Abstract
The worldwide presence of a hitherto-nondescribed group of predominant soil microorganisms related to Bacillus benzoevorans was analyzed after development of two sets of selective primers targeting 16S rRNA genes in combination with denaturing gradient gel electrophoresis (DGGE). The high abundance and cultivability of at least some of these microorganisms makes them an appropriate subject for studies on their biogeographical dissemination and diversity. Since cultivability can vary significantly with the physiological state and even between closely related strains, we developed a culture-independent 16S rRNA gene-targeted DGGE fingerprinting protocol for the detection of these bacteria from soil samples. The composition of the B. benzoevorans relatives in the soil samples from The Netherlands, Bulgaria, Russia, Pakistan, and Portugal showed remarkable differences between the different countries. Differences in the DGGE profiles of these communities in archived soil samples from the Dutch Wieringermeer polder were observed over time during which a shift from anaerobic to aerobic and from saline to freshwater conditions occurred. To complement the molecular methods, we additionally cultivated B. benzoevorans-related strains from all of the soil samples. The highest number of B. benzoevorans relatives was found in the soils from the northern part of The Netherlands. The present study contributes to our knowledge of the diversity and abundance of this interesting group of microbes in soils throughout the world.
The soil ecosystem is known to provide ecological niches for an extremely high number of microorganisms (31). Investigations in Drentse A grassland soils in The Netherlands have demonstrated the abundance of 16S rRNA gene sequences that originated from a group related to Bacillus benzoevorans (6). Analysis of a clone library based on direct recovery of 16S rRNA genes from this Dutch soil indicated that a substantial proportion (32%) of the bacterial sequences belonged to this novel phylogenetic lineage, which includes B. benzoevorans, B. niacini, B. pseudomegaterium, B. jeotgali (34), five recently isolated species comprising B. novalis, B. vireti, B. soli, B. bataviae, and B. drentensis (14), the uncultured Drentse A bacterium DA001, and their relatives (10). Their homogeneous distribution and activity over kilometer distances has been demonstrated previously in the Dutch Drentse A grasslands (7). Hence, it is feasible to broaden the focus for a comparison of more distant landscapes. After its discovery in Dutch soils, more uncultured members of this lineage were reported to be present as the predominant group of Bacillus spp. in soil samples from other countries, including the United States (Wisconsin) (2), Amazonia (3), The Netherlands (8), and the United Kingdom (19).
Since the B. benzoevorans-related bacteria are so abundant, they provide an attractive object for evolutionary studies and the opportunity to examine the microdiversity of one of the most dominant microbial groups in the grassland soil biosphere. Moreover, only a fraction of the bacteria from this cluster has been cultivated thus far (11a, 15). Although new approaches continue to be developed, it is well recognized that the larger part of the soil community remains inaccessible to detailed studies due to inadequate media and methods for cultivation (1, 2, 35).
The purpose of the present study was to develop tools for the fast and sensitive cultivation-independent detection of B. benzoevorans-related bacteria from soil samples. Molecular fingerprinting methods targeting 16S rRNA genes and rRNA, such as denaturing gradient gel electrophoresis (DGGE) or temperature gradient gel electrophoresis can be used for the rapid evaluation of composition and activity of complex microbial assemblages at moderately high temporal and spatial resolution (21). The design and validation of 16S rRNA gene-targeted primer sets specific for the cluster of B. benzoevorans relatives for use in combination with DGGE are presented here. These primers allow for the monitoring of the distribution and succession of these bacteria over time and space. This molecular approach was complemented with cultivation of isolates from several European locations that revealed the presence and sometimes remarkable abundance of these bacteria.
MATERIALS AND METHODS
Bacterial strains, growth conditions, and soil samples.
Reference strains used for validation of the newly developed primers for detection of the predominant B. benzoevorans relatives are listed in Table 1 with their respective sources and growth media. The primers were further tested and validated by using pure culture isolates that were isolated and identified by 16S rRNA gene sequence analysis (Table 1).
TABLE 1.
Reference strains and pure cultures of B. benzoevorans relatives isolated and characterized in this study, growth conditions, and sources used for validation of the new B. benzoevorans cluster-specific primer sets (REX460f-REX1466r and REX576f-REX1446r)
| Species or isolatesa | Mediumb | Collection or GenBank accession no. | Primer setc
|
|
|---|---|---|---|---|
| REX460f-REX1466r | REX576f-REX1446r | |||
| Reference species | ||||
| Bacillus benzoevorans relatives | ||||
| Bacillus benzoevorans | 1 | LMG15526 | + | + |
| Bacillus pseudomegaterium | 1 | LMG18519 | + | + |
| Bacillus niacini | 1 | LMG16677T | + | + |
| Bacillus jeotgali | Marine broth | JCM10885 | + | + |
| Non-Bacillus benzoevorans relatives | ||||
| Paenibacillus polymixa | 51 | LMG6319T | − | NP |
| Bacillus methanolicus | CM0129 | NCIMB13113 | − | NP |
| Bacillus cereus | 1 | LMG6923T | − | NP |
| Bacillus simplex | 1 | LMG11160T | + | − |
| Bacillus megaterium | 1 | LMG7127T | + | − |
| Bacillus mycoides | 1 | LMG7128T | − | NP |
| Bacillus subtilis | 1 | LMG7135T | − | NP |
| Escherichia coli | 1 | LMG18221 | − | NP |
| Staphylococcus hominis | 1 | LMG13348T | − | NP |
| Streptomyces murinus | 78 | LMG10475T | − | NP |
| Sporosarcina globispora | 1 or 51 | LMG6928T | + | − |
| New isolatesd | ||||
| Bacillus benzoevorans relatives | ||||
| Bacillus benzoevorans IDA4715 (97) | Sodium benzoate | AY289495 | + | + |
| Bacillus benzoevorans IDA4919 (98) | Sodium benzoate | AY289496 | + | + |
| Bacillus bataviensis IDA4789 (95) | Sodium benzoate | AY289499 | + | + |
| Bacillus bataviensis IDA4730 (99) | Sodium benzoate | AY289497 | + | + |
| Bacillus bataviensis IDA4740 (97) | Sodium benzoate | AY289498 | + | + |
| Uncultured bacterium DA001 IDA4921 (97) | Sodium benzoate | AY289500 | + | + |
| Non-Bacillus benzoevorans relatives | ||||
| Paenibacillus sp. strain IDA4508 (96) | Sodium acetate | AY289501 | − | NP |
| Paenibacillus sp. strain IDA5358 (99) | Sodium acetate | AY289507 | − | NP |
| Bacillus mycoides IDA4575 (99) | Sodium acetate | AY289502 | − | NP |
| Bacillus megaterium IDA4770 (99) | Sodium benzoate | AY289503 | + | − |
| Bacillus simplex IDA4917 (98) | Sodium benzoate | AY289504 | + | − |
| Bacillus macroides IDA5330 (99) | Sodium benzoate | AY289506 | + | − |
| Bacillus megaterium IDA5367 (99) | Sodium acetate | AY289508 | + | − |
| Streptomyces mirabilis IDA5203 (99) | Sodium acetate | AY289505 | − | NP |
Reference strains were from various strain collections as follows: LMG, Laboratory of Microbiology, University of Ghent, Ghent, Belgium; NCIMB, National Collections of Industrial, Food, and Marine Bacteria, NCIMB, Ltd., Aberdeen, United Kingdom; JCM, Japan Collection of Microorganisms, RIKEN (The Institute of Physical and Chemical Research), Tokyo, Japan. New isolates were from Wageningen University.
Media: 1, see reference 18; medium 51, medium 1 amended with glucose at 1 g/liter; CM0129, tryptone soya broth (Oxoid, Haarlem, The Netherlands); medium 78, BCCM (Bacterial Collection, LMG, University of Ghent).
+, PCR products generated with respective primer pair; −, no PCR products generated with the respective primer pair. After a negative PCR result with REX460f-REX1466r, the subsequent REX576f-REX1466r nested PCR was not performed (NP).
The cultivation temperature of all reference species and new isolates was 30°C, except for B. methanolicus NCIMB13113 (45°C) and S. globispora LMG6928T (15°C). IDA, isolate from Drentse A. The percent similarity to the closest cultured relative is given in parentheses.
The soil samples in the present study can be divided in two main groups: (i) fresh, i.e., stored at 4°C for 6 to 12 months, and (ii) air dried, i.e., collected during different years and from different locations. The locations and soil characteristics of the fresh soils from The Netherlands were as follows: (i) Drentse A, where B. benzoevorans relatives were discovered, an agricultural research area (06°41′E, 53°03′N) with high bacterial diversity (10) representing a 1.5-km stretch of grasslands along the Anlooër Diepje river (11), 0- to 25-cm sampling depth; (ii) Hoeksma and Sikkema farms, Friesland, The Netherlands, 0- to 25-cm sampling depth; (iii) Wageningen University experimental field, sandy soil, 0- to 25-cm sampling depth; (iv) the shore of the river Rhine, Wageningen, The Netherlands, wet clay soil, 0- to 10-cm sampling depth; and (v) the root nodule surface of Alnus glutinosa and the surrounding bulk soil, 0- to 10-cm sampling depth (33). Additional fresh samples were obtained from various locations, including a river site, a grassland, a pine tree forest, a fruit tree garden, and a vegetable garden within Provinzende de Cima, Portugal, and from a vegetable garden in Sofia, Bulgaria. Air-dried samples originated from soil under Coriaria nepalensis plants from the Murree area of the Himalayan Mountains in Pakistan (collected in 1993) (20) and from the Russian tundra (collected in 2001). In addition, air-dried soil samples were used that had been collected from the top 0- to 25-cm layer of nonfertilized areas of an agricultural field in the Wieringermeer polder, The Netherlands. These samples were taken from an extensive archive for technical information and soil samples (TAGA; Alterra B.V., Wageningen, The Netherlands). The collection contains the results of field experiments performed between 1879 and 1998 in The Netherlands (5). Land reclamation by drainage of this former sea bottom was started in 1930 and completed in 1940, followed by transformation into agricultural lands (32). To describe the changes in the total bacterial and B. benzoevorans-related community after land reclamation, samples obtained in the years 1942, 1950, 1951, 1966, 1973, and 1975 were analyzed.
Two nonsoil samples were also analyzed: sludge from an anaerobic wastewater treatment system in Eerbeek, The Netherlands (K. Roest, H. G. H. J. Heilig, H. Smidt, W. M. de Vos, A. J. M. Stams, and A. D. L. Akkermans, unpublished data) and piglet feces (16).
Strain isolation procedure.
Two types of mineral media were used for strain isolation. Briefly, they contained the following (per liter of distilled water): sodium acetate or sodium benzoate, 2 g; yeast extract, 2 g; P-medium, 50 ml; and oligo-element solution (25), 200 μl. Agar (BBL; Becton Dickinson, Le Pont de Claix, France) was added at 20 g per liter for solidification where needed. P-medium comprised (per liter of distilled water): Na2HPO4 · 0.2H2O, 37.75 g; KH2PO4, 19.6 g; MgSO4 · H2O, 0.36 g; and NH4Cl, 2 g. The pH was adjusted to 7.0, and media were autoclaved for 20 min at 121°C. Cells were isolated from 0.1 g of soil via vertical shaking for 60 min at room temperature in 950 μl of 1× phosphate-buffered saline buffer (27). Cell suspensions were diluted to 10−5, and 100 μl was plated on two different media. The plates were incubated for 5 to 7 days at 20°C. A total of 96 colonies were randomly chosen from each medium and from each sample for further investigation.
Platewash.
After the colonies were picked, the plates were washed with 1 ml of 1× phosphate-buffered saline buffer.
DNA isolation, primer design, and PCR conditions.
DNA was isolated from pure cultures (1 ml) and directly from soils (1 g) by using the Fast DNASPIN kit (for soil) (Q BIOgene, Cambridge, United Kingdom) according to the manufacturer's instructions. Preliminary experiments showed that using the kit-based extraction gave the best DNA isolation results (data not shown). DNA from B. novalis, B. vireti, B. soli, B. bataviae, and B. drentensis was kindly provided by Jeroen Heyrman.
All primers used in the present study are listed in Table 2. For primer design, 16S rRNA gene sequences of phylogenetically closely related species were retrieved from GenBank at www.ncbi.nlm.nih.gov and aligned by using CLUSTAL W (30). Two sets of primers (REX primers) were developed for specific detection of B. benzoevorans-related strains. The first primer set, REX460f-REX1466r, excludes Paenibacillus spp., B. methanolicus, B. infernus, Streptococcus mirabilis, and B. macroides. The second primer set, REX576f-REX1446r, excludes B. mycoides, B. cereus, B. megaterium, B. simplex, B. macroides, B. subtilis, B. methanolicus, Paenibacillus spp., S. mirabilis, B. infernus, and B. globisporus (Table 3).
TABLE 2.
Sequences of PCR primers used in this study
| Primer | Sequence | Source or reference |
|---|---|---|
| 27f | 5′-GTT TGA TCC TGG CTC AG-3′ | 17 |
| 1492r | 5′-CGG CTA CCT TGT TAC GAC-3′ | 17 |
| REX460f | 5′-GAG TAA CTG C(T/C)(C/G) GTA CC-3′ | This study |
| REX1466r | 5′-CCA ATC ATC TGT CCC ACC TTA-3′ | This study |
| REX576f | 5′-AAG CGC GCG CAG GCG GTC CT-3′ | This study |
| REX1446r | 5′-CCT TAG GCG GCT GGC TCC TTA-3′ | This study |
| GC-0968f | 5′-CGC CCG GGG CGC GCC CCG GGC GGG GCG GGG GCA CGG GGG GAA CGC GAA CCT TAC-3′ | 23 |
| 1401r | 5′-GCG TGT GTA CAA GAC CC-3′ | 23 |
| SP6 | 5′-ATT TAG GTG ACA CTA TAG-3′ | Promega |
| T7 | 5′-AAT ACG ACT CAC TAT AGG-3′ | Promega |
| 519r | 5′-G(A/T)A TTA CCG CGG C(G/T)G CTG-3′ | 17 |
| BACREXf | 5′-GATCCTGGCTCAGGAYSAACGCTGGCGGC-3′ | 9 |
| LGb1513r | 5′-GGTGATCCAGCCGCACCTTCCGA-3′ | 9 |
| ACB1409r | 5′-CARCCSRCTTTCGTGAT-3′ | 9 |
| VER877r | 5′-GCGACCGTACTYCCCAGGCGGC-3′ | 9 |
| Bmeg180r | 5′-CATCTCCCATGAAGGAGAAGA-3′ | 9 |
| PAE835f | 5′-GGGGTTTCGATACCCTTGGTGC-3′ | 9 |
| REX1426f | 5′-CACCCGAAGTCGGTGGGGTAACCGYA-3′ | 9 |
TABLE 3.
Alignment of primer sets REX460f-REX1466r and REX576f-REX1446r with the respective 16S rRNA gene target and nontargeted sequences from different Bacillus spp.
| Speciesa | Sequence alignmentb for:
|
|||
|---|---|---|---|---|
| REX460f | REX1466r | REX576f | REX1446r | |
| 5′-GAGTAACTGCYSGTACC-3′ | 5′-CCAATCATCTGTCCCACCTTA-3′ | 5′-AAGCGCGCGCAGGCGGTCCT-3′ | 5′-CCTTAGGCGGCTGGCTCCTTA-3′ | |
| Bacillus benzoevorans* | GAGTAACTGCTCGTACC | 3′-GGTTAGTAGATAGGGTGGAAT-5′ | AAGCGCGCGCAGGCGGTCCT | 3′->GGAATCCGCCGACCGAGGATT-5′ |
| Bacillus pseudomegaterium* | GAGTAACTGGGCGTACC | -----------AGGGTGGAAT | AAGCGCGCGCAGGCGGTCCT | GGAATCCGCCGACCGAGGAAT |
| Bacillus niacini* | GAGTAACTGCCGGTACC | GGTTAGTAGACAGGGTGGAAT | AAGCGCGCGCAGGCGGTCCT | GGAATCCGCCGACCGAGGAAT |
| Bacillus jeotgali* | GAGTAACTGCCGGTACC | GGTTAGTAGACAGGGTGGAAT | AAGCGCGCGCAGGCGGTCCT | GGAATCCGCCGACCGAGGTTT |
| Uncultured bacterium DA032* | GAGTAACTGCTGGTGCC | GGTTAGTAGACAGGGTGGAAT | AAGCGCGCGCAGGCGGTCCT | GGAATCCGCCGACCGTGGTAT |
| Bacillus mycoides | GTTGAATAAGCTGGCAC | GGTTAGTAGACAGGGTGGAAT | AAGCGCGCGCAGGTGGTTTC | GGAATCCGCCGACCGAGGTAT |
| Bacillus cereus AH-527 | GTTGAATAAGCTGGCAC | GGTTAGTAGACAGGGTGGAAT | AAGCGCGCGCAGGTGGTTTC | GGAATCCGCCGACCGAGGTAT |
| Bacillus megaterium | GAGTAACTGSTTGTACC | GGTTAGTAGACAGGGTGGAAT | AAGCGCGCGCAGGCGGTTTC | GGAATCCGCCGATCGAGGAAT |
| Bacillus macroides | GAGTAACTGCTGGTACC | ----------------GGAAG | AAGCGCGCGCAGGTGGTTCC | GGAAGCCGCCGACCGAGGTAC |
| Bacillus subtilis | TCGAATAGGGCGGTACC | ---------- | AAGGGCTCGCAGGCGGTTTC | ---------- |
| Bacillus methanolicus | TCGAATAGGGCGGTACT | GGTTAGTAGACAGGGTGGAAT | AAGCGCGCGCAGGCGGTTCC | GGAATCCGCCGACCGAGGAAT |
| Paenibacillus spp. | GAGTAACTGCTCCATAG | ---------- | AAGCGCGCGCAGGCGGCTAA | -----------AACGCCCAAT |
Bacillus species belonging to the cluster of B. benzoevorans relatives are indicated with an asterisk.
Mismatching bases are in boldface and underlined. Incomplete sequence information is indicated by a hyphen.
The specificity of the primers was checked in silico by submitting the sequence to the Check Probe program of the Ribosomal Database Project II (4, 13) and to a BLAST search (www.ncbi.nlm.nih.gov/BLAST/). The decision on the specificity of the primers was made based on the number, type, and position of mismatches with nontarget sequences.
Primers were purchased from MWG Biotech AG (Ebersberg, Germany) and were used in a nested approach of two successive specific PCRs. PCR products that were obtained after amplification with primer pair REX460f-REX1466r were used as a template for PCR with primer pair REX576f-REX1446r at a concentration of 1.0 to 2.6 ng of DNA per 50-μl reaction. Subsequently, REX576f-REX1446r amplicons were used as a template for DGGE-PCR with the primers GC-0968f and 1401r (23).
PCR was performed with Taq polymerase (Life Technologies, Gaithersburg, Md.). DNA was amplified in a Whatman Biometra Thermocycler (Göttingen, Germany) under the following conditions: REX460f-REX1466r, 94°C for 1 min, 35 cycles of 94°C for 10 sec, 56°C for 20 s, and 68°C for 1 min, and finally 68°C for 7 min; and REX576f-REX1446r 94°C for 5 min, 25 cycles of 94°C for 1 min, 71°C for 20 s, and 72°C for 1 min, and finally 72°C for 3 min. PCR with the primer set 0968GCf-1401r was performed as previously described (23). Multiplex PCR was used in the present study to differentiate between B. benzoevorans-related and non-B. benzoevorans-related bacterial species. Seven primers (BACREXf, LGb1513r, ACB1409r, VER877r, Bmeg180r, PE835f, and REX1426f) were used to amplify different organisms. The sequences of the primers are listed in Table 2. The multiplex PCR was performed in 96-well microtiter plates according to a previously published procedure (9). PCR products from multiplex positive isolates were sequenced with primer 519r (Table 2), yielding ca. 500 bp of sequence information.
PCR conditions were optimized for specificity using genomic DNA extracted from reference strains, as well as from pure cultures isolates obtained at Wageningen University (Table 1). The primers were further tested with cloned 16S rRNA genes isolated from soil samples that had been previously demonstrated to contain high numbers of B. benzoevorans-related species (6).
Analysis of PCR products by DGGE and sequencing.
The PCR products obtained after three nested PCRs with the primer sets REX460f-REX1466r, REX576f-REX1446r, and GC-0968f-1401r were separated by DGGE (21, 22). 16S rRNA gene-targeted PCR-DGGE fingerprinting was used to analyze the profiles of the B. benzoevorans relatives from different soil samples and also to identify different clones from clone libraries. A gradient of 38 to 50% of the denaturing chemicals (urea and formamide) was used, and gel electrophoresis was performed according to the method of Heilig et al. (13). The gels were stained with AgNO3 according to the method of Sanguinetti et al. (28), scanned at 400DRI, and analyzed by using the software Molecular Analyst 1.12 (Bio-Rad). The similarity between the DGGE profiles was determined by calculating similarity indices of the densitometric curves of the profiles compared by using Pearson product-moment correlation (12, 36). The UPGMA (unweighted pair-group method with arithmetic averages) algorithm was used as implemented in the analysis software for the construction of dendrograms.
PCR amplicons after performing nested PCR with primer sets REX460f-REX1466r and REX576f-REX1446r were purified by using a Qiaquick PCR purification kit according to the manufacturer's instructions and cloned into E. coli JM109 by using the Promega pGEM-T vector system (Promega, Madison, Wis.) as previously described (13). The size of the inserts was confirmed by PCR with specific vector primers T7 and Sp6 (Table 2). Clones were screened by restriction fragment length polymorphism analysis with the restriction enzymes MspI and CfoI. Plasmids containing unique inserts were further purified, and inserts were sequenced. Sequencing reactions were performed with the Sequenase (T7) sequencing kit (Amersham Life Sciences, Slough, United Kingdom) as previously described (13). Sequences were automatically analyzed on a LI-COR DNA sequencer 4000L (Lincoln, Nebr.) and corrected manually.
Nucleotide sequence accession numbers
The GenBank accession numbers of the 16S rRNA gene sequences of isolates obtained here are given in Table 1.
Statistical analysis.
The Student t test was used for statistical analysis of comparison between similarity indices from the DGGE profiles.
RESULTS
Cultivation and identification of B. benzoevorans relatives and strain isolation.
Samples from Drentse A, Hoeksma and Sikkema farms, the Wageningen University experimental field, A. glutinosa-associated rhizosphere nodules and bulk soil, soil from Bulgaria, five different Portuguese soils, Russian tundra soil, and six Wieringermeer soils were used for cultivation and isolation of potential B. benzoevorans relatives. All soils yielded ca. 104 to 105 CFU per g of soil on either of the two mineral media, except for the Wieringermeer samples, from which only ca. 103 CFU g−1 could be retrieved.
Multiplex PCR was used to tentatively identify B. benzoevorans relatives. A total of 192 CFU from each Dutch soil sample were screened, except for the Wieringermeer samples, in which the possible maximum of CFU was picked, and 48 CFU from the Portuguese, Bulgarian, and Russian soils were analyzed. In general, ca. 20% of all screened colonies retrieved from Dutch soil samples (except the Wieringermeer samples) tested multiplex positive, in contrast to the isolates from Portuguese (2.5%), Bulgarian (no positives), and A. glutinosa-associated (no positives) samples. B. benzoevorans relatives could not be detected in the Russian tundra, Bulgarian vegetable garden, Portuguese river site, grassland, and vegetable garden soil samples or in the Wageningen A. glutinosa rhizosphere soil. Single-colony isolates that were identified as potentially B. benzoevorans-related by multiplex PCR were subjected to 16S rRNA gene sequence analysis for further identification.
The highest number of B. benzoevorans relatives was found in the soils from the northern part of The Netherlands. About 9% of all 384 screened isolates obtained from the Drentse A and Friesland samples were identified as B. benzoevorans relatives, whereas the remaining multiplex-positive isolates belonged to other Bacillus spp. The fraction of B. benzoevorans-related isolates that could be retrieved from Wageningen and Portuguese soil samples was significantly lower (3.5% of the screened 96 CFU and ca. 1.7% of the screened 240 CFU, respectively). The length of the sequences was ∼0.5 kb, and a threshold of 97% sequence similarity was used to consider strains as B. benzoevorans related (data not shown).
The isolates used for validation of the new B. benzoevorans cluster-specific primers are shown in Table 1. They were chosen to represent a range of B. benzoevorans-related and -unrelated microorganisms. Since the 16S rRNA gene sequence similarity of several isolates was <97%, some potentially novel B. benzoevorans relatives were cultivated.
Design and validation of new group specific primers for detection of B. benzoevorans relatives.
Primer sets REX460f-REX1466r and REX576f-REX1446r for the specific detection of B. benzoevorans relatives were designed based on a multiple alignment of the complete 16S rRNA gene sequences of B. benzoevorans relatives, as well as non-B. benzoevorans-related species (Table 3).
The optimal conditions for PCR amplification were experimentally determined by using genomic DNA from 29 bacterial strains as the template (Table 1). The optimum for REX460f-REX1466r was annealing at 56°C and 35 cycles, and for REX576f-REX1446f it was annealing at 71°C and 25 cycles. In agreement with the multiple alignment, the primers discriminated against different Bacillus species, which do not belong to the cluster of B. benzoevorans relatives (Table 3). These primer sets, when applied in a nested approach, specifically amplified 16S rRNA gene fragments of Bacillus benzoevorans, B. niacini, B. pseudomegaterium, B. jeotgali, and uncultured bacterium DA001 as expected (Table 1).
The diversity of B. benzoevorans relatives judged by DGGE was compared between DNA isolated directly from the soil and after the plate was washed (data not shown). The DGGE fingerprints showed different profiles.
Specific DGGE patterns of B. benzoevorans relatives from the Dutch soil samples.
The applicability of the REX primers in combination with DGGE for monitoring of the B. benzoevorans-related community was initially tested with DNA isolated from soil samples from Drentse A and Friesian farms in The Netherlands (Fig. 1, lane II), since these grassland fields contain high numbers of B. benzoevorans-related isolates. PCR products generated by the two-step nested specific REX PCR were used as the template for PCR with universal primers, i.e., GC-0968f-1401r, and products were separated by DGGE. The resulting profiles of B. benzoevorans relatives were compared to the total bacterial community profiles of the two Dutch soil samples from Drentse A and Friesland (lane I). The results revealed that the predominant bands of the upper half of the bacterial DGGE patterns corresponded to bands in the DGGE fingerprint of B. benzoevorans relatives.
FIG. 1.
DGGE profiles of the total bacterial and B. benzoevorans-related communities in Drentse A (A) and Friesian (B) soil samples. Identified bands are indicated with numbers and explained in the table below. Lane I, total bacterial fingerprint (universal DGGE primer set GC-0968f-1401r); lane II, B. benzoevorans-related fingerprint (specific primer sets REX460f-REX1466r and REX576f-REX1446r and universal DGGE primer set GC-0968f-1401r).
The DGGE profiles of B. benzoevorans-related microorganisms were compared in five soil samples (Fig. 1 and 2), which demonstrated differences in the diversity of B. benzoevorans relatives in the different soils. In order to confirm the specificity of the newly developed nested REX PCR approach, cloning and sequence analysis were performed for PCR products that were obtained from several soil samples, including Drentse A and Friesland (Fig. 1), the Wageningen University experimental field, the river Rhine, and Himalaya (Fig. 2). These analyses showed that all of the unique clones, selected after restriction fragment length polymorphism analysis, belonged to the cluster B. benzoevorans relatives. The DGGE bands of 20 of 22 studied clones (Fig. 1 and 2, lanes A1 to A5, B1 to B5, C1 to C3, and D1 to D3) from all tested samples corresponded to the predominant bands in the fingerprint of the B. benzoevorans relatives. The majority of sequences retrieved from the clone libraries were most closely related with clones from uncultured bacteria, also related to B. benzoevorans, detected in Drentse A grassland soils in a previous study (10).
FIG. 2.
DGGE analysis of the B. benzoevorans-related communities in four different soil samples. Identified bands are indicated with numbers and explained in the table below. Lane I, Wageningen University experimental field; lane II, river Rhine shore; lane III, Himalaya, bulk soil; lane IV, rhizosphere soil sample from C. nepalensis, Himalaya; A1 to A5, clones from DNA extracted from sample I; B1 to B5, clones from DNA extracted from sample II; C1 to C3, clones from DNA extracted from sample III; D1 to D3, clones from DNA extracted from sample IV.
Comparison of the diversity of B. benzoevorans relatives in different places.
A remarkable richness of B. benzoevorans relatives in different places was observed based on DGGE profiles (Fig. 3). No B. benzoevorans relatives could be detected in the nonsoil samples, i.e., sludge from an anaerobic wastewater treatment reactor and piglet intestine, nor in the Bulgarian and Russian tundra soil samples (data not shown). In contrast, the other 13 locations showed some similar but also different DGGE profiles. To determine whether the B. benzoevorans-related communities from different soil samples were significantly different or similar, similarity indices of the DGGE profiles were calculated. Both Himalayan rhizosphere soil samples showed almost identical B. benzoevorans-related communities (similarity index of 95). Remarkably, they showed significant similarity (similarity indices from 82.5 to 88.8) with the B. benzoevorans cluster-specific DGGE fingerprints of Dutch soil samples. B. benzoevorans-related communities in the soils from Portugal and also in the samples from A. glutinosa nodules and surrounding soil were less diverse than the Dutch and Himalayan soils (Fig. 3). The B. benzoevorans cluster-specific DGGE banding patterns observed for the different Portuguese samples were significantly less similar and were also distinct from those obtained from the Dutch and Himalayan soils, based on the different positions of the bands in the profiles (Fig. 3b). The similarity indices between the Portuguese and the Himalayan rhizosphere samples varied from 22.2 to 69.7. The DGGE profiles of the samples from the river site and vegetable garden in Portugal showed only two dominant bands. In contrast to other Dutch soil samples, low-complexity B. benzoevorans-related communities were found on A. glutinosa nodules and the surrounding soil, where only one predominant species and possibly some species present in lower numbers could be detected. A Student t test revealed that there was a significant difference between the two separately clustering groups shown on the dendrogram (Fig. 3), since the lowest Ptwo-tail observed was 0.199776 × 10−7 (df = 18).
FIG. 3.
DGGE profile of the B. benzoevorans related community in soil samples from different places (a) and corresponding dendrogram (UPGMA clustering) (b). Lanes: M, marker; 1, Wageningen University experimental field (The Netherlands); 2, river Rhine shore (The Netherlands); 3, Drentse A (The Netherlands); 4, Friesland (The Netherlands); 5, A. glutinosa nodules (The Netherlands); 6, A. glutinosa rhizosphere (The Netherlands); 7, river site (Portugal); 8, grassland (Portugal); 9, pine forest (Portugal); 10, fruit garden (Portugal); 11, vegetable garden (Portugal); 12, Himalaya, bulk soil; 13, Himalaya, rhizosphere soil from C. nepalensis.
Heterogeneity of the 16S rRNA genes of B. benzoevorans relatives.
The heterogeneity of the 16S rRNA genes was studied to determine whether each single band from the DGGE profiles of B. benzoevorans-related members belongs to a different strain. The DGGE fingerprints of 12 pure cultures were compared (Fig. 4). Most of the B. benzoevorans-related isolates showed a single dominant band, often with one to two fainter bands, except B. benzoevorans IDA4919, which had four bands. This indicates that at least some of the species have multiple, heterogeneous 16S rRNA gene alleles. In addition, the positions of the predominant bands obtained for some of the isolates were similar. These findings suggest that the diversity of the B. benzoevorans relatives can be either slightly over- or underestimated, depending on the populations present in an environmental sample.
FIG. 4.
DGGE fingerprint and list of 15 pure cultures of B. benzoevorans-related isolates, showing heterogeneity of their rRNA. M, marker. IDA, isolates from Drentse A. Cultured strains that show a 16S rRNA gene sequence identical to that of the uncultured bacterium DA001 are indicated by a superscript “a.”
Development and succession of B. benzoevorans-related communities in Wieringermeer polder soil samples.
Changes in the B. benzoevorans-related communities over a period of 33 years were investigated in soil samples from an experimental field in the Wieringermeer polder (The Netherlands) with samples obtained in 1942, 1950, 1951, 1966, 1973, and 1975. The soil environment changed from anaerobic to aerobic during the reclamation process (32). B. benzoevorans-specific PCR-DGGE was applied to determine whether B. benzoevorans-related communities were present in these soils and if changes occurred over the years (Fig. 5). In the 1942 soil sample, only two major bands were detected, whereas the number of bands had increased by 1950 to at least four strong bands, indicating an increase in the diversity during these years. No major changes in the diversity of B. benzoevorans-related communities in the period between 1950 and 1975 were detected.
FIG. 5.
DGGE profile of the B. benzoevorans-related community in soil samples from different years (Wieringermeer polder) (a) and the corresponding dendrogram (UPGMA clustering) (b). M, marker; 1942, 1950, 1951, 1966, 1973, or 1975 indicate the year of sampling.
DISCUSSION
Both culture-dependent and independent strategies were developed here to monitor B. benzoevorans-related soil bacteria that are surprisingly predominant in some soils around the world. Specific primers for the group were developed and validated that allowed their rapid detection in environmental samples. The validity of the primers was demonstrated by determining the distribution of the group in different geographical locations and also their diversity over time in various soil samples.
A large number of B. benzoevorans isolates could be retrieved from a variety of soil samples by using mineral media supplemented with sodium acetate or sodium benzoate as substrates. It was possible to obtain and keep the bacteria viable on the mineral media developed in the present study, which will allow for further morphological and physiological characterization of isolates. The majority of B. benzoevorans relatives were isolated from the sodium benzoate medium, whereas sodium acetate amended media proved less suitable for the isolation of B. benzoevorans relatives. The species belonging to this cluster probably have better metabolic potential to utilize benzoate than acetate as a substrate (24). The isolation of a substantial collection of cultured representatives of this novel lineage of Bacillus now offers a solid basis to gain more insight into the physiology of these microorganisms. In this context it is noteworthy that recently obtained pure culture isolates (e.g., IDA1527 and IDA3504) (Fig. 4) of the B. benzoevorans group share an identical 16S rRNA gene sequence (similarities of 99.5 and 99.9%, respectively) with clones of the highly abundant uncultured bacterium DA001 (Jeroen Heyrman, unpublished data).
Nevertheless, the cultivation procedure utilized here was rather time-consuming, and the yield of cultured stains was still low. Fewer than 7% of all isolates obtained in the present study were B. benzoevorans relatives. Moreover, it is likely that not all members of this lineage can be cultivated with similar efficiency, since it is widely accepted that most of the soil microorganisms (estimated to be 99.5 to 99.9%) have not yet been isolated and cultivated on laboratory media (1, 2, 26, 31). To compensate for this well-recognized inability to comprehensively address microbial diversity solely by cultivation, we developed a fast and sensitive method for qualitative analysis of the B. benzoevorans relatives in the soil based on a group-specific PCR and DGGE. Since the present study focused on a fast and easy detection of B. benzoevorans-related bacteria in different soil samples, these two primer sets were constructed to consecutively exclude non-target Bacillus species. Because different Bacillus species showed 16S rRNA sequence similarities that were too high, it was not possible to develop only one primer pair that would be sufficiently specific to amplify only target sequences from B. benzoevorans relatives. Therefore, a two-step nested-PCR approach was necessary for molecular detection of the B. benzoevorans relatives. Indeed, the DGGE profiles showed that most of the predominant B. benzoevorans related microorganisms detected by these primers could not be cultivated.
The products of REX576f-REX1446r PCR were subsequently used as a template for DGGE-PCR. We did not use a GC clamp for either of these primers because the products would have a size of ca. 900 bp, which is not optimal for separation by DGGE (22). We also observed that the high annealing temperature, required for specificity (71°C), negatively influenced the amplification efficiency of the GC clamp primer.
The specificity of the REX primers for B. benzoevorans relatives was demonstrated by cloning and sequencing PCR products obtained with DNA isolated directly from different soil samples. All of the sequenced clones belonged to the B. benzoevorans-related cluster. In most cases, the highest sequence similarity was found with the Drentse A clones corresponding to uncultured relatives of B. benzoevorans, which were retrieved from a Dutch Drentse A soil 16S rRNA gene clone library (10, 11).
The REX primers provide a convenient and fast method not only for detection of B. benzoevorans relative populations in the soil samples but also for monitoring of this part of the bacterial community over time and its distribution in different places. We studied the presence of these microorganisms in 15 soil samples and two nonsoil samples (anaerobic sludge and gut content). In neither of the two nonsoil environments were B. benzoevorans relatives detected, which was an expected result since these ecosystems are not among the preferred habitats for characterized Bacillus species belonging to this cluster. Specifically, the anaerobic conditions in the two environments are probably a limiting factor for the aerobic B. benzoevorans relatives, although the genus Bacillus includes some facultative anaerobic members. In the soil samples, B. benzoevorans relatives were present in 19 of the 21 studied soil samples from different countries. Although B. benzoevorans relatives could be found worldwide (8, 10), they were not detectable in all of the samples from the different countries and thus do not comprise the predominant soil bacteria everywhere. In the Russian and Bulgarian samples they are present, if at all, in numbers that are below the detection threshold of the PCR-DGGE method. The highest predominance of the Bacillus species from this cluster was demonstrated in the Dutch and Himalayan soil samples. In all of the Portuguese habitats B. benzoevorans relatives were present but not as the predominant population. When soil samples from similar environments but from different countries were compared (e.g., river sites in The Netherlands and Portugal), it was found that the B. benzoevorans-related community was also different. Significant differences in the B. benzoevorans-related communities were found not only between samples from the different countries but also between soils collected from different habitats in the same country.
In the period from 1942 to 1950, significant differences in the DGGE profiles of the B. benzoevorans-related community in soil samples from the Wieringermeer polder were observed. These changes in the diversity of the B. benzoevorans relatives can be explained by the fact that in 1945 the Wieringermeer polder was flooded during the war (29). It was subsequently reclaimed and, at that time, agricultural activities were also developed that can be another reason for the observed increase in the bacterial diversity of the reclaimed land. From 1950 onward, the richness did not alter, but the intensity of the DGGE bands slightly varied, suggesting changes in relative abundance of the populations within this bacterial community.
The strategy that combines the B. benzoevorans relatives-specific PCR with DGGE is widely applicable for monitoring of the temporal and spatial diversity of B. benzoevorans-related bacilli and their abundance in the bacterial community. Their abundance suggests that these broadly distributed and, in certain places, predominant microorganisms have an important role in the soil ecosystems, one that will be examined in further studies.
Acknowledgments
This study was supported by a grant from the European Communities EC project Exploration of Genomic and Metabolite Diversity of a Novel Group of Abundant Soil Bacteria (BACREX-project QLK3-2000-01678). The Netherlands Organization for International Cooperation in Higher Education is acknowledged for financial support for Y.L.
We thank Phillip Ehlert, TAGA (ALTERRA B.V., Wageningen, The Netherlands), who allowed us access to the soil collection. We thank Jeroen Heyrman (Ghent University, Ghent, Belgium) for kindly providing DNA samples. Carla Belo is especially acknowledged for help with the experimental part of the study.
REFERENCES
- 1.Amann, R. I., W. Ludwig, and K.-H. Wchleiferm. 1995. Phylogenetic identification and in situ detection of individual microbial cells without cultivation. Microbiol. Rev. 59:143-169. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Borneman, J., P. W. Skroch, K. M. O'Sullivan, J. A. Palus, N. G. Rumjanek, J. L. Jansen, J. Nienhuis, and E. W. Triplett. 1996. Molecular microbial diversity of an agricultural soil in Wisconsin. Appl. Environ. Microbiol. 62:1935-1943. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Borneman, J., and E. W. Triplett. 1997. Molecular microbial diversity in soils from Eastern Amazonia: evidence for unusual microorganisms and microbial population shifts associated with deforestation. Appl. Environ. Microbiol. 63:2647-2653. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Cole, J. R., B. Chai, T. L. Marsh, R. J. Farris, Q. Wang, S. A. Kulam, S. Chandra, D. M. McGarrell, T. M. Schmidt, G. M. Garrity, and J. M. Tiedje. 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. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.De Willigen, P., P. F. A. M. Römkens, and P. A. I. Ehlert. 2001. Data mining TAGA for background levels of heavy metals. Report 261. Alterra, Wageningen, The Netherlands. (In Dutch.)
- 6.Felske, A. 1999. Reviewing the DA001-files: a 16S rRNA chase on suspect #X99967, a Bacillus and Dutch underground activist. J. Microbiol. Methods 36:77-93. [DOI] [PubMed] [Google Scholar]
- 7.Felske, A., and A. D. L. Akkermans. 1998. Spatial homogeneity of the most abundant bacterial 16S rRNA molecules in grassland soils. Microbiol. Ecol. 36:31-36. [DOI] [PubMed] [Google Scholar]
- 8.Felske, A., A. D. L. Akkermans, and W. M. de Vos. 1998. In situ detection of an uncultured predominant Bacillus in Dutch grassland soils. Appl. Environ. Microbiol. 64:4588-4590. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Felske, A., J. Heyrman, A. Balcaen, and P. De Vos. 2003. Multiplex PCR screening of soil isolates for novel Bacillus-related lineages. J. Microbiol. Methods 55:447-458. [DOI] [PubMed] [Google Scholar]
- 10.Felske, A., A. Wolterink, R. van Lis, and A. D. L. Akkermans. 1998. Phylogeny of the main bacterial 16S rRNA sequences in Drentse A grassland soils (The Netherlands). Appl. Environ. Microbiol. 64:871-879. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Felske, A., A. Wolterink, R. van Lis, W. M. de Vos, and A. D. L. Akkermans. 1999. Searching for predominant soil bacteria: 16S rDNA cloning versus strain cultivation. FEMS Microbiol. Ecol. 30:137-145. [DOI] [PubMed] [Google Scholar]
- 11a.Felske, A. D. M., V. Tzeneva, J. Heyrman, M. A. Langeveld, A. D. Akkermans, and P. de Vos. 25March2004. Isolation and biodiversity of hitherto undescribed soil bacteria related to Bacillus niacini. Microbiol. Ecol. 10.1007/s00248-003-2025-4. [DOI] [PubMed]
- 12.Häne, B. G., K. Jäger, and H. Drexler. 1993. The Pearson product-moment correlation coefficient is better suited for identification of DNA fingerprint profiles than band matching algorithms. Electrophoresis 14:967-972. [DOI] [PubMed] [Google Scholar]
- 13.Heilig, H. G. H. J., E. G. Zoetendal, E. E. Vaughan, P. Marteau, A. D. L. Akkermans, and W. M. de Vos. 2002. Molecular diversity of Lactobacillus spp. and other lactic acid bacteria in the human intestine as determined by specific amplification of 16S ribosomal DNA. Appl. Environ. Microbiol. 68:114-123. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Heyrman, J., B. Vanparys, N. A. Logan, A. Balcaen, M. Rodriguez-Diaz, A. Felske, and P. De Vos. 2004. Bacillus novalis sp. nov., Bacillus vireti sp. nov., Bacillus soli sp. nov., Bacillus bataviae sp. nov., and Bacillus drentensis sp. nov., five new species isolated from the Drentse A grasslands. Int. J. Syst. Evol. Microbiol. 54:47-57. [DOI] [PubMed] [Google Scholar]
- 15.Janssen, P. H., P. S. Yates, B. E. Grinton, P. M. Taylor, and M. Sait. 2002. Improved culturability of soil bacteria and isolation in pure culture of novel members of the divisions Acidobacteria, Actinobacteria, Proteobacteria, and Verrucomicrobia. Appl. Environ. Microbiol. 68:2391-2396. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Konstantinov, S. R., W.-Y. Zhu, B. A. Williams, S. Tamminga, W. M. de Vos, and A. D. L. Akkermans. 2003. Effect of fermentable carbohydrates on piglet fecal bacterial communities as revealed by denaturing gradient gel electrophoresis analysis of 16S ribosomal DNA. FEMS Microbiol. Ecol. 43:225-235. [DOI] [PubMed] [Google Scholar]
- 17.Lane, D. J. 1991. 16S/23S rRNA sequencing. John Wiley & Sons, Ltd., London, United Kingdom.
- 18.Lapage, S. P., J. E. Shelton, and T. G. Mitchell. 1970. Nutrient agar, p. 116. In J. R. Norris and D. W. Ribbons (ed.), Methods microbiology, vol. 3A. Academic Press, Ltd., London, United Kingdom. [Google Scholar]
- 19.Macrae, A., C. M. M. Lucon, D. L. Rimmer, and A. G. O'Donnell. 2001. Sampling DNA from the rhizosphere of Brassica napus to investigate rhizobacterial community structure. Plant Soil 233:223-230. [Google Scholar]
- 20.Mirza, M. S., W. M. Akkermans, and A. D. L. Akkermans. 1994. PCR-amplified 16S rRNA sequence analysis to confirm nodulation of Datisca cannabina L. by the endophyte of Coriaria nepalensis Wall. Plant Soil 160:147-152. [Google Scholar]
- 21.Muyzer, G., E. C. de Waal, and A. G. Uitterlinden. 1993. Profiling of complex microbial populations by denaturing gradient gel electrophoresis analysis of polymerase chain reaction-amplified genes coding for 16S rRNA. Appl. Environ. Microbiol. 59:695-700. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Muyzer, G., and K. Smalla. 1998. Application of denaturing gradient gel electrophoresis (DGGE) and temperature gradient gel electrophoresis (TGGE) in microbial ecology. Antonie Leeuwenhoek 73:127-141. [DOI] [PubMed] [Google Scholar]
- 23.Nübel, U., B. Engelen, A. Felske, J. Snaidr, A. Wieshuber, R. I. Amann, W. Ludwig, and H. Backhaus. 1996. Sequence hetereogeneities of genes encoding 16S rRNAs in Paenibacillus polymyxa detected by temperature gradient gel electrophoresis of 16S rDNA fragments. J. Bacteriol. 178:5636-5643. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Pichinoty, F., J. Asselineau, and M. Mandel. 1984. Biochemical characterization of Bacillus benzoevorans sp. nov., a new filamentous, sheated mesophilic species which degrades various aromatic acids and phenols. Ann. Microbiol. 135:209-217. [PubMed] [Google Scholar]
- 25.Pichinoty, F., M. Mandel, B. Greenway, and J.-L. Garcia. 1977. Isolation and properties of a denitrifying bacterium related to Pseudomonas lemignei. Int. J. Syst. Bacteriol. 27:346-348. [Google Scholar]
- 26.Rosswall, T., and E. Kvillner. 1978. Principal-components and factor analysis for the description of microbial populations. Adv. Microbiol. Ecol. 2:1-48. [Google Scholar]
- 27.Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual, vol. 3. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.
- 28.Sanguinetti, C. J., D. Neto, and A. J. G. Simpson. 1994. Rapid silver staining and recovery of PCR products separated on polyacrylamide gels. BioTechniques 17:915-919. [PubMed] [Google Scholar]
- 29.Terpstra, P. 1979. 50 Jaar Wieringermeer. Uitgeverij M. A. van Seijen, Leeuwarden, The Netherlands.
- 30.Thompson, J. D., D. G. Higgins, and T. J. Gibson. 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. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Torsvik, V., J. Goksøyr, and F. L. Daae. 1990. High diversity of DNA of soil bacteria. Appl. Environ. Microbiol. 56:782-787. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.van Schreven, D. A., and G. W. Harmsen. 1968. Soil bacteria in relation to the development of polders in the region of the former Zuider Zee. Liverpool University Press, Liverpool, United Kingdom.
- 33.Wolters, D. J., A. D. L. Akkermans, and C. van Dijk. 1997. Ineffective Frankia strains in wet stands of Alnus glutinosa L. Gaertn in The Netherlands. Soil Biol. Biochem. 29:1707-1712. [Google Scholar]
- 34.Yoon, J.-H., S.-S. Kang, K.-C. Lee, Y. H. Kho, S. H. Choi, and K. H. P. Kang, Y.-H. 2001. Bacillus jeotgali sp. nov., isolated from jeotgal, Korean traditional fermented seafood. Int. J. Syst. Evol. Microbiol. 51:1087-1092. [DOI] [PubMed] [Google Scholar]
- 35.Zengler, K., G. Toledo, M. Rappe, J. Elkins, E. J. Mathur, J. M. Short, and M. Keller. 2002. Cultivating the uncultured. Proc. Natl. Acad. Sci. USA 99:15681-15686. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Zoetendal, E. G., A. D. L. Akkermans, W. M. Akkermans-van Vliet, J. A. G. M. de Visser, and W. M. de Vos. 2001. The host genotype affects the bacterial community in the human gastrointestinal tract. Microb. Ecol. Health Dis. 13:129-134. [Google Scholar]