Nodeomics: pathogen detection in vertebrate lymph nodes using meta-transcriptomics

Abstract

The ongoing emergence of human infections originating from wildlife highlights the need for better knowledge of the microbial community in wildlife species where traditional diagnostic approaches are limited. Here we evaluate the microbial biota in healthy mule deer (Odocoileus hemionus) by analyses of lymph node meta-transcriptomes. cDNA libraries from five individuals and two pools of samples were prepared from retropharyngeal lymph node RNA enriched for polyadenylated RNA and sequenced using Roche-454 Life Sciences technology. Protein-coding and 16S ribosomal RNA (rRNA) sequences were taxonomically profiled using protein and rRNA specific databases. Representatives of all bacterial phyla were detected in the seven libraries based on protein-coding transcripts indicating that viable microbiota were present in lymph nodes. Residents of skin and rumen, and those ubiquitous in mule deer habitat dominated classifiable bacterial species. Based on detection of both rRNA and protein-coding transcripts, we identified two new proteobacterial species; a Helicobacter closely related to Helicobacter cetorum in the Helicobacter pylori/Helicobacter acinonychis complex and an Acinetobacter related to Acinetobacter schindleri. Among viruses, a novel gamma retrovirus and other members of the Poxviridae and Retroviridae were identified. We additionally evaluated bacterial diversity by amplicon sequencing the hypervariable V6 region of 16S rRNA and demonstrate that overall taxonomic diversity is higher with the meta-transcriptomic approach. These data provide the most complete picture to date of the microbial diversity within a wildlife host. Our research advances the use of meta-transcriptomics to study microbiota in wildlife tissues, which will facilitate detection of novel organisms with pathogenic potential to human and animals.

Figure 1. MEGAN comparison of the taxonomic profiles of (A) cDNA transcript-tags from 454 sequencing five individual lymph node samples and two lymph node sample pools and (B) genomic DNA-tags from four individual lymph node samples.

Depicted are assignments with bit score cutoffs ≥50. Circle sizes are scaled logarithmically. Not assigned: sequencing-tags matching to sequences in the NCBI database that are not assigned to taxa; no hits: sequencing-tags not matching to any sequences in the NCBI database.

Figure 2. MEGAN comparison of taxonomic profiles of MD 257 cDNA transcript-tags analyzed against the protein database (red) and the ribosomal database (blue), and of V6 amplicon 16S rRNA-tags analyzed against the ribosomal database (green).

Bit score cutoff for the protein database comparison was set at 50, and confidence cutoffs for the ribosomal database comparisons were set at 80%.

Figure 3. Maximum likelihood trees showing the phylogenetic affiliation of sequences obtained from 454 sequencing with GenBank homologous sequences.

(A) 16S rRNA Helicobacter, (B) rpo-β Helicobacter, (C) 16S rRNA Acinetobacter, (D) rpo-β Acinetobacter. Bootstrap support for each node is indicated.

Figure 4. Maximum likelihood tree inferred from the partial nucleotide sequence data of env gene showing the phylogenetic placement of mule deer (MD) retrovirus.

The two MD PCR sequences reported in the present study are in bold. GenBank accession numbers of reference viruses are mentioned. Bootstrap support for each node is indicated.