abstract

Fecal microbiota were analyzed in seven healthy individuals by 16S rRNA gene libraries (universal library) using universal primers, and the Clostridium coccoides group libraries using the universal primer 27F and the C. coccoides group-specific primer Erec482. The universal libraries were used in our previous studies. The 972 clones obtained from two different primer set libraries belonged to the C. coccoides group and were classified into 139 operational taxonomic units (OTU) (at least 98% sequence similarity). Of these, 41 OTU were detected commonly from universal libraries and C. coccoides group libraries. One hundred and ten OTU were detected from the C. coccoides group libraries. Fifteen new OTU were isolated from the C. coccoides group libraries in human gut. Most of the OTU did not correspond to known species, thus representing as-yet-uncultured bacteria. We also detected OTU that related to the butyrate-producing bacteria. The C. coccoides group consisted of an average of 35 OTU, although there were differences in the number and the type of species in each individual. When fecal microbiota were analyzed using universal libraries, the OTU belonging to the C. coccoides group detected in elderly individuals were fewer than those detected in adult individuals. When C. coccoides group libraries were used, the numbers of OTU in elderly and adult individuals were not different. Interindividual differences in the composition of OTU belonging to the C. coccoides group were also observed in fecal microbiota.

Introduction

The human gut microbiota have traditionally been compared by analyzing isolates using culture-based methods. Consequently, it has been documented that 1012 bacteria cells per gram of content (dry matter) (Langendijk et al., 1995; Suau et al., 1999) and about 500–1000 species inhabit the human gut (Xu & Gordon, 2003; Sonnenburg et al., 2004). The composition and activity of indigenous intestinal microbiota are of paramount importance in human immunology, nutrition, and pathological processes, and hence the health of the individual (Van der Waaij et al., 1971). However, the cultivable bacteria form only 20–30% of the total because of the strict anaerobic and complex environment (Langendijk et al., 1995; Suau et al., 1999; Hayashi et al., 2002a). Recently, the application of molecular-biological techniques has allowed the phylogenetic analysis of bacterial 16S rRNA genes in the human gut. In particular, phylogenetic analysis based on 16S rRNA genes has made it possible to clarify the dominant human fecal microbiota (Wilson and Blitchington 1996; Zoetendal et al., 1998; Suau et al., 1999; Hold et al., 2002; Eckburg et al., 2005). We recently reported that the fecal microbiota in adults, vegetarian, and elderly individuals could be analyzed by universal libraries (Hayashi et al., 2002a, b, 2003, 2005). Among a total of about 2000 clones obtained, c. 75% of the clones were as-yet-uncultured bacteria. A large number of species that have not yet been identified exist in the human gut. Based on the results of phylogenetic analyses, the Clostridium leptum subgroup, the Clostridium coccoides group, and the Bacteroides group were considered the predominant bacteria in adult individuals.

The C. coccoides group is one of the most predominant groups in the human gut. According to previous reports (Suau et al., 1999; Hold et al., 2002; Hayashi et al., 2002a; Matsuki et al., 2004), this group constituted c. 25–60% of the total clones, and about 1010−1011 bacteria cells per gram of content (wet matter) inhabit the human gut. This group includes cultivable species of such genera as Butyrivibrio, Clostridium, Coprococcus, Dorea, Eubacterium, Lachnospira, Roseburia, and Ruminococcus, which are high oxygen-sensitive anaerobes (Suau et al., 1999; Hayashi et al., 2002a, 2003). Some of these species are known butyrate-producing bacteria, thereby contributing to processes important to colonic health (Barcenilla et al., 2000; Pryde et al., 2002). Based on the results of the universal libraries, most of the clones that belong to this group are as-yet-uncultured bacteria (Suau et al., 1999; Hold et al., 2002; Hayashi et al., 2002a, 2003; Eckburg et al., 2005). In addition, 20–30 operational taxonomic units (OTU) belonging to this group have been detected in each adult individual (Suau et al., 1999; Hayashi et al., 2002a). On the other hand, there are a few clones and OTU that belong to this group in elderly individuals (Hayashi et al., 2003). Although there are individual differences in composition OTU of this group, Ruminococcus obeum and relatives were detected at high frequency in adult human gut (Suau et al., 1999; Hayashi et al., 2002a; Zoetendal et al., 2002). However, the composition members and OTU numbers of this group are not known yet.

In the present study, we describe the diversity of the C. coccoides group in human gut microbiota using the C. coccoides group libraries with a C. coccoides-specific primer. We also compare with universal libraries already described in previous studies (Hayashi et al., 2002a, 2003).

Materials and methods

DNA extraction

Stool samples were collected from seven healthy individuals (S: 27-year-old male, O: 28-year-old male, B: 52-year-old male, W: 25-year-old female, OLDA: 94-year-old male, OLDB: 88-year-old female, and OLDC: 75-year-old female) who had not been treated with antibiotics for 3 months before the study. The fresh fecal sample was suspended in buffer A [10 mM Tris-HCl and 50 mM ethylenediaminetetra-acetic acid (EDTA), pH 7.5] and centrifuged at 12 000 g for 5 min. The pellet was resuspended in buffer A and centrifuged at 12 000 g for 5 min. This operation was repeated four times. The following procedures were carried out as described previously (Hayashi et al., 2002a, 2003).

PCR amplification and cloning

Universal primer 27F (5′ AGAGTTTGATCCTGGCTCAG 3′) (Lane, 1991) and the Clostridium coccoides group-specific primer Erec482 (5′ GCT TCT TAG TCA RGT ACC G 3′) (Franks et al., 1998) were used to amplify the 16S rRNA genes of the C. coccoides group. PCR amplification was performed using the following program: 95°C for 3 min, followed by 15 cycles consisting of 95°C for 30 s, 50°C for 30 s, 72°C for 1.5 min, and a final extension period of 72°C for 10 min. The amplified 16S rRNA genes were purified using the UltraClean PCR Clean-up kit (Mo Bio Laboratories Inc., Carlsbad, CA). The purified amplicons from fecal samples were ligated into the plasmid vector pCR®2.1. One Shot® INVαF′ competent cells (Invitrogen, San Diego, CA) were transformed with the ligation mixture. Plasmid DNA of selected transformants was purified using MultiScreen 96-well filter plates (Millipore, Bedford, MA). The stool sample from subject W was analyzed for the microbiota using two-universal primers 27F and 1492R (5′ GGTTACCTTGTTACGACTT 3′) (Lane, 1991).

DNA sequencing and phylogenetic analysis

Plasmid DNAs from the universal libraries were used as templates for sequencing. An equal portion (about 500 bp) of 16S rRNA gene (Escherichia coli position 28–482) was used for sequence analysis. The dideoxy chain termination reaction was conducted with a double-stranded DNA template and 27F or Erec482R primer using the BigDye Terminator Cycle Sequencing Kit (Applied Biosystems, Foster City, CA), and products were analyzed on a model ABI PRISM 3100 DNA analyzer system (Applied Biosystems). Nucleotide sequences were analyzed with the FASTA search and Sequence Match program of the Ribosomal Database Project-II (RDP-II) (Cole et al., 2003). All sequences were examined for possible chimeric artifacts by the CHECK CHIMERA program of the RDP-II (Cole et al., 2003).

Sequence data were aligned with the CLUSTAL W (Thompson et al., 1994) package and corrected by manual inspection. The nucleotide substitution rates (Knuc values) were calculated (Kimura, 1980) and phylogenetic trees were constructed using the neighbor-joining method (Saitou & Nei, 1987). The OTU was previously used to describe clusters of clone sequences that differed from known species by about 2% and were at least 98% similar to members of their cluster (Suau et al., 1999). Identical OTU detected from universal libraries or the C. coccoides group libraries were counted as one OTU.

Nucleotide sequence accession numbers

The GenBank/EMBL/DDBJ accession numbers for the novel 16S rRNA gene sequences, which belonged to the C. coccoides group, detected by this study are AB231643–AB231657.

Results and discussion

The diversity of the bacterial community structure of the Clostridium coccoides group was evaluated by generating universal libraries using universal primers and for the C. coccoides group libraries, using the C. coccoides group-specific primer (Erec482). The primer Erec482 was prepared as a specific probe for the C. coccoides group and has been used to analyze only the bacterial population of this group in in situ hybridization or flow cytometry (Franks et al., 1998; Rigottier-Gois et al., 2002; Zoetendal et al., 2002). The results of these studies showed that although the primer is not sufficient to analyze the diversity of this group, this group is the most predominant group in human fecal microbiota. Thus, we used primer Erec482 to analyze the diversity of the C. coccoides group. About 90 clones were randomly selected from each library and the partial sequence of c. 500 bp was determined for each clone. A total of 139 OTU were obtained from the universal and the C. coccoides group libraries. One hundred and one OTU were detected from the C. coccoides group libraries (Fig. 1). Forty-one OTU were detected commonly from universal libraries and the C. coccoides group libraries. The C. coccoides group exhibited a great diversity in the human gut based on analysis using two primer sets. The analysis of the diversity of the C. coccoides group is not completely accurate when using universal primers alone. In this regard, it is difficult to detect small numbers of OTU (species) using universal primers because of the preferential detection of predominant OTU (species). Actually, ‘Gammaproteobacteria’ and the Clostridium ramosum assemblage OTU were detected at high frequencies in two elderly subjects (OLDB-U and OLDC-U) with small number of OTU belonging to the C. coccoides group (Hayashi et al., 2003). These results suggest that the use of two kinds of primers allows elaborate analysis of the human gut microbiota.

1

Phylogenetic tree showing the relationship of 16S rRNA gene sequences in fecal samples within the Clostridium coccoides group. The tree was constructed using neighbor-joining analysis based on 16S rRNA gene sequences. Bootstrap values (n=100 replicates) of ≥50 are considered as percentage. The scale bar represents 0.02 substitutions per nucleotide position. Clones from the C. coccoides group libraries are shown by each character (O, sample O; B, sample B; S, sample S; W, sample W; OLDA, sample OLDA; OLDB, sample OLDB; OLDC, sample OLDC). Clones from the universal libraries appear as underlined letters in parentheses and are shown by each character (U-ON, sample O; U-BN, sample B; U-NS, sample S; U-W, sample W; U-OLDA, sample OLDA; U-OLDB, sample OLDB; U-OLDC, sample OLDC). New operational taxonomic units (OTU) that were detected using the C. coccoides group-specific primer are shown by open arrows. OTU that were detected from libraries using C. coccoides-specific primer but not universal primers are marked with an asterisk. OTU that relate to the butyrate-producing bacteria are shown by solid arrows (Barcenilla et al., 2000; Louis et al., 2004). The Clostridium propionicum sequence was used as the outgroup for rooting the tree.

1

Phylogenetic tree showing the relationship of 16S rRNA gene sequences in fecal samples within the Clostridium coccoides group. The tree was constructed using neighbor-joining analysis based on 16S rRNA gene sequences. Bootstrap values (n=100 replicates) of ≥50 are considered as percentage. The scale bar represents 0.02 substitutions per nucleotide position. Clones from the C. coccoides group libraries are shown by each character (O, sample O; B, sample B; S, sample S; W, sample W; OLDA, sample OLDA; OLDB, sample OLDB; OLDC, sample OLDC). Clones from the universal libraries appear as underlined letters in parentheses and are shown by each character (U-ON, sample O; U-BN, sample B; U-NS, sample S; U-W, sample W; U-OLDA, sample OLDA; U-OLDB, sample OLDB; U-OLDC, sample OLDC). New operational taxonomic units (OTU) that were detected using the C. coccoides group-specific primer are shown by open arrows. OTU that were detected from libraries using C. coccoides-specific primer but not universal primers are marked with an asterisk. OTU that relate to the butyrate-producing bacteria are shown by solid arrows (Barcenilla et al., 2000; Louis et al., 2004). The Clostridium propionicum sequence was used as the outgroup for rooting the tree.

Of the 139 OTU, only nine possessed sequences closely related to those of cultivated bacteria (Fig. 1). No OTU that belonged to known species were detected commonly in all subjects. Most of the OTU that belonged to the C. coccoides group were uncultured bacteria. Fifteen new OTU were detected from these libraries for the first time. In addition, 11 of these 15 OTU were detected in elderly individuals. Recently, Eckburg (2005) analyzed microbiota in various sites of the large bowel in three individuals using universal libraries, and 13 335 clones of bacterial 16S rRNA genes were extracted from each site of the intestinal tract. However, they were able to detect these new OTU in only a minority of adult subjects. In this regard, we previously reported the existence of noticeable differences in the composition of fecal microbiota between adults and elderly individuals (Hayashi et al., 2002a, 2003, 2005). Most of the novel OTU were detected in elderly individuals, and thus may be specific to elderly individuals. It is necessary to analyze such OTU in more detail and determine the relationship between them and aging.

On average, 35 OTU were detected for each individual based on analysis using two primer sets (Table 1). In addition, major interindividual differences were noted in the composition of the C. coccoides group in fecal microbiota (Fig. 1). Only two OTU were detected from universal libraries in samples OLDB and OLDC. On the other hand, 29 and 25 OTU were detected from C. coccoides group libraries in the above two samples (Table 1). The number of OTU in elderly individuals (OLDA, OLDB, and OLDC) increased two- to threefold, especially when the C. coccoides group-specific primer was used (Table 1). Based on the finding of the universal libraries and T-RFLP using universal primers, the proportion of the C. coccoides group was lower in elderly individuals than in adult individuals (2.5–25.3% of total clones, 2–14 OTU) (Hayashi et al., 2003). The C. coccoides group was detected at high frequency from universal libraries in adult individuals (24–59% of total clones, 31–35 OTU) (Suau et al., 1999; Hayashi et al., 2002a). Matsuki (2004) detected the C. coccoides group at high frequency by real-time PCR in adult individuals (17–41% of total cell count). When the human gut microbiota were analyzed using universal primers, predominant OTU were selectively detected. In other words, the C. coccoides group was not readily detected in certain subjects like elderly individuals, and such group comprised a small number of species. As for the subjects with a few OTU from universal libraries, many OTU were detected from C. coccoides group libraries. Based on the results of the analysis of the C. coccoides group library, the elderly individuals with low-diversity C. coccoides group (as determined by analysis of universal libraries) also showed great diversity, similar to adult individuals. When the C. coccoides group-specific primer was used, we had to pay attention to nonspecific amplification because the 16S rRNA gene sequences other than the C. coccoides group were detected in samples OLDB and OLDC (10 and 17 clones). The sample with a small number of species causes nonspecific amplification. The C. coccoides group exhibited a great diversity in all individuals, although there were differences in the number of species.

1

Number of operational taxonomic unit that belonged to the Clostridium coccoides group in seven individuals

 Number of OTU Number of clones 
O-C 31 90 
O-U 29 127 
O-Total 45 217 
S-C 16 90 
S-U 13 76 
S-Total 20 166 
B-C 18 95 
B-U 20 63 
B-Total 31 158 
W-C 30 85 
W-U 16 55 
W-Total 35 140 
OLDA-C 44 85 
OLDA-U 14 20 
OLDA-Total 55 105 
OLDB-C 29 89 
OLDB-U 
OLDB-Total 29 91 
OLOC-C 25 92 
OLOC-U 
OLDC-Total 25 95 
 Number of OTU Number of clones 
O-C 31 90 
O-U 29 127 
O-Total 45 217 
S-C 16 90 
S-U 13 76 
S-Total 20 166 
B-C 18 95 
B-U 20 63 
B-Total 31 158 
W-C 30 85 
W-U 16 55 
W-Total 35 140 
OLDA-C 44 85 
OLDA-U 14 20 
OLDA-Total 55 105 
OLDB-C 29 89 
OLDB-U 
OLDB-Total 29 91 
OLOC-C 25 92 
OLOC-U 
OLDC-Total 25 95 

O, sample O; S, sample S; B, sample B; W, sample W; OLDA sample, OLDA; OLDB, sample OLDB; OLDC, sample OLDC; C, fecal microbiota analyzed by the Clostridium coccoides group libraries; U, fecal microbiota analyzed by universal libraries; Total, operational taxonomic units (OTU) detected from universal libraries as well as the C. coccoides group libraries were counted as one OUT.

Ruminococcus obeum and relatives were hardly detected in elderly individuals, although they were detected at high frequency from universal libraries in adult individuals (Hayashi et al., 2002a, 2003). Hold (2002) also could not detect these bacteria in colonic tissue samples from elderly individuals. On the other hand, Zoetendal (2002) reported that these microorganisms and their relatives comprised about 16% of the C. coccoides group in their subjects aged 25–32 years. The 207 clones related to R. obeum and relatives were classified into 27 OTU. We also detected them at high frequency from the C. coccoides libraries in all subjects (5.4–27.0% of each total clones) (Fig. 1). Ruminococcus obeum and relatives were one of the most important species in the C. coccoides group. They exist commonly in human gut regardless of age, though the number of these clones tends to diminish with age.

Short-chain fatty acids (SCFA) produced by bacterial fermentation of undigested starches and dietary fiber have important effects on colonic health (Szylit & Andrieux, 1993). Barcenilla (2000) and Louis (2004) reported the isolation of many butyrate-producing bacteria from human feces. Many of these isolates belonged to the C. coccoides group and were microorganisms that have not yet been isolated previously. In the present study, 10 OTU corresponded to these butyrate-producing bacteria (Fig. 1). These OTU were detected in both adult and elderly individuals. There were also OTU detected at high frequencies among these OTU (Eubacterium rectale OTU, Roseburia intestinalis OTU, and O-A12 OTU). Eubacterium rectale and R. intestinalis were detected as butyrate-producing bacteria at high frequencies by culture-based methods (Barcenilla et al., 2000). Butyrate-producing bacteria (OTU) are important members of the C. coccoides group.

In conclusion, our results clearly showed that the C. coccoides group is of great diversity based on the use of their group-specific primer for analysis of the 16S rRNA gene library. Most OTU that belonged to this group were as-yet-uncultured bacteria. The OTU (clones) that belonged to the C. coccoides group were detected by their specific primer in subjects found to be negative by universal primers. We also detected several new OTU that have not been previously identified. Future studies should use various OTU-specific primers designed for use for rapid detection of the C. coccoides group.

Acknowledgement

This study was supported by a grant from the Special Postdoctoral Research Program of RIKEN, Saitama, Japan.

References

Barcenilla
A
Pryde
SE
Martin
JC
Duncan
SH
Stewart
CS
Henderson
C
Flint
HJ
(
2000
)
Phylogenetic relationships of butyrate-producing bacteria from the human gut
.
Appl Environ Microbiol
 
66
:
1654
1661
.
Cole
JR
Chai
B
Marsh
TL
et al
(
2003
)
The Ribosomal database project (RDP-II): previewing a new autoaligner that allows regular updates and the new prokaryotic taxonomy
.
Nucleic Acids Res
 
31
:
442
443
.
Eckburg
PB
Bik
EM
Bernstein
CN
Purdom
E
Dethlefsen
L
Sargent
M
Gill
SR
Nelson
KE
Relman
DA
(
2005
)
Diversity of the human intestinal microbial flora
.
Science
 
308
:
1635
1638
.
Franks
AH
Harmsen
HJM
Raangs
GC
Jansen
GJ
Schut
F
Welling
GW
(
1998
)
Variations of bacterial populations in human feces measured by fluorescent in situ hybridization with group-specific 16S rRNA-targeted oligonucleotide probes
.
Appl Environ Microbiol
 
64
:
3336
3345
.
Hayashi
H
Sakamoto
M
Benno
Y
(
2002a
)
Phylogenetic analysis of the human gut microbiota using 16S rDNA clone libraries and strictly anaerobic culture-based methods
.
Microbiol Immunol
 
46
:
535
548
.
Hayashi
H
Sakamoto
M
Benno
Y
(
2002b
)
Fecal microbial diversity in a strict vegetarian as determined by molecular analysis and cultivation
.
Microbiol Immunol
 
46
:
819
831
.
Hayashi
H
Sakamoto
M
Kitahara
M
Benno
Y
(
2003
)
Molecular analysis of fecal microbiota in elderly individuals using 16S rDNA library and T-RFLP
.
Microbiol Immunol
 
47
:
557
570
.
Hayashi
H
Takahashi
R
Nishi
T
Sakamoto
M
Benno
Y
(
2005
)
Molecular analysis of jejunal, ileal, cecal and recto-sigmoidal human colonic microbiota using 16S rRNA gene libraries and terminal restriction fragment length polymorphism
.
J Med Microbiol
 
54
:
1093
1101
.
Hold
GL
Pryde
SE
Russell
VJ
Furrie
E
Flint
HJ
(
2002
)
Assessment of microbial diversity in human colonic samples by 16S rDNA sequence analysis
.
FEMS Microbiol Ecol
 
39
:
33
39
.
Kimura
M
(
1980
)
A simple method for estimating the evolutionary rate of base substitutions through comparative studies of nucleotide sequences
.
J Mol Evol
 
16
:
111
120
.
Lane
DJ
(
1991
)
16S/23S rRNA sequencing
.
Nucleic Acid Techniques in Bacterial Systematices
  (
Stackebrandt
E.
Goodfellow
M.
, eds), pp.
115
175
.
Wiley
,
New York
.
Langendijk
PS
Schut
F
Jansen
GJ
Raangs
GC
Kamphuis
GR
Wilkinson
MH
Welling
GW
(
1995
)
Quantitative fluorescence in situ hybridization of Bifidobacterium spp. with genus-specific 16S rRNA-targeted probes and its application in fecal samples
.
Appl Environ Microbiol
 
61
:
3069
3075
.
Louis
P
Duncan
SH
McCrae
SI
Millar
J
Jackson
MS
Flint
HJ
(
2004
)
Restricted distribution of the butyrate kinase pathway among butyrate-producing bacteria from the human colon
.
J Bacteriol
 
186
:
2099
2106
.
Matsuki
T
Watanabe
K
Fujimoto
J
Takada
T
Tanaka
R
(
2004
)
Use of 16S rRNA gene-targeted group-specific primers for real-time PCR analysis of predominant bacteria in human feces
.
Appl Environ Microbiol
 
70
:
7220
7228
.
Pryde
SE
Duncan
SH
Hold
GL
Stewart
CS
Flint
HJ
(
2002
)
The microbiology of butyrate formation in the human colon
.
FEMS Microbiol Lett
 
217
:
133
139
.
Rigottier-Gois
L
Le Bourhis
AG
Gramet
G
Rochet
V
Dore
J
(
2002
)
Fluorescent hybridization combined with flow cytometry and hybridization of total RNA to analyze the composition of microbial communities in human faeces using 16S rRNA probes
.
FEMS Microbiol Ecol
 
43
:
237
245
.
Saitou
N
Nei
M
(
1987
)
The neighbor-joining method: a new method for reconstructing phylogenetic trees
.
Mol Biol Evol
 
4
:
406
425
.
Sonnenburg
JL
Angenent
LT
Gordon
JI
(
2004
)
Getting a grip on things: how do communities of bacterial symbionts become established in our intestine?
Nat Immunol
 
5
:
569
573
.
Suau
A
Bonnet
R
Sutren
M
Godon
J-J
Gibson
GR
Collins
MD
Dore
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
.
Szylit
O
Andrieux
C
(
1993
)
Physiological and pathophysiological effects of carbohydrate fermentation
.
World Rev Nutr Diet
 
74
:
88
122
.
Thompson
JD
Higgins
DG
Gibson
TJ
(
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
.
Van der Waaij
Berghuis-de Vries
DJM
Lekkerkerk-van der Wees
JEC
(
1971
)
Colonization resistance of the digestive tract in conventional and antibiotic-treated mice
.
J Hygiene
 
67
:
405
411
.
Wilson
KH
Blitchington
RB
(
1996
)
Human colonic biota studied by ribosomal DNA sequence analysis
.
Appl Environ Microbiol
 
62
:
2273
2278
.
Xu
J
Gordon
JI
(
2003
)
Honor thy symbionts
.
Proc Natl Acad Sci USA
 
100
:
10452
10459
.
Zoetendal
EG
Akkermans
AD
de Vos
WM
(
1998
)
Temperature gradient gel electrophoresis analysis of 16S rRNA from human fecal samples reveals stable and host-specific communities of active bacteria
.
Appl Environ Microbiol
 
64
:
3854
3859
.
Zoetendal
EG
Ben-Amor
K
Harmsen
HJM
Schut
F
Akkermans
ADL
de Vos
WM
(
2002
)
Quantification of uncultured Ruminococcus obeum-like bacteria in human fecal samples by fluorescent in situ hybridization and flow cytometry using 16S rRNA-targeted probes
.
Appl Environ Microbiol
 
68
:
4225
4232
.