Functional Analysis of the Complex trans-Activating Response Element RNA Structure in Simian Immunodeficiency Virus

Abstract

Transcription of human immunodeficiency virus (HIV) and simian immunodeficiency virus (SIV) is activated through binding of the viral Tat protein to the trans-activating response (TAR) element at the 5′ end of the nascent transcript. Whereas HIV type 1 (HIV-1) TAR folds a simple hairpin structure, the corresponding domains of HIV-2 and SIVmac exhibit a more complex structure composed of three stem-loops. This structural polymorphism may be attributed to additional functions of TAR in HIV-2/SIVmac replication. We recently constructed an SIVmac variant that does not require the Tat-TAR interaction for transcription. We used this variant to study additional roles of TAR in SIVmac replication and generated mutants with a truncated TAR structure. We demonstrate that partial or nearly complete removal of TAR does not impair viral transcription, RNA processing, and translation. Moreover, these deletions do not significantly affect virus replication in the PM1 T-cell line and macaque peripheral blood mononuclear cells. These results demonstrate that the complex TAR structure in SIVmac has no other essential function in virus replication in vitro besides its role in Tat-mediated activation of transcription.

Transcription of human and simian immunodeficiency viruses (HIV and SIV, respectively) is activated by binding of the virally encoded Tat trans-activator protein to the trans-activating response element (TAR) at the 5′ ends of the viral RNAs (3, 4, 20, 26, 28). Upon Tat binding to the nascent transcript, the positive transcriptional elongation factor (pTEFb) is recruited to the transcription complex at the 5′ long terminal repeat (LTR) promoter. Consequently, the processivity of the elongating RNA polymerase II is enhanced through phosphorylation of its C-terminal domain by the kinase component of pTEFb (see reference 5 for a review). The assembly of new transcription complexes is also stimulated by pTEFb, through the recruitment of TATA box binding protein to the LTR promoter (27, 39). TAR thus functions as an essential RNA enhancer element to promote the switch from low basal transcription to Tat-activated processive transcription (11, 14).

Most of the studies in which the Tat-TAR interaction was investigated were focused on HIV type 1 (HIV-1), and despite strong similarities between the trans-activation mechanisms and Tat proteins of HIV-1, HIV-2, and SIVmac, some clear differences are apparent. Whereas HIV-1 Tat can efficiently activate transcription of HIV-2 and the closely related virus SIVmac, HIV-2 Tat is a poor activator of HIV-1 transcription (28). This difference in trans-activation capacity may be a consequence of the structural polymorphism between HIV-1 and HIV-2/SIVmac TARs (Fig. 1). HIV-1 TAR RNA folds into a characteristic hairpin structure with a 3-nucleotide (nt) bulge and a 6-nt loop, which facilitate the binding of Tat and the cyclin T1 component of pTEFb, respectively (Fig. 1A). The HIV-2 and SIVmac TAR elements exhibit a more complex RNA structure with three stem-loop domains (SL1 to SL3) (Fig. 1B) (8, 26). SL1 and SL2 contain a 2-nt bulge and a 6-nt loop and resemble the single HIV-1 stem-loop structure. Although both SL1 and SL2 are involved in Tat-mediated activation of transcription (10, 25), SL1 seems to be more important than SL2 for trans-activation (26, 40). This is possibly due to intrinsic folding kinetics of the two structures, as SL1 has a position more proximal to the transcription initiation site. Alternatively, the presence of a purine base in the SL2 bulge may be suboptimal for Tat binding. SL3 does not resemble the typical HIV-1 TAR structure and is not involved in Tat trans-activation (25).

FIG. 1.

Comparison of HIV-1 and SIVmac239 TAR structures. Schematic representations of the HIV-1 (A) and SIVmac239 (B) genomes are shown. The 5′LTR, 3′LTR, and open reading frames are indicated. The untranslated leader RNAs of HIV-1 (positions +1 to +335) and SIVmac239 (positions +1 to +534) consist of several regulatory domains, including the TAR, poly(A) hairpin, primer-binding site (PBS), dimerization initiation site (DIS), splice donor (SD), RNA packaging signal (Ψ), and start codon of gag (AUG) (22, 31). TAR and poly(A) hairpin folding is depicted.

The reason that HIV-2 and SIVmac have such a complex TAR structure with three stem-loop elements, whereas HIV-1 replicates efficiently with only a single TAR hairpin, is currently unclear. Strong conservation of certain characteristics of the TAR structure in different HIV-2 and SIV isolates suggests that all three stem-loops are important for viral replication (8, 26). Possibly, TAR has other functions in the HIV-2/SIV life cycle in addition to its role in the activation of transcription, as previously suggested for HIV-1 (2, 5-7, 9, 13, 29). The best system with which to study additional TAR functions is that of the replicating virus. However, the analysis of TAR mutant viruses is complicated by the fact that TAR is essential for transcription. Since TAR mutations will therefore have a dominant negative effect on viral replication, it is difficult to study any effect on other steps in the viral life cycle.

We recently generated an SIVmac variant in which the Tat-TAR trans-activation mechanism is functionally replaced by the doxycycline (DOX)-inducible Tet-On gene regulation system (18). Transcription of this SIV-rtTA variant is not dependent on the Tat-TAR axis, which makes it an ideal tool to study additional TAR functions. In this study, new SIV-rtTA variants were generated in which TAR was partially or completely deleted. We found that SIV-rtTA can replicate efficiently upon deletion of all three stem-loop domains of TAR, which demonstrates that TAR has no additional essential function in SIVmac replication in vitro.

MATERIALS AND METHODS

Construction of SIV-rtTA variants with deletions in TAR.

Construction of the SIV-rtTA molecular clone was described previously (18). The variant used in this study contains the wild-type tat gene (19). Deletions in TAR were introduced in both the 5′LTR and 3′LTR of the plasmid pSIV-rtTA-Tat^wt through multiple steps of PCR mutagenesis. The primers used in this study are listed in Table 1.

TABLE 1.

Primers used in this study

Primer	Sequence (5′-3′)
5Nde1	GAGGCATATGTTAGATACCCAG
TA016luc	GGTTCCATCTTCCAGCGG
3TARNar1	CTGTTCAGGCGCCAATCTGCTAGGGATTTTCCTG
5WTΔSL1	ATTCAGTCGCTCTGCGGAGAG-SL1 deletion-CTAGCAGGTAGAGCCTGGG
5MutΔSL1	ATTCAGTCGCTCTGCGGAGAG-SL1 deletion-CTAGCAGGAAGAGCATTGG
5+63ΔSL1	ATTCAGTCGCTCTGCGGAGAG-SL1 deletion-CTAGCAGGAGGAGCATTGG
3ΔSL1	CTCCGCAGAGCGACTGAATACAGAG
5ΔSL2	CTCACCAGCACTTGGCCGG
3ΔSL2	CCGGCCAAGTGCTGGTGAGAG-SL2 deletion-TGCTGGAGAGAACCTCC
5ΔSL3	GTTCCCTGCTAGACTCTCA-SL3 deletion-GCAGAGTGACTCCACGCTTG
3ΔSL3	GAGAGTCTAGCAGGGAACACC
5TAR10Δ	CGCTCTGTATTCAGTCGCTCTGC-TAR deletion-AGAGTGACTCCACGCTTGCTTG
3TAR10Δ	GCAGAGCGACTGAATACAGAGCGAAATG
5ΔTAR	CACTGCATTTCGCTCTGTATTCC-TAR deletion-ACGCTTGCTTGCTTAAAGCCCT
3ΔTAR	GGAATACAGAGCGAAATGCAGTG

(i) Introduction of TAR deletions into 3′LTR.

The mutagenesis protocol involved two PCRs with primer sets specific for each TAR deletion (PCR1 and PCR2), followed by a common PCR (PCR3).

(a) Deletion of SL1.

PCR1 was performed with primers 5WTΔSL1 and TA016luc on the template pSIV-LTR-2Δ15tetO-TAR^wt-luc (18) for the construction of the 1^Δ2^wt variant, with primers 5MutΔSL1 and TA016luc on pSIV-LTR-2Δ15tetO-TAR^m-luc (18) for the 1^Δ2^m variant, and with primers 5+63ΔSL1 and TA016luc on pSIV-LTR-2Δ15tetO-TAR^m+63-luc (17) for the 1^Δ2⁺⁶³ variant. PCR2 was performed with primers 5Nde1 and 3ΔSL1 on pSIV-LTR-2Δ15tetO-TAR^m-luc for all three variants.

(b) Deletion of SL2.

PCR1 was performed with primers 5Nde1 and 3ΔSL2 on pSIV-LTR-2Δ15tetO-TAR^wt-luc for the 1^wt2^Δ variant or on pSIV-LTR-2Δ15tetO-TAR^m-luc for the 1^m2^Δ variant. PCR2 was performed with primers 5ΔSL2 and TA016luc on pSIV-LTR-2Δ15tetO-TAR^m-luc for both variants.

(c) Deletion of SL3.

PCR1 was performed with primers 5Nde1 and 3ΔSL3 on pSIV-LTR-2Δ15tetO-TAR^m+63-luc. PCR2 was performed with primers 5ΔSL3 and TA016luc on pSIV-LTR-2Δ15tetO-TAR^m+63-luc.

(d) Complete deletion of TAR.

PCR1 was performed with primers 5Nde1 and 3ΔTAR on pSIV-LTR-2Δ15tetO-TAR^m-luc. PCR2 was performed with primers 5ΔTAR and TA016luc on pSIV-LTR-2Δ15tetO-TAR^m-luc.

(e) Nearly complete deletion of TAR.

PCR1 was performed with primers 5Nde1 and 3TAR10Δ on pSIV-LTR-2Δ15tetO-TAR^m-luc. PCR2 was performed with primers 5TAR10Δ and TA016luc on pSIV-LTR-2Δ15tetO-TAR^m-luc.

The PCR1 and PCR2 products for each TAR variant were purified, mixed, and used as a template in PCR3 with primers 5Nde1 and TA016luc. The PCR3 products were digested with EcoRI and XhoI and used to replace the corresponding fragments in pBS-3′SIV-rtTA, which contains the 3′ half of the SIV-rtTA genome (18). We thus generated pBS-3′SIV-rtTA plasmids with different TAR deletions in the 3′LTR.

(ii) Introduction of TAR deletions into 5′LTR.

PCRs were performed with primers 5Nde1 and 3TARNar1, with the different TAR-deleted pBS-3′SIV-rtTA constructs as templates. The PCR products were digested with EcoRI and NarI and used to replace the corresponding fragment in the pKP-5′SIV construct, which contains the 5′ half of the SIV-rtTA genome (18), resulting in pKP-5′SIV plasmids with the different TAR deletions in the 5′LTR.

(iii) Introduction of TAR deletions into SIV-rtTA molecular clone.

The TAR modifications were introduced into the 3′LTR of the SIV-rtTA molecular clone through ligation of the XmaI-XhoI fragment of each pBS-3′SIV-rtTA TAR deletion variant into the XmaI- and XhoI-digested pSIV-rtTA-Tat^wt plasmid. The TAR modifications were subsequently also introduced into the 5′LTR by ligation of the NotI-NarI fragment of each pKP-5′SIV TAR deletion variant into NotI- and NarI-digested pSIV-rtTA with the corresponding modification in the 3′LTR.

RNA secondary structure prediction.

Computer-assisted RNA secondary structure predictions were performed using the Mfold, version 3.2, algorithm (47) offered by the MBCMR Mfold server (http://mfold.burnet.edu.au). Standard settings were used for all folding jobs (37°C and 1 M NaCl).

rtTA activity assay.

In the pSIV-LTR-2Δ15tetO-TAR^m-luc plasmid, the expression of firefly luciferase is under the control of the LTR-2ΔtetO promoter of SIV-rtTA (18). In this plasmid, the complete LTR region of SIV-rtTA (encompassing U3, R, and U5) is present upstream of the luciferase gene. The EcoRI-XhoI fragment of each pSIV-rtTA TAR deletion variant, spanning the U3 promoter and the TAR region, was used to replace the corresponding sequences in the pSIV-LTR-2Δ15tetO-TAR^m-luc plasmid, which resulted in the TAR deletion variants of this reporter construct. The plasmid pRL-CMV (Promega), in which the expression of Renilla luciferase is controlled by the cytomegalovirus (CMV) immediate-early enhancer/promoter, was cotransfected into C33A cells to allow correction for differences in transfection efficiency.

C33A cells were cultured in 2-cm² wells and transfected with 20 ng pSIV-LTR-2Δ15tetO-TAR-luc construct (TAR^m, TAR^m+63, TAR^wt, or TAR deletion variant), 0.5 ng pRL-CMV, and 0.4 ng rtTA expression plasmid pCMV-rtTA_{F86Y A209T} (19) or 0 to 50 ng SIV Tat expression plasmid pcDNA3-SIV-Tat (18). pBluescript was added to the transfection mix as carrier DNA to a total of 1 μg of DNA. After transfection, C33A cells were cultured for 48 h with 0 to 1,000 ng/ml DOX. Cells were lysed in passive lysis buffer, and firefly and Renilla luciferase activities were determined with a dual-luciferase assay (Promega). The expression of firefly and Renilla luciferases was within the linear range, and no squelching effects were observed. The promoter activity was calculated as the ratio between the firefly and Renilla luciferase activities and corrected for between-session variation (41).

Isolation and Northern blot analysis of viral RNA.

For the isolation of intracellular viral RNA, 293T cells were transfected with 5 μg of the pSIV-rtTA constructs by calcium phosphate precipitation and cultured in 10-cm² wells with 1,000 ng/ml DOX (19). After 48 h, cells were washed with phosphate-buffered saline (PBS), lysed in 350 μl RLT buffer (Qiagen), and homogenized with a QIAshredder column (Qiagen). Total cellular RNA was isolated with an RNeasy kit (Qiagen), and contaminating DNA was removed with RNase-free DNase during isolation (Qiagen). Upon electrophoresis of 5 μg RNA in a 1% agarose gel in 1× morpholinepropanesulfonic acid (MOPS) buffer (40 mM MOPS, 10 mM sodium acetate, pH 7.0) with 6.5% formaldehyde at 100 V, RNAs were transferred onto a positively charged nylon membrane (Boehringer Mannheim) overnight by means of capillary force. RNAs were linked to the membrane in a UV cross-linker (Stratagene). An NdeI fragment spanning part of the rtTA gene and the U3 region of SIV-rtTA was labeled with [α-³²P]dCTP by use of a High Prime DNA labeling kit (Roche). Prehybridization and hybridization of the membrane with the probe were done in ULTRAhyb buffer (Ambion) at 55°C for 1 and 16 h, respectively. The membrane was then washed at room temperature in low-stringency buffer (2× SSC-0.2% sodium dodecyl sulfate [1× SSC is 0.15 M NaCl plus 0.015 M sodium citrate]) and at 55°C in high-stringency buffer (0.1× SSC-0.2% sodium dodecyl sulfate). Images were obtained using a Storm 860 phosphorimager (Amersham Biosciences), and data analysis was performed with the ImageQuant software package. The sizes of SIV-rtTA RNA fragments were estimated from their migration on the Northern blot, using 18S and 28S rRNAs as size markers.

Cell and virus cultures.

293T cells were transfected with 5 μg of the pSIV-rtTA constructs by calcium phosphate precipitation and cultured in 10-cm² wells with 1,000 ng/ml DOX as previously described (19). Culture supernatants were harvested 48 h after transfection and filtered through 0.45-μm-pore-size filters. Virus production was determined by CA-p27 enzyme-linked immunosorbent assay (Coulter).

The PM1 T-cell line was cultured at 37°C and 5% CO₂ in RPMI 1640 containing 10% fetal bovine serum, 100 units/ml penicillin, and 100 units/ml streptomycin. To assay virus replication, PM1 cells were infected by incubation with an equal amount of each SIV-rtTA variant, corresponding to 1 ng CA-p27, for 4 h at 37°C. Cells were then extensively washed with PBS and cultured in 2-cm² wells in complete RPMI medium with 1,000 ng/ml DOX. Cell-free supernatants were collected and viral replication was monitored with a real-time PCR-based reverse transcriptase (RT) assay (18). Each experiment was performed in triplicate.

Peripheral blood mononuclear cells (PBMC) were isolated from cynomolgus macaques (36) and cultured as described previously (17). Cells were activated with 2 μg/ml phytohemagglutinin for 2 days and infected with equal amounts of virus (10 ng CA-p27) for 16 h as described for PM1 cells. Cells were maintained with 100 units/ml recombinant interleukin-2 and 1,000 ng/ml DOX following infection. The RT levels in cell-free supernatants were determined to monitor viral replication.

RESULTS

Construction of SIV-rtTA variants in which TAR is partially or completely deleted.

We previously constructed a conditionally live SIVmac239 variant that replicates exclusively in the presence of DOX (18). In this SIV-rtTA variant, the Tat-TAR transcription activation mechanism was functionally replaced by the DOX-inducible Tet-On gene regulation system (Fig. 2A). Briefly, the Tat-TAR axis was inactivated through mutation of the bulge and loop sequences in the SL1 and SL2 domains of TAR, which prevents the binding of Tat and cyclin T1 (18). The Tet-On system was incorporated into the SIVmac239 genome by insertion of the rtTA transcriptional activator gene in place of the nef gene and tet operator sites (teto) in the LTR promoter (19). In the presence of DOX, rtTA can bind to teto and activate transcription from the downstream transcription initiation site. Transcription and replication of SIV-rtTA are thus critically dependent on DOX. Although SIV-rtTA does not require Tat for transcription, mutational inactivation of Tat abolished viral replication (18). This result indicated that Tat has an additional function in SIVmac replication, and the wild-type tat gene was therefore maintained in SIV-rtTA. Upon long-term culturing, SIV-rtTA acquired an A-to-G substitution at position +63 in the SL2 domain of TAR that strongly improved viral replication. This mutation did not restore Tat binding, and transcription of the new SIV-rtTA variant was still fully dependent on DOX (17).

FIG. 2.

Mutation of the TAR element in the DOX-inducible SIV-rtTA variant. (A) In the DOX-dependent SIV-rtTA variant, the Tat-TAR transcription mechanism was functionally replaced by the Tet-On gene regulation mechanism. TAR was inactivated by mutation of the bulge and loop sequences (TAR^m). The rtTA gene was inserted at the site of the nef gene, and teto binding sites were introduced in both the 5′LTR and 3′LTR promoter regions. (B) We generated a set of partial to complete TAR deletion variants. The constructs used in this study have a wild-type (TAR^wt), mutated (TAR^m; bulge and loop mutations are shown with black circles), or evolved (TAR^m+63; +63^A-G substitution is shown with a gray circle) TAR element. We made novel variants in which either SL1 (1^Δ2^WT, 1^Δ2^m, and 1^Δ2⁺⁶³), SL2 (1^wt2^Δ and 1^m2^Δ), or SL3 (1^m2⁺⁶³3^Δ) was deleted. In addition, we made a ΔTAR variant, in which all TAR sequences were deleted (not shown), and a TAR10Δ variant, in which TAR was truncated to a 20-nt hairpin corresponding to nt +1 to +10 and +114 to +124.

Since SIV-rtTA does not need TAR for transcription from the LTR promoter, it is an ideal tool for studying additional TAR functions in SIVmac replication. We therefore constructed novel SIV-rtTA variants with different TAR configurations (Fig. 2B). Compared to wild-type SIVmac239 (TAR^wt), SIV-rtTA carries mutations in the bulge and loop sequences in both SL1 and SL2 (TAR^m). In this SIV-rtTA context, we deleted either SL1 (1^Δ2^m variant) or SL2 (1^m2^Δ). We also used the improved SIV-rtTA variant with the additional TAR mutation (TAR^m+63), in which either SL1 (1^Δ2⁺⁶³) or SL3 (1^m2⁺⁶³3^Δ) was deleted. We also generated variants in which a short 8-bp stem-loop structure, formed by the first and last 10 nt of TAR, was left (TAR10Δ) or in which the complete TAR sequence was deleted (ΔTAR). As controls, we constructed variants with a wild-type TAR (TAR^wt) and derivatives with a deletion of SL1 (1^Δ2^wt) or SL2 (1^wt2^Δ).

Efficient gene expression upon partial or complete TAR deletion.

To determine the effects of partial and complete TAR deletion on viral gene expression, we generated promoter-reporter constructs in which the TAR mutant LTR promoter of the different SIV-rtTA variants was placed upstream of a luciferase reporter gene (Fig. 3A). To assay DOX responsiveness, C33A cells were transfected with the LTR-luciferase constructs and an rtTA-expressing plasmid and cultured with or without DOX. The luciferase level measured 48 h after transfection reflects the efficiency of viral gene expression. All constructs showed a low level of luciferase expression in the absence of DOX and a high level with DOX (Fig. 3B). Deletion of SL1 (compare 1^Δ2^m with TAR^m, 1^Δ2⁺⁶³ with TAR^m+63, and 1^Δ2^wt with TAR^wt), SL2 (compare 1^m2^Δ with TAR^m and 1^wt2^Δ with TAR^wt), or SL3 (compare 1^m2⁺⁶³3^Δ with TAR^m+63) resulted in an increased level of gene expression. This can be explained by destabilization of the TAR structure, which has been reported to have an inhibitory effect on mRNA translation (24, 42, 43). In agreement with this, the highest luciferase levels were observed upon nearly complete or complete deletion of TAR (TAR10Δ and ΔTAR).

FIG. 3.

Gene expression is not affected by TAR deletions. (A) Promoter-reporter constructs were generated in which the firefly luciferase gene was under the control of the original or TAR-mutated SIV-rtTA LTR promoter. (B) To determine DOX responsiveness, C33A cells were transfected with the reporter constructs and an rtTA-expressing plasmid and cultured with 0 or 1,000 ng/ml DOX. (C) To determine Tat responsiveness, C33A cells were transfected with the reporter constructs and 0 or 50 ng of Tat-expressing plasmid. (B and C) A CMV promoter-Renilla luciferase construct was cotransfected to allow correction for differences in transfection efficiency. The ratios of the firefly and Renilla luciferase activities measured 2 days after transfection reflect the LTR promoter activity. The error bars represent the standard deviations for three experiments.

The Tat responsiveness of the constructs was determined by transfection of C33A cells with the LTR-luciferase constructs and a Tat-expressing plasmid. As previously observed, the TAR^m and TAR^m+63 constructs did not show any Tat stimulation (Fig. 3C), which is in agreement with the finding that SIV Tat cannot bind to TAR^m and TAR^m+63 RNAs (17, 18). As expected, derivatives of the TAR^m and TAR^m+63 constructs in which either SL1 (1^Δ2^m, 1^Δ2⁺⁶³), SL2 (1^m2^Δ), or SL3 (1^m2⁺⁶³3^Δ) was deleted and the variants with a complete or nearly complete TAR deletion also did not respond to Tat. Only the control constructs with wild-type TAR (TAR^wt) and its variants in which either SL1 or SL2 was deleted (1^Δ2^wt and 1^wt2^Δ) showed luciferase expression with Tat. The 1^Δ2^wt and 1^wt2^Δ variants exhibited weaker Tat trans-activation capacities than that of the TAR^wt construct, which is in agreement with previous studies demonstrating that both SL1 and SL2 contribute to optimal Tat activation of transcription (10, 25). The contribution of SL1 is quantitatively more important than that of SL2, which could be due to hairpin characteristics (e.g., the presence of a purine in the SL2 bulge) (26, 40). These results demonstrate that partial or complete deletion of TAR allows efficient, DOX-dependent, Tat-independent gene expression from the SIV-rtTA promoter.

TAR deletions do not affect SIV-rtTA production.

To assess whether partial or complete deletion of TAR affects the production of viral RNA and proteins, we introduced the TAR deletions into both the 5′LTR and 3′LTR of SIV-rtTA. 293T cells were transfected with the SIV-rtTA variants and cultured in the presence of DOX for 48 h. Northern blot analysis of the viral RNAs isolated from the transfected cells revealed no significant quantitative differences in unspliced, singly spliced, and multispliced RNAs of the original and TAR-deleted SIV-rtTA variants (Fig. 4). The observed differences in transcript size are due to the TAR deletion in both the 5′LTR and 3′LTR. This result demonstrates that the partial or complete deletion of TAR in SIV-rtTA does not significantly affect the production and processing of viral RNA. Virus production was determined by measurement of CA-p27 (Fig. 5) and RT activity (not shown) in the culture supernatant. Both analyses revealed similarly high levels of virus production for all SIV-rtTA variants. Whereas increased reporter gene expression was observed upon TAR truncation (Fig. 3B), CA-p27 production of the different TAR-truncated SIV-rtTA variants did not show this trend, possibly due to the presence of the complete untranslated leader in these viral RNAs (Fig. 5). These results demonstrate that partial or complete deletion of TAR does not impair SIV-rtTA virion production.

FIG. 4.

TAR deletions do not affect viral RNA production. 293T cells were transfected with the SIV-rtTA variants and cultured with 1,000 ng/ml DOX for 48 h. Total intracellular RNA was isolated and analyzed by Northern blot analysis with an SIV-specific probe. (A) The positions of multispliced (msRNA), singly spliced (ssRNA), and unspliced (usRNA) RNAs are indicated on the blot. The size differences between the SIVmac239, SIVmac239ΔNef, and SIV-rtTA RNAs are due to deletion of nef sequences, introduction of rtTA and teto sequences, and/or deletion of TAR sequences. (B) Ethidium bromide staining of the corresponding agarose gel demonstrates that the observed differences in RNA signals are due to differences in RNA loading. The 18S (∼1.9 kb) and 28S (∼5.0 kb) rRNAs were used as size markers.

FIG. 5.

TAR deletions do not impair virus production. 293T cells were transfected with the SIV-rtTA variants and cultured for 2 days with 1,000 ng/ml DOX. The virus level in the culture supernatant was determined by CA-p27 enzyme-linked immunosorbent assay. The error bars represent the standard deviations for three experiments.

TAR-deleted SIV-rtTA variants replicate efficiently.

To study the impact of partial and complete TAR deletion on virus replication, PM1 T cells were infected with the SIV-rtTA variants and viral replication was monitored by measurement of the RT level in the culture supernatant. SIV-rtTA-TAR^m+63 replicated more efficiently than the original SIV-rtTA-TAR^m construct and almost as efficiently as the parental strains SIVmac239 and SIVmac239ΔNef (Fig. 6A). Remarkably, the SIV-rtTA-TAR^m derivatives in which either SL1 (1^Δ2^m) or SL2 (1^m2^Δ) was deleted replicated as efficiently as the TAR^m+63 variant (Fig. 6B). Deletion of SL1 (1^Δ2⁺⁶³) or SL3 (1^m2⁺⁶³3^Δ) in the TAR^m+63 variant did not affect viral replication. The TAR10Δ variant, in which TAR was almost completely deleted, also replicated efficiently (Fig. 6C). Only the ΔTAR variant with a complete TAR deletion exhibited a reduced replication potential. This ΔTAR mutation alters the sequence around the transcription start site, which may affect the selection of the initiation site. Sequence analysis of the 5′ ends of the viral transcripts did indeed reveal that transcription of SIV-rtTA-ΔTAR frequently started downstream of the original +1 position (2 or more nt downstream [data not shown]). This aberrant initiation of transcription may explain the reduced replication capacity of the ΔTAR variant. In the TAR10Δ variant, in which only the first and last 10 nt of the TAR sequence were maintained, transcription initiated at the original +1 position, and this virus did indeed replicate efficiently. These results demonstrate that partial or nearly complete deletion of TAR does not impair replication of SIV-rtTA in PM1 cells. In agreement with this, sequence analysis of these TAR-deleted variants upon prolonged culturing (up to 200 days) revealed that the introduced mutations were stably maintained, and none of the variants reverted to a wild-type TAR structure (data not shown).

FIG. 6.

Effects of TAR deletions on virus replication. (A to C) PM1 cells were infected with SIV-rtTA variants (1 ng CA-p27) and cultured with 1,000 ng/ml DOX. Replication of the original (TAR^m) and evolved (TAR^m+63) SIV-rtTA variants and of wild-type and Nef-deleted SIVmac239 (A), of variants with a partial deletion of TAR (B), and of variants with a nearly complete to complete TAR deletion (C) is shown. Viral replication was monitored by quantification of the RT level in the culture supernatant. Each experiment was performed in triplicate, and data for a representative experiment are shown. Similar results were obtained when the infected cells were cultured with 100 ng/ml DOX and when virus cultures were started by transfection of either 1 or 5 μg of the proviral plasmids (data not shown). (D) PBMC isolated from cynomolgus macaques were infected with the TAR10Δ and TAR^m+63 variants and with wild-type and Nef-deleted SIVmac239 (10 ng CA-p27) and then cultured with 1,000 ng/ml DOX.

In addition, we assessed replication of the TAR10Δ and TAR^m+63 variants in PBMC isolated from cynomolgus macaques (Fig. 6D). As observed with PM1 cells, these variants replicated with similar efficiencies in macaque PBMC, which demonstrates that the nearly complete deletion of TAR also does not affect replication of SIV-rtTA in primary cells. Both DOX-controlled viruses replicated somewhat less efficiently than wild-type SIVmac239 and the SIVmac239ΔNef variant, which were included for comparison.

DISCUSSION

In this study, we analyzed the role of the complex TAR RNA element in SIVmac replication. For this purpose, we used a SIV-rtTA variant that does not depend on TAR for transcription, which allowed us to address the issue of whether TAR has an additional function in SIVmac replication. We demonstrated that partial or nearly complete deletion of TAR does not affect viral transcription, RNA processing, production, or replication. Therefore, we conclude that TAR has no essential function in SIVmac replication in vitro apart from its role in Tat-mediated activation of transcription.

We recently showed that TAR deletion is also compatible with replication of a similar DOX-dependent HIV-1 variant (HIV-rtTA) (15). While SIVmac has an ∼124-nt TAR element that folds into a complex three-stem-loop domain, the HIV-1 TAR element is a single hairpin of only ∼57 nt (Fig. 1). Apparently, this relatively simple TAR structure suffices for HIV-1 replication. The strong conservation of the complex TAR structure in different SIV isolates suggested that all three stem-loop elements may play a certain role in SIV replication. Previous studies (10, 25) and our observations (Fig. 3C) did indeed reveal that efficient Tat-mediated activation of transcription involves both SL1 and SL2. No role for SL3 in the Tat trans-activation function has been reported (25). In this study, we showed that none of the stem-loop domains has an essential function in any other replication step. However, it cannot be excluded that SL1, SL2, and/or SL3 may have an accessory function in the SIVmac life cycle under specific conditions or in specific cell types in vivo, since replication of the TAR-deleted SIV-rtTA variants was assayed only in cell culture systems (32).

TAR is part of the repeat (R) sequence present at the 5′ and 3′ ends of the viral genome. This region is essential for the first-strand transfer step of reverse transcription, in which the copy DNA is translocated from the 5′ to the 3′ R sequence. Whereas the R region is 176 nt in wild-type SIVmac239, the efficient replication of the TAR10Δ variant demonstrates that a 72-nt R sequence suffices for this function. This observation is in agreement with previous HIV-1 studies demonstrating that strand transfer is efficient with shortened R regions of only 30 nt (12) and 39 nt (15).

We previously showed that HIV-1 transcripts require a perfect hairpin structure at the extreme 5′ end to avoid the presence of an unpaired 5′ dangling end (15). TAR fulfills this prerequisite in wild-type HIV-1, but a truncated TAR hairpin or an unrelated hairpin also suffices in the context of HIV-rtTA. All SIV-rtTA TAR-mutated transcripts can fold a base-paired stem structure at the 5′ end, and even the severely truncated 8-bp hairpin of the TAR10Δ variant supports efficient replication. However, complete deletion of TAR, as in the ΔTAR variant, shifted the transcription start site further downstream. These aberrantly initiated transcripts start within the poly(A) hairpin and will not fold into a stable stem-loop structure at the 5′ end (Fig. 1B), resulting in the presence of unpaired nucleotides. As recently observed for HIV-1, such unpaired nucleotides may interact with downstream RNA sequences and affect the overall leader RNA folding (30, 37, 44). The secondary structure of this leader RNA region is important for several processes in the lentiviral life cycle, such as dimerization, packaging, splicing, and polyadenylation (1, 16, 21-23, 33-35, 38, 45, 46). Any modification of the leader RNA structure will likely hamper viral replication, which may explain the delayed replication of the ΔTAR variant.

The reduced replication of the original SIV-rtTA variant (TAR^m) may also be due to aberrant folding of the leader RNA. Possibly, the introduced bulge and loop mutations destabilize the local TAR folding or generate a new sequence with complementarity to downstream sequences, which could trigger an interaction between TAR sequences and a downstream leader domain. The beneficial +63^A-G mutation that is selected during SIV-rtTA replication may prevent this unwanted interaction and thus restore viral replication (17). Although details of such a misfolding scenario are currently lacking, this scenario is supported by the observation that the precise removal of SL1 (1^Δ2^m variant), SL2 (1^m2^Δ), or all SL sequences (TAR10Δ) similarly improves SIV-rtTA replication.

SIV-rtTA can be used as a tool in SIV biology studies and vaccine research because both the level and duration of replication can be controlled by DOX administration. We previously demonstrated that SIV-rtTA requires wild-type Tat protein for replication, although gene expression is strictly controlled by the incorporated Tet-On system and not affected by Tat inactivation. This result suggests that Tat, unlike TAR, may have an important second function in the SIVmac life cycle, in addition to its transcriptional role. For this reason, the SIV-rtTA variant used in this study encodes the wild-type Tat protein. Reversion of the bulge and loop mutations in TAR^m, which had been introduced to prevent Tat and cyclin T1 binding, would restore the Tat-TAR mechanism of transcription control. However, such a repair route requires multiple nucleotide substitutions in the bulge and loop of TAR, which are not likely to occur. Indeed, we never observed restoration of the Tat-TAR axis in multiple long-term cultures of SIV-rtTA. Nevertheless, the likelihood of this unwanted evolution route is further reduced by the deletion of TAR sequences, and it seems highly unlikely that the TAR10Δ variant can revert to a Tat-TAR transcription mechanism.