A Role of RNA Helicase A in cis -Acting Transactivation Response Element-mediated Transcriptional Regulation of Human Immunodeficiency Virus Type 1*

RNA helicase A (RHA) has two double-stranded (ds) RNA-binding domains (dsRBD1 and dsRBD2). These domains are conserved with the cis -acting transactivation response element (TAR)-binding protein (TRBP) and dsRNA-activated protein kinase (PKR). TRBP and PKR are involved in the regulation of HIV-1 gene expression through their binding to TAR RNA. This study shows that RHA also plays an important role in TAR-mediated HIV-1 gene expression. Wild-type RHA preferably bound to TAR RNA in vitro and in vivo . Overexpression of wild type RHA strongly enhanced viral mRNA synthesis and virion production as well as HIV-1 long terminal repeat-directed reporter (luciferase) gene expression. Substitution of lysine for glutamate at residue 236 in dsRBD2 (RHA K236E ) reduced its affinity for TAR RNA and im- paired

. We have previously shown that RHA mediates association of the CREB-binding protein (CBP) with RNA polymerase II (pol II) (3) and that RHA links breast cancer-specific tumor suppressor protein (BRCA1) to pol II (4). RHA consists of two type A double-stranded RNA-binding domains (dsRBDs) (5, 6), a classical Walker type NTP-binding site, a DEAH/D helicase domain, and a single-stranded nucleic acid-binding domain characterized by Arg-Gly-Gly (RGG) repeats (7). Trypsin-digested RHA, lacking both dsRBDs and the RGG repeat sequence, has reduced helicase activity, implicating these domains in the unwinding function of RHA (7). Amino acids 1-262 and 255-664 of RHA have proved to be CBP-and pol II-binding sites, respectively (3). RHA shuttles between nucleus and cytoplasm with a cis-acting constitutive transport element in simian retroviruses (8). It has been proposed that RHA is necessary for releasing both constitutive transport element-and HIV-1 Rev response element (RRE)-containing RNA from spliceosomes prior to the completion of splicing (9).
After integration of HIV-1 into the host genome, the nuclear factor-B (NF-B) binds to enhancer elements in the HIV-1 long terminal repeat (LTR) and stimulates the expression of the viral genome in a signal-dependent manner (10,11). HIV-1 expression can be divided into two phases (early and late). In the early phase, the majority of viral mRNA is multiply spliced to produce 2-kb transcripts that encode the regulatory proteins, including Tat and Rev, necessary to activate HIV-1 expression. A transition subsequently occurs to accumulate singly spliced (4-kb) and full-length (9-kb) transcripts encoding the viral structural proteins and providing the genomic RNA (12)(13)(14). Regulation of the viral RNA-splicing transition mechanism is reported to involve the protection and transport of full-length HIV-1 RNA by Rev (15,16).
TAR, a nascent viral leader RNA transcribed from the R region of the LTR, plays an important role in HIV-1 gene expression and forms a unique stem and loop structure (17,18). Tat function is mediated by the TAR RNA and requires the recruitment of a complex consisting of Tat and cyclin T1 component of positive transcription elongation factor b (P-TEFb) bound to TAR (19). A defect of Tat-induced transactivation in murine cells was attributed to the lack of a functional cyclin T1 (20,21). Alternatively, the defect was linked to the reduced abundance of p300 and p300/CBP-associated factor (22). In addition, several cellular cofactors play crucial roles in HIV-1 gene expression through binding to the TAR RNA. For instance, TRBP consists of two type A dsRBDs and another type B dsRBD and activates the HIV-1 LTR (6, 23) through binding of the second type A dsRBD (dsRBD2) to the TAR RNA (24). * This work was supported by grants from the Japanese Ministry of Education, Science, Culture, and Sports (10480196 and 10177230), Japanese Ministry of Health and Welfare, Japan Science and Technology Corporation (PRESTO), Human Health Science Foundation Medicitial Welfide Foundation, and by funds from Memorial Yamanouchi Foundation, Kaken Pharmaceutical Co. Ltd., and Santen Pharmaceutical Co. Ltd. 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.
¶ PKR contains two type A dsRBDs and functions as a cellular antiviral factor by inhibiting eukaryotic initiation factor 2 via phosphorylation of its ␣-subunit (25). A similar function of PKR is mediated by an interaction with several kinds of RNA, leading to PKR autophosphorylation (26). Like TRBP and PKR, RHA possesses dsRBDs and may act as a cellular transcriptional regulator. Of particular interest is whether RHA binds to the TAR RNA and influences HIV-1 gene expression. This study shows that RHA acts as a novel TAR-binding cellular cofactor and enhances HIV-1 LTR-directed gene expression and viral production in vivo.
Plasmids-Plasmid pHyg LTR-Luc encodes a chimeric gene consisting of the HIV-1 LTR containing two intact B elements and a luciferase gene (28). The mNFB LTR-Luc plasmid was constructed by inserting both of the B element-mutated LTR fragments (14) into the SmaI/ HindIII site of PGV-B containing a luciferase gene without a promoter (Toyo-Inki). The pCD-SR␣/tat plasmid, a mammalian expression vector for HIV-1 Tat, was constructed as described previously (28). The HIV-1 pNL4 -3 proviral DNA clone (30) was obtained from the AIDS Research and Reference Reagent Program (Division of AIDS, NIAID, National Institutes of Health) and was contributed by Malcolm Martin. Wildtype (wt) RHA and RHA mATP constructs were prepared as described previously (3). The RHA mATP and RHA W339A 2 constructs contain single amino acid substitutions leading to defects in ATPase/helicase activity and pol II binding ability of RHA, respectively. RHA K236E containing the lysine to glutamate substitution at residue 236 was generated by PCR. The wt RHA construct and three RHA mutants were tagged by hemagglutinin (HA) for Western blot assays and immunoprecipitation studies. To obtain the RNA probes by in vitro transcription, the pcTAR and ⌬loopTAR plasmids were constructed by inserting PCR-generated cDNA fragments (5Ј-GGGTCTCTCTGGTTAGACCAGATCTGAGCCT-GGGAGCTCTCTGGCTAACTAGGGAACCC-3Ј and 5Ј-GGGTCTCTCT-GGTTAGACCAGATCTGAGCGCTCTCTGGCTAACTAGGGAACCC-3Ј) with HindIII linker into the HindIII site of pcDNA3 (Invitrogen). The plasmids as described above were purified using cesium chloride gradients.
A Transient Transfection and Luciferase Assays-HEK 293 and HeLa cells were transiently transfected with 100 ng of pHyg LTR-Luc or mNFB-Luc reporter, 100 ng of RSV-␤-gal control plasmid, 2 or 10 ng pCD-SR␣/tat, and various amounts of RHA, using the calcium phosphate method as described previously (31). To ensure an equal amount of DNA, "empty plasmids" were added in each transfection. In the reporter gene assays, luciferase activity derived from expression of the pHyg LTR-Luc and mNFB LTR-Luc reporter plasmids was normalized to ␤-galactosidase activity from cotransfected Rous sarcoma virus (RSV)-expression plasmid containing the ␤-galactosidase gene (RSV-␤gal) (3). Luciferase activity was measured with AutoLumat (Berthold). All experiments were performed in triplicate, and all results were obtained from at least three separate experiments. Equivalent expression of RHA constructs was verified by Western blot analysis using the anti-HA antibody 12CA5 (Roche Molecular Biochemicals).
TAR Binding Assays-Total cellular extract from Escherichia coli BL21, expressing a series of RHA polypeptides or full-length TRBP fused to GST protein, was electrophoresed on an SDS gel and transferred to a polyvinylidene difluoride membrane (Immobilon P, Millipore). TAR binding assays were performed as reported previously (23). Filters were incubated in binding buffer (20 mM HEPES, pH 7.3, 40 mM KCl, 1.5 mM MgCl 2 , and 1 mM dithiothreitol) with 10 g/ml yeast RNA, 10 g/ml calf thymus DNA, and 200 fmol/ml gel-purified 32 P-labeled TAR RNA probe. Filters were washed with binding buffer and then exposed to x-ray film. Comparable expression of GST-fused protein was determined by Western blotting with an anti-GST antibody (Amersham Pharmacia Biotech) and Coomassie Brilliant Blue staining.
Wild-type TAR, ⌬loop TAR RNA, and ssRNA were obtained by in vitro transcription using BamHI-cleaved pcTAR or ⌬loopTAR plasmid and XhoI-cleaved pcDNA3, respectively. GST and RHA-(1-262) polypeptide blots were incubated in binding buffer with 10 g/ml calf thymus DNA, 1 pmol/ml 32 P-labeled wt TAR, ⌬loop TAR RNA or ssRNA probe, and 300 pmol/ml unlabeled RNA or yeast tRNA (Sigma) as a competitor. The filters were washed with binding buffer and exposed to x-ray film.
Immunoprecipitation and Slot Blot Assays-HEK 293 cells were transfected with 3 g of pHyg LTR-Luc or G5b-Luc (32), 150 ng of pCD-SR␣/tat or 300 ng Gal4-VP16 (33), and 3 g of each RHA construct. After 24 h of transfection, the cells were lysed in 1 ml of 0.65% Nonidet P-40 lysis buffer (20 mM HEPES, pH 7.3, 150 mM KCl, 1.5 mM MgCl 2 , 0.65% Nonidet P-40). Cell lysates were pre-cleared with 5 g of normal mouse IgG (Santa Cruz Biotechnology) conjugated to protein G-Sepharose beads (Amersham Pharmacia Biotech) for 1 h at 4°C. After a brief centrifugation, the lysates were mixed with 5 g of normal mouse IgG or 5 g of anti-HA antibody conjugated to protein G-Sepharose beads. After an overnight incubation at 4°C, the beads were washed three times with lysis buffer. RNA was extracted from the beads with ISO-GEN (Nippon Gene) and then blotted onto GeneScreen Plus membrane (PerkinElmer Life Sciences) using a slot blotting apparatus (HYBRI-SLOT MANIFOLD, Life Technologies, Inc.). The membrane was hybridized with a 32 P-labeled luciferase gene probe generated by PCR

FIG. 1. The effects of RHA on HIV-1 LTR-directed gene expression. HEK 293 (A) and HeLa cells (B)
were transfected with 100 ng of pHyg LTR-Luc reporter, and 2 ng (for HEK 293 cells) or 10 ng (for HeLa cells) of pCD-SR␣/tat (tat) or "empty vector" (Ϫtat), together with various amounts of the wt RHA expression vector as indicated. To ensure an equal amount of DNA, the control plasmid was added in each transfection. Luciferase activity was normalized to transfection efficiency via cotransfection with 100 ng of RSV-␤-gal. All experiments were performed in triplicate, and results were taken from three separate experiments. *, lanes 5 and 6 versus 4, p Ͻ 0.05, by multiple comparisons using the t test.
using the primer pair 5Ј-GGATGGAACCGCTGGAGAG-3Ј and 5Ј-GTT-TCATAGCTTCTGCCAACCG-3Ј and exposed to x-ray film. Equivalency for immunoprecipitation of HA-tagged RHA was verified by Western blot analysis.
p24 Assays-HEK 293 cells were transfected with 10 ng of pNL4-3, 20 ng of PGV-C control plasmid containing the SV40-luciferase gene (Toyo-inki, Tokyo Japan), and various amounts of RHA construct. To ensure an equal amount of DNA, empty plasmid was added in each transfection. After 24 and 48 h of transfection, culture supernatants were collected and tested for HIV-1 p24 antigen levels using a p24 antigen detection kit (Retro-tek, Zepto Metrix Corporation, Buffalo, NY). Equivalent transfection efficiency was verified by luciferase activity derived from the cotransfected PGV-C control plasmid. All experiments were performed in triplicate.
Northern Blot Assays-HEK 293 cells were transfected with 500 ng of pNL4-3, 1 g of PGV-C control plasmid, and 5 g of each RHA construct. After 24 h of transfection, polyadenylated RNA was extracted from the transfected cells using an mRNA purification kit (Amersham Pharmacia Biotech), run on a 0.8% agarose gel, and blotted onto Gene-Screen Plus membrane. The membrane was hybridized with a 32 Plabeled HIV-1 LTR probe and then exposed to x-ray film. The probe was amplified by PCR using the primers 5Ј-AGTGCTCAAAGTAGTGT-GTG-3Ј and 5Ј-GATCTCCTCTGGCTTTACTTT-3Ј and pNL4-3 as a template. Equivalent transfection efficiency was determined by measuring the amount of luciferase mRNA derived from the cotransfected PGV-C control plasmid.

RESULTS
The Effects of RHA on HIV-1 LTR-directed Gene Expression-To assess the regulatory effects of RHA on HIV-1 LTRdirected gene expression, several human cell lines were transfected with the pHyg LTR-Luc reporter plasmid. In HEK 293 cells, wt RHA markedly enhanced Tat-induced reporter activity in a dose-dependent manner (Fig. 1A). In contrast, overexpression of wt RHA did not affect HIV-1 LTR-directed luciferase activity in HeLa cells (Fig. 1B) as described previously (9).
Determination of the RHA Regions Required for TAR Binding-Northwestern assays were conducted to determine whether RHA binds to the TAR RNA. First, we divided fulllength RHA into four fragments ( Fig. 2A) and examined each fragment for their TAR RNA binding activity. Only RHA-(1-262), which contains both dsRBD1 and dsRBD2, could bind to the TAR RNA (Fig. 2B, left panel). Second, we constructed three deletion mutants of RHA-(1-262) ( Fig. 2A). Unexpectedly, only RHA-(160 -262) (containing dsRBD2) and not RHA-(1-90) (containing dsRBD1) bound to the TAR RNA (Fig. 2B,  middle panel). Amino acids 235-249 in the RHA dsRBD2 are well conserved among the dsRNA-binding protein family (Fig.  2C). In particular, the lysine at residue 211 (Lys-211) in the  (20). Similarly, it is important to define the region of TAR that interacts with the RHA dsRBD and to confirm its relative specificity. Presumably, RHA dsRBD binds to the stem region of TAR RNA, since dsRBD binds to dsRNA. The ⌬loop TAR RNA probe illustrated in Fig. 3A was therefore constructed. GST-fused RHA-(1-262) and GST alone (negative control) were blotted onto filters, and the filters were incubated with an equivalent amount of 32 P-labeled wt TAR, ⌬loop TAR RNA, or ssRNA. Interestingly, the binding affinity of ⌬loop TAR RNA to RHA-(1-262) was significantly stronger than that of wt TAR RNA. ssRNA bound weakly to RHA-(1-262). None of the tested probes bound to GST (Fig. 3B). In competition assay (Fig. 3C), the binding of wt TAR RNA to RHA-(1-262) was almost completely inhibited by wt TAR or ⌬loop TAR RNA competitor at a concentration of 300-fold higher than that of the 32 P-labeled probe. In contrast, the ssRNA competitor and yeast tRNA slightly reduced the probe binding. These results suggest that the RHA dsRBD preferably recognizes the stem of the TAR RNA.
Association of RHA with TAR in Vivo-To test for in vivo interaction of the TAR RNA with the RHA complex, coimmunoprecipitation assays were performed on whole cell lysates from HEK 293 cells cotransfected with pHyg LTR-Luc (to express TAR-containing luciferase mRNA) or G5b-Luc (to express TAR-negative luciferase mRNA) and each RHA construct. Except for "no-transfectant," a comparable amount of TAR-con-taining and TAR-negative mRNA was detected in each cell lysate (Fig. 4, middle panel). Equal amounts of wt RHA and RHA K236E were precipitated with anti-HA antibody (Fig. 4,  lower panel). The TAR-containing mRNA was present in the wt  TAR). B, the element in TAR that binds to RHA dsRBD. Filters transferred with various amounts of RHA-(1-262) as indicated and 100 pmol of GST proteins were incubated with 1 pmol/ml 32 P-labeled wt TAR, ⌬loop TAR RNA, or ssRNA probe (see "Experimental Procedures"). C, relative specificity for RHA binding to TAR RNA. The same filters as indicated above were incubated with 1 pmol/ml 32 P-labeled wt TAR RNA alone (no competitor) or together with 300 pmol/ml of either unlabeled wt TAR, ⌬loop TAR, ssRNA, or yeast tRNA (competitor).
RHA precipitant (Fig. 4, upper panel). However, although significantly less mRNA was coprecipitated with RHA K236E . No precipitant was identified in the negative controls (no-transfectant and "mock-transfectant"). Furthermore, the association of wt RHA with the TAR-negative luciferase mRNA was as weak as that of RHA K236E and the TAR-containing luciferase mRNA. These results indicate that the TAR-containing mRNA existed in the RHA complex in vivo and that the complex was formed primarily via the binding of RHA dsRBD2 to the TAR RNA. Although RHA-(1-262/K236E) could not bind to the TAR RNA in vitro (Fig. 2B), the TAR-containing mRNA was weakly associated with RHA K236E complex in vivo. Thus, it cannot be excluded that RHA also binds to mRNA in a TAR-independent fashion, as in the case of RHA binding to ssRNA through the RGG motif (7).
Enhancement of TAR-dependent Gene Expression by RHA-The finding that RHA binds to the TAR RNA in vivo prompted a study of the role of RHA in TAR-mediated gene expression. Transient transfections of the pHyg LTR-Luc reporter plasmid were conducted to analyze the effect of RHA K236E on HIV-1 LTR-directed gene expression. RHA K236E , but neither RHA mATP nor RHA W339A , enhanced the reporter activity (Fig.  5, lanes 7-12), although the effect of RHA K236E was significantly lower than that of wt RHA (ϳ40%). This reduced effect of RHA K236E may be due to the lack of TAR-binding ability. Alternatively, RHA may be functionally bound to the CBP⅐NF-B complex on the B elements of HIV-1 LTR (35,36). To exclude this possibility, transient transfection assays with the mNFB LTR-Luc reporter plasmid that contained two mutated B elements were also performed. Mutation of the B elements strongly reduced luciferase activity (Fig. 5, lane 4  versus 16), as described previously (11). Wild-type RHA markedly enhanced the Tat-induced reporter activity in a dose-dependent manner (Fig. 5, lanes 16 -18), and although both RHA mATP and RHA W339A slightly increased the luciferase activity, RHA K236E had no effect (Fig. 5, lanes 19 -24). These results indicate that the association of RHA with the TAR RNA is required for the RHA-induced transactivation in the mutated B LTR. Conversely, the higher activity of RHA K236E in intact B LTR compared with mutated B LTR suggests that RHA can also interact with HIV-1 LTR through the intact B elements in a TAR-independent fashion.
The Effects of RHA on HIV-1 Production-To elucidate the role of TAR-binding ability of RHA in HIV-1 viral replication, HEK 293 cells were cotransfected with pNL4-3 and different amounts of wt RHA or RHA K236E (Fig. 6A). After 24 h of transfection, HIV-1 p24 production was enhanced ϳ5-fold in the wt RHA transfectants in a dose-dependent manner (Fig.  6A, left panel), as previously demonstrated in HeLa cells (9). The effect of RHA K236E on HIV-1 p24 production was significantly less than that of wt RHA (ϳ30%), which was consistent with the results in the reporter gene assays (Fig. 5). Similar results were also obtained after 48 h of transfection, except in the cells expressing for higher p24 antigen levels (Fig. 6A, right  panel). These results suggest that the association of RHA with the TAR RNA is partly required for the RHA-enhanced HIV-1 gene expression.
Finally, to elucidate the role of RHA in transcriptional regulation of HIV-1, HEK 293 cells were transfected with pNL4-3 and each RHA construct. Fig. 6B demonstrates that three different lengths of HIV-1 mRNA (ϳ2-, 4-, and 9-kb transcripts) were detected by Northern blot assay. Equivalent transfection efficiency was confirmed by the amount of luciferase mRNA derived from each cotransfected PGV-C control plasmid (data not shown). The amounts of all HIV-1 mRNA transcripts were increased by wt RHA coexpression (Fig. 6B). No other RHA mutants affected HIV-1 mRNA synthesis or p24 production. These results strongly support a functional role of RHA in the transcriptional activation of HIV-1. DISCUSSION RHA belongs to the dsRNA-binding protein family, which includes TRBP and PKR. These proteins display variable RNAbinding characteristics. For instance, the PKR dsRBD1 displays higher affinity than the PKR dsRBD2 for several kinds of RNA, including the TAR RNA (37,38). The TRBP dsRBD2 alone can bind to TAR RNA independent of other dsRBDs (39). Two polypeptides containing each of the RHA dsRBDs bind to poly(rI⅐rC) dsRNA with similar affinity, and it has been considered that the two RHA dsRBDs cooperate to interact with dsRNA, such as with poly(rI⅐rC) (7). The results presented here show that only dsRBD2, and not dsRBD 1, could bind to TAR RNA in vitro (Fig. 2B). This RNA-binding property of RHA is similar to that of TRBP. Gel mobility shift assays with polypep- tides containing each or both RHA dsRBD were used to confirm the contribution of dsRBD1 in binding to the TAR RNA. This approach was unsuccessful due to aggregation of purified dsRBD polypeptide (data not shown).
The dsRBDs interact with highly structured RNA and dsRNA, generally without obvious RNA sequence specificity (40). In fact, RHA could bind to various RNA species, including ssRNA in vitro. Like TRBP (24), RHA bound to the TAR RNA with higher affinity than to ssRNA or yeast tRNA and preferably bound to TAR-containing mRNA rather than TAR-negative mRNA. Thus, it appears that TAR RNA is one of the binding targets of RHA dsRBDs. These findings further demonstrate that the conserved amino acids 235-249 in the RHA dsRBD2 are essential for TAR binding. It was previously reported that the Lys-211 residue of TRBP dsRBD2 is important for interaction with TAR RNA (34). It was confirmed by x-ray crystallography that the corresponding residue of Xlrbpa, the Xenopus homologue of TRBP, directly interacts with dsRNA (41). These residues correspond to the Lys-236 residue of RHA dsRBD2, and the importance of Lys-236 was demonstrated here the using a RHA-(1-262/K236E) mutant. Together, the results confer in that the characteristics of RHA for TARbinding are similar to those of TRBP.
It is known that transcription factor levels differ between cell lines. For instance, murine NIH3T3 cells express a small amount of p300 compared with HeLa cells. Thus, the failure of HIV-1 LTR-directed transactivation by Tat in the murine cells could be attributed to the small amount of p300 (22). Western blot analysis showed that the amount of RHA in HEK 293 cells was less than 50% that in HeLa cells, although an equivalent amount of CBP or CREB was identified in both cell lines. 3 The HIV-1 LTR-directed luciferase activity was enhanced by exogenous RHA in HEK 293 cells but not in HeLa cells. HEK 293 cells are widely used in transfection experiments with HIV-1 proviral DNA clones (29,42). Taken together, the results indicate HEK 293 cells to be a favorable host cell system to assess the functions of RHA in HIV-1 gene expression by transient transfection. RHA was reported to bind weakly to HIV-1 RRE and be involved in post-transcriptional regulation of HIV-1 (9). However, the reporter gene constructs used in this study do not contain RRE, and the results in the reporter gene assays accorded well with those in the p24 assays, suggesting that the reduced activity of RHA K236E in the p24 assays was due primarily to the lack of the TAR-binding ability of RHA in HEK 293 cells. This discrepancy may be due to either the different cell types used or to the distinct roles of RHA in transcription and post-transcriptional regulation in HIV-1 gene expression.
We have previously suggested that both the ATPase/helicase activity and pol II binding ability of RHA contribute to the CREB-dependent transcription (3). It can be proposed from the current studies that these functions of RHA may also be important for HIV-1 transcription. Moreover, the data demonstrate that RHA enhances HIV-1 LTR-directed gene expression in a B element-dependent and -independent fashion. NF-B functionally binds to the B elements in the HIV-1 LTR (10,11) and associates with CBP (35,36). Furthermore, RHA was shown to mediate the association of CBP with pol II (3). The RHA K236E mutant was capable of binding to CBP in GST pulldown assays (data not shown). Therefore, the certain activity of RHA K236E in the reporter assays using intact B elements but not in the assays using the mutated B elements may indicate that RHA also mediates the association of CBP with pol II on the B-dependent HIV-1 preinitiation complex. Conversely, it was shown that RHA enhances HIV-1 gene expression, at least in part, through its TAR binding. TAR RNA is required for the recruitment of a complex consisting of Tat and the cyclin T1 component of P-TEFb (19 -21). It is possible that both the ATPase/helicase activity and pol II binding ability of RHA may be required for the unwinding of highly structured RNA, such as TAR RNA, and following HIV-1 transcription.
FIG. 6. The effect of RHA on HIV-1 production. A, implication of TAR-binding of RHA in HIV-1 replication. HEK 293 cells were transfected with 10 ng of pNL4-3, 20 ng of PGV-C plasmid, and various amounts of each RHA construct as indicated. To ensure an equal amount of DNA, the control plasmid was added in each transfection. After 24 and 48 h of transfection, culture supernatants were tested for p24 antigen levels. Equivalent transfection efficiency was verified by the luciferase activity derived from cotransfected PGV-C plasmid. All experiments were performed in triplicate. (*1, lanes 2 and 3 versus 1, p Ͻ 0.04; *2, lane 5 versus 3, p ϭ 0.0006 by multiple comparisons using the t test). B, the effect of RHA on HIV-1 mRNA synthesis. HEK 293 cells were transfected with 500 ng of pNL4-3, 1 g of PGV-C control plasmid, and 5 g of each RHA construct as indicated. After 24 h of transfection, polyadenylated RNA was extracted, electrophoresed on an agarose gel, and blotted onto GeneScreen Plus membrane. The membrane was hybridized with the 32 P-labeled HIV-1 LTR probe. To compare the amount of RNA in each lane, the replica membrane was hybridized with a glyceraldehyde-3-phosphate dehydrogenase (GAPDH) cDNA probe. Equivalent transfection efficiency was verified by measuring the amount of luciferase mRNA derived from each cotransfected PGV-C control plasmid. The p24 antigen levels in culture supernatants of the transfected cells were determined.