To infect nondividing cells, HIV-1 needs to cross the nuclear membrane. The importin transportin-SR2 (TRN-SR2 or transportin-3) has been proposed to mediate HIV-1 nuclear import, but the detailed mechanism remains unresolved. The direct interaction of TRN-SR2 with HIV-1 integrase (IN) has been proposed to drive HIV-1 nuclear import. Alternatively, TRN-SR2 may play an indirect role by mediating nuclear import of cleavage and polyadenylation specificity factor 6 (CPSF6). To unravel the role of TRN-SR2, we designed CRISPR/Cas9 guide RNAs targeting different exons of TNPO3. Although this approach failed to generate full knockouts, monoallelic knockout clones were generated with indel mutations. HIV-1 replication was hampered in those clones at the level of HIV-1 nuclear import without an effect on the cellular distribution of the TRN-SR2 cargoes CPSF6 or alternative splicing factor1/pre-mRNA splicing factor SF2 (ASF/SF2). Recombinant ΔV105 TRN-SR2 expressed in clone 15.15 was 2-fold impaired for interaction with HIV-1 IN and classified as an interaction mutant. Our data support a model whereby TRN-SR2 acts as a cofactor of HIV-1 nuclear import without compromising the nuclear import of cellular cargoes. CRISPR/Cas9-induced mutagenesis can be used as a method to generate interface mutants to characterize host factors of human pathogens. IMPORTANCE Combination antiretroviral therapy (cART) effectively controls HIV-1 by reducing viral loads, but it does not cure the infection. Lifelong treatment with cART is a prerequisite for sustained viral suppression. The rapid emergence of drug-resistant viral strains drives the necessity to discover new therapeutic targets. The nuclear import of HIV-1 is crucial in the HIV-1 replication cycle, but the detailed mechanism remains incompletely understood. This study provides evidence that TRN-SR2 directly mediates HIV-1 nuclear import via the interaction with HIV-1 integrase. The interaction between those proteins is therefore a promising target toward a rational drug design which could lead to new therapeutic strategies due to the bottleneck nature of HIV-1 nuclear import.

Moloney murine leukemia virus (MLV) infects BALB/c mice and induces T-cell lymphoma in mice. Retroviral integration is mediated by the interaction of the MLV integrase (IN) with members of the bromodomain and extraterminal motif (BET) protein family (BRD2, BRD3, and BRD4). The introduction of the W390A mutation into MLV IN abolishes the BET interaction. Here, we compared the replication of W390A MLV to that of wild-type (WT) MLV in adult BALB/c mice to study the role of BET proteins in replication, integration, and tumorigenesis in vivo. Comparing WT and W390A MLV infections revealed similar viral loads in the blood, thymus, and spleen cells. Interestingly, W390A MLV integration was retargeted away from GC-enriched genomic regions. However, both WT MLV- and W390A MLV-infected mice developed T-cell lymphoma after similar latencies represented by an enlarged thymus and spleen and multiorgan tumor infiltration. Integration site sequencing from splenic tumor cells revealed clonal expansion in all WT MLV- and W390A MLV-infected mice. However, the integration profiles of W390A MLV and WT MLV differed significantly. Integrations were enriched in enhancers and promoters, but compared to the WT, W390A MLV integrated less frequently into enhancers and more frequently into oncogene bodies such as Notch1 and Ppp1r16b. We conclude that host factors direct MLV in vivo integration site selection. Although BET proteins target WT MLV integration preferentially toward enhancers and promoters, insertional lymphomagenesis can occur independently from BET, likely due to the intrinsically strong enhancer/promoter of the MLV long terminal repeat (LTR). IMPORTANCE In this study, we have shown that the in vivo replication of murine leukemia virus happens independently of BET proteins, which are key host determinants involved in retroviral integration site selection. This finding opens a new research line in the discovery of alternative viral or host factors that may complement the dominant host factor. In addition, our results show that BET-independent murine leukemia virus uncouples insertional mutagenesis from gene enhancers, although lymphomagenesis still occurs despite the lack of an interaction with BET proteins. Our findings also have implications for the engineering of BET-independent MLV-based vectors for gene therapy, which may not be a safe alternative.