CROI 2018 Abstract eBook

Abstract eBook

Poster Abstracts

380 WITHDRAWN 381 HIV INTEGRATES IN GENES REGULATING CELL CYCLE, DNA DAMAGE, & VIRAL TRANSPORT Sulggi Lee 1 , Kevin G. Haworth 2 , Cassandra Thanh 1 , Lauren E. Schefter 2 , Zachary K. Norgaard 2 , Christopher D. Pilcher 1 , Frederick M. Hecht 1 , Jeffrey N. Martin 1 , Timothy J. Henrich 1 , Steven G. Deeks 1 , Hans-Peter Kiem 1 1 University of California San Francisco, San Francisco, CA, USA, 2 University of Washington, Seattle, WA, USA Background: HIV integration is a key step in the viral replication cycle. Prior studies disagree on whether HIV preferentially integrates in oncogenes leading to clonal proliferation. We performed HIV integration site analysis from HIV+ participants and a humanized mouse model. Methods: HIV+ adults on effective antiretroviral therapy (ART) were sampled from the UCSF SCOPE cohort. We extracted human DNA from PBMCs, enriched for CD4+ T cells, and quantified cell-associated HIV total DNA and unspliced RNA by qPCR. HIV integration sites were identified using nested PCR (Illumina MiSeq) with primers for HIV genome and linker sequences. A novel humanized mouse model challenged with a CCR5-tropic virus allowed in vivo comparisons of HIV integration. Sequences were aligned using UCSC Genome BLAT (hg38 assembly). Gene set enrichment analyses were performed using COSMIC Cancer Gene and Gene Ontology Consortium databases. Results: Twenty-six (96%male) participants were included with median age of 46 years, timing of ART from HIV infection of 2.6 years, and ART suppression of 5 years. Among a total of 31,890 detected integration sites, there was a high degree of similarity in integration site loci (7,504 sites) and frequency between clinical and humanized mouse samples (Figure 1). Only 4.5% of integrations occurred within oncogenes. Integrations occurred in genes involved in cell cycle, DNA damage, membrane and nuclear envelope disassembly, viral transport, and DNA unwinding (all >2.26-fold enrichment, false discovery-adjusted P<1.21x10 2). The most frequently observed gene, NPLOC4, regulates spindle disassembly during mitosis, negatively regulates type I interferon to promote degradation of RIG-I, an important antiviral factor. The frequency of unique integration sites was inversely correlated with log 10 copies of unspliced HIV RNA (Spearman R=-0.52, P=0.0075) but not with total HIV DNA (R= 0.27, P=0.18). Timing of ART was also inversely associated with the frequency of unique sites (N=22, R= 0.44, P=0.039) but not with the %expanded clones. Conclusion: We observed consistent integration site patterns between HIV+ patient and in vivo humanized mouse samples. The majority of insertions occurred in genes that may serve to promote viral latency, supported by the observation that a higher number of unique integration sites correlated with lower levels of residual viral transcription, and the observation that integrations were enriched in genes associated with cell cycle, response to DNA damage, and viral transport.

Poster Abstracts

382 INTEGRATION SITE-INDEPENDENT ENHANCEMENT OF LATENCY REVERSAL BY HIV-1 NEF Xiao Mei T. Kuang , Steven W. Jin, Tristan M. Markle, Mark Brockman Simon Fraser University, Burnaby, BC, Canada Background: Nef is a crucial accessory protein that enhances HIV

pathogenesis, in part through its ability to evade host immunity, but Nef’s contribution viral latency reversal remain elusive. Nef is reported to modulate T cell signaling events, which may alter cellular reactivation in the context of latent infection. To investigate this, we examined the reactivation efficiency of latent HIV+ T cell lines harboring functional or defective Nef. Methods: Latent CEM-A02 (CLat) T cell clones were generated using NL4.3∆Env viruses encoding Nef and GFP marker (SF2GFP, G2AGFP or NL4.3IeG). The nef gene was subsequently disrupted in selected CLat clones using CRISPR/Cas9. Viral reactivation induced by TNFα, panobinostat and/or prostratin was assessed using flow cytometry to measure GFP or intracellular Gag-p24 expression. Results: We generated a panel of CLat-SF2GFP [N=11] and CLat-G2AGFP [N=38] clones and tested latency reversal using a combination of LRAs (TNFα, panobinostat and prostratin). In general, viral reactivation was lower in clones encoding defective G2A Nef (GFP MFI=149 [IQR 123-187]) compared to those encoding WT SF2 Nef (GFP MFI=375 [272-474]; p<0.0001), resulting in a reduced number of Gag-p24+ cells. Similar results were obtained when these clones were treated with individual LRAs. Consistent with previous literature, we showed that HIV reactivation efficiency was dependent in part on the proviral DNA integration site. To overcome potential bias associated with these differences between clones, we used two pairs of sgRNA to disrupt the Nef gene in four different CLat clones encoding either SF2GFP or NL4.3-IeG. Total reactivation and Gag-p24 expression were reduced in all four bulk-Nef KO cell lines. To investigate further, we generated and characterized Nef KO clones that lacked the ability to downregulate CD4 and HLA-A02. These KO clones displayed variable reactivation profiles, but notably, the reactivation intensity (GFP MFI) and % Gag-p24+ cells were lower in Nef KO clones compared to their corresponding parental cell line (see table). Additional sequence characterization revealed Nef inversion and/or large HIV sequence deletion in majority of KO clones, an unexpected outcome using CRISPR/Cas9.

CROI 2018 133