|Home | About | Journals | Submit | Contact Us | Français|
HIV-1 Gag assembles into virus particles predominantly at the plasma membrane (PM). Previously, we observed that phosphatidylinositol-(4,5)-bisphosphate [PI(4,5)P2] is essential for Gag binding to the plasma membrane and virus release in HeLa cells. In the current study, we found that PI(4,5)P2 also facilitates Gag binding to the PM and efficient virus release in T cells. Notably, serial passage of HIV-1 in an A3.01 clone that expresses polyphosphoinositide 5-phosphatase IV (5ptaseIV), which depletes cellular PI(4,5)P2, yielded an adapted mutant with a Leu-to-Arg change at matrix residue 74 (74LR). Virus replication in T cells expressing 5ptaseIV was accelerated by the 74LR mutation relative to replication of wild type HIV-1 (WT). This accelerated replication of the 74LR mutant was not due to improved virus release. In control T cells, the 74LR mutant releases virus less efficiently than does the WT, whereas in cells expressing 5ptaseIV, the WT and the 74LR mutant are similarly inefficient in virus release. Unexpectedly, we found that the 74LR mutation increased virus infectivity and compensated for the inefficient virus release. Altogether, these results indicate that PI(4,5)P2 is essential for Gag-membrane binding, targeting of Gag to the PM, and efficient virus release in T cells, which in turn likely promotes efficient virus spread in T cell cultures. In T cells with low PI(4,5)P2 levels, however, the reduced virus particle production can be compensated for by a mutation that enhances virus infectivity.
Particle formation of retroviruses, including HIV-1, is driven by the precursor polyprotein Gag. HIV-1 Gag consists of four major domains, matrix (MA), capsid (CA), nucleocapsid (NC), and p6, as well as two spacer peptides, SP1 and SP2 (1). MA is required for Gag targeting and binding to the plasma membrane (PM). CA and NC are essential for Gag multimerization. p6 recruits the cellular ESCRT complexes that facilitate release of virions. These domains give rise to individual mature Gag proteins upon proteolytic cleavage mediated by viral protease, which occurs during or immediately after virus particle release.
MA is composed of five major α-helices and a three-stranded β-sheet, which form a single globular domain (34). The N terminus of MA is modified by myristoylation, which is essential for membrane binding (7, 24, 27, 63). The myristoyl moiety is normally sequestered in the MA globular domain. Nuclear magnetic resonance (NMR) studies suggest that upon Gag multimerization or MA-phosphatidylinositol-(4,5)-bisphosphate [PI(4,5)P2] binding, the myristoyl moiety is exposed and mediates binding of Gag to the membrane (60, 65). In addition, the MA highly basic region (HBR), which is comprised of residues 17 to 31 in MA, interacts with acidic phospholipids in the inner leaflet of the membrane and is required for targeting of Gag to the PM in HeLa and T cells (33, 36, 53, 56, 71). Several amino acids within the MA globular domain are also involved in incorporation of the viral envelope glycoprotein into virions and the postentry process (5, 6, 8, 10, 15, 16, 21-23, 31, 32, 37, 38, 45, 48, 49, 55, 69, 70). Interaction of MA with the cytoplasmic tail of Env also regulates fusogenicity of Env (47, 68).
Accumulating evidence suggests that the MA domain of HIV-1 Gag, in particular the HBR, interacts with a PM-specific acidic phospholipid, PI(4,5)P2, thereby facilitating proper localization of Gag to the PM and efficient Gag-membrane binding. Overexpression of polyphosphoinositide 5-phosphatase IV (5ptaseIV), which removes the phosphate group at the D5 position of the inositol ring of PI(4,5)P2, significantly reduces HIV-1 release from HeLa and HEK293T cells (11, 51). This virus release defect is due to Gag mislocalization to intracellular compartments and reduced membrane binding (13, 51). Similarly, 5ptaseIV expression inhibits release of HIV-2, murine leukemia virus (MLV), and Mason-Pfizer monkey virus (MPMV) but not equine infectious anemia virus (EIAV) from cell lines including HeLa, HEK293, and 293T cells (11, 29, 59, 64). In the case of EIAV, PI(3)P and PI(3,5)P2 are required for virus release (19). Consistent with the possibility that HIV-1 MA mediates Gag-PI(4,5)P2 interactions, HIV-1 particles are enriched in PI(4,5)P2 in an MA-dependent manner (11). These findings indicate that PI(4,5)P2 interacts with MA and plays an important role in virus particle production at least in adherent cell lines, such as HeLa cells. However, the role of PI(4,5)P2 in HIV-1 assembly in T cells, a natural host cell type for HIV-1, still remains to be determined.
Several groups have examined interactions between PI(4,5)P2 and retroviral MA and/or Gag using various in vitro assays (2, 4, 12-14, 29, 59, 61, 62). NMR, surface plasmon resonance, and protein footprinting studies of HIV-1 MA/Gag-PI(4,5)P2 complexes revealed that the basic residues in HBR associate with water-soluble derivatives of PI(4,5)P2 (4, 61, 62). Interaction of HIV-1 Gag and MA with membrane-associated PI(4,5)P2 in its natural cellular form was demonstrated using liposomes or a lipid monolayer system (3, 13, 14). Furthermore, site-directed mutagenesis studies of the HBR showed that the basic residues in the HBR are important for binding to PI(4,5)P2 in liposomes as well (13, 14). Altogether, these reports strongly suggest that basic residues in the MA HBR play a key role in MA-PI(4,5)P2 binding. However, it remains to be determined whether residues outside of the HBR are also required for PI(4,5)P2 dependence of Gag.
In this study, we found that Gag membrane binding, Gag localization to the PM, and virus release efficiency were impaired in T cells expressing 5ptaseIV, demonstrating that PI(4,5)P2 facilitates HIV-1 assembly in T cells. Using T cell clones that express 5ptaseIV upon HIV-1 infection, we obtained a mutant virus adapted to low PI(4,5)P2 conditions. The adaptive Leu-to-Arg mutation at MA residue 74 impaired virus release in T cells with normal PI(4,5)P2 levels but allowed virus production at efficiencies comparable to that of wild type (WT) HIV-1 under low-PI(4,5)P2 conditions, suggesting that Leu74 is involved in PI(4,5)P2 dependence. Unexpectedly, this mutation increased infectivity of progeny virions, allowing HIV-1 to replicate with faster kinetics than the WT in cell clones expressing 5ptaseIV. These results demonstrate that HIV-1 can overcome the disadvantage imposed by the lack of PI(4,5)P2 through enhancement of virus infectivity in T cells.
pNL4-3/GagVenus, encoding Gag fused to the mVenus variant of yellow fluorescent protein (YFP) fusion at the C terminus of Gag, was described previously (13). pNL4-3/GagVenus/398/IRES-Myc-5ptaseIV was constructed using standard molecular cloning techniques using pNL4-3/GagVenus. pHIV-Myc-5ptaseIV and pHIV-Myc-5ptaseIVΔ1 were also described previously (13). Expression of Myc-5ptaseIV and Myc-5ptaseIVΔ1 genes in these plasmids is driven by the HIV-1 long terminal repeat (LTR) promoter that is transactivated by HIV-1 Tat. pNL4-3/KFS/398/IRES-Myc-5ptaseIV and pNL4-3/KFS/398/IRES-Myc-5ptaseIVΔ1 encode Myc-5ptaseIV and an inactive deletion mutant, Myc-5ptaseIVΔ1, respectively, following an internal ribosome entry site (IRES) sequence in place of the nef gene. A parental plasmid, pNL4-3/KFS/398, has the nef gene sequence replaced with a sequence containing multiple restriction sites derived from a plasmid, p398-6 (a kind gift from K. T. Jeang). These constructs also contain a frameshift mutation (KFS), which disrupts Env expression (23). pSV-A-MLV-env was obtained through the AIDS Research and Reference Reagent Program, Division of AIDS, NIAID, NIH, from Nathaniel Landau and Dan Littman (42). pNL4-3/MA74LR was constructed by PCR mutagenesis.
The human CD4+ T cell line A3.01 was cultured and maintained in RPMI 1640 (Invitrogen) supplemented with 10% heat-inactivated fetal bovine serum (FBS) (HyClone) (RPMI-10). To establish cell clones that express Myc-5ptaseIV or Myc-5ptaseIVΔ1 in a Tat-dependent manner, A3.01 cells were cotransfected with linearized plasmids encoding a puromycin resistance gene (pBABE-puro; a kind gift from V. KewalRamani) and myc-5ptaseIV or myc-5ptaseIVΔ1 genes driven by the HIV-1 LTR promoter (pHIV-Myc-5ptaseIV or pHIV-Myc-5ptaseIVΔ1) at a ratio of 1:10 by Nucleofection (Amaxa), cultured for 2 days, and then subjected to selection in the presence of 3 μg/ml puromycin. After puromycin-resistant cell clones were isolated by limiting dilution, these cells were maintained in RPMI-10 containing 3 μg/ml puromycin (Sigma). A CEM-GFP cell line, which expresses a green fluorescent protein (GFP) driven by the HIV-1 LTR promoter, was obtained through the AIDS Research and Reference Reagent Program, Division of AIDS, NIAID, NIH, from Jacques Corbeil (25) and maintained in RPMI-10 containing 500 μg/ml Geneticin (Invitrogen). HeLa cells were cultured in Dulbecco's modified Eagle's medium (DMEM) (Lonza) supplemented with 5% FBS (DMEM-5). The MAGI cell line, which is a HeLa cell clone expressing CD4 and HIV-LTR-β-galactosidase (β-Gal), was obtained through the AIDS Research and Reference Reagent Program, Division of AIDS, NIAID, NIH, from Michael Emerman (39) and cultured in DMEM-5.
Stocks of WT and 74LR mutant HIV-1 were prepared by transfecting 5.6 × 105 HeLa cells with 4 μg of pNL4-3 or pNL4-3/MA74LR. To prepare HIV-1 pseudotyped with murine leukemia virus (MLV)-Env, 1.6 μg of pNL4-3-derived molecular clones, 1.6 μg of pCMVNLGagPol-RRE (54), and 0.4 μg of pSV-A-MLV-env were cotransfected into 5.6 × 105 HeLa cells. Two days posttransfection, virus-containing supernatants were filtered through a 0.45-μm filter and used to inoculate T cells. Amounts of viruses were determined by a reverse transcriptase (RT) assay as described previously (24).
A virus release assay was performed as previously described (51). Briefly, T cells were transfected with molecular clones or infected with HIV-1 pseudotyped with MLV-Env. At 28 h posttransfection or 3 days postinfection, culture medium was changed to RPMI-1640 lacking both methionine (Met) and cysteine (Cys) and supplemented with 2% FBS [RPMI-2 (−Met/−Cys)] and incubated for 30 min. Subsequently, these cells were metabolically labeled with [35S]Met/Cys (Perkin-Elmer) in fresh RPMI-2 (−Met/−Cys) for 2 h. Cell and virion lysates were prepared and subjected to immunoprecipitation with HIV-Ig antiserum (NIH AIDS Research and Preference Reagent Program). The virus release efficiency was calculated as the amount of virion-associated Gag as a fraction of the total amount of Gag synthesized during the labeling period.
T cells were infected with WT or 74LR mutant HIV-1. At 2 days postinfection, these cells were labeled with [35S]Met/Cys in fresh RPMI-2 (−Met/−Cys) for 12 h. Virion lysates were prepared and subjected to immunoprecipitation with HIV-Ig antiserum. Env incorporation was compared based on the ratios of the amount of Env versus that of p24 CA in virions.
A liposome binding assay was performed as previously described (13).
Cells infected with MLV-Env-pseudotyped HIV-1 encoding GagVenus and either 5ptaseIV or 5ptaseIVΔ1 were fixed with 4% paraformaldehyde (Electron Microscopy Sciences) in phosphate-buffered saline (PBS) for 30 min at room temperature, permeabilized in PBS containing 0.2% (wt/vol) saponin and 10% FBS (abbreviated as 0.2% saponin) for 10 min at room temperature, and washed once with 10% FBS in PBS (10% FBS/PBS). The cells were incubated in 0.2% saponin containing mouse monoclonal anti-c-Myc antibody (9E10; Santa Cruz Biotechnology Inc.) for 1 h at room temperature, washed twice with 0.2% saponin, and stained with goat anti-mouse IgG conjugated to Alexa Fluor 594 (Invitrogen) for fluorescence microscopy or Alexa Fluor 647 (Invitrogen) for flow cytometry in 0.2% saponin for 30 min at room temperature. For fluorescence microscopy, these immunostained cells were subsequently washed with 10% FBS/PBS and mounted. Z-stack images of 50 fields were recorded using a Nikon TE2000 microscope, and cells positive for full-length 5ptaseIV or the Δ1 mutant were evaluated for a Gag localization pattern. At least 100 cells expressing GagVenus and 5ptaseIV were examined for each condition and counted using the ImageJ software program, v. 1.40g (NIH).
For flow cytometry, cells immunostained as described above were examined with a FACSCanto flow cytometer (BD Biosciences) and analyzed with the FlowJo software program, v. 8.7.1 (Tree Star Inc.).
For analysis of infectivity, virus stocks were prepared by infection of parental A3.01 cells or derivatives with HIV-1. Supernatants of infected cells were collected 3 days postinfection, and viruses in the supernatants were pelleted by ultracentrifugation and resuspended in RPMI-10. These virus stocks were treated with 1 U/μl of DNase I (Roche) for 30 min at room temperature, and equal amounts based on RT activity (200,000 cpm RT activity) were used to inoculate 2 × 105 of A3.01 cells or derivative clones for 2 h at 37°C. Cells were then washed and incubated at 37°C for 6 h.
Total DNA was purified from infected cultures using the DNeasy Blood & Tissue kit (Qiagen) and eluted in a total volume of 200 μl Qiagen AE buffer. Three microliters of DNA was used for real-time quantitative PCR analysis. For measurement of late reverse transcription products, viral DNAs spanning nucleotides 557 to 699 of HIV-1 were detected using the following primers and probe: forward primer (5′-TGTGTGCCCGTCTGTTGTGT-3′), reverse primer (5′-TCGCCTCGCCTCTTGCCGTGCGC-3′), and probe (5′-6-carboxyfluorescein (FAM)-CAGTGGCGCCCGAACAGGGA-BHQ-1-3′). This primer pair does not amplify the fragment encoding LTR-driven 5ptaseIV that was integrated in chromosomes of A3.01-derived cell clones used in this study. For measurement of early transcription products, viral DNAs spanning nucleotides 497 to 574 of HIV-1 were detected using the following primers and probe: forward primer (5′-GCTAACTAGGGAACCCACTGCTT-3′), reverse primer (5′-ACAACAGACGGGCACACACTAC-3′), and probe (5′-FAM-AGCCTCAATAAAGCTTGCCTTGAGTGCTTC-BHQ-1-3′). Reactions were performed in triplicate using iQ Supermix (Bio-Rad), 0.5 pmol/μl of each primer, and 0.125 pmol/μl of probe. After 1.5 min at 95°C, reactions were cycled for 15 s at 95°C followed by 1 min at 60°C for 40 repeats on an iCycler thermal cycler (Bio-Rad).
For analysis of virus infectivity using CEM-GFP cells, 5 × 105 cells were inoculated with virus stocks normalized by RT activity (200,000 cpm RT activity) for 2 h. To block the second-round infection, the CD4 blocking antibody Leu3a (0.25 μg/ml) (BD Biosciences) and the reverse transcriptase inhibitor zidovudine (AZT) (1 μM) (Sigma) were added to the medium 12 h postinfection. Two days postinfection, cells were fixed in 4% paraformaldehyde in PBS and analyzed with a FACSCanto flow cytometer and FlowJo software, v. 8.7.1.
For analysis of virus infectivity using MAGI cells, 4 × 104 cells were inoculated with virus stocks normalized by RT activity (200,000 cpm RT activity) for 2 h. To block the second-round infection, the CD4 blocking antibody Leu3a (0.25 μg/ml) and reverse transcriptase inhibitor AZT (1 μM) were added to the medium 12 h postinfection. Two days postinfection, cells were fixed in PBS containing 1% formaldehyde and 0.2% glutaraldehyde for 5 min at 4°C. After cells were washed twice, cells were stained in PBS containing 1 mg/ml 5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside (X-Gal) (Roche), 5 mM potassium ferrocyanide (Sigma), 5 mM potassium ferricyanide (Sigma), and 2 mM MgCl2 for 30 min at 37°C. Cells were then washed, and the blue foci exhibiting nuclear staining were counted.
For the analysis of virus replication phenotypes, 2 × 105 A3.01 cells or A3.01-derived cell clones were infected with HIV-1 for 2 h at 37°C. Every 2 days, culture supernatants were harvested for measurement of RT activity, and one-third of the cells were retained for continued culture. Virus-containing supernatants near the peak of RT activity were used for the next round of infection. DNA was prepared from infected cells near the peak of replication using a DNeasy Blood & Tissue kit (Qiagen). Proviral DNAs were amplified using specific primers, and sequences of the entire viral genomes were determined.
We demonstrated previously that PI(4,5)P2 is essential for Gag-PM binding and virus release in HeLa cells (13, 51). To determine whether PI(4,5)P2 is also required for Gag localization to the PM in T cells, A3.01 cells were transfected with pNL4-3 derivatives encoding GagVenus and either Myc-tagged 5ptaseIV or its inactive derivative 5ptaseIVΔ1, lacking the 5-phosphatase catalytic domain. At 28 h posttransfection, cells were fixed and immunostained with anti-Myc antibody. In cells expressing 5ptaseIVΔ1, Gag localized predominantly to the PM (Fig. (Fig.11 A and B). In contrast, in the presence of full-length 5ptaseIV, Gag localized less to the PM and more to intracellular compartments or in the cytosol, suggesting that PI(4,5)P2 depletion causes defects in Gag localization and membrane binding. We previously observed that when membrane binding was restored by adding the N-terminal sequence of Fyn kinase to Gag, mislocalization of this Gag derivative to intracellular compartments was still observed in 5ptaseIV-expressing HeLa cells (13). In the case of 5ptaseIV-expressing T cells, the Fyn-derived sequence increased localization of Gag to the PM but some mislocalization to intracellular compartments was still observed (Fig. (Fig.1B).1B). To investigate the effect of 5ptaseIV expression on the efficiency of virus release from A3.01 cells, we transfected A3.01 cells with pNL4-3 encoding full-length 5ptaseIV or 5ptaseIVΔ1. At 28 h posttransfection, these cells were metabolically labeled with [35S]Met/Cys for 2 h, and the amount of 35S-labeled Gag in the viral and cell lysates was examined by immunoprecipitation. The efficiency of virus release from A3.01 cells expressing full-length 5ptaseIV was 2.5-fold lower than that from A3.01 cells expressing 5ptaseIVΔ1 (Fig. 1C and D). As expected, the Fyn-derived membrane-binding sequence partially restored virus release efficiency in the presence of 5ptaseIV (Fig. 1C and D). The small reduction in release of Fyn(10)fullMA Gag in 5ptaseIV-expressing cells is consistent with the mislocalization of this Gag derivative shown in Fig. Fig.1B.1B. Altogether, these data indicate that PI(4,5)P2 facilitates Gag membrane binding and specific localization to the PM, thereby enhancing virus release in T cells.
In order to identify Gag regions contributing to PI(4,5)P2 dependence by a forward genetics approach, we sought to isolate HIV-1 mutants that are adapted to low-PI(4,5)P2 conditions. To this end, we established A3.01 cell clones that express 5ptaseIV under the control of the HIV-1 LTR promoter. These cell clones express 5ptaseIV at a high level only upon HIV-1 infection and following Tat expression (Fig. (Fig.22 A). Therefore, they allow us to examine HIV-1 replication under low-PI(4,5)P2 conditions without long-term expression of 5ptaseIV, which would induce cell death (39, 40). We first examined whether these clones express Myc-tagged 5ptaseIV or 5ptaseIVΔ1 upon infection with MLV-Env-pseudotyped HIV-1 encoding GagVenus (Fig. (Fig.2B).2B). Parental A3.01 cells showed no Myc-5ptaseIV signal in the GagVenus-positive population after HIV-1 infection. In contrast, in clones 04 and 07, 90% and 73%, respectively, of cells expressing a high level of GagVenus showed the Myc-5ptaseIV signals after HIV-1 infection.
Data described in Fig. Fig.11 suggest that depletion of PI(4,5)P2 reduces efficient virus release in T cells. To examine whether virus release is also reduced in clones 04 and 07, parental A3.01 cells and the A3.01-derived cell clones were infected with MLV-Env-pseudotyped HIV-1 that lacks HIV-1 Env expression. Three days postinfection, clones 04 and 07 showed a 40 to 50% reduction in virus release efficiency compared to that with parental A3.01 cells and Myc-5ptaseIVΔ1-expressing clone 13 (Fig. (Fig.33 A and B). Altogether these results indicate that in clones 04 and 07, HIV-1 infection induces expression of 5ptaseIV, which in turn impairs the late phase of the virus life cycle.
To isolate the HIV-1 mutants adapted to low-PI(4,5)P2 conditions, we inoculated the cell clones described above with HIV-1 and passaged the virus 5 times. In each round, culture supernatants near the peak of replication as monitored by RT activities were collected and used for the next round of infection. In parental A3.01 cells, replication kinetics of virus that was adapted to A3.01 cells through 4 rounds of infection (dark-gray squares) was slightly accelerated compared to that of the original NL4-3 (light-gray squares) (Fig. (Fig.44 A). The A3.01-adapted virus showed a similar acceleration of replication in clone 04 (Fig. (Fig.4B).4B). Importantly, however, in clone 04, the replication kinetics of virus adapted to clone 04 (black triangles) was further accelerated compared to that of the A3.01-adapted virus (Fig. (Fig.4B).4B). These observations suggest that the virus adapted to clone 04 may have acquired mutations that accelerate its replication kinetics in the presence of 5ptaseIV.
Sequence analysis of DNA derived from cell lysates prepared at virus replication peaks in clone 04 showed an emergence of a Leu-to-Arg change at MA amino acid 74 (74LR) (Fig. (Fig.4C)4C) and mutations in Vpu (data not shown). Similar Vpu mutations appeared even in viruses adapted to parental A3.01 cells (data not shown), while the 74LR mutation was observed only in the clone-04-adapted virus. Sequences encoding arginine (Arg74) instead of leucine (Leu74) at MA residue 74 were detected as early as the second round of replication. By the fourth round, the MA amino acid 74 became predominantly arginine (Fig. (Fig.4C).4C). Altogether, these results suggest a possibility that the 74LR mutation confers a fitness advantage to viruses replicating in T cells expressing 5ptaseIV.
To investigate whether the 74LR mutation is sufficient for accelerating HIV-1 replication in clones 04 and 07, we introduced this mutation to WT pNL4-3. Each cell clone was inoculated with WT and 74LR viruses prepared by transfection of HeLa cells with pNL4-3 and pNL4-3/MA74LR, respectively, and virus replication was monitored as described above. In parental A3.01 cells and 5ptaseIVΔ1-expressing clone 13, replication kinetics were not appreciably different between WT and 74LR viruses (Fig. (Fig.55 A and D). However, in 5ptaseIV-expressing clones 04 and 07, virus replication kinetics were markedly accelerated by the 74LR mutation (Fig. 5B and C). To investigate whether the presence of a basic amino acid at residue 74 is important for virus replication under low-PI(4,5)P2 conditions, we constructed additional MA mutants: 74LK, 74LD, 74LI, and 74LF. The 74LI and 74LF mutations have been found in clinical virus isolates described in the Los Alamos National Laboratory HIV Sequence Database. Interestingly, 74LI and 74LF mutants displayed replication kinetics similar to that of the WT, while 74LK and 74LD mutants failed to replicate even in parental A3.01 cells (data not shown). These results indicate that the 74LR change is involved in adaptation to the low-PI(4,5)P2 condition in a manner that is not solely dependent on charge.
Even though increased positive charge may not account for the effect of the 74LR change, it was possible that the 74LR Gag mutant had a higher affinity to PI(4,5)P2 than the WT, thereby allowing Gag to bind the PM even at lower PI(4,5)P2 levels. To determine whether the 74LR mutation affects PI(4,5)P2 binding, we performed a liposome binding assay. Liposome-bound and non-liposome-bound Gag proteins were separated by equilibrium flotation centrifugation using sucrose gradients. The amount of Gag in each fraction was quantified, and the percentage of Gag bound to the liposomes versus total Gag was calculated (Fig. (Fig.66 A and B). Previous studies showed that a 20LK mutation in MA increases Gag-membrane binding (37, 38, 52) and the affinity for PI(4,5)P2 (60). Consistent with these reports, we observed that this 20LK mutation increased Gag binding to PI(4,5)P2-containing liposomes, particularly when liposomes contained a smaller amount (3.52 mol%) of PI(4,5)P2 (Fig. 6A and B). In contrast, the liposome-binding efficiency of 74LR Gag was the same as or slightly lower than that of WT Gag. These results suggest that binding affinity for PI(4,5)P2 is not significantly altered by the 74LR mutation.
It is possible that the 74LR mutation restored efficient virus release via a PI(4,5)P2-independent mechanism, perhaps by facilitating Gag-PM binding or by augmenting other assembly steps. To test this possibility, we analyzed the impacts of the 74LR mutation on Gag localization and virus release in parental A3.01 cells and the cell clones described above. In the majority (75 to 85%) of A3.01 cells and clone 13, we observed that WT Gag localized at the PM (Fig. (Fig.77 A). In contrast, as observed in A3.01 cells transiently expressing 5ptaseIV (Fig. (Fig.1),1), WT Gag expressed in clones 04 and 07 remained in the cytosol in a large number of cells (approximately 60%). Interestingly, the 74LR mutation modestly reduced the cell population in which Gag localized at the PM in parental A3.01 cells and clone 13. In clones 04 and 07, however, this mutation did not cause a further reduction in PM localization of Gag (Fig. (Fig.7A).7A). Moreover, numbers of Gag puncta at the PM were generally lower with 74LR Gag than with WT Gag in parental A3.01 cells and clone 13 but were comparable between WT and 74LR Gag in clones 04 and 07 (Fig. (Fig.7B).7B). These results indicate that the 74LR mutation does not facilitate PM localization of Gag under low-PI(4,5)P2 conditions and reduces this under the natural PI(4,5)P2 conditions.
To examine the effect of the 74LR mutation on virus release efficiency, cell clones and parental A3.01 cells were infected with MLV-Env-pseudotyped HIV-1 encoding WT or 74LR Gag and cultured for 72 h. Subsequently, these infected cells were labeled with [35S]Met/Cys for 2 h, and viral and cell lysates were analyzed by immunoprecipitation (Fig. 7C and D). The release efficiency of the 74LR mutant was reduced in parental A3.01 cells and clone 13 compared to that of the WT. In clones 04 and 07, virus release efficiency of the WT showed a 2-fold reduction compared to that in A3.01 cells. Notably, however, the 74LR mutant showed no further reduction in virus release efficiency compared to the WT in clones 04 and 07. Altogether, these results indicate that Leu74 is important for efficient virus release in the presence of PI(4,5)P2 but not under low-PI(4,5)P2 conditions. Conversely, these results suggest that 74LR is defective in the PI(4,5)P2-dependent mode of virus release but is still capable of releasing virus in a PI(4,5)P2-independent manner.
It is possible that virus infectivity may be increased by the 74LR mutation, resulting in the accelerated replication kinetics under-low PI(4,5)P2 conditions compared to that of the WT (Fig. (Fig.5).5). Therefore, we examined the infectivity of the 74LR mutant generated from parental A3.01 cells or the clones analyzed above. Six hours postinfection, cells were collected, the copy numbers of newly synthesized HIV-1 late reverse transcription products and β-actin DNA were quantified by quantitative PCR (qPCR), and HIV-1 DNA copies normalized by β-actin DNA copies were calculated. When we utilized A3.01 cells as target cells, the infectivity of the 74LR mutant virus was higher than the infectivity of the WT (Fig. (Fig.88 A). A similar enhancement was observed when clones 04, 07, and 13 were used as target cells (Fig. (Fig.8A).8A). Synthesis of early reverse transcription products was also enhanced by the 74LR mutation in A3.01 cells (Fig. (Fig.8B).8B). Likewise, when CEM-GFP cells encoding green fluorescent protein under the control of the HIV-1 LTR were used for infectivity analysis, we observed that the infectivity of the 74LR mutant virus was higher than that of the WT (Fig. (Fig.8C).8C). In this assay, slightly larger differences between WT and the 74LR mutant viruses were observed when input viruses were prepared in clones 04 and 07. However, in both qPCR and CEM-GFP experiments, the enhanced infectivity of the 74LR mutant in T cells was observed regardless of T cell clones that produced input viruses (Fig. 8A to C), suggesting that a reduction in PI(4,5)P2 levels of producer cells is not essential for the effect of the 74LR change on infectivity. Env incorporation in 74LR mutant virus particles was not higher than that in WT virus particles (Fig. (Fig.8D),8D), indicating that the 74LR mutation enhances virus infectivity without changing the amount of Env incorporated into the virions.
In an attempt to compare mutant phenotypes between T cells and HeLa-derived cell lines, we tested infectivities of WT and 74LR mutant viruses in MAGI cells, a HeLa-derived cell line encoding β-Gal under the control of the HIV-1 LTR. Unexpectedly, in MAGI cells, infectivity of virions produced by A3.01 or A3.01-derived cell clones was decreased by the 74LR mutation (Fig. (Fig.8E).8E). These results indicate that the 74LR mutation alters virus infectivity in a cell type-specific manner. Altogether, these results suggest that the 74LR change enhances virus infectivity when target cells are T cells and that this enhancement leads to the acceleration of the virus replication in T cells in which virus production is inefficient due to low PI(4,5)P2 levels.
In this study, we observed that PI(4,5)P2 facilitates Gag binding to the PM and virus release in T cells, a natural host cell type for HIV-1. In order to obtain viruses adapted to low-PI(4,5)P2 conditions, we established novel cell clones 04 and 07, which express 5ptaseIV upon HIV-1 infection. In these cells, Gag largely remained in the cytosol, and virus particle production was reduced compared to that of parental A3.01 cells. HIV-1 adapted to clone 04 contained a MA mutation, 74LR. The 74LR change accelerated virus replication in clones 04 and 07 but did not improve virus release efficiency in these cell clones. Intriguingly, the 74LR mutation in MA instead enhanced infectivity of progeny virions. These data suggest that HIV-1 is capable of adapting to low-PI(4,5)P2 conditions through enhancement of virus infectivity in T cells.
The severity of virus release defects due to overexpression of 5ptaseIV was different between HeLa and T cells. As we reported previously, virus release efficiency was reduced 5- to 10-fold by overexpression of 5ptaseIV in HeLa cells. However, in T cells overexpressing 5ptaseIV, virus release efficiency was reduced up to 2.5-fold (Fig. (Fig.11 and and3).3). Perhaps due to this modest defect, it took 4 passages for the 74LR mutant to become dominant in the culture. There are at least two possibilities that could explain this somewhat reduced impact of 5ptaseIV expression on HIV-1 release in T cells compared to that in HeLa cells. First, some levels of PI(4,5)P2 may still remain in T cells in spite of 5ptaseIV overexpression. The GFP-tagged pleckstrin homology (PH) domain of phospholipase Cδ1, which binds specifically to PI(4,5)P2, localized at the PM in parental A3.01 cells and uninfected cell clones, but in the infected clones 04 and 07, this GFP construct exhibited diffused localization in the cytosol (data not shown). Therefore, PI(4,5)P2 is indeed reduced in clones 04 and 07 after HIV-1 infection. However, it remains unknown whether PI(4,5)P2 is completely depleted in these infected cell clones. Second, there may be an alternative, PI(4,5)P2-independent pathway of Gag localization and virus release in T cells. For instance, according to a lipidomics study, the concentration of PI(4,5)P2 in HIV-1 virions is significantly higher than the concentration of PI(4,5)P2 in the PM of the T cells those viruses were released from. However, this enrichment is not as significant as that observed in viruses released from monocyte-derived macrophages (11), indicating that PI(4,5)P2 may be less required for virus assembly in T cells. Therefore, it is conceivable that in T cells, a population of HIV-1 Gag may bind to the PM by a PI(4,5)P2-independent mechanism. Alternatively, virus could be released from not only the PM but also intracellular compartments in T cells. Consistent with this possibility, it has recently been shown that Gag that has been mistargeted to endosomes in T cells can still be released extracellularly (36).
In addition to the 74LR change in MA, mutations in the Vpu sequence were identified in HIV-1 adapted to T cell clones harboring the 5ptaseIV gene. Interestingly, similar changes occurred in viruses adapted to parental A3.01 cells, suggesting that changes in Vpu generally increase replication fitness of HIV-1 in A3.01 cells. This is consistent with a previous report that truncation of Vpu accelerates the cell-to-cell spread of virus in Jurkat cells (28). The cytoplasmic tail of Vpu is important for downmodulation of tetherin on the cell surface (18, 44, 46, 66, 67). Tetherin, a host defense protein, inhibits virus release by tethering the mature virus to the PM (17, 20, 30, 50, 57). Therefore, it is possible that tetherin-induced accumulation of completed viruses on the cell surface might accelerate the cell-to-cell spread of the virus, thereby providing an advantage to Vpu-truncated mutants for replication in A3.01 cells. However, whether tetherin plays a positive or negative role in the virus transmission via cell-cell contacts is still a controversial issue (9, 35, 41, 72). Thus, it remains to be investigated how truncated Vpu accelerates virus replication in A3.01 cells.
WT and 74LR Gag showed similar PI(4,5)P2-binding efficiencies in the liposome binding experiments (Fig. 6A and B). This result indicates that Leu74 does not play an important role in PI(4,5)P2 binding in vitro. However, Leu74 enhances Gag binding to the PM and virus release in a PI(4,5)P2-dependent manner in vivo (Fig. 7A to D). The apparent difference in effects of the 74LR mutation between in vivo and in vitro experiments might be due to the presence or absence of other proteins that interact with PI(4,5)P2. In liposome binding experiments using rabbit reticulocyte lysates, interaction of HIV-1 Gag with PI(4,5)P2 was examined perhaps in the presence of relatively low levels of other proteins that bind PI(4,5)P2 strongly, since membranes and hence proteins stably bound to membranes are largely removed from lysate preparations. In cells, however, there are cellular proteins, such as myristoylated alanine-rich C-kinase substrate (MARCKS), that interact with and sequester PI(4,5)P2 (26, 43). Moreover, it was recently reported that HIV-1 Tat also binds to PI(4,5)P2 with high affinity (58). These cellular and viral molecules likely compete with Gag for PI(4,5)P2 binding in cells. If there is a slight difference in PI(4,5)P2 affinity between WT and 74LR Gag, which may be undetectable in the liposome binding assay, it is conceivable that other PI(4,5)P2-binding proteins impair PI(4,5)P2 interaction of 74LR but not WT Gag.
The 74LR mutation enhanced an entry or early postentry step, as is evident from increased viral DNA synthesis in T cells (Fig. 8A to C). This effect of the 74LR change may be accounted for by one of several mechanisms in which MA affects viral infectivity. First, although involvement of MA in nuclear import of preintegration complexes is unsubstantiated (reviewed in reference 31), several reports showed that alteration of MA residues modulates other steps of the postentry process (5, 8, 10, 16, 31, 32, 37, 38, 45). For example, substitutions of Leu20 in MA impair viral DNA synthesis early postentry (37, 38). It is possible that the 74LR mutation modulates the same postentry process directly or affects MA residues involved in postentry steps (e.g., Leu20) and thereby enhances the viral infectivity after virus-cell fusion. Second, several MA residues are known to play key roles in Env incorporation into virions (5, 6, 10, 15, 16, 21-23, 48, 49, 55, 69, 70). However, we did not observe a significant impact of the 74LR mutation on the amount of Env in released virus particles (Fig. (Fig.8D).8D). Therefore, it is unlikely that the 74LR change increases virion infectivity via this mechanism. Finally, the interaction between immature Gag and Env has been shown to regulate virus-cell membrane fusion through the restriction of Env fusogenicity (47, 68). It is possible that in the context of mature MA, the 74LR mutation might affect Env conformation or free movement of Env on the virion surface in a manner opposite to what immature Gag does, thereby enhancing the membrane fusion process.
Surprisingly, the infectivity of the 74LR mutant produced from A3.01 cells and A3.01-derived cell clones was 2-fold reduced compared to that of the WT when examined using MAGI cells, a HeLa-derived cell line (Fig. (Fig.8E).8E). These results indicate that the enhanced infectivity of the 74LR mutant is a cell type-dependent phenotype. In light of this finding, it will be interesting to reexamine infectivity phenotypes of MA mutants, which were previously examined using either HeLa-derived indicator cell lines or T cells but not both. As for replication of the 74LR mutant, this cell type-specific difference in progeny virion infectivity prevented us from examining the 74LR mutant for a growth disadvantage relative to the WT in CD4-expressing cell lines derived from HeLa cells where virus assembly is largely PI(4,5)P2-dependent.
Since PI(4,5)P2 is required for efficient virus release in T cells, the MA-PI(4,5)P2 interaction may serve as a potential target for anti-HIV-1 drugs. However, it is of concern that the 74LR change may render HIV-1 resistant to compounds inhibiting MA-PI(4,5)P2 interactions. The replication kinetics of the 74LR mutant was comparable to that of WT in parental A3.01 cells (Fig. (Fig.4),4), likely because the enhanced infectivity of this mutant (Fig. (Fig.8)8) offsets the defect in virus release (Fig. (Fig.7).7). Therefore, if the 74LR mutant preexists in the virus population, it could quickly become dominant in the presence of an inhibitor of MA-PI(4,5)P2 interactions. However, in the Los Alamos National Laboratory HIV Sequence Database, Leu74 is relatively conserved, and other amino acids are rare at this position (L = 88.6%; I = 8.1%; F = 2.0%; V = 1.1%; R = 0.1%; M = 0.1%). Therefore, it is possible that Arg74 imposes lower fitness in vivo. We speculate that such a fitness defect may be due to an assembly or infectivity defect not evident in A3.01 cultures. For example, it is possible that the 74LR mutation may occur only in limited circumstances, such as A3.01 cells with low PI(4,5)P2 levels wherein a PI(4,5)P2-independent mode of virus assembly is available. Further analysis of the 74LR mutant in different cell types could elucidate the nature of PI(4,5)P2-dependent and -independent modes of virus assembly and the cell type-specific role of MA at early steps in HIV-1 life cycle.
We thank Eric Freed and the members of our laboratory for helpful discussions and critical reviews of the manuscript. We also thank Eric Freed, Kuan-Teh Jeang, and Vineet KewalRamani for providing plasmids. The following reagents were obtained through the AIDS Research and Reference Reagent Program, Division of AIDS, NIAID, NIH: HIV-Ig from NABI and NHLBI, pSV-A-MLV-env from Nathaniel Landau and Dan Littman, CEM-GFP from Jacques Corbeil, and HeLa-CD4-LTR-β-gal (MAGI) from Michael Emerman.
This work is supported by National Institutes of Health grant R01 AI071727 and amfAR research grant 107449-45-RGHF to A.O.
Published ahead of print on 26 January 2011.