Interaction between the human immunodeficiency virus type 1 Gag matrix domain and phosphatidylinositol-(4,5)-bisphosphate is essential for efficient gag membrane binding

Abstract

Human immunodeficiency virus type 1 (HIV-1) particle assembly mediated by the viral structural protein Gag occurs predominantly on the plasma membrane (PM). Although it is known that the matrix (MA) domain of Gag plays a major role in PM localization, molecular mechanisms that determine the location of assembly remain to be elucidated. We observed previously that overexpression of polyphosphoinositide 5-phosphatase IV (5ptaseIV) that depletes PM phosphatidylinositol-(4,5)-bisphosphate [PI(4,5)P(2)] impairs virus particle production and redirects processed Gag to intracellular compartments. In this study, we examined the impact of PI(4,5)P(2) depletion on the subcellular localization of the entire Gag population using Gag-fluorescent protein chimeras. Upon 5ptaseIV overexpression, in addition to perinuclear localization, Gag also showed a hazy cytosolic signal, suggesting that PI(4,5)P(2) depletion impairs Gag membrane binding. Indeed, Gag was less membrane bound in PI(4,5)P(2)-depleted cells, as assessed by biochemical analysis. These observations are consistent with the hypothesis that Gag interacts with PI(4,5)P(2). To examine a putative Gag interaction with PI(4,5)P(2), we developed an in vitro binding assay using full-length myristoylated Gag and liposome-associated PI(4,5)P(2). Using this assay, we observed that PI(4,5)P(2) significantly enhances liposome binding of wild-type Gag. In contrast, a Gag derivative lacking MA did not require PI(4,5)P(2) for efficient liposome binding. To analyze the involvement of MA in PI(4,5)P(2) binding further, we examined MA basic amino acid substitution mutants. These mutants, previously shown to localize in perinuclear compartments, bound PI(4,5)P(2)-containing liposomes weakly. Altogether, these results indicate that HIV-1 Gag binds PI(4,5)P(2) on the membrane and that the MA basic domain mediates this interaction.

FIG. 1.

Overexpression of 5ptaseIV relocalizes Gag from the PM to perinuclear compartments and cytosol. HeLa cells expressing GagVenus (A and B) or GagmRFP (B) along with full-length 5ptaseIV (FL) or the Δ1 derivative (Δ1) were immunostained with anti-Myc (for 5ptaseIV) and analyzed with a Nikon TE2000 microscope. The number of cells with (i) Gag localized at the PM (small arrows, gray), (ii) Gag localized predominantly at intracellular compartments (arrowheads, white), and (iii) Gag localized only in the cytosol (large arrows, black) were counted. At least 100 5ptaseIV-positive, Gag-positive cells were examined for each condition. (C) Lysates of HeLa cells transfected with pNL4-3/PR⁻, pNL4-3/GagVenus, or pNL4-3/GagmRFP were subjected to SDS-PAGE and analyzed by immunoblotting using HIV-Ig, anti-GFP, or anti-DsRed. Note that predominantly single bands were detected using antibodies against fluorescent proteins.

FIG. 2.

Membrane binding of WT Gag is reduced in 5ptaseIV-overexpressing cells. HeLa cells expressing WT or Fyn(10)fullMA Gag along with full-length 5ptaseIV (FL) or the Δ1 mutant were pulse-labeled for 5 min and chased for 20 min. Postnuclear supernatants of cell homogenates were subjected to membrane flotation centrifugation. (A) Membrane-bound (M) and non-membrane-bound (NM) Gag was recovered by immunoprecipitation and subjected to SDS-PAGE. (B) The labeled Gag protein was quantified by phosphorimager analysis, and the membrane binding efficiency was calculated. Data from six (WT) or four [Fyn(10)fullMA] different experiments are shown as means ± standard deviations. P values were determined using Student's t test. ns, not significant; **, P < 0.005.

FIG. 3.

Gag binds PI(4,5)P₂ specifically in a concentration-dependent manner. ³⁵S-labeled WT Gag was synthesized in the in vitro transcription and translation system by using reticulocyte lysates and was incubated with liposomes containing various amounts of acidic phospholipids. (A and B) PC:PS (2:1) liposomes containing various amounts of PI(4,5)P₂ were used. (C and D) Liposomes containing increased levels of PS were compared with liposomes containing 62.75 mol% PC, 30 mol% PS, and 7.25 mol% PI(4,5)P₂. (E and F) A concentration of 7.25 mol% of PI(4,5)P₂, PI(3,5)P₂, and PI(3,4,5)P₃ was compared in the context of the PS:PS (2:1) liposomes. The reaction mixtures were subjected to membrane flotation centrifugation, and five 1-ml fractions were collected from each tube. M, liposome bound; NM, not liposome bound. Fractions were subjected to SDS-PAGE followed by autoradiography (A, C, and E). The amount of labeled Gag in each fraction was quantified using a phosphorimager, and the percentage of labeled Gag in the membrane-bound fraction versus total amount of labeled Gag was calculated (B, D, and F). For panel B, data from seven (0 and 7.25 mol%), five (0.63 mol%), and four (2.1, 3.52, and 5.71 mol%) different experiments are shown as means ± standard deviations. P values were determined between 0% PI(4,5)P₂ and various percentages of PI(4,5)P₂ using Student's t test (ns, not significant; **, P < 0.005; ***, P < 0.001). In panels D and F, data from three and four different experiments, respectively, are shown (ns, not significant; *, P < 0.05; **, P < 0.01).

FIG. 4.

Myristoylation-defective mutant Gag does not bind liposomes efficiently even in the presence of a high percentage of PI(4,5)P₂. (A) WT or 1GA Gag was synthesized in the reticulocyte lysate system in the presence of [³⁵S]methionine-cysteine or [³H]myristic acid. The major band (labeled Pr55^Gag) comigrated with Gag immunoprecipitated from lysates of HeLa cells transfected with pNL4-3/PR⁻ in a parallel experiment (data not shown). In addition to the Pr55^Gag band, three minor bands (asterisk) were detected in the in vitro translation reaction of WT Gag performed in the presence of the ³⁵S-labeled amino acids. Two of the minor bands that are also labeled with ³H likely represent Gag species arising from premature termination of translation, whereas the one band labeled only with ³⁵S may be formed by internal ribosomal entry. In vitro translation reaction mixtures for 1GA Gag as well as other Gag derivatives used in this study (data not shown) contained the same minor bands with similarly low abundance. (B and C) ³⁵S-labeled WT Gag or myristoylation mutant (1GA) Gag was synthesized as for Fig. 3 and incubated with liposomes containing or not containing PI(4,5)P₂. The reaction mixtures were subjected to membrane flotation centrifugation, and fractions were analyzed as for Fig. 3 (B). The amount of labeled Gag in each fraction was quantified using a phosphorimager, and the percentage of labeled Gag in the membrane-bound fraction versus the total amount of labeled Gag was calculated (C). The relative liposome-binding efficiency for each condition was calculated in comparison to the binding efficiency of WT Gag with PI(4,5)P₂-containing liposomes. Data from three different experiments are shown as means ± standard deviations. P values were determined using Student's t test. ***, P < 0.001.

FIG. 5.

Addition of an efficient membrane-binding sequence does not eliminate the requirement for PI(4,5)P₂ in Gag liposome binding. (A) Schematic representation of Gag derivatives with the Fyn N-terminal sequence. Myristoylation (m) and palmitoylation (palm) sites are shown. (B) ³⁵S-labeled WT Gag, Fyn(10)fullMA Gag, and Fyn(10)ΔMA Gag were synthesized and incubated with liposomes as for Fig. 4. The reaction mixtures were subjected to membrane flotation centrifugation, and the fractions were analyzed as for Fig. 3. (C) The amount of labeled Gag in each fraction was quantified using a phosphorimager, and statistical analysis was performed as for Fig. 4. Data from four different experiments are shown as means ± standard deviations. P values were determined using Student's t test. ***, P < 0.001; *, P < 0.05; ns, not significant.

FIG. 6.

Virus release efficiency and PM localization of Fyn(10)fullMA Gag are still sensitive to PI(4,5)P₂ depletion in cells. (A) HeLa cells expressing WT, Fyn(10)fullMA, or Fyn(10)ΔMA GagVenus along with full-length 5ptaseIV (FL) or the Δ1 mutant (Δ1) were immunostained with anti-Myc antibody and analyzed with a Nikon TE2000 microscope. (B) The number of cells with (i) Gag localized at the PM (small arrows, gray), (ii) Gag localized predominantly at intracellular compartments (arrowheads, white), and (iii) Gag localized only in the cytosol (large arrows, black) were counted as for Fig. 1. Note that for cells expressing Fyn(10)ΔMA Gag, a majority of cells showing Gag signals at the intracellular vesicles (category 2) also displayed Gag signals at the plasma membrane. (C) HeLa cells expressing WT, Fyn(10)fullMA, or Fyn(10)ΔMA Gag along with full-length 5ptaseIV (FL) or Δ1 mutant (Δ1) were metabolically labeled for 120 min. Cell- and virus-associated Gag were recovered by immunoprecipitation and analyzed by SDS-PAGE. (D) Signal intensity of labeled Gag was quantified by phosphorimager analysis. Virus release efficiency was calculated as the amount of virus-associated Gag as a fraction of total Gag synthesized during the labeling period and normalized to the virus release efficiency in 5ptaseIVΔ1-expressing cultures. The average virus release efficiencies by cells expressing WT, Fyn(10)fullMA, and Fyn(10)ΔMA Gag along with 5ptaseIVΔ1 were 15.6%, 27%, and 21.3%, respectively. Data from four different experiments are shown as means ± standard deviations. P values were determined using Student's t test. ***, P < 0.001; **, P < 0.005; *, P < 0.05.

FIG. 7.

The basic domain of MA is important for efficient Gag-PI(4,5)P₂ interaction. ³⁵S-labeled WT Gag and basic domain mutants (29KE/31KE or 29KT/31KT) were synthesized and incubated with liposomes as for Fig. 4. (A) The reaction mixtures were subjected to membrane flotation centrifugation, and the fractions were analyzed as for Fig. 3. (B) The amount of labeled Gag in each fraction was quantified using a phosphorimager, and the statistical analysis was performed as for Fig. 4. Data from three different experiments are shown as means ± standard deviations. P values were determined using Student's t test. **, P < 0.005; *, P < 0.05; ns, not significant.