TFII-I Regulates Induction of Chromosomally Integrated Human Immunodeficiency Virus Type 1 Long Terminal Repeat in Cooperation with USF

Recombinant DNA molecules.Details of plasmids for the expression of proteins in insect and mammalian cells and of the reporter genes will be described elsewhere. Plasmids expressing USF-2 (pCDNA-USF2) and the dominant negative USF (pCMV566-A-USF), TFII-I (pEB-P70), and I-κB (pSVK3-IKBα2N) proteins were described previously (7, 23, 33).

Cell culture, transfections, metabolic labeling, and immunoprecipitation.T-lymphoblastoid Jurkat cells were grown in RPMI 1640, and COS-1 cells were grown in Dulbecco's modified Eagle medium supplemented with 10% heat-inactivated fetal bovine serum at 37°C in a 5% CO₂ atmosphere. Jurkat cell lines with stably integrated wild-type (WT) and RBEIII mutant HIV-1 LTR-luciferase reporters were established by transfection of the pGLneo-WT or pGLneo-M3 plasmid by use of the SuperFect transfection reagent (QIAGEN). The cells were washed with phosphate-buffered saline 48 h later, transferred to medium containing 800 μg of G-418 sulfate (Promega) per ml, and maintained with selection for 1 month. Cells were then grown as pools (WP and MP) in medium containing 100 μg of G-418 per ml or were cloned by limiting dilution. Jurkat cells stably expressing USF1-Flag were generated similarly by transfection with pcDNA-USF1. Transient transfections of Jurkat and COS-1 cells were performed by use of the SuperFect transfection reagent, and transfection efficiencies were normalized by cotransfection of an internal control plasmid, pCMV-galactosidase; activities were measured by use of a luciferase assay system (Promega), a luminescent β-galactosidase detection kit II (Clontech), and a microplate luminometer (Turner Designs) from a minimum of three separate transfections. To enrich for cells expressing dominant negative proteins, we cotransfected cells with the relevant expression plasmid and pEGFP (Clontech) at a 10:1 molar ratio. At 48 h posttransfection, green fluorescent protein (GFP)-expressing cells were isolated by fluorescence-activated cell sorting (FACS), and the luciferase activity was measured after treatment with phorbol-12-myristate-13-acetate (PMA; Sigma) and ionomycin (Sigma). Cells were stimulated with PMA at 25 ng/ml, ionomycin at 1 μM, tumor necrosis factor alpha (TNF-α; Sigma) at 10 ng/ml, and trichostatin A (TSA; Sigma) at 50 ng/ml. Stimulation by T-cell receptor cross-linking was performed as described previously (48) by use of the monoclonal antibody C305 against CD3. Cells were stimulated for 24 h before harvesting and measurement of luciferase activity. T-cell activation was monitored by FACS analysis with an anti-human CD69-phycoerythrin (PE) conjugate (BD Biosciences) (48).

For metabolic labeling, 7 × 10⁶ Jurkat cells were washed in phosphate-depleted medium (Gibco) and incubated in the same medium for 2 h prior to the addition of 0.1 mCi of [³²P]orthophosphate (ICN) per ml. Cells were labeled for a total of 3 h, with the addition of PMA 1 h prior to harvesting. The cells were lysed in RIPA buffer containing phosphatase inhibitors and protease inhibitor cocktail (Sigma), and USF1, USF2, and TFII-I were immunoprecipitated and analyzed by tryptic phosphopeptide analysis (39).

Protein expression and purification.Nuclear extraction from Jurkat cells and the purification of RBF-2 were performed as described previously (2, 14). USF proteins were produced by in vitro transcription and translation with the TNT T7 system (Promega), using template DNAs amplified from pFastbac-USF vectors with the primers indicated in Table 2. For the production of recombinant proteins in insect cells, SF9 cells were infected at a multiplicity of infection of 10 and harvested 4 days later. The cells were washed in phosphate-buffered saline and then lysed in buffer C (10 mM HEPES [pH 7.9], 20% [vol/vol] glycerol, 1.5 mM MgCl₂, 420 mM KCl, 0.2 mM EDTA, 0.5 mM dithiothreitol, 0.5 mM phenylmethylsulfonyl fluoride) by sonication. An equal volume of buffer D (10 mM HEPES [pH 7.9], 20% [vol/vol] glycerol, 0.2 mM EDTA, 0.5 mM dithiothreitol, 0.5 mM phenylmethylsulfonyl fluoride) was added, and the extracts were clarified by centrifugation.

Immunoblotting, electrophoretic mobility shift, and chromatin immunoprecipitation assays.Rabbit polyclonal antibodies against USF1, USF2, c-fos, JunB, Myc-1, YY-1, ERK-1, Non-O, PSF, STAT-1, SRF, XBP-1, NF-κB, β-actin, and LexA and a mouse monoclonal antibody against Myc-1 were purchased from Santa Cruz Biotech. Polyclonal antibodies against TFII-I were produced against a glutathione S-transferase--TFII-I fusion protein (36) in mice. Electrophoretic mobility shift assay (EMSA) binding reactions were performed as described previously (2) and typically contained 2 pmol of labeled oligonucleotide probe, 2 μg of poly(dI-dC) (Sigma), 2 mg of bovine serum albumin, 10 mM HEPES [pH 7.9], 100 mM KCl, 5 mM MgCl₂, 5% glycerol, and 2 to 5 μg of Jurkat nuclear extract, 0.1 to 1 μg of purified protein fractions, 4 μl of in vitro translation mix, or SF9 extract, as indicated, in a total volume of 20 μl. Labeled oligonucleotide probes were typically added to binding reactions after a 20-min preincubation of nuclear or insect cell extracts with the reaction components plus competitor oligonucleotides or specific antibodies. The sequences of the WT and mutant RBEIII-containing oligonucleotides P3 and C1 are shown in Fig. 2A. The E-box-containing oligonucleotide oJC033 (5′-CTAGAAAGACACGTGACATCGAGCTTTCTCCAAG-3′) was used as a probe and competitor for the experiment shown in Fig. 4. Chromatin immunoprecipitations were performed with the W26 and M2 Jurkat cell lines bearing WT or mutant HIV-1 LTR-luciferase reporters as described previously (3). Primers for the amplification of the HIV-1 LTR upstream region and β-globin promoter are indicated in Table 2 and produce 220- and 321-bp products, respectively.

• Open in new tab
• Download powerpoint

FIG. 2.

Antibodies against USF1, USF2, and TFII-I inhibit DNA binding of RBF-2. (A) Sequences of oligonucleotides used for EMSAs. The location of the RBEIII element is boxed. Nucleotide substitutions in the CM1 and P3M mutant oligonucleotides are indicated in lowercase bold letters. (B) Purified RBF-2 was analyzed by EMSA with the P3 oligonucleotide probe. The unlabeled competitor P3 (lane 2) or P3M (lane 3) oligonucleotide was added at a 100-fold molar excess. Antibodies against the indicated proteins (top) were added to the binding reactions (lanes 4 to 15). The migration of RBF-2 is indicated (left). (C) A Jurkat nuclear extract was analyzed by EMSA with the P3 oligonucleotide probe. The unlabeled competitor C1 (lane 2), CM1 (lane 3), P3 (lane 4), or P3M (lane 5) oligonucleotide was added at a 100-fold molar excess. Antibodies against the indicated proteins (top) were added to the binding reactions (lanes 6 to 13). In addition to RBF-2, YY1 bound to the 3′ end of the P3 oligonucleotide from Jurkat nuclear extracts (indicated with arrows).

RBF-2 is comprised of USF1, USF2, and TFII-I.RBF-2 was identified as a nuclear factor that binds specifically to the highly conserved RBEIII cis element (ACTGCTGA) (Fig. 1A), which was shown to be necessary for the responsiveness of HIV-1 to oncogenic protein tyrosine kinases and v-Ras (2, 12). Purified RBF-2 (14) bound specifically to RBEIII oligonucleotides in EMSAs (Fig. 1B, lane 8, and Fig. 2B, lane 1). Mutation of the core RBEIII sequence to ACTGCACT prevents the interaction of purified RBF-2 with RBEIII, as an oligonucleotide containing this mutation (P3M) (Fig. 2A) was unable to compete for binding with a WT probe in EMSAs (Fig. 2B, compare lanes 2 and 3). The P3M mutation impairs the responsiveness of transiently transfected LTR reporter templates to v-Ras (2, 12), consistent with the hypothesis that RBF-2 is a downstream target of tyrosine kinase and Ras signaling.

A fortuitous observation that an RBEIII oligonucleotide could compete with the binding of nuclear extract USF and purified recombinant USF-1 to an E-box element, along with the presence of RBF-2 preparations of a polypeptide equivalent in size to the USF-interacting TFII-I (14, 36, 37), suggested to us that RBF-2 might be composed, at least in part, of USF-1/USF-2 and TFII-I. Indeed, an immunoblot revealed this to be the case (Fig. 1C, lane 8). We observed SRF, STAT1, and Myc, proteins which are known to interact with TFII-I (35), in Jurkat T-cell nuclear extracts (Fig. 1C, lane 1) and RBF-2-containing fractions eluted from heparin agarose (lane 2), but STAT1 and SRF were lost upon purification by Mono-Q chromatography (lane 3), and Myc was not retained specifically by RBEIII oligonucleotide affinity (lanes 4 to 8).

Consistent with the observation that USF1, USF2, and TFII-I can be detected in purified fractions by immunoblotting, we also found that antibodies recognizing these factors prevented the binding of purified RBF-2 to an RBEIII oligonucleotide in EMSAs (Fig. 2B, lanes 4 to 6). In contrast, antibodies against a variety of other factors had no effect on the binding of purified RBF-2 (lanes 7 to 15). Additionally, the binding of RBF-2 from Jurkat nuclear extracts to RBEIII oligonucleotides was prevented by antibodies against USF1, USF2, and TFII-I (Fig. 2C, lanes 6 to 8). Taken together, these results demonstrate that RBF-2 is comprised, minimally, of USF1, USF2, and TFII-I.

RBF-2 can be produced in vitro as a USF1/USF2 heterodimer.We noticed that RBEIII is weakly related to an E box within the core sequence (ACTGCTGA). USF1 and USF2 produced by in vitro transcription and translation were each capable of binding to the RBEIII P3 oligonucleotide in an EMSA (Fig. 3A, lanes 2 and 4). USF1 and -2 exist predominately as heterodimers in most cell types (42), and accordingly, USF1/USF2 heterodimers were also capable of binding RBEIII in vitro (Fig. 3A, lane 6). When cotranslated, USF1 and USF2 produced a complex with an intermediate mobility (lane 6) compared to those of USF1 and USF2 produced individually (lanes 2 and 4). The mobility of the complex formed by in vitro-translated USF1/USF2 heterodimers was identical to that of the complex formed by RBF-2 from Jurkat nuclear extracts (Fig. 3A, compare lanes 6 and 10). Furthermore, the binding of both recombinant USF1/2 (see below) and RBF-2 could be competed by a WT RBEIII oligonucleotide (Fig. 3A, lane 11) but not by the mutant P3M oligonucleotide (lane 12). Importantly, however, the binding of RBF-2 to RBEIII was sensitive to antibodies against TFII-I (Fig. 3A, lane 15, and Fig. 2B and C) in addition to antibodies against USF1 and -2 (Fig. 3A, lanes 13 and 14), whereas in vitro-translated USF1/2 was sensitive to antibodies against USF1 and -2 (Fig. 3A, lanes 7 and 8) but not to antibodies against TFII-I (lane 9). We demonstrate below that the requirement for TFII-I is necessitated by the weak affinity of USF1/2 for the RBEIII element.

• Open in new tab
• Download powerpoint

FIG. 3.

Recombinant USF1 and USF2 bind RBEIII in vitro. (A) RBF-2 can be produced as a USF1/USF2 heterodimer. USF1 and USF2 were produced by translation in vitro, either separately (lanes 2 and 3 and lanes 4 and 5, respectively) or by cotranslation (lanes 6 to 9), and were assayed for binding to a labeled P3 oligonucleotide. For comparison, Jurkat nuclear extracts were analyzed with the same probe (lanes 10 to 15). Binding reaction mixtures contained a 100-fold molar excess of unlabeled P3 (lane 11) or P3M (lane 12) oligonucleotide or antibodies against USF1 (lanes 3, 7, and 13), USF2 (lanes 5, 8, and 14), or TFII-I (lanes 9 and 15). Migration of USF1 and USF2 homodimers, USF1/USF2 heterodimers, and RBF-2 is indicated. Lanes 6 to 9 and 10 to 13 are identical gels but were exposed for 5 or 1 h, respectively. (B) USF1/USF2 heterodimers produced in insect cells bind specifically to RBEIII in vitro. USF1 and USF2 coexpressed in insect cells were assayed for binding to the RBEIII P3 oligonucleotide probe. Binding reaction mixtures contained 0 to 100 pM total USF protein (lanes 1 to 8) or a 100 pM concentration of total USF protein (lanes 9 to 12) or a lysate from uninfected SF9 cells (lane 13). Antibodies against USF1 (lane 9), USF2 (lane 10), and the competitor oligonucleotide P3 (lane 11) or P3M (lane 12) were added to the binding reactions.

RBEIII is a low-affinity binding site for recombinant USF1 and USF2 in vitro.The coexpression of USF1 and USF2 in insect cells produced homodimers as well as USF1/2 heterodimers capable of binding RBEIII in EMSAs (Fig. 3B, lanes 1 to 8). However, this interaction was considerably weaker than binding to a consensus E box, as an unlabeled E box competitor oligonucleotide was found to compete for binding of USF proteins to the RBEIII element more avidly than the WT RBEIII oligonucleotide (Fig. 4A, compare lanes 4 and 5 to lanes 7 to 9). Conversely, an unlabeled RBEIII oligonucleotide was not as efficient at competing for binding to an E box probe (Fig. 4A, lanes 16 to 18). Competition titration experiments indicated that USF binds approximately 160-fold less efficiently to an RBEIII oligonucleotide than to a consensus E-box-containing oligonucleotide (Fig. 4B). Thus, although recombinant USF proteins are capable of binding the RBEIII element in vitro, their affinity for this element is significantly less than that for a canonical E box element.

• Open in new tab
• Download powerpoint

FIG. 4.

RBEIII is a low-affinity binding site for USF in vitro. (A) EMSAs were performed with 50 pM USF1 and USF2, produced by coinfection of SF9 cells, and a labeled oligonucleotide P3 probe (lanes 1 to 9) or E box probe (lanes 10 to 18). An unlabeled E box competitor oligonucleotide was added at a 12.5- or 25-fold (lanes 4 and 5) or 25-, 50-, or 100-fold molar excess (lanes 13 to 15). The RBEIII oligonucleotide P3 competitor was added in a 25-, 50-, or 100-fold molar excess (lanes 7 to 9 and 16 to 18). Antibodies against USF2 or USF1 were included in lanes 1 and 2, respectively. (B) Competition experiments by EMSA with coexpressed USF1 and USF2 and a labeled RBEIII oligonucleotide P3 probe. Binding reactions contained 50 pM USF1/USF2 and an unlabeled RBEIII P3 competitor (○, □, and ▵) or E box competitor (•, ▪, ▴) oligonucleotide at the indicated molar excess (x axis). The amounts of probe bound to USF1 (▪ and □), USF2 (▴ and ▵), and the USF1/USF2 heterodimer (• and ○) were quantitated from phosphorimaging scans and presented as the proportions of probe bound in the sample without competitor. Inset, expanded view of data for E box competitor from 0 to 0.8-fold molar excess.

TFII-I promotes interaction of USF1/USF2 with RBEIII in vitro.TFII-I is capable of binding DNA on its own to initiator and E box consensus elements (36). We found that recombinant TFII-I is also capable of binding to RBEIII in EMSAs (not shown). Moreover, as indicated above, TFII-I was specifically retained on an RBEIII oligonucleotide affinity column, although significant amounts were detected in the flowthrough and early wash fractions (Fig. 1B, lanes 4 to 6). Although capable of a weak interaction with RBEIII on its own, recombinant TFII-I promoted the binding of USF1, as well as USF1/USF2 heterodimers, to the RBEIII oligonucleotide in vitro (Fig. 5). EMSA reaction mixtures containing 5 pmol of USF1 produced a barely detectable complex with the labeled P3 oligonucleotide (Fig. 5A, lane 1), whereas recombinant TFII-I showed a dose-dependent stimulation of binding of USF1 to RBEIII (Fig. 5A, lanes 2 to 9, and D). In contrast, the addition of TFII-I to EMSA reaction mixtures containing USF2 did not cause an increase in binding to the P3 RBEIII oligonucleotide (Fig. 5B, lanes 2 to 9, and D). The addition of TFII-I to coexpressed USF1/USF2 enhanced the formation of USF1 homodimer, USF1/USF2 heterodimer, and to a lesser extent, USF2 homodimer complexes with the P3 oligonucleotide (Fig. 5C, lanes 2 to 7, and D). Thus, the stimulatory effect of TFII-I on USF binding seems to be mediated predominately through USF1. This observation may be consistent with the finding that an interaction between USF1 and TFII-I, but not between USF2 and TFII-I, can be detected in Jurkat cells by coimmunoprecipitation (see below; also data not shown). In EMSA binding reactions in which TFII-I was limiting, the inclusion of antibodies against TFII-I inhibited the binding of all USF species to RBEIII (Fig. 5C, lane 10). These results are consistent with the observation that the binding of purified RBF-2 and RBF-2 present in Jurkat nuclear extracts to RBEIII can be inhibited by TFII-I antibodies (Fig. 2 and 3A), and they demonstrate that TFII-I can promote the binding of USF1 and USF1/USF2 heterodimers to the RBEIII element in vitro.

• Open in new tab
• Download powerpoint

FIG. 5.

TFII-I promotes binding of USF1 and USF1/USF2 to RBEIII in vitro. (A) EMSA binding reactions with labeled RBEIII oligonucleotide P3 were performed with 5 pM USF1-Flag and TFII-I at increasing concentrations (0 to 10 pM, lanes 1 to 9) or with 10 pM TFII-I (lanes 11 to 14). Binding reactions contained antibodies against USF1 (lane 10), the Flag epitope (lane 11), or TFII-I (lane 12) or an unlabeled P3 (lane 13) or P3M (lane 14) oligonucleotide. (B) EMSA binding reactions with RBEIII oligonucleotide P3 probe were performed with 5 pM USF2 and either 0 to 10 pM (lanes 1 to 9) or 10 pM (lanes 10 to 14) TFII-I. Binding reactions contained antibodies against USF2 (lane 10) or TFII-I (lane 11) or an unlabeled P3 (lane 12) or P3M (lane 13) oligonucleotide. (C) EMSA reactions with 10 pM coexpressed USF1 and USF2 and the RBEIII oligonucleotide P3 probe contained 0 to 10 pM TFII-I (lanes 1 to 6) or 5 pM TFII-I (lanes 7 to 10). The USF1/USF2 heterodimer complex and TFII-I complex migrated identically in the EMSA (indicated with an asterisk). Binding reactions contained antibodies against USF1 (lane 8), USF2 (lane 9), or TFII-I (lane 10). (D) EMSA reactions were performed as described for panels A to C. The relative amounts of RBEIII probe bound to USF1/2 heterodimers (less the contribution of signal from the binding of TFII-I [○]), USF1(□ and ▪), and USF2 (▵ and ▴) were quantitated from phosphorimager scans. Reactions contained the indicated molar ratios of TFII-I/total USF protein (x axis) and 10 pM coexpressed USF1/USF2 (○, □, and ▵) or 5 pM USF1 (▪) or USF2 (▴).

RBEIII is necessary for the response of integrated LTRs to T-cell activation.To determine the role of RBEIII in regulating the LTR in the context of an integrated provirus, we produced Jurkat cell lines with stably integrated WT or RBEIII mutant LTR-luciferase reporter genes (Fig. 6A). In general, stably transfected cloned Jurkat lines with RBEIII mutant LTR reporters had slightly higher basal expression than clones with WT LTR-luciferase reporter genes in unstimulated cells but were relatively unresponsive in cells stimulated with PMA and ionomycin in combination with the histone deacetylase inhibitor TSA (Fig. 6A). Jurkat clones with stably integrated WT LTR-luciferase reporters were typically induced 20- to 40-fold, but the mutant LTRs were induced only 2- to 3-fold (Fig. 6A). We chose representative WT (W26) and RBEIII mutant (M2) integrated LTR clones for further analysis (Fig. 6B). The activation of T cells through engagement of the T-cell receptor causes an induction of the WT integrated HIV-1 LTR through a combination of signaling mechanisms (18). We found that the RBEIII mutation significantly impairs induction of the integrated LTR in cells stimulated by cross-linking of the T-cell receptor (TCR) (Fig. 6B). Cell lines bearing the wild-type and mutant integrated LTR reporter genes were both stimulated by TCR cross-linking, as indicated by the expression of CD69, an early T-cell activation marker (Fig. 6C), but expression of the RBEIII mutant LTR was induced only approximately 4-fold, in contrast to a 25-fold induction for the WT integrated LTR (Fig. 6B). The RBEIII mutation also slightly impaired induction in cells stimulated with PMA alone (Fig. 6B). In contrast, the wild-type and RBEIII mutant integrated LTRs were induced to similar levels by treatment with TNF-α, alone or in combination with PMA (Fig. 6B). TNF-α activates NF-κB through IKK (29, 41), and thus this result demonstrates that the RBEIII mutation selectively impairs the activation of the LTR by RBF-2 but not by NF-κB.

• Open in new tab
• Download powerpoint

FIG. 6.

RBEIII is required for response of integrated HIV-1 LTR to T-cell activation signals. (A) Luciferase activities were measured from stable clones of Jurkat cells transfected with WT LTR-luciferase (W1, W2, W7, W14, W24, and W26) or RBEIII mutant LTR-luciferase (M1, M2, M3, M4, M10, and M20) or from pools of cells with stably transfected wild-type (WP) or RBEIII mutant (MP) reporters. Cells were untreated or were stimulated with PMA, TSA, and ionomycin. (B) Luciferase activities was measured in representative clones of stably transfected Jurkat cells bearing WT LTR-luciferase (W26, closed bars) or RBEIII mutant LTR-luciferase (M2, open bars) following treatment with TSA, TNF-α, PMA, PMA-TSA-ionomycin, T cell receptor cross-linking (TCR), or TNF-α--PMA. The results are presented as fold activities relative to untreated samples. (C) Stably transfected Jurkat clones bearing WT LTR-luciferase (W26, WT) or RBEIII mutant LTR-luciferase (M2, Mut.) were analyzed by FACS with anti-CD69 antibodies. Cells were unstimulated or were treated with PMA-TSA-ionomycin (PMA) or by cross-linking of the T-cell receptor (TCR).

USF and TFII-I are necessary for induction of the integrated HIV LTR in T cells.To address whether USF1, USF2, and TFII-I can function through the RBEIII element in vivo, we used luciferase reporter genes with four tandem repeats of the wild-type (pTA-P3WT) or mutant (pTA-P3mut) RBEIII element upstream of the minimal herpes simplex virus thymidine kinase core promoter. The WT RBEIII reporter had a significantly higher basal level of expression than the RBEIII mutant in transient transfections of COS-1 cells (Fig. 7A, vector). Cotransfection of a USF1 expression vector, alone or in combination with a USF2 expression vector, caused significantly elevated expression from the WT, but not the mutant, RBEIII reporter, indicating that USF1 can function in vivo through the RBEIII element.

• Open in new tab
• Download powerpoint

FIG. 7.

USF and TFII-I are required for induction of the integrated HIV-1 LTR. (A) Luciferase activities were measured from COS-1 cells cotransfected with minimal TK-luciferase reporter constructs bearing four upstream direct tandem repeats of the WT RBEIII oligonucleotide P3 (pTA-P3WT, closed bars) or the RBEIII mutant oligonucleotide P3M (pTA-P3mut, open bars) and plasmids expressing USF1, USF2, TFII-I, both USF1 and USF2, both USF1 and TFII-I, or a vector control (vector). (B) Jurkat clone W26 bearing the stably integrated WT LTR-luciferase reporter gene was cotransfected with a vector control or a plasmid expressing dominant negative USF (DN-USF), TFII-I (DN-TFII-I), or I-κb (DN-IκB) and a GFP expression plasmid at a 10:1 molar ratio. Transfected cells were enriched by sorting for GFP fluorescence, and luciferase activities were measured from unstimulated cells or following treatment with PMA-TSA-ionomycin. Activities are presented as fold stimulations relative to the unstimulated vector control.

Consistent with the observation that RBEIII is necessary for the response of the integrated LTR to T-cell activation signals, we found that the expression of dominant negative forms of USF and TFII-I prevented the stimulation of the LTR in cells treated with PMA, TSA, and ionomycin (Fig. 7B). In contrast, expression of the dominant negative IκB, which was previously shown to prevent the activation of NF-κB in Jurkat cells by TCR cross-linking (23, 46), had no effect (Fig. 7B). These results demonstrate that RBF-2 is necessary for activation of the integrated LTR by PMA and T-cell receptor engagement, but not for induction by TNF-α through the activation of NF-κB.

RBF-2 components are phosphorylated in PMA-stimulated Jurkat cells.USF1, USF2, and TFII-I are components of RBF-2 and are required for induction of the integrated HIV-1 LTR in response to PMA treatment. Accordingly, we observed the phosphorylation of USF1 and TFII-I in response to PMA treatment in Jurkat cells, as determined by metabolic labeling (Fig. 8A, lane 2, USF1 ³²P and TFII-I ³²P). An analysis of the labeled proteins from PMA-treated cells indicated that both proteins are phosphorylated at three independent sites (Fig. 8C). USF2 is phosphorylated in unstimulated Jurkat cells (Fig. 8A, lane 1, USF2 ³²P) and is hyperphosphorylated in response to PMA treatment (lane 2, USF2 ³²P). Phosphopeptide analysis indicated that USF2 is phosphorylated at two different sites and that PMA treatment does not cause the appearance of novel phosphopeptides (not shown). We observed an additional labeled protein of approximately 120 kDa in immunoprecipitates of USF1 from PMA-treated Jurkat cells (Fig. 8A, lane 2, “*”). This protein appears to be TFII-I, as it generated an identical two-dimensional phosphopeptide fingerprint to that of TFII-I (data not shown; Fig. 8C, right panel). Also, we observed an interaction of TFII-I with USF1 in Jurkat cells by coimmunoprecipitation with Flag-tagged USF1 (Fig. 8B). Under similar conditions, we were not able to observe an interaction between USF2 and TFII-I (data not shown; Fig. 8A). Thus, all of the factors identified as components of RBF-2 become phosphorylated in response to the stimulation of Jurkat T cells with PMA, which is consistent with a role in regulating the response of HIV-1 to T-cell activation signals.

• Open in new tab
• Download powerpoint

FIG. 8.

USF1, USF2, and TFII-I are phosphorylated in PMA-stimulated T cells. (A) Jurkat cells were labeled with [³²P]orthophosphate, and USF1 (top), USF2 (center), or TFII-I (bottom) was recovered by immunoprecipitation from unstimulated cells (lane 1) or cells treated with PMA (lane 2). The migration of labeled USF1 (USF1 ³²P), USF2 (USF2 ³²P), and TFII-I (TFII-I ³²P) is indicated with arrows. The band indicated with an asterisk from USF1 immunoprecipitates was found to represent labeled TFII-I (not shown). Parallel unlabeled samples were analyzed by immunoblotting with antibodies against USF1 (α-USF1 WB), USF2 (α-USF2 WB), TFII-I (α-TFII-I WB), or actin (α-Actin WB). (B) In vivo interaction between USF1 and TFII-I. Lysates from Jurkat cells (lanes 1 and 3) or Jurkat cells stably expressing USF-Flag (lanes 2 and 4) were immunoprecipitated with anti-Flag antibodies, and the samples were analyzed by immunoblotting with antibodies against USF1 (lanes 1 and 2) or TFII-I (lanes 3 and 4). (C) Labeled USF1 (left) and TFII-I (right) from PMA-stimulated Jurkat cells were digested with trypsin, and phosphopeptides were analyzed by 2D fingerprinting. The origin is labeled (○), and phosphopeptides observed in PMA-treated cells are numbered (1 to 3).

TFII-I and USF1/2 are bound to the RBEIII region of the LTR in unstimulated cells.Using chromatin immunoprecipitation, we found that USF1, USF2, and TFII-I are bound to the upstream region (−220 to −1) (Fig. 1A) of the WT LTR in both unstimulated cells and cells treated with PMA (Fig. 9A, panels iii and iv). In contrast, TFII-I and USF2 were not observed on the RBEIII mutant LTR under either condition, and only a small amount of USF1 was observed in stimulated cells, likely reflecting an interaction with the upstream E box (Fig. 9A, panel ii). This demonstrates that RBEIII is necessary for the interaction of USF1, USF2, and TFII-I with the upstream LTR region in vivo. Furthermore, since USF1, TFII-I, and USF2 were bound in unstimulated cells, these factors must be available for recruiting complexes that are necessary for induction, concomitant with their modification by phosphorylation. Unlike the USFs and TFII-I, we only observed bound NF-κB in PMA-treated cells (Fig. 9A, panels iii and iv), and furthermore, the RBEIII mutation did not affect the interaction with the LTR in vivo (Fig. 9A, panels i and ii). Thus, although the RBEIII mutation impaired the binding of the RBF-2 components in vitro and in vivo, as well as induction by PMA and ionomycin, it did not prevent the binding of NF-κB. This observation supports the view that TFII-I, in combination with USF1 and USF2, plays an important role in the LTR response to T-cell activation signals independently of NF-κB.

• Open in new tab
• Download powerpoint

FIG. 9.

USF1, USF2, and TFII-I are bound to RBEIII on the integrated WT HIV-1 LTR in unstimulated cells. (A) Chromatin immunoprecipitation was performed with formaldehyde cross-linked stable Jurkat clones bearing the RBEIII mutant LTR-luciferase (i and ii) or WT LTR-luciferase (iii and iv) reporter by the use of antibodies against USF1 (lane 3), USF2 (lane 4), TFII-I (lane 5), or NF-κB (lane 6). Samples were analyzed by PCR with oligonucleotides specific for the HIV-1 LTR flanking the RBEIII region (−220 to −1) and the β-globin promoter. A 100-bp ladder is shown in lane 1, and the input samples are shown in lane 2. Cells were untreated (i and iii) or stimulated with PMA-TSA-ionomycin (ii and iv).(B) The results from panel A were quantitated from a phosphorimager scan and normalized relative to the input sample and a β-globin internal control.