Translation Is Not Required To Generate Virion ... - Journal of Virology

JOURNAL OF VIROLOGY, Dec. 2000, p. 11531–11537 0022-538X/00/$04.00⫹0 Copyright © 2000, American Society for Microbiology. All Rights Reserved.

Vol. 74, No. 24

Translation Is Not Required To Generate Virion Precursor RNA in Human Immunodeficiency Virus Type 1-Infected T Cells MELINDA BUTSCH1,2

AND

KATHLEEN BORIS-LAWRIE1,2,3,4,5*

1

Center for Retrovirus Research, Departments of Veterinary Biosciences4 and Molecular Virology, Immunology and Medical Genetics,3 The Ohio State Biochemistry Program,2 and Comprehensive Cancer Center,5 The Ohio State University, Columbus, Ohio 43210-1093 Received 18 July 2000/Accepted 25 September 2000

The retroviral primary transcription product is a multifunctional RNA that is utilized as pre-mRNA, mRNA, and genomic RNA. The relationship between human immunodeficiency virus type 1 (HIV-1) unspliced transcripts used as mRNA for viral protein synthesis and as virion precursor RNA (vpRNA) for encapsidation remains an important question. We developed a biochemical assay to evaluate the hypothesis that prior utilization as mRNA template for protein synthesis is necessary to generate vpRNA. HIV-1-infected T cells were treated with translation inhibitors under conditions that maintain virus production. Immunoprecipitation of newly synthesized HIV-1 Gag protein revealed that de novo translation is not necessary to sustain assembly, release, or processing of Gag structural protein. Both newly synthesized protein and steady-state Gag are competent for assembly, and the extracellular accumulation of Gag is proportional to the intracellular abundance of Gag. As early as 2 h after transcription, newly synthesized RNA is detectable in cell-free virions and encapsidation is sustained upon inhibition of host cell translation. Results of both [3H]uridine incorporation assays and HIV-1-specific RNase protection assays (RPAs) indicate that translation inhibition reduces the absolute amounts of both cytoplasmic and virion-associated RNA. Evaluation of encapsidation efficiency by RPA revealed that the cytoplasmic availability of vpRNA is increased, indicating that HIV-1 unspliced mRNA can be rerouted to function as vpRNA. Our data contrast with results from the HIV-2 and murine leukemia virus systems and indicate that HIV-1 unspliced RNA constitutes a single functional pool that can function interchangeably as mRNA and as vpRNA. The genomes of RNA viruses are multifunctional molecules. In retroviruses, including human immunodeficiency virus type 1 (HIV-1), the primary RNA transcript functions as premRNA for splicing, mRNA for synthesis of viral protein, and virion precursor RNA (vpRNA) for packaging into infectious virions. The unspliced HIV-1 mRNA and vpRNA are physically indistinguishable and are defined experimentally by their association with ribosomes and virions, respectively. The relationship between mRNA and vpRNA remains poorly understood, and its characterization may yield a new strategy to inhibit production of infectious HIV-1 and to improve lentiviral vector systems for gene transfer applications. Initial investigation of the relationship between retroviral unspliced mRNA and vpRNA focused on cells productively infected with the genetically simple murine leukemia virus (MLV) (11, 15, 20). Levin and colleagues (10, 11) analyzed cells treated with the transcription inhibitor actinomycin D (actD) and showed that viral mRNA remains available to direct viral protein synthesis, but the particles do not contain genomic RNA. These data implied that MLV transcripts segregate into two functionally distinct populations of mRNA for translation or vpRNA for encapsidation (11). Stoltzfus et al. (23) applied isotopic equilibrium assay to cells infected with avian sarcoma virus (ASV) and observed not two but rather a single RNA population that functions as both ASV mRNA and vpRNA. Sonstegard and Hackett (22) came to similar conclusions in their studies of Rous sarcoma virus (RSV) vector RNAs. Transfection studies with vectors that contain or lack * Corresponding author. Mailing address: Center for Retrovirus Research, Department of Veterinary Biosciences, The Ohio State University, Columbus, OH 43210-1093. Phone: (614) 292-1392. Fax: (614) 292-6473. E-mail: [email protected].

most of the RSV encapsidation signal, ␺, indicate that interaction of Gag with ␺ autogenously modulates competition between the translational machinery and assembling viral proteins. The data indicate that equilibrium exists between vector RNA destined for translation or encapsidation, which is determined by the cytoplasmic availability of Gag protein and ribosomes (22). Investigation of the fate of vpRNA from genetically complex retroviruses has been largely limited to genetic studies with HIV vectors and has not been pursued for RNA expressed from HIV-1 provirus in human T cells. Studies with HIV-1based vectors have shown that the RNA structure inherent in the HIV-1 encapsidation signal inhibits efficient translation (6, 17). These results imply that HIV-1 encapsidation and translation are competing processes. McBride et al. (13) evaluated a subgenomic HIV-1 vector that contains a premature gag stop codon and found that encapsidation remained efficient. These data are consistent with the successful use of HIV-1 as a gene transfer vector (9, 18) and eliminate a requirement for ongoing Gag protein synthesis. However, the question of whether or not it is necessary for vpRNA to serve as mRNA template prior to encapsidation remains open. Contrasting results were obtained in a study of HIV-2-based vectors that contain deletions at the 3⬘ end of the gag open reading frame. These results indicated that Gag protein translation from vector template was necessary to generate HIV-2 vpRNA (8). HIV-1 differs from HIV-2 in that the complete encapsidation signal exists only on unspliced viral RNA and not on spliced RNAs as in HIV-2 (14). The requirement for prior translation of HIV-2 gag mRNA is a potential mechanism for selective encapsidation of HIV-2 unspliced RNA into progeny virions (8). The primary goal of this project was to evaluate the hypothesis that translation is a prerequisite to generate HIV-1

11531

11532

BUTSCH AND BORIS-LAWRIE

vpRNA in chronically infected human T cells. Definition of the relationship between HIV-1 unspliced mRNA and vpRNA will shed light on whether the transcripts constitute a single RNA pool or two functionally independent pools of RNA that are dedicated as either mRNA or vpRNA. If HIV-1 unspliced transcripts function as a single pool of RNA, inhibition of protein synthesis is expected to increase the availability of vpRNA and augment encapsidation efficiency. However, in the case that prior utilization as mRNA is necessary to generate vpRNA, encapsidation efficiency would be decreased upon inhibition of protein synthesis. If HIV-1 unspliced transcripts represent two independent pools of RNA that are committed to either translation or encapsidation, translation inhibition would not alter generation of vpRNA. To examine these possibilities, HIV-1-infected human T cells were treated with translation inhibitors under conditions that maintain virus assembly. Comparison of ribosomal profiles of HIV-1-infected and mock-infected T cells verifies that HIV-1 infection does not mediate shutoff of host cell translation or disrupt the mechanistic effects of pactamycin (pac), cycloheximide (chx), or anisomycin (aniso). Analysis of newly synthesized HIV-1 protein and vpRNA after short-term treatment with the inhibitors established that de novo translation is not necessary to maintain assembly, release, and processing of Gag precursor protein or encapsidation of vpRNA. RNase protection assays (RPAs) demonstrate that HIV-1 encapsidation efficiency is increased upon 80 to 90% inhibition of de novo translation. The data indicate that prior translation of HIV-1 unspliced RNA is not a prerequisite to generate vpRNA. HIV-1 unspliced transcripts constitute a single population of RNA that can be selected interchangeably as vpRNA and as mRNA. MATERIALS AND METHODS Cells and translation inhibitors. CEM(A) T cells infected with HIV-1NL4-3 [CEM(A)/HIV-1 cells] were cultured in RPMI 1640 medium supplemented with 10% fetal calf serum and 1% antibiotic-antimycotic (Gibco-BRL). Cell viability in response to chx, pac, or aniso was assayed by propidium iodide and flow cytometry at 4 h after treatment (3). In subsequent experiments we used maximal concentrations having minimal cytopathic effect during a 4-h incubation: 5 ⫻ 10⫺8 M pac (Pharmacia & Upjohn, Kalamazoo, Mich.), 0.5 ␮g of chx (Sigma, St. Louis, Mo.) per ml, and 0.1 ␮g of aniso (Sigma) per ml. Protein analysis. CEM(A)/HIV-1 T cells were lysed in radioimmunoprecipitation assay (RIPA) buffer (0.05 M Tris-HCl [pH 8], 0.1% sodium dodecyl sulfate [SDS], 1% Triton X-100, 2 mM phenylmethylsulfonyl fluoride, 0.15 M NaCl) containing 1% deoxycholic acid, and the nuclei were removed after centrifugation at 13,400 ⫻ g for 10 min. Total protein concentration was determined by the Bio-Rad DC protein assay (Bio-Rad Laboratories, Hercules, Calif.). Virioncontaining medium was clarified by centrifugation at 2,000 ⫻ g for 10 min, and virions were collected by centrifugation at 156,000 ⫻ g for 1.5 h at 4°C in a Beckman SW41 rotor. Gag enzyme-linked immunosorbent assay (ELISA) was performed as specified by the manufacturer (Beckman-Coulter, Brea, Calif.). 35 S-labeling experiments were performed by incubating CEM(A)/HIV-1 T cells in cysteine/methionine-free RPMI medium with 5% dialyzed fetal bovine serum for 30 min, followed by addition of pac, chx, or aniso coincident with [35S]cysteine/methionine (10 ␮Ci/ml; 1,175 Ci/mmol, 43.5 MBq/ml; ICN Biochemicals, Irvine, Calif.). The cells were lysed in RIPA buffer containing 1% deoxycholic acid. Fifty nanograms of cytoplasmic lysate and virion lysates equivalent to 30 ng of Gag were precipitated by trichloroacetic acid (TCA) (20% TCA, 0.1 mg of bovine serum albumin per ml) onto 25-mm-diameter glass fiber filters (type A/C; Pall Corp., Ann Arbor, Mich.), washed three times in 10% TCA and in 100% ethanol, and subjected to scintillation counting. For pulse-chase experiments, CEM(A)/HIV-1 T cells were incubated first for 30 min in cysteine/methioninefree RPMI medium with 5% dialyzed fetal bovine serum and then for 1 h in [35S]cysteine/methionine-supplemented medium with or without pac, chx, or aniso. The cells were washed, and complete RPMI medium was added. At 1, 2, 4, 6, or 8 h postchase, cells were harvested as above. For immunoprecipitation, the 35S-labeled lysates were incubated for 16 h with protein A-Sepharose beads (Pharmacia) and Gag p24 antibody (gift of N. Panganiban) (12). The beads were washed once in high-salt RIPA buffer (1 M NaCl) and once in low-salt RIPA buffer (0.15 M NaCl) and then boiled to elute the proteins. The 35S-labeled precipitated proteins were subjected to SDS-polyacrylamide gel electrophoresis (PAGE), visualized, and quantified by PhosphorImager analysis (Molecular Dy-

J. VIROL. namics, Inc., Sunnyvale, Calif.) with ImageQuaNT software version 4.2 (Molecular Dynamics). RNA analysis. CEM(A)/HIV-1 cells were plated in T150 flasks, cultured overnight to 80% confluence, and incubated for 4 h in medium with [3H]uridine (30 ␮Ci/ml; Amersham, Piscataway, N.J.) and with or without pac, chx, or aniso. Cytoplasmic extracts were prepared in 0.9 ml of cold cell lysis buffer (10 mM Tris [pH 8.3], 150 mM NaCl, 1.5 mM MgCl2) and 0.1 ml of 5% NP-40. Following centrifugation to pellet the nuclei, the supernatant was mixed with TriReagent LS, and cytoplasmic RNA was isolated as specified by the manufacturer (Molecular Research, Cincinnati, Ohio). One microgram of cytoplasmic [3H]RNA and virions equivalent to 30 ng of extracellular Gag were applied to glass fiber filters, which were washed four times with 5% TCA containing 20 mM sodium pyrophosphate and once with 100% ethanol, dried, and subjected to scintillation counting. For ribosome profiles, clarified cytoplasmic extract from three 80% confluent T150 flasks was layered onto a 10-ml linear gradient of 15 to 45% sucrose (21). The gradient was centrifuged 225,000 ⫻ g for 2.25 h at 4°C in a Beckman SW41 rotor. Gradients were fractionated and monitored for A254 on an ISCO (Lincoln, Neb.) model 160 gradient fractionator. To prepare virion RNA, cell medium was clarified by centrifugation at 2,000 ⫻ g for 10 min, and virions were pelleted by centrifugation at 156,000 ⫻ g for 2.5 h at 4°C in a Beckman SW28 rotor, lysed in 1 ml Trizol Reagent, and isolated as specified by the manufacturer (Gibco BRL, Gaithersburg, Md.). 32 P-labeled antisense RNA probes were generated by in vitro transcription of pGEM(600–900), which contains the 5⬘ untranslated region of HIV-1NL4-3 (12), and pGAPDH, which contains the human glyceraldehyde dehydrogenase (gapdh) gene (2). Following digestion of pGEM(600–900) with NotI and pGAPDH with NcoI, antisense runoff RNA transcripts were synthesized with MAXIscript T7 RNA polymerase (Ambion, Austin, Tex.), and the probes were isolated by gel elution. RPA was performed using RPA III (Ambion) according to the instruction manual, with some modifications (2). Typically, 10 ␮g of cytoplasmic RNA or viral RNA from virions equivalent to 250 ng of Gag was precipitated by ethanol with 2 ⫻ 105 cpm of HIV-1 probe and 2 ⫻ 104 cpm of gapdh probe. Samples were resuspended in 10 ␮l of hybridization buffer, denatured at 94°C for 3 min, and hybridized at 42°C overnight. RNase A/T1 was diluted 1:100 in Ambion RNase digestion buffer, and 150 ␮l was added to each sample and incubated at 37°C for 30 min. SDS and proteinase K were added to final concentrations of 1% and 0.5 mg/ml, respectively, and samples were incubated at 37°C for 30 min. A 100-bp 32P-labeled DNA fragment was added to virion RNA samples, followed by phenol-chloroform and chloroform extraction and precipitation with ethanol in the presence of 10 ␮g of yeast RNA. Following centrifugation, the pellets were dissolved in 6 ␮l of loading buffer, denatured at 94°C for 3 min, and subjected to 5% denaturing PAGE RNase protection products were visualized and quantified by PhosphorImager analysis.

RESULTS HIV-1 infection does not alter host cell response to pac, chx, or aniso. Our goal was to evaluate trafficking of HIV-1 unspliced RNA expressed from provirus in T cells. A genetic approach using conventional transfection methods was of limited utility for this purpose because overexpression of RNA from transfected DNA may saturate the assembly process and obscure the natural relationship between HIV-1 unspliced mRNA and vpRNA that is exhibited by authentic provirus in an infected T cell. Therefore, we developed a biochemical approach that limits de novo translation of HIV-1 RNA under conditions that maintain virus production. CEM(A)/HIV-1 cells were subjected to short-term incubation with three mechanistically distinct biochemical antagonists of translation: pac, chx, and aniso. To determine the magnitude and onset of inhibition of protein synthesis, [35S]cysteine/ methionine incorporation into whole cell protein was evaluated after a 4-h incubation with pac, chx, or aniso. Comparison with mock-treated cells indicated that incorporation of [35S] cysteine/methionine into whole cell protein was inhibited 80 to 90%; the reduction in protein synthesis commenced by 30 min posttreatment and continued up to 4 h posttreatment (Fig. 1). Propidium iodide staining and flow cytometry detected no overt cytopathic effects on the cells during the 4-h incubation period. Relative to the mock control, cell viability remained 100% in response to pac and 96% in response to chx but was reduced to 86% in response to aniso. Ribosomal profile analysis of the cells indicated that the treatments exert the expected mechanistic effects on the translational machinery (Fig. 2). Pac

VOL. 74, 2000

CYTOPLASMIC TRAFFICKING OF HIV-1 UNSPLICED RNA

FIG. 1. Incorporation of [35S]cysteine/methionine is inhibited by incubation with pac, chx, and aniso. Inhibition of [35S]cysteine/methionine incorporation occurs 0.5 h posttreatment and is sustained over a 4-h period. CEM(A)/HIV-1 T cells were incubated with cysteine/methionine-free RPMI medium for 0.5 h, followed by the addition of [35S]cysteine/methionine with pac, chx, or aniso. Total cell lysates were collected at 0.5, 2, or 4 h posttreatment, and [35S]cysteine/methionine incorporation was quantified by TCA precipitation assay. Average results of at least four experiments are shown; error bars indicate standard deviations.

produced an accumulation of 80S monosomes, which is attributable to interference with translation initiation (5, 7, 24). Chx resulted in the accumulation of polyribosomes in response to a block in EF-2-dependent peptide translocation (4, 19, 24).

11533

Aniso reduced polyribosome abundance associated with defective peptide bond formation during elongation of the polypeptide (24). Comparison with the rRNA profiles of mockinfected CEM(A) T cells (Fig. 2) indicated that HIV-1 infection did not change the rRNA profile of CEM(A) T cells (1), nor did it alter the response to pac, chx, or aniso. These results indicate that HIV-1 infection does not disrupt the host translation machinery. De novo translation is not necessary for virion production or Gag processing. Pulse-chase labeling was used to define the onset and duration of accumulation of newly synthesized Gag in virions. Cells were incubated for 0.5 h in medium lacking cysteine/methionine, followed by incubation for 1 h in [35S] cysteine/methionine-supplemented medium, washing, and incubation with medium without [35S]cysteine/methionine; 1 to 8 h later, virions were isolated by ultracentrifugation and quantified by Gag ELISA. One microgram of whole cell lysate or virions equivalent to 30 ng of Gag were subjected to precipitation assay with TCA to determine [35S]cysteine/methionine incorporation into newly synthesized proteins. Nonspecific accumulation of [35S]cysteine/methionine was quantified in control cultures that were treated with pac to inhibit protein synthesis. Level of background incorporation of [35S]cysteine/ methionine into the pac-treated whole cell lysates and virion samples were similar at each time point: 1,000 cpm or less for whole cell lysate and 30 cpm or less for virion samples (Fig. 3). Incorporation of 35S-labeled protein into virions was maximal at 1 h postchase, indicating that newly synthesized Gag is readily incorporated into virions (Fig. 3). Production of 35Slabeled virions continued for 6 h postchase, indicating that

FIG. 2. Ribosomal profile analysis in response to pac, chx, and aniso. HIV-1 infected or uninfected CEM(A) T cells were treated for 4 h with or without pac, chx, or aniso, and cytoplasmic extracts were placed on 10-ml linear gradients of 15 to 45% sucrose. After ultracentrifugation, the gradients were fractionated and monitored at A254 using an ISCO gradient fractionation system.

11534


J. VIROL.

FIG. 3. Pulse-chase analysis of Gag incorporation into virions. CEM(A)/HIV-1 T cells were incubated with cysteine/methionine-free medium for 0.5 h, followed by a 1-h incubation with of 10 ␮Ci of [35S]cysteine/methionine per ml and 5 ⫻ 10⫺8 M pac in control plates. Cells were washed and incubated in complete medium with or without pac. Cell lysates and virion lysates were collected at intervals between 1 and 8 h postchase. TCA precipitation assay was performed with 50 ng of cellular protein or virion lysate equivalent to 30 ng of Gag. Representative results of three experiments are shown.

newly synthesized Gag is not required for continued virion production. [35S]cysteine/methionine incorporation into virions diminished over time, a trend that matched the decline in intracellular 35S-labeled protein. The results indicate a concentration-dependent relationship between intracellular and extracellular Gag. Gag immunoprecipitation assay was used to evaluate the effects of pac, chx, and aniso on Gag protein synthesis and processing. We also evaluated the effect of actD, an RNA synthesis inhibitor that is known to disrupt shuttling of HIV-1 Rev between the nucleus and the cytoplasm (16). The cells were incubated for 4 h in medium supplemented with [35S]cysteine/methionine. Subsequent Gag ELISA of cell-free supernatant indicated that Gag production was reduced but not abrogated in response to the translation inhibitors. Compared to the mock-treated cells, levels of Gag production from pac-, chx-, aniso-, and actD-treated cells were 68% ⫾ 11%, 69% ⫾ 16%, 71% ⫾ 19%, and 93% ⫾ 15%, respectively. These results indicate that the inhibitor treatments do not prevent previously synthesized Gag from being released from the cell. Immunoprecipitation assays detected similar levels of 35Slabeled Gag in mock-treated and actD-treated cells (100 and 120%, respectively) (Fig. 4). Similar levels of 35S incorporation were also observed in virion samples, indicating that de novo RNA synthesis is not necessary for synthesis and processing of HIV-1 Gag. Treatment with pac, chx, and aniso reduced intracellular 35S-labeled Gag levels to 28, 40, and 28%, respectively. Extracellular 35S-labeled Gag levels were similarly reduced to 20, 20, and 26%, respectively, indicating that extracellular 35Slabeled Gag levels are proportional to the cytoplasmic abundance of 35S-labeled Gag. Each virion sample displayed fully processed Gag p24, indicating that de novo translation is not required for Gag protein processing. Minor differences observed in the intracellular ratio of unprocessed Gag p55 to Gag p24 may be attributable to variation among the cells in the intracellular concentration of Gag p55. The observation that the extracellular accumulation of Gag is proportional to the intracellular abundance of Gag validates a concentration-de-

pendent relationship between extracellular and intracellular Gag. Newly synthesized RNA accumulates in virions within 2 h. [3H]uridine labeling was performed to evaluate the time course in which newly synthesized vpRNA becomes available for encapsidation. CEM(A)/HIV-1 cells were incubated with [3H]uridine over a 6-h period, and cytoplasmic and virion-associated RNAs were isolated and subjected to the TCA precipitation assay. One microgram of cytoplasmic RNA and virion RNA equivalent to 30 ng of Gag were analyzed. By 2 h postlabeling, 3 H-labeled RNA was present in the cytoplasm of the mocktreated cells and was incorporated into virions (Fig. 5). These results indicate that as early as 2 h postlabeling, changes in cytoplasmic RNA are manifested in virions. As a negative control for nonspecific incorporation of [3H]uridine, RNA synthesis was inhibited by treatment with actD. The actD-treated samples exhibited low-level [3H]uridine incorporation at each

FIG. 4. Pac, chx, and aniso significantly inhibit Gag synthesis. CEM(A)/HIV1 T cells were incubated with cysteine/methionine-free RPMI medium for 0.5 h, followed by the addition of 10 ␮Ci of [35S]cysteine/methionine per ml with or without pac, chx, or aniso. Cell lysates and cell-free supernatants were collected 4 h posttreatment, and virions were isolated by centrifugation. Fifty-nanogram samples of 35S-labeled cell lysate and virions equivalent to 100 ng Gag were subjected to radioimmunoprecipitation assay with Gag p24 antibody followed by SDS-PAGE and PhosphorImager analysis.

VOL. 74, 2000

CYTOPLASMIC TRAFFICKING OF HIV-1 UNSPLICED RNA

11535

FIG. 5. Incorporation of [3H]uridine into newly synthesized virions. CEM(A)/HIV-1 cells were incubated for 2, 4, and 6 h in medium containing [3H]uridine (30 ␮Ci/ml) with or without actD (0.5 ␮g/ml). [3H]uridine levels in cytoplasmic RNA and in virion RNA were quantified by TCA precipitation analysis. Representative results of at least three experiments are shown.

time point: 1,000 cpm in cytoplasmic RNA and 100 cpm in virion RNA (Fig. 5). To evaluate the possibility that de novo translation is a prerequisite for generation of vpRNA, [3H]uridine incorporation into virion RNA was evaluated with or without treatment with pac, chx, or aniso. Again treatment with actD was used to determine the value of background incorporation of [3H]uridine into virion preparations. [3H]uridine incorporation into cellular RNA of the actD-treated cells was reduced to 15% of the level in mock-treated cells. As expected, background [3H]uridine incorporation into virions was minimal, 8% or less at each time point (Fig. 6). [3H]uridine incorporation into cytoplasmic RNA of pac-, chx-, or aniso-treated cells was 53% ⫾ 8%, 70% ⫾ 21%, or 52% ⫾ 21%, respectively, of the level in mock-treated cells. These reductions in [3H]uridine incorporation are in part attributable to turnover of short-lived proteins that facilitate the stability of steady-state cellular RNA. [3H]uridine incorporation into virion RNA displayed coincident reduction to 48% ⫾ 8%, 30% ⫾ 5%, or 31% ⫾ 9%, respectively, of the mock-treated control level. These results indicate that inhibition of de novo translation decreases but does not abrogate the supply of vpRNA. Encapsidation efficiency is sustained upon translation inhibition. To evaluate the effect of translation inhibition on encapsidation efficiency of HIV-1 vpRNA, RPAs were performed with an RNA probe complementary to the HIV-1 5⬘ untranslated region. To control for variation in cytoplasmic RNA loading, cytoplasmic RNA samples were also hybridized to a probe complementary to cellular gapdh RNA. To monitor for possible variation in virion RNA processing, the virion samples were supplemented with a 100-bp 32P-labeled DNA following probe hybridization and digestion and before phenol extraction and ethanol precipitation. Four independent RPAs were performed using 10 ␮g of cytoplasmic RNA and virion RNA equivalent to 250 ng of Gag p24. A representative RPA is shown in Fig. 7A, and the data from the four RPAs are summarized in Fig. 7B. Encapsidation efficiency was calculated as the level of HIV-1 virion RNA relative to the level of cytoplasmic HIV-1 unspliced RNA.

Consistent with the [3H]uridine results, the absolute abundance of cytoplasmic HIV-1 unspliced RNA was reduced upon treatment with pac, chx, or aniso. The absolute amount of HIV-1 RNA in virions was also reduced. Compared to the mock-treated control, the levels of vpRNA encapsidation efficiency were 165% ⫾ 22%, 147% ⫾ 52%, and 97% ⫾ 21% in response to pac, chx, and aniso, respectively (Fig. 7B). Encapsidation efficiency was also increased by treatment with actD (200% [Fig. 7B]). These data indicate that vpRNA remains available for encapsidation during inhibition of de novo translation and that prior translation of the HIV-1 unspliced transcripts is not necessary for generation of vpRNA. The actD

FIG. 6. HIV-1 vpRNA remains available for encapsidation during translation inhibition. CEM(A)/HIV-1 cells were incubated for 4 h in medium containing 30 ␮Ci of [3H]uridine per ml with or without pac, chx, or aniso. [3H]uridine levels in cytoplasmic RNA and in virion RNA were quantified by TCA precipitation analysis. Average results of three experiments are shown; error bars indicate standard deviations.

11536


FIG. 7. Encapsidation efficiency is not reduced upon inhibition of de novo translation. (A) Representative RPA of cytoplasmic and virion RNA that was harvested after 4 h of incubation with or without pac, chx, or aniso. Labels indicate the sizes of the protected RNAs and control 100-bp DNA fragment used to control for viral RNA sample processing (virus control), cell treatment, and undigested probes. (B) Summary of four RPAs. Average results are shown; error bars indicate standard deviations. Encapsidation efficiency was determined by dividing virion RNA level by the corresponding cytoplasmic RNA level.

data indicate that de novo RNA synthesis and Rev shuttling are not necessary for encapsidation of vpRNA. DISCUSSION We developed a biochemical assay to examine the relationship between HIV-1 unspliced mRNA and vpRNA. The assay uses three mechanistically distinct translation antagonists to inhibit protein synthesis in HIV-1-infected T cells under conditions that maintain virion production. Our ribosomal profile analysis comparing HIV-1-infected and mock-infected T cells agrees with the ribosomal profiles of Agy et al. (1) with the exception that we do not observe a significant reduction in the overall abundance of rRNA in response to HIV-1 infection. Our experiments show that uninfected and HIV-1-infected cells exhibit the expected distinct profiles in response to pac, chx, or aniso, which selectively inhibit either the initiation or

J. VIROL.

elongation step of translation. These data confirm that HIV-1 infection does not disrupt the translation machinery. Pulse-chase experiments and immunoprecipitation assays established that de novo translation is not necessary for HIV1 particle assembly and release, and that a concentration-dependent relationship exists between cell-associated Gag and virion-associated Gag. The newly synthesized Gag can be readily assembled into virions, but steady-state Gag is also sufficient to produce virions. Immunoprecipitation results also indicate that inhibition of protein synthesis does not interfere with processing of Gag precursor protein. Examination of cytoplasmic and virion RNA by RPA and [3H]uridine labeling demonstrated that de novo translation is not required for encapsidation of vpRNA. The absolute level of virion RNA is reduced upon translation inhibition. The magnitude of this reduction was greater when measured by the [3H]uridine labeling approach, which detects both host and viral transcripts, than by the HIV-1-specific RPA. One possible explanation for this difference is that the 3H-labeling technique detects changes in encapsidation of host RNAs. Treatment with the inhibitors increased the cytoplasmic availability of vpRNA and yielded increased encapsidation efficiencies, indicating that HIV-1 mRNA can also be utilized as vpRNA. This ability to increase vpRNA availability indicates that generation of vpRNA does not require prior utilization of the HIV-1 unspliced RNA as the mRNA template for protein synthesis. We speculate that disruption of the protein synthesis machinery reduces competition by ribosomes, and the HIV-1 mRNA is rerouted to function as vpRNA. Our data imply that HIV-1 vpRNA and mRNA do not follow a separate intracellular RNA pathway. Instead HIV-1 unspliced RNA constitutes a single functional pool that can function interchangeably as mRNA and as vpRNA. Our results are similar to results with ASV and RSV in which Stoltzfus et al. (23) and Sonstegard and Hackett (22) concluded that a single metabolic pool of viral RNA exists that functions as both mRNA and vpRNA. Contrasting results in the MLV system suggested that there are two nonequilibrating pools of MLV RNA, each functioning as either mRNA or vpRNA (11). In the MLV system, actDtreated cells produced virions without genomic RNA. In our HIV-1 system, actD-treated cells sustain production of virions with genomic RNA and exhibit increased encapsidation efficiency. Our biochemical results from HIV-1-infected human T cells are in agreement with genetic analysis of HIV-1-based vectors (13) and indicate that translation of HIV-1 vector mRNA is not a rate-limiting step in production of vector virus. Our data contrast with HIV-2 experiments in which continued protein synthesis was required for encapsidation of vpRNA (8). This feature of the HIV-2 system is presumed to be necessary for sorting HIV-2 genomic RNA because the RNA encapsidation signal is present on both the HIV-2 unspliced vpRNA and spliced mRNA (8, 14). We speculate that for HIV-1, interaction of HIV-1 Gag protein with the RNA encapsidation signal modulates the competition between host translational machinery and virus assembly complexes, similar to the mechanism originally proposed from study of RSV (22). ACKNOWLEDGMENTS CEM(A), from Mark Wainberg and James McMahon, was obtained through the AIDS Research and Reference Reagent Program, Division of AIDS, NIAID, NIH. This work was supported by grants from the National Institute Allergy and Infectious Diseases (R29AI40851) and the National Cancer Institute (P30CA16058), Bethesda, Md.

VOL. 74, 2000

CYTOPLASMIC TRAFFICKING OF HIV-1 UNSPLICED RNA REFERENCES

1. Agy, M. B., M. Wambach, K. Foy, and M. G. Katze. 1990. Expression of cellular genes in CD4 positive lymphoid cells infected by the human immunodeficiency virus, HIV-1: evidence for a host protein synthesis shut-off induced by cellular mRNA degradation. Virology 177:251–258. 2. Butsch, M., S. Hull, Y. Wang, T. M. Roberts, and K. Boris-Lawrie. 1999. The 5⬘ RNA terminus of spleen necrosis virus contains a novel posttranscriptional control element that facilitates human immunodeficiency virus Rev/ RRE-independent Gag production. J. Virol. 73:4847–4855. 3. Coder, D. M. 1997. Assessment of cell viability, p. 9.2.1–9.2.2. In J. P. Robinson et al. (ed.), Current protocols in cytometry, vol. 1. John Wiley & Sons, Inc., New York, N.Y. 4. Felicetti, L., B. Colombo, and C. Baglioni. 1966. Inhibition of protein synthesis in reticulocytes by antibiotics. II. The site of action of cycloheximide, streptovitacin A and pactamycin. Biochim. Biophys. Acta 119:120–129. 5. Fresno, M., L. Carrasco, and D. Vazquez. 1976. Initiation of the polypeptide chain by reticulocyte cell-free systems. Survey of different inhibitors of translation. Eur. J. Biochem. 68:355–364. 6. Geballe, A. P., and M. K. Gray. 1992. Variable inhibition of cell-free translation by HIV-1 transcript leader sequences. Nucleic Acids Res. 20:4291– 4297. 7. Goldberg, I. H. 1974. Mechanisms of action of antitumor antibiotic inhibitors of protein synthesis. Cancer Chemother. Rep. 58:479–489. 8. Kaye, J. F., and A. M. Lever. 1999. Human immunodeficiency virus types 1 and 2 differ in the predominant mechanism used for selection of genomic RNA for encapsidation. J. Virol. 73:3023–3031. 9. Kim, V. N., K. Mitrophanous, S. M. Kingsman, and A. J. Kingsman. 1998. Minimal requirement for a lentivirus vector based on human immunodeficiency virus type 1. J. Virol. 72:811–816. 10. Levin, J. G., P. M. Grimley, J. M. Ramseur, and I. K. Berezesky. 1974. Deficiency of 60 to 70S RNA in murine leukemia virus particles assembled in cells treated with actinomycin D. J. Virol. 14:152–161. 11. Levin, J. G., and M. J. Rosenak. 1976. Synthesis of murine leukemia virus proteins associated with virions assembled in actinomycin D-treated cells: evidence for persistence of viral messenger RNA. Proc. Natl. Acad. Sci. USA 73:1154–1158. 12. McBride, M. S., and A. T. Panganiban. 1996. The human immunodeficiency virus type 1 encapsidation site is a multipartite RNA element composed

13.

14. 15. 16. 17.

18. 19.

20. 21. 22. 23. 24.

11537

of functional hairpin structures. J. Virol. 70:2963–2973. (Erratum, 71:858, 1997.) McBride, M. S., M. D. Schwartz, and A. T. Panganiban. 1997. Efficient encapsidation of human immunodeficiency virus type 1 vectors and further characterization of cis elements required for encapsidation. J. Virol. 71: 4544–4554. McCann, E. M., and A. M. Lever. 1997. Location of cis-acting signals important for RNA encapsidation in the leader sequence of human immunodeficiency virus type 2. J. Virol. 71:4133–4137. Messer, L. I., J. G. Levin, and S. K. Chattopadhyay. 1981. Metabolism of viral RNA in murine leukemia virus-infected cells; evidence for differential stability of viral message and virion precursor RNA. J. Virol. 40:683–690. Meyer, B. E., and M. H. Malim. 1994. The HIV-1 Rev trans-activator shuttles between the nucleus and the cytoplasm. Genes Dev. 8:1538–1547. Miele, G., A. Mouland, G. P. Harrison, E. Cohen, and A. M. Lever. 1996. The human immunodeficiency virus type 1 5⬘ packaging signal structure affects translation but does not function as an internal ribosome entry site structure. J. Virol. 70:944–951. Naldini, L., U. Blomer, P. Gallay, D. Ory, R. Mulligan, F. H. Gage, I. M. Verma, and D. Trono. 1996. In vivo gene delivery and stable transduction of nondividing cells by a lentiviral vector. Science 272:263–267. Obrig, T. G., W. J. Culp, W. L. McKeehan, and B. Hardesty. 1971. The mechanism by which cycloheximide and related glutarimide antibiotics inhibit peptide synthesis on reticulocyte ribosomes. J. Biol. Chem. 246:174– 181. Paskind, M. P., R. A. Weinberg, and D. Baltimore. 1975. Dependence of Moloney murine leukemia virus production on cell growth. Virology 67:242– 248. Roberts, T. M., and K. Boris-Lawrie. 2000. The 5⬘ RNA terminus of spleen necrosis virus stimulates translation of nonviral mRNA. J. Virol. 74:8111– 8118. Sonstegard, T. S., and P. B. Hackett. 1996. Autogenous regulation of RNA translation and packaging by Rous sarcoma virus Pr76gag. J. Virol. 70:6642– 6652. Stoltzfus, C. M., K. Dimock, S. Horikami, and T. A. Ficht. 1983. Stabilities of avian sarcoma virus RNAs: comparison of subgenomic and genomic species with cellular mRNAs. J. Gen. Virol. 64:2191–2202. Vazquez, D. 1979. Inhibitors of protein biosynthesis. Mol. Biol. Biochem. Biophys. 30:1–14, 36–38, 136–138, 155–159.