Draft Genome Sequences of Violacein-Producing Duganella sp. Isolates from a Waterway in Eastern Pennsylvania

Five Duganella sp. bacterial isolates that synthesize violacein were cultured from a central Pennsylvania waterway.

W ater samples were obtained from streams in eastern Pennsylvania, along the Marcellus Shale formation, where salamanders have been impacted by Batrachochytrium dendrobatidis (1). B. dendrobatidis can cause chytrid infections and contributes to the decline in worldwide amphibian populations. Bacterial strains BJB475, BJB476, BJB480, BJB488, and BJB489 were isolated by plating a single water sample (150 to 200 l) from Crooked Run in North Union Township, Pennsylvania, on Reasoner's 2A (R2A) agar and incubating at 22 to 25°C for 48 h. Five violet-pigmented colonies were subcultured for genomic analysis.
Genomic DNA extraction was completed with the Gentra Puregene yeast/bacteria kit (Qiagen) following the manufacturer's protocol. Library preparation was performed using Illumina's Nextera XT library preparation kit. The multiplexed, paired-end Illumina libraries (150 bp) were run using HiSeq sequencing technology on the Illumina HiSeq 4000 instrument. Data were then demultiplexed by sample, and raw data were sent for analysis (Wright Labs, Huntington, PA). Reads were assembled using a previously published local pipeline (2)(3)(4). Sequences were quality filtered using BBDuk from the BBMap package version 37.50, maintaining a Q-score cutoff of 10 (https://sourceforge .net/projects/bbmap). A draft whole-genome assembly was built using SPAdes version 3.11.0 (5) with k-mer sizes of 21, 33, 55, 77, 99, and 127. Contigs shorter than 500 bp, or consisting of fewer than four reads, were filtered out of the assembly.
Draft whole-genome assemblies of the five strains averaged 40.6 contigs, with a high of 48 (BJB489) and a low of 35 (BJB475) ( Table 1). The average N 50 value for all five assemblies was 576,048 bp ( Table 1). The average genome size is predicted to be 7.207 Mb, with an average GϩC content of 64.358% (Table 1), comparable to the OxaII cluster of Duganella previously described (6). The three genomes of BJB480, BJB488, and BJB489 are nearly identical in length and GϩC content and are likely closely related or clonal isolates.
Assembled contigs were annotated using three methods, a local pipeline running Prokka (7), RASTtk, via the PATRIC pipeline (8,9), and the NCBI Prokaryotic Genome Annotation Pipeline (PGAP) (10). BLAST search results for fragments of 16S rRNA for all five isolates were 99 to 100% identical to those of other Duganella species, specifically HH01 (6), and RAxML analysis further clustered these isolates with those in the earlier study (6). Annotations across the three annotation platforms yielded an average of 6,292 coding DNA sequences (CDS), with a high of 6,401 (BJB489) and a low of 6,234 (BJB475). As expected, the violacein biosynthetic operon (vioABCDE) was present in all annotations for all strains. Additionally, all genomes contained genes involved in swarming and gliding motility, as well as biofilm production, correlating with the growth phenotypes observed on solid agar growth medium.
Future work may reveal if different phylogenetic groupings of violacein-producing strains provide unique phenotypic benefits when colonizing particular environments. Research into native violacein-producing strains may also suggest optimal bioremediation strain candidates for amphibians, should chytrid infections worsen in this watershed.

ACKNOWLEDGMENTS
We thank members of the Lamendella laboratory for collecting water samples. B.A.J. was provided funding for sequencing by the New York State Water Resources Institute/New York State Hudson River Estuary Program. Support for training was provided by GCAT-SEEK. GCAT-SEEK is supported by U.S. National Science Foundation award DBI-1248096: RCN-UBE-GCAT-SEEK: The Genome Consortium for Active Undergraduate Research and Teaching Using Next-Generation Sequencing.
R.L. is the cofounder of Wright Labs.