SLFN14 functions as a P-TEFb inhibitor to modulate the transcription of HIV-1 and cellular genes Authors: Shu-min Chen, Ling Ma, Zhen Wang, Quan-jie Li, Rui Zhou, Jing Wang, Ji-wei Ding, Jian-yuan Zhao, Dong-rong Yi, Yong-xin Zhang, Sai-sai Guo, Xin-lu Wang, Xiao-yu Li, Zhen-long Liu, Shan Cen Published: 2025 ABSTRACT The Schlafen (SLFN) gene family encodes different growth regulatory factors known to participate in cell growth, differentiation, tumorigenesis, and the immune response against virus infection. However, the molecular mechanisms underlying their functions remain largely incomplete. In this study, we report that the SLFN family member SLFN14 has potent anti–HIV-1 activity. Specifically, SLFN14 binds to the positive transcription elongation factor b (P-TEFb), thereby preventing HIV-1 Tat-mediated recruitment of P-TEFb to the HIV-1 promoter and causing an arrest of viral transcription. Furthermore, this P-TEFb–targeted inhibiting activity also allows SLFN14 to regulate the transcription of a subset of cellular genes. Thus, our data identify SLFN14 as an inhibitor of the important transcription elongation factor P-TEFb complex, which unveils the crucial role of SLFN14 in both cellular antiviral defense and other important cellular processes that rely on the function of this complex. INTRODUCTION The Schlafen (SLFN) family comprises of a number of genes, which were initially characterized for their involvement in thymocyte maturation in mice (1). Mouse has 10 SLFN genes, while human has 6: SLFN5, SLFN11, SLFN12, SLFN12L, SLFN13, and SLFN14. Except for SLFN12 and SLFN12L, human SLFN proteins share a common core region that contains a divergent AAA domain (AAA_4) with adenosine 5′-triphosphate–binding activity and a putative RNA helicase motif belonging to the superfamily I helicases (2). SLFN family members have been investigated for their essential functions in cell growth (3), differentiation (4), and tumorigenesis (5, 6). In addition, SLFNs are preferentially expressed in lymphoid tissues, are inducible by interferons (IFNs) (7, 8) and Toll-like receptor (TLR) ligands (9) and inhibit viral replication (9, 10). Previous studies have demonstrated that SLFN11 inhibits replication of HIV-1 (11) and equine infectious anemia virus (12). SLFN14 is a family member shown to play several fundamental roles, such as platelet formation and function. Mutations of SLFN14 cause a form of inherited thrombocytopenia exhibiting excessive bleeding (13) and defects in platelet secretion (14), leading to the development of atrial septal defect (15). Furthermore, ectopically expressed SLFN14 inhibits DNA and RNA virus replication (9). SLFN14 is thought to lead to a delay in the transfer of influenza virus particles from the cytoplasm to the nucleus and enhancement of retinoic acid-inducible gene I (RIG-I)–mediated IFN-β signaling (9). Although a recent work propose that SLFN14 impairs the expression of HIV-1 transcripts rich in rare codons, in a SLFN11-like manner, the detailed mechanism remains unexplored (16). The positive transcription elongation factor b (P-TEFb) regulates the transcription of genes involved in many physiological and pathological processes. It contains CDK9 and the regulatory subunit CyclinT1 (17). P-TEFb rescues transcription from abortive elongation through CDK9, which hyperphosphorylates the C-terminal domain (CTD) of the largest subunit of RNA polymerase II (Pol II), thus enhancing the processivity of elongation (18). P-TEFb is often targeted to specific promoters by adaptor proteins. For example, the cellular protein Brd4 binds to the active form of P-TEFb and directs it to the transcription elongation machinery (19). Similarly, the HIV-1 Tat protein hijacks the P-TEFb complex to stimulate HIV-1 gene transactivation and viral replication (20). Tat recruits P-TEFb to the transactivation response region (TAR), an RNA stem-loop constituted by the +1 to +59 region of nascent HIV-1 RNA, to hyperphosphorylate the second serine residue of the Pol II CTD. This posttranslational modification, in turn, overcomes the block that causes low level production of full-length transcripts from the long terminal repeat (LTR) promoter due to stalled Pol II after initiation (21). Here, we report that SLFN14 blocks HIV-1 transcription by preventing Tat from recruiting P-TEFb to stimulate viral transcription elongation by decreasing the occupation of phosphorylated Ser2 (Ser2p) by RNA Pol II (RNAPII) on the HIV-1 promoter. We further found that SLFN14 interacts with P-TEFb and suppresses transcription of a subset of cellular genes. These data unravel a critical function of SLFN14 in regulating the transcription of viral and cellular genes, suggesting that SLFN14 should be considered as a potential target for antiviral therapies. RESULTS SLFN14 inhibits HIV-1 production To examine the potential antiviral function of SLFN14, we first assessed the effect of ectopically expressed SLFN14 on the production of HIV-1 and simian immunodeficiency virus (SIV) by transfecting human embryonic kidney (HEK) 293T cells with proviral DNA and a SLFN14 expression vector. We tested different HIV-1 strains, including commonly used strains NL4-3 and BH10, and primary isolates YU-2, 89.6, A/G, and subtype C. The amounts of infectious viruses in the culture supernatants were determined by infecting the TZM-bl cell line, which contains the HIV-1 LTR-Luc reporter cassette. We observed more than 10-fold inhibition of all HIV-1 strains and 2-fold inhibition in viral production for the SIVtan proviral DNA (Fig. 1A). This marked inhibition on HIV-1 was further demonstrated in experiments showing a dose-dependent effect of SLFN14 on both infectious viral particles of NL4-3 and infection of vesicular stomatitis virus (VSV)–G–pseudotyped HIV-1 (single-cycle infection assay) (Fig. 1, B and C), in which SLFN14 was overexpressed in virus-producing and virus-infected cells, respectively. To determine the role of endogenous SLFN14 in restricting HIV-1 infection, we depleted SLFN14 in HIV-1–producing HEK293T cells with vectors expressing SLFN14-specific short hairpin RNA (shRNA; shSLFN14-1 or shSLFN14-2) and then quantified the production of progeny viruses. Knockdown of SLFN14 increased the level of infectious viruses by about 1.5 to 1.8 times (Fig. 1D and fig. S1). Notably, SLFN14 exerted reduced efficacy against VSV-pseudotyped HIV-1 compared to wild-type viruses (Fig. 1, A to C). The mechanistic basis for this differential activity requires further investigation. SLFN14 inhibits HIV-1 replication. (A) Levels of infectious HIV-1and SIV viral particles produced in HEK293T cells transfected with SLFN14-Flag expressing vector and viral DNA constructs, as determined by infection of TZM-bl cells and quantification of firefly luciferase activity (top). Expression of SLFN14 was determined by Western blot (bottom). (B) Levels of infectious wild-type HIV-1NL4-3 viruses from HEK293T cells transfected with viral DNA and various amounts of SLFN14-Flag expressing vector. Levels of SLFN14 were determined by Western blot (bottom). (C) The infectivity of VSV-G–pseudotyped HIV-1NL4-3-E-R- in HEK293T cells transfected with SLFN14-expressing vector determined by luciferase activity in infected cells. Expression of SLFN14 protein was determined by Western blot (bottom). (D) Quantification of the abundance of VSV-G–pseudotyped HIV-1NL4-3-E-R—luc produced by HEK293T cells transfected with either 0.5 μg of shSLFN14 or green fluorescent protein-specific shRNA (shGFP), as determined by infection of SupT1 cells and quantification of firefly luciferase activity (top). The level of endogenous SLFN14 was determined by Western blot (bottom). (E and F) Flow cytometry analysis showing the effect of transiently expressed SLFN14-Flag protein on infection in both PBMCs (E) and Jurkat (F) by VSV-G–pseudotyped HIV-1NL4-3-E-R-. The relative infection rate was calculated from the results of three experiments. (G) The infectivity of VSV-G–pseudotyped HIV-1NL4-3-E-R—luc in Jurkat cells transfected with shSLFN14 or control shGFP vectors determined by luciferase activity in infected cells (top). The level of endogenous SLFN14 was determined by Western blot (bottom). (H) Flow cytometry analysis of p24 positive cell frequency in SLFN14 knockout Jurkat cells. The gRNA1, gRNA2, and gRNA3 represent three gRNAs targeting different sites of SLFN14, respectively. Quantitative data are presented as means ± SD of three independent experiments with statistical significance (n.s., not significant; *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001) calculated using Student’s t test. To confirm this finding in HIV-1–permissive cells, we first investigated the anti-HIV activity of overexpressed SLFN14 in both Jurkat cells and peripheral blood mononuclear cells (PBMCs). The results demonstrated that ectopic expression of SLFN14 markedly diminished the number of HIV-1–positive Jurkat cells and PBMCs, respectively (Fig. 1, E and F, and fig. S1). Furthermore, similar observations were made when SLFN14 was knocked down with the above shRNA-expressing vectors in Jurkat cells, followed by infection of HIV-1 (Fig. 1G and fig. S1). SLFN14 depletion with the CRISPR/guide RNA system leads to more than twofold increase of viral replication (Fig. 1H and fig. S1). These data support the antiviral activity of SLFN14 in both typical human T lymphocyte cell line and primary HIV-1 target cells. Together, we demonstrate the notable inhibitory potential of SLFN14 on HIV-1 infection. SLFN14 negatively regulates HIV-1 RNA synthesis We next sought to identify the step of the HIV-1 replicative cycle, which was inhibited by SLFN14. HEK293T cells were cotransfected with the luciferase-expressing envelope-deficient pNL4-3-Luc.R-E- proviral DNA and expression vectors of VSV-G and Flag-tagged SLFN14. We then quantified protein and mRNA levels of HIV-1 gag gene in transfected cells and determined virus production at 48 hours posttransfection by measuring capsid (p24) levels in the cell supernatant. Results showed that the SLFN14 expression markedly reduced levels of viral protein and RNA (Fig. 2, A and B), virus production (Fig. 2C), and the infectivity of progeny viruses in an equal volume of culture supernatant (Fig. 2D) in a dose-dependent fashion, which suggested that SLFN14 mainly acts at the RNA level. In line with this idea, we found that SLFN14 reduced the levels of luciferase activity and gag mRNA in HIV-infected cells but did not affect the early events of HIV-1 replication, including reverse transcription, nuclear import of viral DNA, and viral DNA integration (fig. S2). Furthermore, we observed a gradual reduction in HIV-1 Vif, Rev, and Vpu mRNA levels, inversely correlating with an increase in SLFN14 levels (Fig. 2E), supporting that SLFN14 functions as a negative regulator of HIV-1 RNA expression. SLFN14 inhibits HIV-1 RNA expression. (A to D) HEK293T cells were cotransfected with pNL4-3-Luc.R-E-, the VSV-G expression vector, and increasing levels of the SLFN14-expressing vector. Levels of viral proteins in transfected HEK293T cells were determined by Western blot (A). Gag mRNA was quantified in HEK293T cells by RT-qPCR and normalized with glyceraldehyde-3-phosphate dehydrogenase (gapdh) mRNA levels (B). HIV-1 p24 levels were measured in supernatants by enzyme-linked immunosorbent assay (C). Infectious viruses in an equal volume of culture supernatant were quantified by measuring luciferase activity in infected SupT1 cells at 48 hours postinfection (D). (E) Levels of different viral RNA were determined by RT-qPCR in HEK293T cells cotransfected with pNL4-3-Luc.R-E- and VSV-G, and SLFN14 expression vectors (right). Levels of SLFN14 and Gag proteins were determined by Western blot (left). (F) Domains of SLFN14 and truncated regions of mutants. (G to I) HEK293T cells were cotransfected with pNL4-3-Luc.R-E- and expression vectors of VSV-G and wild-type SLFN14 or mutated versions. Expression of HIV-1 Gag and SLFN14 was analyzed by Western blot (G). Gag mRNA was quantified in HEK293T cells by RT-qPCR (H). Levels of progeny viruses were determined by infecting SupT1 cells and subsequent quantification of luciferase activity (I). Quantitative data are presented as means ± SD of three independent experiments with statistical significance (n.s., not significant; * P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001) calculated using Student’s t test. To further understand how SLFN14 inhibits HIV-1 production, we tested some SLFN14 deletion mutants. We first generated three Flag-tagged truncated SLFN14 mutants—dC1, dC2, and dN3—which retained amino acids 1 to 571, 1 to 391, or 572 to 912, respectively, to remove the C-terminal RNA helicase motif, SWI/SNF-associated domain-like (SWADL) motif, and N terminus of SLFN14 (Fig. 2F). Only dC1 reduced HIV-1 Gag expression (Fig. 2G), Gag mRNA level (Fig. 2H), and viral infectivity (Fig. 2I), to an even greater degree than that of wild-type SLFN14, indicating a critical role for the SWADL motif in SLFN14-mediated anti–HIV-1 activity, independent of its RNA helicase function. These results further support the conclusion that SLFN14 exerts its anti–HIV-1 activity by suppressing viral gene expression. SLFN14 inhibits the release of transcription pausing at HIV-1 LTR SLFN14 may reduce the level of HIV-1 RNA either by inhibiting the synthesis of viral RNA from the viral LTR or by inducing viral RNA degradation. To test these possibilities, we first examined whether SLFN14 inhibits HIV-1 transcription using a reporter construct, in which luciferase expression is under the control of the HIV-1 LTR promoter and can be stimulated by HIV-1 Tat. SLFN14 significantly inhibited Tat transactivation of the LTR in a dose-dependent manner, with a maximum inhibition of 80% of the control group, while exerting a modest yet significant reduction of 50% on basal LTR-driven transcription (Fig. 3A and fig. S3). In contrast, it showed no effect on the control cytomegalovirus (CMV) promoter (Fig. 3B), suggesting a selective inhibition by SLFN14 of transcription from HIV-1 promoter. In support of this, we further demonstrated that shRNA-mediated SLFN14 depletion (confirmed by quantification of SLFN14 mRNA; Fig. 3C, right) increased HIV-1 LTR activity (Fig. 3C, left), as opposed to results obtained for CMV-driven reporter gene expression (fig. S1D). These data demonstrated that SLFN14 suppresses RNA synthesis from the HIV-1 LTR promoter. SLFN14 inhibits HIV-1 transcription by impeding LTR pausing-release. (A and B) HEK293T cells were cotransfected with the SLFN14-Flag and Tat-Myc expression vectors and HIV-1 LTR-luc (A) or CMV-luc (B) (top) and subsequently assessed for luciferase activity. SLFN14 expression was analyzed in HEK293T cells by Western blot (bottom). (C) LTR activity in SLFN14 knockdown HEK293T cells (left). SLFN14 RNA was quantified by RT-qPCR (right). (D) The effect of wild-type SLFN14 versus its mutants on HIV-1 LTR activity was assessed by measuring luciferase activity in transfected HEK293T cells. (E) HEK293T cells were cotransfected with pNL4-3-Luc.R-E- and increasing amounts of the SLFN14 expression vector. Levels of HIV-1 RNA at the promoter-proximal (1 to 69) and promoter-distal (1726 to 1819) regions were determined by RT-qPCR. Results were normalized with that of the control group and presented as fold reduction. (F) ChIP-qPCR analysis of RNAPII and Ser2p-RNAPII occupation of the HIV-1 promoter-proximal region, in HEK293T cells overexpressing wild-type SLFN14 or mutants. The amount of immunoprecipitated DNA was normalized to the input DNA and expressed as a percentage of the input. Quantitative data are presented as means ± SD of three independent experiments with statistical significance (n.s., not significant; *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001) calculated using Student’s t test. To test whether SLFN14 induces degradation of HIV-1 RNA, we measured the decay kinetics of HIV-1 RNA in SLFN14-expressing cells using actinomycin D to block transcription RNAPII. Similar decay kinetics of HIV-1 RNA was detected with or without SLFN14 expression (fig. S4), although moderately reduced viral RNA levels persisted in SLFN14-expressing cells, suggesting that RNA instability induced by SLFN14 is a contributing factor to the overall reduction in viral RNA. Given the substantially stronger effect on total level of viral mRNA, these results support that SLFN14 reduces HIV-1 RNA abundance primarily through transcriptional suppression. Further, the dC1, but not dC2 and dN3 mutants, profoundly inhibited Tat-mediated LTR transcription (Fig. 3D), supporting that SLFN14 inhibits HIV-1 gene expression at the transcription level. Of note, the dC1 mutant had a much less pronounced effect on basal LTR-driven transcription than wild-type SLFN14 (fig. S5A), suggesting that the C-terminal RNA helicase motif may contribute to the inhibition of HIV-1 basal transcription. HIV-1 transcription consists of two steps, initiation, and elongation. Following transcription initiation, RNAPII shortly pauses downstream of the transcription initiation site. Release of the paused RNAPII and onset of productive transcription elongation are key rate-limiting events to ensure the synthesis of full-length viral RNA. To understand how SLFN14 inhibits HIV-1 transcription, we quantified the viral RNA transcripts generated at regions proximal and distal to the HIV-1 LTR by reverse transcription quantitative polymerase chain reaction (RT-qPCR), which reflects the transcription initiation and elongation steps of HIV-1 transcription, respectively. The result showed that HIV-1 transcription elongation was severely inhibited in SLFN14- and dC1-overexpressing cells, while transcription initiation was much less affected (20- to 100-fold reduction in levels of distal transcripts versus 2- to 4-fold reduction in levels of proximal transcripts upon SLFN14 overexpression; Fig. 3E and fig. S5B). These data suggest that SLFN14 reduces HIV-1 RNA synthesis mainly by blocking transcription elongation. Because efficient transcription elongation depends on the phosphorylation of residues Ser2 (Ser2p) of RNAPII for the release of paused RNA polymerases, we next investigated whether SLFN14 affected this posttranslational modification. First, the presence of RNAPII at proximal and distal regions of the HIV-1 LTR was assessed by chromatin immunoprecipitation (ChIP)–qPCR using antibodies specific for RNAPII or Ser2p-RNAPII. SLFN14 overexpression selectively decreased the occupancy of Ser2p-RNAPII rather than total RNAPII (Fig. 3F). In contrast, the functional-deficient dN3 did not affect Ser2p-RNAPII occupation (Fig. 3F). Furthermore, SLFN14 did not modify levels of RNAPII or Ser2p-RNAPII (fig. S6A), which excludes the possibility that reduced RNAPII occupancy could be attributed to lower RNAPII expression. To further characterize the impact of SLFN14 on transcriptional pausing, we quantified the RNAPII pausing index (PI). The PI serves as an established metric for promoter-proximal RNAPII stalling and quantitatively reflects the extent of polymerase pausing. We performed ChIP assays using anti–Pol II antibodies, followed by qPCR amplification of distal HIV gene regions. SLFN14-expressing cells exhibited a 5.5-fold increase in PI values compared to controls, while dN3-expressing cells showed no significant change (fig. S6, B and C). These results demonstrate that SLFN14 impairs the release of paused RNAPII during HIV-1 proviral transcription. Collectively, these data unfolded the molecular mechanism that SLFN14 specifically inhibits Ser2p-RNAPII occupancy of the HIV-1 LTR, thus impairing the release of paused RNAPII and the subsequent transcription elongation. SLFN14 blocks the interaction between Tat and P-TEFb It is known that HIV-1 Tat binds to and recruits P-TEFb to the HIV-1 TAR RNA stem-loop to catalyze Ser2 phosphorylation of RNAPII and then to release transcription pausing, a critical regulatory step in HIV-1 gene expression. We thus examined whether SLFN14 affects the recruitment of P-TEFb by HIV-1 Tat. Western blot analyses of SLFN14-overexpressing HEK293T cells first showed that SLFN14 did not affect the expression of CDK9 and CyclinT1 (two components of P-TEFb) or their interaction (Fig. 4A), suggesting no effect of SLFN14 on the formation of P-TEFb complex. In contrast, SLFN14 expression led to an inhibition of the association of Tat with CDK9 by more than 10-fold compared to the control group (Fig. 4, A and B), which demonstrates potent SLFN14-dependent inhibition of the recruitment of P-TEFb by HIV-1 Tat. SLFN14 blocks the interaction between Tat and P-TEFb. (A) Interaction between CDK9 and Tat in HEK293T cells transfected with increasing SLFN14 DNA, as determined by coimmunoprecipitation (co-IP). (B) Analyses of the ratios of immunoprecipitated Tat and CyclinT1 in the co-IP assay are shown in (A). (C) Effect of wild-type SLFN14 and dC2 mutant on the interaction between CDK9 and Tat in HEK293T cells treated with MG132 for 6 hours. (D) Analyses of the ratios of immunoprecipitated Tat and CyclinT1 in the co-IP assay shown in (E) Confocal microscopy images from PLA assay performed in HeLa cells expressing wild-type SLFN14 or various mutants with antibodies specific to Tat and CDK9. The interaction between Tat and CDK9 was examined with PLA and detected as red puncta in cells. Nontransfected cells were used as a negative control. Nuclei were stained with 4′,6-diamidino-2-phenylindole (DAPI; blue). (F) The number of Tat and CDK9 interaction signals was quantified (in E) and is expressed per cell. Results from 15 cells for each condition are presented. Quantitative data are presented as means ± SD of three independent experiments with statistical significance (n.s., not significant; *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001) calculated using Student’s t test. We noticed that the level of Tat gradually reduced concurrent with increased expression of SLFN14 (Fig. 4A). This raised the possibility that the lower amount of Tat associated with P-TEFb may be due to a decrease in the level of Tat in the presence of SLFN14. Previous work showed that disrupting the interaction between Tat and P-TEFb caused a proteasome-dependent degradation of free Tat (22). Consistent with this report, we also found that treatment with the proteasome inhibitor MG132 notably elevated the level of Tat in SLFN14-expressing cells (fig. S7). Although the presence of similar levels of Tat in MG132-treated cells with or without SLFN14 expression (Fig. 4C), the amount of Tat associated with CDK9 was still reduced in cells expressing SLFN14, but not in cells expressing mutant dC2 (Fig. 4, C and D). These data demonstrated that SLFN14 inhibits the interaction of P-TEFb and HIV-1 Tat and that free Tat undergoes degradation by proteasomes. Notably, the incomplete restoration of Tat levels by MG132 in SLFN14-overexpressing cells (Fig. 4C and fig. S7) suggests that additional mechanisms beyond proteasomal degradation contribute to Tat reduction, warranting further investigation. We next examined the effect of SLFN14 on the interaction of P-TEFb and Tat by performing the in situ proximity ligation assay (PLA) (Fig. 4E). In this assay, protein interaction is visualized as an individual fluorescent dot. As shown in Fig. 4E, red fluorescent dots representing the interaction between P-TEFb and Tat were only found in Tat-expressing HEK293T cells, but not in cells without Tat. Upon expression of either SLFN14 or dC1, the number of fluorescent dots markedly reduced, suggesting that the interaction between P-TEFb and Tat was disrupted, while mutant dC2 (or dN3) showed no effect (Fig. 4E). The quantification analysis of the PLA data revealed a fivefold reduction in Tat and P-TEFb interaction caused by SLFN14 (Fig. 4F). Together, these data demonstrated that SLFN14 disrupts the interaction between P-TEFb and Tat, which blocks the RNAPII pausing release and subsequent transcription elongation. SLFN14 interacts with the P-TEFb complex We next investigated how SLFN14 inhibits the interaction between P-TEFb and Tat. In the above experiments, SLFN14 was detected in immunoprecipitates using anti-CDK9 antibodies, and Tat and SLFN14 abundance was inversely correlated (Fig. 4A). This suggests that SLFN14 could bind and prevent P-TEFb from being recruited by Tat. Confirming this scenario, both CyclinT1 and CDK9 were detected in extracts from HEK293T cells expressing exogenous SLFN14-Flag immunoprecipitated with anti-Flag antibodies (Fig. 5A). In addition, we did not observe SLFN14 association with other transcription complexes such as CyclinH/CDK7 (Fig. 5A), which argued for selectivity in the binding of SLFN14 to P-TEFb. The association of SLFN14 with P-TEFb was further supported by PLA experiments showing that SLFN14 interacted with endogenous CDK9 in the cytoplasm and the nucleus (Fig. 5B). The interaction between endogenous SLFN14 and P-TEFb was subsequently confirmed in both Jurkat and HEK293T cells by coimmunoprecipitation (co-IP) with anti-CDK9 antibodies (Fig. 5C). Together, these data demonstrated an interaction of SLFN14 with P-TEFb. SLFN14 interacts with the P-TEFb complex. (A) Co-IP assay was performed to detect the interaction of overexpressed wild-type SLFN14 or mutants with endogenous CDK9, CyclinT1, CDK7, and CyclinH in HEK293T cells. (B) PLA assay were performed in HeLa cells expressing SLFN14 with antibodies specific to SLFN14 and CDK9. Control cells were transfected with an empty vector. Nuclei were stained with DAPI (blue). The association of SLFN14 with CDK9 was observed as red puncta in cells. (C) Co-IP assay was performed to detect the interaction between endogenous SLFN14 and CDK9 in Jurkat (left) and HEK293T (right) cells. (D) In vitro analyses of the binding of purified SLFN14 to the P-TEFb complex and isolated CDK9 or CyclinT1 using the Octet assay. (E and F) Co-IP analyses of the interaction of SLFN14 with CyclinT1 and CDK9 following knockdown of CDK9 (E) or CyclinT1 (F) in HEK293T cells. (G) SLFN14 distribution of fractions from a 5 to 30% glycerol gradient of HEK293T cell lysates measured by Western blot. (H) Co-IP assay was performed to detect the interaction between CDK9 and SLFN14 in HEK293T cells with ribonuclease (RNAse) treatment. (I) Structural model of the SLFN14–P-TEFb (CDK9/CyclinT1) complex predicted using AlphaFold 3. (J) Structural model of the SLFN14–P-TEFb (CDK9/CyclinT1)–HEXIM1–7SK snRNA complex predicted using AlphaFold 3. SLFN14 is shown with the dC1 region highlighted in red and the dC2 region highlighted in salmon. CDK9 is colored green, CyclinT1 is colored in blue, and the HEXIM1 dimer is represented in yellow and magenta, respectively. The 7SK snRNA is shown in gray. The structure was visualized using the PyMOL Molecular Graphics System (version 3.1.4). Quantitative data are presented as means ± SD of three independent experiments with statistical significance (***P ≤ 0.001) calculated using Student’s t test. Consequently, the association of SLFN14 with P-TEFb in vitro was characterized using the Octet technique with purified recombinant proteins. Binding responses [nanometer‌ (nm) shift on the y axis] reflect the mass of proteins bound to the sensor surface and thus indicate complex formation capacity. Results showed stronger binding to pre-assembled P-TEFb compared with isolated CDK9 or CyclinT1 (Fig. 5D), demonstrating preferential association of SLFN14 with the intact complex. Next, the co-IP experiments were conducted in HEK293T cells, in which endogenous CDK9 or CyclinT1 was individually silenced. Upon immunoprecipitation of extracts from transfected HEK293T cells with either anti-CDK9 or anti-CyclinT1 antibodies, SLFN14 was almost completely diminished when the other component of the P-TEFb complex was depleted (Fig. 5, E and F). These data indicated that SLFN14 binds to the P-TEFb complex rather than one of its individual components, CDK9 or CyclinT1. As opposed to full-length SLFN14 and dC1, dC2 and dN3 mutants did not interact with P-TEFb (Fig. 5A). Similar results were obtained in co-IP experiments using anti-CDK9 antibodies (fig. S9). On the basis of the data from these mutants, we thus proposed that the SWADL domain of SLFN14 could be playing an essential role by facilitating its interaction with P-TEFb (Fig. 2F). This conclusion is consistent with SLFN14 domains functional analysis (Fig. 2H). Thus, as truncated mutants dC2 and dN3 were unable to interact with P-TEFb and could not inhibit HIV-1 transcription, our data strongly argued that SLFN14 exerted its anti–HIV-1 activity by its interaction with P-TEFb. P-TEFb exists in two forms, the free active form and the 7SK small nuclear ribonucleoprotein (snRNP)–bound inactive form. To determine which form of P-TEFb SLFN14 binds to, we first performed a glycerol gradient sedimentation analysis of cell lysates. We thus observed two distinct pools of endogenous P-TEFb, the light active form at the top of the gradient and the heavy inactive form at the bottom of the gradient (Fig. 5G). As expected, HEXIM1, a component of 7SK snRNP, was mainly found associate to inactive P-TEFb fractions. SLFN14 was only detected in fractions containing the heavy form of the P-TEFb complex, suggesting that SLFN14 predominantly binds to the inactive P-TEFb complex (Fig. 5G). In support to these results, SLFN14 could be immunoprecipitated from HEK293T cell extracts with an antibody specific for HEXIM1, a component of inactive complex. To further test the association of SLFN14 with the inactive form of P-TEFb, cell lysates were treated with ribonuclease to disassemble the 7SK snRNP complex, and, upon co-IP with anti-CDK9 antibodies, only modest levels of SLFN4 were detected (Fig. 5H), which further lent support to its main association to the inactive P-TEFb complex. These data suggest that SLFN14 acts as a bona fide anti–P-TEFb factor, inhibiting activation and recruitment of P-TEFb by HIV-1 Tat. To further characterize the interaction between SLFN14 and P-TEFb, we used AlphaFold 3 to model SLFN14 interactions within two complexes: SLFN14/CDK9/CyclinT1 (Fig. 5I) and SLFN14/CDK9/CyclinT1/7SK RNA/HEXIM1 (Fig. 5J). In the active P-TEFb complex, the N-terminal domain (residues 1 to 391) of SLFN14 engages both CDK9 and CyclinT1 (Fig. 5I). This aligns with our in vitro and cellular data showing that SLFN14 binds the P-TEFb complex, but not individual CDK9 or CyclinT1 subunits (Fig. 5, A to E). These interactions likely prevent Tat binding to the active P-TEFb, thereby inhibiting P-TEFb recruitment by Tat. In the inactive P-TEFb complex, the N-terminal domain of SLFN14 preferentially engages 7SK RNA via its canonical RNA-binding region (Fig. 5J), consistent with our observation that SLFN14 mainly associates with the inactive P-TEFb complex in cells (Fig. 5, G and H). Critically, when bound to the 7SK ribonucleoprotein complex, SLFN14 appears to enhance the stability of the inhibitory complex composed of P-TEFb (CDK9/CyclinT1), HEXIM1, and 7SK snRNA, thereby maintaining P-TEFb in an inactive state. Consequently, Tat is unable to effectively displace HEXIM1 to induce the activation of P-TEFb. Collectively, these results support the dual inhibitory mechanism of SLFN14 on inhibiting Tat-mediated activation and recruitment of P-TEFb, consequently impairing viral transcription and replication. SLFN14 expression is stimulated by TLR3 agonists The potent inhibition of HIV-1 expression by SLFN14 proposes a possible role of SLFN14 in host innate immunity, including IFN-mediated immune response mediated by IFN-stimulated genes (23). However, we did not detect any change in the expression of SLFN14 after various IFN treatments (fig. S10, A to D). However, polyinosinic polycytidylic acid [poly(I:C)] treatment, which activates the TLR3 signaling pathway, was observed to up-regulate the level of SLFN14 mRNA by more than sixfold in HeLa cells and induction comparable to the one observed for the IFN-β gene (Fig. 6A). Induced expression of SLFN14 was also detected at the protein level (Fig. 6B). An increase in SLFN14 mRNA expression following poly(I:C) treatment was also confirmed in other cell types, including a fourfold increase in PBMCs (fig. S10E). Our results thus indicated that SLFN14 expression responds to the stimulation by TLR3 agonists. SLFN14 contributes to anti-HIV innate immunity induced by the TLR3 agonist poly(I:C). (A) RT-qPCR analyses of IFN-β and SLFN14 mRNA in HeLa cells after treatment with different amounts of poly(I:C), determined by. (B) Western blot analysis of SLFN14 in HeLa cells following poly(I:C) treatment. SLFN14 levels were analyzed using ImageJ and plot the data, with results representing three independent experiments. (C and D) SLFN14 was knocked down in HeLa cells by transfection shRNA expression vector. After poly(I:C) treatment, HeLa cells were infected with VSV-G–pseudotyped HIV-1NL4-3-R-E- for 120 hours. Gag mRNA was quantified by RT-qPCR and is expressed as a percentage of nontreated cells (C). Viral infectivity was further assessed by luciferase activity (D). Quantitative data are presented as means ± SD of three independent experiments with statistical significance (*P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001) calculated using Student’s t test. IgG, immunoglobulin G. TLR3 recognizes double-stranded RNA and triggers multiple intracellular events that are responsible for innate antiviral response in macrophages. As activation of TLR3 by poly(I:C) can stimulate IFN (24) and apolipoprotein B mRNA editing enzyme catalytic polypeptide 3G (APOBEC3G) (25) expression, we next tested the potential implication of SLFN14 in the anti–HIV-1 response induced by TLR3 activation. HeLa cells were first transfected with a specific shRNA-expressing vector to silence SLFN14, then treated with poly(I:C), and infected with VSV-G–pseudotyped luciferase-expressing NL4-3 viruses. Results demonstrated that both levels of viral mRNA and luciferase activity increased in cells silenced for SLFN14 expression, suggesting that SLFN14 knockdown alleviated poly(I:C)-mediated inhibition of HIV-1 infection (Fig. 6, C and D). Hence, these results implied that SLFN14 contributed to the anti–HIV-1 response dependent on TLR3 activation. SLFN14 suppresses the transcription of a subset of cellular genes Efficient transcription of specific cellular genes such as interleukin-8 (IL-8), carbamoyl-phosphate synthetase 2, aspartate transcarbamoylase, dihydroorotase (CAD), and major histocompatibility complex class II (MHC II) depends on the recruitment of P-TEFb to their promoters by transactivators RelA, Myc, and class II major histocompatibility complex transactivator (CIITA), respectively (18). In a similar manner to HIV-1 Tat, SLFN14 might inhibit the expression of these genes by negatively acting on P-TEFb. We observed that SLFN14 expression profoundly reduced the amounts of RelA, Myc, and CIITA coimmunoprecipitated with CDK9, indicating potent impairment of their association with P-TEFb (fig. S11A). Notably, SLFN14 expression reduced cellular abundances of Myc and CIITA (fig. S11), potentially explaining their decreased recovery in immunoprecipitates. However, IL-8, CAD, and MHC II expression was only moderately decreased in SLFN14-expressing HEK293T cells (fig. S11, B and C). One possibility is that the basal expression level of these genes may not depend on P-TEFb, thereby not being sensitive to SLFN14. To reveal this possibility, IL-8 expression was induced with tumor necrosis factor–α (TNF-α) in SLFN14-expressing cells as previously described. Results showed that SLFN14 suppressed TNF-α–induced IL-8 expression (Fig. 7A). In addition, we observed that PSA expression, which highly depends on the recruitment of P-TEFb by the androgen receptor when stimulated with dihydrotestosterone (26), was reduced by 10 folds in the presence of SLFN14 (fig. S11D). Together, these data suggested that SLFN14 also negatively modulates P-TEFb–sensitive genes’ expression in induced conditions. SLFN14 inhibits the transcription of a subset of cellular genes. (A) RT-qPCR analyses of IL-8 expression in HEK293T cells transfected with a SLFN14 expression vector and treated with TNF-α for 1 hour. (B) Volcano plot of genes differentially expressed between HeLa cells transfected with a SLFN14 expression vector and the empty vector. FC, fold change. (C) KEGG analysis of pathways significantly down-regulated by SLFN14 in transfected HeLa cells. (D) Promoter-proximal and promoter-distal regions of selected genes were analyzed by RT-qPCR in SLF14-overexpressing HeLa cells and normalized to mRNA signals of corresponding regions from control HeLa cells. (E) ChIP-qPCR analysis of transfected HeLa cells of promoter-proximal regions of selected genes using antibodies specific to RNAPII and Ser2p-RNAPII. The amount of immunoprecipitated DNA was normalized to the input DNA, and relative percentages were presented. (F and G) Quantification of mRNA levels of selected genes in HeLa cells transfected with empty vector (F) or the shSLFN14 expressing vector (G) and treated with poly(I:C). Experimental replicates independent experiments are now specified in figure legends. Quantitative data are presented as means ± SD with statistical significance (n.s., not significant; *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001) calculated using Student’s t test. To obtain a broad view of how cellular genes respond to SLFN14 overexpression, we conducted a transcriptomic analysis of HeLa cells in which SLFN14 was overexpressed. A total of 682 genes were shown to be differentially expressed [log2(fold changes) > 1.8 or < −1.8 and P < 0.5] (Fig. 7B). Among these genes, 96% (656) were down-regulated by SLFN14 (dataset S1). Gene set enrichment analyses [Kyoto Encyclopedia of Genes and Genomes (KEGG)] determined that these down-regulated genes were enriched in pathways mainly involved in the cell cycle and protein processing in the endoplasmic reticulum (Fig. 7C). We next selected 20 genes (see the Supplementary Materials) from the top 40 genes showing the highest levels of down-regulation and assessed the effect of SLFN14 on their transcription initiation and elongation by RT-qPCR quantification. Results showed that SLFN14 inhibited transcription elongation of nine genes (ZMAT1, CP, PLA2G7, TMED7, ZNF441, ZNF501, ZNF658, EAF2, and ZNF181) to a different extent (from 30 to 90% reduction), without significantly affecting transcription initiation (Fig. 7D). This suggests that SLFN14 suppresses the release of paused RNAPII for the expression of these genes and thus the productive transcription elongation at the basal level. In agreement with this conclusion, ChIP-qPCR analyses revealed that SLFN14 significantly decreased the occupancy of Ser2p-RNAPII, but not of total RNAPII (Fig. 7E) for at least four genes—CP, ZNF441, EAF2, and PLA2G7—accompanied by >2-fold increases in PI values (fig. S12). This indicated that SLFN14 inhibits transcription elongation by reducing the level of gene-associated Ser2p-RNAPII. Because TLR3 agonist poly(I:C) stimulated the expression of SLFN14 (Fig. 6), we thus examined the effect of induced endogenous SLFN14 on the transcription of the above genes. Results of RT-qPCR analyses showed that poly(I:C) significantly reduced mRNA levels of CP, ZNF441, EAF2, and PLA2G7 (Fig. 7F) and that this inhibition was undetectable when SLFN14 expression was blocked by shRNA (Fig. 7G). Hence, these data demonstrated that SLFN14 is a crucial negative regulator of the transcription of a subset of cellular genes. DISCUSSION A number of host proteins have been shown to restrict HIV-1 infection. These include APOBEC3G, TRIM5α, BST-2, MxB, SAMHD1, and SERINC5 (27), which constitute critical players of antiviral innate immunity and operate via distinct molecular mechanisms. Our study suggests that SLFN14 is a host restriction factor that inhibits the function of HIV-1 Tat protein. Tat is indispensable for HIV-1 gene expression and acts by recruiting P-TEFb to the TAR motif of viral RNA and markedly stimulating viral transcription. However, none of the host factors have yet been reported to target the recruitment of P-TEFb by Tat. In this study, SLFN14 was shown to interact with P-TEFb and blocked its association with HIV-1 Tat, thus impeding the elongation of viral transcription and halting subsequent HIV-1 replication. Furthermore, we also found that SLFN14 was induced by the TLR3 agonist poly(I:C), but not by IFN, contributing to the inhibition of HIV-1 replication by poly(I:C). Our results thus uncover a critical cellular restriction mechanism that inhibits HIV-1 replication at the step of viral gene transcription. The anti–HIV-1 activity of SLFN14 was shown to be effective against different HIV-1 strains, including primary HIV-1 isolates. SLFN14 inhibitory properties were further confirmed in different cell lines and PBMCs. Poly(I:C)-induced SLFN14 expression also effectively suppressed HIV-1 production. Our study also revealed that SLFN14 inhibits HIV-1 RNA expression. SLFN14 was recently reported as a ribosome-associated endoribonuclease, which participates in ribosomal RNA and ribosome-bound mRNA degradation in particular cell type (28). However, in our study, SLFN14 did not change the decay kinetic of HIV-1 RNA, which argues against the involvement of endoribonuclease activity of SLFN14 in changes in HIV-1 RNA level. Instead, our data suggest that SLFN14 inhibits HIV-1 transcription at the elongation step. On the basis of our co-IP and PLA analyses, we propose that SLFN14 disrupts the interaction between P-TEFb and Tat, thus averting the RNAPII pausing release and blocking efficient elongation of HIV-1 transcription. SLFN14 overexpression has been reported to inhibit the replication of varicella-zoster virus and influenza virus (9), although these studies did not further investigate the exact mechanism underlying this inhibition. P-TEFb has been shown to act as an adapter to mediate the interaction of viral ribonucleoprotein (vRNP) and RNAPII and promote transcription of influenza viral genes (29). Furthermore, the transmembrane glycoprotein of the varicella-zoster virus was reported to be phosphorylated by P-TEFb, which may create signals for internalization and trafficking of viral glycoproteins (30). These reports thus raised the possibility that SLFN14 may suppress the infection of varicella-zoster virus and influenza virus by disrupting the function of P-TEFb. Our data suggest that SLFN14 interacts with P-TEFb and disrupts the association between Tat and P-TEFb, thus inhibiting HIV-1 transcription. It remains to be defined how SLFN14 binding to P-TEFb blocks Tat interaction with P-TEFb. Previous work has shown that phosphorylation of CDK9 regulates P-TEFb activity and function. Specifically, phosphorylation of amino acids Thr186 or Ser175 of CDK9 is essential for assembly of functional P-TEFb and modulates the competitive binding of Tat to P-TEFb (31, 32). We did not observe any changes in the phosphorylation of CDK9 at Thr186 upon overexpression of SLFN14 (fig. S13), suggesting that SLFN14 may not affect CDK9 phosphorylation. Because we could not detect phosphorylation of CDK9 at Ser175, we cannot exclude the possibility that SLFN14 may affect the phosphorylation of this residue. Given that SLFN14 impairs the binding of P-TEFb to several different transactivators, including RelA, Myc, CIITA, and HIV-1 Tat (fig. S11), it is likely that SLFN14 association with P-TEFb physically prevents their interaction. Furthermore, our data demonstrate that SLFN14 mainly associates with the inactive P-TEFb complex in cells (Fig. 5), which may stabilize the inactive P-TEFb/7SK snRNP complex and impede its dissociation induced by HIV-1 Tat and other transactivators. Not unexpectedly, in addition to suppressing HIV-1 transcription, SLFN14 also inhibited the transcription of a subset of cellular genes, which are enriched in pathways involved in the cell cycle and protein processing in the endoplasmic reticulum (Fig. 7, B and C). Similar outcomes were described for the SLFN14-mediated inhibition of viral and cellular gene transcription, such as the inhibition of transcription elongation (Fig. 7D), decreased occupancy of Ser2p-RNAPII (Figs. 3 and 7E), and blockade of P-TEFb recruitment to promoters. In agreement with the role of P-TEFb in ensuring a high level of gene transcription, SLFN14 did not affect the basal expression of P-TEFb–sensitive genes, including IL-8, CAD, and MHC II, but somewhat strongly inhibited induced expression of genes, such as IL-8 and PSA. SLFN14 thus regulates the response of a group of genes to external stimuli by controlling the availability of active P-TEFb. SLFN14 has been reported to play a fundamental role in platelet formation and function (13, 14). The heterozygous missense mutations in SLFN14 were linked to inherited thrombocytopenias, a group of disorders characterized by a low platelet count and frequent excessive bleeding (33). Studies suggest that SLFN14 may be involved in ribosome degradation during platelet formation and maturation. Expression of the mutated SLFN14 in patients and in vitro is markedly reduced as a result of posttranslational degradation, which contributes to defective megakaryocyte maturation and platelet dysfunction (34). In agreement with the role of SLFN14 in platelet function, results of our KEGG analysis revealed that SLFN14 down-regulates the expression of genes involved in platelet activation, which includes PLA2G7, PRKACB, RAP1A, and PLA2G4A (dataset S2). This finding suggests that SLFN14 may regulate platelet formation and maturation by controlling the transcription of genes involved in platelet activation. In support of this possibility, SFLN14 mutations found in patients suffering from inherited thrombocytopenias, i.e., K218E, K219N, V220D, and R223W, are all part of the N terminus of SLFN14, which is required for SLFN14 association with P-TEFb. Further investigation is needed to test whether SLFN14 regulates platelet formation and maturation through the modulation of the activity of P-TEFb. In summary, our data demonstrate that SLFN14 functions as an inhibitor of P-TEFb, leading to the repression of HIV-1 and cellular genes expression. We have thus unfolded a critical cellular mechanism that SLFN14 regulates the function of the transcription activator complex P-TEFb. This finding not only advances our understanding of the regulation of HIV-1 transcription and SLFN14-related genetic diseases but also suggests avenues to develop therapeutics to treat viral infections and other related diseases. MATERIALS AND METHODS Cell culture, transfection, and treatments HEK293T, HeLa, and TZM-bl cells (Table 1) were cultured in Dulbecco’s modified Eagle’s medium (Gibco) supplemented with 10% fetal bovine serum (FBS) (Gibco). SupT1, Jurkat, and THP-1 cells were maintained in RPMI 1640 (Gibco) containing 10% FBS. PBMCs were purchased from Beijing Reader Biotechnology Co. Ltd. and were isolated from buffy coats of healthy volunteers using density gradient centrifugation on Lymphoprep. PBMCs were cultured in complete RPMI medium containing 20% FBS, streptomycin (100 mg/ml), penicillin (100 U/ml), concanavalin A (50 μg/ml; TargetMol), and 1:1000 recombinant human IL-2 (R&D Systems) for 3 days and were transfected using the Amaxa P3 Primary Cell 4D-Nucleofector X Kit. Transfection of 293T cells was performed using Lipofectamine 2000 (Invitrogen) according to the manufacturer’s instructions. HeLa and TZM-bl cells were transfected with Transeasy (Invitrogen) according to the manufacturer’s instructions. Transfection of SupT1, Jurkat, and THP-1 was performed with the Neon transfection Kit according to manufacturer’s instructions. Lipofectamine RNAiMAX reagent (Invitrogen) was used to transfect small interfering RNA (siRNA) (Table 1). HEK293T cells were reverse transfected with the indicated siRNAs (20 nM) for 24 hours, followed by plasmid DNA transfection according to the manufacturer’s instructions. To analyze protein stability and degradation, cells were treated with proteasome inhibitor MG132 (10 μM, Sigma-Aldrich) for 6 hours and harvested at the indicated time. Table 1.Key resources.NA, not applicable; ELISA, enzyme-linked immunosorbent assay; DAPI, 4′,6-diamidino-2-phenylindole; RNase H, ribonuclease H; DMSO, dimethyl sulfoxide; GST, glutathione S-transferase; ATCC, American Type Culture Collection.Reagent or resourceSourceIdentifier Antibodies Anti-SLFN14AbcamCatalog no.ab106406RRID: AB_10890814Monoclonal ANTI-FLAG M2 antibodySigma-AldrichCatalog no. F1804RRID: AB_262044Anti-Myc tagAbcamCatalog no. ab9132RRID: AB_307033Anti-HA tagAbcamCatalog no. ab9110RRID: AB_307019Anti–RNA polymerase II CTD repeat YSPTSPS antibodyAbcamCatalog no. ab5095RRID: AB_304749Anti–RNA polymerase II CTD repeat YSPTSPSAbcamCatalog no. ab26721RRID: AB_777726Anti-HEXIM1AbcamCatalog no. ab25388RRID: AB_2233058Antihuman CDK9AbcamCatalog no. ab6544RRID: AB_305557Antimouse monoclonal (IMD-26) to Cdk7AbcamCatalog no. ab130001RRID: AB_11155158Anti-CyclinT1 (EPR17982)AbcamCatalog no. ab184703RRID: AB_2814653Anti-CyclinHCell Signaling TechnologyCatalog no. 2927RRID: AB_2259936Anti-CADSigma-AldrichCatalog no. HPA057266RRID: AB_2683387Anti–MHC IIAbcamCatalog no. ab23206RRID: AB_2042341Anti–IL-8 (807)AbcamCatalog no. ab18672RRID: AB_444617Anti–β-actinAbcamCatalog no.ab8227RRID: AB_2305186Anti-tubulinAbcamCatalog no. ab56676RRID: AB_945996Anti-p24This paperNAAnti–HIV-1 Tat monoclonal (15.1)NIH Aids Reagent ProgramCatalog no. 1974Anti–CDK9-186p(31)NA Bacterial and virus strains BL21 (DE3) chemically competent cellTransGen BiotechCatalog no. CD601-03 Chemicals, peptides, and recombinant proteins Poly(I:C)Sigma-AldrichCatalog no. P1530DAPIThermo Fisher ScientificCatalog no. D1306Human IFN-αPeproTechCatalog no. 300-01A-50Human IFN-βPeproTechCatalog no. 100-280Human IFN-γPeproTechCatalog no. 300-02-20Human TNF-αPeproTechCatalog no. 300-01ARNase HThermo Fisher ScientificCatalog no. 18021071MG132AbsinCatalog no. abs47025245bActinomycin DAbsinCatalog no. abs810015aLipofectamine 2000 transfection reagentThermo Fisher ScientificCatalog no. 11668019TRIzol reagentThermo Fisher ScientificCatalog no. 15596018HisTrap excel GE HealthcareCatalog no. 17-3712-05DMSOSigma-AldrichCatalog no. D2650Triton X-100Carl RothCatalog no. 3051.4IL-2R&D SystemsCatalog no. 202-IL-010Concanavalin ATargetMolCatalog no. TM-T4004EZ-Link NHS-Biotin reagentThermo Fisher ScientificCatalog no. 21343P-TEFb-HisThermo Fisher ScientificCatalog no. PV4131CDK9-GSTAbcamCatalog no. ab85603 Cell lines HEK293T cellsATCCCatalog no. CRL-11268HeLa cellsATCCCatalog no. CCL-2TZM-bl cellsNIH Aids Reagent ProgramCatalog no. 8129SupT1 cellsATCCCatalog no. CRL-1942Human adult purified PBMC-cryopreserved vialBeijing Reader Biotechnology Co. Ltd.Catalog no. PB010COligonucleotidessiCDK9: 5′-GGUGCUGAUGGAAAACGAGTT-3′This paperNAsiCyclinT1: 5′-CCCACATACTCATGTAGTATT-3′This paperNAsiTLR3: 5′-GAACUAAAGAUCAUCGATT-3′This paperNAPrimer sequencesSee table S2NA Recombinant DNA HIV-1 NL4-3 infectious molecular clone (pNL4-3)NIH Aids Reagent ProgramCatalog no. 114HIV-1 YU2 infectious molecular cloneNIH Aids Reagent ProgramCatalog no. 1350HIV-1 89.6 infectious molecular clone (p89.6)NIH Aids Reagent ProgramCatalog no. 3552HIV-1 BH10 noninfectious molecular clone (pBH10)NIH Aids Reagent ProgramCatalog no. 90HIV-1 NL4-3 ΔEnv Vpr luciferase reporter vector (pNL4-3-Luc.R-E-)NIH Aids Reagent ProgramCatalog no. 3418HIV-2ROD phageNIH Aids Reagent ProgramCatalog no. 207SIVcpz TAN1.910 infectious molecular cloneNIH Aids Reagent ProgramCatalog no. 11496SLFN14-FlagThis paperNASLFN14-truncationsThis paperNAshSLFN14Sigma-AldrichNATat-FlagG. Gao, Chinese Academy of SciencesNA pCMV-luc G. Gao, Chinese Academy of SciencesNANF-kB p65 (RELA) (NM_021975) Human-tagged ORF cloneOriGeneCatalog no. RC220780c-Myc (MYC) (NM_002467) human-tagged ORF cloneOriGeneCatalog no. RC201611CIITA (NM_000246) human-tagged ORF cloneOriGeneCatalog no. RC222253 Software and algorithms GraphPad Prism 5GraphPad softwareNAImageJNational Institutes of Health https://imagej.nih.gov/ij/ FlowJoBD www.flowjo.com/ Other RNA-seq analysesOebiotechNAELISAWanTai BioPharmCatalog no. HIV-0896PrimeScript RT Master KitTakaraCatalog no. RR036ASE Cell Line 4D X Kit L (24 RCT)LonzaCatalog no. V4XC-1024Duolink In Situ Red Starter Kit mouse/rabbitSigma-AldrichCatalog no. DUO9210117-371EZ ChIP KitMilliporeCatalog no. 17-371 Key resources. NA, not applicable; ELISA, enzyme-linked immunosorbent assay; DAPI, 4′,6-diamidino-2-phenylindole; RNase H, ribonuclease H; DMSO, dimethyl sulfoxide; GST, glutathione S-transferase; ATCC, American Type Culture Collection. Plasmid DNA constructs The lentiCRISPR(v2) plasmid DNA was ordered from Addgene (no. 52961). The SLFN14 single guide RNA (sgRNA) fragments were cloned into lentiCRISPR(v2) as described (35). The SLFN14 gRNA target sites are shown in the following: sgRNA1, TCTACTATTACCTCAGGATA; sgRNA2, TACTGAAATGCCGTATCCTG; and sgRNA3, TTAACAGTGCACATATAGCC. The SLFN14 overexpression plasmid pQCXIP-SLFN14 was constructed by inserting the SLFN14 gene into Not I/Mlu I sites of the pQCXIP vector. Lentivirus and retrovirus packaging and transduction The lentiCRISPR(v2) DNA, with or without the SLFN14 sgRNA sequences, was cotransfected into HEK293T cells together with psPAX2 (Addgene, no. 12260) and pVSV-G (Addgene, no. 138479). At 48 hours after transfection, viruses in the culture supernatants were harvested to infect Jurkat cells via spin infection with polybrene (10 μg/ml) for 45 min at 1800 rpm, room temperature. For packaging retroviruses carrying the SLFN14 gene, the pQCXIP-SLFN14 plasmid was cotransfected with pMLV-gag/pol and pVSV-G into HEK293T cells to produce retrovirus. HIV-1 infection At 48 hours after transducing Jurkat cells, cells were infected with HIV-1 equivalent to 100 ng of p24. The inoculated viruses were washed off 16 hours postinfection. After 48 hours of infection, cells were collected for the following assay. The percentage of HIV-1–infected Jurkat cells was determined by immunostaining of viral p24 using fluorescein isothiocyanate–conjugated anti-p24 antibodies (1:200 dilution; catalog no. 6604665, Beckman Coulter), followed by flow cytometry analysis. The SLFN14 expression level was measured by Western blotting. Single-cycle viral infection assay HIV-Luc reporter pseudotyped particles were produced in HEK293T cells by transfecting a VSV-G expression vector and pNL4-3-Luc.R-E-. Supernatants were passed through 0.45-μm filters and stored at −80°C. For single-round infection assays, 5 × 105 SupT1 cells/ml were infected with pseudotyped viruses at a volume ratio of 20:1. At 48 hours postinfection, cells were lysed with cell culture lysis buffer (Promega) for 20 min at 37°C. The lysate (5 μl) was mixed with 40 μl of substrate, and luciferase activity was measured using the 960 luminometer (Berthold Technologies). Western blots Cells were lysed in a buffer containing 1% SDS, and protein extracts were subjected to SDS–polyacrylamide gel electrophoresis and then transferred onto nitrocellulose membranes. Membranes were blocked for 1 hour in tris-buffered saline containing 0.1% Tween [phosphate-buffered saline (PBS)] and 5% nonfat dried skim milk, followed by incubation for 2 hours at room temperature or overnight at 4°C with primary antibodies against proteins (Table 1). After extensive washing with PBS, membranes were incubated for 1 hour with an appropriate horseradish peroxidase–conjugated secondary antibody. Antibody binding was detected using an enhanced chemiluminescence system (GE Healthcare). All Western blotting (WB) experiments were performed in three independent replicates, and quantification was performed using ImageJ software. Coimmunoprecipitation HEK293T cells (1 × 107) were collected and lysed in 200 μl of ice-cold Glycerol Gradient Lysis Buffer containing 25 mM tris (pH 7.4), 150 mM NaCl, 1% NP-40, 1 mM EDTA, and 5% glycerol (Thermo Fisher Scientific), supplemented with a protease inhibitor cocktail (Roche). Lysates were incubated with anti-CyclinT1, anti-Flag, anti-HEXIM1, or anti-CDK9 antibodies (Table 1) for 4 hours, followed by incubation with protein A/G agarose (Beyotime Biotechnology) overnight at 4°C. Beads were washed with PBS four times, and samples were analyzed by Western blot. Quantification of RNA and integrated DNA Quantification of RNA and integrated DNA were set up followed previous protocols (36, 37). Cells were directly lysed by TRIzol and harvested for RNA extraction according to the instruction of the TRIzol reagent. The cDNA was synthesized using the Primescript RT Master Kit (Takara) with random and oligo(dT) primers and M-MLV Reverse Transcriptase (Promega) with specific primers. Quantitate viral cDNA intermediates using sequences for U5-gag, which represents completion of reverse transcription. RT-qPCR was performed using the SYBR premix Ex Taq II kit (Takara) according to the manufacturer’s instructions. The comparative CT (2−∆∆Ct) method was used to calculate the relative RNA levels against gapdh RNA levels, used as a control for quantification. For integrated and 2-LTR DNA, total DNA was isolated using a DNeasy DNA isolation kit (QIAGEN), and Alu-LTR was performed to measure integrated provirus cDNA. The first PCR was carried out by using nested primers, Alu-LTR 5′ from conserved sequences of human Alu. Following the initial PCR, a second nested PCR amplification was carried out by using an aliquot equivalent of the 22-cycle PCR products. The second nested PCR, which allows amplification of a portion of the LTR region of HIV-1 DNA, was performed by using SsoFast EvaGreen Supermix (Bio-Rad), nested primers. Efavirenz and Raltegravir served as positive controls, targeting reverse transcriptase and integrase, respectively. All the PCR primers were listed in the Supplementary Materials (table S2). RNA stability assay HEK293T cells were treated with actinomycin D (Absin) at the indicated time points. Total cellular RNA was extracted and analyzed by RT-qPCR using primers specific to Gag and gapdh, as described above. Enzyme-linked immunosorbent assay Viral stocks and quantification of viruses produced in kinetic experiments were monitored with an enzyme-linked immunosorbent assay kit specific for HIV-1 p24 (WanTai BioPharm) following the manufacturer’s instructions. Flow cytometric analysis For intracellular flow cytometry, infected PBMCs were washed twice with PBS and resuspended in Fixation/Permeabilization solutions (BD) for 20 min at 4°C. Next, cells were washed twice in Perm/Wash buffer (BD) and then resuspended cells in Perm/Wash buffer for 15 min at 4°C. Next, cells were left in FBS (2%) for 15 min followed by an incubation in the presence of antibodies in Perm/Wash solution for 30 min in the dark. Antibodies included Alexa Fluor 647 anti-Flag (BioLegend) and a rabbit immunoglobulin G isotype control [Alexa(R) 647 Conj.] (Cell Signaling Technology). Cells were washed twice with Perm/Wash buffer and subjected to flow cytometry analysis with the FACSCalibur housed in the Flow Cytometry Core Lab (Cancer Hospital Chinese Academy of Medical Sciences). Data analyses were performed using FlowJo version 7.6. Expression, purification of proteins, and Octet binding The SLFN14 and CyclinT1 constructs (SLFN14, residues 1 to 571; and CyclinT1, residues 1 to 292) were expressed in Escherichia coli BL21 (DE3) (TransGen Biotech) as His-tagged proteins in the pET28a vector. Cell pellets were resuspended in 40 ml of buffer containing 25 mM tris (pH 8.0), 200 mM NaCl, 8 M urea, and 20 mM imidazole per liter of culture and lysed by sonication. The soluble fraction was isolated by centrifugation. The supernatant was purified using Ni Sepharose excel (GE Healthcare) by nickel affinity chromatography. Pure SLFN14-dC1-His and CyclinT1 1-292-His proteins were lyophilized and stored at −80°C. Lyophilized proteins were dissolved in 50 to 100 mM tris buffer (pH 8) and dialyzed overnight against buffer containing 20 mM tris (pH 6.8), 50 mM NaCl, and 2 mM dithiothreitol before use. The Octet RED (ForteBio), equipped with super streptavidin biosensor chips (ForteBio), was used to analyze protein interactions in a fluidics-free system. The binding of SLFN14 protein to purified CyclinT1 1-292-His or CDK9 (Abcam) or P-TEFb (Sigma-Aldrich) was performed by ForteBio Inc. Company (www.fortebio.com). The purified SLFN14-dC1-His protein was biotinylated at a 3:1 ratio to biotin in a 30-min incubation at room temperature. Streptavidin biosensors (ForteBio) were coated in a solution containing 1 μM biotinylated protein for 10 min at 27°C. A duplicate set of sensors was incubated in an assay buffer with 2% FBS without proteins and served as a binding control. Both sets of sensors were blocked with biocytin (10 mg/ml) for 5 min. The binding of purified CyclinT1 or CDK9 protein or P-TEFb to sensors was measured over a period of 120 s. Data analysis on the ForteBio Octet RED instrument was performed using the reference subtraction in the ForteBio data analysis software. Chromatin immunoprecipitation ChIP was conducted using the EZ ChIP Chromatin Immunoprecipitation Kit (Millipore, 17-371RF) according to the manufacturer’s instructions. Cross-linking of chromatin was induced with 1% formaldehyde for 10 min at room temperature and stopped by adding glycine for 5 min at room temperature. Cells were washed twice with cold PBS, resuspended in 1 ml of SDS lysis buffer containing 1× Protease Inhibitor Cocktail II, and incubated for 10 min. At 4°C, nuclear pellets were sonicated to generate 200– to 1000–base pair (bp) DNA fragments, and, after centrifugation, the supernatant was aliquoted and stored at −20°C before use. For each ChIP reaction, nonspecific binding was eliminated by incubation of cross-linked DNA with protein G agarose (60 μl), from which 10 μl of the supernatant was saved as an input following centrifugation of the beads. Antibody was next incubated overnight at 4°C in the presence of precleared chromatin, followed by adding 60 μl of protein G agarose to each IP and another 1 hour of incubation at 4°C with rotation. Beads were sequentially washed with the following buffers: Low-Salt Immune Complex Wash Buffer, High-Salt Immune Complex Wash Buffer, LiCl Immune Complex Wash Buffer, and Tris-EDTA (TE) Buffer. Chromatin immunocomplexes were eluted by incubation for 10 min at 65°C with 1% SDS and 100 mM NaHCO3, and cross-linking was reversed by incubation in 200 mM NaCl and proteinase K for 1 hour at 65°C. RT-qPCR was subsequently performed for the quantification of the targeted DNA. Duolink PLA The Duolink PLA was performed according to the manufacturer’s protocol (Sigma-Aldrich). Briefly, HeLa cells were plated on glass coverslips, rinsed three times with PBS, and fixed in 4% formaldehyde in PBS for 10 min. Cells were permeabilized in 0.5% Triton X-100 for 10 min and blocked for 60 min at 37°C. After blocking, cells were incubated with antibodies against CDK9, HEXIM1, or FLAG for 60 min at 37°C, followed by incubation with corresponding secondary antibodies conjugated with PLA probes for 60 min at 37°C. Next, cells were incubated with ligase buffer for 60 min at 37°C; signals were amplified with DNA polymerase for 100 min at 37°C. Cells were washed three times in wash buffer B. The nuclei were stained with 4′,6-diamidino-2-phenylindole. Confocal microscopy and image acquisition were performed with the Nikon inverted microscope using 100× objective and the Volocity software (Nikon). Library construction and transcriptome sequencing Total RNA was isolated and purified from transfected HeLa cells using TRIzol reagent (Invitrogen, Carlsbad, CA, USA). To monitor RNA quality, RNA integrity number score was analyzed using an Agilent 2100 Bioanalyzer System. RNA sequencing libraries were generated from 150 ng of RNA using the Illumina’s TruSeq Stranded mRNA Sample Prep Kit following the manufacturer’s instructions, modifying the shear time to suitable time. RNA libraries were multiplexed and sequenced with 50-bp single reads (SR50) to a depth of ~30 million reads per sample on an Illumina HiSeq Xten/NovaSeq6000. Statistical analysis All statistical analyses were carried out using Prism 9 statistical software package. Bar graphs represent the means of results of experiments minimally performed three times, and error bars represent SDs. Comparisons between groups were carried out using a two-tailed paired Student’s t test. In all cases, a P value less than 0.05 was considered significant. The P values are reported on all graphs with asterisks following the GraphPad style: n.s., not significant; *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001.