Abstract
Efficient HIV-1 transcription requires the induction of cellular transcription factors, such as NF-κB, and the viral factor Tat, which through the recruitment of P-TEFb enhances processive transcription. However, whether cellular signals repress HIV-1 transcription to establish proviral latency has not been well studied. Previously, it has been shown that the receptor tyrosine kinase RON inhibits HIV transcription. To gain insights into the biochemical mechanisms by which RON inhibits transcription we examined the binding of transcription factors to the HIV provirus long terminal repeat using chromatin immunoprecipitation. RON expression decreased basal levels of NF-κB and RNA polymerase II (Pol II) binding to the HIV provirus long terminal repeat but did not prevent the induction of these complexes following treatment with cytokines. However, RON did decrease efficient transcription elongation because reduced RNA Pol II was associated with HIV-1 genomic sequences downstream of the transcriptional start site. There was a correlation between RON expression and increased binding of factors that negatively regulate transcription elongation, NELF, Spt5, and Pcf11. Furthermore, the ability of RON to inhibit HIV-1 transcription was sensitive to a histone deacetylase inhibitor and was associated with nucleosome remodeling. These results indicate that RON represses HIV transcription at multiple transcriptional check points including initiation, elongation and chromatin organization and are the first studies to show that cellular signaling pathways target Pol II pausing to repress gene expression.
The HIV long terminal repeat (LTR),5 the primary transcriptional regulatory element for the provirus, is often divided into four functional regions; modulatory element, enhancer, promoter, and Tat-activating region (TAR). The promoter, enhancer, and modulatory elements contain crucial binding sites for cellular transcription factors, such as Sp-1, C/EBPβ, and NF-κB, which are instrumental in initiating HIV-1 transcription. The TAR element forms a stable RNA stem-loop structure essential for the recruitment of the HIV-encoded transactivator Tat, which enhances HIV-1 transcription by increasing polymerase processivity through interaction with P-TEFb (positive transcription elongation factor b) (1, 2, 3). Cellular transcription factors, as well as Tat, recruit coactivators such as histone acetyltransferases and SWI/SNF complexes to the HIV LTR (4, 5, 6), which modify and displace a positioned nucleosome to further promote HIV-1 transcription. Transcription initiation, elongation, and chromatin organization have all been demonstrated to be critical check points for HIV-1 transcription. Cellular signals that negatively regulate these processes would result in inefficient provirus transcription and possibly contribute to viral latency.
The existence of long-lived stable HIV reservoirs was demonstrated by the rebound of virus replication following highly active antiretroviral therapy interruption (7, 8, 9). These latent reservoirs, which include quiescent memory T cells and tissue-resident macrophages (10, 11), represent a subset of cells with decreased or inactive proviral transcription. Studies with chronically and acutely infected cells show that mutations in Tat (12), lack of cellular transcription factors (13, 14, 15), microRNA machinery (16), and proviral integration into transcriptionally silent sites (17, 18) contribute to postintegration latency. Transcription of latent HIV-1 is induced by treating cells with cytokines or mediators that activate cells (7, 9, 19), and the signals necessary for HIV-1 reactivation include protein kinase C, NFAT, Src kinases, and NF-κB (15, 19). Insufficient or inappropriate signaling may extinguish HIV transcription (11) as suggested by reports of PI3K (20) and MAPK (21) signaling repressing HIV-1 transcription. The exact mechanisms by which cellular signals establish HIV latency are not fully understood and such studies have been hampered by the rarity and inaccessibility of latently infected cells. To gain an appreciation of signaling events that actively repress HIV-1 transcription will require experimental systems in which HIV expression can be extinguished and the biochemical processes can be characterized.
We have previously demonstrated that the receptor tyrosine kinase RON represses HIV-1 transcription (22). RON, which is expressed on tissue resident macrophages, regulates macrophage differentiation, apoptosis, and cell movement and in general suppresses activities associated with inflammation (23). Macrophages from RON knockout mice produce elevated levels of NO in response to IFN-γ and LPS, and exhibit increased inflammation, tissue damage, and death due to endotoxic shock upon LPS challenge (24). The ability of RON to maintain homeostasis of macrophage inflammatory activity is in part due to its ability to repress the transcription of genes such as IL-12p40 and inducible NO synthase. To gain insights into how RON represses HIV-1 provirus transcription, we have characterized the binding of transcription factors as well as the behavior of RNA polymerase II (Pol II) on the HIV-1 LTR in the absence and presence of RON. Our results indicate that RON, in addition to influencing NF-κB signaling, regulates transcription elongation and chromatin organization to inhibit HIV-1 transcription. These findings provide new insights into how cellular signals contribute to HIV provirus latency as well as more general implications into how RON, by targeting transcription elongation, controls gene expression.
Materials and Methods
Cells and cell lines
Cell lines used in these studies were the monocytic cell line U937 and U937-RON, which overexpress RON (22). RON mutations were introduced using a QuikChange Mutagenesis kit (Stratagene) and then were subcloned into murine stem cell virus-derived retroviral transfer vector MSCV2.1 as previously described (25, 26). Receptor expression was confirmed by immunoblotting (22) using a rabbit anti-RON polyclonal Ab (Santa Cruz Biotechnology) and mouse anti-GAPDH (Fitzgerald) as a loading control. U937-derived cells were cultured in RPMI 1640 medium supplemented with 10% FCS, 100 U/ml penicillin, 100 mg/ml streptomycin, and 0.2 M l-glutamine. 293T human embryonic kidney cells were grown in DMEM supplemented with 10% FCS, 100 U/ml penicillin, 100 mg/ml streptomycin, and 0.2 M l
HIV-1 infections and luciferase assays
The pNL43-Luc+Env− virus, a replication-defective virus (27), was generated by transfecting 293T cells with 15 μg of pNL43-Luc+Env− DNA, 3 μg of VSV-G-Env (vesicular stomatitis virus glycoprotein envelope) DNA, and 3 μg of Rev DNA with calcium phosphate transfection (22). One milliliter of virus stock was added to 1.0 × 106 cells for 18 and 48 h postinfection virus transcription was measured using a luciferase assay (Promega).
Chromatin immunoprecipitation (ChIP)
ChIP was performed as previously described (28, 29). Briefly, U937 or U937-RON cells were infected with NL4-3 Luc+Env− virus pseudotyped with VSV-G-env for 18 h and either stimulated or left unstimulated as indicated. The 37% formaldehyde was added to a final concentration of 1% for cross-linking and the reaction was stopped with glycine. Following washes, cells were resuspended in 3 ml of sonication buffer (1% SDS, 0.5 M EDTA, 1 M Tris (pH 8.0), and PMSF) and sonicated for 20 min. Debris was removed by centrifugation and an equal volume of urea was added to the samples for dialysis in buffer (10 mM Tris (pH 8.0), 1 mM EDTA, 0.5 mM PMSF, 10% glycerol, 0.1% sodium deoxycholate, and 1% Triton X-100). Chromatin samples were spun at 4°C for 5 min and precleared with protein A-Sepharose beads for 30 min at 4°C and stored at −70°C before use for immunoprecipitations.
Immunoprecipitations were performed by incubating 20 μg of chromatin with 5 μg of Ab at 4°C overnight. Abs used include anti-p50, anti-p65, anti-NELF-E, anti-Pol II, and normal rabbit IgG (Santa Cruz Biotechnology). Rabbit anti-Pcf11 antiserum is previously described (29). Following incubation with Ab, 35 μl of protein A-Sepharose beads was added and samples were incubated for 2 h at 4°C. Beads were washed six times in low-salt buffer (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris-HCl (pH 8.1), 150 mM NaCl) and three times with high-salt buffer (0.1% SDS, 1% Triton X-100, 2 mM EDTA, 20 mM Tris-HCl (pH 8.1), 500 mM NaCl), followed by an overnight wash in lithium wash buffer (0.25 M LiCl, 1% Nonidet P-40, 1% sodium deoxycholate, 1 mM EDTA, 10 mM Tris-HCl (pH 8.1)) at 4°C. Beads were washed one final time in 1 ml of TE buffer (10 mM Tris (pH 7.5), 1 mM EDTA) before eluting protein/DNA complexes by incubating twice with 250 μl elution buffer (1% SDS, 0.1 M NaHCO3) for 15 min at room temperature. Cross-links were reversed and proteins digested by adding 0.2 M NaCl and incubating the samples at 65°C for 4 h followed by the addition of 10 mM EDTA, 40 mM Tris-Cl (pH 6.5), and 40 μg/ml proteinase K and incubating at 45°C for 1 h. DNA was purified by phenol/chloroform extraction followed by ethanol precipitation and was resuspended in 0.5× TE buffer.
Immunoprecipitated or input DNA was used in a 50-μl PCR containing 10 pmol of each primer, 200 μM dNTP, 4 μCi [α-32P]dCTP, 12.5 U of Taq polymerase, and 1× Taq buffer (Gene Choice). The following PCR conditions were used to amplify DNA: 20 cycles at 94°C for 1 min, 57°C for 55 s, 72°C for 50 s, and a final extension of 72°C for 10 min. The following primers were used to amplify different regions of the HIV-1 gene: LTR (−155 to +79) 5′-CCGAGAGCTGCATCCGGAGT-3′ and 5′-TTGAGGCTTAAGCAGTGGGTTCC-3′; and LTR (+2415 to +2690) 5′-GGTACAGTATTAGTAGGACCTACACC-3′ and 5′-CCCTTCCTTTTCCAT CTCTGTACA-3′.
PCR products were run on a 6% acrylamide nondenaturing gel and detected by radioactivity using a PhosphorImager (Molecular Dynamics). The intensity of each band was quantified by volume analysis using ImageQuant software.
Permanganate footprinting
Permanganate (KMnO4) footprinting was performed as previously described (28). Briefly, 4 × 106 infected or uninfected U937 or U937-RON cells were washed with PBS and treated with permanganate by adding 20 mM KMnO4. The permanganate reaction was incubated on ice for 1 min and stopped with stop solution (20 mM Tris-HCl (pH 7.5), 20 mM NaCl, 40 mM EDTA, 1% SDS, 400 mM 2-ME). The samples were immediately shaken until all coloration had vanished and treated with 50 μg of proteinase K for at least 1 h at 37°C, then extracted with a sequence of phenol, phenol-chloroform-isoamyl alcohol (49.5:49.5:1), and chloroform. DNA was precipitated with 0.6 M sodium acetate (pH 6.0) and ethanol, washed with 75% ethanol and dissolved in TE buffer. To determine the pattern of permanganate reactivity, 500 ng of DNA sample was diluted to 15 μl in TE buffer (pH 7.5) and 75 μl of H2O plus 10 μl of piperidine was added, and the samples were incubated at 90°C for 30 min. Following the incubation, samples were diluted with H2O and extracted three times with isobutanol and one time with ether. The DNA was precipitated with ethanol, dissolved in TE buffer (pH 7.5), transferred to fresh siliconized tube, and used in a primer extension reaction.
Restriction enzyme accessibility assay
A total of 2 × 106 infected U937 or U937-RON cells cultured in the absence or presence of 10 ng/ml TNF-α plus 10 ng/ml IL-6 for 16 h were harvested and resuspended in 1 ml of buffer A (10 mM Tris (pH 7.40), 10 mM NaCl, 3 mM MgCl2, and 0.3 M sucrose) and incubated for 5 min on ice. Cells were lysed with 10% Nonidet P-40 to a final concentration of 0.3%, and nuclei were spun down and resuspended in buffer B (10 mM Tris (pH 7.9), 10 mM MgCl2, and 50 mM NaCl, 1 mM DTT, 0.1 mM PMSF, and 100 μg/ml BSA). The 40 U of HindIII (New England Biolabs) were added and samples were incubated at 37°C for 1 h. Proteins were digested by adding 200 μg/ml proteinase K in buffer (100 mM Tris (pH 7.5), 200 mM NaCl, 2 mM EDTA, and 1% SDS) and incubating samples at 45°C for 1 h. DNA was extracted with phenol-chloroform-isoamyl alcohol (25:24:1) and precipitated with ethanol, resuspended in H2O, and digested with 40 U of PvuII (New England Biolabs) at 37°C for 1 h. The digested samples were extracted with phenol-chloroform-isoamyl alcohol (25:24:1) and precipitated with ethanol. Genomic DNA from HIV-1-infected U937 cells was used as a control.
Products from restriction digest were detected by ligation-mediated PCR (LM-PCR) using a series of nested primers LM-1, LM-2, and LM-3, spanning sequences +131 to +186 of the HIV-1 LTR where +1 is the transcription start site as previously described (28, 29). Nested primer sequences are as follows: LM-1, 5′-ACTGCTAGAGATTTTCCACACT-3′; LM-2, 5′-CCACACTGACTAAAAGGGTCTG-3′; LM-3, 5′-GGGTCTGAGGGATCTCTAGTTACCA-3′; Linker A, 5′-GCTGTGACAAGGAGATCTGAATTC-3′; and Linker B, 5′-GAATTCAGATC-3′.
Nested primer LM-3 was radiolabeled with 300 μCi of [γ-32P]ATP, 10 U of kinase (Promega), and 1× kinase buffer. The labeling reaction was incubated at 37°C for 1 h and purified with a NucTrap (Stratagene) column. Linker DNA was generated by annealing 20 μM Linker A and 20 μM Linker B in 250 mM Tris-Cl (pH 7.7), by heating to 95°C for 5 min, and by allowing the primers to slowly cool to room temperature.
Digested genomic DNA (500 ng) or chromatin was used in a first strand synthesis reaction that included first strand buffer (0.2 M NaCl, 0.05 M Tris-Cl (pH 8.9), 25 mM MgSO4, 0.05% gelatin), 0.5 pmol LM-1, 250 μM dNTPs, and 0.5 U of vent polymerase (New England Biolabs). The first strand synthesis reaction was 95°C for 5 min; 54°C for 30 min; and 76°C for 10 min. Following synthesis, 17.5 μl of modified ligation dilution solution (0.126 M Tris-Cl (pH 7.5), 20.57 mM MgCl2, 57 mM DTT), 25 μg of BSA, 9.25 μl of modified ligation premix (26.93 mM MgCl2, 54 mM DTT), 1.25 μg of BSA, 3 mM rATP, 100 pM annealed linker pair, and 2 U of T4 DNA ligase (Upstate Biotechnology) was added and ligation mixtures were incubated overnight at 15°C. Ligated DNA was precipitated, resuspended in H2O, and amplified with 20 μl of 5× vent amplification buffer (0.2 M NaCl, 0.1 M Tris-Cl (pH 8.9), 25 mM MgSO4, 0.05% gelatin, 0.5% Triton X-100), 10 pM LM-2, 10 pM Linker A, 0.29 mM dNTPs, and 1 U of vent polymerase. The following PCR conditions were used: 95°C for 3 min; 58°C for 2 min; 76°C for 3 min; 10 cycles of 95°C for 1 min; 58°C for 2 min; 76°C for 3 min; and 12 cycles of 95°C for 1 min; 58°C for 2 min; 76°C for 4 min). PCR products were detected by primer extension using 9.5 μl of H2O, 4 μl of 5× vent amplification buffer, 1 nM dNTPs, 2 pM radiolabeled LM-3, and 1 U of vent polymerase using the following reaction conditions: 95°C for 3 min; 62°C for 2 min; 76°C for 10 min; 95°C for 1 min; 62°C for 2 min; and 76°C for 10 min. The labeling reaction was stopped with 14 μl of stop solution (71 mM Na2EDTA (pH 8.0), 2.57 M sodium acetate (pH 7.0)) added to each reaction, and DNA was extracted with phenol-chloroform-isoamyl alcohol (25:24:1) and precipitated with cold ethanol. Products were resolved on a 6% acrylamide nondenaturing gel and visualized using a PhosphorImager (Molecular Dynamics).
Results
Signals emanating from RON inhibit HIV transcription
RON is primarily expressed on human tissue resident macrophages which are not readily accessible or amendable to biochemical studies (22, 23, 30, 31). Therefore, to gain further insight into mechanisms by which RON repress HIV-1 provirus transcription we used site-directed mutagenesis to generate several mutations in the RON cytoplasmic domain (25, 26). In particular, we mutated Y1353 and Y1360, which serve as a multifunctional docking site for several signaling molecules including Gab-1, Grb2, PI3K, phospholipase C, and Src homology protein-2 (32). In addition, we disrupted the kinase activity of RON. U937 monocytic cells expressing these mutant RON receptors or wild-type RON were infected with the HIV clone HIV-Luc, a replication incompetent HIV-1 clone in which the envelope gene (Env) was replaced with a luciferase reporter gene (27) and HIV transcription was monitored by measuring luciferase activity. Immunoblots indicated that similar levels of RON and mutant RON were expressed in the cells (Fig. 1⇓). Consistent with previous studies (22), wild-type RON inhibited HIV transcription; however, mutating either Y1353, Y1360, or the kinase domain ablated the ability of RON to repress HIV-1 transcription (Fig. 1⇓) demonstrating that signals emanating from RON actively repress provirus transcription.
RON docking sites and kinase domain are required for repression of HIV transcription. A, U937 cells stably expressing RON, RON harboring mutations in Y1353, Y1360, the kinase domain or mock-transduced with murine stem cell virus (MSCV), were infected with the HIV clone HIV-Luc. Provirus expression was monitored by measuring luciferase activity 24 h postinfection. Data are presented as the percentage of luciferase, with luciferase activity in the cells lacking RON being set at 100%. Each data point represents three independent infections and error bars show the SD of these replicates. B, Immunoblots showing RON and mutant RON expression for transduced cells.
RON inhibits binding of NF-κB to the HIV LTR
The ability of RON to inhibit HIV transcription provides a unique experimental system in which signaling events and transcriptional targets that actively repress HIV-1 provirus transcription can be identified. Initially, we explored whether RON was affecting the recruitment of NF-κB to the HIV LTR in HIV-1 infected cells because it has been suggested that RON inhibits NF-κB signaling and translocation of the p65 subunit to the nucleus (22, 33, 34). To analyze direct binding of NF-κB subunits to the HIV-1 enhancer in RON expressing cells we used ChIP. Chromatin was prepared from HIV-infected U937 and U937-RON cells. It should be noted that following infection, there was no selection for HIV expression; therefore, these experiments were performed on a heterogeneous population of infected cells with a spectrum of replicative properties and not a clonal population of HIV-infected cells. ChIP was performed using Abs against NF-κB subunits p65 and p50. In agreement with previous studies showing that the NF-κB heterodimer p50:p65 is essential for the activation of HIV-1 transcription (1, 14), we observed binding of both p65 and p50 to the LTR in U937 cells (Fig. 2⇓). In U937 cells expressing RON, the association of both of these subunits with the LTR was reduced by greater than 75%, suggesting that RON inhibits HIV-1 transcription in part through decreasing p50:p65 recruitment and binding to the HIV-1 LTR in unstimulated cells (Fig. 2⇓). To gain insight into RON-dependent repression of HIV transcription in the presence of inflammatory mediators, we measured p50 and p65 binding to the HIV-1 LTR in U937 and U397-RON cells stimulated with IL-6 and TNF-α following HIV infection. Cytokine treatment induced binding of p65 and p50 to the HIV LTR by 7-fold in the presence of RON, so that the binding of NF-κB was comparable in control U937 and U937-RON cells. However, RON signaling represses HIV-1 transcription even in the presence of inflammatory stimuli (22 and data not shown), indicating that the recruitment of NF-κB to the HIV LTR in response to cytokines is not sufficient to induce HIV-1 transcription in U937-RON cells.
Binding of Rel transcription factors to the HIV LTR in the absence and presence of RON. A, ChIP for Rel proteins. Soluble chromatin was prepared from U937 or U937-RON cells infected with NL4-3 virus and immunoprecipitated with the indicated specific Abs or nonspecific isotype control Ab (N.S. Ab). Some cells were stimulated (STM) with 10 ng/ml IL-6 and 10 ng/ml TNF-α for 6 h following infection. DNA that was immunoprecipitated was detected using primers that amplify −155 to +79 of the HIV LTR. Lanes labeled 5% and 1% are products generated from input DNA that had not been precipitated. All data in the composite were taken from the same gel and exposure. B, Densitometry was used to quantify LTR-associated bands. Background was subtracted out and the intensity of each band was normalized to 1% input DNA to yield a total percentage of precipitation. Data include results from three ChIP experiments ± SD.
RON regulates RNA Pol II activity
The fact that induction of NF-κB was not sufficient to induce HIV-1 transcription in the presence of RON prompted us to explore alternative mechanisms by which RON repressed transcription. Efficient HIV-1 transcription elongation requires recruitment as well as the phosphorylation and subsequent activation of the RNA Pol II complex at the promoter region of the HIV-LTR. To explore the possibility that RON signaling decreases RNA Pol II processivity we initially assessed the presence of Pol II at the HIV-1 LTR, as well as sequences several thousand bases downstream of the transcription start site. The association of Pol II with HIV-1 provirus in infected U937 and U937-RON cells was determined by ChIP. In the absence of cytokine stimulation, Pol II associated with the LTR was decreased by 70% in U937-RON cells compared with control U937 cells, suggesting that RON inhibits basal transcription initiation (Fig. 3⇓A). Following cytokine treatment, Pol II binding was induced to similar levels in the U937 and U937-RON cells (Fig. 3⇓B). Because the levels of Pol II associated with the LTR in the two cell lines following cytokine stimulation were nearly equivalent, yet transcription of HIV was inhibited by >80% (22 and data not shown), we concluded that RON was not interfering with formation of the preinitiation complex, but was repressing HIV-1 transcription by inhibiting transcription elongation. To analyze the ability of Pol II to be released from the transcription start site on the HIV-1 LTR, we again used ChIP, but investigated the association of Pol II with a region spanning +2415 to +2690 bp downstream from the start site. In U937-RON cells there was 65% less Pol II associated with these downstream sequences compared with U937 cells (Fig. 3⇓A). Following cytokine treatment, there was a 3.5-fold increase of Pol II found associated with this same region (+2415 to +2690) in U937 cells, indicating a release of paused Pol II and an increase in transcription elongation. Although in the U937-RON cells there was a modest increase in Pol II associated with downstream sequences, the amount of Pol II associated with the downstream sequences was still reduced by 50% compared with stimulated U937 controls, consistent with less Pol II being released from its paused state. The observation that RNA Pol II association with downstream HIV-1 sequences in the absence and presence of stimulatory factors is decreased in U937-RON cells compared with U937 cells suggests that RON signaling inhibits HIV-1 transcription by suppressing RNA Pol II processivity.
Binding of RNA Pol II to the HIV LTR in the absence and presence of RON. A, Soluble chromatin was prepared from U937 or U937-RON cells infected with NL4-3-Luc that were either unstimulated or stimulated with 10 ng/ml IL-6 and 10 ng/ml TNF-α for 6 h. ChIP was performed using anti-Pol II or nonspecific control Ab (N.S. Ab). PCR was performed using primers to amplify LTR-associated product (−155 to +79) or downstream proviral sequences spanning (+2415 to +2690) to detected precipitated DNA. Lanes labeled 5% and 1% are PCR products generated from input DNA. All data in the composite were taken from the same gel and exposure. B, Densitometry was used to quantify LTR and downstream associated sequences. Background was subtracted out and the intensity of each band was normalized to 1% input DNA to yield a total percentage of precipitation. Data include results from three ChIP experiments ± SD.
To more precisely map the position and distribution of the polymerase complex throughout the HIV-1 LTR and investigate whether Pol II is indeed paused at the HIV-1 promoter, we used a previously described technique of permanganate footprinting (28, 35, 36). Permanganate footprinting permits the distribution of Pol II on DNA in vivo to be visualized at a resolution of 10–20 bp by oxidizing the C5–C6 double bond in single-stranded thymines located within the transcription bubble with permanganate. Oxidized bases are cleaved with piperidine and then amplified using LM-PCR. Hypersensitive sites represent regions with an abundance of Pol II (37). Fig. 4⇓ shows the permanganate reactivity of thymines in HIV-infected unstimulated U937 and U937-RON cells. Comparing the patterns of reactive thymines from U937 with that observed in purified DNA reveals an even distribution of RNA Pol II across the HIV LTR with reactive thymines at positions +45, +65, +75, +85, and +95. Analysis of the permanganate reactivity of thymines in infected U937-RON cells reveals strong promoter proximal pausing of Pol II at position +43 with only modest amounts of RNA Pol II detected downstream of this site, as indicated by the lack of hypersensitive thymines downstream from position +43. These data are consistent with the RNA Pol II ChIP results showing that RON signaling results in RNA Pol II pausing. This pausing occurs before the completion of the TAR element, which requires transcription through +61 (38), and despite the recruitment NF-κB and Pol II to the HIV LTR (Figs. 2⇑ and 3⇑). Importantly, these data are the first to show that signals emanating from a cellular receptor repress HIV transcription by establishing a paused Pol II complex.
RON induces Pol II pausing on the HIV LTR. Permanganate footprinting, as described in Materials and Methods, was performed on unstimulated U937 and U937-RON cells following infection with HIV-1 NL4-3. Cleavage events were detected by LM-PCR using primers spanning LTR sequences from +1 to +131. Values at right indicate the location of thymines on the nontranscribed strand. The arrowhead indicates a hypersensitive thymine cleave site consistent with a paused Pol II complex at this location.
Negative elongation factor (NELF) and DRB sensitivity-inducing factor (DSIF) are two protein complexes that have been shown to inhibit transcription elongation by maintaining Pol II pausing in the promoter proximal region of the 70-kD heat shock protein gene in Drosophila as well as the HIV LTR (28, 39, 40, 41, 42). Furthermore, Pcf11, a transcription termination factor has been recently demonstrated to negatively regulate HIV-1 transcription by prematurely terminating transcription in the HIV-1 LTR (29). ChIP analysis was done to investigate the binding of NELF-E, Spt5 (a subunit of DSIF), and Pcf11 to the HIV-1 LTR using specific Abs with chromatin prepared from U937 and U937-RON cells that were infected with HIV-1 and stimulated with cytokines or left unstimulated. In unstimulated U937-RON cells there was a >3-fold increase in NELF-E and Spt5 associated with the HIV-1 LTR compared with the U937 cells (Fig. 5⇓), whereas a very modest increase in binding of Pcf11 was observed in U937-RON cells. However, if we consider the levels of these factors normalized to bound Pol II (see Fig. 3⇑), then more dramatic changes become apparent. In the context of infected U937-RON cells, there is 5-fold more Pcf11, 20-fold more NELF, and 60-fold more Spt5 associated with the HIV-1 LTR relative to Pol II vs U937 cells (Fig. 5⇓C). Following cytokine stimulation, there was an overall increase in the amount of NELF-E, Spt5, and Pcf11 that precipitated with HIV LTR sequences. The ratio of Spt5 to Pol II and Pcf11 to Pol II bound to the HIV-LTR were similar in both U937 and U937-RON cells, whereas 2-fold more NELF was bound to the HIV LTR in U937-RON cells following cytokine stimulation. Taken together, these data suggest that RON signaling inhibits HIV transcription in unstimulated cells by promoting the association of negative elongation factors, NELF, DSIF, and Pcf11, with the Pol II. In stimulated cells, RON expression correlates with increased NELF binding at the HIV LTR and decreased Pol II associated with downstream HIV genomic sequences.
Binding of elongation factors to the HIV LTR in the absence and presence of RON. A, Soluble chromatin was prepared from U937 or U937-RON cells infected with NL4-3-Luc that were either unstimulated or stimulated with 10 ng/ml IL-6 and 10 ng/ml TNF-α for 6 h. ChIP was performed using anti-NELF-E, Spt5, Pcf11, or nonspecific control Ab (N.S. Ab). PCR was performed using primers to amplify LTR-associated product (−155 to +79) to detected precipitated DNA. Lanes labeled 5% and 1% are PCR products generated from input DNA. Data in the composite were taken from the same gel and exposure. B, Densitometry was used to quantify LTR-associated bands. Background was subtracted out and the intensity of each band was normalized to 1% input DNA to yield a total percentage of precipitation. Data include results from three ChIP experiments ± SD. C, The densitometry values obtained from ChIPs shown in A were normalized to a percentage of Pol II associated with the LTR (−155 to +79) in Fig. 3⇑. It should be noted that the Pol II ChIP in Fig. 3⇑ and this experiment were performed with the same chromatin preparation, amplified at the same time, resolved on the same gel, and compared under the same exposure.
RON alters the nucleosome organization
Induction of HIV-1 transcription in latently infected cell lines correlates with changes in histone acetylation and the displacement of a positioned nucleosome (nuc-1) (43). Furthermore, nucleosomes potentially regulate RNA Pol II pausing (44) and it has been suggested that nuc-1, which spans +10 to +155 (43), contributes to inefficient transcription elongation by acting as a natural pausing site for Pol II at the HIV-1 LTR. To investigate a possible role for chromatin in RON-dependent suppression of HIV-1 transcription, we infected U937 and U937-RON cells with HIV-Luc and 18 h postinfection cells were treated with the histone deacetylase inhibitor TSA, which has been shown to induce repressed HIV-1 transcription (45). The addition of TSA to HIV-1-infected U937-RON cells abrogated the ability of RON to inhibit HIV-1 transcription. In the presence of TSA, HIV transcription was induced 5-fold in the U937-RON cells returning expression levels to those observed in U937 controls (Fig. 6⇓). U937 cells lacking RON, which support active HIV-1 transcription, did not respond to TSA, arguing against TSA having a global effect on transcription by inducing cellular transcription factors such as NF-κB. Therefore, RON signaling appears to negatively regulate histone acetylation thus inducing a chromatin organization that does not favor efficient HIV-1 transcription.
RON influences nucleosome organization at the HIV LTR. A, U937 cells mock-transduced with murine stem cell virus (U937-MSCV) and U937-RON cells were infected with the HIV clone HIV-Luc. At 24 h postinfection, cells were treated with ethanol vehicle control, or 250 nM TSA for 12 h. Cells were collected and provirus transcription was measured by luciferase activity. Data are from a single experiment in which three independent infections were performed for each treatment. Error bars are SDs between the triplicate experiments. B, Diagram of the locations of restriction sites and positioned nucleosomes on the HIV LTR. C, Restriction enzyme accessibility in the HIV LTR in the absence or presence of RON. Nuclei were isolated from HIV-Luc-infected U937 and U937-RON cells and digested with 20 U of HindIII. Genomic DNA was isolated and cut to completion with PvuII. Digested DNA products were then analyzed by LM-PCR. As a positive control, purified genomic DNA was digested with either PvuII or HindIII. Experiment is representative of at least three independent experiments.
A restriction enzyme accessibility assay was used to determine whether RON impacts the association of nuc-1 with the HIV LTR. Dissociation of nuc-1 from HIV-1 DNA allows access to the previously protected HindIII restriction site, and the cutting frequency of HindIII provides a measure of the accessibility of the underlying DNA (43, 46). If RON represses HIV-1 transcription by permitting a nucleosome to remain positioned at the transcriptional start site, then we would predict that HindIII would cut less efficiently, because this site falls within the region protected by the nucleosome. To examine the presence of nuc-1 at the HIV-1 LTR in U937-RON cells, nuclei were isolated from HIV-1-infected U937 and U937-RON cells and digested with HindIII. DNA was prepared from the digested nuclei and cut to completion with PvuII. HindIII cutting was determined by LM-PCR using nested primers spanning a 54 nt region downstream of the HindIII and PvuII cut sites. The frequency of HindIII digestion of the HIV-1 LTR in U937-RON cells was reduced by 25–30% vs control U937 cells regardless of cytokine treatment, suggesting that RON interferes with nuc-1 remodeling (Fig. 6⇑ and data not shown). Overall, these data indicate that RON signaling induces changes in chromatin that correlate with the establishment of a paused Pol II and repressed HIV-1 provirus transcription in monocytic cells.
Discussion
HIV-1 transcription is regulated at multiple levels, including recruitment of cellular transcription factors, chromatin organization, transcription initiation, and transcription elongation. Despite knowing a great deal about the factors and signals that activate HIV-1 transcription, whether cellular biochemical pathways actively repress HIV transcription is not clear. We have taken advantage of the ability of RON to repress HIV transcription to determine whether cellular signals target these transcriptional check points to inhibit HIV-1 transcription.
NF-κB is a critical regulator of HIV-1 transcription that is required for efficient transcription initiation and remodeling of chromatin structure. NF-κB positively and negatively regulates HIV-1 transcription and latency. Induction of p65 and recruitment of histone acetyltransferases are key events for reversing proviral transcription latency (15, 19), whereas p50 homodimers recruit histone deacetylases to promote a repressive chromatin state at the HIV-LTR. RON has been shown to decrease NF-κB signaling (22, 33, 34), and we observed a modest decrease in the NF-κB p50 and p65 binding to the HIV LTR in the presence of RON. RON did not increase p50 binding on the HIV LTR, suggesting that RON repression of HIV transcription is independent of p50. However, RON did not inhibit the induction of NF-κB binding to the HIV-LTR in response to cytokines or PMA nor was this increase in NF-κB binding sufficient to rescue HIV-1 transcription in the presence of RON. Therefore, there are multiple mechanisms, both NF-κB-dependent and -independent, that negatively regulate HIV-1 transcription.
Using RON to repress HIV transcription allowed us to identify transcription elongation as a key regulatory step in provirus transcription. To our knowledge, this is the first report to suggest that Pol II pausing is targeted by cellular signals. In the presence of RON, we observed less Pol II processivity as indicated by the abundance of Pol II at the HIV-LTR. RON signaling could repress HIV-1 transcription by establishing a paused Pol II complex. We observed in HIV-1-infected cells expressing RON that there was more NELF, DSIF, and Pcf11 associated with the LTR, suggesting that these factors are participating in establishing RON-dependent Pol II pausing and premature termination. NELF and DSIF are associated with the Pol II complex and inhibit transcription elongation by pausing Pol II (42, 47, 48, 49). The inhibitory action of these two proteins is overcome by P-TEFb (42, 50). Importantly, NELF and DSIF have been implicated in regulating HIV-1 transcription elongation. Both DSIF and NELF contribute to Tat-dependent activation of HIV-1 transcription and are phosphorylated by P-TEFb (39, 51). Furthermore, the E subunit of NELF associates with the TAR element (39, 52), and ectopic expression of this NELF subunit inhibits transient basal transcription from the HIV LTR (39). Depleting NELF in the context of U1 cells, which harbor latent HIV-1 provirus, released the Pol II paused at ∼+42 and increased processive HIV transcription (28). In addition, the ratio of Pcf11 to Pol II is 5-fold higher in unstimulated cells expressing RON than in control cells suggesting that repression of HIV-1 by RON in unstimulated cells is re-enforced by Pcf11-mediated premature termination of transcription. This observation is consistent with recent findings by our groups that showed Pcf11 negatively regulates HIV-1 transcription (29).
We also show that signals emanating from RON inhibit transcription in part by influencing the chromatin structure of the HIV LTR. RON-mediated inhibition of HIV-1 transcription correlated with changes in restriction enzyme accessibility and the histone deacetylase inhibitor TSA induced HIV-1 transcription in the presence of RON indicating that displacement of the positioned nucleosome located at the 5′ untranslated leader region (5′-UTR) of the HIV LTR ∼+110 bp downstream of the transcriptional start site limits provirus transcription. This positioned nucleosome has been proposed to block transcription initiation or elongation (53, 54) and by acting as a barrier to the elongation complex may re-enforce proximal RNA Pol II pausing. In addition, TSA may be influencing the acetylation of other critical transcription factors including NF-κB subunits (55, 56), C/EBPβ (57), and Tat (58). However, the ability of TSA to rescue HIV transcription in the presence of RON does not exclude a model in which Pol II processivity and chromatin organization are coupled or coordinately regulated. RNA Pol II pausing may facilitate recruitment of histone acetyltransferases and nucleosome remodeling complexes to the HIV LTR (59). This model is consistent with findings from our labs that demonstrated that releasing paused RNA Pol II increased nucleosome remodeling and HIV transcription in a latently infected cell line (28).
Based on the literature and our recent findings we propose that HIV-1 transcription is actively repressed by proximal promoter pausing and premature transcription termination. The paused Pol II is established by negative elongation factors NELF and DSIF and a positioned nucleosome at the 5′-UTR. Pcf11 is also recruited to the promoter mediating premature termination (29). Processive HIV transcription requires Tat, which recruits P-TEFb to the HIV LTR. P-TEFb phosphorylates NELF, DSIF and the C-terminal domain of RNA Pol II (60). Tat also facilitates HIV transcription by recruiting histone acetyltransferases and Swi/Snf chromatin remodeling complexes to the HIV LTR, which extensively reorganize the chromatin allowing for productive HIV-1 transcription. Therefore for efficient HIV provirus transcription at least three critical check points must be overcome, transcription initiation, transcription elongation/premature pausing, and repressive chromatin organization. Importantly, we show that signaling from cellular receptors can either direct efficient HIV-1 transcription and replication or actively repress HIV-1 transcription by targeting each of these check points.
We show that the docking sites and kinase domain in the intracellular domain of RON are required for repression of HIV-1 transcription suggesting that signaling cascades downstream of RON that directly impact HIV transcription might include Grb-2, Ras/MAPK pathway, and MAPK kinase (32). Upon activation and autophosphorylation of tyrosines in the docking site, RON recruits the adapter proteins Grb-2 and Gab-1. Gab-1 has binding sites for the p85 regulatory subunit of PI3K, which serves to recruit PI3K to the signalsome and amplifies RON activity. Although PI3K is a critical factor in regulating RON function (61, 62) and in the context of T cells PI3K inhibits HIV-1 transcription (20), studies with specific inhibitors including LY294002 and wortmannin suggest that PI3K is not required for RON mediated inhibition (data not shown). Similarly, the ERK1/2 inhibitor U0126 did not influence HIV-1 transcription in infected cells expressing RON (data not shown). Other signal transduction molecules that have been suggested to inhibit HIV transcription that are downstream of RON are MEK1 and p38 (21). However, we have not formally tested whether these pathways are responsible for the ability of RON to inhibit HIV-1 transcription.
Given the physiological expression pattern of RON, which is restricted to tissue resident macrophages such as Kuppfer cells, Langerhans cells, alveolar macrophages, and microglia, we speculate that different macrophage subsets would have varying capacities for supporting HIV-1 replication. This is consistent with recent observations that not all macrophage subsets efficiently support HIV transcription. For example, CD16+ monocytes were shown to be more permissive for HIV-1 infection and replication than CD14+CD16− monocytic cells (63, 64). We would predict that some tissue resident macrophage populations are infected but do not readily support replication, possibly as a result of negative cellular signals, from receptors such as RON. These repressive signals create a microenvironment that establishes latent cellular reservoirs enabling the persistence of HIV-1. With time these negative signals are compromised as a result of chronic infection and inflammation (65, 66), thus disrupting the normal homeostasis of the tissue microenvironment and resulting in robust virus transcription, tissue destruction and disease progression. Consistent with such a model, we have observed that RON protein is decreased in brains of patients with AIDS (22). Understanding how HIV-1 is regulated in the different subsets of tissue resident macrophages will be critical for devising ways to purge virus from different tissues.
Disclosures
The authors have no financial conflict of interest.
Footnotes
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
↵1 This work was supported by Penn State Tobacco Formula Funds (to A.J.H.) and by Grants R01 HL066571 (to P.A.H.), GM47477 (to D.S.G.), and AI62467 (to A.J.H.) from the National Institutes of Health.
↵2 A.K. and Z.Z. contributed equally to this work.
↵3 Current address: Division of Basic Sciences, Immunobiology Working Group, Fox Chase Cancer Center, Philadelphia, PA 19111-2497.
↵4 Address correspondence and reprint requests to Dr. Andrew J. Henderson, Department of Medicine, Section of Infectious Diseases, Center for HIV/AIDS Care and Research, Evans Biomedical Research Center, Boston University School of Medicine, 650 Albany Street, Boston, MA 02118-2393. E-mail address: andrew.henderson{at}bmc.org
↵5 Abbreviations used in this paper: LTR, long terminal repeat; TAR, Tat-activating region; ChIP, chromatin immunoprecipitation; NELF, negative elongation factor; Pol II, polymerase II; LM-PCR, ligation-mediated PCR; DSIF, DRB sensitivity-inducing factor; TSA, trichostatin A.
- Received October 12, 2007.
- Accepted November 19, 2007.
- Copyright © 2008 by The American Association of Immunologists