Cohesins are important for chromosome structure and chromosome segregation during mitosis and meiosis. Cohesins are composed of two structural maintenance of chromosomes (SMC1-SMC3) proteins that form a V-shaped heterodimer structure, which is bridged by a α-kleisin protein and a stromal antigen (STAG) protein. Previous studies in mouse have shown that there is one SMC1 protein (SMC1β), two α-kleisins (RAD21L and REC8) and one STAG protein (STAG3) that are meiosis-specific. During meiosis, homologous chromosomes must recombine with one another in the context of a tripartite structure known as the synaptonemal complex (SC). From interaction studies, it has been shown that there are at least four meiosis-specific forms of cohesin, which together with the mitotic cohesin complex, are lateral components of the SC. STAG3 is the only meiosis-specific subunit that is represented within all four meiosis-specific cohesin complexes. In Stag3 mutant germ cells, the protein level of other meiosis-specific cohesin subunits (SMC1β, RAD21L and REC8) is reduced, and their localization to chromosome axes is disrupted. In contrast, the mitotic cohesin complex remains intact and localizes robustly to the meiotic chromosome axes. The instability of meiosis-specific cohesins observed in Stag3 mutants results in aberrant DNA repair processes, and disruption of synapsis between homologous chromosomes. Furthermore, mutation of Stag3 results in perturbation of pericentromeric heterochromatin clustering, and disruption of centromere cohesion between sister chromatids during meiotic prophase. These defects result in early prophase I arrest and apoptosis in both male and female germ cells. The meiotic defects observed in Stag3 mutants are more severe when compared to single mutants for Smc1β, Rec8 and Rad21l, however they are not as severe as the Rec8, Rad21l double mutants. Taken together, our study demonstrates that STAG3 is required for the stability of all meiosis-specific cohesin complexes. Furthermore, our data suggests that STAG3 is required for structural changes of chromosomes that mediate chromosome pairing and synapsis, DNA repair and progression of meiosis.

Three Prime Repair Exonuclease 1 (TREX1) is the most abundant 3′→5′ exonuclease in mammalian cells. It has been suggested that TREX1 degrades HIV-1 DNA to enable the virus evade the innate immune system. However, the exact role of TREX1 during early steps of HIV-1 infection is not clearly understood. In this study, we report that HIV-1 infection is associated with upregulation, perinuclear accumulation, and nuclear localization of TREX1. However, TREX1 overexpression did not affect reverse transcription or nuclear entry of the virus. Surprisingly, HIV-1 DNA integration was increased in TREX1-overexpressing cells, suggesting a role of the exonuclease in post-nuclear entry step of infection. Accordingly, preintegration complexes (PICs) extracted from TREX1-overexpressing cells retained higher levels of DNA integration activity. TREX1 depletion resulted in reduced levels of proviral integration and PICs formed in TREX1-depleted cells retained lower DNA integration activity. Addition of purified TREX1 to PICs also enhanced DNA integration activity suggesting that TREX1 promotes HIV-1 integration by stimulating PIC activity. To understand the mechanism, we measured TREX1 exonuclease activity on substrates containing viral DNA ends. These studies revealed that TREX1 preferentially degrades the unprocessed viral DNA but the integration competent 3’-processed viral DNA remains resistant to degradation. Finally, we observed that TREX1 addition stimulates the activity of HIV-1 intasomes assembled with the unprocessed viral DNA but not of intasomes containing the 3’-processed viral DNA. These biochemical analyses provide a mechanism by which TREX1 directly promotes HIV-1 integration. Collectively, our study demonstrates that HIV-1 infection upregulates TREX1 to facilitate viral DNA integration. Importance Productive HIV-1 infection is dependent on a number of cellular factors. Therefore, a clear understanding of how the virus exploits the cellular machinery will identify new targets for inhibiting HIV-1 infection. The three prime repair exonuclease 1 (TREX1) is the most active cellular exonuclease in mammalian cells. It has been reported that TREX1 prevents accumulation of HIV-1 DNA and enables the virus to evade the host innate immune response. Here, we show that HIV-1 infection results in the upregulation, perinuclear accumulation, and nuclear localization of TREX1. We also provide evidence that TREX1 promotes HIV-1 integration by preferentially degrading viral DNAs that are incompatible to chromosomal insertion. These observations identify a novel role of TREX1 in a post-nuclear entry step of HIV-1 infection.

HIV-1 infection is dependent on integration of the viral DNA into the host chromosomes. The preintegration complex (PIC) containing the viral DNA, the virally encoded integrase (IN) enzyme, and other viral/host factors carries out HIV-1 integration.

HIV-1 DNA integration into the host chromosomes is carried out by the preintegration complex (PIC). The PIC contains the viral DNA, virally encoded integrase enzyme and other critical viral/host factors. The PIC-associated viral DNA is preferentially integrated into gene bodies of actively transcribing genes. Here, we identify a biochemical mechanism underlying the preference of PIC-mediated viral DNA integration (PIC-VDI). Specifically, we observed that the PIC-VDI into human chromatin is preferred over the genomic DNA. Surprisingly, nucleosome core particles without any histone modifications were not preferred for PIC-VDI when compared to the analogous naked DNA. However, PIC-VDI was markedly enhanced with nucleosomes containing the trimethylated histone 3 lysine 36 (H3K36me3), an epigenetic mark linked to HIV-1 DNA integration preference. Interestingly, we observed that nucleosomes with flanking linker DNA promoted PIC-VDI in the presence of LEDGF/p75. We also discovered that nucleosomes with linker DNA and H3K36me3 served as the optimal substrate for PIC-VDI. Mapping of the integration sites within these substrates identified preference of specific regions of the nucleosome core DNA for integration. Finally, we provide biochemical and genetic evidence that histone H1 protein, that condenses the chromatin, negatively regulates HIV-1 DNA integration, consistent with the integration preference for open chromatin structure. Collectively, these results identify the role of specific chromatin marks that drive HIV-1 integration preference and define the optimal substrate requirement for efficient DNA integration by the PIC.