FEMS Microbiology Letters 183 (2000) 43^47
www.fems-microbiology.org
Rapid identi¢cation and quanti¢cation of Collinsella aerofaciens
using PCR
Akiko Kageyama *, Mitsuo Sakamoto, Yoshimi Benno
Japan Collection of Microorganisms, The Institute of Physical and Chemical Research (RIKEN), Wako-shi, Saitama 351-0198, Japan
Received 27 August 1999 ; received in revised form 10 November 1999; accepted 29 November 1999
Abstract
The number and incidence of Collinsella aerofaciens in the human intestine are the highest among Gram-positive non-spore-forming
bacilli. Identification of this species is very difficult and requires considerable time. A PCR-based identification system using C. aerofaciensspecific primers is described. Using this PCR method, we identified 181 C. aerofaciens-like species isolated from human feces. These 181
strains were identified using the traditional method in past studies. Results of both methods matched. The direct detection method was
performed using human feces samples from seven adults. Nested PCR was applied directly to the samples and all seven samples were
positive. Quantification studies were performed using LightCycler1. The assay uses a double-stranded DNA dye to continuously monitor
product formation and in a short time is able to quantify samples to 5 log units in concentration. ß 2000 Federation of European
Microbiological Societies. Published by Elsevier Science B.V. All rights reserved.
Keywords : Collinsella aerofaciens; Polymerase chain reaction ; 16S rDNA; Identi¢cation; Quanti¢cation
1. Introduction
The genus Eubacterium, an anaerobic Gram-positive
non-spore-forming bacillus, is one of the predominant microorganisms in the human intestinal micro£ora. However, its fastidious growth requirements coupled with an
intrinsically slow growth rate restrict rapid identi¢cation
in fecal £ora analysis [1,2]. Growth and physiological
characteristics of the eubacteria are not remarkable, and
therefore are not very helpful in identi¢cation. Some isolates belonging to the genus Eubacterium are frequently
identi¢ed at the genus level. Identi¢cation of these species
is very di¤cult and requires considerable time. Therefore,
it is necessary to establish an easy and rapid method for
identi¢cation of Eubacterium species.
Recently molecular techniques based on sequence composition of nucleic acids can be used to provide molecular
characterization [3^5]. The highly conserved regions of the
* Corresponding author. Tel. : +81 (48) 462-1111, ext. 5132 ;
Fax: +81 (48) 462-4619; E-mail: [email protected]
rRNA molecule can serve as primer binding sites for in
vitro ampli¢cation by PCR [6,7]. The more variable sequence regions are more appropriate for genus-, speciesand sometimes even strain-speci¢c probes [8^10].
E. aerofaciens, which is the predominant microorganism
in the human intestine [11^13], had been a member of the
genus Eubacterium, but since its phylogenetic position was
far from other Eubacterium spp. and its character was
unique, this species was recently transferred to Collinsella
aerofaciens [14]. The genus Collinsella has one species, C.
aerofaciens, but many strains belonging to this species
were found and their phylogenetic and phenotypic characters were investigated; 181 strains were identi¢ed using the
traditional method [14]. Since the identi¢cation of C. aerofaciens [15] is very di¤cult, we consider that an accurate
method for identi¢cation of this microorganism is necessary. For rapid identi¢cation of C. aerofaciens, a PCRbased [16] identi¢cation system is useful.
In the present study, we attempted to establish a PCRbased identi¢cation system using primers designed from
the 16S rRNA sequence of C. aerofaciens and related species and 181 strains of C. aerofaciens-like species isolated
from human feces. The PCR results matched those from
the studies using the traditional method [14]. Furthermore,
nested PCR using human fecal samples and quanti¢cation
of human fecal samples were also studied.
0378-1097 / 00 / $20.00 ß 2000 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved.
PII: S 0 3 7 8 - 1 0 9 7 ( 9 9 ) 0 0 6 2 7 - 8
FEMSLE 9201 18-1-00
44
A. Kageyama et al. / FEMS Microbiology Letters 183 (2000) 43^47
2. Materials and methods
2.1. Bacterial strains studied and cultivation
A total of 178 strains of C. aerofaciens isolated from
human feces and three strains of C. aerofaciens (JCM
10188T , JCM 7790 and JCM 7791), Eggerthella lenta
JCM 9979T , Coriobacterium glomerans JCM 10262T, Atopobium minutum JCM 1118T , E. barkeri JCM 1389T , E.
limosum JCM 6421T , E. multiforme JCM 6484T , E. nitritogenes JCM 6485T , E. tenue JCM 6486T , E. desmolans
JCM 6566T , E. cylindroides JCM 7786, Lactobacillus plantarum JCM 1149T , Bi¢dobacterium bi¢dum JCM 1255T ,
Propionibacterium propionicus JCM 5830T and Pseudoramibacter alactolyticum JCM 6480T were used in the
present study. The purity of all isolated strains was
checked by subculturing for 2 days at 37³C on Eggerth
Gagnon agar (Eiken, Tokyo, Japan) in an anaerobic jar
with 100% CO2 .
2.2. Primer design
Comparing the 16S rRNA sequences of E. aerofaciens
JCM 10188T , JCM 7790 and JCM 7791 and Eg. lenta
JCM 9979T , we designed one set of primers, AERO-F
(5P-CTTTCAGCAGGGAAGAGTCAA-3P) with positions
436^466 and AERO-R (5P-AGCCATGCACCACCTGTATGG-3P) with positions 1060^1039.
2.3. Preparation of bacterial DNA template
Bacterial DNA was prepared by suspending the cultured
colony in distilled water, heating at 100³C for 5 min and
then cooling. Bacterial DNA from human fecal samples
was prepared by the method described by Zhu et al. [17].
Brie£y, bacterial cells were suspended in extraction mixture (450 Wl) consisting of 250 Wl of extraction bu¡er (100
WM Tris-HCl, 40 WM EDTA, pH 9.0), 50 Wl of 10% SDS,
and 150 Wl of benzyl chloride. Following incubation at
50³C for 30 min with shaking, DNA was obtained by
isopropanol precipitation. DNA (100 Wl) was extracted
from 0.1 g of fecal sample. This extracted DNA was
used for the following examinations. For conventional
PCR and nested ¢rst PCR we used 10 Wl of extracted
DNA in 100 Wl of reaction mixture.
2.4. PCR conditions
PCR was performed at an annealing temperature of
63³C and 25 cycles using a DNA thermal cycler (PerkinElmer Cetus, Foster City, CA, USA). The PCR reaction
was conducted using a positive control of C. aerofaciens
and negative control of E. lenta. After PCR, an 8-Wl aliquot of ampli¢ed sample from each PCR tube was electrophoresed through 1 or 2% agarose gel (Sigma, St. Louis,
MO, USA) in TAE bu¡er for 30 min at 100 V. Ampli¢-
cation products were visualized and photographed under a
UV light transilluminator after 30 min of ethidium bromide staining. The molecular masses of the amplicons
were determined by comparison with commercial DNA
molecular mass markers.
When nested PCR was applied, we added the following
step before the above experiment : PCR was performed at
an annealing temperature of 55³C and 35 cycles using
prokaryotic 16S rDNA universal primers 27F (5P-AGAGTTTGATCCTGGCTCAG-3P) and 1492R (5P-GGTTACCTTGTTACGACTT-3P). The second PCR was done
using 10 Wl of the ¢rst PCR product.
2.5. Primer species speci¢city
Primer speci¢city was de¢ned as the ability of a primer
to anneal speci¢cally to only the C. aerofaciens 16S rRNA.
The speci¢city of the primers was tested against the following organisms: three C. aerofaciens strains (JCM
10188T , JCM 7790 and JCM 7791), three C. aerofacienslike species (A1-75, RCA57-66 and RCA57-64) isolated
from human feces, seven other Eubacterium species (E.
barkeri JCM 1389T , E. limosum JCM 6421T , E. multiforme
JCM 6484T , E. nitritogenes JCM 6485T , E. tenue JCM
6486T , E. desmorans JCM 6566T , and E. cylindroides
JCM 7786), and seven other genera (Eg. lenta JCM
9979T , P. alactolyticus JCM 6480T , C. glomerans JCM
10262T , L. plantarum JCM 1149T , B. bi¢dum JCM
1255T , Propionibacterium propionicus JCM 5830T and Atopobium minutum JCM 1118T ).
2.6. Ampli¢cation of isolated strains
Many strains were ¢rst isolated from human feces by
cultures on EG agar plates under anaerobic conditions
and many colonies were obtained. Cell morphology and
sugar fermentation were examined to check if these strains
were C. aerofaciens in past studies. Finally a total of 178
C. aerofaciens-like strains isolated from human feces were
tested using this PCR method.
2.7. Real-time quantitative PCR
PCR was performed in a £uorescence temperature cycler (LightCycler1, Roche, Mannheim, Germany).
Ampli¢cation was performed in a 20 Wl ¢nal volume
containing 4 mM MgCl2 , 2 Wl Mastermix (LightCyclerDNA Master SYBR Green I (Roche, Mannheim, Germany) containing Taq DNA polymerase, reaction bu¡er,
dNTP mix, and SYBR Green I dye), 2 Wl DNA template
(10 ng Wl31 ), 0.5 WM primer and 176 ng of TaqStart antibody (ClonTech, Palo Alto, CA, USA). The ampli¢cation
program included 45 cycles of three steps each, comprised
of heating at 20³C/s to 96³C with 0 s hold, cooling at
20³C/s to 66³C with 3 s hold, heating at 20³C/s to 72³C
with 24 s hold, and heating it 20³C/s to 87³C. Fluorescent
FEMSLE 9201 18-1-00
A. Kageyama et al. / FEMS Microbiology Letters 183 (2000) 43^47
product was detected at the last step of each cycle. After
ampli¢cation, a melting curve was obtained by heating at
20³C/s to 96³C, cooling at 20³C/s to 70³C, and slowly
heating at 0.2³C/s to 96³C with £uorescence collection at
0.2³C intervals. Melting curves were used to determine the
speci¢city of the PCR [18].
3. Results and discussion
To complete comprehensive studies on the e¡ect of dietary changes, aging, and health on the composition of intestinal micro£ora, rapid, sensitive, and speci¢c methods
for the detection of intestinal microbes are needed. A PCR
method with a genus- and species-speci¢c primer set was
studied. A database search revealed that C. aerofaciens
was close to Eg. lenta [19,20] and C. glomerans [21]; there-
45
fore, the 16S rRNA sequence of three C. aerofaciens
strains (JCM 10188T , JCM 7790 and JCM 7791) and the
published sequence of Eg. lenta were compared, and a set
of C. aerofaciens-speci¢c primers, AERO-F and AERO-R,
was designed. A database search revealed that these sequences did not match those of other species. The PCR
amplicon was about 590 bp and the optimum PCR conditions were 63³C for annealing temperature and 25 cycles.
To check the speci¢city of these primers, we performed
PCR with the following organisms : three strains (JCM
10188T , JCM 7790 and JCM 7791) of C. aerofaciens, three
strains (A1-75, RCA57-66 and RCA57-64) of C. aerofaciens-like microorganisms, other Eubacterium species (E.
barkeri JCM 1389T , E. limosum JCM 6421T , E. multiforme
JCM 6484T , E. nitritogenes JCM 6485T , E. tenue JCM
6486T , E. desmolans JCM 6566T , E. cylindroides JCM
7786) and other related genera (Eg. lenta JCM 9979T ,
Fig. 1. Fluorescent ampli¢cation pro¢les for C. aerofaciens PCR product in the presence of SYBR Green I. A: Serially diluted samples of C. aerofaciens
cells containing an estimated 109 , 108 , 107 , 106 , 105 , or 0 starting templates were prepared and ampli¢ed for 45 cycles as described in Section 2 (solid
lines). Uncounted samples 1^7 were also prepared and ampli¢ed for 45 cycles (dashed lines). B: Melting curves of PCR products. Serially diluted samples of C. aerofaciens cells containing as estimated 109 , 108 , 107 , 106 , 105 , or 0 starting templates are shown as solid lines and those of uncounted samples 1^7 as dashed lines. The rate of £uorescence change with changing temperature (3dF/dT) was plotted as a function of temperature.
FEMSLE 9201 18-1-00
46
A. Kageyama et al. / FEMS Microbiology Letters 183 (2000) 43^47
Fig. 2. SYBR Green I standard curve with unknown samples. The
crossing point is calculated using the LightCycler software. A line is ¢t
to the log-linear portion of each ampli¢cation curve. This ¢gure depicts
the crossing point plotted as a function of the starting copy number.
Pseudoanaerobacterium alactolyticum JCM 6480T , Coriobacterium glomerans JCM 10262T , A. minutum JCM
1118T , Lactobacillus plantarum JCM 1149T , Bi¢dobacterium bi¢dum JCM 1255T , Propionibacterium propionicus
JCM 5830T ). Three strains of C. aerofaciens and three
C. aerofaciens-like strains were all positive and the others
were all negative. These results showed that this PCR
method was very useful in identi¢cation of C. aerofaciens-like organisms. The 16S rRNA sequences of C. aerofaciens strains A1-75, RCA57-66 and RCA57-64 were also
determined and con¢rmed to be the same as those of the
designed primers. Furthermore, we con¢rmed that three
strains of C. aerofaciens, JCM 10188T , JCM 7790 and
JCM 7791, and C. aerofaciens-like organism A1-75 were
one species by DNA-DNA hybridization [14].
From 1976 to 1987, many anaerobic strains were isolated from human feces of Japanese and Canadians, patients with colon cancer, patients with gastric cancer, patients with ulcerative colitis, and patients with Crohn's
disease. These isolates were identi¢ed as C. aerofaciens
using cell morphology and sugar fermentation patterns.
A total of 178 strains of C. aerofaciens-like microorganisms were isolated. This traditional method is very timeconsuming and requires skill, and wrong results are occasionally obtained. In this study, the same 178 C. aerofaciens-like strains were identi¢ed using a PCR-based identi¢cation system. The PCR results matched those from the
studies using the more laborious and time-consuming traditional method.
This examination used isolated strains ; no fermentation
test was found to be very useful or rapid, but isolation was
also troublesome. Therefore, investigating a speci¢c PCR
method using direct samples was important. The direct
sample PCR presents various problems. Feces contain
many compounds, such as bilirubin and bile salts, which
can inhibit PCR analysis [22,23], and these inhibitors had
to be removed by washing the cells many times. We attempted to identify C. aerofaciens in seven adult human
fecal samples. The DNA from human fecal samples was
obtained by the method described by Zhu et al. [17].
Firstly, normal PCR with primers AERO-F and AEROR was studied. The result showed that the method was
successful for six samples, but one sample was not ampli¢ed. The amplicons were very weak, and this PCR should
be improved. Secondly, nested PCR was investigated.
Nested PCR consisted of two steps; the ¢rst PCR was
undertaken with universal primers, and a second PCR
was undertaken with speci¢c primers AERO-F and
AERO-R using the ¢rst PCR amplicon. This method
was very e¡ective and all seven samples were strongly
ampli¢ed. From these results, nested PCR appears to be
useful for identi¢cation of C. aerofaciens strains in human
fecal samples. However, nested PCR does not permit
quanti¢cation. Finally, we tried to quantify C. aerofaciens
in human fecal samples. Real-time PCR, a method based
on the incorporation of the £uorescent SYBR Green I dye
into ampli¢ed product during rapid cycle PCR, and detection of £uorescent dye release during melting of the product immediately after ampli¢cation, in a LightCycler1, is
useful for this purpose. Quanti¢cation of C. aerofaciens in
human fecal samples was performed with an externalstandard-based PCR. Measuring the ampli¢cation of targeted 16S rDNA from samples containing serially diluted
C. aerofaciens DNA was used to test the dynamic range of
this assay. Cycle-by-cycle collection of £uorescence generated a series of sigmoidal ampli¢cation pro¢les (Fig. 1A).
The £uorescent signal had three phases: an insu¤cient
dsDNA production phase, a log-linear phase, and a plateau phase. With fewer starting cells, more cycles were
required for detection. However, increasing the cycles
causes the speci¢city for C. aerofaciens to decrease, and
samples that contain few cells are not suitable for this
method. The data showed that the detection limit for
this method is more than 5 log cells in starting samples
(Fig. 1A). Following PCR, the products were melted, and
the release of £uorescent dye was measured to generate
melting curves from which Tm was calculated, another
feature of this assay. The melting point for the targeted
Table 1
Detection of C. aerofaciens in human fecal samples by three di¡erent
PCR methods
Sample
Normal PCRa
Nested PCRa
LightCyclerb
1
2
3
4
5
6
7
+
+
+
+
+
3
+
+
+
+
+
+
+
+
3.4U108
1.7U109
7.5U109
2.9U109
1.7U109
2.5U106
1.4U109
a
b
+, positive; 3, negative.
Cell number in 1 g of human feces.
FEMSLE 9201 18-1-00
A. Kageyama et al. / FEMS Microbiology Letters 183 (2000) 43^47
C. aerofaciens gene was 91.1³C while a nonspeci¢c melting
at around 78³C was assumed to be a primer dimmer.
To determine if nonspeci¢c product contributed to the
total £uorescent signal, data were collected at 87³C (Fig.
1B). C. aerofaciens in seven samples of human feces was
identi¢ed and quanti¢ed by the LightCycler1 PCR assay.
The standard curve is the plot of crossing point versus the
log of cell number. Fig. 2 shows a standard curve of C.
aerofaciens PCR product from 105 ^109 cells ml31 . The
cells were counted using microscopy. Each uncounted human fecal sample was calculated using this standard curve.
The results showed that all seven samples contained C.
aerofaciens, and the cell numbers were 2.5U106 ^7.5U109
g31 (Table 1, Fig. 2). Melting curves showed that amplicons were actually C. aerofaciens (Fig. 1B). In this study,
we used extracted DNA from isolated culture as the external standard. But for the future, it is necessary to examine the PCR quanti¢cation method using the standard
curve determined by similar background DNA to extracted DNA from fecal samples.
The PCR results for the seven samples in this study are
listed in Table 1. Samples 1^5 and 7 contained large numbers of C. aerofaciens, and all three methods were positive.
Because sample 6 did not contain such large numbers of
microorganisms, the normal PCR method was not useful
for identi¢cation. C. aerofaciens is the predominant microorganism in the human intestine, and using the LightCycler1 PCR assay, we can identify and quantify almost all
C. aerofaciens in human fecal samples.
In conclusion, PCR procedures for identi¢cation of C.
aerofaciens predominant in human feces were developed.
The PCR method with simple fecal sample preparation
may be useful for monitoring intestinal bacteria in humans.
Acknowledgements
This work was supported by a grant from the Yakult
Foundation for Bioscience Research.
References
[1] Lewis, R.P. and Sutter, V.L. (1981) The genus Eubacterium. In: The
Prokaryotes (Starr, M.P., Stolp, H., Truper, H.G., Balows, A. and
Schlegel, H.G., Eds.), pp. 1903^1911. Springer Verlag, Berlin.
[2] Moore, W.E.C. and Holdemann Moore, L.V. (1986) Genus Eubacterium Prevot 1938. In: Bergey's Manual of Systematic Bacteriology,
Vol. 2 (Sneath, P.H.A., Mair, N.S. Sharpe, M.E. and Holt, J.G.,
Eds.), pp. 1353^1373. Williams and Wilkins, Baltimore, MD.
[3] Olsen, G.J. and Woese, C.R. (1993) Ribosomal RNA : a key to phylogeny. FASEB J. 7, 113^123.
[4] Woese, C.R. (1987) Bacterial evolution. Microbiol. Rev. 51, 221^271.
[5] Woese, C.R. and Fox, G.E. (1977) Phylogenetic structure of the
prokaryotic domain : the primary kingdoms. Proc. Natl. Acad. Sci.
USA 74, 5088^5090.
47
[6] Ludwig, W., Dorn, S., Springer, N., Kirchhof, G. and Schleifer, K.H.
(1994) PCR-based preparation of 23S rRNA-targeted group-speci¢c
polynucleotide probes. Appl. Environ. Microbiol. 60, 3236^3244.
[7] Weisburg, W.G., Barns, S.M., Pelletier, D.A. and Lane, D.J. (1991)
16S ribosomal DNA ampli¢cation for phylogenetic study. J. Bacteriol. 1173, 697^703.
[8] Lin, C. and Stahl, D.A. (1995) Taxon-speci¢c probes for the cellulolytic genus Fibrobacter reveal abundant and novel equine-associated
populations. Appl. Environ. Microbiol. 61, 1348^1351.
[9] Odenyo, A.A., Mackie, R.I., Stahl, D.A. and White, B.A. (1994) The
use of 16S rRNA-targeted oligonucleotide probes to study competition between ruminal ¢brolytic bacteria : development of probes for
Ruminococcus species and evidence for bacteriocin production. Appl.
Environ. Microbiol. 60, 3688^3696.
[10] Stahl, D.A. and Amann, R. (1991) Development and application of
nucleic acid probes. In: Nucleic Acid Techniques in Bacterial Systematics (Stackebrandt, E. and Goodfellow, M., Eds.), pp. 205^248. John
Wiley and Sons, Chichester.
[11] Benno, Y., Suzuki, K., Suzuki, K., Narisawa, K., Bruce, W.R. and
Mitsuoka, T. (1986) Comparison of the fecal micro£ora in rural
Japanese and urban Canadians. Microbiol. Immunol. 30, 521^531.
[12] Finegold, S.M. and Sutter, V.L. (1978) Fecal £ora in di¡erent populations with special reference to diet. Am. J. Clin. Nutr. 27, 1456^
1469.
[13] Moore, W.E.C. and Holdeman, L.V. (1974) Human fecal £ora: the
normal £ora of 20 Japanese-Hawaiians. Appl. Microbiol. 27, 961^
979.
[14] Kageyama, A., Benno, Y. and Nakase, T. (1999) Phylogenic and
phenotypic evidence for the transfer of Eubacterium aerofaciens to
the genus Collinsella as Collinsella aerofaciens gen. nov., comb.
nov. submitted for publication. Int. J. Syst. Bacteriol. 49, 557^565.
[15] Moore, W.E.C., Cato, E.P. and Holdeman, L.V. (1971) Eubacterium
aerofaciens (Eggerth) Prevot 1938: Emendation of description and
designation of the neotype strain. Int. J. Syst. Bacteriol. 21, 307^310.
[16] Saiki, R.K., Gelfand, D.H., Sto¡el, S., Scharf, S.J., Higuchi, R.,
Horn, G.T., Mullis, K.B. and Erlich, H.A. (1988) Primer-detected
enzymatic ampli¢cation of DNA with a thermostable DNA polymerase. Science 239, 487^491.
[17] Zhu, H.l., Qu, F. and Zhu, L.-H. (1993) Isolation of genomic DNA
from plants, fungi and bacteria using benzyl cloride. Nucleic Acids
Res. 21, 5279^5280.
[18] Ririe, K.M., Rasmussen, R.P. and Wittwer, C.T. (1997) Product
di¡erentiation by analysis of DNA melting curves during the polymerase chain reaction. Anal. Biochem. 245, 154^160.
[19] Wade, W.G., Downes, J., Dymock, D., Hiom, S.J., Weightman, A.J.,
Dewhirst, F.E., Paster, B.J., Tzellas, N. and Coleman, B. (1999) The
family Coriobacteriaceae : reclassi¢cation of Eubacterium exiguum
(Poco et al., 1996) and Peptostreptococcus heliotrinreducens (Lanigan
1976) as Slackia exigua gen. nov., comb. nov. and Slackia heliotrinireducens gen. nov., comb. nov., and Eubacterium lentum (Prevot 1938)
as Eggerthella lenta gen. nov., comb. nov.. Int. J. Syst. Bacteriol. 49,
595^600.
[20] Kageyama, A., Benno, Y. and Nakase, T. (1999) Phylogenetic evidence for the transfer of Eubacterium lentum to the genus Eggerthella
as Eggerthella lenta gen. nov., comb. nov. Int. J. Syst. Bacteriol. 49,
in press.
[21] Hass, H. and Ko«nig, H. (1988) Coriobacterium glomerans gen. nov.,
sp. nov. from the intestinal tract of the red soldier bug. Int. J. Syst.
Bacteriol. 38, 382^384.
[22] Ktrader, C.A. (1995) Design and evaluation of Bacteroides DNA
probes for the speci¢c detection of human fecal pollution. Appl.
Environ. Microbiol. 61, 1171^1179.
[23] Wilde, J., Eiden, J. and Yolken, R. (1990) Removal of inhibitory
substances from human fecal specimens for detection of group A
rotaviruses by reverse transcriptase and polymerase chain reactions.
J. Clin. Microbiol. 28, 1300^1307.
FEMSLE 9201 18-1-00