PCID2 dysregulates transcription and viral RNA processing to promote HIV-1 latency

HIV-1 latency results from tightly regulated molecular processes that act at distinct steps of HIV-1 gene expression. To elucidate the molecular players that govern latency, we previously performed a dCas9-chromatin immunoprecipitation coupled with mass spectrometry (Catchet-MS) and identified the interactome of the latent HIV-1 LTR. Here we characterize the Catchet-MS-identified PCI domain-containing 2 (PCID2) protein, a component of the TREX2 complex, to play a dual role in promoting HIV-1 latency by enforcing both transcriptional repression and post-transcriptional blocks to HIV-1 gene expression. PCID2 bound the latent HIV-1 LTR and repressed transcription initiation during latency. Depletion of PCID2 remodelled the chromatin landscape at the HIV-1 promoter and resulted in transcriptional activation and reversal of latency. Immunoprecipitation coupled to Mass Spectrometry identified PCID2-interacting proteins to include members of the spliceosome, including negative viral RNA (vRNA) alternative splicing regulators, and PCID2 depletion resulted in over-splicing of intron-containing vRNA and misregulated expression of vRNA splice variants in cell lines and primary cells obtained from people living with HIV-1. We demonstrate that MCM3AP and DSS1, two other RNA-binding TREX2 complex subunits that comprise the dock of the complex also inhibit transcription initiation and viral RNA alternative splicing during latency and similarly to PCID2 function as prominent latency associated repressors of HIV-1 gene expression. Thus, PCID2 is a novel HIV-1 latency-promoting factor, which in context of the TREX2 sub-complex PCID2-DSS1-MCM3AP blocks transcription and dysregulates vRNA processing. 2 million shControl and shPCID2 J-Lat 11.1 cells were collected for RNA sequencing analysis in duplicate. Total RNA was isolated as described above and cDNA libraries were generated using the 3’ mRNA-seq Library Prep Kit Protocol for Ion Torrent (QuantSeq-LEXOGEN) following manufacturer’s instructions. Library quality was assessed on a Bionalayzer using the DNA High Sensitivity Kit reagent and protocol (Agilent Technologies). Libraries were pooled and template and enriched using an Ion Proton One Touch system with Ion PI Hi-Q OT2 200 Kit (Thermo Fisher Scientific). Sequencing was performed using the Ion PI Hi-Q Sequencing 200 Kit on Ion Proton PI V2 chips (Thermo Fisher Scientific) according to manufacturer’s protocols. Reads were mapped on the UCSC hg19 reference genome and processed for normalization as described before (elife paper). Differentially expressed gene analysis of obtained reads was performed using edgeR package under Galaxy (https://usegalaxy.eu/). Cut-off for false discovery rate was 0.01 and a 1.5 fold change for differentially expressed. Volcano plots were generated using Galaxy (https://usegalaxy.eu/).