Tristetraprolin inhibits HIV-1 production by binding to genomic RNA

Microbes and Infection 8 (2006) 2647e2656 www.elsevier.com/locate/micinf

Original article

Tristetraprolin inhibits HIV-1 production by binding to genomic RNA Masae Maeda a,b,c, Hirofumi Sawa c,d,e, Minoru Tobiume a, Kenzo Tokunaga a, Hideki Hasegawa a, Takeshi Ichinohe a, Tetsutaro Sata a, Masami Moriyama b, William W. Hall f, Takeshi Kurata a, Hidehiro Takahashi a,* a

Department of Pathology, National Institute of Infectious Diseases, Toyama 1-23-1, Shinjuku-ku, Tokyo 162-8640, Japan b Department of Microbiology and Immunology, School of Medicine, Keio University, Tokyo, Japan c CREST, JST, Japan d Department of Molecular Pathobiology, Hokkaido University Research Center for Zoonosis Control, Hokkaido University, Sapporo, Japan e 21st Century COE Program for Zoonosis, Hokkaido University, Sapporo, Japan f Department of Medical Microbiology, Conway Institute of Biomolecular and Biomedical Research, University College Dublin, Belfield, Dublin, Ireland Received 2 February 2006; accepted 18 July 2006 Available online 8 August 2006

Abstract HIV-1 genome has an AU-rich sequence and requires rapid nuclear export by Rev activity to prevent multiple splicing. HIV-1 infection occurs in activated CD4þ T cells where the decay of mRNAs of cytokines and chemokines is regulated by the binding of AU-rich elements to the mRNA-destabilizing protein tristetraprolin. We here investigated the influence of tristetraprolin on the replication of HIV-1. Treatment of siRNA against tristetraprolin in a latently HIV-1 infected cell line increases HIV-1 production following stimulation. A chloramphenicol acetyltransferase and luciferase assay revealed that exogenous tristetraprolin reduced HIV-1 virion production and in contrast increased the multiply spliced products. Furthermore, quantitative RTePCR analysis showed tristetraprolin increases the ratio of multiple-spliced RNAs to un-, single-spliced RNA. Moreover, an electrophoretic mobility shift assay showed that tristetraprolin binds to synthesized HIV-1 RNA with AU-rich sequence but not to RNA with less AU sequence. These results suggest that tristetraprolin is a regulator of HIV-1 replication and enhances splicing by direct binding to AU-rich sequence of HIV-1 RNAs. Ó 2006 Elsevier Masson SAS. All rights reserved. Keywords: Human immunodeficiency virus type 1; Tristetraprolin; AU-rich element

1. Introduction Expression of human immunodeficiency virus type 1 (HIV-1) genes is regulated by several posttranscriptional mechanisms,

Abbreviations: HIV-1, human immunodeficiency virus type 1; TTP, tristetraprolin; ARE, AU-rich element; RRE, Rev-responsive element; nt, nucleotide; RT, reverse transcription; PCR, polymerase chain reaction; CAT, chloramphenicol acetyltransferase; VSV-G, vesicular stomatitis virus envelope glycoprotein. * Corresponding author. Tel.: þ81 3 5285 1111; fax: þ81 3 5285 1189. E-mail address: [email protected] (H. Takahashi). 1286-4579/$ - see front matter Ó 2006 Elsevier Masson SAS. All rights reserved. doi:10.1016/j.micinf.2006.07.010

which include RNA splicing, stability, transport and translation. For example, HIV-1 RNA contains inhibitory sequences (INS) negatively regulating expression [1], which is also known to be AU-rich [2]. Replacement of AU residues of HIV-1 mRNA has been reported to result in a marked increase in expression of HIV-1 Gag, Pol, and Env proteins independent of the Crm1/Rev/Rev-responsive element (RRE) export pathway [3]. In addition, another nuclear export system, constitutive transport elements (CTE)/Tap, can substitute Rev/ RRE [4], suggesting that effective nuclear export is required to complement the character of AU-rich RNA genome. As we have previously reported that HIV-1 RNA is fragmented

2648

M. Maeda et al. / Microbes and Infection 8 (2006) 2647e2656

in viral particles and in vitro-synthesized HIV-1 RNA is cleaved between U and A or C and A [5], effective nuclear export of HIV-1 RNA by Crm1/Rev/RRE might be required to rescue a potentially unstable RNA genome, and the balance between RNA decay and nuclear export could be an important facet of HIV-1 replication and pathogenesis. Whereas activated CD4þ T cells are a primary target for productive HIV-1 infection [6], HIV-1 production does not occur in resting CD4þ T cells because of a low level of reverse transcriptase activity [7]. In activated cells, including CD4þ T cells that produce cytokines, chemokines and other proteins in response to inflammation and infection, mRNAs are degraded rapidly following transient activation [8]. These mRNAs contain the AU-rich element (ARE) in their 3 0 -untranslated region (3 0 UTR) that binds with tristetraprolin (TTP, also known as TIS11, Nup475, or G0S24) [9]. TTP is the prototype of a family of proteins that possess a pair of closely spaced zinc fingers of the CCCH class and are capable to binding AU-rich elements (ARE) in the 3 0 UTR and subsequent destruction of transcripts of pro-inflammatory mediators such as tumor necrosis factor a, granulocyte/macrophage colony stimulating factor, and cyclooxygenase 2 [10]. We propose the hypothesis that TTP might interact with the AU-rich regions of HIV-1 RNA and that the regulation of the HIV-1 RNA genome could play important roles in the posttranscriptional control of HIV-1 replication in activated or memory T cells. We show that TTP reduced HIV-1 virion production by enhancing multiple splicing through binding to HIV-1 AU-rich RNA. 2. Materials and methods 2.1. Cell culture The human cell lines HEK293T and HeLa were maintained under an atmosphere of 5% CO2 at 37  C in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal bovine serum. U1 cells [11] were cultured in complete medium (RPMI 1640 medium supplemented with 10% fetal bovine serum). 2.2. RNAi inhibition siRNA treatment was achieved using synthetic oligonucleotides which were purchased from Ambion. Sense sequence of TTP siRNA is CCC AUA AAU CAA UGG GCC Ctt (small capital displays deoxyribonucleic acid) and antisense is GGG CCC AUU GAU UUA UGG Gtg. Two rounds of siRNA treatments were performed. siRNA was transfected by using the electric transfection apparatus, amaxa in U1 cells. 2.3. Cloning, mutagenesis, and plasmid construction for TTP or rev expression vectors Polyadenylated RNA was isolated from HeLa cells and subjected to reverse transcription (RT) with SuperScript RNaseH reverse transcriptase (Invitrogen Life Technologies) and a 15-nucleotide poly(dT) primer. The coding region of

human TTP cDNA was amplified by polymerase chain reaction (PCR) with the reverse-transcribed cDNA as the template and with the primers 5 0 -CGT GAATTC ATG GAT CTG ACT GCC ATC TAC GAG AGC CT-3 0 (TTPF, EcoRI site underlined) and 5 0 -GAC CGG GCA G GCGGCCGC TCA CTC AGA AAC AGA GAT-3 0 (TTPR, NotI site underlined). The PCR product was digested with EcoRI and NotI and then cloned into pcDNA4/HisMax-C (Invitrogen Life Technologies), a mammalian expression vector containing the Xpress epitope tag sequence, or into pCAGGS-IRES-EGFP [12], yielding pchTTPwt or pCAGGSTTP, respectively. The resulting plasmids were sequenced with a DNA sequencer (model 377A; Perkin Elmer, Norwalk, CT). Expression vectors for TTP deletion mutants, including 1e101, 76e189, and 176e 320, were similarly constructed with pchTTPwt as the template and with the primers 5 0 -TCT CTG AG GCGGCCGC TTA TAG CTC AGT CTT GTA GCG CGA -3 0 (R1e101), 5 0 -CTG GCT GAATTC CTG GGC CCT GAG CTG TCA CCCT-3 0 (F76e189), 5 0 -AGG CCT GGT GCGGCCGC TTA GGT CCG GCG GCC AGA GGG CA-3 0 (R76e189), and 5 0 -CCT GTG GAATTC CAG AGC ATC AGC TTC TCC GGC CT-3 0 (F176e320). The resulting expression vectors were designated pcD1e101, pcD76e189, and pcD176e 320, respectively. For construction of vectors for Xpress-tagged Rev, the Rev cDNA was amplified by PCR with pSRaRev [13] as the template and with the primers 5 0 -AAA AAA AGATCT ATG GCA GGA AGA-3 0 (Rev-BglII) and 5 0 -AAA AAAGTCGAC CTA TTC TTT AGTT-3 0 (Rev-SalI). The PCR product was digested with BglII and SalI and then cloned into pEGFP-C1 (BD Biosciences Clontech, Palo Alto, CA) or pcDNA4/HisMax-C that had been digested with BamHI and XhoI; the resulting vectors were designated pEGFP-Rev or pcMax-Rev. 2.4. Luciferase, chloramphenicol acetyltransferase (CAT), and p24 enzyme-linked immunosorbent (ELISA) assays HEK293T cells were cotransfected with HIV-1 proviral DNA (pNL43 [14], pNL-Luc-ERþ [15], or pNL-enCAT [16]) and TTP expression vectors (pchTTPwt, pcD1e101, pcD76e189, or pcD176e320) with the use of Fugene 6 (Roche, Mannheim, Germany) or by the calcium phosphate method. The amount of virus in culture supernatants was quantified by measurement of p24 antigen with a p24 Gag antigen capture ELISA assay (ZeptoMetrix, Buffalo, NY). For measurement of luciferase activity or CAT [17], cells were lysed and assayed with a Luciferase Assay System (Promega, Madison, WI) and a Lumat LB 96V luminometer (Perkin Elmer) or with a CAT ELISA (Roche). Expression of wild-type and mutant TTP was examined by immunoprecipitation, SDSepolyacrylamide gel electrophoresis and immunoblot with mouse monoclonal antibodies to the Xpress epitope (Invitrogen) analysis or goat polyclonal antibodies to TTP (Santa Cruz Biotechnology, Santa Cruz, CA). The antibodies to Nef and to p24 also used for immunoblot analysis were kind gifts from Dr. Ikuta. Signals were obtained

M. Maeda et al. / Microbes and Infection 8 (2006) 2647e2656

Cellular total RNA or virion RNA was subjected to RT with a 15-nucleotide poly(dT) primer and the use of an Omniscript kit (Qiagen, Valencia, CA). The resulting cDNA was subjected to quantitative analysis with a real-time, 5 0 -exonuclease PCRbased assay (TaqMan) and primereprobe combinations that were selected with the use of Primer Express software (Applied Biosystems, Foster City, CA) as specific for the coding regions of the tat, env, or luciferase genes. The reaction was performed with a QuantiTect Probe PCR kit (Qiagen) and an amplification protocol comprising incubation at 95  C for 15 min followed by 45 cycles of 95  C for 15 s and 60  C for 1 min. The primers 1459F (5 0 -GGT CCT ATG ATT ATG TCC GGT TATG-3 0 ) and 1535R (5 0 -TGT AGC CAT CCA TCC TTG TCAA-3 0 ), and the probe 1491P (5 0 -FAM-TCC GGA AGC GAC CAA CGC CTT-TAMRA-3 0 , in which FAM indicates 6-carboxyfluorescein and TAMRA indicates N,N,N 0 ,N 0 -tetramethyl-6-carboxyrhodamine) were used for detection of luciferase cDNA; the primers 7484F (5 0 -ATA AAC ATG TGG CAG GAA GTA GGAA-3 0 ) and 7572R (5 0 -AGC AGC CCA GTA ATA TTT GAT GAAC-3 0 ), and the probe 7513P (5 0 -FAM-AAT GTA TGC CCC TCC CAT CAG TGG ACA-TAMRA-3 0 ) were used for that of env cDNA; the primers 6016F (5 0 -CAG ACT CAT CAA GCT TCT CTA TCA AAG-3 0 ) and 8465R (5 0 -CGT TCA CTA ATC GAA TGG ATC TGT-3 0 ), and the probe 8392P (5 0 FAM-ACC CGA CAG GCC CGA AGG AAT AGA ATAMRA-3 0 ) were used for that of rev, tat and nef cDNA; the primers 5918F (5 0 -GTT GCT TTC ATT GCC AAG TTTG-3 0 ) and 6017R (5 0 -TGA CTG TTC TGA TGA GCT CTT CGT-3 0 ), and the probe 5945P (5 0 -FAM-TGA CA A AAG CCT TAG GCA TCT CCT ATG GC-TAMRA-3 0 ) were used for that of the second exon of tat; the primers TTP 405F (5 0 -GCT GCG CCA GGC CAA TC-3 0 ) and TTP 477R (5 0 -GCA GCG GCC CTG GAG GTA-3 0 ), and the probe TTP 423 (5 0 -FAM- CCA CCC CAA ATA CAA GAC GGA ACT CTG TC-3 0 ) were used for that of TTP cDNA. The transcripts detected by the env region primers and probe include both unspliced (9 kb) and singly-spliced (4 kb) RNA molecules, whereas those detected by the tat, rev and nef cDNA primers and probe include multiply-spliced transcripts (2 kb) [15]. The RNA copy numbers of luciferase, HIV-1 transcripts and TTP were determined with pGL3-promoter, pNL-LucERþ, pSRaRev and pchTTPwt as a standard, respectively. Vesicular stomatitis virus envelope-glycoprotein (VSV-G)pseudotyped HIV-1 was produced by cotransfection of pNLLuc-ERþ and the VSV-G expression vector pHIT/G [18] in HEK293T cells. VSV-G-pseudotyped NL-Luc virus was infected to HEK293T cells transfected with TTP-expression vector or an empty vector. Following exposure to actinomycin D, an inhibitor of transcription, the copy numbers of multiple

*

7

Fold increase of TTP mRNA

2.5. RTePCR and virus infection

A

B

Control siRNA TTP siRNA

6 5 4 3 2 1 0

0

4

10 hours

2 1.8

Fold increase of control P24

with the ECL detection system (Amersham Pharmacia) and digitized with an LAS1000 imaging system (Fujifilm, Tokyo, Japan).

2649

1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0

Control siRNA

TTP siRNA

Fig. 1. Reduction of endogenous TTP increased HIV-1 production in U1 cells. (A) U1 cells were treated with two rounds of siRNA for TTP and controls. Following exposure to 7.5 ng/ml of PMA for 0, 4, or 10 h, the amount of TTP mRNAs were determined by RTePCR and represented as the relative ratio against that of the cells before exposure to PMA. Data are means  S.D. of values from three independent experiments. (B) Viral products (p24) of the supernatants of similarly treated cells as in (A) after 48 h post PMA stimulation were measured. *P < 0.05.

spliced or un-, singly spliced transcripts were determined by RTePCR.

2.6. Preparation of RNA substrates Substrate RNAs were prepared by in vitro transcription with T7 RNA polymerase and DNA amplified by the PCR as the template. DNA fragments that include the T7 promoter and encode RNA substrates of 75 and 93 nucleotides (nt) corresponding to the p7 and RRE were thus synthesized by PCR with the following primers: p7F (5 0 -TAA TAC GAC TCA CTA TAG GGA ACA AAT CCA GCT ACC ATA ATG ATA-3 0 ) (underlining denotes T7 promoter) and P7R (5 0 -ATT GAA ACA CTT AAC AGT CTT-3 0 ), corresponding to nt 1900 to 1971 of pNL43, GenBank accession number, M19921; stem II of RRE; U-GF (5 0 - AAA TAA TAC GAC TCA CTA TA GGGAG CAG CAG GAA GCA CTA TGG GCT-3 0 ) and UGR (5 0 -TTG TTC TGC TGC TGC ACT ATA TCA GAC AAT-3 0 ), corresponding to nt 7783 to 7875. The RNA molecules were labeled at the 5 0 end with [g-32P]ATP and

M. Maeda et al. / Microbes and Infection 8 (2006) 2647e2656

2650

2.8. Immunofluorescence analysis

polynucleotide kinase after dephosphorylation with calf intestinal phosphatase. They were purified by electrophoresis on denaturing 15% polyacrylamide gels containing 7 M urea with the use of a Mini Whole Gel Eluter (Bio-Rad, Hercules, CA). The concentration of RNAs was determined by spectrophotometer (NanoDrop technology, Wilmington, DE).

HeLa cells were transfected with pEGFP-Rev and either pcDNA4/HisMax-C, pchTTPwt, pcD1e101, pcD76e189, or pcD176e320. After 24 h, the transfection mixture was removed and the cells were washed twice with phosphate-buffered saline (PBS), treated with 0.5% trypsin in 0.1 mM EDTA, suspended in culture medium, and plated on coverslips. The cells were then incubated for 2 h at 37  C, fixed for 1 h at room temperature with 4% paraformaldehyde in PBS, and permeabilized for 5 min with 0.1% Triton X-100 in PBS. After quenching with 0.1 M NH4Cl at room temperature for 10 min, the cells were incubated with 10% fetal bovine serum in PBS before consecutive exposures for 1 h at room temperature first to antibodies to the Xpress epitope in PBS containing 0.5% gelatin, 10 mM glycine, 10 mM EDTA, and 0.05% NaN3 and then to Alexa Fluor 543-conjugated goat polyclonal antibodies to mouse immunoglobulin

2.7. Analysis of proteins and enzyme purification Recombinant TTP protein was produced by overexpression in Escherichia coli using an His6-tagged prokaryotic expression vector, pQE-9 (Qiagen) encoding TTP cDNA (GenBank accession number, M63625) [19]. Recombinant proteins were purified using an HiTrap chelating column (Amersham Pharmacia, Biotech, Freiburg, Germany) [20]. After staining with SYPRO (Molecular Probes), the polyacrylamide gel was visualized with FLA 2000 (Fujifilm).

Supernatant p24 (%)

A

**

150

NES wt TTP Xpress

50 0

0.4 0.02 0.05 0.1 0.2 0.4 (µg)

CCCH

1-101

Xpress

326

76-189

Xpress

CCCH

CCCH

176-320 Xpress

TTP

E

B

**

kD 50 -

200

20 P or 89 ct t TT 101 6-1 76-3 e 1 7 V w 1

33 26 19 -

100 0

NLS CCCH

102-121 140-159

100

Mock

Viral RNA (%)

D

0.4 0.02 0.05 0.1 0.2 0.4 (µg) Mock

Anti-Xpress

TTP

F p24 (ng/ml)

C

TTP-

0.4 0.02 0.05 0.1 0.2 0.4 (µg) Mock

200 160

**

120 80 40 0

TTP 5

Ve

w cto t r

1-1

1 7 01 6-18 76-3 9 2

0

Fig. 2. Suppression of HIV-1 production by exogenous TTP. HEK293T cells (5  10 per well in a 12-well dish) were cotransfected with HIV-1 proviral DNA (0.04 mg of pNL-Luc-ERþ) and either the indicated amounts of an expression vector for TTP tagged with the Xpress epitope (pchTTPwt) or the corresponding empty vector (0.4 mg). The cells were subsequently cultured for 48 h in fresh medium, after which the amounts of p24 (A) and HIV-1 genomic RNA (the second exon of tat) (B) in the culture supernatants were determined; data are expressed as a percentage of the values for cells transfected with the empty expression vector and are means  S.D. of values from three independent experiments. Cell lysates were also subjected to immunoblot analysis with an antibody to the Xpress epitope tag (C). **P < 0.02. (D) Schematic representation of TTP mutants. Xpress tag, zinc finger domain (CCCH), nuclear export signal (NES), nuclear localization sequence (NLS) and the number of amino acid (326) are indicated. (E) Immunoprecipitation using an antibody to the Xpress epitope of the lysates from HEK293T cells expressing Xpress-tagged wild-type (wt) or mutant forms of TTP, followed by immunoblot analysis with the same antibody. (F) HEK293T cells (5  105) were cotransfected with pNL-Luc-ERþ(0.04 mg) and either an expression vector for wild-type or mutant TTP (0.4 mg) or the corresponding empty vector. The cells were subsequently cultured for 48 h in fresh medium, after which the amount of p24 in the culture supernatants was determined. Data are means  S.D. of values from three independent experiments. **P < 0.02.

M. Maeda et al. / Microbes and Infection 8 (2006) 2647e2656

2651

G (Molecular Probes, Eugene, OR). Fluorescence signals were detected with a confocal laser-scanning microscope (LSM 410; Carl Zeiss, Oberkochen, Germany).

U1 cells transfected with TTP siRNA were found to be increased (Fig. 1B), suggesting that TTP decreases HIV-1 production.

3. Results

3.2. Suppression of HIV-1 production by exogenous TTP

3.1. Reduction of endogenous TTP mRNA levels increased HIV-1 production

To determine whether exogenous TTP is able to affect HIV1 production, we co-expressed human TTP and the HIV-1 clone pNL-Luc-ERþ [24] in HEK293T cells. The amounts of p24 and genomic RNA in culture supernatants were markedly reduced by TTP (maximal inhibition of 97% and 86%, respectively) in a concentration-dependent manner (Fig. 2AeC). We investigated the functional domains of TTP responsible for the inhibition of HIV-1 production by co-expressing various TTP mutants (Fig. 2D,E) with pNL-Luc-ERþ in HEK293T cells. A TTP mutant (1e101) that lacks the COOH-terminal region containing the nuclear localization sequence (NLS) inhibited HIV-1 virion production to an extent similar to that observed with the wild-type protein (Fig. 2F). In contrast, the NH2-terminal deletion mutants 76e189 and 176e320, both of which lack the nuclear export sequence (NES), did not significantly inhibit HIV-1 production (Fig. 2F). The expression level of TTP mutant (176e320) was much lower than other recombinant wild-type and mutants TTPs (1e101 and 76e189) and this result was also observed when the mutants were subcloned into another plasmid, pcDNA4/HisMax-C used in the Fig. 6, suggesting the amino acid residue (1e175) might play a role in stable

We initially examined whether reduction of endogenous TTP in a latently HIV-1 infected cell line would increase HIV-1 production following stimulation. The U1 cell line is transcriptionally latent in terms of HIV-1 expression, and is activated by treatment with several reagents [21,22]. Under normal culture conditions, less than 5% of the cells continuously express HIV-1 antigens [22]. Stimulation of U1 cells with phorbol 12-myristate 13-acetate (PMA) significantly increased the population of HIV-1 antigen-expressing cells up to 40% at 10 h. At the present time, no antibody capable of detecting human endogenous TTP expression is available (personal communication). Therefore, we examined TTP mRNA levels in U1 cells transfected with TTP siRNA or control siRNA. Following treatment of PMA, TTP mRNA peaked at 4 h in U1 cells transfected with control siRNA (Fig. 1A), as previously reported [23]. In contrast, in U1 cells transfected with TTP siRNA PMA stimulation increased TTP mRNA significantly less than cells transfected with control siRNA. Following 48 h stimulation by PMA, p24 levels in the supernatant in

A

rev

pNL43

LTR

gag

pol

vif vpr

tat vpu env

pNL-enCAT

nef

LTR

CAT

pNL-Luc-E-R+ Luc

Unspliced (9 kb) RRE

Single-spliced (4 kb) Multiple-spliced (2 kb)

C 500 400 300 200 100 0

Ve

1 7 w 1 cto t TT -101 6-18 76-3 20 9 P r

D 200

**

150 100 50 0

w 7 1 1 cto t TT -101 6-18 76-3 9 20 P r

Ve

CAT antigen (%)

*

CAT antigen (%)

Luciferase activity (%)

B

150

**

100 50 0

Ve

w 1 7 1 cto t TT -101 6-18 76-3 9 20 P r

Fig. 3. Promotion of multiple splicing of HIV-1 RNA by TTP. (A) Schematic representations of pNL43, pNL-enCAT, and pNL-Luc-ERþ. The sites of PCR primers and probes for quantitative PCR were indicated as arrows. (B) HEK293T cells (5  105) were transfected with an expression vector for wild-type or mutant TTP (0.4 mg) and then infected for 48 h with VSV-G-pseudotyped NL-Luc virus (1 ng of p24) (B) or NL-enCAT virus (1 ng of p24) (C), after which the amounts of luciferase activity (B) or CAT antigen (C) in cell lysates were determined. Data are means  S.D. of values from three independent experiments. (D) HEK293T cells were cotransfected with pNL-enCAT (0.04 mg) and an expression vector for wild-type or mutant TTP (0.4 mg), and the amount of cellular CAT antigen was determined after incubation for 48 h. Data are means  S.D. of values from three independent experiments. *P < 0.05, **P < 0.02.

M. Maeda et al. / Microbes and Infection 8 (2006) 2647e2656

2652

expression of the protein, because the expression plasmid of this mutant has the same promoter as those of other mutants.

containing the HIV-1 chimeric genome of pNL-enCAT (Fig. 3A,C) or transfected with pNL-enCAT (Fig. 3A,D), both of which contain the CAT reporter gene in place of the env open reading frame [15]. In both instances, the expression of CAT was markedly inhibited by wild-type TTP or the 1e101 mutant but not by the NH2-terminal deletion mutants 76e189 or 176e320. These results indicated that TTP promoted multiple splicing of HIV-1 transcripts, thereby increasing the products of the 2-kb spliced RNA and decreasing that of unspliced (9 kb) and single-spliced (4 kb) RNAs. Although the 1e101 mutant decreased the expression of CAT, the mutant did not significantly affect on the luciferase activity, suggesting that the mechanism of suppression of CAT activity by wild-type and the 1e101 mutant TTP seemed to be different. To confirm that TTP affects HIV-1 replication through regulation of splicing, we co-expressed TTP and the HIV-1 clone pNL43 in HEK293T cells and examined expression levels of

3.3. TTP influences the production of un-, single- and multiple-spliced transcripts To examine which steps of HIV-1 replication were influenced by TTP, we next used HIV-1 reporter viruses and distinguished viral products associated with un-, single- and multiply spliced transcripts (Fig. 3A). HEK293T cells were transfected with an expression vector for wild-type or mutant TTP and then infected with the pseudotyped NL-Luc virus. The luciferase activity of the cells was increased significantly (4-fold) by wild-type TTP and to a lesser extent (1.5-fold) by the 1e101 mutant, whereas the NH2-terminal deletion mutants 76e189 and 176e320 did not affect the luciferase activity (Fig. 3B). We then examined the expression of CAT in cells either inoculated with the NL-enCAT pseudovirus

A

B 0 0.8 0.4 0.2 (µg) 0 0.2 0.4 0.8 (µg)

pNL43 pCAGGSTTP

Pr55-

pCMV128

CMV

CAT SD

Nefp24-

1

2

CAT (ng/ml)

C 300 4

3

Nef/Pr55 (fold increase)

TTP4 3 2

0

+ Ve cto

r

+ wt TT

4

D

P

+ 1-1 01

+ 76 -1

+ 17 6-3 89 20

E 300

200

** 100

0

Mo ck

wt

1-1

01

17

76

-18

9

6-3 20

Un-, singly spliced RNA (%)

Un-, singly spliced RNA (%)

**

100

1

3

SA

200

Rev:

0 Lane 2

LTR

RRE

120

Virus

100

1

ng

0.2 ng

80

0.03 ng

60 40 20 0

Mock

TTP

Fig. 4. Effects of TTP on HIV-1 RNA splicing. (A) HEK293T cells were cotransfected with the indicated amounts of pNL43 and of a TTP expression vector (pCAGGSTTP), after which cell lysates were subjected to immunoblot analysis with antibodies to p24 and to Nef (upper panel) or with those to TTP (lower panel). The Nef/Pr55gag signal ratio for the blot was determined and expressed as fold increase relative to the value for lane 2. (B) Schematic representation of the Rev reporter plasmid pCMV128. The CMV promoter (CMV), splice donor (SD), splicing acceptor (SA), and HIV-1 long terminal repeat (LTR) are indicated. (C) HEK293T cells were cotransfected with pCMV128 (0.1 mg) and an expression vector for wild-type or mutant TTP (0.4 mg) in the absence or presence of pcMax-Rev (0.4 mg). The cells were then incubated for 48 h in fresh medium before determination of the amount of intracellular CAT antigen. Data are means  S.D. of values from three independent experiments. **P < 0.02. (D) HEK293T cells were transfected as in (Fig. 3B) and the relative ratio of the number of un-, singly spliced transcripts (9 and 4 kb) to the number of those reactive with multiply-spliced transcripts (2 kb) was determined. Data are means  S.D. of values from three independent experiments. (E) Cells transfected with pchTTPwt (0.4 mg) and then infected with VSV-G-pseudotyped NL-Luc virus (1, 0.2, or 0.03 ng of p24 antigen) were assayed for viral RNA splicing as in (A). Data are from a representative experiment. **P < 0.02.

M. Maeda et al. / Microbes and Infection 8 (2006) 2647e2656

Gag and Nef proteins. Although the intracellular levels of Gag proteins (Pr55, p24) was substantially decreased by co-expression of TTP, those of Nef were not affected (Fig. 4A). The ratio of the amount of Nef to that of Pr55gag was found to be increased by TTP in a concentration-dependent manner (Fig. 4A). Furthermore, we investigated the effects of TTP with the use of different expression system, pCMV128, a Rev reporter plasmid that encodes most of the env sequence and a Rev-responsive element (RRE) downstream of the CAT gene, with splice donor and splice acceptor sites at the 5 0 and 3 0 ends of the construct, respectively [25] (Fig. 4B). Both wild-type TTP and the 1e101 mutant markedly suppressed CAT expression in HEK293T cells transfected with pCMV128 and a Rev expression plasmid, whereas the NH2terminal deletion mutants of TTP (76e189, 176e320) had no effect (Fig. 4C). These results thus indicated that TTP and its 1e101 mutant suppressed the Rev-induced gene expression. 3.4. Effects of TTP and its mutants on the splicing of HIV-1RNA Because TTP decreased the products of the un-, single spliced transcripts of HIV-1 and increased multiple spliced transcripts, we examined the effects of TTP and its mutants on the splicing of HIV-1 RNA in cells transfected with pNLLuc-ERþ (Fig. 3A). Wild-type TTP induced significant decrease (86%) in the proportion of unspliced HIV-1 RNA (assessed from the ratio of the abundance of transcripts detected by the env probe to that of those detected by the tat, rev and nef cDNA probe) compared with the cells transfected with the corresponding mock vector (Fig. 4D). The TTP mutant 1e101 induced a smaller decrease (56%) in the proportion of unspliced HIV-1 RNA, whereas the NH2-terminal deletion mutants (76e189 and 176e320) had no effect (Fig. 4D). Similar results were obtained with HEK293T cells inoculated with the VSV-G-pseudotyped NL-Luc virus (Fig. 4E). The proportion of unspliced HIV-1 RNA in cells infected with a low titer (0.03 ng of p24) of the pseudotyped NLLuc virus was reduced by 10% by co-expression of TTP (Fig. 4E). These results thus suggested that TTP enhanced the multiple splicing of HIV-1 transcripts. To examine whether TTP destabilizes HIV-1 RNA, we compared the effects of TTP on the RNA stability of wildtype HIV-1. However, TTP did not significantly affect the rate of RNA decay of both multiply spliced and un-, singlyspliced RNA for 180 min (data not shown). 3.5. TTP binds to HIV-1 AU-rich RNA That AU residues of inhibitory sequence in of HIV-1 RNA can complement Rev activity [26] and replacement of AU residues of HIV-1 mRNA without alteration of the encoded proteins has been reported to result in marked expression of HIV-1 Gag, Pol, and Env proteins independent of the Rev/RRE/Crm1, suggesting that the AU-rich negative elements are necessary for the responsiveness of late HIV-1 transcripts to Rev [2,3]. As TTP binds to AU-rich elements [10], the P7 region,

2653

which has high ratio of AU residues (63%), was chosen in the HIV-1 sequence. Meanwhile, the RRE is known as a structural region bound to Rev and has relatively low AU sequence (48%). Therefore, as a control, we used a synthesized RNA of the RRE. To analyze the binding in an acrylamide gel we chose a stem loop region of the RRE, stem II, which is a minimal RNA element for Rev binding [27]. We synthesized both RNA of p7 and stem II of the RRE, which was found to contain small amount of longer substrates (Fig. 5B, black arrow). The binding assays revealed that TTP bound to P7 sequence (Fig. 5A) much more strongly than that of stem II of the RRE (Fig. 5B). To confirm that TTP specifically bound to the P7 sequence, we examined whether unlabeled P7 inhibited the binding of TTP. Unlabeled P7 RNA inhibited the binding of TTP to labeled P7 in a concentration-dependent manner, but stem II of the RRE did not (Fig. 5C). 3.6. Effects of TTP on the subcellular localization of Rev To investigate the mechanism of suppression of HIV-1 replication by wild-type and 1e101 mutant TTP, Rev fused with

A

B TTP:

-

TTP:

TTP-RNA complex TTP-RNA complex

Cold probe : TTP:

TTP-RNA complex

-

RRE -

P7-

C

-

P7

RRE

- - ++++++++++ +

-

P7-

Fig. 5. Binding of TTP to the p7 RNA. Native electrophoretic mobility-shift assay of the binding of TTP (0.05, 0.1, 0.25, 0.5, or 1 mg/ml) to the 32P-labeled P7 (A) or stem II of RRE (B) RNA (10 ng) in a final volume of 10 ml. Control reactions were performed in the absence of TTP(-). The positions of the free RNA molecule of P7 (75 nt), of RRE (93 nt) and of the TTPeRNA complex are indicated. (C) The binding of TTP (0.25 ng) to the 32P-labeled P7 RNA (3 ng), which was mixed with different amounts of unlabeled P7 and unlabeled stem II of RRE RNA (0, 1.25, 2.5, 5, 10, and 20 ng in each lane).

M. Maeda et al. / Microbes and Infection 8 (2006) 2647e2656

2654

GFP-Rev

Merge

TTP

Wt TTP

A

B

D

E

F

G

H

I

J

K

L

M

N

C

+LMB

TTP (1-101)

TTP (76-189)

TTP (176-320)

Fig. 6. Effects of TTP on the subcellular localization of Rev. HeLa cells (2.5  104 per well in a 12-well dish) were cotransfected with an expression plasmid for GFP-Rev (0.1 mg) and 0.3 mg of pchTTPwt (A, B, C), pcD1e101 (F, G, H), pcD76e189 (I, J, K) or pcD176e320 (L, M, N). Similarly, HeLa cells were transfected with 0.4 mg of GFP-Rev (D) or pchTTPwt (E). The cells were incubated for 90 min with leptomycin B (5 ng/ml) before analysis (D, E). The subcellular localization of GFP-Rev was monitored by GFP fluorescence (A, D, F, I, L). Immunofluorescent signals of wild-type TTP (B, E) or the mutant 1e101 (G), 76e189 (J), and 176e320 TTP (M) were examined by anti-Xpress antibody. Merged images of Rev and wild-type and mutant TTP are shown in C, H, K, and N.

marker protein, green fluorescent protein (GFP-Rev) and wildtype or the mutant TTP were co-expressed in HeLa cells (Fig. 6AeC). Both GFP-Rev and wild-type TTP were found to be localized predominantly in the cytoplasm and to a lesser extent in the nucleus and nucleolus. Singly expressed GFPRev or wild-type TTP showed the similar localization (data not shown). Rev and TTP are known to shuttle between the

nucleolus and the cytoplasm [28]. Treatment of the cells with leptomycin B, which inhibits the CRM1-mediated nuclear export [28], resulted in the accumulation of GFP-Rev or wild-type TTP in the nucleus and nucleolus (Fig. 6D,E). GFP-Rev was restricted to the nucleus and nucleolus in the cells expressing the TTP mutant 1e101, which itself was largely restricted to the cytoplasm (Fig. 6FeH). Both the TTP

M. Maeda et al. / Microbes and Infection 8 (2006) 2647e2656

mutant 76e189 and 176e320 did not affect on the localization of GFP-Rev (Fig. 6IeN). As has been reported previously [28], the 76e189 mutant localized in the nucleus (Fig. 6J).

4. Discussion In the present study we have demonstrated that TTP inhibits HIV-1 production by promoting multiple splicing by direct binding to AU-rich sequences. Because TTP increased luciferase activity of NL-luc virus, which is derived from multiple spliced transcripts, inhibition of HIV-1 production was not due to nonspecific damage of the cells induced by overexpression of TTP, but due to a specific suppression by TTP binding to HIV-1 RNA. Although the TTP mutant 1e101 lacking NLS could suppress the production derived from un-, single-spliced RNA, it did not increase the product of multiple-spliced RNA. Among the wild-type TTP and its mutants, the 1e101 mutant exclusively suppressed nuclear export of Rev similarly to leptomycin B. It is suggested that the NES sequence of the 1e101 mutant plays an important role in suppression of nuclear export of Rev. Because the 1e101 mutant is not capable of binding with RNA, marked suppression of CAT expression by the 1e101 mutant (Fig. 4C) seems to be mainly dependent on competition with Rev for binding to CRM1 which shuttles between cytoplasm and nucleus. In contrast, wild-type TTP, which directly binds to AU-rich sequence of HIV-1 RNAs, did not affect subcellular localization of Rev and increased the production of multiple-spliced RNA. Although the 76e 189 mutant, which predominantly localizes in the nucleus (Fig. 6), has RNA binding activity, this mutant could not enhance the splicing HIV-1 RNA, suggesting that the nuclear localization might be related to abrogation of splicing HIV-1 RNA. Turnover of mRNA for certain cytokines containing an ARE (AUUUA) in their 3 0 -untranslated region is regulated by cis elements and trans-acting factors, TTP [10]. Although sequences of HIV-1 do not contain ARE, as we found that ARE of GM-CSF mRNA was cleaved between U and A residues we expected that TTP might destabilize HIV-1 RNA, which contains AU-rich sequences [2]. However, TTP did not significantly affect HIV-1 RNA stability (data not shown). This result may be explained by the following reasons. Firstly, HIV-1 AU-rich sequences were not so effectively cleaved by TTP as ARE in GM-CSF mRNA, so we could not obtain the significant decrease of HIV-1 RNA using our assay system even in the presence of TTP. Secondly, actinomycin D in the assay system could not completely suppress the transcription of HIV-1 RNA. Other experiments would be required to demonstrate destabilization of HIV-1 RNA by TTP. Given that TTP is exclusively expressed in activated CD4þ T cells, which are targets of HIV-1, TTP might inhibit virion production in these cells and thereby inhibit the presentation of viral antigens by MHC class I molecules on the cell surface, thus contributing to persistence of an HIV-1 reservoir in T cells. The resulting inactivated cells harboring a silent HIV-1

2655

genome might be subsequently reactivated by various stimuli and produce HIV-1 virions. ZAP, a family of RNA binding proteins carrying CCCHtype zinc fingers, also inhibit retroviral RNA production [29]. Zinc fingers of TTP and ZAP may play an important role in nuclear retention or decay of genomic RNA with a different mechanism as RNA interference. Further studies are required to reveal the precise mechanism by which TTP affects HIV-1 RNA trafficking mediated by Rev [30].

Acknowledgments We thank Dr. B.R. Cullen (Duke University) for providing pCMV128, pDM128/CTE and Dr. K. Ikuta (Osaka University) for the gifts of U1 cells and antibodies to Nef and p24. We also thank Dr. H. Shida (Hokkaido University) for providing pSRaRev. This work was supported in part by grants from the Ministry of Education, Science, Technology, Sports, and Culture of Japan, the Ministry of Health, Labor, and Welfare of Japan, the Japan Human Science Foundation, and the Japanese Foundation for AIDS Prevention.

References [1] A.W. Cochrane, K.S. Jones, S. Beidas, P.J. Dillon, A.M. Skalka, C.A. Rosen, Identification and characterization of intragenic sequences which repress human immunodeficiency virus structural gene expression, J. Virol. 65 (1991) 5305e5313. [2] W. Tan, S. Schwartz, The Rev protein of human immunodeficiency virus type 1 counteracts the effect of an AU-rich negative element in the human papillomavirus type 1 late 3 0 untranslated region, J.Virol. 69 (1995) 2932e2945. [3] M. Graf, A. Bojak, L. Deml, K. Bieler, H. Wolf, R. Wagner, Concerted action of multiple cis-acting sequences is required for Rev dependence of late human immunodeficiency virus type 1 gene expression, J.Virol. 74 (2000) 10822e10826. [4] A.S. Zolotukhin, A. Valentin, G.N. Pavlakis, B.K. Felber, Continuous propagation of RRE(-) and Rev(-)RRE(-) human immunodeficiency virus type 1 molecular clones containing a cis-acting element of simian retrovirus type 1 in human peripheral blood lymphocytes, J. Virol. 68 (1994) 7944e7952. [5] H. Takahashi, H. Sawa, H. Hasegawa, T. Sata, W.W. Hall, K. Nagashima, T. Kurata, Reconstitution of cleavage of human immunodeficiency virus type-1 (HIV-1) RNAs, Biochem. Biophys. Res. Commun. 293 (2002) 1084e1091. [6] J.S. McDougal, A. Mawle, S.P. Cort, J.K. Nicholson, G.D. Cross, J.A. Scheppler-Campbell, D. Hicks, J. Sligh, Cellular tropism of the human retrovirus HTLV-III/LAV.I. Role of T cell activation and expression of the T4 antigen, J. Immunol. 135 (1985) 3151e3162. [7] J.A. Zack, A.M. Haislip, P. Krogstad, I.S. Chen, Incompletely reversetranscribed human immunodeficiency virus type 1 genomes in quiescent cells can function as intermediates in the retroviral life cycle, J. Virol. 66 (1992) 1717e1725. [8] A. Clark, Post-transcriptional regulation of pro-inflammatory gene expression, Arthritis Res. 2 (2000) 172e174. [9] A. Bevilacqua, M.C. Ceriani, S. Capaccioli, A. Nicolin, Post-transcriptional regulation of gene expression by degradation of messenger RNAs, J. Cell Physiol. 195 (2003) 356e372. [10] N. Xu, C.Y. Chen, A.B. Shyu, Modulation of the fate of cytoplasmic mRNA by AU-rich elements: key sequence features controlling mRNA deadenylation and decay, Mol. Cell. Biol. 17 (1997) 4611e4621.

2656

M. Maeda et al. / Microbes and Infection 8 (2006) 2647e2656

[11] T.M. Folks, J. Justement, A. Kinter, C.A. Dinarello, A.S. Fauci, Cytokine-induced expression of HIV-1 in a chronically infected promonocyte cell line, Science 238 (1987) 800e802. [12] M. Tobiume, M. Takahoko, T. Yamada, M. Tatsumi, A. Iwamoto, M. Matsuda, Inefficient enhancement of viral infectivity and CD4 downregulation by human immunodeficiency virus type 1 Nef from Japanese long-term nonprogressors, J. Virol. 76 (2002) 5959e5965. [13] Y. Hakata, M. Yamada, H. Shida, A multifunctional domain in human CRM1 (exportin 1) mediates RanBP3 binding and multimerization of human T-cell leukemia virus type 1 Rex protein, Mol. Cell. Biol. 23 (2003) 8751e8761. [14] A. Adachi, H.E. Gendelman, S. Koenig, T. Folks, R. Willey, A. Rabson, M.A. Martin, Production of acquired immunodeficiency syndrome-associated retrovirus in human and nonhuman cells transfected with an infectious molecular clone, J. Virol. 59 (1986) 284e291. [15] K. Tokunaga, A. Kojima, T. Kurata, K. Ikuta, H. Akari, A.H. Koyama, M. Kawamura, R. Inubushi, R. Shimano, A. Adachi, Enhancement of human immunodeficiency virus type 1 infectivity by Nef is producer celldependent, J. Gen. Virol. 79 (Pt 10) (1998) 2447e2453. [16] R. Shibata, H. Sakai, M. Kawamura, K. Tokunaga, A. Adachi, Early replication block of human immunodeficiency virus type 1 in monkey cells, J. Gen. Virol. 76 (Pt 11) (1995) 2723e2730. [17] T.M. Ross, B.R. Cullen, The ability of HIV type 1 to use CCR-3 as a coreceptor is controlled by envelope V1/V2 sequences acting in conjunction with a CCR-5 tropic V3 loop, Proc. Natl. Acad. Sci. USA 95 (1998) 7682e7686. [18] R.A. Fouchier, B.E. Meyer, J.H. Simon, U. Fischer, M.H. Malim, HIV-1 infection of non-dividing cells: evidence that the amino-terminal basic region of the viral matrix protein is important for Gag processing but not for post-entry nuclear import, EMBO J 16 (1997) 4531e4539. [19] G.A. Taylor, W.S. Lai, R.J. Oakey, M.F. Seldin, T.B. Shows, R.L. Eddy Jr., P.J. Blackshear, The human TTP protein: sequence, alignment with related proteins, and chromosomal localization of the mouse and human genes, Nucleic Acids Res. 19 (1991) 3454.

[20] Z. Tsuchihashi, P.O. Brown, DNA strand exchange and selective DNA annealing promoted by the human immunodeficiency virus type 1 nucleocapsid protein, J. Virol. 68 (1994) 5863e5870. [21] D.P. Bednarik, T.M. Folks, Mechanisms of HIV-1 latency, AIDS 6 (1992) 3e16. [22] M. Tobiume, K. Fujinaga, M. Kameoka, T. Kimura, T. Nakaya, T. Yamada, K. Ikuta, Dependence on host cell cycle for activation of human immunodeficiency virus type 1 gene expression from latency, J. Gen. Virol. 79 (Pt 6) (1998) 1363e1371. [23] A.M. Fairhurst, J.E. Connolly, K.A. Hintz, N.J. Goulding, A.J. Rassias, M.P. Yeager, W. Rigby, P.K. Wallace, Regulation and localization of endogenous human tristetraprolin, Arthritis Res. Ther. 5 (2003) R214e R225. [24] K. Tokunaga, M.L. Greenberg, M.A. Morse, R.I. Cumming, H.K. Lyerly, B.R. Cullen, Molecular basis for cell tropism of CXCR4-dependent human immunodeficiency virus type 1 isolates, J. Virol. 75 (2001) 6776e 6785. [25] Y. Luo, H. Yu, B.M. Peterlin, Cellular protein modulates effects of human immunodeficiency virus type 1 Rev, J. Virol. 68 (1994) 3850e3856. [26] I. Mikaelian, M. Krieg, M.J. Gait, J. Karn, Interactions of INS (CRS) elements and the splicing machinery regulate the production of Revresponsive mRNAs, J. Mol. Biol. 257 (1996) 246e264. [27] K.S. Cook, G.J. Fisk, J. Hauber, N. Usman, T.J. Daly, J.R. Rusche, Characterization of HIV-1 REV protein: binding stoichiometry and minimal RNA substrate, Nucleic Acids Res. 19 (1991) 1577e1583. [28] T. Murata, Y. Yoshino, N. Morita, N. Kaneda, Identification of nuclear import and export signals within the structure of the zinc finger protein TIS11, Biochem. Biophys. Res. Commun. 293 (2002) 1242e1247. [29] G. Gao, X. Guo, S.P. Goff, Inhibition of retroviral RNA production by ZAP, a CCCH-type zinc finger protein, Science 297 (2002) 1703e1706. [30] N. Sanchez-Velar, E.B. Udofia, Z. Yu, M.L. Zapp, hRIP, a cellular cofactor for Rev function, promotes release of HIV RNAs from the perinuclear region, Genes Dev. 18 (2004) 23e34.