Representative Freshwater Bacterioplankton Isolated from Crater Lake, Oregon
Kathleen A. Page, Stephanie A. Connon, and Stephen J. Giovannoni
Isolation and characterization of ecologically important prokaryotes are essential to understanding the vital roles they play in geochemical cycles. We used HTC methods to isolate typical lake bacterioplankton from Crater Lake. 16S rRNA-based phylogenetic reconstructions and comparison of isolate sequences with clone library sequences were used to determine the relationship of each isolate to phylogenetic groups thought to be abundant in lakes. Low-nutrient media were used to obtain the isolates described in this study. The success of this strategy was evidenced by tests that revealed the inability of most of the isolates to grow on 10% R2A, a traditional bacterioplankton isolation medium (34). The liquid, low-nutrient media employed resulted in cultures rich in isolates belonging primarily to Bacteroidetes and Alpha-, Beta-, and Gammaproteobacteria divisions. We did not isolate representatives of Cyanobacteria, Actinomycetes, Verrucomicrobiales, green nonsulfur bacteria, Planctomycetales, Crenarchaeota, or candidate division OP10, even though these groups are represented in rRNA gene clone libraries from Crater Lake (44) and other freshwater environments (12, 53). Recently, Bruns et al. (3) reported the isolation of previously uncultured freshwater actinobacteria from a eutrophic lake when the signal molecule cyclic AMP was added to liquid enrichment culture media. Thus, in some cases, a simple addition to culture medium formulation can dramatically improve culturability. Hahn et al. used stepwise culture acclimatization techniques to enrich ultramicrobacteria classified as actinobacteria (15). No single isolation medium can isolate all of the prokaryotes from a diverse environment, such as Crater Lake. However, it is possible that the application of HTC methods using a series of rationally designed media will eventually yield isolates representing all dominant bacterioplankton groups. Thirty-six of the 55 isolates had 16S rRNA gene sequences that were highly similar to Crater Lake and/or other freshwater 16S rRNA gene clone library sequences, thus validating the HTC method as a way to obtain isolates that are representative of numerically significant clades.
Three different groups of Crater Lake isolates, HTCC528 to HTCC539, HTCC540, and HTCC541 to HTCC542, were found to be members of the beta I freshwater cluster. puf operon sequences were PCR amplified from HTCC528 but not HTCC540 or HTCC541, suggesting that HTCC528 has the ability to carry out photosynthesis. The beta I cluster is a very diverse and widely distributed bacterial group that includes both photosynthetic (Rhodoferax fermentans and Variovorax paradoxus) and nonphotosynthetic (Hydrogenophaga palleronii, Comamonas testosteroni, and Brachymonas denitrificans) bacteria (12). Sequences that hybridize to beta I bacterial probes have been found abundantly associated with macroscopic organic aggregates (lake snow) from Lake Constance, providing further evidence for the heterotrophic role played by at least some beta I bacteria (38). While beta I bacteria are frequently found to be abundant in lakes, relatively few isolates are available. The addition of HTCC528, HTCC540, and HTCC541 to the current set of beta I isolates provides an important resource for elucidation of the ecological roles played by this widespread group of bacterioplankton.