Abstract Body

Background: Next-generation sequencing (NGS) has the potential to revolutionize strategies for HIV surveillance, genome characterization, and minor variant analysis. Moreover, application of this technology in an unbiased manner offers the promise of simultaneous detection and characterization of viral co-infections. In this study, we examined the utility of high-throughput sequence analysis for characterization of HIV and detection of known and novel viruses. Methodology: Thirty-five plasma specimens from HIV-1-infected Cameroonian blood donors were selected for analysis based on partial genome characterization (gag p24, pol IN and env IDR) using traditional Sanger sequencing. NGS cDNA libraries prepared using nuclease pre-treatment followed by RNA extraction and random-primed reverse transcription were multiplexed (4-5 libraries/run) and run on a MiSeq instrument. Sequence data was analyzed using a customized bioinformatics pipeline for rapid pathogen identification, SURPI, to identify and bin all viral sequences. Phylogenetic analysis was performed with PHYLIP and HIV-1 recombinants were characterized using SimPlot. Results: HIV-1 related sequences ranged from 0.002% to 4.11% of total sequences yielding >90% coverage for 83% of samples. Complete genomes were assembled for 26 strains including 7 CRF11_cpx and 2 CRF13_cpx, 1 CRF18_cpx, 2 CRF22_01A1, and 1 CRF37_cpx, circulating recombinant forms with relatively limited data available as well as 8 unique recombinant forms. Average depth of genome coverage ranged from 121- to 8972-fold. Bioinformatics analysis revealed the presence of numerous other viruses. Notably, GBV-C (HPgV) co-infection was detected in 65% of samples. Eight samples yielded complete GBV-C genomes with read numbers ranging from 0.19% to 4.40%. These GBV-C genomes formed a unique cluster in genotype 1 with Ghanian sequences distinct from S. African reference strains. Additional detected viruses included hepatitis B virus, adeno-associated virus, hepatitis delta virus, human parvovirus B, human rhinovirus C, and viral sequences with distant homology (<70% aa identity) to small circular DNA viruses (circoviruses, cycloviruses, and geminiviruses). Conclusions: These data demonstrate the utility of unbiased NGS for HIV-1 surveillance, rapid viral genome characterization, and for simultaneous screening of known and potentially novel viruses. Moreover, the extent of HIV-1 genome coverage substantially increases confidence in phylogenetic classification of strains. This technology coupled with powerful bioinformatics represents an unparalleled opportunity to characterize the virome and decipher interrelationships between HIV-1 and other viral infections.