 19), and terminal restriction fragment length polymorphism (34, 36) have allowed the analysis of the composition of the predominant bacterial community that has not been described previously by the culture method. DNA microarray technology, with its ability to detect and measure thousands of distinct DNA sequences simultaneously, has been developed as a potentially valuable tool for the high-throughput and detailed analysis of microbial diversity (38, 39, 56). The fluorescent in situ hybridization (FISH) method, combined with microscopy (16, 48) or flow cytometry (9) with rRNA-targeted oligonucleotide probes, has enabled culture-independent observation of the in situ localization of bacteria (49). In these cases, the quantitative PCR (qPCR) method with rRNA gene-targeted primers (2, 13-15, 32, 33) is one of the most popular methods for the quantitative analysis of predominantly anaerobic bacterial populations, such as the Clostridium coccoides group, the Clostridium leptum subgroup, the Bacteroides fragilis group, Bifidobacterium, the Atopobium cluster, and Prevotella. However, subdominant populations, such as those of Enterobacteriaceae, Lactobacillus, Enterococcus, and Staphylococcus, which are important intestinal genera, have been neglected in previous examinations because of the lower population levels of such inhabitants (10² to 10⁸ cells per g of contents), which are difficult to quantify by these methods.

We recently reported that the reverse transcription-qPCR (RT-qPCR) targeting of rRNA molecules yields results for the quantification of bacteria in the environment, such as the intestines, that are more than 100-fold sensitive and specific than those of qPCR alone because of the high copy number of targeted rRNA molecules (30). In addition to this quantification method, here we developed new sets of primers to quantify intestinal subdominant populations, such as Enterococcus, Staphylococcus, and Clostridium perfringens, and also constructed primer sets for Lactobacillus species subgroups to cover the Lactobacillus genus to analyze entire bacterial populations that inhabit a wide and dynamic range of the human gastrointestinal tract. Fecal microbiota of 40 healthy adults were analyzed by RT-qPCR with the combined use of previously reported primer sets for Enterobacteriaceae, Pseudomonas (30), and six predominant anaerobe groups (31, 33). They also were analyzed with newly developed primer sets for subdominant bacterial populations with detection limits of 10² to 10⁴ cells per g of feces.

MATERIALS AND METHODS

Reference strains and culture conditions.

The bacterial strains listed in Table 2 were used in this study. Ruminococcus productus JCM 1471^T, Faecalibacterium prausnitzii ATCC 27768^T, Bacteroides vulgatus ATCC 8482^T, Bifidobacterium adolescentis ATCC 15703^T, Bifidobacterium breve ATCC 15700^T, Collinsella aerofaciens DSM 3979^T, Prevotella melaninogenica ATCC 25845^T, and C. perfringens JCM 1290^T were cultured anaerobically in modified Gifu anaerobic medium broth (Nissui Pharmaceutical Co., Ltd., Tokyo, Japan) supplemented with 1% glucose at 37°C. Lactobacillus casei ATCC 334^T, Lactobacillus acidophilus ATCC 4356^T, Lactobacillus plantarum ATCC 14917^T, Lactobacillus reuteri JCM 1112^T, Lactobacillus ruminis JCM 1152^T, Lactobacillus sakei subsp. sakei JCM 1157^T, Lactobacillus brevis ATCC 14869^T, Lactobacillus fermentum ATCC 14931^T, Lactobacillus fructivorans JCM 1117^T, and Enterococcus faecalis ATCC 19433^T were grown anaerobically in MRS broth (Becton Dickinson, Sparks, MD) at 37°C. The anaerobic manipulation of these strains was performed in an anaerobic glove box (Coy Laboratory Products Inc., Grass Lake, MI). Escherichia coli JCM 1649^T, Staphylococcus aureus ATCC 12600^T, and Pseudomonas aeruginosa IFO 12689^T were grown aerobically in brain heart infusion broth (Becton Dickinson) at 37°C. Total bacterial cell counts were determined by the 4′,6-diamidino-2-phenylindole (DAPI) staining method according to the method of Jansen et al. (23).

TABLE 2.

Specificity of the newly developed primer sets for Lactobacillus

Taxon	Strain	Subgroupa	Reaction with primer setb:
			sg-Lcas-F/sg-Lcas-R	sg-Lgas-F/sg-Lgas-R	sg-Lpla-F/sg-Lpla-R	sg-Lreu-F/sg-Lreu-R	sg-Lrum-F/sg-Lrum-R	sg-Lsak-F/sg-Lsak-R	s-Lbre-F/s-Lbre-R	LFer-1/LFer-2	s-Lfru-F/s-Lfru-R
L. casei	ATCC 334T	L. casei	+	−	−	−	−	−	−	−	−
L. rhamnosus	ATCC 7469T	L. casei	+	−	−	−	−	−	−	−	−
L. zeae	ATCC 393	L. casei	+	−	−	−	−	−	−	−	−
L. acidophilus	ATCC 4356T	L. gasseri	−	+	−	−	−	−	−	−	−
L. amylovorus	JCM 1126T	L. gasseri	−	+	−	−	−	−	−	−	−
L. crispatus	JCM 1185T	L. gasseri	−	+	−	−	−	−	−	−	−
L. delbrueckii subsp. bulgaricus	ATCC 11842T	L. gasseri	−	+	−	−	−	−	−	−	−
L. delbrueckii subsp. delbrueckii	ATCC 9649T	L. gasseri	−	+	−	−	−	−	−	−	−
L. delbrueckii subsp. lactis	ATCC 12315T	L. gasseri	−	+	−	−	−	−	−	−	−
L. gasseri	DSM 20243T	L. gasseri	−	+	−	−	−	−	−	−	−
L. helveticus	ATCC 15009T	L. gasseri	−	+	−	−	−	−	−	−	−
L. jensenii	ATCC 25258T	L. gasseri	−	+	−	−	−	−	−	−	−
L. johnsonii	JCM 2012T	L. gasseri	−	+	−	−	−	−	−	−	−
L. plantarum	ATCC 14917T	L. plantarum	−	−	+	−	−	−	−	−	−
L. pentosus	JCM 1558T	L. plantarum	−	−	+	−	−	−	−	−	−
L. reuteri	JCM 1112T	L. reuteri	−	−	−	+	−	−	−	−	−
L. oris	NCFB 2160T	L. reuteri	−	−	−	+	−	−	−	−	−
L. panis	JCM 11053T	L. reuteri	−	±	−	+	±	−	−	−	−
L. vaginalis	JCM 9505T	L. reuteri	−	−	−	+	−	−	−	−	−
L. ruminis	JCM 1152T	L. ruminis	−	−	−	−	+	−	−	−	−
L. animalis	JCM 5670T	L. ruminis	−	−	−	−	+	−	−	−	−
L. mali	JCM 1116T	L. ruminis	−	−	−	−	+	−	−	−	−
L. salivarius	ATCC 11741T	L. ruminis	−	−	−	−	+	−	−	−	−
L. satsumensis	JCM 12392T	L. ruminis	−	−	−	−	+	−	−	−	−
L. sakei subsp. sakei	JCM 1157T	L. sakei	−	−	−	−	−	+	−	−	−
L. sakei subsp. carnosus	JCM 11031T	L. sakei	−	−	−	−	−	+	−	−	−
L. curvatus	JCM 1096T	L. sakei	−	−	−	−	−	+	−	−	−
L. graminis	NRIC 1775T	L. sakei	−	−	−	−	±	+	−	−	−
L. brevis	ATCC 14869T	L. brevis	−	−	−	−	−	−	+	−	−
L. fermentum	ATCC 14931T	L. fermentum	−	−	−	±	−	−	−	+	−
L. fructivorans	JCM 1117T	L. fructivorans	−	−	−	−	−	−	−	−	+
L. coryniformis subsp. coryniformis	JCM 1164T	L. coryniformis	−	−	−	−	−	−	−	−	−
L. vaccinostercus	JCM 12184	L. vaccinostercus	−	−	−	−	−	−	−	−	−
L. farciminis	JCM 1097T	L. farciminis	−	−	−	−	−	−	−	−	−
L. sharpeae	JCM 1186T	L. sharpeae	−	−	−	−	−	−	−	−	−

^aThe species of Lactobacillus were differentiated into 13 subgroups based on 16S rRNA gene sequences.

^bThe specificity of the RT-qPCR assay for target bacteria with each primer was investigated using RNA extracts corresponding to 10⁵ cells from each strain described. Specificity was judged using the criteria described in Materials and Methods. In addition, negative PCR results were obtained for the following bacterial strains: Ruminococcus productus JCM 1471^T, Ruminococcus obeum ATCC 29174^T, Faecalibacterium prausnitzii ATCC 27768^T, Clostridium orbiscindens DSM 6740^T, Bacteroides ovatus JCM 5824^T, Bacteroides vulgatus ATCC 8482^T, Bifidobacterium adolescentis ATCC 15703^T, Bifidobacterium longum ATCC 15707^T, Collinsella aerofaciens DSM 3979^T, Eggerthella lenta ATCC 25559^T, Prevotella melaninogenica ATCC 25845^T, Escherichia coli JCM 1649^T, Acinetobacter calcoaceticus JCM 6842^T, Pseudomonas aeruginosa IFO 12689^T, Enterococcus faecalis ATCC 19433^T, Streptococcus anginosus GIFU 8327, Staphylococcus aureus ATCC 12600^T, Lactococcus lactis subsp. lactis ATCC 19435^T, Bacillus cereus JCM 2152^T, Campylobacter jejuni subsp. jejuni DSM 4688^T, and Candida albicans IFO 1385^T.

Development of rRNA gene-targeted primers.

All primers used in this study are listed in Table 1. Through the use of 16S rRNA gene sequences obtained from DDBJ/GenBank/EMBL databases for bacteria detectable in the human intestinal tract, we constructed a multiple alignment of the target groups and reference organisms with the Clustal X program (50). After analyzing the sequences, potential primer target sites for group- or species-specific detection were identified. We designed primers and checked their specificity with the database by submitting the sequence to the Probe Match program of the Ribosomal Database Project (RDP-II) (http://rdp.cme.msu.edu/) (28). As for Lactobacillus, we first differentiated Lactobacillus species into 13 subgroups based on 16S rRNA gene sequences and designated each subgroup as shown in Table 2. We constructed subgroup- or species-specific primers for 9 of the 13 subgroups, the L. casei subgroup, the L. gasseri subgroup, the L. plantarum subgroup, the L. reuteri subgroup, the L. ruminis subgroup, the L. sakei subgroup, L. brevis, L. fermentum (57), and L. fructivorans, which comprise species that have been reported to be detected in the human intestines (19, 45, 54, 55).

TABLE 1.

16S or 23S rRNA gene-targeted primers used in this study

Targeta	Primer	Sequence (5′-3′)	Product sizeb (bp)	Annealing temp (°C)	Reference or source
Clostridium coccoides group	g-Ccoc-F	AAATGACGGTACCTGACTAA	440	55	33
	g-Ccoc-R	CTTTGAGTTTCATTCTTGCGAA
Clostridium leptum subgroup	sg-Clept-F	GCACAAGCAGTGGAGT	239	55	33
	sg-Clept-R3	CTTCCTCCGTTTTGTCAA
Bacteroides fragilis group	g-Bfra-F2	AYAGCCTTTCGAAAGRAAGAT	495	50	31
	g-Bfra-R	CCAGTATCAACTGCAATTTTA
Bifidobacterium	g-Bifid-F	CTCCTGGAAACGGGTGG	552	55	33
	g-Bifid-R	GGTGTTCTTCCCGATATCTACA
Bifidobacterium breve	BiBRE-1	CCGGATGCTCCATCACAC	288	55	32
	BiBRE-2	ACAAAGTGCCTTGCTCCCT
Atopobium cluster	c-Atopo-F	GGGTTGAGAGACCGACC	190	55	33
	c-Atopo-R	CGGRGCTTCTTCTGCAGG
Prevotella	g-Prevo-F	CACRGTAAACGATGGATGCC	513	55	33
	g-Prevo-R	GGTCGGGTTGCAGACC
Clostridium perfringensc	s-Clper-F	GGGGGTTTCAACACCTCC	170	60	This study
	ClPER-R	GCAAGGGATGTCAAGTGT			24
Enterobacteriaceae	En-lsu-3F	TGCCGTAACTTCGGGAGAAGGCA	428	60	30
	En-lsu-3′R	TCAAGGACCAGTGTTCAGTGTC
Lactobacillus casei subgroup	sg-Lcas-F	ACCGCATGGTTCTTGGC	296	60	This study
	sg-Lcas-R	CCGACAACAGTTACTCTGCC
Lactobacillus gasseri subgroup	sg-Lgas-F	GATGCATAGCCGAGTTGAGAGACTGAT	197	60	This study
	sg-Lgas-R	TAAAGGCCAGTTACTACCTCTATCC
Lactobacillus plantarum subgroup	sg-Lpla-F	CTCTGGTATTGATTGGTGCTTGCAT	54	60	This study
	sg-Lpla-R	GTTCGCCACTCACTCAAATGTAAA
Lactobacillus reuteri subgroup	sg-Lreu-F	GAACGCAYTGGCCCAA	289	60	This study
	sg-Lreu-R	TCCATTGTGGCCGATCAGT
Lactobacillus ruminis subgroup	sg-Lrum-F	CACCGAATGCTTGCAYTCACC	182	60	This study
	sg-Lrum-R	GCCGCGGGTCCATCCAAAA
Lactobacillus sakei subgroup	sg-Lsak-F	CATAAAACCTAMCACCGCATGG	303	60	This study
	sg-Lsak-R	TCAGTTACTATCAGATACRTTCTTCTC
Lactobacillus brevis	s-Lbre-F	ATTTTGTTTGAAAGGTGGCTTCGG	289	55	This study
	s-Lbre-R	ACCCTTGAACAGTTACTCTCAAAGG
Lactobacillus fermentum	LFer-1	CCTGATTGATTTTGGTCGCCAAC	414	55	57
	LFer-2	ACGTATGAACAGTTACTCTCATACGT
Lactobacillus fructivorans	s-Lfru-F	TGCGCCTAATGATAGTTGA	452	55	This study
	s-Lfru-R	GATACCGTCGCGACGTGAG
Enterococcus	g-Encoc-F	ATCAGAGGGGGATAACACTT	337	55	This study
	g-Encoc-R	ACTCTCATCCTTGTTCTTCTC
Staphylococcus	g-Staph-F	TTTGGGCTACACACGTGCTACAATGGACAA	79	60	This study
	g-Staph-R	AACAACTTTATGGGATTTGCWTGA
Pseudomonas	PSD7F	CAAAACTACTGAGCTAGAGTACG	215	60	30
	PSD7R	TAAGATCTCAAGGATCCCAACGGCT

^aSpecific primer sets were developed by using 16S rRNA gene sequences, except for En-lsu-3F/En-lsu-3′R, which target 23S rRNA genes.

^bRNAs extracted from the standard strains described in Materials and Methods were used as the RT-qPCR controls.

^cThe forward primer s-Clper-F was newly developed in this study and was used in the combination with the reverse primer ClPER-R, which was previously reported (24).

Collection and preparation of fecal samples.

Immediately after defecation, fecal samples were kept at 4°C anaerobically by using Anaero Pouch-Anaero (Mitsubishi Gas Chemical Company, Inc., Tokyo, Japan), which were examined within 24 h after sampling. Each fecal sample was weighed and suspended in 9 volumes of sterilized anaerobic transfer medium [KH₂PO₄, 0.0225% (wt/vol); K₂HPO₄, 0.0225% (wt/vol); NaCl, 0.045% (wt/vol); (NH₄)₂SO₄, 0.0225% (wt/vol); CaCl₂, 0.00225% (wt/vol); MgSO₄, 0.00225% (wt/vol); Na₂CO₃, 0.3% (wt/vol); l-cysteine hydrochloride, 0.05% (wt/vol); resazurin, 0.0001% (wt/vol); Lab Lemco powder (Oxoid Co., Ltd., Basingtoke, United Kingdom), 1.0% (wt/vol); glycerol (Wako Pure Chemical Industries, Ltd., Osaka, Japan), 10% (wt/vol)] to make a fecal homogenate in an anaerobic glove box.

Cultivation of fecal bacteria.

After the serial dilution of the fecal homogenates with anaerobic diluting solution [0.0225% (wt/vol) KH₂PO₄, 0.0225% (wt/vol) K₂HPO₄, 0.045% (wt/vol) NaCl, 0.0225% (wt/vol) (NH₄)₂SO₄, 0.00225% (wt/vol) CaCl₂, 0.00225% (wt/vol) MgSO₄, 0.3% (wt/vol) Na₂CO₃, 0.05% (wt/vol) l-cysteine hydrochloride, 0.0001% (wt/vol) resazurin, 0.05% (wt/vol) BACTO agar (Difco), 0.05% (wt/vol) Tween 80], 50-μl portions of the appropriate diluents were spread onto the following agar culture media: nonselective medium 10 (M10) (20) for total anaerobes; modified transgalactosylated oligosaccharide propionate (MTP) (Eiken Chemical Co., Ltd., Tokyo, Japan) for Bifidobacterium; Clostridium welchii agar (Nikken Biomedical Laboratory, Inc., Kyoto, Japan) for lecithinase-positive Clostridium spp. [designated Clostridium (L+)]; deoxycholate hydrogen sulfide lactose (DHL; Nikken Biomedical Laboratory, Inc.) for Enterobacteriaceae; Lactobacillus selection (LBS; Becton Dickinson) for Lactobacillus; colistin-oxolinic acid-blood agar (COBA) (40) (Nikken Biomedical Laboratory, Inc.) for Enterococcus; and yolk-supplemented mannit salt A (Nikken Biomedical Laboratory, Inc.) for Staphylococcus. M10, MTP, C. welchii, and LBS agar plates were incubated at 37°C anaerobically for 4, 3, 1, and 3 days, respectively. DHL, COBA, and yolk-supplemented mannit salt A agar plates were incubated at 37°C aerobically for 1, 2, and 2 days, respectively. Colonies on the agar plates were counted, and CFU counts for target bacteria per gram of feces (wet weight) were calculated. The lower limit of bacterial detection with this procedure was 200 CFU per g of feces.

Enumeration of predominant bacteria in fecal samples by FISH.

FISH analyses with 16S rRNA-targeted group-specific oligonucleotide probes were carried out according to the procedure of Takada et al. (48). Briefly, oligonucleotide probes were synthesized with a Cy5-reactive fluorescent dye at the 5′ end (Sigma Genosys, Hokkaido, Japan). The following probes were used to enumerate the target bacterial groups in fecal samples: Eub338 (5′-GCT GCC TCC CGT AGG AGT-3′) (1) for total bacteria, Erec482 (5′-GCT TCT TAG TCA RGT ACC G-3′) for the C. coccoides group (12), Clept1240 (5′-GTT TTR TCA ACG GCA GTC-3′) for the C. leptum subgroup (43), Bfra602 (5′-GAG CCG CAA ACT TTC ACA A-3′) for the B. fragilis group (12), Bif153 (5′-ACC ACC CGT TTC CAG GAG-3′) for Bifidobacterium (48), ATO291 (5′-GGT CGG TCT CTC AAC CC-3′) for the Atopobium cluster (17), and Prev496 (5′-CGG AAT TAG CCG GTC CTT AT-3′) for Prevotella (this study). Fecal homogenates were added to 3 volumes of 4% paraformaldehyde-PBS (PFA) solution, which were incubated for 16 h at 4°C. A PFA-treated fecal sample was smeared on a MAS-coated slide glass (Matsunami, Osaka, Japan), which was hybridized with the probes or stained with Vectashield mounting medium with DAPI (Vector Laboratories, Inc., Burlingame, CA). The observation and acquisition of fluorescent images were performed with a Leica imaging system (an automatic fluorescent microscope [Leica DM6000], image acquisition software [QFluoro], and a cooled black-and-white charge-coupled display camera [Leica DFC3500FX]) (Leica Microsystems GmbH, Wetzlar, Germany). The Leica DM6000 was equipped with an HCX PLAN APO objective (magnification, ×100; numerical aperture, 1.35), a Leica EL6000 external light source, and two fluorescent filters, A4 (excitation, 360/40 nm; dichroic mirror, 400 nm; and emission, 470/40 nm [numbers after each slash represent bandwidths for each bandpass filter]) and Y5 (excitation, 620/60 nm; dichroic mirror, 660 nm; and emission, 700/75 nm). The fluorescent images obtained were analyzed using image analysis software (Image-Pro Plus v. 4.5; Media Cybernetics, Inc., Bethesda, MD) to determine the fluorescent cells in each fecal sample. Microscopic counts were determined from 10 images, and a minimum of 50 cells per image was counted.

Total RNA isolation.

For RNA stabilization, fresh cultures of each bacterial (50 μl) or fecal (200 μl) homogenate were added to 2 volumes of RNAprotect bacterial reagent (Qiagen GmbH, Hilden, Germany) and then incubated for 5 min at room temperature. After the centrifugation of the mixture at 5,000 × g for 10 min, the supernatant was discarded, and pellets were stored at −80°C until used to extract RNA. RNA was isolated using a modified method of acid guanidinium thiocyanate-phenol-chloroform extraction (4). Briefly, the thawed sample was resuspended in a solution containing 346.5 μl of RLT buffer (Qiagen Sciences, Germantown, MD), 3.5 μl of β-mercaptoethanol (Sigma-Aldrich Co., St. Louis, MO), and 100 μl of Tris-EDTA buffer. Glass beads (300 mg; diameter, 0.1 mm) (BioSpec Products, Inc., Bartlesville, OK) were added to the suspension, and the mixture was vortexed vigorously for 60 s using a FastPrep FP 120 (BIO 101, Vista, CA) at a power level of 5.0. Acid phenol (500 μl; Wako Pure Chemical Industries, Ltd.) was added, and the mixture was incubated for 10 min at 60°C. After incubation, the mixture was cooled on ice for 5 min and added to 100 μl of chloroform-isoamilalcohol (24:1). After centrifugation at 12,000 × g for 10 min at 4°C, 450 μl of supernatant was collected and added to an equal volume of chloroform-isoamilalcohol. After centrifugation at 12,000 × g for 5 min, 400 μl of supernatant was collected and subjected to isopropanol precipitation. Finally, the nucleic acid fraction was suspended in 1 ml of nuclease-free water (Ambion, Inc., Austin, TX). The following DNase treatment was skipped in this study, because we confirmed that untreated and DNase-treated samples showed identical results in the preliminary experiments, indicating that contaminating DNA does not affect RT-qPCR quantification (see Table S1 in the supplemental material).

RT-qPCR.

RT-qPCR was conducted in a one-step reaction using a Qiagen OneStep RT-PCR kit (Qiagen GmbH). Each reaction mixture (10 μl) was composed of 1× Qiagen OneStep RT-PCR buffer, 0.5× Q-solution, each deoxynucleoside triphosphate at a concentration of 400 μM, a 1:100,000 dilution of SYBR green I (Molecular Probes, Eugene, OR), 0.4 μl of Qiagen OneStep RT-PCR enzyme mix, each specific primer at a concentration of 0.6 μM (except for g-Bfra-F2/g-Bfra-R at 1.2 μM and sg-Lsak-F/sg-Lsak-R at 2.4 μM), and 5 μl of template RNA. The reaction mixture was incubated at 50°C for 30 min for reverse transcription. The continuous amplification program consisted of one cycle at 95°C for 15 min; 40 cycles at 94°C for 20 s, 50, 55, or 60°C (Table 1) for 20 s, and 72°C for 50 s; and finally one cycle at 94°C for 15 s. Fluorescent products were detected in the last step of each cycle. Melting curve analysis was performed after amplification to distinguish the target from nontargeted PCR products. The melting curve was obtained by slow heating at temperatures from 60 to 95°C at a rate of 0.2°C/s with continuous fluorescent collection. Amplification and detection were performed in 384-well optical plates on an ABI PRISM 7900HT sequence detection system (Applied Biosystems, Foster City, CA).

DNA extraction and qPCR.

DNA extraction and qPCR were performed according to the method described by Matsuki et al. (32). qPCR amplification and detection were performed in 384-well optical plates on an ABI PRISM 7900HT sequence detection system (Applied Biosystems).

Determination of RT-PCR sensitivity.

RNA dilutions corresponding to bacterial counts ranging from 10⁻³ to 10⁵ cells for C. perfringens JCM 1290^T, L. casei ATCC 334^T, L. acidophilus ATCC 4356^T, L. plantarum ATCC 14917^T, L. reuteri JCM 1112^T, L. ruminis JCM 1152^T, L. sakei subsp. sakei JCM 1157^T, L. brevis ATCC 14869^T, L. fermentum ATCC 14931^T, L. fructivorans JCM 1117^T, E. faecalis ATCC 19433^T, and S. aureus ATCC 12600^T were added to a series of RT-PCR mixtures with specific primer sets for each strain (Table 1). RNA concentrations were confirmed to be linear with the cycle number at which the product fluorescence surpassed a defined cyclic threshold (C_T value) over the range corresponding to 10⁻³ to 10⁵ cells per reaction for C. perfringens, L. acidophilus, L. plantarum, L. reuteri, L. ruminis, L. sakei subsp. sakei, L. brevis, and L. fructivorans; 10⁻² to 10⁵ cells for L. casei, E. faecalis, and S. aureus; and 10⁻¹ to 10⁵ cells for L. fermentum (R² > 0.99). R. productus JCM 1471^T, F. prausnitzii ATCC 27768^T, B. vulgatus ATCC 8482^T, B. adolescentis ATCC 15703^T, C. aerofaciens DSM 3979^T, and P. melaninogenica ATCC 25845^T also were subjected to this analysis with their specific primers. Identical results were obtained across the range of RNA concentrations corresponding to 10⁻³ to 10⁵ cells per reaction for R. productus, F. prausnitzii, and B. vulgatus, 10⁻² to 10⁵ cells for B. adolescentis and P. melaninogenica, and 10⁻¹ to 10⁵ cells for C. aerofaciens (R² > 0.99). These data indicate that the linear range for the procedures used in this study theoretically is 10², 10³, or 10⁴ to 10¹⁰ cells per g of feces.

Determination of primer specificity.

The specificity of the newly developed primers was determined as follows. Total RNA fractions extracted from the bacterial cells of each strain shown in Table 2 at a dose corresponding to 10⁵ cells were assessed by RT-qPCR using group- or species-specific primers (Table 1). Using the standard curve for the representative strain of each group, obtained as described above, the amplified signal was judged as positive when it was more than that of 10⁴ standard cells and as negative when less than that of 10⁻¹ standard cells. The signal between 10⁻¹ and 10⁴ standard cells was defined as positive-negative. The amplified signal was defined as negative when the corresponding melting curve showed a different peak from that of the standard strain.

Determination of bacterial count by RT-qPCR.

The standard curve was generated with RT-qPCR data, C_T values, and the corresponding cell count, which was determined microscopically with the DAPI staining method, of the dilution series of the following standard strains: R. productus JCM 1471^T (for the C. coccoides group), F. prausnitzii ATCC 27768^T (for the C. leptum subgroup), B. vulgatus ATCC 8482^T (for the B. fragilis group), B. adolescentis ATCC 15703^T (for Bifidobacterium), B. breve ATCC 15700^T (for B. breve), C. aerofaciens DSM 3979^T (for the Atopobium cluster), P. melaninogenica ATCC 25845^T (for Prevotella), C. perfringens JCM 1290^T (for C. perfringens), E. coli JCM 1649^T (for Enterobacteriaceae), L. casei ATCC 334^T (for the L. casei subgroup), L. acidophilus ATCC 4356^T (for the L. gasseri subgroup), L. plantarum ATCC 14917^T (for the L. plantarum subgroup), L. reuteri JCM 1112^T (for the L. reuteri subgroup), L. ruminis JCM 1152^T (for the L. ruminis subgroup), L. sakei subsp. sakei JCM 1157^T (for the L. sakei subgroup), L. brevis ATCC 14869^T (for L. brevis), L. fermentum ATCC 14931^T (for L. fermentum), L. fructivorans JCM 1117^T (for L. fructivorans), E. faecalis ATCC 19433^T (for Enterococcus), S. aureus ATCC 12600^T (for Staphylococcus), and P. aeruginosa IFO 12689^T (for Pseudomonas). For the determination of the target bacteria present in fecal samples, three serial 10-fold dilutions of extracted RNA sample (corresponding to 1/2,000, 1/20,000, and 1/200,000 of the amount of RNA extracted from 20 mg of feces) were applied to RT-qPCR, and C_T values in the linear range of the assay were applied to the standard curve generated in the same experiment to obtain the corresponding bacterial count in each nucleic acid sample. The results were converted into the count per sample.

Quantification of bacteria spiked into human feces by RT-qPCR.

Fecal samples were collected from three healthy adult volunteers who had been confirmed in advance not to have B. breve, C. perfringens, L. fructivorans, or Pseudomonas as their indigenous intestinal populations. Each fecal sample was weighed and suspended in 9 volumes of sterilized anaerobic transfer medium to make a fecal homogenate. Serial dilutions of B. breve ATCC 15700^T, C. perfringens JCM 1290^T, L. fructivorans JCM 1117^T, and P. aeruginosa IFO 12689^T were spiked into the fecal homogenates to obtain final concentrations ranging from 10² to 10⁹ cells per g of feces. RNA fractions extracted from each sample were assessed by RT-qPCR with the specific primer sets for B. breve, C. perfringens, L. fructivorans, and Pseudomonas (Table 1). The C_T values obtained were applied to the analytical curve of B. breve ATCC 15700^T, C. perfringens JCM 1290^T, L. fructivorans JCM 1117^T, or P. aeruginosa IFO 12689^T to determine RT-qPCR counts. The bacterial cell counts of the cultures spiked were enumerated by the DAPI staining method.

Sequencing of RT-PCR-amplified rRNA genes.

RT-PCR products generated with the primer set of En-lsu-3F/En-lsu-3′R or s-Clper-F/ClPER-R were purified using Montage PCR centrifugal filter devices (Millipore Corporation, Bedford, MA) and used for the sequence analysis of 16S or 23S rRNA gene fragments. Cycle-sequencing reactions were performed with a BigDye Terminator v3.1 cycle sequencing kit (Applied Biosystems) according to the manufacturer's specifications. Sequences were automatically analyzed on an ABI PRISM 3130 genetic analyzer (Applied Biosystems). The rRNA gene sequences obtained were analyzed using the BLAST program of the DNA Data Bank of Japan (http://blast.ddbj.nig.ac.jp/) to assign a strain to a particular species.

Statistical analysis.

We employed SPSS 14.0 software (SPSS Japan Inc., Tokyo, Japan). Regression analysis was performed to determine the statistical correlation of the results, and Pearson's product-moment correlation coefficient was calculated. P < 0.05 was regarded as significant.

RESULTS

Primer specificity.

The specificity of the newly developed primers was evaluated. Total RNA fractions extracted from 56 bacterial strains (Table 2) corresponding to 10⁵ cells were assessed by RT-qPCR using Lactobacillus subgroup- or species-specific primers (Table 1). As shown in Table 2, each primer set gave positive RT-qPCR results only for the corresponding target bacterial species. The primer sets of sg-Lgas-F/sg-Lgas-R, sg-Lreu-F/sg-Lreu-R, and sg-Lrum-F/sg-Lrum-R cross-reacted with some of the nontarget microorganisms tested at negligible levels that have little effect on the specific enumeration of target bacteria. The specificity of s-Clper-F/ClPER-R (for C. perfringens) (24), g-Encoc-F/g-Encoc-R (for Enterococcus), and g-Staph-F/g-Staph-R (for Staphylococcus) also was confirmed (data not shown).

Quantification of bacteria spiked into human feces by RT-qPCR.

Serial dilutions of B. breve, C. perfringens, L. fructivorans, or P. aeruginosa were spiked into the fecal samples of volunteers who were found in advance not to have the corresponding indigenous populations. RT-qPCR could detect B. breve, C. perfringens, L. fructivorans, and P. aeruginosa with a lowest concentration of 10^2.7, 10^2.4, 10^2.8, and 10^2.6 cells per g of feces, respectively, and the concentration of the spiked bacteria determined by the DAPI staining method (x axis) and the RT-qPCR counts (y axis) were found to correlate well across the range of bacterial concentrations, from 10² to 10⁹ cells per g of feces (R² > 0.99) (Fig. 1). Based on these results, it was suggested that rRNA-targeted RT-qPCR could determine the bacterial count in human feces sensitively, with a lower detection limit of 10² to 10³ cells per g of feces.

FIG. 1.

Quantification of bacteria spiked into human feces by RT-qPCR. Fecal samples from three individuals spiked with serial dilutions of B. breve ATCC 15700^T (A), C. perfringens JCM 1290^T (B), L. fructivorans JCM 1117^T (C), or P. aeruginosa IFO 12689^T (D) to final concentrations of 10² to 10⁹ cells per g were assessed by RT-qPCR. Bacterial counts were determined by the DAPI staining method and plotted against the corresponding RT-qPCR counts; data are expressed as the means and standard deviations of three individuals.

Quantification of bacteria in human feces by RT-qPCR, qPCR, and the culture method.

RT-qPCR analysis was performed to enumerate individual bacterial groups in fecal samples collected from 40 healthy adult volunteers (24 males and 16 females; ages, 20 to 65 years [average, 41 ± 13 years]) (Table 3). The compositions of both the predominant and subdominant bacterial populations were analyzed by RT-qPCR, with the lower detection limits ranging from 10² to 10⁴ cells per g of feces, and the total population of these 11 bacterial groups and one species was log₁₀(10.6 ± 0.5) (means ± standard deviations) cells per g of feces. These results were compared to those obtained by qPCR and the culture method. The bacterial counts and detection rates of the C. coccoides group, the C. leptum subgroup, the B. fragilis group, Bifidobacterium, and the Atopobium cluster obtained by RT-qPCR were equivalent to those by qPCR. On the other hand, the population levels of Prevotella, Lactobacillus, and Enterococcus species detected by RT-qPCR were lower than those by qPCR, which can be explained in terms of the difference in sensitivity between RT-qPCR and qPCR. Because RT-qPCR is much more sensitive than qPCR, RT-qPCR data include the populations around 10³ to 10⁶ cells per g of feces, which were below the lower detection limit of qPCR for these species. The RT-qPCR detection rates were equivalent to (Enterobacteriaceae, Lactobacillus, and Staphylococcus) or greater than (C. perfringens and Enterococcus) those determined by the culture method, while qPCR detected C. perfringens, Lactobacillus, Staphylococcus, and Pseudomonas at a far lower incidence than RT-qPCR or the culture method.

TABLE 3.

Comparison of RT-qPCR counts to qPCR counts and to cultural counts from human feces

Bacterial population	RT-qPCR		qPCR		Culture
	Log10 cells/g fecesa	Detection ratio (%)	Log10 cells/g fecesa	Detection ratio (%)	Log10 CFU/g fecesa	Detection ratio (%)
Clostridium coccoides group	10.2 ± 0.5	100	10.1 ± 0.5	100	NT
Clostridium leptum subgroup	9.5 ± 0.7	100	9.4 ± 0.8	100	NT
Bacteroides fragilis group	9.8 ± 0.6	100	9.4 ± 0.5	100	NT
Bifidobacterium	9.5 ± 1.0	98	9.1 ± 0.9	98	9.5 ± 0.5	95
Atopobium cluster	9.1 ± 0.8	100	8.9 ± 0.7	100	NT
Prevotella	6.9 ± 1.9	88	7.7 ± 1.4	63	NT
Clostridium perfringens/Clostridium (L+)d	4.3 ± 1.6	83	6.4	6	4.4 ± 1.5	48
Enterobacteriaceae	7.1 ± 1.2	100	8.2 ± 0.5	100	6.8 ± 1.1	100
Lactobacillus	6.3 ± 1.5e	98	7.1 ± 0.9e	39	5.6 ± 1.6	90
Enterococcus	6.2 ± 1.4	90	8.0 ± 1.4	56	6.8 ± 1.5	45
Staphylococcus	5.3 ± 0.5	85	ND	0	3.7 ± 1.1	70
Pseudomonas	4.3 ± 0.7	30	ND	0	NT
Total	10.6 ± 0.5b	100	10.5 ± 0.4b	100	10.5 ± 0.5c	100

^aData are expressed as the means and standard deviations of fecal samples collected from 40 adult volunteers. NT, not tested. ND, not detected.

^bThe total count of bacteria obtained by RT-qPCR or qPCR is expressed as the sum of the counts of 11 groups and one species.

^cThe total count of bacteria cultivated with M10 plates.

^dIn the culture method, Clostridium (L+) was selectively cultivated on C. welchii agar plates.

^eThe count of Lactobacillus species organisms obtained by RT-qPCR or qPCR is expressed as the sum of the counts of six subgroups and three species.

To further confirm the validity of the quantification of Enterobacteriaceae and C. perfringens by RT-qPCR, a sequence analysis of the RT-qPCR products was performed. The dominant 0.42-kb fragments generated by the En-lsu-3F/En-lsu-3′R primer set from 33 subjects resulted in sequences most similar to the 23S rRNA genes of known Enterobacteriaceae species, with identities exceeding 97%. These include 26 strains of E. coli, 4 strains of Klebsiella pneumoniae, 2 Enterobacter species, and 1 strain of Citrobacter freundii. For the RT-qPCR products generated by the s-Clper-F/ClPER-R primer set (0.17-kb fragments) from 31 subjects, all were confirmed to be C. perfringens, which was the targeted species (data not shown).

Composition of Lactobacillus subgroups determined by RT-qPCR.

Table 4 shows the diversity of the population levels of different Lactobacillus species subgroups in fecal samples of 40 subjects determined by RT-qPCR. Lactobacillus species subgroups were detected in 98% of the samples, and the mean total count of Lactobacillus species organisms as the sum of six subgroups and three species was log₁₀(6.3 ± 1.5) per g of feces. The average number of Lactobacillus species subgroups detected per subject was 4.6 ± 1.7, and the L. reuteri, L. plantarum, L. ruminis, and L. gasseri subgroups were detected as the major groups, with an incidence of 80, 80, 78, and 78% and population levels of log₁₀(5.1 ± 1.5), log₁₀(4.3 ± 1.1, log₁₀(5.2 ± 1.8), and log₁₀(5.1 ± 1.4), respectively. L. brevis and L. fructivorans were found to be relatively less common (13 and 5%) and were present at the lower population levels of log₁₀(4.3 ± 1.1) and log₁₀ 3.6, respectively.

TABLE 4.

Quantification of Lactobacillus species subgroups in human feces by RT-qPCR

Bacterial population	Log10 cells/g fecesa	Detection ratio (%)
L. casei subgroup	4.7 ± 1.1	48
L. gasseri subgroup	5.1 ± 1.4	78
L. plantarum subgroup	4.3 ± 1.1	80
L. reuteri subgroup	5.1 ± 1.5	80
L. ruminis subgroup	5.2 ± 1.8	78
L. sakei subgroup	5.2 ± 1.3	23
L. brevis	4.3 ± 1.1	13
L. fermentum	5.5 ± 1.4	60
L. fructivorans	3.6	5
Totalb	6.3 ± 1.5	98

^aData are expressed as the means and standard deviations from fecal samples collected from 40 adult volunteers.

^bThe total count of Lactobacillus species organisms is expressed as the sum of the counts of six subgroups and three species.

Linear regression analysis for bacterial counts obtained by RT-qPCR, qPCR, and the culture method.

The bacterial counts obtained by RT-qPCR correlated well with those by qPCR for predominant populations such as the C. coccoides group, the C. leptum subgroup, Bifidobacterium, and the Atopobium cluster (P < 0.01) (Fig. 2). Good correlations between the bacterial counts determined by RT-qPCR and the culture methods were found for Bifidobacterium, C. perfringens, Enterobacteriaceae, and Lactobacillus (P < 0.01) (Fig. 3).

FIG. 2.

Correlation between RT-qPCR and qPCR counts in human feces. Total RNA and DNA fractions extracted from 40 human fecal homogenates were assessed by RT-qPCR (total RNA) and qPCR (DNA) assays to enumerate indigenous population levels of the C. coccoides group (A), the C. leptum subgroup (B), Bifidobacterium (C), and the Atopobium cluster (D). qPCR counts were plotted against RT-qPCR counts.

FIG. 3.

Correlation between RT-qPCR and cultural counts in human feces. Total RNA fractions extracted from 40 human fecal homogenates were assessed by RT-qPCR to enumerate indigenous population levels of Bifidobacterium (A), C. perfringens (B), Enterobacteriaceae (C), and Lactobacillus (D). CFU counts were determined by culturing the same fecal samples on MTP (Bifidobacterium), C. welchii (C. perfringens), DHL (Enterobacteriaceae), and LBS (Lactobacillus) agar plates and were plotted against RT-qPCR counts.

Comparison of RT-qPCR with FISH for quantification of bacterial counts in human feces.

The RT-qPCR method was compared to FISH for the enumeration of six predominant bacterial groups in fecal samples collected from four healthy adult volunteers (three males and one female) (Table 5). Overall, the populations of the bacterial groups determined by RT-qPCR and FISH were found to be at equivalent levels. However, the population levels of Bifidobacterium determined by FISH were lower than those by RT-qPCR (P < 0.05, as determined by a paired t test).

TABLE 5.

Comparison of bacterial counts in human feces determined by RT-qPCR and FISH

Bacterial population	Log10 bacterial cells/g feces for subjecta:
	A		B		C		D
	RT-PCR	FISH	RT-PCR	FISH	RT-PCR	FISH	RT-PCR	FISH
Clostridium coccoides group	10.3	10.3	10.3	10.5	10.3	10.4	10.4	10.4
Clostridium leptum subgroup	10.1	9.7	10.3	10.1	10.6	10.7	10.0	9.9
Bacteroides fragilis group	10.5	10.3	10.2	10.4	10.2	10.5	10.2	10.5
Bifidobacterium	10.6	10.1	10.3	10.0	10.6	10.2	10.3	9.8
Atopobium cluster	9.8	9.8	9.5	9.9	8.7	8.8	9.7	9.9
Prevotella	ND	ND	ND	ND	ND	6.6	6.0	ND
Sum of six groups	11.0	10.8	10.9	10.9	11.1	11.1	10.9	10.9
Total count of bacteriab	NT	11.1	NT	11.0	NT	11.3	NT	11.1
Total count of cells (by DAPI staining)		11.3		11.2		11.6		11.2

^aFecal samples collected from four adult volunteers were used for the comparisons. NT, not tested. ND, not detected.

^bThe total count of bacteria was determined by hybridization with probe Eub338 (1).

DISCUSSION

This is the first evidence showing the quantitative examination of subdominant as well as predominant bacterial populations in human fecal specimens without culture, which was achieved by the rRNA-targeted RT-qPCR method with 20 group- and species-specific primer sets with a sensitivity of 10² to 10⁴ cells per g of feces. Recently, the sequencing approach of 16S rRNA gene clones (8, 11, 27) or microbial genomes (25, 29) was extensively conducted to investigate the composition of human intestinal microbiota. The comprehensive sequencing analysis of >13,000 16S rRNA gene clones revealed that human intestinal microbiota were dominated by Firmicutes and Bacteroidetes, which made up more than 99% of the identified phylogenetic types, and the composition differed between fecal and mucosal communities (8). The comparative metagenomic approach with human fecal samples indicated that the complexity of Firmicutes is reduced in Crohn's disease patients compared to that in healthy subjects (29). Although such studies have well described the diversity of human intestinal microbiota and their abundance, especially in predominant bacterial populations, sequence information about subdominant bacteria is limited and is not appropriate for the accurate understanding of these bacteria. On the other hand, the rRNA-targeted RT-qPCR method could quantify the abundance of targeted bacterial populations, including subdominant bacteria, with high resolution, and it has several advantages, such as sensitivity, rapidity, and accuracy. These aspects of the RT-qPCR method will contribute to a more accurate understanding of a bacterium's relations with the human host.

Although the bacterial counts of the predominant populations in fecal samples enumerated by two different molecular biological methods targeting rRNA molecules, i.e., RT-qPCR and FISH, were similar, the population levels of Bifidobacterium determined by the RT-qPCR method were significantly higher than those determined by the FISH method (Table 5). Although the exact reason for this is not clear from the present results, it may be explained by the difference in the sensitivity between these two methods: the RT-qPCR method could measure smaller amounts of rRNA in the bacterial cells than the FISH method. On the other hand, the RT-qPCR method gave higher values than the qPCR method in some samples at lower counts, which was reflected by the high intercept values in linear regression analysis (Fig. 2A to D). This seems to reflect the different sensitivity levels of the RT-qPCR and qPCR methods. On the other hand, the population levels of Enterobacteriaceae and Lactobacillus enumerated by RT-qPCR correlated well with the CFU counts, with the exception of samples with lower CFU counts (Fig. 3C and D), which seems to be due to the presence of bacterial cells losing their colony-forming ability on selective media but still maintaining certain levels of rRNA molecules (37). These results suggest that the rRNA-targeted RT-qPCR method enables the enumeration of both predominant and subdominant bacteria more sensitively and accurately than the usual methods. On the other hand, the rRNA sequence gives information that is limited to those on nucleotide sequences for group- or species-specific quantifications. For the detailed analysis of fecal microbiota, such as differentiations at the strain level, the qPCR method targeting the differences in the conservative gene sequences would be useful.

There have been several reports concerning the genus- or species-specific detection of Lactobacillus by PCR (7, 19, 45, 54, 55), but few reports on the quantitative analysis cover all inhabitants of the human intestines (14). It is widely acknowledged that the taxonomy of Lactobacillus is unsatisfactory due to the phylogenetic heterogeneity of this large assembly of microorganisms (5, 6). Additionally, 16S rRNA gene sequences of this genus are similar among species of Streptococcus, Lactococcus, and Enterococcus. Because of the limited information on homologous nucleotide sequences for the genus-specific detection of Lactobacillus, we used a strategy to differentiate Lactobacillus species into 13 subgroups based on 16S rRNA gene sequences and developed subgroup- or species-specific primers for 9 out of 13 subgroups to cover indigenous Lactobacillus inhabitants of the human intestines (Tables 1 and 2). RT-qPCR using these nine sets of primers quantified indigenous Lactobacillus species with higher frequency (98% in total) and higher population levels than the culture method (Table 3). L. gasseri, L. salivarius, L. ruminis, L. casei/L. paracasei, and L. rhamnosus have been reported as the major Lactobacillus species (42, 52, 53) in the adult intestines. The present results clearly showed that L. reuteri, L. plantarum, L. ruminis, L. gasseri, and L. casei subgroups are the predominant Lactobacillus subgroups present in the adult intestines (Table 4), which is comparable to the results of previous reports, and it is strongly suggested that newly developed subgroup- and species-specific primers for Lactobacillus cover indigenous Lactobacillus species in the human intestine. On the other hand, a wide variety of Lactobacillus species, e.g., L. casei, L. rhamnosus, L. acidophilus, L. delbrueckii subsp. bulgaricus, L. gasseri, L. johnsonii, L. plantarum, and L. reuteri, are in foods or are used as probiotics (21, 41). Therefore, it should be noted that the high detection frequency of these subgroups on a single occasion might reflect the transient bacteria ingested with food, and multiple occasions should be taken to signify a member of Lactobacillus of the bowel community.

RT-qPCR gave higher detection rates and higher counts for C. perfringens than those obtained by the culture method with some samples (Table 3, Fig. 3B). These differences might be because only lecithinase-C (α-toxin)-positive colonies on the C. welchii agar plate (egg-yolk medium) were identified as C. perfringens by the culture method, while RT-qPCR counted C. perfringens organisms irrespective of their lecithinase-C activity. C. perfringens is a species that causes food poisoning outbreaks with diarrhea and severe abdominal pain (18, 44). For the characterization of such opportunistic pathogens in the intestines, the evaluation of unique virulence factors such as the production of toxins, apparatus for invasion, and drug resistance is essential. Therefore, the identification of certain functions of bacteria should be the next objective after the quantification of their exact population levels by RT-qPCR, and it should lead to information about the activities of commensal bacteria in the corresponding environments. By using total RNA extracts, various information on bacterial functions will be available from the viewpoint of mRNA expressions. For a more precise analysis of C. perfringens, a combination assay of population size by rRNA-targeted RT-qPCR and mRNA expression of the phospholipase C gene (lecithinase-C) may be needed.

In this study, we developed an analytical system for the precise evaluation of human intestinal microbiota using rRNA-targeted RT-qPCR with group- and species-specific primer sets that allows the sensitive and accurate quantification of the target bacteria, especially for subdominant populations. This RT-qPCR method will enable the large-scale, systematic, quantitative, and comparative analysis of human intestinal microbiota, which should be effective for investigating several in situ aspects: the effects of probiotics or prebiotics and the relationship between microbiota and digestive diseases such as inflammatory bowel disease, infectious diseases, and colon cancer.