Identifying the Source of Unknown Microcystin Genes and Predicting Microcystin Variants by Comparing Genes within Uncultured Cyanobacterial Cells

While multiple phylogenetic markers have been used in the culture-independent study of microcystin-producing cyanobacteria, in only a few instances have multiple markers been studied within individual cells, and in all cases these studies have been conducted with cultured isolates. Here, we isolate and evaluate large DNA fragments (>6 kb) encompassing two genes involved in microcystin biosynthesis (mcyA2 and mcyB1) and use them to identify the source of gene fragments found in water samples. Further investigation of these gene loci from individual cyanobacterial cells allowed for improved analysis of the genetic diversity within microcystin producers as well as a method to predict microcystin variants for individuals. These efforts have also identified the source of the novel mcyA genotype previously termed Microcystis-like that is pervasive in the Laurentian Great Lakes and they predict the microcystin variant(s) that it produces.Microcystin-producing cyanobacteria are common nuisance organisms in harmful algal blooms in freshwaters around the world (4). This genetically diverse group (based on 16S rRNA, mcyA, mcyD, and mcyE gene sequences 6, 10, 15, 16, 22]) ranges in morphology from unicellular and colonial cocci to large filamentous strands. Many species can produce a variety of secondary metabolites that can act as hepatatoxins upon ingestion by animals (e.g., variants of microcystin) (4, 33). Microcystin production reduces the water quality in reservoirs used by human populations and fishery resources, and production of these toxins by this group of cyanobacteria makes them important organisms for continued observation and study (4, 33, 36). Much effort has been expended over the past 15 years to characterize the genomic and structural components of the microcystin (mcy) synthetase operon responsible for the production of microcystins. Several complete DNA sequences of the mcy synthetase operon are currently available in GenBank (3, 11, 29, 31).Although the mechanisms of microcystin production are now better understood, recent analyses of mcyA gene fragments from Lakes Erie and Ontario indicated a microcystin toxin producer of unknown phylogeny (7, 28). This discrepancy suggested a need for improved molecular characterization of naturally occurring microcystin producers, which spurred our research to identify the source of several unusual mcyA fragments from the cyanobacterial community (7, 28). It was apparent from initial sequence data that these mcyA gene fragments, termed Microcystis-like, were highly similar to those from Microcystis spp. (colonial or unicellular cocci). However, they contained a 6-nucleotide insert consistent with mcyA genes from filamentous cyanobacteria (e.g., Anabaena, Nostoc, and Planktothrix) (28). These preliminary findings suggested that these unusual mcyA fragments either came from (i) a novel species or strain, (ii) an ancestral Microcystis, (iii) the highly unlikely hybridization of colonial cocci and filamentous cyanobacteria, or (iv) a chimera of cocci and filamentous PCR products. To identify the source of these mcyA gene fragments from uncultured cyanobacteria, we used culture-independent methods to amplify and isolate long regions of the mcy synthetase operon for the simultaneous analysis of two genes, mcyA and mcyB, in one individual from a population. This approach ensures that both genes are contained on the same DNA molecule, thus allowing for more continuous sequence information to use in comparative phylogenetic analyses than previously described. We also envisioned that this mcy gene combination would provide an improved diagnostic tool for determining the genetic potential of naturally occurring cyanobacteria to produce specific microcystin variants by comparing the phylogenetic marker in mcyA to the predictor of amino acid incorporation (via an adenylation domain) in mcyB1.