Virus Genes (2007) 35:215–223 DOI 10.1007/s11262-007-0109-9
Induction of chromosomally integrated HIV-1 LTR requires RBF-2 (USF/TFII-I) and RAS/MAPK signaling Tom Malcolm Æ Jiguo Chen Æ Carol Chang Æ Ivan Sadowski
Received: 30 March 2007 / Accepted: 30 April 2007 / Published online: 2 June 2007 Springer Science+Business Media, LLC 2007
Abstract The HIV-1 LTR is regulated by multiple signaling pathways responsive to T cell activation. In this study, we have examined the contribution of the MAPK, calcineurin-NFAT and TNFa-NF-jB pathways on induction of chromosomally integrated HIV-1 LTR reporter genes. We find that induction by T-cell receptor (CD3) cross-linking and PMA is completely dependent upon a binding site for RBF-2 (USF1/2-TFII-I), known as RBEIII at –120. The MAPK pathway is essential for induction of the wild type LTR by these treatments, as the MEK inhibitors PD98059 and U0126 block induction by both PMA treatment and CD3 cross-linking. Stimulation of cells with ionomycin on its own has no effect on the integrated LTR, indicating that calcineurin-NFAT is incapable of causing induction in the absence of additional signals, but stimulation with both PMA and ionomycin produces a synergistic response. In contrast, stimulation of NF-jB by treatment with TNFa causes induction of both the wild type and RBEIII mutant LTRs, an effect that is independent of MAPK signaling. USF1, USF2 and TFII-I from unstimulated cells are capable of binding RBEIII in vitro, and furthermore can be observed on the LTR in vivo by chromatin imunoprecipitation from untreated cells. DNA binding activity of USF1/2 is marginally stimulated by PMA/ ionomycin treatment, and all three factors appear to remain associated with the LTR throughout the course of induction. These results implicate major roles for the MAPK pathway and RBF-2 (USF1/2-TFII-I) in coordinating events necessary for
T. Malcolm J. Chen C. Chang I. Sadowski (&) Department of Biochemistry and Molecular Biology, Molecular Epigenetics, LSI, University of British Columbia, 2350 Health Sciences Mall, Vancouver, BC, Canada V6T 1Z3 e-mail:
[email protected]
transition of latent integrated HIV-1 to active transcription in response to T cell signaling. Keywords RBF-2 USF1 USF2 TFII-I Ras MAPK T cell signaling
Introduction Human immunodeficiency virus type-1 (HIV-1) replication is tightly coupled to T cell activation, and this relationship contributes to pathology of acquired immunodeficiency syndrome (AIDS) [1, 2]. During the course of infection, a fraction of activated T cells are thought to revert to a resting G0 state where transcription of chromosomally integrated HIV-1 becomes repressed to produce a latently infected population [2]. The mechanisms that cause silencing of the integrated provirus are not understood, but are likely to involve formation of repressive chromatin [3], which can be influenced by the site of chromosomal integration [4–7]. Accordingly, the LTR is known to have phased nucleosomes at –160 and immediately downstream of the transcriptional start site (see Fig. 1A, Nuc 0 and Nuc 1) under conditions where transcription is repressed [8, 9]. Despite intense scrutiny focused on the LTR [10–12], the molecular mechanisms which enable transition from a repressed latent state to active transcription in response to T cell signaling are not understood. One reason for this is that most experiments examining transcriptional regulation of the LTR have been performed with transiently transfected templates, which are not efficiently assembled into chromatin [13, 14]. Latent chromosomally integrated provirus can be induced in response to engagement of the T-cell receptor (TCR) during antigen presentation, or by stimulation of
123
216
Fig. 1 (A) Schematic representation of the integrated LTR. Binding sites are indicated for USF, GABP/Ets, NFAT/ NF-kB and SP1. RBEIII binds a USF1/USF2 hetrodimer in conjuction with TFII-I (RBF-2). RBEI, overlapping the initiator binds a factor with properties similar to RBF-2. The location of phased nucleosomes on transcriptionally repressed LTR are indicated (Nuc 0 and Nuc 1). (B) RBEIII-containing oligos bind USF1, USF2 and TFII-I. EMSA was performed using Jurkat nuclear extracts and labeled oTM045/ oTM046 probe (lane 1). Binding reactions were performed in the presence of 20-fold molar excess wild type RBEIII (lane 2) or RBEIII mutant (lane 3) oligo, or antibodies against USF-1 (lane 4), USF-2 (lane 5), TFII-I (lane 6), or YY1 (lane 7). The positions of complexes containing USF1/2, TFII-I and YY1 are indicated with arrows
T-cells with TNFa [2]. Induction of transcription is accompanied by disruption of nucleosome phasing [15]. Stimulation of the TCR activates three parallel downstream signaling pathways through generation of the second messengers inositol triphosphate (IP3) and diacylglycerol (DAG) [16, 17]. PKCh stimulates the Ras-Raf-MEK-ERKMAPK pathway through Ras-GRP [18, 19], and also activates NF-jB by a mechanism involving IjB kinase (IKK) and degradation of the cytoplasmic inhibitor IjB [20]. The RAS/MAPK pathway is known to stimulate AP1 and GABP/ Ets by direct phosphorylation [21]. IP3 causes release of intracellular calcium and activation of calcineurin, which dephosphorylates NFAT to allow nuclear accumulation and activation of responsive genes [22, 23]. Consequently, multiple parallel signals downstream of the TCR regulate activity of four different transcriptional activators that bind elements within the HIV-1 LTR, including GABP/Ets, AP1, NFAT and NF-jB (Fig. 1A), some of which have been shown to function cooperatively for induction of LTR-directed transcription [24, 25]. Additionally, stimulation of T cells with TNFa, can cause induction of the LTR by activation of NF-jB, independently of the other factors [26, 27]. Most studies examining
123
Virus Genes (2007) 35:215–223
T cell signaling pathways and LTR transcription have focused on NFAT and NF-jB, which bind overlapping elements within the enhancer region (Fig. 1A) [28–34]. HIV-1 replication is highly error-prone, and conservation of cis-acting sequences within the 5¢ LTR is likely to reflect a corresponding requirement of their cognate cellular transcription factors for the viral life cycle. In this regard, we have shown that a binding site for RBF-2 at – 120, designated RBEIII, is one of the most highly conserved sequences in viral isolates from AIDS patients [35, 36], along with the core promoter and binding sites for NFAT/NF-jB (Fig. 1A). We have previously shown that RBF-2 is comprised of a USF1/USF2 heterodimer, whose interaction with RBEIII, a non-consensus USF binding site (E box), requires TFII-I as a co-factor [35]. The relationship between the RBEIII element and USF/TFII-I in controlling LTR directed transcription in response to T-cell signaling is not clear, given the defined role of the transcription factors discussed above. Consequently, in this study we have examined the contribution of the parallel signaling pathways downstream of the T cell receptor on expression of an integrated HIV-1 LTR reporter gene and association of USF1/2 and TFII-I at RBEIII during this response. Our results indicate that the RBEIII element is essential for induction of chromosomally integrated LTR by the MAP kinase pathway, but not for induction by TNFa. Furthermore, USF1, USF2 and TFII-I are bound to the LTR in unstimulated cells, and remain associated upon induction. The MAP kinase pathway is essential and sufficient for induction by TCR signaling, but the calcineurin-NFAT pathway produces a synergistic response. Given the previously defined functions for USF and TFII-I, we argue that these factors are necessary to coordinate reorganization of the transcriptionally repressed latent LTR to enable activation by the signal-responsive transcriptional activators.
Materials and methods The HIV-1 LTR-luciferase reporter gene and RBEIII mutant (ACTGCTGA to ACTGCact) were described previously [35]. Jurkat T-lymphoblastoid cells were grown in RPMI 1640 (Gibco) supplemented with 10% heat-inactivated fetal bovine serum (FBS, Gibco) at 37C in a 5% CO2 incubator. Stably integrated LTR-Jurkat cells were established by transfection of linearized wild type and RBEIII mutant LTR reporter plasmids using SuperFect Transfection Reagent (QIAGEN, Germany) according to the manufacturer’s instruction. G418 resistant clones were selected and cloned by limiting dilution. 24 independent clones for each were examined for transcriptional responses to PMA, and one representative clone bearing the wild type and
Virus Genes (2007) 35:215–223
RBEIII mutant LTR reporters were selected for further experiments [35]. Cells were stimulated for the indicated time periods with 50 ng/ml, phorbol-12-myristate-13-acetate (PMA, Sigma) , 25 ng/ml, ionomycin (Sigma) , 1 mM tumor necrosis factor a (TNFa, Calbiochem), and 50 ng/ml trichostatin A (TSA, Sigma). Stimulation by CD3 crosslinking was performed in 96 well plates (Nunc), which were pre-coated with 1.5 lg C305 anti-CD3 antibodies (Cell Signaling, Upstate, NY) in 1X PBS for 2 h at 37C, and then washed 3 times with 1X PBS. 2.4 · 105 Jurkat T cells bearing the wild type and mutant integrated LTR-luciferase reporter constructs were added per well and the plates incubated at 37C for the specified times prior to measuring luciferase activity. The MEK inhibitors PD98059 (Calbiochem) or U0126 (Calbiochem) were added at the specified concentrations 1 h prior to stimulation, and cyclsosporin A (Sigma) at 3 lM 15 min prior to stimulation. Cell lysates were prepared using a luciferase assay kit (Promega, WI) and activities were measured using a microplate luminometer (Turner Designs, CA). Luciferase results represent averages from assays performed on at least three independent cell cultures. Cells for preparation of nuclear extracts from Jurkat cells, treated as described above, were collected by centrifugation, washed twice in ice-cold phosphate buffered saline, and suspended in two packed cell volumes of lowsalt HEPES Buffer (10 mM HEPES pH 7.9, 1.5 mM MgCl2, 10 mM KCl, and 0.5 mM DTT). The cells were gently lysed by passing through a 27 1/2 gauge needle 8 times, and the nuclei collected by pulse centrifugation at 13,000 rpm in a microfuge. Hi-salt HEPES Buffer (10 mM HEPES pH 7.9, 20% v:v glycerol, 1.5 mM MgCl2, 420 mM KCl, 0.2 mM EDTA, 0.5 mM PMSF and 0.5 mM DTT) was added to a volume of 0.6 packed nuclei volumes. The extract was incubated on ice for 15 min prior to the addition of one volume of no-salt HEPES Buffer (10 mM HEPES pH 7.9, 20% v:v glycerol, 0.2 mM EDTA, 0.5 mM PMSF and 0.5 mM DTT). The samples were microfuged 13,000 rpm for 1 min and the nuclear extract removed to a fresh tube. EMSA binding reactions contained 2 pmol labeled double stranded RBEIII probe (annealed oligos oTM045; GATCCTTCAAGAACTGCTG ACATCGAGC TTTCTC and oTM046; GATCGAGAAAGCTCGATGTCAGCAGT TCTTGAAG) 2 lg of poly dI-dC (AP Biotech), 2 lg of bovine serum albumin (NEB) 10 mM HEPES pH 7.9, 100 mM KCl, 5 mM MgCl2, 5% glycerol, and 5 lg nuclear extract protein in a total volume of 20 ll. Unlabeled wild type (oTM045/ oTM046) and RBEIII mutant (annealed oligos P3 M RBEIIIm(+); GATCCTT CAAGAACTGCACTCATCG AGCTTTCTC and P3 M RBEIIIm(-) GATCGAGAAAGCTCGATGAGTGCAGTT CTTGAAG) were added at 20 fold molar excess. Samples were preincubated on ice for 30 min prior to addition of
217
labeled probe, and binding reactions were performed at room temperature for 30 min. Reactions were resolved on 4.5% nondenaturing polyacrylamide gels containing 0.5 · TBE, 20% glycerol at 200 V for 4 h. Gels were fixed in methanol/ acetic acid prior to dying and exposing to X-ray film. Immunoblots were performed using 5 lg nuclear extract protein per lane. Mouse a-TFII-I antibodies were as described previously [35], USF-1, USF-2, YY1 and HDAC3 antibodies were obtained from Santa Cruz Biotechnology, and a-actin antibodies from MPI Biomedicals Inc. Chromatin immunopreciptiation (ChIP) assays were performed as previously described [37]. Jurkat cells bearing stably integrated WT and RBEIII mutant LTR reporters, treated as described above, were washed three times with PBS, and cross linked in PBS containing 1% formaldehyde for 10 min. The cross-linking reaction was quenched by the drop-wise addition of a 1/10 volume of 1.25 M glycine. Cells were washed twice in PBS, once in PBS supplemented with 0.5 mM PMSF, and then resuspended in lysis buffer (50 mM HEPES [pH 7.9], 140 mM NaCl, 1% Triton X-100, 0.5 mM PMSF 1x Protease Inhibitor Cocktail (Sigma)). DNA was sheared by sonication, and the suspension was centrifuged at maximum speed in a microcentrifuge for 10 min at 4C. The supernatant, designated soluble chromatin, was subjected to immunoprecipitation overnight at 4C with antibodies against USF1, USF2, TFII-I, and HDAC3. Immune complexes were washed three times sequentially in: high salt buffer (50 mM HEPES [pH 7.9], 500 mM NaCl, 1% Triton X-100), wash buffer (50 mM HEPES [pH 7.9], 250 mM LiCl, 0.5% NP-40), and then TE (10 mM Tris [pH 8.0], 1 mM EDTA) all containing 1x PIC. DNA was eluted twice in 50 mM Tris [pH 8.0], 1% SDS, and 10 mM EDTA for 10 min at 65C, the eluates were pooled and DNA precipitated with ethanol. Purified DNA was dissolved in 30 ll TE, and 5 ll was used per PCR reaction. HIV-1 LTR primer sequences were ChIP-F: 5¢ CGCGGAGAAAGAA GTGTTAG-3¢ and ChIP-R: 5¢-GAGCTCCCAGGCTCAG ATCT-3¢, and the b-globin control primers were globin-F; TATCTTAGAGGGAGGGCTGAGGGTTTG, and globinR; CCAACTTCATCCCAGTTCACCTTGC.
Results TFII-I promotes binding of USF1/2 heterodimers to RBEIII The upstream binding site for RBF-2 at –120, designated RBEIII, is one of the most highly conserved cis-elements on the 5¢ LTR in patients with AIDS [36] (Fig. 1A). We have previously shown that RBF-2 is comprised of a USF1/
123
218
Virus Genes (2007) 35:215–223
USF2 heterodimer, whose interaction with the non-canonical E box RBEIII (Fig. 1A) requires the multifunctional factor TFII-I [35]. USF1/2 heterodimers from Jurkat T-cell nuclear extracts form complexes with oligonucleotides spanning RBEIII, which can be super shifted with antibodies against USF1 or USF2 (Fig. 1B, lanes 4 and 5). Binding of USF1/2 to RBEIII is inhibited by mutation of the RBEIII consensus ACTGCTGA to ACTGCact (mRBEIII), as this complex can be eliminated by competitor wild type oligo (lane 2) but not mutant competitor oligo (lane3). Furthermore, binding of USF1/2 to RBEIII can be prevented in vitro by antibodies against TFII-I (lane 6), consistent with our previous observations that TFII-I can stimulate binding of USF1/2 to the non-canonical RBEII element [35]. RBEIII is necessary for induction of integrated HIV-1 LTR by T cell signaling To determine the requirement of USF/TFII-I bound at RBEIII for transcriptional response of the LTR to T-cell signaling pathways, we constructed Jurkat cell lines bearing integrated HIV-1 LTR-luciferase reporter genes with wild type or mutant RBEIII element (RBEIIImut). We consistently observe that integrated RBEIII mutant LTRs produce higher basal expression than the wild type (Fig. 2A, Untreated), but both the wild type and RBEIII mutant LTRs are induced to comparable levels upon stimulation with TNFa, which causes activation of NF-jB (Fig. 2A, TNFa) [27]. In contrast, the RBEIII mutation prevents induction in response to T-cell receptor crosslinking (TCR), which mimics engagement of the receptor during antigen presentation (Fig. 2A) [38]. The LTR can be induced to maximal levels by treatment with PMA and ionomycin, which stimulate the RAS-MAPK and calcineurin-NFAT pathways, in combination with the histone deacetylase inhibitor trichostatin A [30]. Interestingly, we observed that unlike the wild type LTR, the integrated RBEIII mutant LTR is unresponsive to such treatment (Fig. 2A), indicating that USF/TFII-I bound at RBEIII is necessary for response to stimulation of the RAS-MAPK and calcineurin-NFAT pathways. The wild type LTR can be induced approximately by 5-fold treatment with PMA alone, but is unresponsive to separate treatment with either ionomycin or trichostatin A, or with ionomycin in combination with TSA (Fig. 2B). Treatment with PMA in combination with either ionomocyin or TSA produces an intermediate level of expression, which supports the view that these agonists stimulate parallel pathways for induction of RBEIII-dependent expression from the HIV-1 LTR.
123
Fig. 2 (A) RBEIII is necessary for induction of integrated LTR by T cell signaling. Representative Jurkat T cell lines bearing integrated wild type (WT LTR, open bars) or RBEIII mutant LTR (RBEIIImut LTR, solid bars) -luciferase reporter genes were left untreated, or stimulated with a combination of PMA/TSA/ionomycin, CD3 TCR crosslinking, or TNFa. (B) PMA, but not ionomycin, can induce wild type integrated HIV-1 LTR. Jurkat cells bearing the integrated wild type LTR-luciferase reporter were treated with combinations of PMA, Ionomocyin and TSA as indicated below. Cells were harvested 4 h post treatment for measurement of luciferase activity
The MAPK pathway is essential for induction of integrated HIV-1 LTR by CD3 cross-linking and stimulation with PMA We compared induction kinetics of the integrated wild type LTR by treatment with PMA and cross-linking with antibodies against CD3 (TCRc/e), and found that although PMA caused more robust expression overall, both produced maximal induction at approximately 4–8 h of post treatment (Fig. 3A). In T-cells, PMA causes activation of the RAS-MAPK pathway through RAS-GRP [19, 39], but can additionally activate NF-jB through PKCh [17, 20]. Induction of the wild type LTR by PMA and CD3 crosslinking appears to predominately require the RAS-MAPK pathway, because both were inhibited in a dose-dependent manner by PD98059 and U0126, small molecule inhibitors of the MAP/ERK Kinases (MEK) (Fig. 3B, C). U0126
Virus Genes (2007) 35:215–223
219
Fig. 3 MAPK/ERK kinase (MEK) inhibitors prevent induction of the wild type HIV-1 LTR by PMA and CD3 crosslinking. (A) Jurkat cells bearing the wild type LTR-luciferase reporter were stimulated with PMA (s) or CD3 cross-linking (h) for the indicated times prior to harvesting cells and measurement of luciferase activity. (B) and (C)
Wild type LTR-luciferase Jurkat cells were pre-treated with the indicated concentration of MEK inhibitors PD98059 (d) or U0126 (n), and then stimulated with PMA (Panel B) or CD3 crosslinking (Panel C) for 4 h prior to measuring luciferase activity
completely blocks induction by PMA at 40 lM, and by CD3 crosslinking at 120 lM. These results support the view that activation of the RAS-MAPK pathway is critical for induction of latently integrated LTR [28, 33], and consistent with the fact that RBEII was originally identified by its requirement for response of the LTR to oncogenic vHa-Ras [40].
response which can be blocked by the calcineurin inhibitor cyclosporin A (Fig. 4). Induction of the LTR by PMA/ ionomycin is limited to approximately by 8-fold in cells treated with PD98059 in combination with cyclosporin A (Fig. 3C). This residual induction in the presence of both inhibitors likely reflects the extent that PMA activates NFjB through PKCh [20]. These results demonstrate the parallel pathways downstream of the T cell receptor that contribute to induction of the integrated HIV-1 LTR to produce the transition from a repressed latent state, and that of these, the MAPK pathway is essential for induction by PMA and engagement of the T cell receptor.
Calcineurin-NFAT causes synergistic induction of the LTR with the MAPK pathway Dephosphorylation of NFAT by calcinuerin allows translocation to the nucleus to enable activation of responsive genes. We find that activation of calcinuerin-NFAT by ionomycin on its own has no effect on expression from the integrated wild type LTR (Fig. 2B, Fig. 4, Ion). However, costimulation of cells bearing the wild type LTR with PMA in combination with ionomycin produces a synergistic
Fig. 4 MAPK and calcineurin-NFAT cause synergistic induction of the LTR. Wild type LTR-luciferase Jurkat cells were stimulated with PMA and/or ionomycin in the presence of the inhibitors PD9805, U0126, and cyclosporin A as indicated below. Cells were harvested 4 h poststimulation for measurement of luciferase activity
RBEIII is constitutively occupied by USF/TFII-I Having established that induction of the integrated LTR requires RBEIII, we examined whether occupancy of this site was affected by signals produced by treatment with PMA and/or ionomcyin. For this purpose, nuclear extracts prepared from Jurkat cells stimulated with PMA or ionomycin were used in EMSA with an oligonucleotide probe spanning the RBEIII element. We found that PMA and ionomycin produced a small but detectable increase in the USF and TFII-I-specific complexes with the RBEIII oligo after 4 and 8 h post treatment (Fig. 5A, lanes 4–9). Additionally, cells treated with ionomycin for 24 h, alone or in combination with PMA (Fig. 5A, lanes 11 and 12), were found to produce significantly more USF1/2 heterodimer and TFII-I-specific complexes. This appears to be a consequence of accumulation of USF1 protein, since a significantly greater amount of USF1 could be detected in these extracts by immunoblotting (Fig. 5B, lanes 9 and 10). In contrast, expression levels of USF2 and TFII-I remain constant until at least 48 h post treatment (Fig. 5B). We have not determined the mechanism for accumulation of USF1 under these conditions, but because this effect is only
123
220
Virus Genes (2007) 35:215–223
Fig. 6 RBF-2 (USF1/2-TFII-I) are associated with the integrated wild type HIV-1 LTR in unstimulated cells. Chromatin immunopreciptiation was performed with Jurkat cells bearing the wild type (Wt) or RBEIII mutant (Mu) LTR reporter genes. Cells were untreated (–) or stimulated with PMA/ Ionomycin/ TSA (+) for 8 h prior to formaldehyde crosslinking. Extracts were immunoprecipitated with antibodies against USF1 (lanes 1–4), USF2 (lanes 5–8), TFII-I (lanes 9–12), or HDAC3 (lanes 13–16). Immunoprecipitated template was analyzed by PCR with oligos specific for the HIV-1 LTR (RBEIII) and b-globin. Lane 17 contains a representative chromatin input sample from unstimulated Jurkat cells bearing the wild type LTR
Fig. 5 RBF-2 (USF1/2-TFII-I) from unstimulated nuclear extracts binds RBEIII. (A) Nuclear extracts were prepared from unstimulated Jurkat cells (lanes 1–3), or cells stimulated with PMA (lanes 4, 7, 10, 13), Ionomycin (lanes 5, 8, 11, 14), or both PMA and Ionomycin (lanes 6, 9, 12, 15) for the indicated times. EMSA was performed with labeled oTM045/oTM046 probe (lane 1). 20fold molar excess unlabeled wild type RBEIII or RBEIII mutant competitor oligo was added to the binding reactions in lanes 2 and 3, respectively. The positions of complexes containing USF1/2, TFII-I and YY1 are indicated with arrows. (B) Nuclear extracts described in Panel A were analysed by immunoblotting with antibodies against TFII-I, USF-1, USF-2, and actin
observed 16–20 h after transcription levels have peaked (Fig. 3A) it must not contribute to induction of the LTR in response to T-cell signaling. Nevertheless, because the levels of USF1 appear to parallel the USF1/2 heterodimer and TFII-I DNA binding activities in these extracts, it seems USF1 may be the limiting factor for formation of these complexes in the Jurkat cell line. TFII-I, USF1 and USF2 were also found to be constitutively bound to the LTR in vivo as determined by chromatin immunopreciptiation. In unstimulated cells, we observed all three factors bound to the wild type LTR, but significantly less on the RBEIII mutant LTR (Fig. 6). This result is consistent with those of Fig. 1B, demonstrating that interaction of these factors with this upstream site requires the RBEIII consensus sequence. Stimulation of cells with PMA/ionomycin and TSA does not alter interaction of the factors with the wild type LTR, but does cause a
123
detectable amount of recruitment to the mutant LTR (Fig. 5). This likely reflects the fact that the RBEIII mutation does not affect the upstream consensus binding site for USF (Fig. 1A), and also that TFII-I/ USF likely interact with RBEI (Fig. 1A), which overlaps the initiator element [41, 42]. TFII-I is known to have a repressive effect on transcription though a direct physical interaction with histone deacetylase 3 (HDAC 3) [43]. Consequently, we also examined association of HDAC3 with the integrated LTRs, and we observed interaction with both the wild type and RBEIII mutant LTRs in unstimulated cells. This indicates that RBEIII is not specifically required for recruitment of HDAC3, although as noted above, TFII-I is also bound to the initiator element. Treatment with PMA/ ionomycin/TSA caused loss of HDAC3 from both the wild type and RBEIII mutant LTRs (Fig. 6). Consequently, HDAC3 can apparently become dissociated from the LTR independently of additional events required to cause induction. In this respect, it is also interesting to note that treatment of cells with the HDAC inhibitor trichostatin on its own, has no effect on the integrated wild type LTR. This reflects the fact that transition of the LTR from a repressed latent state is likely to involve multiple alterations in chromatin modification and organization, in addition to recruitment of co-activator and general transcription factor complexes. Importantly, the fact that USF1/2 and TFII-I are bound to the repressed LTR, and remain associated upon induction implies that these factors must play important functions in coordinating these events.
Discussion A hallmark of HIV-1 infection is the production of CD4+ T lymphocytes bearing transcriptionally repressed provirus,
Virus Genes (2007) 35:215–223
which represent a reservoir that is impenetrable by highly active antiretroviral therapy (HAART) [44, 45]. Understanding the mechanisms controlling the establishment and maintenance of latency, as well as those enabling reactivation of viral transcription by T-cell signaling will be important for developing strategies for purging this latent population. In this study we have examined signaling pathways that control induction of a chromosomally integrated HIV LTR and find that activation of the MAPK pathway is essential for induction of the LTR in cells stimulated by treatment with PMA or T cell receptor crosslinking. Stimulation of calcineurin-NFAT produces a synergistic response in combination with the MAPK pathway, but is incapable of causing induction of LTR transcription on its own. Induction of the LTR by PMA or T cell receptor cross-linking is also completely dependent upon an intact RBEIII element, the upstream binding site for RBF-2, represented by a USF1/USF2 heterodimer and TFII-I [35]. In contrast, treatment with TNFa, which stimulates NF-jB through IKK, could cause induction of the LTR, albeit to a lesser extent, largely independent of MAPK signaling, and the upstream RBEIII element. These results reveal several important features of the transcription factors known to bind the 5¢ LTR in vivo. Several previous reports have implicated a central role for MAP kinases in regulation of HIV-1 transcription. In U1 cells, which harbor a latent HIV-1 proviral genome, the MEK inhibitors PD98059 and U0126 were shown to inhibit induction of HIV-1 gene expression by both PMA and TNFa [46]. Responsiveness of the LTR to MAPK signaling was reported to involve cooperative interaction between AP-1 and NF-jB within the enhancer region [33]. Similarly, the p38 MAP kinase inhibitor SB203580 was found to inhibit induction of transiently transfected HIV-1 LTR reporter constructs by UV, IL1, TNFa and high-osmolarity treatment of HeLa cells; for induction by cytokines this effect requires the enhancer region containing the binding sites for NFAT/NF-jB [47] (see Fig. 1A). However, signaling of the MAPK pathway(s) to the HIV-1 LTR cannot be limited to effects on AP-1/NF-jB. LSF, which binds near the transcriptional start site in conjunction with YY1, plays an important role in regulating establishment of latency by recruitment of HDAC1 [48]. Interestingly, LSF is differentially regulated by p38 MAPK and ERK phosphorylation; Erk inhibits, but p38 enhances binding of LSF to the LTR in vitro and in vivo [49]. Therefore, the function of MAPK signaling for regulation of transcription from the LTR involves multiple transcription factors, and importantly is not limited to control by factors bound to the enhancer region (Fig. 1A). Our results provide additional evidence that MAPK signaling must regulate multiple distinct events on the LTR to enable transcription in response to T-cell signaling. The
221
RBEIII element is essential for response of the LTR to stimulation of T cells with PMA or CD3 cross-linking, and specifically to signals transmitted by the MAPK pathway. This is consistent with the fact that RBEIII was initially identified as one of four elements necessary for response of the LTR to cotransfected v-Ha-Ras expression plasmids [40]. RBEI overlaps the initiator element (Fig. 1A), and binds a factor with identical sequence specificity and mobility as RBF2, bound at RBEIII [40]. This is consistent with the observation that TFII-I was observed to be bound cooperatively with USF to the initiator elements of the AdML and HIV-1 LTR promoters [41, 42]. Two further elements, designated RBEII, flanking the NFAT/ NF-jB sites within the enhancer, and RBEIV, positioned further upstream at –180 were found to bind the Ets family member GABP [36], consistent with the defined role of GABP/ Ets factors in mediating tyrosine kinase-Rasresponsive transcription [50]. Identification of RBF-2 as a USF1/ USF2 heterodimer was surprising,considering that these factors had not previously been implicated in mediating response to Ras or MAPK signaling. However, a dominant negative USF1 derivative was shown to block oncogenic transformation by v-Ras [51], which supports a view that these factors must be downstream of MAPK signaling for regulation of some growth factor-responsive genes. Binding of USF1/ USF2 to RBEIII is dependent upon TFII-I, and furthermore, TFII-I itself forms complexes with RBEIII-containing oligonucleotides (Fig. 1B), although we have yet to identify the precise residues mediating this interaction. Unlike USF, TFII-I has been shown to be involved in tyrosine-kinase - RAS responsive transcription, in cooperation with the serum response factor (SRF) [52–57], although mechanistic details of this relationship have not been elucidated. The RBEIII element on its own inserted upstream of a minimal promoter has only a marginal effect on transcription, even when concatamerized [35], which implies that USF1/2 and TFII-I bound at that element likely do not function as classical transcriptional activators, but rather must perform function(s) that are dependent upon a specific promoter context. The fact that RBEIII appears to be stringently conserved, both spatially and in sequence, suggest that USF1/2 and TFII-I bound specifically at –120 may play a critical role in organizing the integrated LTR into a specific conformation, perhaps by regulating positioning or modification of nucleosomes [36] (Fig. 1A). Consistent with this possibility, USF was found to be present in complexes bound to the chicken b-globin 5¢HS4 insulator element, which prevents spreading of heterochromatin [58]. Furthermore, TFII-I has also been shown to participate in chromatin modification and repression through direct interaction with HDAC3 [43, 59].
123
222
Our results demonstrate distinct differences among the various transcription factors known to regulate the LTR with respect to their ability to promote induction from latency. Stimulation of cells with ionomycin on its own has no effect on induction of the wild type LTR, which implies that NFAT is either incapable of binding to the integrated repressed LTR, or is unable to recruit factors necessary to promote changes in chromatin organization that would enable induction. We suspect that the former possibility is more likely, because ionomycin in combination with trichostatin A is also incapable of causing induction (Fig. 2B). In contrast, TNFa is capable of causing transcription from the LTR, largely independently of MAPK signaling or the RBEIII element (Fig. 2A and data not shown). This suggests that, unlike NFAT, NF-jB must be capable of binding the repressed LTR, and on its own promotes recruitment of factors sufficient to induce transition from latency. AP1 and GABP/ Ets, known downstream targets of the MAPK pathway [21, 60], stimulated by PMA and TCR cross-linking [21, 33], represent an intermediate capability. These factors appear to be dependent upon RBEIII for induction, but otherwise must be capable of binding and activating transcription of the repressed LTR. Considering the available evidence, it seems that induction of chromosomally integrated HIV-1 from latency requires a combination of molecular events to relieve the repressive effects of phased nucleosomes, cause recruitment of the general transcription factor complexes and promote initiation and elongation from the LTR promoter. How might USF1/2 and TFII-I bound at RBEIII contribute to this process? In a previous study, we found that USF1, USF2 and TFII-I all become hyperphosphorylated in response to treatment with PMA [35]. These proteins are bound to the LTR in vivo in unstimulated cells (Fig. 6), their DNA binding activity is not altered by treatment with PMA (Fig. 5A), and they remain associated with the LTR during induction (Fig. 6). Chromosomally integrated RBEIII mutant LTRs consistently produce elevated basal expression in unstimulated cells relative to wild type (Fig. 2A) [35], and consequently USF/TFII-I bound to RBEIII may contribute to repression of the latent LTR. Phosphorylation of these proteins in response to T-cell signaling and the MAP kinase pathway might modify their function to cause relief of repression and enable recruitment of factors necessary to promote reorganization of nucleosomes on the LTR, subsequently allowing transactivation by NFAT, AP1, GABP/Ets and NF-jB [24, 25, 33]. Our observations suggest a role for RBF-2 (USF/TFIII) in the establishment of HIV-1 latency through a repressive function in unstimulated cells, and in chromatin remodeling in response to activation of the Ras/MAPK pathway in activated T-cells. Consequently, this transcription factor complex may represent a potentially
123
Virus Genes (2007) 35:215–223
important target for therapies to eliminate the latent viral reservoir. Acknowledgments This research was supported by funds from the Canadian Institutes for Health Research (MOP-77807). We thank Hung-Sia Teh and Ed Kim for advice on CD3 cross-linking experiments.
References 1. Y. Han, M. Wind-Rotolo, H.C. Yang, J.D. Siliciano, R.F. Siliciano, Nat. Rev. Microbiol. 5, 95 (2007) 2. D. Persaud, Y. Zhou, J.M. Siliciano, R.F. Siliciano, J. Virol. 77, 1659 (2003) 3. A. Jordan, P. Defechereux, E. Verdin, Embo J. 20, 1726 (2001) 4. A.R. Schroder, P. Shinn, H. Chen, C. Berry, J.R. Ecker, F. Bushman, Cell 110, 521 (2002) 5. H. Soudeyns, G. Pantaleo, Immunol. Today 20, 446 (1999) 6. J.M. McCune, Cell 82, 183 (1995) 7. A. Jordan, D. Bisgrove, E. Verdin, Embo J. 22, 1868 (2003) 8. S.A. Stanfield-Oakley, J.D. Griffith, J. Mol. Biol. 256, 503 (1996) 9. G. He, L. Ylisastigui, D.M. Margolis, DNA Cell. Biol. 21, 697 (2002) 10. K. A. Jones, B.M. Peterlin, Annu. Rev. Biochem. 63, 717 (1994) 11. G. J. Nabel, Human Retroviruses, (Oxford university press, 1993), p. 49 12. L.A. Pereira, K. Bentley, A. Peeters, M.J. Churchill, N.J. Deacon, Nucleic Acids Res. 28, 663 (2000) 13. T.K. Archer, C.J. Fryer, H.L. Lee, E. Zaniewski, T. Liang, J.S. Mymryk, J. Steroid Biochem Mol. Biol. 53, 421 (1995) 14. C. Fryer, T. Archer, Nature 393, 88 (1998) 15. E. Verdin, P. Paras Jr., C. Van Lint, Embo J. 12, 3249 (1993) 16. Y. Huang, R.L. Wange, J. Biol. Chem. 279, 28827 (2004) 17. N. Isakov, A. Altman, Annu. Rev. Immunol. 20, 761 (2002) 18. J.O. Ebinu, S.L. Stang, C. Teixeira, D.A. Bottorff, J. Hooton, P.M. Blumberg, M. Barry, R.C. Bleakley, H.L. Ostergaard, J.C. Stone, Blood 95, 3199 (2000) 19. J.O. Ebinu, D.A. Bottorff, E.Y. Chan, S.L. Stang, R.J. Dunn, J.C. Stone, Science 280, 1082 (1998) 20. N. Coudronniere, M. Villalba, N. Englund, A. Altman, Proc. Natl. Acad. Sci. USA 97, 3394 (2000) 21. E. Flory, A. Hoffmeyer, U. Smola, U.R. Rapp, J.T. Bruder, J. Virol. 70, 2260 (1996) 22. N.A. Clipstone, G.R. Crabtree, Nature 357, 695 (1992) 23. A. Rao, C. Luo, P.G. Hogan, Annu. Rev. Immunol. 15, 707 (1997) 24. M.H. Sieweke, H. Tekotte, U. Jarosch, T. Graf, Embo J. 17, 1728 (1998) 25. A.G. Bassuk, R.T. Anandappa, J.M. Leiden, J Virol 71, 3563 (1997) 26. B.A. Antoni, A.B. Rabson, A. Kinter, M. Bodkin, G. Poli, Virology 202, 684 (1994) 27. O.N. Ozes, L.D. Mayo, J.A. Gustin, S.R. Pfeffer, L.M. Pfeffer, D.B. Donner, Nature 401, 82 (1999) 28. D.G. Brooks, P.A. Arlen, L. Gao, C.M. Kitchen, J.A. Zack, Proc. Natl. Acad. Sci. USA 100, 12955 (2003) 29. E. Flory, C.K. Weber, P. Chen, A. Hoffmeyer, C. Jassoy, U.R. Rapp, J. Virol. 72, 2788 (1998) 30. A. El Kharroubi, G. Piras, R. Zensen, M.A. Martin, Mol. Cell. Biol. 18, 2535 (1998) 31. A.M. Lemieux, M.E. Pare, B. Audet, E. Legault, S. Lefort, N. Boucher, S. Landry, T. van Opijnen, B. Berkhout, M.H. Naghavi, M.J. Tremblay, B. Barbeau, J. Biol. Chem. 279, 52949 (2004)
Virus Genes (2007) 35:215–223 32. V. Quivy, E. Adam, Y. Collette, D. Demonte, A. Chariot, C. Vanhulle, B. Berkhout, R. Castellano, Y. de Launoit, A. Burny, J. Piette, V. Bours, C. Van Lint, J. Virol. 76, 11091 (2002) 33. X. Yang, Y. Chen, D. Gabuzda, J. Biol. Chem. 274, 27981 (1999) 34. M.J. Giffin, J.C. Stroud, D.L. Bates, K.D. von Koenig, J. Hardin, L. Chen, Nat. Struct. Biol. 10, 800 (2003) 35. J. Chen, T. Malcolm, M.C. Estable, R.G. Roeder, I. Sadowski, J. Virol. 79, 4396 (2005) 36. I. Sadowski, D.A. Mitchell, Eur. J. Cancer. 41, 2528 (2005) 37. B. Bell, E. Scheer, L. Tora, Mol. Cell. 8, 591 (2001) 38. C. Barat, M.J. Tremblay, J. Biol. Chem. 277, 28714 (2002) 39. R. Marais, Y. Light, C. Mason, H. Paterson, M.F. Olson, C.J. Marshall, Science 280, 109 (1998) 40. B. Bell, I. Sadowski, Oncogene 13, 2687 (1996) 41. A.L. Roy, H. Du, P.D. Gregor, C.D. Novina, E. Martinez, R.G. Roeder, Embo J. 16, 7091 (1997) 42. A.L. Roy, M. Meisterernst, P. Pognonec, R.G. Roeder, Nature 354, 245 (1991) 43. Y.D. Wen, W.D. Cress, A.L. Roy, E. Seto, J. Biol. Chem. 278, 1841 (2003) 44. R.C. Desrosiers, Nat. Med. 5, 723 (1999) 45. R.A. Koup, Nat. Med. 7, 404 (2001) 46. Y. Qyang, X. Luo, T. Lu, P.M. Ismail, D. Krylov, C. Vinson, M. Sawadogo, Mol. Cell. Biol. 19, 1508 (1999) 47. S. Kumar, M.J. Orsini, J.C. Lee, P.C. McDonnell, C. Debouck, P.R. Young, J. Biol. Chem. 271, 30864 (1996)
223 48. J.J. Coull, F. Romerio, J.M. Sun, J.L. Volker, K.M. Galvin, J.R. Davie, Y. Shi, U. Hansen, D.M. Margolis, J. Virol. 74, 6790 (2000) 49. L. Ylisastigui, R. Kaur, H. Johnson, J. Volker, G. He, U. Hansen, D. Margolis, J. Virol. 79, 5952 (2005) 50. A. Verger, M. Duterque-Coquillaud, Bioessays 24, 362 (2002) 51. C. Aperlo, K.E. Boulukos, P. Pognonec, Eur. J. Biochem. 241, 249 (1996) 52. V. Cheriyath, Z.P. Desgranges, A.L. Roy, J. Biol. Chem. 277, 22798 (2002) 53. D. A. Grueneberg, R.W. Henry, A. Brauer, C.D. Novina, V. Cheriyath, A.L. Roy, M. Gilman, Genes Dev. 11, 2482 (1997) 54. D.W. Kim, V. Cheriyath, A.L. Roy, B.H. Cochran, Mol. Cell. Biol. 18, 3310 (1998) 55. D.W. Kim, B.H. Cochran, Mol. Cell. Biol. 20, 1140 (2000) 56. D.W. Kim, B.H. Cochran, Mol. Cell. Biol. 21, 3387 (2001) 57. C. Sacristan, M.I. Tussie-Luna, S.M. Logan, A.L. Roy, J. Biol. Chem. 279, 7147 (2004) 58. A.G. West, S. Huang, M. Gaszner, M.D. Litt, G. Felsenfeld, Mol. Cell. 16, 453 (2004) 59. M.I. Tussie-Luna, D. Bayarsaihan, E. Seto, F.H. Ruddle, A.L. Roy, Proc. Natl. Acad. Sci. USA 99, 12807 (2002) 60. S.H. Yang, A.D. Sharrocks, A.J. Whitmarsh, Gene 320, 3 (2003)
123
