Download Comparison of real-time PCR with SYBR Green I or

Microbiology (2003), 149, 269–277
DOI 10.1099/mic.0.25975-0
Comparison of real-time PCR with SYBR Green I
or 59-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
[email protected]
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.
Received 30 August 2002
Accepted 1 October 2002
INTRODUCTION
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
Abbreviation: GI, gastrointestinal.
0002-5975 G 2003 SGM
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 (59-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 59-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
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 12 Aug 2017 11:37:46
Printed in Great Britain
269
E. Malinen and others
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
Dot-blot hybridization. The oligonucleotide probes (Table 1) were
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
270
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.
59-end labelled with [g-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 ml, consisting of 16T4 polynucleotide kinase buffer (70 mM
Tris/HCl, 10 mM MgCl2, 5 mM DTT, pH 7?6), 20 U T4 polynucleotide kinase, 1 mM [g-33P]ATP [specific activity .3000 Ci mmol21
(.111 TBq mmol21)] and 1 mM 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 210 ˚C to Tm
+ 10 ˚C, and 100 ˚C using the equation Tm581?5 ˚C + 16?6 ˚C6
log10[Na + ] + 0?416(mol% G + C)2600/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). Prehybridization at 65 ˚C for 1 h in a buffer containing 56Denhart’s
solution, 56SSC, 0?5 % SDS and 100 mg denatured Herring Sperm
DNA ml21 (Sigma) was followed by the addition of the 59-end labelled
oligonucleotide probe to a final concentration of 7610212 mol ml21.
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 26SSC and 0?5 % SDS, four times at room
temperature (5 min) and once at Ti (15 min), followed by a rinse at
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 12 Aug 2017 11:37:46
Microbiology 149
Quantification of selected faecal bacteria
Table 1. Dot-blot hybridization probes and washing temperatures used in this study
Target bacterium
Bacteroides fragilis
Bifidobacterium lactis
Bifidobacterium longum3
E. coli4
L. acidophilus1
R. productus||
Oligonucleotide probe
Ti (˚C)*
59-GAAACATGTCAGTGAGCAATCACC-39
59-GTGGAGACACGGTTTCCCTT-39
59-GTTCCAGTTGATCGCATGGTCTT-39
59-GTTAATACCTTTGCTCATTGA-39
59-GATAGAGGTAGTAACTGGCCTTTA-39
59-GACATCC CTCTGACCGTCCCGT-39
56
51
55
54
56
54
*Ti refers to the stringent washing temperature, as defined in Methods.
3Bifidobacterium longum probe also detects the closely related species Bifidobacterium infantis, Bifidobacterium longum, Bifidobacterium
pseudolongum and Bifidobacterium suis.
4E. coli probe was designed for the Escherichia subgroup, composed of E. coli, Hafnia alvei and Shigella spp.
1L. acidophilus probe detects L. acidophilus, L. amylovorus, L. amylolyticus, L. crispatus, L. gasseri and L. johnsonii.
||R. productus probe also detects C. coccoides.
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).
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.
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 mM each dNTP, 1 mM
each primer and 0?03 U Dynazyme Polymerase II ml21 (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
Quantitative PCR was performed using an iCycler iQ apparatus (BioRad) associated with the ICYCLER OPTICAL SYSTEM INTERFACE software
(version 2.3; Bio-Rad). All PCRs were performed in triplicate in a
volume of 25 ml, 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
Table 2. PCR primers and 59-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.
PCR assay (amplicon size)
Oligonucleotide sequence (59R39)
Bacteroides fragilis (176 bp)
F: 59-GAAAGCATTAAGTATTCCACCTG-39
R: 59-CGGTGATTGGTCACTGACA-39
59-(HEX)-TGAAACTCAAAGGAATTGACGGGG-(DABCYL)-39
F: 59-CCCTTTCCACGGGTCCC-39
R: 59-AAGGGAAACCGTGTCTCCAC-39
59-(HEX)-AAATTGACGGGGGCCCGCACAAGC-(DABCYL)-39
F: 59-CAGTTGATCGCATGGTCTT-39
R: 59-TACCCGTCGAAGCCAC-39
59-(FAM)-TGGGATGGGGTCGCGTCCTATCAG-(TAMRA)-39
F: 59-GTTAATACCTTTGCTCATTGA-39
R: 59-ACCAGGGTATCTAATCC TGTT-39
59-(FAM)-CGTGCCAGCAGCCGCGGTA-(DABCYL)-39
F: 59-AGAGGTAGTAACTGGCCTTTA-39
R: 59-GCGGAAACCTCCCAACA-39
59-(FAM)-CGTGCCAGCAGCCGCGGTA-(DABCYL)-39
F: 59-GGTGGCAAAGCCATTCGGT-39
R: 59-GTTACGGGACGGTCAGAG-39
59-(HEX)-TGAAACTCAAAGGAATTGACGGGG-(DABCYL)-39
Bifidobacterium lactis (194 bp)
Bifidobacterium longum (106 bp)
E. coli (340 bp)
L. acidophilus (391 bp)
R. productus (182 bp)
http://mic.sgmjournals.org
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 12 Aug 2017 11:37:46
271
E. Malinen and others
Table 3. Optimized reaction conditions for SYBR Green I (SG) and TaqMan (TM) assays with the six oligonucleotide sets used
PCR test for
Tm (˚C)
Mg2 + (mM)
SG
Bacteroides fragilis
Bifidobacterium lactis
Bifidobacterium longum
E. coli
L. acidophilus
R. productus
58
58
58
60
58*
60
3
2
2?5
3
3
2
TM
2
1?5
4
6
4
4
dNTP (mM)
SG
TM
200
200
200
200
200
200
300
200
300
300
300
300
TM probe (nM)
400
100
80
200
200
400
Polymerase (U ml21)3
SG
TM
0?08
0?08
0?08
0?08
0?08
0?08
0?024
0?024
0?024
0?036
0?024
0?036
*Tm of 60 ˚C was used with the SYBR Green I assay.
3In the SYBR Green I assay, BlueTaq was used; in the TaqMan assay, Dynazyme II was used.
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 hotstart 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.
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 mM each dNTP, 0?5 mM each primer, 0?08 U BlueTaq ml21
and either 10 ml 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 59-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 mM each dNTP, 1 mM each primer,
80–400 nM fluorescent probe, 0?024–0?036 U Dynazyme II ml21 and
either 10 ml 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 nontarget 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.
272
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. 2–56106
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
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
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 12 Aug 2017 11:37:46
Microbiology 149
Quantification of selected faecal bacteria
Table 4. Detection of target bacteria using dot-blot hybridization with the six oligonucleotide probes designed
Species
Hybridization signals (¡SD)* with target DNA of
0 ng
Bacteroides fragilis
Bifidobacterium lactis
Bifidobacterium longum
E. coli
L. acidophilus
R. productus
261
602
148
278
91
245
(45)
(34)
(12)
(19)
(4)
(12)
10 ng
219
841
167
269
91
249
(5)
(65)
(14)
(5)
(1)
(9)
30 ng
217
1279
209
287
100
270
(5)
(196)
(6)
(18)
(3)
(6)
90 ng
250
2298
371
342
113
278
(6)
(196)
(8)
(19)
(4)
(7)
270 ng
335
4806
786
479
168
304
(29)
(57)
(25)
(5)
(3)
(23)
NC3
Cut-off4
810 ng
590
8952
1718
1429
334
305
(12)
(1399)
(8)
(123)
(25)
(13)
211
1150
164
277
96
236
(0)
(14)
(10)
(9)
(2)
(15)
312
1170
178
300
98
257
*In hybridization assays with 33P-labelled probe the signal intensities were quantified by autoradiography (QUANTITY ONE; Bio-Rad) as the sum of
the intensities of the pixels within the dot volume boundary by the pixel area (counts6mm2).
3The negative control (NC) was pooled from 270 ng of DNA from each of the other five species used in this study.
4The cut-off values were determined as the upper 95 % confidence intervals of the highest negative or non-template hybridization signal intensity
values. Values above the cut-off line are displayed in bold.
required for the generation of the positive hybridization
signal (Table 4).
Applicability of real-time PCR for the detection
of rDNA
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).
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 26105–46106 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)5
[10(21/slope)21], 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 hotstart 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.
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.
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 mg
of faecal DNA. &, Bacteroides fragilis assay; n, Bifidobacterium
lactis assay; $, Bifidobacterium longum assay; m, E. coli
assay; #, L. acidophilus assay; h, R. productus assay. Mean
values ¡SD values for three parallel samples are shown.
http://mic.sgmjournals.org
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
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 12 Aug 2017 11:37:46
273
E. Malinen and others
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
(2dRFU/dT) against temperature, is shown. $, Dynazyme II; #,
BlueTaq (hot-start polymerase). The mean values ¡SD values
from two replicates are displayed.
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) 59-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. &, Bacteroides fragilis assay; n, Bifidobacterium
lactis assay; $, Bifidobacterium longum assay; m, E. coli
assay; #, L. acidophilus assay; h, R. productus assay.
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.
274
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) 59-nuclease (TaqMan) assays. Purified genomic
DNA from the test target species was added to 20 ng of nontarget DNA pools in quantities of 10 000, 1000, 100, 10, 1
and 0?1 pg. &, Bacteroides fragilis assay; n, Bifidobacterium
lactis assay; $, Bifidobacterium longum assay; m, E. coli
assay; #, L. acidophilus assay; h, 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
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 12 Aug 2017 11:37:46
Microbiology 149
Quantification of selected faecal bacteria
Fig. 5. Detection of specific target DNAs mixed in faecal DNA
with the (a) SYBR Green I and (b) 59-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. &, Bacteroides fragilis assay; n,
Bifidobacterium lactis assay; $, Bifidobacterium longum assay;
m, E. coli assay; #, L. acidophilus assay; h, R. productus assay.
with the Bifidobacterium lactis PCR primer set (Table 2).
Quantification of the target bacteria was performed successfully when 3?16107 c.f.u. Bifidobacterium lactis cells
were introduced into 1 g of faeces. According to real-time
PCR results, parallel DNA preparations contained 1?876107
(0?705¡0?924 pg target DNA in 100 ng faecal DNA) and
8?946107 (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.
DISCUSSION
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
http://mic.sgmjournals.org
Fig. 6. Detection of specific target DNAs with the SYBR
Green I (m) or 59-nuclease (TaqMan; h) 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.
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 healthstatus-induced differences between or within populations.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 12 Aug 2017 11:37:46
275
E. Malinen and others
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 g21) 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 hotstart 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
276
(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
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.
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.
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.
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.
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.
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.
Heid, C. A., Stevens, J., Livak, K. J. & Williams, P. M. (1996). Real
time quantitative PCR. Genome Res 6, 986–994.
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.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 12 Aug 2017 11:37:46
Microbiology 149
Quantification of selected faecal bacteria
Hein, I., Lehner, A., Rieck, P., Klein, K., Brandl, E. & Wagner, M.
(2001). Comparison of different approaches to quantify Staphylococcus
Pearson, W. & Lipman, D. (1988). Improved tools for biological
aureus cells by real-time quantitative PCR and application of this
technique for examination of cheese. Appl Environ Microbiol 67,
3122–3126.
Sakata, S., Kitahara, M., Sakamoto, M., Hayashi, H., Fukuyama, M.
& Benno, Y. (2002). Unification of Bifidobacterium infantis and
Hopkins, M. J., Sharp, R. & Macfarlane, G. T. (2001). Age and disease
sequence comparison. Proc Natl Acad Sci U S A 85, 2444–2448.
Bifidobacterium longum. Int J Syst Evol Microbiol Papers in Press,
http://dx.doi.org/10.1099/ijs.0.02221-0
related changes in intestinal bacterial populations assessed by cell
culture, 16S rRNA abundance, and community cellular fatty acid
profiles. Gut 48, 198–205.
Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning:
Langendijk, P. S., Schut, F., Jansen, G. J., Raangs, G. J., Kamphuis,
G. R., Wilkinson, M. H. & Welling, G. W. (1995). Quantitative
Satokari, R. M., Vaughan, E. L., Akkermans, A. D. L., Saarela, M. &
De Vos, W. M. (2001a). Bifidobacterial diversity in human feces
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.
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.
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.
Malinen, E., Mättö, J., Salmitie, M., Alander, M., Saarela, M. &
Palva, A. (2002). PCR–ELISA. II: Analysis of Bifidobacterium popula-
tions in human faecal samples from a consumption trial with Bifidobacterium lactis Bb-12 and a galacto-oligosaccharide preparation.
Syst Appl Microbiol 25, 249–258.
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.
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.
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.
a Laboratory Manual, 2nd edn. Cold Spring Harbor, NY: Cold
Spring Harbor Laboratory.
detected by genus-specific PCR and denaturing gradient gel
electrophoresis. Appl Environ Microbiol 67, 504–513.
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.
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.
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.
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.
Suzuki, M. T., Taylor, L. T. & DeLong, E. F. (2000). Quantitative
analysis of small-subunit rRNA genes in mixed microbial populations via 59-nuclease assays. Appl Environ Microbiol 66, 4605–4616.
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.
Muttray, A. F. & Mohn, W. W. (2000). Quantitation of the population
Walter, J., Hertel, C., Tannock, G. W., Lis, C. M., Munro, K. &
Hammes, W. P. (2001). Detection of Lactobacillus, Pediococcus,
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.
Leuconostoc, and Weissella species in human feces by using groupspecific PCR primers and denaturing gradient gel electrophoresis.
Appl Environ Microbiol 67, 2578–2585.
Nadkarni, M. A., Martin, F. E. & Jacques, N. A. (2002). Determination
Wang, R.-F., Cao, W.-W. & Cerniglia, C. E. (1996). PCR detection and
of bacterial load by real-time PCR using a broad-range (universal)
probe and primers set. Microbiology 148, 257–266.
quantitation of predominant anaerobic bacteria in human and
animal fecal samples. Appl Environ Microbiol 62, 1242–1247.
Nogva, H. K., Bergh, A., Holck, A. & Rudi, K. (2000). Application of
Zoetendal, E. G., Akkermans, A. D. L. & De Vos, W. M. (1998).
the 59-nuclease PCR assay in evaluation and development of
methods for quantitative detection of Campylobacter jejuni. Appl
Environ Microbiol 66, 4029–4036.
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.
http://mic.sgmjournals.org
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Sat, 12 Aug 2017 11:37:46
277