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

Abstract

Viruses rely on host cellular machinery for replication. After entering the nucleus, the HIV genome accumulates in nuclear niches where it undergoes reverse transcription and integrates into neighbouring chromatin, promoting high transcription rates and new virus progeny. Despite antiretroviral treatment, viral genomes can persist in these nuclear niches and reactivate upon treatment interruption, raising the possibility that they could play a role in the establishment of viral reservoirs. The post-nuclear entry dynamics of HIV remain unclear, and understanding these steps is critical for revealing how viral reservoirs are established. In this study, we elucidate the formation of HIV-induced CPSF6 puncta and the domains of CPSF6 essential for this process. We also explore the roles of nuclear speckle (NS) scaffold factors, SON and SRRM2, in the biogenesis of these puncta. Through genetic manipulation and depletion experiments, we demonstrate the key role of the intrinsically disordered region of SRRM2 in enlarging NSs in the presence of the HIV capsid. We identify the FG domain of CPSF6 as essential for both puncta formation and binding to the viral core, which serves as the scaffold for CPSF6 puncta. While the low-complexity regions modulate CPSF6 binding to the viral capsid, they do not contribute to puncta formation, nor do the disordered mixed charge domains of CPSF6. Interestingly, the FG peptide facilitates viral replication. These results demonstrate how HIV evolved to hijack host nuclear factors, enabling its persistence in the host. Of note, this study provides new insights into the underlying interactions between host factors and viral components, advancing our understanding of HIV nuclear dynamics and offering potential therapeutic targets for preventing viral persistence.

Keywords: HIV; cell biology; condensates; infectious disease; microbiology; nuclear speckles; nucleus; post-nuclear entry; viruses.

Figure 1.. Role of HIV-induced CPSF6 puncta in the nuclear reverse transcription upon removal of NVP.

(A) THP-1 cells, infected with VSV-G ∆env HIV-1 (NL4.3) ∆R LUC (MOI 10) in presence or not of Nevirapine (10 µM) for 5 days, or in presence of Nevirapine (10 µM) for 2 days and then the remaining 3 days without drug or in presence of Nevirapine (10 µM) for 2 days then in presence of PF74 (25 µM). Confocal microscopy images, to verify the presence of CPSF6 puncta, the cells are stained with anti-CPSF6 antibody (green). Nuclei are stained with Hoechst (blue). Scale bar 10 µm. (B) Luciferase assay, to verify luciferase expression in the aforementioned samples. Luciferase values were normalized by total proteins revealed with the Bradford kit. One-way ANOVA statistical test with multiple comparison was performed (****p < 0.0001; *p < 0.05; ns, p > 0.05). Data are representative of two independent experiments. (C) Western blots demonstrate CPSF6 depletion using a specific antibody against CPSF6 in THP-1 cells subjected to CRISPR–Cas9 methods: CRISPR–Cas9 bulk (left), and CRISPR–Cas9 clones selected by limiting dilution (right). Each condition is normalized for actin labelling. The ratio between the intensity signal of CPSF6 and actin was analysed via ImageJ and is plotted below each western blot. (D) Confocal microscopy images of THP-1 ctrl CRISPR clone 2 cells (Ctrl 2) and THP-1 duplex1-2-3 CRISPR clone 4 cells (CPSF6 KO 4) infected with VSV-G/HIV-1ΔEnvIN_HA LAI (BRU) -vpx (MOI 10) in the presence of Nevirapine (10 µM). The cells are stained 30 hr p.i. with anti-CPSF6 antibody and anti-HA antibodies to detect HA tagged integrase (IN).

Figure 1—figure supplement 1.. Multiple examples of confocal microscopy images of THP-1 ctrl CRISPR clone 2 cells and THP-1 CPSF6 KO 4 cells infected with VSV-G/HIV-1ΔEnvIN_HA LAI (BRU) -vpx (MOI 10) in the presence of Nevirapine (10 µM).

The cells are stained 30 hr p.i. with CPSF6 (green) and HA (red) antibodies to detect integrase (IN).

Figure 1—figure supplement 2.. Multiple examples of THP-1 CPSF6 KO clone 4 cells transduced with different LVs carrying CPSF6 WT or mutants and stained with CPSF6 and HA antibody 30 hr p.i.

Scale bar 5 µm.

Figure 2.. Role of CPSF6 domains in HIV-induced CPSF6 puncta.

(A) Schema of CPSF6 isoform 588 aa deletion mutants. (B) Confocal microscopy images of THP-1 CPSF6 KO cells, transduced with different mutants of CPSF6, infected with VSV-G/HIV-1ΔEnvIN_HA LAI (BRU) -vpx (MOI 10) in presence of Nevirapine (10 µM). The cells are stained with CPSF6 and HA antibody 30 hr p.i. Scale bar 5 µm. (C) Analysis of the number of CPSF6 puncta in THP-1 CPSF6 KO cells transduced with different mutants of CPSF6, not infected or infected in the presence of Nevirapine (10 µM) (the number of analysed cells is shown under the x-axis). Statistical test: ordinary one-way ANOVA (****p < 0.0001; ***p < 0.001; *p < 0.05; ns, p > 0.05). (D) The plot compares the number of CPSF6 puncta per cell in THP-1 CPSF6 KO cells transduced with different mutants of CPSF6, infected with HIV-1 in the presence of Nevirapine (10 µM). Statistical test: ordinary one-way ANOVA (****p < 0.0001; ns, p > 0.05). (E) Confocal microscopy images of THP-1 CPSF6 KO clone 4, non-transduced and non-infected or transduced with WT CPSF6 and CPSF6 3xNLSΔMCD and infected with VSV-G/HIV-1ΔEnvIN_HA LAI (BRU) -vpx (MOI 10) in presence of Nevirapine (10 µM). Immuno-RNA FISH: the cells are stained with CPSF6 (green) antibody and with 24 probes against HIV-1 Pol sequence (grey) (RNA-FISH) 25 hr p.i. Nuclei are stained with Hoechst (blue). Scale bar 10 µm. Violin plot presenting the percentage of CPSF6 puncta colocalizing with the viral RNA in THP-1 CPSF6 KO clone 4 cells transduced with LVs expressing CPSF6 WT or CPSF6 3xNLSΔMCD (respectively, n = 73 and n = 103) and infected with VSV-G/HIV-1ΔEnvIN_HA LAI (BRU) -vpx (MOI 10) in presence of Nevirapine (10 µM). A total of 198 CPSF6 WT puncta and 264 CPSF6 3xNLSΔMCD puncta were counted. Statistical test: unpaired t-test, ns, p > 0.05. (F) Confocal microscopy images of THP-1 KO CPSF6 cells transduced with WT CPSF6 and CPSF6 ∆MCD without NLS, with 3xNLS or with PY NLS, respectively. Cells were differentiated for 3 days, transduced with CPSF6 lentiviral vectors (MOI 1) for 3 days and infected for 24 hr with VSV-G/HIV-1ΔEnvIN_HA LAI (BRU) -vpx (MOI 10) in the presence of Nevirapine (10 µM). The panels show transduced and uninfected cells. CPSF6 and the IN tagged with the HA are labelled with anti-CPSF6 (green) and anti-HA (white) antibodies, respectively. Nuclei are stained with Hoechst (blue). The arrows show CPSF6 puncta in colocalization with IN-HA. Scale bar 10 µm.

Figure 2—figure supplement 1.. Per-residue intrinsic disorder propensity of the CPSF6 isoform 588 aa evaluated by the Rapid Intrinsic Disorder Analysis Online platform (RIDAO) (Dayhoff and Uversky, 2022) that yields results for IU-Pred_short (yellow line), IUPred_long (blue line), PONDR VL3 (green line), PONDR VLXT (black line), PONDR VSL2 (red line), and PONDR FIT (pink line) and computes a mean disorder score for each residue based on these predictors (MDP, thick, dark pink, dashed line).

Light pink shadow represents MDP error distribution. The thin black line at the disorder score of 0.5 is the threshold between order and disorder, where residues/regions with disorder scores above 0.5 are disordered, and residues/regions below 0.5 are ordered. The dashed line at the disorder score of 0.15 is the threshold between order and flexibility, where residues/regions above 0.15 are flexible, and residues/regions below 0.15 are highly ordered (upper). Schema of the deletion mutants of CPSF6 (bottom).

Figure 2—figure supplement 2.. Lentiviral vector transduction of phorbol 12-myristate 13-acetate (PMA)-differentiated THP-1 cells expressing CPSF6 ∆MCD fused to mNeonGreen (left), CPSF6 NLS∆MCD fused to mNeonGreen (centre), and CPSF6 3xNLS∆MCD fused to mNeonGreen (right).

CPSF6 is represented in green, and nuclei are stained in blue. Scale bar 5 µm.

Figure 3.. Evaluation of CPSF6 deletion mutants’ binding capacity to the viral core.

(A) Ability of wild-type and mutant CPSF6 proteins to bind to the HIV-1 core. Cellular extracts derived from human 293T cells expressing similar levels of the indicated CPSF6 proteins (INPUT) were incubated with HIV-1 capsid stabilized tubes for 1 hr at room temperatures in the presence and absence of 10 µM PF74, as described in materials and methods. As a carrier control, we utilized DMSO. Subsequently, HIV-1 capsid stabilized tubes were washed, and the bound proteins were eluted 1X Laemmli buffer 1X. The BOUND fractions were analysed by western blotting using antibodies against neon-GFP and the HIV-1 capsid. (B) Experiments were repeated at least three times and the average BOUND fraction relative to the INPUT fraction normalized to wild-type binding is shown for the different CPSF6 mutants. *** indicates a p-value <0.001; **** indicates a p-value <0.0001; and ns indicates no significant difference as determined by unpaired t-tests.

Figure 4.. Comparison of second structures of ADD2 and low-complexity region (LCR).

(A) Physicochemical characteristics of the LCR-FG and ADD2-FG sequences. Intrinsic disorder predispositions evaluated by PONDR VLXT. Position of the FR segment within the LCR-FG and ADD2-FG sequences is shown as grey shaded area. (B) Linear distribution of the net charge per residue (NCPR) within the LCR-FG sequence evaluated by CIDER. (C) Linear distribution of the NCPR within the ADD2-FG sequence evaluated by CIDER. (D) Secondary structure propensity of the LCR-FG sequence evaluated by PSIPRED. (E) Secondary structure propensity of the ADD2-FG sequence evaluated by PSIPRED. (F) Analysis of the peculiarities of the amino acid compositions of the intrinsically disordered C-terminal domain (residues 261–358) of human CPSF6 and its different mutants. Relative abundance of prion-like LCR defining uncharged residues in analysed protein segments. (G) Relative abundance of proline residues in analysed protein segments. (H) Relative abundance of charged residues in analysed protein segments. The values were calculated by dividing numbers of prion-like LCR defining uncharged (Ala, Gly, Val, Phe, Tyr, Leu, Ile, Ser, Thr, Pro, Asn, Gln, Pro) and charged (Asp, Glu, Lys, Arg) residues by the total number of amino acids in the respective protein fragments. Corresponding values for all protein sequences deposited in the UniProtKB/Swiss-Prot database, PDB Select25, and DisProt are shown for comparison.

Figure 4—figure supplement 1.. Sequences of FG and low-complexity regions (LCRs) or substituted amino acid sequences analysed in Figure 4.

Figure 5.. Role of FG motif in viral replication.

(A) WB showing CPSF6 protein from several single clones derived from CPSF6 KO clone obtained upon complementation with CPSF6 ∆FG and normalized with beta-actin. (B) Quantification of the expression of CPSF6 ∆FG protein in different single clones compared to CPSF6 WT (value 1). (C) Infectivity assay using a single-round infectious virus carrying the cDNA of Luciferase as reporter gene. Values are expressed as % of RLU compared to WT cells. (D) Infectivity assay of a replication competent virus: vRNA from new viruses produced after infection of WT, CPSF6 KO, and CPSF6∆FG cells was analysed and shown in the histograms as % of vRNA copies compared to vRNA in WT THP-1 cells considered 100%.

Figure 6.. Depletion of mixed charge domain (MCD) or low-complexity region (LCR) does not affect the formation of HIV-induced CPSF6 puncta.

(A) Epifluorescence microscopy images of both infected and non-infected differentiated THP-1 cells showing the presence of CPSF6 puncta only in the infected condition. CPSF6 and SC35 are labelled with anti-CPSF6 (green) and anti-SC35 (red) antibodies, respectively. Nuclei are stained with Hoechst (blue). Scale bar 10 µm. (B) Confocal microscopy images of THP-1 KO CPSF6 cells, differentiated for 3 days, transduced with CPSF6 lentiviral vector (MOI 1) (specifically WT CPSF6, CPSF6 ∆LCRs, CPSF6 ∆MCD with 3xNLS, without NLS, or with PY NLS) for 3 days and infected for 24 hr with VSV-G/HIV-1ΔEnvIN_HA LAI (BRU) -vpx MOI 10 in presence of Nevirapine (10 µM). CPSF6 and nuclear speckles were labelled with anti-CPSF6 (green) and anti-SC35 (red) antibodies, respectively. Nuclei are stained with Hoechst (blue). Scale bar 10 µm. The percentage of CPSF6 puncta associated with SC35 per field of view is shown in the graph. N cells were counted in each condition and a one-way ANOVA statistical test with multiple comparison was performed; ns, p-value >0.05.

Figure 7.. Dynamics of the HIV-induced CPSF6 puncta formation and their fusion with NSs.

(A) Time course of infection of THP-1, 6, 9, 12, 30 h.p.i., or non-infected. Cells were stained with antibodies against CPSF6 (green) and SC35 (red). (B) The graph shows the percentage of CPSF6 puncta associated with nuclear speckle (NS) or adjacent to NS or isolated from NS at different time post-infection. (C) The graph shows the progression of CPSF6 puncta associated with NS during the time post-infection. N indicates the number of cells analysed. One-way ANOVA statistical test with multiple comparison was performed; ns, p-value >0.05; **** indicates a p-value <0.0001.

Figure 7—figure supplement 1.. THP-1 infected with VSV-G/HIV-1ΔEnvIN_HA LAI (BRU) -vpx were labelled after 6, 9, 12, 30 h.p.i. with specific antibodies against CPSF6 (green) and SC35 (red), nuclei are stained with Hoechst (blue).

Scale bar 10 µm.

Figure 7—figure supplement 2.. Timelapse of Figure 7—video 2 which recapitulates the key steps of the dynamics of HIV-induced-CPSF6 puncta biogenesis.

CPSF6 was visualized transducing HEK293 SRRM2 HaloTag cells for 24 hr with CPSF6-mNeonGreen lentiviral vector (MOI 0.5). SRRM2-HaloTag allowed the use of TMR-Halo Tag Ligand to visualize the SRRM2 membraneless organelles (MLOs). After 9 hr after VSV-G/ΔEnvHIV-1 LAI (BRU) (MOI 10) infection in the presence of Nevirapine, CPSF6 puncta formed independently (9 h 01 m), migrating afterwards to the SRRM2 MLOs and leading to a final fusion between CPSF6 puncta and SRRM2 MLOs (9h14m).

Figure 8.. Role of SRRM2 and SON in the formation of HIV-induced CPSF6 puncta.

(A) Depletion of SON and SRRM2 in THP-1 cells using AUMsilence ASO technology. The degree of depletion is quantified by WB and the mean intensity through immunofluorescence using antibodies against SON and SRRM2, respectively. Scale bar 5 µm. (B) The percentage of CPSF6 puncta formation is quantified by IF in THP-1 cells knocked down for SON, SRRM2, and control (Ctrl) infected with HIV-1 (MOI 25) for 48 hr. CPSF6 is stained with an antibody against CPSF6 (green), and nuclei are stained with Hoechst (blue). The graph on the right reports the percentage of CPSF6 puncta calculated from more than 100 cells. Scale bar 10 µm. Experiments were performed at least twice. (C) (Top panels) Confocal microscopy images of ∆IDR HaloTag SRRM2 HEK293 and HaloTag SRRM2 HEK293 cells stained with the halo tag ligand (red), and nuclei (blue). Scale bar 10 µm. (Bottom panels) Confocal microscopy images of HaloTag SRRM2 HEK293 and ∆IDR HaloTag SRRM2 HEK293 cells, both labelled with anti-SRRM2 (red) and anti-SON (grey) antibodies. Nuclei are stained with Hoechst (blue). Scale bar 10 µm. Statistical studies are summarized in the violin plot which displays the distribution of the number of SON puncta per cell in the two conditions. N cells were counted and Kolmogorov–Smirnov test was performed, ns, p > 0.05. (D) Confocal microscopy images of HaloTag SRRM2 HEK293 and ∆IDR HaloTag SRRM2 HEK293 cells, either non-infected or infected for 24 hr with VSV-G/HIV-1ΔEnvIN_HA LAI (BRU) (MOI 10) in the presence of Nevirapine (10 µM). CPSF6 and SC35 are labelled with anti-CPSF6 (red) and anti-SC35 (grey) antibodies, respectively. Nuclei are stained with Hoechst (blue). The plot shows the mean ± SD of the percentage of cells with CPSF6 puncta calculated in n fields of view (n = 24, 29, 32); N is the number of cells analysed for each of the three different cell lines; an unpaired t-test was performed, ****p < 0.0001; ns, p > 0.05. Scale bar 10 µm. Experiments were performed at least twice.

Figure 8—figure supplement 1.. Analysis of the mean intensity: SRRM2 depletion is confirmed using antibodies against both (A) SC35 and (B) SRRM2 by IF.

Statistical analysis: one-way ANOVA (****p < 0.0001; *p < 0.05; ns, p > 0.05).

Figure 8—figure supplement 2.. HEK 293 HaloTag SRRM2 or ∆IDR were labelled with Halo ligand (red) and an anti-SON antibody (grey), nuclei are stained by Hoechst (blue).

Scale bar 10 µm.

Figure 8—figure supplement 3.. Multiple examples (A–E) of confocal microscopy images of HaloTag SRRM2 HEK293 and ∆IDR HaloTag SRRM2 HEK293 cells, either non-infected or infected for 24 hr with VSV-G/HIV-1ΔEnvIN_HA LAI (BRU) (MOI 10) in the presence of Nevirapine (10 µM).

CPSF6 and SC35 are labelled with anti-CPSF6 (red) and anti-SC35 (grey) antibodies, respectively. Nuclei are stained with Hoechst (blue). Scale bar 10 µm.