1201. Sarcoidosis Candidate Microbes Identified by Next Generation Sequencing

BACKGROUND: Sarcoidosis is an autoimmune disease characterized by granulomatous lung disease with very prominent mediastinal adenopathy. Acid-fast bacteria, fungi, and viruses have been considered as possible causes of sarcoidosis. We used next-generation or deep sequencing to characterize the microbial content of diseased mediastinal lymph nodes from 10 sarcoidosis patients compared to a set of 10 negative-controls. METHODS: RNA was extracted from fixed paraffinized mediastinal lymph nodes (MLN) from 12 diseased specimens taken from 10 sarcoidosis patients and 2 positive control subjects (TB, MAI), and normal appearing MLN from 10 negative-control subjects (mostly cancer patients). The extracted RNA was sequenced on the Illumina 2500, yielding 125-bp paired-end reads. These reads were aligned to the human genome, human transcriptome, and a nonredundant panmicrobial database. Each experimental sample were compared against the set of 10 negative-controls using the false discovery rate method (q-value). Directed qPCR was performed on all the samples. RESULTS: 100-153 million read-pairs were obtained from the 24 sequenced samples (12 sarcoidosis, 10 negative-control, 2 positive-control). Among these, 0.01-1.32% of the reads were microbial, with a trend towards fewer microbial reads in the sarcoidosis group compared to controls (means 66K vs. 457K, p=0.09). Mycobacterial sequence was significantly enriched (q< 0.05) in the MAI but not the TB sample compared to the negative-controls. Among the 12 sarcoidosis samples, sequence mappings were significantly enriched (q< 0.05) for the following genera: fungal, Magnaporthe (N=4 samples) and Debaromyces (1); bacteria, Odoribacter (1) and Granulicella (1); and viral, Roseolovirus (6) and Mardivirus (6). Further metagenomic analysis eliminated Magnaporthe as a candidate. qPCR confirmed the presence of Odoribacter in 2 specimens and Debaromyces in 1. Roseolovirus (HHV6) could not be detected by qPCR in any of the samples. CONCLUSION: We conclude that sequencing is a feasible method for identifying candidate microbes that might trigger sarcoidosis in human subjects. Further research is required to establish or refute the pathogenicity of these organisms in patients with sarcoidosis. DISCLOSURES: All Authors: No reported disclosures