Abstract
A water sample from a noncontaminated site at the source of the Woluwe River (Belgium) was analyzed by culture-dependent and -independent methods. Pseudomonas isolates were identified by sequencing and analysis of the rpoD gene. Culture-independent methods consisted of cloning and pyrosequencing of a Pseudomonas rpoD amplicon from total DNA extracted from the same sample and amplified with selective rpoD gene primers. Among a total of 14,540 reads, 6,228 corresponded to Pseudomonas rpoD gene sequences by a BLAST analysis in the NCBI database. The selection criteria for the reads were sequences longer than 400 bp, an average Q40 value greater than 25, and >85% identity with a Pseudomonas species. Of the 6,228 Pseudomonas rpoD sequences, 5,345 sequences met the established criteria for selection. Sequences were clustered by phylogenetic analysis and by use of the QIIME software package. Representative sequences of each cluster were assigned by BLAST analysis to a known Pseudomonas species when the identity with the type strain was greater than or equal to 96%. Twenty-six species distributed among 12 phylogenetic groups or subgroups within the genus were detected by pyrosequencing. Pseudomonas stutzeri, P. moraviensis, and P. simiae were the only cultured species not detected by pyrosequencing. The predominant phylogenetic group within the Pseudomonas genus was the P. fluorescens group, as determined by culture-dependent and -independent analyses. In all analyses, a high number of putative novel phylospecies was found: 10 were identified in the cultured strains and 246 were detected by pyrosequencing, indicating that the diversity of Pseudomonas species has not been fully described.
INTRODUCTION
Pseudomonas bacteria are ubiquitous in soil and water ecosystems, occupying numerous and diverse ecological niches. The taxonomy of the genus is complex, comprising at least 139 recognized species at present (List of Prokaryotic Names with Standing in Nomenclature [www.bacterio.net]) (1). These species are metabolically versatile, are extremely heterogeneous from a nutritional perspective, and participate in the carbon and nitrogen cycles (2). Some of these species are well known for their beneficial roles in plants, and others are used for bioremediation and as biocontrol agents, while still others are plant or animal pathogens (3). On the basis of the data derived from cultivation-dependent analysis, Pseudomonas appears to occupy a prominent position in nature (4).
Previous work in our laboratory with the species Pseudomonas stutzeri (5–9) permitted us to generate the tools needed to extend our study to other Pseudomonas species. This work allowed us to select appropriate genes for sequencing (6, 9), to improve PCR protocols, to design appropriate primers (3), and to create the PseudoMLSA database, which is now available, in order to compile all of these gene sequences for the characterization and taxonomic identification of Pseudomonas strains (10). The rpoD gene encodes the sigma 70 factor of RNA polymerase. It exhibits a high number of polymorphic sites (70.4%) distributed evenly throughout the sequence, and therefore, it is a good choice for phylogenetic and taxonomic analyses of Pseudomonas species (11).
Pseudomonas species are generally easy to isolate in standard growth media, and several types of selective media have been tested for the isolation and quantification of these bacteria in environmental samples. However, it is still not known how well these bacterial isolates represent the total environmental population of Pseudomonas spp. (12), and therefore, culture-independent methods are being developed. A highly selective pair of rpoD primers, PsEG30F–PsEG790R (760-bp amplicon), has been designed for the Pseudomonas genus and has been used previously to generate clone libraries from environmental DNA obtained from sand samples of the intertidal shore (3). These primers have advantages over other Pseudomonas-selective primers that have been designed for similar purposes for 16S rRNA genes (12, 13), because the rpoD amplicon is longer and the sequence much better at distinguishing between Pseudomonas species. The limitations of 16S rRNA gene sequence analysis and the advantages of new approaches based on other housekeeping genes have been discussed by Pontes and collaborators (14). These molecular methods used in microbial ecology studies are often limited to the analysis of a relatively small number of clones and thus identify only a small fraction of the microbial diversity that has been unraveled by previous studies (15). In the past few years, pyrosequencing, a cost-effective, high-throughput DNA sequencing method, has been successfully applied in the determination of bacterial diversity within complex environmental ecosystems, such as deep ocean water, hydrothermal vents, and soil (16–18).
The Woluwe River, located in Brussels, Belgium, is a typical lowland river with an altitude difference (i.e., slope from source to mouth) of only 30 m. It is 1 to 3 m wide and 5 to 50 cm deep. The sources of the river are groundwater fed and are situated in the protected areas of a forest, the Sonian Forest (forêt de Soignes). The central part of the river is highly fragmented by diverse park and pond systems (flowthrough and overflow) and by vaulted stretches. In these waters, the Pseudomonas aeruginosa population has been studied using culture-dependent methodologies (19), and Pseudomonas species diversity has been studied using DNA-based molecular methodologies, with the primers for amplification of the oprI and oprL genes, for the identification of the cultured strains (20).
To assess the diversity of Pseudomonas species in a nonpolluted sample obtained from the Woluwe River, we used two approaches. In the culture-dependent method, the water sample was cultured in Pseudomonas-selective medium, and the isolates were identified by sequence analysis of the rpoD gene. In the culture-independent analysis, total DNA from the same sample was purified, the DNA was amplified with selective rpoD gene primers, and the amplicons were either cloned into Escherichia coli or directly sequenced by pyrosequencing. The results of the two methods were compared, and the enormous diversity of phylotypes was analyzed.
MATERIALS AND METHODS
Water sampling.
The water sample was obtained from the Woluwe River at sampling site W5 (19) in Brussels. W5 is located in the forest near the source. Surface water samples were collected (by skimming the surface) in sterile 1-liter bottles in September 2011 and were immediately filtered, first through a 0.5-μm filter (type SVLP) and then through a 0.2-μm filter (type GTTP) (both from Millipore, USA).
Total cell count by DAPI.
For cell-counting purposes, a water sample (1 ml) was filtered through white polycarbonate filters (type GTTP; pore size, 0.2 μm; diameter, 22 mm; Millipore, USA). Total cells were counted after being stained with 4′,6-diamidino-2-phenylindole (DAPI) and visualized using epifluorescence, as described previously (21).
Culture, isolation, and identification of Pseudomonas spp.
The presence of Pseudomonas was determined by plate counts on a selective medium, CFC medium. CFC medium consisted of Pseudomonas agar F (Difco) containing cephaloridine (50 mg liter−1), fucidin (10 mg liter−1), and cetrimide (10 mg liter−1) (CFC supplement; Sigma). Overnight incubation was performed at 30°C.
A total of 162 randomly selected colonies were identified by sequencing of the rpoD gene. The template DNA for the PCR was obtained with the Wizard genomic DNA purification kit (Promega, Madison, WI, USA) from an overnight culture of a fresh colony inoculated in 4 ml Luria broth. Vials of DNA were subsequently stored at −20°C. The rpoD gene was amplified by PCR as described previously (volumes, nucleotides, and primer concentrations) (3) in an Eppendorf thermocycler. The degenerate selective primers used were PsEG30F (ATYGAAATCGCCAARCG) and PsEG790R (CGGTTGATKTCCTTGA) (3). Amplicons were purified with MultiScreen PCR 96-well plates (Millipore) and were then sequenced directly using the ABI Prism BigDye Terminator cycle sequencing kit (version 3.1; Applied Biosystems). Amplicons were sequenced on a 3130 Genetic Analyzer (Applied Biosystems). Strains were identified by analysis of the sequences using the NCBI database and an in-house rpoD database. Colonies grown on CFC medium whose DNA could not be amplified by the selective rpoD primers were identified by sequencing and analysis of their 16S rRNA genes (6).
DNA extraction from the water sample and PCR conditions.
Total DNA was extracted from 5 liters of freshwater by using the MasterPure Complete DNA and RNA purification kit (Epicentre Biotechnologies, Madison, WI, USA). DNA was purified using the Wizard DNA Clean-Up system (Promega, Madison, WI, USA). Purified environmental DNA was used to amplify the rpoD gene with the PsEG30F–PsEG790R primer pair (3). PCR amplification was performed with a thermocycler (Eppendorf). The PCR was carried out in a final volume of 50 μl using the KAPA2G Robust HotStart PCR kit (Kapa Biosystems) with 5 U of Taq DNA polymerase. The individual reaction mixtures contained 10 μl of PCR KAPA2G buffer B, 0.4 μl of a deoxynucleotide mixture at 100 μM (Fermentas), 0.5 μl of each primer at a concentration of 100 μM, 10 μl of 5× KAPA Enhancer 1, 0.4 μl of Taq DNA polymerase, and 1 μl of template DNA. The PCR program used consisted of an initial denaturation step of 4 min at 95°C; 35 cycles of denaturation at 95°C for 30 s, annealing at 52°C for 20 s, and extension at 72°C for 45 s; and incubation for 10 min at 72°C. The PCR fragments obtained were purified using the Qiagen PCR purification kit.
Clone library.
The purified rpoD PCR products were used to generate the clone library by using the CloneJET PCR cloning kit (Fermentas) according to the manufacturer's instructions. The PCR products were ligated into the pJET1.2/blunt cloning vector and were transformed into Escherichia coli XL1-Blue competent cells prepared in the laboratory. Transformants were cultured on LB agar plates supplemented with 30 μg ampicillin per ml and incubated at 37°C for approximately 16 to 24 h. Plasmid DNA of the transformed cells was obtained by boiling the cells for 5 min in 100 ml TE buffer (pH 8) (10 mM Tris-HCl, 1 mM EDTA), followed by centrifugation at 16,100 × g for 5 min, after which time the supernatant was collected. The inserts of each clone were amplified with the same primers used for the initial amplification. Five microliters of the amplified PCR products was analyzed by electrophoresis on a 1.5% (wt/vol) agarose gel and was stained with ethidium bromide. Amplicons were purified and sequenced as described above using primer PsEG30F.
Sequence analysis of amplified genes.
The partial rpoD gene sequences obtained from the cultured isolates and clones were compared with sequences from an in-house rpoD gene database. The partial 16S rRNA gene sequences obtained from the cultured isolates were compared by BLAST analysis with sequences in the NCBI gene database. All rpoD gene sequences were aligned, and phylogenetic analysis was performed as described previously (11). A similarity of ≥96% to a known Pseudomonas type strain allowed a sequence to be assigned to that species, as proposed previously (22, 23). A similarity of <96% to any known Pseudomonas species type strain indicated that the sequence corresponded to a possible new species in the Pseudomonas genus.
rpoD gene pyrosequencing.
The rpoD amplicon (760 nucleotides [nt]) of the total DNA purified from the water sample was also analyzed by pyrosequencing with standard 454/Roche GS-FLX Titanium equipment. Reads were obtained in both orientations (forward and reverse sequences). The sequences obtained from the pyrosequencing protocol were compared by BLAST analysis with the sequences from the NCBI gene database. Sequences were filtered according to their closest matches. In the first step, only the rpoD gene sequences were selected (first filtration). In the second step, those with a length of >400 bp, an identity to the closest related species of >85%, and an average Q40 value of >25 were selected (second filtration). A previous test in silico with known sequences demonstrated that sequences longer than 400 nt were correctly assigned to their species. Next, phylogenetic reconstructions of rpoD gene sequences similar to those of known Pseudomonas species, as first matches in the NCBI database, were performed as described previously (3). Sequences with ≥96% similarities that were in the same phylogenetic branch were grouped. For each group, a deeper analysis was performed that compared the rpoD gene sequences to those from an in-house database by local BLAST. At least one representative sequence of each group was selected for further analysis, by following the criteria of maximum length, maximum coverage, and a high average Q40 value. When possible, two forward and two reverse sequences from the same group were selected as representative sequences. Similarities greater than or equal to 96% to a known Pseudomonas type strain allowed the assignment of the representative sequences (and the corresponding group) to the known Pseudomonas species. Values lower than 96% indicated that the representative sequence (and its corresponding group) belonged to a possible new species in the genus Pseudomonas. The coverage of Pseudomonas rpoD sequences located in a branch distant from the main group of Pseudomonas rpoD sequences was checked in order to discard chimeras. Sequences with a coverage lower than 60% were discarded (third filtration). The remaining rpoD gene sequences were grouped according to their genera and were compared to sequences in the NCBI database.
Sequences were also clustered by the QIIME software package for the comparison and analysis of microbial communities (24), with an operational taxonomic unit (OTU) clustering similarity threshold of 96%; the UCLUST method was used for picking OTUs. The most abundant sequence in each OTU was selected as representative and was analyzed by local BLAST with an in-house database.
Statistical analysis.
Each different sequence was considered a single phylotype, and several phylotypes were grouped into a single phylospecies when they affiliated in the same phylogenetic branch and the similarity with the rpoD gene was greater than 96% (11). Sequence information obtained from the pyrosequences and isolates were used to calculate coverage, diversity indices, and phylotype and phylospecies richness with the PAST software package (25). The rarefaction curves were also obtained with the PAST software package.
Nucleotide sequence accession numbers.
The sequences generated in this study have been deposited in GenBank/EMBL/DDBL under accession numbers HG933830 to HG933969. The DNA sequences from this metagenomic project have been deposited in the NCBI Sequence Read Archive under accession number SRP036143.
RESULTS
Total cell count.
Counting of cells using epifluorescence microscopy after DAPI staining of the filtered sample indicated a population of 1.02 × 109 cells/ml. A minimum of 500 cells per filter were counted, with an average of 29.34 cells/field.
Culture-dependent analysis.
In total, 136 type strains of Pseudomonas species were tested for growth on CFC medium. Of these, 125 strains were able to grow on this selective medium at their optimal growth temperatures after 24 to 72 h of incubation (see Table S1 in the supplemental material). In all, 188.4 CFU/ml was counted on CFC medium in water from the Woluwe River, representing a small fraction of the total cell count (1 in 5.3 × 106 cells detected by DAPI). In total, 162 randomly selected colonies were identified by sequencing of the rpoD gene (126 colonies), or by sequencing of the 16S rRNA genes when the Pseudomonas-selective primers did not generate a product (36 colonies). Of the 126 rpoD gene sequences, 114 corresponded to Pseudomonas rpoD gene sequences and 12 to species in the Aeromonas genus. The 36 colonies identified by sequencing of the 16S rRNA genes were assigned to the following genera: Aeromonas (7 colonies), Bacillus (4 colonies), Brevundimonas (1 colony), Hafnia (2 colonies), Morganella (3 colonies), Staphylococcus (1 colony), and Yersinia (18 colonies). These results indicated that 70% of the colonies belonged to the genus Pseudomonas.
The Pseudomonas isolates with similarities of ≥96% to a type strain (the species threshold considered) were assigned to the corresponding species (24 isolates distributed among 7 Pseudomonas species). The other Pseudomonas isolates were assigned to the phylogenetic group or subgroup that included the type strain closest to the new isolate (90 isolates in 10 groups or subgroups). Isolates that showed an rpoD gene similarity to the closest type strain below the 96% species threshold were considered to be putative novel species (Table 1; see also Table S2 in the supplemental material).
TABLE 1.
rpoD gene sequence similarities of isolates
| Species of closest Pseudomonas type strain | Similarity (%) | No. of isolates | Species or group assignment | Probable new species |
|---|---|---|---|---|
| P. protegens | 94.6–95.9 | 51 | P. chlororaphis SG | I |
| P. brenneri | 94.5–94.6 | 14 | P. gessardii SG | II |
| P. saponiphila | 93.9–93.8 | 11 | P. chlororaphis SG | III |
| P. rhodesiae | 98.0–98.4 | 8 | P. rhodesiae | |
| P. proteolytica | 99.2–99.2 | 8 | P. proteolytica | |
| P. frederiksbergensis | 90.5–91.5 | 6 | P. mandelii SG | IV |
| P. lurida | 99.4–99.8 | 4 | P. lurida | |
| P. japonica | 80.7 | 2 | P. putida G | V |
| P. protegens | 88.6–88.8 | 2 | P. chlororaphis SG | VI |
| P. stutzeri (gv1) | 99.2 | 1 | P. stutzeri | |
| P. rhodesiae | 91.8 | 1 | P. fluorescens SG | VII |
| P. simiae | 99.8 | 1 | P. simiae | |
| P. lurida | 93.9 | 1 | P. fluorescens SG | VIII |
| P. grimontii/P. marginalis | 95.6 | 1 | P. fluorescens SG | IX |
| P. grimontii | 99.8 | 1 | P. grimontii | |
| P. moraviensis | 97.3 | 1 | P. moraviensis | |
| P. moraviensis | 94.2 | 1 | P. koreensis SG | X |
Clone library analysis.
Cloning of the PCR product resulted in a library of 115 clones. Sixty-four clones with the expected insert size were sequenced, while 51 clones did not generate an amplicon of the expected size or lacked an insert. Of the 64 clones analyzed, 26 showed high similarities to Pseudomonas species (Table 2). Most of the clones belonged to the P. syringae group: 18 clones were similar to P. cichorii (88.6 to 89.0% similarity), and 2 clones were identified as P. viridiflava (97.1% similarity). The other 6 clones belonged to the P. fluorescens group. Only 2 of these clones were identified at the species level, and these were classified as P. grimontii, with 99.3 to 99.6% similarity. The phylogenetic distances of the remaining clones were high enough to represent different Pseudomonas species: 1 clone was similar to P. frederiksbergensis (92.2% similarity), 1 was close to P. grimontii (87.1% similarity), 1 was close to P. grimontii and P. marginalis (93.4% similarity), and 1 was close to P. lurida (95.5% similarity) (Table 2; see also Table S2 in the supplemental material). The remaining 38 sequences were analyzed using the NCBI database and did not correspond to Pseudomonas species and were discarded.
TABLE 2.
Assignment of rpoD gene clones to species
| Species of closest Pseudomonas type strain | Similarity (%) | No. of clones | Species or group assignment | Probable new species |
|---|---|---|---|---|
| P. cichorii | 88.6–89.0 | 18 | P. syringae G | I |
| P. grimontii | 99.3–99.6 | 2 | P. grimontii | |
| P. viridiflava | 97.1 | 2 | P. viridiflava | |
| P. grimontii/P. marginalis | 93.4 | 1 | P. fluorescens SG | II |
| P. grimontii | 87.1 | 1 | P. fluorescens SG | III |
| P. frederiksbergensis | 92.2 | 1 | P. mandelii SG | IV |
| P. lurida | 95.5 | 1 | P. fluorescens SG | V |
Pyrosequencing of the rpoD gene amplicon.
A total of 14,540 sequences were obtained by pyrosequencing of the rpoD amplicon from the DNA extracted from the water sample (see Data set S1 in the supplemental material). Of these, 12,691 sequences (87.3%) were longer than 400 bp; 6,935 sequences (47.7%) were 400 to 599 bp, and 5,750 sequences (39.6%) were 600 to 799 bp (Table 3).
TABLE 3.
Sequences obtained by pyrosequencing of the rpoD gene amplicon
| Length (bp) | No. (%) of: |
|
|---|---|---|
| All sequences | Pseudomonas rpoD sequences | |
| <200 | 619 (4.3) | 169 (2.7) |
| 200–399 | 1,230 (8.5) | 457 (7.3) |
| 400–599 | 6,935 (47.7) | 2,463 (39.5) |
| 600–799 | 5,750 (39.5) | 3,138 (50.3) |
| ≥800 | 6 (0.04) | 1 (0.02) |
| Total | 14,540 | 6,228 |
When all of the sequences were compared with sequences in the NCBI database, 7,031 sequences (48.3%) resulted in an rpoD gene as the closest match; of these, 6,228 (88.6%) corresponded to Pseudomonas rpoD gene sequences. The phylogenetic assignment of the 7,031 sequences to genera and their abundance levels are shown in Table 4 (see also Fig. S1 in the supplemental material).
TABLE 4.
Phylogenetic assignment of the rpoD gene sequences to genera and numbers of sequences after each filtration step
| Genus | Family | Class | No. of sequences after the following filtration step: |
||
|---|---|---|---|---|---|
| 1st | 2nd | 3rd | |||
| Pseudomonas | Pseudomonadaceae | Gammaproteobacteria | 6,228 | 5,345 | 5,165 |
| Acidovorax | Comamonadaceae | Betaproteobacteria | 351 | 176 | 170 |
| Azoarcus | Rhodocyclaceae | Betaproteobacteria | 143 | 102 | 0 |
| Uncultured | 123 | 82 | 8 | ||
| Aeromonas | Aeromonadaceae | Gammaproteobacteria | 83 | 62 | 56 |
| Xanthomonas | Xanthomonadaceae | Gammaproteobacteria | 24 | 12 | 0 |
| Stenotrophomonas | Xanthomonadaceae | Gammaproteobacteria | 15 | 11 | 0 |
| Alcanivorax | Alcanivoracaceae | Gammaproteobacteria | 13 | 11 | 0 |
| Bordetella | Alcaligenaceae | Betaproteobacteria | 9 | 5 | 0 |
| Pseudoxanthomonas | Xanthomonadaceae | Gammaproteobacteria | 8 | 5 | 1 |
| Halomonas | Halomonadaceae | Gammaproteobacteria | 6 | 3 | 0 |
| Burkholderia | Burkholderiaceae | Betaproteobacteria | 5 | 2 | 0 |
| Chromohalobacter | Halomonadaceae | Gammaproteobacteria | 4 | 1 | 0 |
| Curvibacter | Comamonadaceae | Betaproteobacteria | 3 | 0 | 0 |
| Alteromonas | Alteromonadaceae | Gammaproteobacteria | 2 | 0 | 0 |
| Cupriavidus | Burkholderiaceae | Betaproteobacteria | 2 | 1 | 0 |
| Dechloromonas | Rhodocyclaceae | Betaproteobacteria | 2 | 2 | 0 |
| Rhodococcus | Nocardiaceae | Actinobacteria | 2 | 2 | 0 |
| Sphingomonas | Sphingomonadaceae | Alphaproteobacteria | 2 | 2 | 0 |
| Variovorax | Comamonadaceae | Betaproteobacteria | 2 | 0 | 0 |
| Achromobacter | Alcaligenaceae | Betaproteobacteria | 1 | 0 | 0 |
| Diaphorobacter | Comamonadaceae | Betaproteobacteria | 1 | 1 | 0 |
| Psychrobacter | Moraxellaceae | Gammaproteobacteria | 1 | 0 | 0 |
| Rhodobacter | Rhodobacteraceae | Alphaproteobacteria | 1 | 1 | 0 |
Of the 6,228 Pseudomonas rpoD gene sequences, 5,345 sequences were longer than 400 bp and had an average Q40 value greater than 25 and a percentage of identity to a Pseudomonas species greater than 85% (the minimal identity between two type strains in the same Pseudomonas group or subgroup). These 5,345 sequences were analyzed further. Of these sequences, 2,619 were derived from the forward sequence (5′ to 3′) and 2,726 from the reverse sequence (3′ to 5′). Of the 5,345 sequences selected, only 5,165 were analyzed after the third filtration. Of the 5,165 sequences analyzed, only 716 (10.8%) were assigned to 1 of 26 known Pseudomonas species (Table 5). The remaining 4,449 sequences were assigned to 247 phylospecies in a group or subgroup in the genus Pseudomonas (see Table S2 and Data set S2 in the supplemental material). Singletons were detected in high proportions in these phylospecies (189 sequences [76.52%]).
TABLE 5.
Assignment of the rpoD gene sequences to known species by phylogenetic analysis and by the QIIME software package
| Pseudomonas species | Phylogenetic analysis |
QIIME analysis |
||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Representative sequence | No. of reads | F/Ra | Length (bp) | % Identity | Representative sequence | No. of reads | F/R | Length (bp) | % Identity | |
| P. abietaniphila | A8KH0 | 2 | F | 694 | 98.5 | A8KH0 | 2 | F | 694 | 98.5 |
| P. alcaligenes | CM0GW | 7 | F | 648 | 99.4 | CM0GW | 7 | F | 648 | 99.4 |
| P. asplenii | EJ3H4 | 2 | F | 697 | 99.3 | EJ3H4 | 2 | F | 697 | 99.3 |
| P. asturiensis | B3IWI | 4 | R | 699 | 99.3 | B3IWI | 4 | R | 699 | 99.3 |
| P. baetica | EWMUM | 1 | R | 548 | 97.5 | EWMUM | 1 | R | 548 | 97.5 |
| P. cichorii | AHSPY | 48 | F | 749 | 98.2 | CD2FF | 45 | R | 700 | 96.1 |
| P. flavescens | CFLR2 | 21 | F | 716 | 99.8 | BDM65 | 23 | F | 671 | 98.8 |
| P. frederiksbergensis | AJ492 | 1 | F | 561 | 97.4 | AJ492 | 1 | F | 561 | 97.4 |
| P. grimontii | BYB7J | 435 | F | 745 | 99.6 | BDYW0 | 436 | F | 729 | 99.9 |
| P. jessenii | B8Y2U | 7 | R | 699 | 98.8 | EPPXD | 9 | F | 702 | 97.5 |
| P. lini | DRM6I | 3 | F | 501 | 97.3 | DRM6I | 3 | F | 501 | 97.3 |
| P. lurida | C1YFC | 4 | F | 635 | 98.4 | C1YFC | 3 | F | 635 | 98.4 |
| P. mandelii | EXEKH | 1 | R | 457 | 97.5 | EXEKH | 2 | R | 457 | 97.5 |
| P. meridiana | BS01Y | 1 | R | 593 | 96.3 | BYA6T | 2 | R | 551 | 96.7 |
| P. mohnii | EB9W5 | 1 | R | 519 | 97.4 | EB9W5 | 1 | R | 519 | 97.4 |
| P. orientalis | B7DUG | 1 | F | 731 | 97.3 | B7DUG | 1 | F | 731 | 97.8 |
| P. peli | C2QJX | 6 | F | 748 | 98.4 | DO7FD | 6 | F | 728 | 98.7 |
| P. proteolytica | A7TBV | 1 | F | 536 | 99.4 | A7TBV | 1 | F | 536 | 99.4 |
| P. punonensis | AGAMS | 2 | F | 731 | 96.6 | AGAMS | 2 | F | 731 | 96.6 |
| P. putida | AHLG8 | 6 | F | 739 | 96.2 | AHLG8 | 6 | F | 739 | 96.4 |
| P. rhodesiae | B6BYO | 9 | F | 746 | 96.3 | B6BYO | 9 | F | 746 | 96.3 |
| P. salomonii | D2729 | 1 | F | 634 | 97.9 | D2729 | 1 | F | 634 | 97.9 |
| P. syringae | D9JMS | 3 | F | 699 | 97.6 | D9JMS | 3 | F | 699 | 97.7 |
| P. trivialis | BS6B3 | 1 | R | 698 | 96.6 | BS6B3 | 1 | R | 698 | 96.6 |
| P. veronii | DTPTN | 5 | F | 703 | 98.5 | D2HIX | 5 | F | 581 | 96.6 |
| P. viridiflava | AP8C3 | 143 | F | 697 | 98.8 | BK4UX | 147 | R | 521 | 97.6 |
F, forward sequence; R, reverse sequence.
The 5,165 sequences selected were also grouped and analyzed with the QIIME software package. Two hundred fifty-three clusters were obtained at the 96% threshold (Table 6). As indicated in Table 5, the assignment of the representative sequences of each OTU to known species was highly concordant with the previous phylogenetic assignment. The representative sequence of each cluster selected by QIIME was the same as that selected in the previous phylogenetic analysis for 19 of the 26 species detected.
TABLE 6.
Statistical indices
| Statistic | Culture results | Pyrosequencing results |
|||
|---|---|---|---|---|---|
| Phylogenetic analysis |
QIIME analysis |
||||
| All phylospecies | FR phylospeciesa | All phylospecies | FR phylospeciesa | ||
| Total no. of sequences | 114 | 5,165 | 4,688 | 5,165 | 4,379 |
| No. of phylospecies | 17 | 272 | 32 | 253 | 23 |
| No. of singletons | 8 | 189 | 0 | 144 | 0 |
| Simpson index | 0.239 | 0.532 | 0.633 | 0.253 | 0.351 |
| Simpson diversity index | 0.760 | 0.468 | 0.367 | 0.747 | 0.649 |
| Coverage index | 0.529 | 0.305 | 1 | 0.431 | 1 |
| Chao1 index | 26.33 | 907 | 32 | 524 | 23 |
| Shannon diversity index (H) | 1.96 | 1.585 | 0.968 | 2.218 | 1.316 |
FR phylospecies are those identified by forward and reverse sequences.
In screening data sets for variants, one key criterion for determining whether a variant determination is valid is whether it is supported by both forward and reverse reads. Therefore, this more stringent criterion for assignment of the sequences to a known or putative phylospecies was considered. Forward and reverse sequences not assigned to a species type strain, with similarities lower than 96%, but located in the same phylogenetic branch close to a type strain, were assigned to the same putative phylospecies. Twelve known Pseudomonas species and 20 putative novel species were detected by forward sequences and the corresponding reverse sequences (a total of 4,688 sequences) and are termed “FR phylospecies” (Table 6). Thus, 32 FR phylospecies (known or putative novel species) were distributed in 12 of the 19 phylogenetic groups or subgroups in the genus Pseudomonas.
Diversity indices.
The number of clones in the library was low, and for statistical analysis, only the results of culture and pyrosequencing were considered (Table 6). The coverage index for the cultured isolates was estimated at 0.53 (53%). It reached 0.30 (30%) in the analysis of all phylospecies, and the value was 1 (100%) when the conservative criterion of high confidence in species assignment was applied (FR phylospecies). The Chao1 index was used as an estimate of the total number of phylotypes or species present (Table 6). The total number of phylospecies detected by pyrosequencing reached 907 in the analysis of all phylospecies but only 32 with the conservative criterion.
Rarefaction curves for the rpoD gene sequences of the amplicons were constructed to determine the extent of the diversity of phylotypes and phylospecies detected (Fig. 1). In total, 5,165 phylotypes were detected and were distributed among all 272 phylospecies. The rarefaction curve for all phylospecies did not reach saturation. However, the rarefaction curve for the 32 phylospecies detected with the conservative criterion (FR phylospecies), which were represented by 4,688 sequences, reached saturation, indicating that the total diversity in the sample was detected.
FIG 1.
Rarefaction curves of the pyrosequencing results. The rarefaction curve of FR phylospecies (phylospecies to which reads were assigned by both forward and reverse representative sequences) is indicated by a solid line. The dashed line indicates the rarefaction curve of all phylospecies.
DISCUSSION
Several types of selective media have been used for the isolation of Pseudomonas by other authors: Gould's S1 medium, King's B medium (26), and a nutrient-poor medium supplemented with naphthyl acetic acid (nutrient-poor NAA) (12). However, CFC medium is considered to be highly selective and is the selective medium most widely used for Pseudomonas (20). It has also been used previously in a Pseudomonas analysis of the Woluwe River. We tested 136 Pseudomonas species type strains for growth on CFC plates, and only 11 type strains were unable to grow (7%). In the sample analyzed, we also detected colonies of other Gram-negative bacteria, as has been shown by other authors (20, 27), and, to a lesser extent, Gram-positive bacteria. The most frequent non-Pseudomonas species detected were Aeromonas spp. In a previous study, only 17% of the cultured Pseudomonas isolates could be assigned to a known species by using the primers for the amplification of the oprI and oprL genes for identification (20). Similarly, in our study, 22% of the isolates were identified at the species level. In water from sampling point W2, close to sampling point W5 of our study, 9 Pseudomonas species were cultured (20), 4 of which were not detected in our study (P. chlororaphis, P. fluorescens, P. fragi, and P. marginalis). The Pseudomonas species found in the present study are comparable to the microorganisms found in a previous study (20), which belong predominately to the P. fluorescens group of microorganisms, indicators of noncontaminated water. More than 50% of the putative novel species detected in the present study also belong to the P. fluorescens group. This group of Pseudomonas species was also found and described at sampling point W2 more than 4 years ago (20), indicating the presence of a stable population over a distance of 3.5 km (from sampling point W5 to sampling point W2), but not 8 km away (at sampling point W15), where a gradient of pollution has been reported (28) and P. aeruginosa detected (19).
The high selectivity of the Pseudomonas rpoD primers published previously (3) was confirmed in the present study. Only a few cultured non-Pseudomonas strains, of the genus Aeromonas, were amplified with this primer set. Analysis of the sequences of the rpoD amplicon also confirmed the selectivity of the primers: only 4.3% of the rpoD gene sequences obtained were not assigned to the Pseudomonas genus after the third filtration.
In addition to sequencing errors and chimeras, PCR-based methods introduce biases that can affect the results of microbial community structure analyses (29). Such PCR biases can be particularly significant, because the current procedure for amplicon sequencing involves at least two different amplification steps: PCR amplification during the initial sample preparation using template-specific primers, followed by emulsion PCR (emPCR) on Roche's 454 instrument or bridge PCR on the Illumina platform prior to sequencing (30). Thus, the presence of species, but not their abundance, could be determined in the present study, because specific bias corrections, as suggested by Sikaroodi and Gillevet (31), were not introduced in the procedure.
A comparison of the three methods used here indicated that all the sequences of the Pseudomonas isolates and all the sequences obtained in the clone library that were assigned to species or phylogenetic groups or subgroups were also detected by the pyrosequencing method, except for P. stutzeri, P. simiae, and P. moraviensis, which were detected only in the cultured strains. Species of the P. fluorescens group, not the P. aeruginosa group, predominate in unpolluted waters, in accordance with the location of sampling point W5, near the source, upstream of sewage water entrances. These data are concordant with the data obtained by Matthijs and collaborators (20) for sampling point W2, which is also located near the source.
The data obtained from the pyrosequences indicated the presence of 21 known species of Pseudomonas not found by the culture or cloning method. All the type strains of these species that were tested grew on CFC medium. These data emphasize the high sensitivity of the pyrosequencing method.
The number of known species detected in the sample by the three different methods is remarkable: 29 species were distributed among 12 phylogenetic groups or subgroups within the genus. Three species detected in the isolates were not detected by pyrosequencing. Only members of three Pseudomonas phylogenetic groups were not detected: the P. oryzihabitans group, the P. pertucinogena group, and the P. oleovorans group. Of the 29 species, 26 were detected by pyrosequencing; the exceptions were P. stutzeri, P. moraviensis, and P. simiae. P. grimontii was the only species detected by all three methods. Pyrosequencing of the rpoD amplicon rendered a huge number of phylotypes. Only 9.5% (26 species) could be assigned to a known species with the proposed species threshold of 96% similarity. Additionally, 246 possible phylospecies not assigned to a known species were detected, indicating that the real number of existing Pseudomonas species could be at least 2 times the number of Pseudomonas species currently known. Interestingly, two sequences obtained in this study and considered putative novel species showed identities greater than 96% with a recently described Pseudomonas strain of the P. syringae group (strain 42B) isolated from the phyllosphere and not yet classified at the species level (32). It can be expected that the putative novel phylospecies detected in our study will be isolated and described in the future.
The 29 known species detected have previously been isolated mainly from aquatic habitats, but also from soil habitats and in association with plants. It is difficult to obtain ecological conclusions based on the presence of these species in the aquatic habitat studied, but comparison with a previous study of oil-contaminated coastal marine sand samples (22) demonstrates that the Pseudomonas community differs strongly with the habitat. Most isolates and rpoD gene clones detected in that previous study (22) were related to the P. putida, P. anguilliseptica, P. oleovorans, or P. stutzeri phylogenetic group, and only 2 phylospecies (<2%) were related to the P. fluorescens or P. syringae group, the predominant groups in the present study of a noncontaminated freshwater habitat. These results demonstrate the high potential of rpoD gene pyrosequencing for detecting the enormous diversity of Pseudomonas species and the need to isolate putative novel species for their physiological and biochemical characterization.
Supplementary Material
ACKNOWLEDGMENTS
Financial support was obtained from the Spanish MINECO through projects CGL2011-24318 and Consolider CSD2009-00006, as well as from funds for competitive research groups from the Government of the Balearic Islands (the last two funds with FEDER cofunding). D. Sánchez was the recipient of a predoctoral fellowship from the Conselleria d'Interior, Direcció General de Recerca, Desenvolupament Tecnològic i Innovació del Govern de les Illes Balears (FPI09), and the European Social Fund (ESF). M. Gomila is the recipient of a postdoctoral contract from the University of the Balearic Islands, funded by the Spanish Ministry of Education, Culture and Sports through the International Excellence Campus Program.
Footnotes
Published ahead of print 23 May 2014
Supplemental material for this article may be found at http://dx.doi.org/10.1128/AEM.00412-14.
REFERENCES
- 1.Parte AC. 2014. LPSN—List of Prokaryotic Names with Standing in Nomenclature. Nucleic Acids Res. 42:D613–D616. 10.1093/nar/gkt1111 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Silby MW, Winstanley C, Godfrey SAC, Levy SB, Jackson RW. 2011. Pseudomonas genomes: diverse and adaptable. FEMS Microbiol. Rev. 35:652–680. 10.1111/j.1574-6976.2011.00269.x [DOI] [PubMed] [Google Scholar]
- 3.Mulet M, Bennasar A, Lalucat J, García-Valdés E. 2009. An rpoD-based PCR procedure for the identification of Pseudomonas species and for their detection in environmental samples. Mol. Cell. Probes 23:140–147. 10.1016/j.mcp.2009.02.001 [DOI] [PubMed] [Google Scholar]
- 4.Moore ERB, Tindall BJ, Martins Dos Santos VAP, Pieper DH, Ramos J-L, Palleroni NJ. 2006. Nonmedical: Pseudomonas, p 646–703 In Dworkin M, Falkow S, Rosenberg E, Schleifer K-H, Stackebrandt E. (ed), The Prokaryotes: a handbook on the biology of bacteria, 3rd ed, vol 6 Springer, New York, NY [Google Scholar]
- 5.García-Valdés E, Castillo MM, Bennasar A, Guasp C, Cladera AM, Bosch R, Engesser KH, Lalucat J. 2003. Polyphasic characterization of Pseudomonas stutzeri CLN100 which simultaneously degrades chloro- and methylaromatics: a new genomovar within the species. Syst. Appl. Microbiol. 26:390–403. 10.1078/072320203322497428 [DOI] [PubMed] [Google Scholar]
- 6.Cladera A, Bennasar A, Barceló M, Lalucat J, García-Valdés E. 2004. Comparative genetic diversity of Pseudomonas stutzeri genomovars, clonal structure, and phylogeny of the species. J. Bacteriol. 186:5239–5248. 10.1128/JB.186.16.5239-5248.2004 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Cladera AM, García-Valdés E, Lalucat J. 2006. Genotype versus phenotype in the circumscription of bacterial species: the case of Pseudomonas stutzeri and Pseudomonas chloritidismutans. Arch. Microbiol. 184:353–361. 10.1007/s00203-005-0052-x [DOI] [PubMed] [Google Scholar]
- 8.Cladera AM, Sepúlveda-Torres LC, Valens-Vadell M, Meyer J, Lalucat J, García-Valdés E. 2006. A detailed phenotypic and genotypic description of Pseudomonas strain OX1. Syst. Appl. Microbiol. 29:422–430. 10.1016/j.syapm.2005.11.007 [DOI] [PubMed] [Google Scholar]
- 9.Mulet M, Gomila M, Gruffaz C, Meyer J, Palleroni NJ, Lalucat J, García-Valdés E. 2008. Phylogenetic analysis and siderotyping as useful tools in the taxonomy of Pseudomonas stutzeri: description of a novel genomovar. Int. J. Syst. Evol. Microbiol. 58:2309–2315. 10.1099/ijs.0.65797-0 [DOI] [PubMed] [Google Scholar]
- 10.Bennasar A, Mulet M, Lalucat J, García-Valdés E. 2010. PseudoMLSA: a database for multigenic sequence analysis of Pseudomonas species. BMC Microbiol. 10:118. 10.1186/1471-2180-10-118 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Mulet M, Lalucat J, García-Valdés E. 2010. DNA sequence-based analysis of the Pseudomonas species. Environ. Microbiol. 12:1513–1530. 10.1111/j.1462-2920.2010.02181.x [DOI] [PubMed] [Google Scholar]
- 12.Li L, Abu Al-Soud W, Bergmark L, Riber L, Hansen LH, Magid J, Sørensen SJ. 2013. Investigating the diversity of Pseudomonas spp. in soil using culture dependent and independent techniques. Curr. Microbiol. 67:423–430. 10.1007/s00284-013-0382-x [DOI] [PubMed] [Google Scholar]
- 13.Widmer F, Seidler RJ, Gillevet PM, Watrud LS, Di Giovanni GD. 1998. A highly selective PCR protocol for detecting 16S rRNA genes of the genus Pseudomonas (sensu stricto) in environmental samples. Appl. Environ. Microbiol. 64:2545–2553 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Pontes DS, Lima-Bittencourt CI, Chartone-Souza E, Amaral Nascimento AM. 2007. Molecular approaches: advantages and artifacts in assessing bacterial diversity. J. Ind. Microbiol. Biotechnol. 34:463–473. 10.1007/s10295-007-0219-3 [DOI] [PubMed] [Google Scholar]
- 15.Lenchi N, İnceoğlu Ö, Kebbouche-Gana S, Gana ML, Llirós M, Servais P, García-Armisen T. 2013. Diversity of microbial communities in production and injection waters of Algerian oilfields revealed by 16S rRNA gene amplicon 454 pyrosequencing. PLoS One 8:e66588. 10.1371/journal.pone.0066588 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Sogin ML, Morrison HG, Huber JA, Welch DM, Huse SM, Neal PR, Arrieta JM, Herndl GJ. 2006. Microbial diversity in the deep sea and the underexplored “rare biosphere.”. Proc. Natl. Acad. Sci. U. S. A. 103:12115–12120. 10.1073/pnas.0605127103 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Roesch LF, Fulthorpe RR, Riva A, Casella G, Hadwin AKM, Kent AD, Daroub SH, Camargo FA, Farmerie WG, Triplett EW. 2007. Pyrosequencing enumerates and contrasts soil microbial diversity. ISME J. 1:283–290. 10.1038/ismej.2007.53 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Poulsen PHB, Al-Soud WA, Bergmark L, Magid J, Hansen LH, Sørensen SJ. 2013. Effects of fertilization with urban and agricultural organic wastes in a field trial—prokaryotic diversity investigated by pyrosequencing. Soil Biol. Biochem. 57:784–793. 10.1016/j.soilbio.2011.12.023 [DOI] [Google Scholar]
- 19.Pirnay JP, Matthijs S, Colak H, Chablain P, Bilocq F, Van Eldere J, De Vos D, Zizi M, Triest L, Cornelis P. 2005. Global Pseudomonas aeruginosa biodiversity as reflected in a Belgian river. Environ. Microbiol. 7:969–980. 10.1111/j.1462-2920.2005.00776.x [DOI] [PubMed] [Google Scholar]
- 20.Matthijs S, Coorevits A, Gebrekidan TT, Tricot C, Vander Wauven C, Pirnay JP, De Vos P, Cornelis P. 2013. Evaluation of oprI and oprL genes as molecular markers for the genus Pseudomonas and their use in studying the biodiversity of a small Belgian river. Res. Microbiol. 164:254–261. 10.1016/j.resmic.2012.12.001 [DOI] [PubMed] [Google Scholar]
- 21.Glöckner FO, Amann R, Alfreider A, Pernthaler J, Psenner R, Trebesius K, Schleifer KH. 1996. An in situ hybridization protocol for detection and identification of planktonic bacteria. Syst. Appl. Microbiol. 19:403–406. 10.1016/S0723-2020(96)80069-5 [DOI] [Google Scholar]
- 22.Mulet M, David Z, Nogales B, Bosch R, Lalucat J, García-Valdés E. 2011. Pseudomonas diversity in crude-oil-contaminated intertidal sand samples obtained after the Prestige oil spill. Appl. Environ. Microbiol. 77:1076–1085. 10.1128/AEM.01741-10 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Scotta C, Gomila M, Mulet M, Lalucat J, García-Valdés E. 2013. Whole-cell MALDI-TOF mass spectrometry and multilocus sequence analysis in the discrimination of Pseudomonas stutzeri populations: three novel genomovars. Microb. Ecol. 66:522–532. 10.1007/s00248-013-0246-8 [DOI] [PubMed] [Google Scholar]
- 24.Caporaso JG, Kuczynski J, Stombaugh J, Ittinger K, Ushman FD, Costello EK, Fierer N, Peña AG, Goodrich JK, Gordon JI, Huttley GA, Kelley ST, Knights D, Koenig JE, Ley RE, Lozupone CA, McDonald D, Muegge BD, Pirrung M, Reeder J, Sevinsky JR, Turnbaugh PJ, Walters WA, Widmann J, Yatsunenko T, Zaneveld J, Knight R. 2010. QIIME allows analysis of high-throughput community sequencing data. Nat. Methods 7:335–336. 10.1038/nmeth.f.303 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Hammer Ø, Harper DAT, Ryan PD. 2001. PAST: paleontological statistics software package for education and data analysis. Palaeontol. Electron. 4(1):art. 4 http://palaeo-electronica.org/2001_1/past/issue1_01.htm [Google Scholar]
- 26.Johnsen K, Nielsen P. 1999. Diversity of Pseudomonas strains isolated with King's B and Gould's S1 agar determined by repetitive extragenic palindromic-polymerase chain reaction, 16S rDNA sequencing and Fourier transform infrared spectroscopy characterisation. FEMS Microbiol. Lett. 173:155–162. 10.1111/j.1574-6968.1999.tb13497.x [DOI] [PubMed] [Google Scholar]
- 27.Bridson EY. 2006. The Oxoid manual, 9th ed, p 2-287–2-289 Oxoid Ltd, Basingstoke, Hampshire, England [Google Scholar]
- 28.Triest L, Kaur P, Heylen S, De Pauw N. 2001. Comparative monitoring of diatoms, macroinvertebrates and macrophytes in the Woluwe River (Brussels, Belgium). Aquat. Ecol. 35:183–194. 10.1023/A:1011468615246 [DOI] [Google Scholar]
- 29.Acinas SG, Sarma-Rupavtarm R, Klepac-Ceraj V, Polz MF. 2005. PCR-induced sequence artifacts and bias: insights from comparison of two 16S rRNA clone libraries constructed from the same sample. Appl. Environ. Microbiol. 71:8966–8969. 10.1128/AEM.71.12.8966-8969.2005 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Pinto AJ, Raskin L. 2012. PCR biases distort bacterial and archaeal community structure in pyrosequencing datasets. PLoS One 7:e43093. 10.1371/journal.pone.0043093 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Sikaroodi M, Gillevet PM. 2012. Quality control in multi-tag pyrosequencing of microbial communities. Biotechniques 53:381–383. 10.2144/000113967 [DOI] [PubMed] [Google Scholar]
- 32.Humphrey PT, Nguyen TT, Villalobos MM, Whiteman NK. 2014. Diversity and abundance of phyllosphere bacteria are linked to insect herbivory. Mol. Ecol. 23:1497–1515. 10.1111/mec.12657 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.