Human embryonic kidney fibroblasts (HEK 293T) were cultured in Dulbecco's Modified Eagle Medium (Invitrogen) supplemented with 10% cosmic calf serum (HyClone) and 1% penicillin/streptomycin antibiotics (Multicell). Transfections were carried out using either the calcium phosphate precipitation method or the Lipofectamine 2000 reagent (Invitrogen). For Western blots, mouse and rabbit HRP-coupled secondary antibodies were purchased from Dako Cytomation and signals were detected using the Western Lightning Chemiluminescence Reagent Plus (PerkinElmer Life Sciences). Signals were detected with a Fluor-S MultiImager apparatus (Bio-Rad). Anti-Na-K ATPase antibodies were kindly provided by Dr. Michel Bouvier.

The construction of pcDNA3-RSV-Stau1⁵⁵-HA₃, pcDNA3-RSV-Stau1^F135A-HA₃, pcDNA-RSV-Stau1^ΔNt88-HA₃, pCMV-Stau1⁵⁵-YFP, pCMV-Stau1^F135A-YFP, pCMV-Stau1^ΔNt88-YFP, pCMV-Stau1^ΔdsRBD3-YFP, pCMV-Pr55^Gag-Rluc, pCMV-Pr55^Gag-YFP, pCMV-NC-p1-YFP and pCMV-CA-p2-NC-p1-Rluc was reported before 5152535459. The HxBRU PR-provirus and the Rev-independent Pr55^Gag expressor were described before 515360.

To construct pcDNA-RSV-Stau1^ΔNt37-HA₃, a polymerase chain reaction (PCR) was performed using pcDNA3-RSV-Stau1⁵⁵-HA₃ as template, sense (5'-ATCAGGTACCATGGGTCCATTTCCAGTTCCACCTTT-3') and anti-sense (5'-CACATCTAGATCATTTATTCAGCGGCCGCACTGAGCAGCGT-3') oligonucleotide primers and the Phusion DNA polymerase (New England Biolabs). The PCR product was purified and digested with KpnI and XbaI restriction enzymes (Fermentas) and then cloned into the KpnI/XbaI cassette of pCDNA3-RSV.

To generate pcDNA-RSV-Stau1⁵⁵-Flag plasmid, oligonucleotides (5'-GGCCTTGATTACAAGGATGACGATGACAAG-3' and 5'-GGCCCTTGTCATCGTCATCCTTGTAATCAA-3') were hybridized and then inserted into the NotI sites of pcDNA-RSV-Stau1⁵⁵-HA₃ in replacement of the HA-tag. For the construction of pcDNA-RSV-Stau1^ΔNt88-Flag, the EcoRI fragment of pcDNA-RSV-Stau1^ΔNt88-HA₃ that contained the mutated Stau1 sequence, was cloned into EcoRI-digested pcDNA-RSV-Stau1⁵⁵-Flag plasmid.

The expressors of NC-p1-YFP and Pr55^Gag-YFP mutants were PCR amplified using the PCR all-around technique 59 to generate the following mutations: the C15S mutation was introduced with the primer pair 5'-AAGAGTTTCAATTGTGGCAAA-3' and 5'-GAAACTCTTAACAATCTTTCT-3'; the C49S mutation was generated with the primer pair 5'-GATAGTACTGAGAGACAGGCT-3' and 5'-AGTACTATCTTTCATTTGGTG-3'; R7S, R10S and K11S mutations (R7 mutant) were introduced with the primer pair 5'-TTTAGCAACCAAAGCTCGATTGTTAAGTGTTTC-3' and 5'-AATCGAGCTTTGGTTGCTAAAATTGCCTCTCTG-3'. PCR reactions were carried out with the Phusion enzyme (New England Biolabs) at 95°C for 50 s, 55°C for 60 s and 72°C for 90 s, for 18 cycles. Resulting products were incubated with 10 units of DpnI enzyme (Fermentas) and then transformed into competent bacteria. Positive clones containing the mutation(s) were screened by restriction and sequencing analyses. The double zinc fingers mutant expressors (pCMV-Pr55^{Gag C15–49S}-YFP and pCMV-NC-p1^C15–49S-YFP) were generated by PCR with the oligonucleotide primer pair for the C49S mutation using the corresponding plasmids that contain the C15S mutation.

Forty hours post-transfection, cell extracts were prepared by passing the cells 20 times through a 23G1 syringe in TE (10 mM Tris pH7.4, 1 mM EDTA pH 8) containing 10% sucrose and proteases inhibitors (Roche). Nuclei were removed by centrifugation at 1,000 × g. Resulting cytoplasmic extracts were separated using the membrane flotation assay as previously described 51. Membrane-associated complexes were collected (fractions 2 and 3). Membranes were solubilized by treating these complexes with 0.5% Triton X-100 at room temperature for 5 minutes and samples were subjected to S100/P100 fractionation as previously described 51 by ultracentrifugation at 100,000 × g for 1 h at 4°C. Supernatants (S100 fractions) and pellets (P100 fractions) were collected and analyzed by Western blotting using anti-CA, anti-HA and anti-Na-K ATPase mouse antibodies.

293T cells were transfected in 6-well plates with constant amounts of the Rluc-fused energy donor expressor (25–75 ng), increasing amounts of YFP-fused acceptor expressor (0.25–2 μg) and Stau1-HA₃-expressing plasmid (1–1.5 μg) when indicated. 48 hours post-transfection, cells were collected in PBS-EDTA 5 mM and diluted to approximately 2 × 10⁶ cells/mL. BRET assays were performed as described before 5152 using a Fusion α-FP apparatus (Perkin-Elmer). In this interaction assay, an X-Rluc fusion protein is used as an energy donour whereas a Y-YFP fusion protein is an energy acceptor. When the two fusion proteins are in close proximity (< 100Å), non-radiative resonance energy is transferred from X-Rluc to Y-YFP which in turn emits measurable fluorescence. This can be quantified by the calculation of the BRET ratio which allows detection of protein-protein interactions. The BRET ratio was defined as [(emission at 510 to 590 nm)-(emission at 440 to 500 nm) × Cf]/(emission at 440 to 500 nm), where Cf corresponds to (emission at 510 to 590 nm)/(emission at 440 to 500 nm) when Rluc fused protein is expressed alone. The total YFP activity/Rluc activity ratio reflects the relative levels of the two fusion proteins in the cells. The BRET ratio increases with the total YFP activity/Rluc activity ratio since more YFP-fused molecules bind to Rluc-fused proteins. For Pr55^Gag multimerization assays, in order to avoid misinterpretation due to variations in relative levels of the Pr55^Gag fusion proteins, changes in the Pr55^Gag/Pr55^Gag BRET ratios following Stau1 overexpression were always analyzed at similar total YFP activity/Rluc activity ratio.

When Pr55^Gag/Pr55^Gag BRET assays were performed following membrane flotation assays, the Rluc substrate coelenterazine H (NanoLight Technology) was added to 90 μL of each fraction and BRET ratio was determined as in live cells. BRET ratios in fractions 1, 3, 4, 5 and 6 were not considered because luciferase activity was too low in these fractions and hence, did not lead to the determination of a reliable BRET ratio.

For CA-p2-NC-p1-Rluc/Stau1-YFP and Stau1⁵⁵-Rluc/NC-p1-YFP interaction assays, BRET ratios were always compared at similar total YFP activity/Rluc activity ratio. The BRET ratio determined in the context of the expression of the unfused YFP protein (YFP) corresponds to non specific interactions between the energy donor and the YFP. Hence, this background BRET ratio was always subtracted from all BRET ratios and was set to 0%. The BRET ratio determined following co-expression of the energy donor and the wild type energy acceptor was set to 100%.

For dose-response Pr55^Gag/Pr55^Gag BRET assays, 293T cells were transfected with fixed amounts of pCMV-Pr55^Gag-Rluc and pCMV-Pr55^Gag-YFP and increasing amounts (0.25–2 μg) of different Stau1-HA₃ expressors. BRET assays were performed 48 hours post-transfection as described above.

293T cells were transfected with Stau1⁵⁵-flag and Gag expressors using Lipofectamine 2000 (Invitrogen). Twenty hours post-transfection, cells were collected in lysis buffer (150 mM NaCl, 50 mM Tris pH 7.4, 1 mM EDTA, 1% Triton X-100) containing proteases inhibitors (Roche). Each cell lysate (1.5 mg of proteins) was pre-cleared with IgG-agarose (Sigma-Aldrich) for 1 h at 4°C and then subjected to immunoprecipitation using 15 μL of anti-Flag M2 affinity gel (Sigma-Aldrich) for 2 h at 4°C. Immune complexes were washed 3 times during 5 minutes with cold lysis buffer, eluted with the Flag peptide (Sigma-Aldrich), resolved in SDS-containing acrylamide gels and analyzed for their content in Stau1 and Gag proteins by Western blotting using mouse monoclonal anti-Flag (Sigma-Aldrich), anti-GFP (Roche) and anti-CA antibodies.

293T cells were transfected with Stau1⁵⁵-HA₃ and Gag expressors using Lipofectamine 2000 (Invitrogen). Twenty hours post-transfection, supernatants were collected and cleared through a 0.45 μm filter. VLPs were pelleted through a sucrose cushion (20% in Tris-NaCl buffer) by ultracentrifugation during 1 hour at 220,000 × g. VLPs were resuspended in Tris-NaCl buffer and analyzed by Western blotting using anti-CA antibodies. Pr55^Gag signals in the VLPs and the cell extracts were quantitated using the Quantity One (version 4.5) software (Bio-Rad).

The interaction between Stau1 and Pr55^Gag is likely a critical determinant for Stau1 function in HIV-1 assembly. Indeed we previously showed that a single point mutation in the third double-stranded RNA-binding domain of Stau1 (Stau1^F135A) prevented both the association of the mutant to Pr55^Gag and the Stau1-mediated increase of HIV-1 assembly 515253. Moreover, we showed that Stau1/Pr55^Gag interaction required the NC domain 52 that contains motifs involved in several steps during HIV-1 assembly. As a first step, to understand the molecular mechanisms underlying Stau1 influence on HIV-1 assembly, we identified which NC sub-domain is required for Pr55^Gag/Stau1 association using the BRET assay with Stau1⁵⁵-Rluc and wild type or mutant NC-p1-YFP fusion proteins. Four NC mutants were constructed. Point mutations were introduced in the NC-p1-YFP fusion protein to disrupt the first zinc finger (NC-p1^C15S-YFP), the second (NC-p1^C49S-YFP), both zinc fingers (NCp1-YFP^C15–49S) or the N-terminal basic residues (NCp1-YFP^R7)(Figure 1A). For this mutant, Arg⁷, Arg¹⁰ and Lys¹¹ were substituted for serines (Figure 1A). Mutations in this basic region were previously reported to severely affect HIV-1 assembly 24. Constructs encoding the wild type and mutants NC fusion proteins were transfected in 293T cells and their expression patterns were analyzed by Western blotting using an anti-GFP antibody. Figure 1B shows that wild type and mutant NC-p1-YFP proteins were well expressed and have the expected molecular weight. However, for unknown reasons, NC-p1^C15–49S-YFP was always slightly less expressed than the other NC-p1-YFP proteins.

These proteins were then tested for their capacity to interact with Stau1⁵⁵ using the BRET assay in live 293T cells (Figure 2A). This technique allows us to detect protein-protein interaction in live cells between Rluc-fused Stau1 and NC-p1-YFP molecules (Figure 2A). Indeed, when the two fusion proteins are in close proximity (≤ 100Å) as a consequence of Stau1-NC interaction, non-radiative resonance energy is transferred from the emitting Rluc to YFP which becomes excited and in turn emits fluorescence. A BRET ratio is calculated for each condition (see Methods). To perform BRET saturation experiments, we transfected 293T cells with constant amounts of pCMV-Stau1⁵⁵-Rluc plasmid and increasing amount of different NC-p1-YFP expressors. BRET assays were performed 48 hours post-transfection (Figures 2A, B). BRET saturation experiments allowed us to compare BRET ratios at the same relative ratio between fusion proteins (comparable total YFP/Rluc ratio) (Figure 2B). As expected, we readily detected a specific BRET between wild type NC-p1-YFP and Stau1⁵⁵-Rluc (arbitrarily set to 100% in Figure 2B) as compared to co-expression of Stau1⁵⁵-Rluc and YFP alone (Figures 2A, B). Mutations that modify the NC N-terminal basic region did not affect the binding of NC to Stau1 since the saturation profile for Stau1/NC-p1^R7-YFP BRET was almost identical to the one obtained with Stau1/NC-p1-YFP (Figures 2A, B). In contrast, when the two zinc fingers were mutated (NC-p1^C15–49S-YFP), the BRET saturation profile was comparable to that obtained with YFP alone and hence, mostly attributable to background (Figure 2A). When compared to NC-p1-YFP at the same total YFP/Rluc ratio, the corrected BRET ratio was decreased by 80% (Figure 2B). This suggests that NC-p1^C15–49S-YFP lost almost completely its ability to interact with Stau1. Mutations in individual zinc finger (NC-p1^C15S-YFP and NC-p1^C49S-YFP) only affected the BRET ratio by 30–40% and these two mutants showed an intermediate profile (Figures 2A, B).

This suggests that the integrity of at least one NC zinc fingers is required for Stau1/NC interaction.

We used a second technique to confirm the involvement of both zinc fingers in Stau1-NC interaction in the context of full-length Pr55^Gag. We generated a Pr55^Gag-YFP-expressing plasmid in which both zinc fingers were mutated (Pr55^{Gag C15–49S}-YFP)(see below). As shown in Figure 2C, this mutant was expressed to the same level as the wild-type Pr55^Gag-YFP and migrated in SDS-containing acrylamide gels at the expected molecular weight (80 kDa). Following co-expression of Flag-tagged Stau1⁵⁵ with wild type or mutated Pr55^Gag-YFP in 293T cells (Figure 2D, upper panel), Stau1⁵⁵-Flag-containing complexes were immunoprecipitated using anti-Flag antibodies. Immunopurified material was analyzed by Western blot using monoclonal anti-GFP and anti-Flag antibodies (Figure 2D, lower panel). As expected, Pr55^Gag-YFP successfully co-precipitated with Stau1⁵⁵-Flag. In contrast, despite similar levels of expression in the cell (Figure 2D, upper panel), the Pr55^{Gag C15–49S}-YFP mutant was not efficiently co-immunoprecipitated with Stau1⁵⁵-Flag as compared to wild type (Figure 2D, lower panel) suggesting that the association between this mutant and Stau1⁵⁵ is impaired. Pr55^{Gag C15S}-YFP and Pr55^{Gag C49S}-YFP mutants retained some association with Stau1⁵⁵-Flag although they displayed lower binding capability than the wild type Pr55^Gag (not shown), consistent with the BRET assay. Altogether, these results show that the two zinc fingers within the NC domain of Pr55^Gag mediate its association with Stau1. Moreover, this suggests that Stau1 influences those assembly processes that depend on NC zinc fingers.

The fact that Stau1 influences HIV-1 Pr55^Gag multimerization and associates with NC zinc fingers is consistent with previous reports showing that these structural motifs are important in HIV-1 assembly 293046. To confirm this hypothesis in a system that tests direct interaction, we evaluated the consequence of mutations in the Pr55^Gag zinc fingers on VLP release and on Pr55^Gag dimerization using the BRET assay. 293T cells were transfected with Pr55^Gag-YFP and Pr55^{Gag C15–49S}-YFP expressors (Figure 3A). Twenty-four hours post-transfection, VLPs were collected from the supernatant and cells were collected. In the cell extracts, Pr55^Gag-YFP and Pr55^{Gag C15–49S}-YFP were present at similar levels (Figure 3B, left panel). In contrast, the release of Pr55^{Gag C15–49S}-YFP in the cell supernatant was reduced by 95.1% (+/- 3.4 S.D.; n = 3) as compared to that of Pr55^Gag-YFP (Figure 3B, right panel) suggesting that this mutant failed to efficiently assemble. Used as a negative control, MA-CA^{WM184–185AA}-YFP (Figure 3A), a Pr55^Gag mutant that was shown to be almost completely monomeric in the cell and unable to generate VLPs 2651, was not detected in the cell supernatant although it was expressed at higher levels than Pr55^Gag-YFP and Pr55^{Gag C15–49S}-YFP in the cell (Figure 3B).

Then, we determined whether mutations in the zinc fingers affect Pr55^Gagmultimerization. Using the BRET assay in live cells, we tested the capacity of Pr55^{Gag C15–49S}-YFP to dimerize with Pr55^Gag-Rluc, the wild-type Pr55^Gag-YFP being used as control (Figure 3A). As shown in Figure 3C, Pr55^Gag homo-dimerization was readily detectable with a BRET ratio of 0.09 at saturation. In contrast, Pr55^{Gag C15–49S}-YFP failed to interact with Pr55^Gag-Rluc in the BRET assay since its saturation curve was similar to the one obtained with the monomeric Gag mutant MA-CA^{WM184–185AA}-YFP. Altogether, these results clearly show that, in the context of VLP assembly, Pr55^Gag zinc fingers are important for multimerization and release. This suggests that Stau1, through its binding to the NC zinc fingers could influence crucial processes that are controlled by these motifs during HIV-1 assembly.

We previously showed that Stau1 over-expression or depletion from cells enhanced Pr55^Gag multimerization. To determine if the binding of Stau1 to NC is sufficient for Pr55^Gag multimerization or whether other determinants within Stau1 are required for this process, we tested several Stau1 deletion mutants for their capacity to enhance assembly (Figure 4A). To this end, we used the previously described Pr55^Gag/Pr55^Gag BRET assay in live 293T cells as a sensor for changes in Pr55^Gag multimerization (Figure 4B)51. Indeed, Rluc- and YFP-fused Pr55^Gag co-expression generates a positive BRET ratio in live cells as a consequence of Pr55^Gag multimerization. In order to compare BRET ratio changes at the same relative levels of Pr55^Gag fusion proteins, we performed BRET saturation experiments. As previously reported, when Stau1⁵⁵-HA₃ was co-expressed with Pr55^Gag-Rluc and Pr55^Gag-YFP in 293T cells, the Pr55^Gag/Pr55^Gag BRET ratio increased as a consequence of enhanced Pr55^Gag multimerization (Figures 4B)51. Several HA-tagged Stau1 deletion mutants were then tested for their capacity to enhance Pr55^Gag/Pr55^Gag BRET ratio and hence, Pr55^Gag multimerization. Interestingly, Stau1^ΔNt88-HA₃, a mutant that lacks the dsRBD2 as a consequence of the deletion of the first N-terminal 88 amino acids (Figure 4A) was unable to significantly increase the Pr55^Gag/Pr55^Gag BRET ratio in live cells [1.29 (+/-0.13 S.D. n = 4)-fold induction] as compared to Stau1⁵⁵-HA₃ [2.04 (+/-0.09 S.D. n = 4)-fold induction] (Figures 4B, D). Western blot analyses showed that Stau1^ΔNt88 was expressed at levels comparable to that of wild type Stau1-HA₃ (Figure 4C). Nevertheless, a moderate increase in Pr55^Gag multimerization was seen when Stau1^ΔNt88 was highly over-expressed although its effect on Pr55^Gag multimerization was always weaker than that obtained with Stau1⁵⁵-HA₃ (see below). In contrast, mutants with deletion in dsRBD4, dsRBD5 or tubulin-binding domain (TBD) all enhanced the Pr55^Gag/Pr55^Gag BRET ratio at levels comparable to that obtained with Stau1⁵⁵-HA₃ (data not shown). As control, Stau1^F135A-HA₃, a Stau1 mutant that does not bind Pr55^Gag, failed to stimulate Pr55^Gag multimerization (data not shown)51. Therefore, the Stau1-mediated enhancement of Pr55^Gag multimerization requires two determinants: dsRBD3 for the association with NC and the N-terminus.

To test the ability of Stau1^ΔNt88 to interact with Pr55^Gag, we performed BRET assays between Stau1^ΔNt88-YFP and a truncated Pr55^Gag (CA-p2-NC-p1-Rluc) that was previously shown by the co-immunoprecipitaton assay to interact as efficiently with Stau1 as full-length Pr55^Gag 52 and (Figure 5A). To verify efficiency between these molecules, Stau1^ΔNt88-YFP and Stau1⁵⁵-YFP were expressed (Figure 5B) and the BRET saturation profiles determined (Figure 5C). Curves obtained with Stau1⁵⁵-YFP and with Stau1^ΔNt88-YFP were almost similar suggesting that Stau1^ΔNt88mutant retains its capacity to bind to Pr55^Gag. The BRET ratios were specific since the Gag-binding deficient mutant Stau1^ΔdsRBD3-YFP showed reduced BRET ratios. In independent saturation experiments (Figure 5D), the specific BRET ratio following co-expression of Stau1^ΔNt88-YFP and CA-p2-NC-p1-Rluc was comparable [105.7 (+/- 18.1 S.D.)% of CA-p2-NC-p1/Stau1⁵⁵ corrected BRET ratio] to that obtained with wild type Stau1⁵⁵-YFP at similar total YFP/Rluc ratio. We could not detect a specific BRET signal when Stau1^F135A-YFP 52 was used [1.3 (+/- 22.1 S.D.)% of CA-p2-NC-p1/Stau1⁵⁵ corrected BRET ratio].

The ability of Stau1^ΔNt88 to associate with Pr55^Gag was confirmed in co-immunoprecipitation assays (Figure 5E). Pr55^Gag and flag-tagged Stau1 or Stau1^ΔNt88 were co-expressed in 293T cells (Figure 5E, left panel) and proteins in the cell extracts were immunoprecipitated using anti-flag antibody. Western blot analyses of the immune complexes showed that Pr55^Gag successfully co-precipitated in a specific manner with both Stau1⁵⁵-flag and Stau1^ΔNt88-FLAG (Figure 5E, right panel). These results show that, although Stau1^ΔNt88 is unable to stimulate Pr55^Gag multimerization, its interaction with Pr55^Gag was maintained. This result suggests that Stau1 association to Pr55^Gag is not sufficient to influence HIV-1 assembly and that Stau1 N-terminus contains a regulatory element that is important for its function during this process.

We previously showed that the Stau1-mediated enhancement of Pr55^Gag multimerization occurs in membrane compartments 51. Therefore, to test whether Stau1^ΔNt88-HA₃ reaches the membranes and whether the whole cell analysis described above masked an effect of Stau1^ΔNt88-HA₃ on assembly, membrane-associated virus assembly was analyzed in the context of Stau1^ΔNt88-HA₃ or Stau1⁵⁵-HA₃ over-expression. Cytoplasmic extracts from transfected 293T cells were analyzed by the membrane flotation assay (Figure 6A)51. This assay allows the separation of membrane-associated complexes (fraction 2; M) from the cytosolic ones (fractions 7, 8 and 9; Cy). First, Western blot analysis indicated that Stau1^ΔNt88-HA₃ was both over-expressed and present in membranes at the same levels as Stau1⁵⁵-HA₃ (data not shown). As previously described 51, Pr55^Gag/Pr55^Gag BRET was readily detected in the membrane fraction (BRET ratio of 0.33) but not in the cytosolic fractions consistent with the fact that HIV-1 assembly occurs on cellular membranes (Figure 6B)106162. Moreover, as reported before, Stau1⁵⁵-HA₃ over-expression led to an increase of 1.6-fold in the Pr55^Gag/Pr55^Gag BRET ratio in the membrane fraction but not in the cytosolic fraction 51. In contrast, there was little change in the ability of Pr55^Gag to multimerize in the membrane fraction when the mutant Stau1^ΔNt88-HA₃ was over-expressed.

The inability of Stau1^ΔNt88-HA₃ to modulate Pr55^Gag multimerization was then tested in the context of provirus-driven immature HIV-1 production. We had previously shown that Stau1-mediated increase in Pr55^Gag multimerization correlated with a partial resistance to mild detergent treatment of membrane-associated Pr55^Gag complexes 51. Stau1 expressors and the protease-defective provirus HxBRU PR-were cotransfected in 293T cells. Forty hours post-transfection, membrane flotation assays on cytoplasmic extracts were performed and fractions containing membrane-associated complexes were collected. The resulting complexes were then subjected to ultracentrifugation at 100,000 × g for 1 hour (Figure 6C). In this assay, insoluble or high-density complexes are found in the pellet (P100) whereas proteins that are soluble or components of small complexes are retained in the supernatant (S100). Resulting P100 and S100 fractions were analyzed by Western blot. As previously shown 51, Pr55^Gag as well as the membrane marker, sodium potassium (Na-K) ATPase, were primarily found in the P100 fraction because Pr55^Gag was membrane-associated (data not shown). This was observed whether Stau1 proteins were over-expressed or not (data not shown). To separate Pr55^Gag complexes from membranes, membranes were solubilized with 0.5% Triton-X100. These experiments were done at room temperature to also solubilize lipid rafts and other membrane compartments that are detergent-resistant at 4°C. As reported before, membrane solubilization prior to S100/P100 fractionation resulted in a complete shift of Pr55^Gagcomplexes and Na-K ATPase into the S100 fraction (Figure 6D)51. Stau1⁵⁵-HA₃ overexpression led to a partial resistance of 33% of Pr55^Gag complexes to Triton-X100 treatment, likely as a consequence of enhanced Pr55^Gag multimerization (Figure 6D). In contrast, Stau1^ΔNt88-HA₃, as well as Stau1^F135A-HA₃ used as control 5152 failed to increase the density of Pr55^Gag complexes. Altogether, these results support the conclusion that the N-terminus of Stau1 is required for Stau1-mediated increase of Pr55^Gag assembly during HIV-1 assembly.

To more precisely map the functional determinant within the N-terminal region that is involved in the regulation of Pr55^Gag multimerization, we generated two additional deletion mutants, Stau1^ΔNt25-HA₃ and Stau1^ΔNt37-HA₃ and confirmed their expression (Figure 7A). Using the Pr55^Gag/Pr55^Gag BRET assay in live 293T cells, Stau1^ΔNt25-HA₃ enhanced the BRET ratio like wild type Stau1⁵⁵-HA₃ suggesting that the first 25 amino acids are not required for this function (Figure 7B). In contrast, Stau1^ΔNt37-HA₃ did not significantly enhance the BRET ratio and generated a BRET curve similar to that obtained with Stau1^ΔNt88-HA₃. Expression of increasing amounts of Stau1⁵⁵-HA₃ or Stau1^ΔNt25-HA₃ (Figure 7C) led to a comparable increases of Pr55^Gag/Pr55^Gag BRET ratio up to 2.16–2.34-fold (Figure 7D). In contrast, expression of Stau1^ΔNt37-HA₃ or Stau1^ΔNt88-HA₃ had no effect on Pr55^Gag/Pr55^Gag BRET ratio when their expression is relatively low. However, when highly expressed, they slightly enhanced the BRET ratio by 1.34–1.47-fold. These results suggest that the sequence located between amino acids 26 and 37 is important for Stau1-mediated enhancement of Pr55^Gag multimerization and show that Stau1 acts on Pr55^Gag multimerization in a dose-dependent manner.

We already showed that Stau1 over-expression, likely as a consequence of its role in multimerization, also increases VLP release from the cell 51. Therefore, considering the inability of Stau1^ΔNt88 to influence Pr55^Gag multimerization (Figure 4), we suspected that over-expression of Stau1^ΔNt88 would not stimulate VLP production. To test this idea, Pr55^Gag and either Stau1⁵⁵-HA₃ or Stau1^ΔNt88-HA₃ were co-expressed in 293T cells. Twenty-four hours post-transfection, the levels of VLP release were analyzed by Western blotting (Figure 8A). As expected, a 2-fold increase in the expression of Stau1⁵⁵-HA₃ as compared to endogenous Stau1 resulted in a 2.5-fold increase in VLP release while the cellular level of Pr55^Gag remained unchanged. In contrast, over-expression of Stau1^ΔNt88-HA₃ did not stimulate VLP production. To confirm these results, dose-response assays were performed. Constant amounts of Pr55^Gag were co-expressed with increasing amounts of either Stau1⁵⁵-HA₃ or Stau1^ΔNt88-HA₃ (Figure 8B). As previously shown 51, Stau1⁵⁵-HA₃ over-expression stimulated VLP production in a dose-dependent manner up to 10-fold with no significant change in the intracellular level of Pr55^Gag. In contrast, Stau1^ΔNt88-HA over-expression did not lead to a significant increase of VLP release, except at high expression levels (Figure 8C). However, the stimulation of VLP production by Stau1^ΔNt88-HA₃ was always less than that produced by Stau1⁵⁵-HA₃, when expressed at the same level. Analyses of VLP production in the presence of Stau1^ΔNt25-HA and Stau1^ΔNt37-HA indicated that, whereas Stau1^ΔNt25-HA enhanced VLP release, Stau1^ΔNt37-HA did not (Figure 8D) consistent with the conclusion that Stau1 molecular determinant for enhanced Pr55^Gag multimerization is carried by amino acids 26 to 37.