Direct HIV-1 CA labeling by genetic code expansion using the amber suppression system

HIV-1 infects CD4+ T cells and macrophages as specific target cells where it replicates by hijacking host cell components. This capsid containing the viral RNA genome consists of ~ 1,200-1,500 capsid proteins (CA) and is transported towards the nucleus to integrate into the host genome. Before integration, the capsid needs to be removed in a process termed uncoating. However, the exact nature and intracellular location of uncoating is not known. Microscopy-based analysis would be capable to address this issue by detecting individual particles. The lack of an experimental system for direct fluorescent labeling of the capsid structure impaired microscopic analyses of CA content within subviral complexes. Experiments relying on immunofluorescence staining depend on epitope accessibility, while CA fusions to fluorescent proteins severely affect infectivity. To overcome these issues, a site-specific and minimally invasive labeling strategy of HIV-1 CA based on genetic code expansion via amber suppression followed by click-labeling with silicon rhodamine (SiR) was developed. 18 amino acids within the protein were screened to identify positions permissive for modification. Out of these candidates, alanine 14 allowed labeling of > 95 % of particles within a preparation while retaining ~50% infectivity without the need for complementation with the wild-type protein. Therefore, this strategy enabled reliable CA content analysis for subviral complexes in different subcellular compartments and single virus tracking by confocal imaging, super-resolution STED nanoscopy, and correlative light electron microscopy (CLEM). Directly labeled particles were detected inside the nucleus of HeLa-derived cells and primary CD4+ T cells with intensities corresponding to the amount of a completely assembled CA lattice. CLEM revealed several closely adjacent cone-shaped objects inside the nucleus of cell cycle arrested T cells correlating to positions of CA SiR signals. These data strongly argue for the translocation of largely intact capsids through nuclear pore complexes (NPCs) into the nucleus in both cell types. Additionally, the analysis of cytoplasmic and nuclear signals by STED imaging suggested nuclear capsid trafficking on common routes resulting in dose-dependent capsid clustering inside the nucleus. The unique properties of the direct labeling approach allowed to clarify the controversially discussed effect of the CA-targeting inhibitor PF74. Upon adding the drug, the complete CA-content of nuclear complexes was retained, indicating that PF74 has a stabilizing effect on the capsid lattice. Together these data support a model in which intact capsids enter the nucleus through NPCs, where uncoating must occur before productive integration. In conclusion, the establishment and thorough validation of this HIV-1 CA labeling method provides a versatile tool to investigate uncoating dynamics. The approach may be applicable to the capsids of other viruses