The deubiquitination enzyme USP8 is a human T cell-specific factor regulating T-cell maturation and functions (21, 22), and CD4⁺ T cells are primary targets of HIV-1 infection. HIV-1 is potently inhibited by several human cytidine deaminases that are members of human APOBEC3 proteins. HIV-1 Vif neutralizes some APOBEC3’s antiviral functions by forming viral-specific CRL5 E3 ubiquitin ligase complexes (12, 13, 23, 24) to promote the polyubiquitination and degradation of APOBEC3 substrates ( Figure 1A ). Whether any host factors involved in the deubiquitination pathway can influence Vif function has not been explored.

We observed that USP8 is highly expressed in primary CD4⁺ T cells and myeloid cells compared with HEK293T cells and RD cells ( Figures 1B, C ). Primary CD4⁺ T cells and myeloid cells are the preferred target cells of HIV-1. We screened a library of mammalian expression vectors that encode 32 USPs twice and discovered USP8 is distinguished from the group. USP8 significantly impaired Vif-triggered APOBEC3G (A3G) degradation ( Supplementary Figures S1H, I ). To investigate whether endogenous USP8 in T cells has an effect on Vif-induced A3G degradation, we generated USP8 knocking down stable cell line in H9 T cells ( Supplementary Figure S1A ). shCon and shUSP8 H9 cells were infected with wild-type (WT) HIV-1 at the same titer or mock infected. As shown in Figure 1D , silencing USP8 has no effect on endogenous A3G expression (lanes 1 and 3). A3G was degraded during HIV-1 infection (lanes 1 and 2). Surprisingly, A3G degradation was enhanced in virus-infected shUSP8 H9 cells (lanes 2 and 4). In addition, cell supernatants were collected at different time points, and virus infectivity was assessed using TZM-BL indicator cells. We found that USP8 knocking down had a positive effect on HIV-1 progeny virus replication ( Figure 1E ). A similar result was also determined in primary CD4⁺ cells ( Supplementary Figure S1B ). Since USP8 expression in HEK293T cells is lower than that in CD4⁺ T cells, we detected whether increasing the USP8 expression could affect HIV-1 Vif function in HEK293T cells. As a result, increasing the USP8 expression effectively inhibited HIV-1 Vif-induced A3G degradation ( Figure 1F ). In the absence of exogenous USP8 expression, HIV-1 Vif induced a >70% reduction in A3G ( Figure 1F , lane 5) when compared to control cells ( Figure 1F , lane 4). The Vif-induced degradation of A3G was essentially abolished in the presence of increasing amounts of USP8 ( Figure 1F , lanes 6 and 7). In the absence of Vif, A3G expression was not affected by USP8 ( Figure 1F , lanes 1–3). Consistent with previous reports, we also observed that USP8 could rescue Vif expression in a dose-dependent manner, since Vif was ubiquitinated and degraded via the proteasome pathway (25) ( Figure 1F ). Furthermore, silencing endogenous USP8 expression ( Supplementary Figure S1C ) enhanced HIV-1 Vif-mediated A3G degradation in HEK293T cells ( Supplementary Figures S1C, D ). The ability of HIV-1 Vif to counteract A3G is quite conserved among different HIV-1 subtypes. To investigate whether USP8 blocking of anti-A3G activity of Vif is universal, seven viral isolates from patients representing seven HIV-1 subtypes Vif molecules were selected (26, 27), except NL4-3 Vif. We validated that USP8 has the ability to block the anti-A3G activity of Vif molecules from diverse HIV-1 subtypes ( Figure 1G ).

To examine the possible effect of USP8 on HIV-1 infection, we first co-expressed increasing amounts of USP8 with HIV-1 (NL4-3). USP8 expression had no detectable effect on viral protein expression, as determined by the expression of intracellular HIV-1 Pr55Gag in HEK293T cells ( Supplementary Figure S1E ). Virus release ( Supplementary Figure S1F ) and the infectivity of the released virus ( Supplementary Figure S1G ) were also unaffected by USP8 expression in the absence of A3G. In contrast, in the presence of USP8, intracellular A3G expression ( Figure 1H ) and virion incorporation of A3G ( Figure 1I ) were increased, even in the presence of HIV-1 Vif. At the same time, the infectivity of the released HIV-1 was reduced in the presence of USP8 ( Figure 1J ), which is consistent with the increased amount of A3G virion packaging. Combined with these data, these results indicate that USP8 exerts its potent antiviral effect through Vif. USP8 antagonized the ability of HIV-1 Vif to suppress the antiviral function of A3G, resulting in an enhanced antiviral activity of A3G against Vif-containing HIV-1.

In addition to A3G, A3H-Haplotype II (hap II), which is overcome by HxB2 Vif (26), and A3F are also potent inhibitors of HIV-1 that are neutralized by Vif (13, 28, 29). We observed that USP8 can efficiently inhibit HIV-1 Vif-induced degradation of A3F ( Figures 2A, B ) and A3H-HapII ( Figures 2C, D ). The degradation of other known HIV-1 Vif targets, such as A3C ( Supplementary Figures S2A, B ) and A3DE ( Supplementary Figures S2C, D ), was also inhibited by USP8. Non-primate lentiviral Vif from bovine immunodeficiency virus (BIV) targets cow APOBEC3 proteins (30). Unlike HIV-1 Vif, non-primate lentivirus BIV Vif assembles with CUL2 and the ElOB/C E3 component ( Supplementary Figure S2E ) and is not regulated by CBFβ to trigger the degradation of bovine APOBEC3 (31). Since USP8 sequence is highly conserved between cow and human ( Supplementary Figure S2F ), human USP8 was also able to block BIV Vif-induced degradation of cow APOBEC3 in HEK293T cells ( Supplementary Figures S2G, H ). Importantly, we observed that diverse Vif expression could be enhanced by USP8 ( Supplementary Figures S2A, C, G ). Taken together, USP8 targets diverse Vif viral substrate receptors but not the E3 cellular components.

To identify whether the deubiquitinating enzymatic activity of USP8 is required for its antagonism of Vif function, a key residue Cysteine 786 of USP8 is mutated to abolish its enzymatic activity (21) ( Figure 2E ). We found that the active site mutant USP8 C786D lost the ability to inhibit Vif-mediated A3G degradation ( Figure 2F , lane 4) when compared to wild-type USP8 ( Figure 2F , lane 3). In the presence of USP8 C786D, Vif-mediated A3G degradation was as efficient as the no-USP8 control ( Figure 2G , bar 4 vs. bar 2). USP8 is vital for the development and homeostasis of T cells and is cleaved in activated CD4⁺ T cells by caspases. The C-terminal fragment of USP8, but not N-terminal fragment, possesses the enzymatic activity (21). Somatic mutations in USP8 enhance its proteolytic cleavage and are associated with Cushing’s disease resulting from elevated deubiquitinase activity (22). We observed that the C-terminal fragment (C40) but not the N-terminal fragment (N90) of USP8 maintained the inhibitory activity against HIV-1 Vif ( Figures 2H, I ). Collectively, enzymatic activity of USP8 is critical for the inhibitory effect of Vif function.

To explore the mechanism of USP8-mediated Vif inhibition, we employed co-immunoprecipitation (co-IP) assay to determine whether USP8 could interact with HIV-1 Vif. In the absence of Vif, USP8 was not detected in the co-IP sample ( Figure 3A , lane 1), indicating the specificity of the assay system. In the presence of Vif, co-precipitation of USP8 with Vif was detected ( Figure 3A , lane 3). Gads is a binding partner of USP8, which is required for USP8 in regulating the development and homeostasis of T cells (21). The interaction with HIV-1 Vif reduced the ability of USP8 binding with its functional cofactor Gads ( Supplementary Figures S3A, B ). These data establish a specific interaction between HIV-1 Vif and USP8.

Vif forms a viral-specific E3 ubiquitin ligase complex with cellular proteins CUL5, ElOB/C, and CBFβ to trigger polyubiquitination of substrates (12). Vif interacts with USP8; we next assessed whether USP8 disrupts Vif–Cullin 5–E3 ubiquitin ligase complex assembly. Interestingly, USP8 did not interact with CUL5, ElOB, or CBFβ ( Supplementary Figure S3C ) in the absence of Vif. The ability of Vif to interact with CUL5, ElOB, and CBFβ ( Figure 3B , lane 2; Figure 3C ) was not affected by USP8 overexpression ( Figure 3B , lane 3; Figure 3C ). The interaction of Vif with the target protein A3G ( Figure 3D , lane 3; Figure 3E ) was also not affected by USP8 ( Figure 3D , lane 4; Figure 3E ). However, HIV-1 Vif-induced polyubiquitination of A3G ( Figure 3F , lane 1) was significantly reduced in the presence of USP8 ( Figure 3F , lane 2). In contrast, enzymatically defective USP8 C786D, which could not inhibit Vif-triggered A3G degradation, was disabled in suppressing HIV-1 Vif-induced polyubiquitination of A3G ( Figure 3F , lane 3) when compared to the WT USP8 ( Figure 3F , lane 2). Additionally, we confirmed that USP8 repressed Vif-mediated A3G polyubiquitination in vitro study ( Supplementary Figures S3D–F ). Ubiquitinated A3G was purified from HEK293T cells transfected with Ub-Flag, Vif-HA, and A3G-V5 using anti-V5 affinity purification. WT USP8 and functional dominant truncation C40, which purified from HEK293T cells and incubated with ubiquitinated A3G, decreased A3G polyubiquitination, while enzymatic mutation C786D was disabled ( Supplementary Figure S3D ). USP8 C40 purified from Escherichia coli reduced A3G polyubiquitination in a time-dependent fashion ( Supplementary Figure S3E ). Likewise, USP8 C40^C786D, an enzymatic mutant, lost the ability of A3G deubiquitination ( Supplementary Figure S3F ). Together, these results indicated that USP8 specifically interacts with Vif and deubiquitinates the ubiquitination of A3G mediated by Vif.

Although USP8 exhibits strong inhibitory activity against Vif-mediated A3G degradation ( Figure 1F ), the anti-HIV-1 function of USP8 could be compromised during HIV-1 replication. Because USP8 expression was reduced in HIV-1-infected CD4⁺ T cells, which is based on mass spectrometry data published by several groups (17, 32, 33). In both H9 ( Figures 4A, B and Supplementary Figure S4A ) and Jurkat ( Figures 4C, D and Supplementary Figure S4B ) CD4⁺ T cells, HIV-1 infection resulted in depletion of USP8 proteins significantly. Moreover, the USP8 mRNA level was also reduced in HIV-1-infected T cells ( Figures 4E, F ), suggesting that down-modulation of USP8 expression by HIV-1 is at least partially the result of an alteration in USP8 mRNA transcription or RNA stability. The C-terminal fragment of USP8 contains the deubiquitining enzymatic activity (22) and is a potent inhibitor of HIV-1 Vif function ( Figure 2H ). HIV-1 Vif-triggered A3G degradation was blocked slightly more efficiently by USP8 C40 than by full-length USP8 ( Figure 4G ), which was correlated with the inhibition of HIV-1 Vif-induced A3G polyubiquitination ( Supplementary Figures S3D , S4C, D ). To further validate USP8 C40 function during HIV-1 infection, USP8 C40 was transduced into CD4⁺ T cells ( Figure 4H ). Transduction with USP8 C40 in H9 cells that express endogenous APOBEC3 antiviral proteins ( Figure 4I , lanes 3 and 4) significantly inhibited HIV-1 replication ( Figure 4J ). In contrast, USP8 C40 ( Figure 4I , lanes 1 and 2) has no effect on HIV-1 replication ( Figure 4K ) when transduced in APOBEC3-negative Jurkat cells. These data demonstrate that modulation of USP8 function in CD4⁺ T cells can enhance the antiviral activity of APOBEC3 cytidine deaminases against HIV-1 even in the presence of Vif.

Ubiquitin modification is linked to many cellular processes. Diverse DUBs are involved in reversing protein ubiquitination and therefore modulate the outcome of this posttranslational modification. HIV-1 Vpr, HIV-1 Vpu, and HIV-2 Rod/SIVmac Vpx also form virus-specific E3 ubiquitin ligase complexes to mediate the polyubiquitination and degradation of various target proteins ( Supplementary Figures S5A–C ). However, in contrast to USP8-mediated Vif inhibition, USP8 had a little effect on HIV-1 Vpr-mediated HLTF degradation ( Supplementary Figure S5G ), HIV-1 Vpu-mediated BST-2 depletion ( Supplementary Figure S5H ), or HIV-2Rod/SIVmac Vpx-mediated SAMHD1 degradation ( Supplementary Figures S5I, J ). Although USP8 had little effect on HIV-1 Vpr, HIV-1 Vpu, or HIV-2 Rod/SIVmac Vpx function, we identified other potent DUB inhibitors of HIV-1 Vpr, HIV-1 Vpu, and HIV-2 Rod/SIVmac Vpx ( Figures 5A–D ) after screening 32 USPs ( Supplementary Figures S5D–F ). USP7 inhibited HIV-1 Vpr-mediated HLTF degradation ( Figure 5A ), USP33 inhibited HIV-1 Vpu-mediated BST-2 depletion ( Figure 5B ), and USP37 inhibited HIV-2 Rod/SIVmac Vpx-mediated SAMHD1 degradation ( Figures 5C, D ). Moreover, we investigated different USP functions in suitable cell lines as previously reported (16, 20, 34). Silencing USP7 in H9 T cells, HLTF degradation was more efficient in shUSP7 cells in the presence of HIV-1 infection compared with its control cells ( Figure 5E , lanes 2 and 4). As a result, silencing USP7 promoted HIV-1 infectivity ( Supplementary Figure S5K ). Meanwhile, similar result was demonstrated in shUSP33 HeLa cells that BST2 degradation was enhanced during HIV-1 infection ( Figure 5F , lanes 2 and 4) and subsequently promoted HIV-1 infectivity ( Supplementary Figure S5L ). HIV-1 Pr55Gag was determined to demonstrate silencing USPs has no effect on virus replication ( Figures 5E, F ). SAMHD1 was degraded in THP-1 cells when infected with SIVmac virus. As expected, SIV-induced SAMHD1 degradation was more efficient in the absence of USP37 THP-1 cells ( Figure 5G ). Collectively, specific DUBs play important roles in virus–host relationships.

USP8, USP8 N90, and USP8 C40 were constructed by PCR amplification from USP8-HA/Flag (#79639; Addgene) and then inserted between the SalI and BamHI sites of a C-terminal HA/Flag tag VR1012 vector. pET28a-USP8 C40 was constructed by PCR amplification from USP8-HA/Flag and then inserted into the pET28a-Plus vector with a 6xHis-tag at the N terminus. pET28a-USP8 C40 C786D was made from pET28a-USP8 C40 by site-directed mutagenesis. All the USP plasmids were purchased from Addgene. The infectious molecular clone pNL4-3 (WT HIV-1) (38), A3G-HA (13), A3G-V5 (13), HIV-1 NL4-3-Vif-HA (Vif-HA) (38), BIV-Vif-HA (39), A3Z2Z3-HA (39), HIV-2RodVpx (40), SIVmac Vpx-HA (40, 41), SAMHD1-HA (41), A3F-V5 (28), A3C-HA (42), A3DE-HA (42), Vpu-Myc (43) and BST-2-HA (43), A3H-hapII-HA (44), and renilla (45) were as previously described. Vifs from different HIV-1 subtypes have been previously described (26). Vpr-HA, HLTF-Flag, Vif (HxB2)-HA, Vif-Myc, and Vpx-Myc were obtained from the Institute of Virology and AIDS Research, First Hospital of Jilin University, and were as previously described (13, 38, 41).

HEK293T (CRL-11268; ATCC), HeLa (CRM-CCL-2; ATCC), and TZM-bl (PTA-5659; ATCC) cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM; HyClone) containing 10% heat-inactivated fetal bovine serum (FBS, 04-001-1; Biological Industries) and penicillin/streptomycin. H9 (HTB-176; ATCC), Jurkat (TIB-152; ATCC), and THP1 (TIB-202; ATCC) cells were purchased from the ATCC and maintained in Roswell Park Memorial Institute 1640 (RPMI-1640) medium (HyClone) with 10% FBS and penicillin/streptomycin. The peripheral blood mononuclear cells (PBMCs) were isolated through Ficoll gradient centrifugation, and the CD4⁺ T lymphocytes were then purified from the PBMCs with anti-CD4-specific antibody-coated microbeads (Miltenyi Biotec, Germany) according to the manufacturer’s instructions. CD4⁺ T lymphocytes were maintained in RPMI-1640 medium (HyClone) with 10% FBS and penicillin/streptomycin.

DNA transfection was carried out using Lipofectamine 3000 (Invitrogen) according to the manufacturer’s instructions. H9 and Jurkat cells were transfected using the Amaxa Cell Line Nucleofector Kit V (Lonza, Switzerland) with the program G-014 or X-001 according to the manufacturer’s instructions.

The antibodies used in this study are as follows: β-tubulin monoclonal antibody (NMS-410P; Covance), Anti-CUL5 (sc-13014; Santa Cruz Biotechnology), anti-EloB (sc-1144; Santa Cruz Biotechnology), anti-CBFb (sc-166142; Santa Cruz Biotechnology), anti-Vif (GTX80393; GeneTex), anti-USP8 (A7031; ABclonal), anti-HA (901513; Biolegend), anti-Myc (AHO0052; Invitrogen), anti-V5 (R960-25; Invitrogen), anti-SAMHD1 (TA502024; OriGene), anti-HLTF (14786-1-AP; Proteintech), anti-BST2 (13560-1-AP; Proteintech), anti-A3G (D221663;Sangon Biotech), anti-USP7 (26948-1-AP; Proteintech), anti-USP33 (20445-1-AP; Proteintech), anti-USP37 (18465-1-AP; Proteintech), anti-His (sc-8036; Santa Cruz Biotechnology). CAp24 mAb (1513) was purchased from the NIH AIDS Reagents Program. Secondary antibodies were alkaline phosphatase-conjugated anti-rabbit (115–055–045; Jackson ImmunoResearch) and anti-mouse (115–055–062; Jackson ImmunoResearch), HRP-conjugated anti-rabbit (NA934V; GE) and anti-mouse (sc-2005; Santa Cruz Biotechnology).

For immunoblot analysis of cell-associated proteins, whole cell lysates were prepared as follows: Cells were collected in culture medium and centrifuged at 5,000 rpm for 5 min. Each supernatant was mixed with an appropriate volume of lysis buffer (50 mM Tris–HCl, pH 7.8, with 150 mM NaCl, 1% NP-40, 1% sodium deoxycholate, and 4 mM EDTA) and corresponding 4× loading buffer (8% sodium dodecyl sulfate in 320 mM Tris–HCl, pH 6.8, with 40% glycerol and 0.002% bromophenol blue). Proteins were solubilized by heating for 30 min at 100°C, with occasional vortexing to shear cellular DNA. Cell lysates were subjected to sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE). Proteins were transferred to NC membranes (10401396; GE Whatman) and reacted with appropriate antibodies as described in the text, e.g., 1:500 rabbit polyclonal anti-USP8 to detect USP8. Membranes were then incubated with the corresponding secondary antibody, and proteins were visualized using a hypersensitive ECL chemiluminescence detection kit (Proteintech) according to the manufacturer’s protocol; the assembled HIV-1 in the culture supernatants was evaluated by immunoblotting in Figure 1I and Supplementary Figure S1F . HIV-1 in the culture supernatants was filtered through a 0.45-μm filter and mixed with corresponding 4× loading buffer (described above). Proteins were solubilized by heating for 15 min at 100°C with occasional vortexing, and the HIV-1 supernatants were subjected to immunoblotting.

In Figure 3 and Supplementary Figures S3 and S4C , HEK293T cells were co-transfected with an expression vector as indicated. In Figures 3D, F and Supplementary Figure S4C , cells were treated with 10 mM MG132 (Sigma) 12 h prior to harvesting (12, 46, 47). Cells were harvested and lysed in lysis buffer (50 mM Tris, pH 7.5, with 150 mM NaCl, 1% NP-40, and complete protease inhibitor cocktail tablets) at 4°C for 1 h, then centrifuged at 10,000×g for 30 min. Precleared cell lysates were mixed with anti-V5 antibody ( Figures 3D, F and Supplementary Figures S3A , S4C ), anti-Myc antibody ( Figures 3A, B ) or anti-HA antibody ( Supplementary Figure S3C )-conjugated protein G agarose beads and incubated at 4°C overnight. The second day, the beads were washed six times with washing buffer (20 mM Tris, pH 7.5, with 100 mM NaCl, 0.1 mM EDTA, and 0.05% Tween-20) and centrifuged at 800×g for 1 min each time. The beads were eluted with elution buffer (0.1 M glycine-HCl, pH 3.5). The eluted materials were then analyzed by SDS-PAGE and immunoblotting as previously described.

Total RNA was extracted with TRIzol reagent (15596-026; Invitrogen) according to the manufacturer’s instructions. RNA reverse transcription used EasyScript First-Strand cDNA Synthesis SuperMix (AE301; TransGen Biotech) according to the manufacturer’s instructions. The quantitative real-time polymerase chain reaction (qPCR) was carried out on an Mx3005P instrument (Agilent Technologies, Stratagene, USA) using Power SYBR^® Green PCR Master Mix (2x) (4367659; ABI). The primers used in this study are as follows: glyceraldehyde 3-phosphate dehydrogenase (GAPDH)-F: GCAAATTCCATGGCACCGT; GAPDH-R: TCGCCCCACTTGATTTTGG; USP8-RT-F: CTGAAAGACTCTCTGAAAGCCT; USP8-RT-R: CCTTTCTCTTTGGTCTCACAT. Data were normalized to the housekeeping GAPDH gene, and the relative abundance of the transcripts was calculated using Ct models.

To generate knocking down USP8 cell lines, chemically synthesized short interfering RNA (siRNA) and a nonspecific control were purchased from RiboBio Co. Ltd. (Guangzhou, China). The siUSP8 sequences are as follows: sense, #1: GCATAAAGGTGAAGTGGCA; #2: GAAAACAGGAAGAGAGGAT; #3: GCAAAGAGGGGCAAAGAAA.

USP7/USP8/USP33/USP37-specific shRNAs with the following target sites were cloned in the lenti-retroviral vector pLKO.1-puro (Addgene). The shRNA sequences are as follows: USP7 shRNA: 5-CCGGCCTGGATTTGTGGTTACGTTACTCGAGTAACGTAACCACAAATCCAGGTTTTTG-3; USP8 shRNA: 5-CCGGTAAGAGTTATGTGCACAGTGCCCTCGAGGGCACTGTGCACATAACTCTTTTTTTG-3; USP33 shRNA 5-CCGGTCTCGACAGTGGCTTAATTAACTCGAGTTAATTAAGCCACTGTCGAGATTTTTG-3; USP37 shRNA 5-CCGGCCGGATTTGCAGAAGATGATACTCGAGTATCATCTTCTGCAAATCCGGTTTTTG-3. HEK293T cells were co-transfected with sh-USP7/USP8/USP33/USP37-pLKO.1 or pLKO.1 plus RRE, REV, and VSV-G expression vectors by using Lipofectamine 3000. At 48 h after transfection, supernatants containing packaged lentivirus were harvested and used to infect H9, Hela, or THP1 cells for 96 h as indicated. Puromycin (3 μg/ml for HEK293T, 5 μg/ml for Hela, and 1.5 μg/ml for THP1) was then added into the culture to screen for stable cell lines.

HEK293T cells were transfected with pNL4-3 (WT HIV-1) plasmid; 48 h later, the supernatant was collected and filtered through a 0.45-μm filter. Then, cells were infected with pNL4-3 virus in the presence of diethylaminoethyl-dextran hydrochloride (DEAE; 20 μg/ml) for 30 h. The culture medium was changed 30 h after infection and replaced with fresh medium and harvested at the indicated time. HIV-1 infectivity was assessed using TZM-BL indicator cells. LTR-luciferase was activated when TZM-BL cells were infected by HIV-1. TZM-BL cells were seeded in 24-well format plates (2 × 104 cells/well); 24 h later, the cells were infected with the equivalent of 0.5 ng of HIV-1 p24 antigen in the presence of 20 μg/ml DEAE. Cells were collected and lysed after 48 h. LTR-luciferase activity was measured with the Dual-Luciferase Reporter Assay System (E1910; Promega) according to the manufacturer’s protocol.

Ubiquitinated A3G was isolated from HEK293T cells transfected with expression vectors of Ub-Flag, Vif-HA, and A3G-V5 and then purified from the cell extracts with anti-V5 antibody-conjugated protein G agarose beads. USP8 or its mutant was purified from HEK293T cells overexpressing USP8-HA or its mutant using anti-HA-Agarose antibody in IP buffer [1% (vol/vol) Triton X-100, 50 mM Tris–HCl pH 7.4, 50 mM EDTA, 150 mM NaCl, 10 mM NaF, 10% glycerol, and fresh protease inhibitor cocktail]. For in vitro deubiquitination assay, ubiquitinated A3G protein was incubated with USP8 or its mutant in the deubiquitination buffer (20 mM Tris–HCl pH 8.0, 200 mM NaCl, 1 mM EDTA, 10 mM DTT, 5% glycerol) for 1 h at 37°C. The ubiquitinated A3G was analyzed by immunoblotting.

pET28a-USP8 C40 and pET28a-USP8 C40 C786D fusion protein were expressed in the strain of BL21(DE3) and purified by metal-affinity chromatography on chelation resin. For in vitro deubiquitination assay, ubiquitinated A3G protein was incubated with USP8 C40 or its mutant in the deubiquitination buffer (mentioned above) for the indicated time courses at 37°C. The ubiquitinated A3G was analyzed by immunoblotting.

Data from the protein quantitative analysis and luciferase reporter assays are presented as means and standard derivations. Differences among groups were analyzed by ANOVA test (Stata Corp., College 251 Station, TX, USA) (NS, not significant; *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001).