Phylogenetic Diversity of Archaea and Bacteria in the Anoxic Zone of a Meromictic Lake (Lake Pavin, France)

The compositions of archaeal and bacterial populations at differentdepths (60 m [mixolimnion-chemocline interface], 70 m [chemocline-subchemoclineinterface], 90 m, and 92 m [the water-sediment interface]) inthe anoxic zone of the water column in Lake Pavin, a freshwaterpermanently stratified mountain lake in France, were determined.Phylogenetic trees were constructed from sequences to assessarchaeal and bacterial diversity at the four sites.

INTRODUCTION

Top

Introduction

Permanent anoxic layers in natural freshwater basins are rareand of considerable interest to microbial ecologists becauseof their potential undisturbed climax microbial communitiesand because of their relationship to an earlier biosphere. LakePavin in France provides such an environment. It is unusualbecause the water column has been stratified for a very longperiod and there has been a lack of mixing (meromixis) and itsanoxic zone is in steady state (2). Despite its unique character,information on the distribution of microbial communities inthe anoxic water column of Lake Pavin is limited to a terminalrestriction fragment length polymorphism study of populations(16). Lehours et al. (16) found that the structures of boththe bacterial and archaeal communities changed with depth. Theresults suggested that communities at interfaces played a predominantrole in the water column. To obtain detailed phylogenetic informationon the diverse populations in the differing anaerobic communities,16S rRNA genes in samples collected at three interface layersin the anoxic water column of Lake Pavin were amplified, cloned,sequenced, and analyzed. Bacterial and archaeal clone librarieswere also constructed from a sample collected at a depth of90 m adjacent to the sediment to determine whether sedimentfluxes influenced lake bottom community composition.

Sample collection and library construction.

Top

Sample collection and library...

Samples from depths of 60 m, 70 m, and 90 m in the water columnof Lake Pavin were collected in August 2004 using an 8-literhorizontal Van Dorn bottle; samples were also collected fromthe sediment-water interface (Inter) at a depth of 92 m usinga Jenkin-Mortimer multiple corer (21) (see reference 16 forsite characteristics). Microbial samples were prepared on sitefrom water samples (500 ml) by filtration through polycarbonatemembrane filters (GTTP; Millipore) (47 mm diameter; pore size,0.2 µm) and stored at –80°C. The sites sampledin the water column are shown in Fig. 1.

View larger version (27K):
[in this window]
[in a new window]

FIG. 1. Frequencies of phylotypes affiliated with phylogenetic groups in libraries of samples collected from depths of 60 m (A), 70 m (B), and 90 m (C) and from the water-sediment interface (92 m) (D).

DNA was extracted as previously described (14). 16S rRNA geneswere amplified using the archaean-specific 21f (5) and bacterium-specific27f (8) forward primers and the universal primer 1492r (13).PCR products were cloned using a TOPO TA cloning kit accordingto the manufacturer's instructions (Invitrogen Corporation,San Diego, CA). Cloned inserts were PCR amplified using theM13 forward and reverse primers, and amplicons were digestedwith the restriction endonuclease HaeIII (Qbiogene). The bandingpatterns were grouped according to similarity, and plasmid DNAfrom a single representative of each unique restriction fragmentlength polymorphism pattern was isolated using a QIAprep plasmidpurification kit (QIAGEN, Chatsworth, CA). Clones sequencedin one direction by MWG Biotech (Roissy CDG, France) yieldedreadable sequences of 800 bp on average. Because of possiblesequence errors from PCR (27) or cloning (24), a conservativevalue of 97% sequence similarity was chosen for grouping intooperational taxonomic units (OTUs). Clone libraries were screenedfor chimeric sequences with the Mallard program available athttp://www.cardiff.ac.uk/biosi/research/biosoft/ (4), and 23sequences identified as chimeras were excluded. The remaining203 bacterial and 131 archaeal sequences were used in analyses.

Analyses of diversity and comparisons of libraries.

Top

Analyses of diversity and...

Calculations were performed on a personal computer with thefreeware program aRarefactWin (S. Holland, University of Georgia,Athens). Shannon-Weiner index (H') (11), Sorensen similarityindex (C_s) (20), and S_Chao1 (the nonparametric Chao speciesrichness estimator) (12) values were computed using EstimateSsoftware version 7.5 (K. Colwell; http://purl.oclc.org/estimates).Coverage (C) and Margalef index (D_Mg) values were calculatedas previously described (11, 22).

Rarefaction analyses showed that there was a broad diversityof bacteria in samples but less diversity of archaea (see Fig.S.1.a in the supplemental material). Coverage values for bacteriain all four libraries were found to be low ( $≤$ 57%) (Table 1).This, and results from rarefaction curve analyses (see Fig.S.1.b in the supplemental material), indicated that a greaternumber of clones would have provided more robust informationon bacterial diversity within sites. Our data showed that bacterialdiversity increased with depth (Table 1) in agreement with findingsfrom a previous terminal restriction fragment length polymorphismstudy of Lake Pavin populations (16). Coverage values for archaeawere high ( $≥$ 92%, Table 1) at all four sites, and rarefactioncurves showed that asymptotes were almost reached (see Fig.S.1.b in the supplemental material). The diversity indices forarchaea ranged from 0.8 to 1.4 for H' and 0.8 to 1.6 for D_Mg,indicating low diversities at all sites (Table 1). The resultswith respect to composition of archaea did not differ greatlybetween sites ( $≥$ 67%), whereas Sorensen similarity indices ofthe bacterial phylotypes were low, ranging from 11.1% to 27.3%(Table 2).

View this table:
[in this window]
[in a new window]

TABLE 1. Properties of the distribution of phylotypes in clone libraries from the anoxic zone of Lake Pavin^a

View this table:
[in this window]
[in a new window]

TABLE 2. Similarity matrix for the compositions of bacterial and archaeal sequences in samples collected at different depths in Lake Pavin

Phylogenetic analyses.

Top

Phylogenetic analyses.

By the use of the phylogenetic software package ARB (19), neighbor-joiningtrees were constructed and partial sequences from clone librariesinserted while overall tree topology was retained. The robustnessof inferred topologies was tested by bootstrap analysis usingPHYLIP (PHYLIP [Phylogeny Inference Package] version 3.5c, 1993;J. Felsenstein, Department of Genetics, University of Washington,Seattle, WA) and 1,000 resamplings of trees. The sequences ofthe cloned inserts were deposited in the GenBank database (http://www.ncbi.nlm.nih.gov/GenBank/).For accession numbers, see Fig. 2 and Fig. S.2.A to S.2.E inthe supplemental material.

View larger version (24K):
[in this window]
[in a new window]

FIG. 2. Neighbor-joining tree showing inferred phylogenetic relationships of 16S rRNA archaeal gene sequences cloned from the anoxic zone of Lake Pavin. Scale bars indicate Jukes-Cantor distances. Bootstrap values of >50 (for 1,000 iterations) are shown. GenBank accession numbers are given in parentheses.

Unusual archaeal phylotypes were not detected except for oneOTU at the 70-m depth (Pav-Arc-010) affiliated with the phylumCrenarchaeota (see Table S.1 in the supplemental material).The archaeal communities were dominated by sequences relatedto methanogens belonging to the Methanosarcinales and the Methanomicrobiales(see Fig. 2 and Table S.1 in the supplemental material). Thisfinding is consistent with the results of in situ hybridizationanalyses performed in a previous study of Lake Pavin water samples(16). Lehours et al. (16) also noted that methane concentrationsin the anoxic zone of Lake Pavin correlated with detection ofmethanogens belonging to Methanosarcinales. In the present study,71% of methanogen sequences were related (>95%) to Methanosaetaconcilii (Fig. 2). We postulate that acetoclastic methanogenesisis an important process in the anoxic zone of the water columnin Lake Pavin.

Bacterial sequences were found to align with 16 of the 52 phylogeneticdivisions (see Table S.1 in the supplemental material). Themajority of the 113 OTUs were most closely related to bacteriaor clone-library sequences associated with Proteobacteria, Bacteroidetes,Verrucomicrobia, and candidate division OP11 (Fig. 1). Becauseof the wide diversity, the inferred relatedness of bacterialOTUs was examined in five separate phylogenetic trees (see Fig.S.2.A to S.2.E in the supplemental material). In the presentstudy, many sequences were found to be only distantly relatedto previously cultivated organisms (see Table S.1 in the supplementalmaterial). Nevertheless, some sequences were sufficiently relatedto known bacteria to enable reasonable hypotheses of functionto be formulated. At the 60-m depth, three sequences were related( $≥$ 98.9%) to sequences from known methylotrophs, suggesting thatsome bacteria are involved in methane oxidation in the upperpart of the chemocline. Three sequences (Pav-008) closely related(97.5%) to that of the microaerophilic iron-oxidizing bacteriumGallionella ferruginea (10) were retrieved from the 60-m library.Because high concentrations of ferric and ferrous iron havebeen detected in the monimolimnion of Lake Pavin (16, 25), itis likely that microaerophilic iron-oxidizing bacteria are involvedin recycling ferrous iron in the chemocline in Lake Pavin.

In the present study, bacterial sequences were dominated bysequences from ${delta}$ -proteobacteria; a number of bacteria in thissubdivision are known to reduce Fe (III) (7). DissimilatoryFe (III) reduction appears to be a significant biological processin systems containing high concentrations of ferrous iron (18).Three sequences (Pav-087) were found to be related (94.4% to94.6%) to those of Geothrix fermentans, an Fe (III)-reducingbacterium isolated from a petroleum-contaminated aquifer (17).A feature of our study was the finding that many sequences (atleast 40%) had their best matches with clone sequences obtainedfrom contaminated sites and sediments (1, 3, 6, 9, 15, 23, 26;see Table S.1 in the supplemental material). This included archaealsequences closely related (97%) to strains of Methanosaeta conciliirecovered from hydrocarbon and chlorinated-solvent contaminatedsystems (Fig. 2). This apparent relationship was unexpected,because Lake Pavin is a mountain lake in a protected environment.We suggest that the link is ferric-iron reduction, a processoften involved in detoxification processes (18).

Concluding remarks.

Top

Concluding remarks.

Because of prolonged meromixis, the anoxic zone of Lake Pavinrepresents an unusual microbial ecosystem. The anoxic communitieshave evolved since the formation of Lake Pavin 6,000 years ago,and there has been little influence from the external environmentexcept for the sedimentation fluxes. Only 30% of the sequencesof the bacterial OTUs retrieved in this study were more similarto sequences from isolated bacteria or published clone sequencesthan to those represented by the 93% genus proxy cutoff. Itappears that the microbial communities which inhabit the anoxiczone of Lake Pavin are rich in new types of bacteria. Culture-dependentstudies will be necessary to reveal their ecological roles.

ACKNOWLEDGMENTS

We gratefully acknowledge the French Embassy in New Zealand,the New Zealand Ministry for Research Science and Technology,and the French Ministry for Foreign Affairs for financial supportfor scientist exchanges (G.F., K.J., and A.-C.L.).

We thank J. C. Romagoux and G. Demeure for their skilled technicalassistance.

FOOTNOTES

* Corresponding author. Mailing address: Laboratoire de Biologie des Protistes, UMR CNRS 6023, Université Blaise Pascal, 63177 Aubière Cedex, France. Phone: (33) 473407712. Fax: (33) 473407670. E-mail: A-Catherine.LEHOURS{at}univ-bpclermont.fr.

Published ahead of print on 19 January 2007.

Supplemental material for this article may be found at http://aem.asm.org/.

REFERENCES

Top