Decoding the biogenesis of HIV-induced CPSF6 puncta and their fusion with nuclear speckles

Tomasini et al., eLife, 2026 asked how HIV builds the nuclear compartments it replicates in. After the capsid crosses the nuclear pore, the virus drives the host protein CPSF6 into puncta where reverse transcription finishes and the pre-integration complex assembles. The question was which part of CPSF6 creates those puncta, and which nuclear speckle proteins the puncta then fuse with. Working in THP-1 cells differentiated into macrophage-like cells, the authors showed that a single disordered stretch of CPSF6, the FG peptide, is required both to bind the viral core and to nucleate the puncta, while the low complexity regions and the mixed charge domain are dispensable. A time course showed the puncta forming first on their own and then merging with nuclear speckles, the fused fraction rising from 61% at 6 hours after infection to 75% at 30 hours.

To test which speckle scaffold matters, the authors knocked down SRRM2 and SON with AUMsilence sdASOs against each messenger RNA, with a matched AUM scramble control, at 10 μM added straight to the medium of the differentiated macrophage-like cells. No transfection reagent was used with the oligonucleotides. The transfections in the methods are plasmid work in HEK293T cells, for virus production and for the capsid binding assay. Depletion was confirmed by western blot and by immunofluorescence intensity.

The knockdowns separated the two scaffolds. Among infected control cells about 78% carried CPSF6 puncta. After SRRM2 knockdown that fell to about 43%, while SON knockdown gave a milder reduction to about 66%. A genetic experiment agreed: in HEK293 cells engineered to express SRRM2 without its disordered C-terminal region, the proportion of cells with puncta fell from about 27% to about 11%. Together the oligonucleotide and the genetic result place the disordered region of SRRM2 at the centre of the fusion step that stabilises the nuclear niches HIV uses.