An improved protocol for quantification of freshwater actinobacteria by fluorescence in situ hybridization

We tested a previously described protocol for fluorescence in situ hybridization of marine bacterioplankton with horseradish peroxidase-labeled rRNA-targeted oligonucleotide probes and catalyzed reporter deposition (CARD-FISH) in plankton samples from different lakes. The fraction of Bacteria detected by CARD-FISH was significantly lower than after FISH with fluorescently monolabeled probes. In particular, the abundances of aquatic Actinobacteria were significantly underestimated. We thus developed a combined fixation and perme-abilization protocol for CARD-FISH of freshwater samples. Enzymatic pretreatment of fixed cells was opti-mized for the controlled digestion of gram-positive cell walls without causing overall cell loss. Incubations with high concentrations of lysozyme (10 mg ml1) followed by achromopeptidase (60 U ml1) successfully per-meabilized cell walls of Actinobacteria for subsequent CARD-FISH both in enrichment cultures and environ-mental samples. Between 72 and>99 % (mean, 86%) of all Bacteria could be visualized with the improved assay in surface waters of four lakes. For freshwater samples, our method is thus superior to the CARD-FISH protocol for marine Bacteria (mean, 55%) and to FISH with directly fluorochrome labeled probes (mean, 67%). Actinobacterial abundances in the studied systems, as detected by the optimized protocol, ranged from 32 to>55 % (mean, 45%). Our findings confirm that members of this lineage are among the numerically most important Bacteria of freshwater picoplankton. The taxonomic composition of picoplankton communitie