Summary Human immunodeficiency virus (HIV-1) remains a major health threat. Viral capsid uncoating and nuclear import of the viral genome are critical for productive infection. The size of the HIV-1 capsid is generally believed to exceed the diameter of the nuclear pore complex (NPC), indicating that capsid uncoating has to occur prior to nuclear import. Here, we combined correlative light and electron microscopy with subtomogram averaging to capture the structural status of reverse transcription-competent HIV-1 complexes in infected T cells. We demonstrated that the diameter of the NPC in cellulo is sufficient for the import of apparently intact, cone-shaped capsids. Subsequent to nuclear import, we detected disrupted and empty capsid fragments, indicating that uncoating of the replication complex occurs by breaking the capsid open, and not by disassembly into individual subunits. Our data directly visualize a key step in HIV-1 replication and enhance our mechanistic understanding of the viral life cycle.

Nuclear entry of HIV-1 replication complexes through intact nuclear pore complexes is critical for successful infection. The host protein cleavage-and-polyadenylation-specificity-factor-6 (CPSF6) has been implicated in different stages of early HIV-1 replication. Applying quantitative microscopy of HIV-1 reverse-transcription and pre-integration-complexes (RTC/PIC), we show that CPSF6 is strongly recruited to nuclear replication complexes but absent from cytoplasmic RTC/PIC in primary human macrophages. Depletion of CPSF6 or lack of CPSF6 binding led to accumulation of HIV-1 subviral complexes at the nuclear envelope of macrophages and reduced infectivity. Two-color stimulated-emission-depletion microscopy indicated that under these circumstances HIV-1 complexes are retained inside the nuclear pore and undergo CA-multimer dependent CPSF6 clustering adjacent to the nuclear basket. We propose that nuclear entry of HIV-1 subviral complexes in macrophages is mediated by consecutive binding of Nup153 and CPSF6 to the hexameric CA lattice.

shRNA overexpression from viral gene therapy vectors can trigger cytotoxicity leading to organ failure and lethality in mice and rats. This process likely involves saturation of endogenous cellular RNAi factors including exportin-5 (Xpo-5). Here, we have shown that Xpo-5 overexpression enhanced shRNA efficiency in the liver of adult mice but increased hepatotoxicity. We identified the 4 members of the human Argonaute (Ago) protein family as downstream factors involved in saturation of endogenous cellular RNAi, all of which were able to interact with shRNAs in cells and mice. In Ago/shRNA coexpression studies, Ago-2 (Slicer) was the primary rate-limiting determinant of both in vitro and in vivo RNAi efficacy, toxicity, and persistence. In adult mice, vector-based Ago-2/Xpo-5 coexpression enhanced U6-driven shRNA silencing of exogenous and endogenous hepatic targets, reduced hepatotoxicity, and extended RNAi stability by more than 3 months. Use of weaker RNA polymerase III promoters to minimize shRNA expression likewise alleviated in vivo toxicity and permitted greater than 95% persistent knockdown of hepatitis B virus and other transgenes in mouse liver for more than 1 year. Our studies substantiate that abundant small RNAs can overload the endogenous RNAi pathway and reveal possible strategies for reducing hepatotoxicity of short-and long-term clinical gene silencing in humans.

Its remarkable ease and efficiency make the CRISPR (clustered regularly interspaced short palindromic repeats) DNA editing machinery highly attractive as a new tool for experimental gene annotation and therapeutic genome engineering in eukaryotes. Here, we report a versatile set of plasmids and vectors derived from adeno-associated virus (AAV) that allow robust and specific delivery of the two essential CRISPR components - Cas9 and chimeric g(uide)RNA - either alone or in combination. All our constructs share a modular design that enables simple and stringent guide RNA (gRNA) cloning as well as rapid exchange of promoters driving Cas9 or gRNA. Packaging into potent synthetic AAV capsids permits CRISPR delivery even into hard-to-transfect targets, as shown for human T-cells. Moreover, we demonstrate the feasibility to direct Cas9 expression to or away from hepatocytes, using a liver-specific promoter or a hepatic miRNA binding site, respectively. We also report a streamlined and economical protocol for detection of CRISPR-induced mutations in less than 3 h. Finally, we provide original evidence that AAV/CRISPR vectors can be exploited for gene engineering in vivo, as exemplified in the liver of adult mice. Our new tools and protocols should foster the broad application of CRISPR technology in eukaryotic cells and organisms, and accelerate its clinical translation into humans.