A computational protocol to characterize elusive Candidate Phyla Radiation bacteria in oral environments using metagenomic data
Several studies have documented the diversity and potential pathogenic associations of organisms in the human oral cavity. Although much progress has been made in understanding the complex bacterial community inhabiting the human oral cavity, our understanding of some microorganisms is less resolved due to a variety of reasons. One such little-understood group is the candidate phyla radiation (CPR), which is a recently identified, but highly abundant group of ultrasmall bacteria with reduced genomes and unusual ribosomes. Here, we present a computational protocol for the detection of CPR organisms from metagenomic data. Our approach relies on a self-constructed dataset comprising published CPR genomic sequences as a filter to identify CPR sequences from metagenomic sequencing data. After assembly and functional prediction, the taxonomic affiliation of CPR contigs can be identified through phylogenetic analysis with publically available 16S rRNA gene and ribosomal proteins, in addition to sequence similarity analyses (e.g., average nucleotide identity calculations and contig mapping). Using this protocol, we reconstructed two draft genomes of organisms within the TM7 superphylum, that had genome sizes of 0.594 Mb and 0.678 Mb. Among the predicted functional genes of the constructed genomes, a high percentage were related to signal transduction, cell motility, and cell envelope biogenesis, which could contribute to cellular morphological changes in response to environmental cues.\n\nImportanceCandidate phyla radiation (CPR) bacterial group is a recently identified, but highly diverse and abundant group of ultrasmall bacteria exhibiting reduced genomes and limited metabolic capacities. A number of studies have reported their potential pathogenic associations in multiple mucosal diseases including periodontitis, halitosis, and inflammatory bowel disease. However, CPR organisms are difficult to cultivate and are difficult to detect with PCR-based methods due to divergent genetic sequences. Thus, our understanding of CPR has lagged behind that of other bacterial component. Here, we used metagenomic approaches to overcome these previous barriers to CPR identification, and established a computational protocol for detection of CPR organisms from metagenomic samples. The protocol describe herein holds great promise for better understanding the potential biological functioning of CPR. Moreover, the pipeline could be applied to other organisms that are difficult to cultivate.