ABSTRACT
HIV-1 Nef promotes virus spread and disease progression by altering host cell transport and signaling processes through interaction with multiple host cell proteins. The N-terminal region in HIV-1 Nef encompassing residues 12 to 39 has been implicated in many Nef activities, including disruption of CD4 T lymphocyte polarization and homing to lymph nodes, antagonism of SERINC5 restriction to virion infectivity, downregulation of cell surface CD4 and major histocompatibility complex class I (MHC-I), release of Nef-containing extracellular vesicles, and phosphorylation of Nef by recruitment of the Nef-associated kinase complex (NAKC). How this region mediates these pleiotropic functions is unclear. Characterization of a panel of alanine mutants spanning the N-terminal region to identify specific functional determinants revealed this region to be dispensable for effects of Nef from HIV-1 strain SF2 (HIV-1SF2Nef) on T cell actin organization and chemotaxis, retargeting of the host cell kinase Lck to the trans-Golgi network, and incorporation of Nef into extracellular vesicles. MHC-I downmodulation was specific to residue M20, and inhibition of T cell polarization by Nef required the integrity of the entire region. In contrast, downmodulation of cell surface CD4 and SERINC5 antagonism were mediated by a specific motif encompassing residues 32 to 39 that was also essential for efficient HIV replication in primary CD4 T lymphocytes. Finally, Nef phosphorylation via association with the NAKC was mediated by two EP repeats within residues 24 to 29 but was dispensable for other functions. These results identify the N-terminal region as a multifunctional interaction module for at least three different host cell ligands that mediate independent functions of HIV-1SF2Nef to facilitate immune evasion and virus spread.
IMPORTANCE HIV-1 Nef critically determines virus spread and disease progression in infected individuals by acting as a protein interaction adaptor via incompletely defined mechanisms and ligands. Residues 12 to 39 near the N terminus of Nef have been described as an interaction platform for the Nef-associated kinase complex (NAKC) and were recently identified as essential determinants for a broad range of Nef activities. Here, we report a systematic mapping of this amino acid stretch that revealed the presence of three independent interaction motifs with specific ligands and activities. While downmodulation of cell surface MHC-I depends on M20, two EP repeats are the minimal binding site for the NAKC, and residues 32 to 39 mediate antagonism of the host cell restriction factor SERINC5 as well as downmodulation of cell surface CD4. These results reveal that the N-terminal region of HIV-1SF2Nef is a versatile and multifunctional protein interaction module that exerts essential functions of the pathogenicity factor via independent mechanisms.
INTRODUCTION
Nef is a myristoylated 25- to 34-kDa protein encoded by the lentiviruses human immunodeficiency virus type (HIV-1), HIV-2, and simian immunodeficiency virus (SIV). While dispensable for virus spread in cultured cells, Nef potently increases virus replication and thus serves as a pathogenicity factor that accelerates disease progression in the infected host (1–3). Nef lacks enzymatic activity but mediates its functions through a large set of interactions with cellular proteins. By virtue of this adaptor function, Nef affects many central processes in HIV target cells. This encompasses modulation of cellular transport pathways leading to downregulation of an array of receptors from the surface of infected cells (4–6). These include the entry receptor CD4 and its chemokine coreceptors, which leads to lower rates of superinfection (7–9). In addition, Nef alters immune recognition of infected cells by lowering cell surface levels of major histocompatibility complex class I (MHC-I) molecules and natural killer cell ligands to reduce lysis by cytotoxic T or NK cells, respectively (10–13). In addition to cell surface receptors, Nef also modulates vesicular transport of peripheral membrane proteins such as the Src kinase Lck, which is retargeted from the plasma membrane to the trans-Golgi network (TGN) by Nef-mediated inhibition of its anterograde transport toward the cell surface (14–17). Together with the ability to inhibit stimulus-induced actin reorganization and chemotaxis (18–21), Nef also alters the response of infected CD4 T lymphocytes to stimulation via the T cell receptor (TCR) to increase virus replication in the infected host (15, 16, 22–26). Finally, Nef also enhances the infectivity of cell-free virus particles (10, 27, 28). This effect is most pronounced when virus particles are produced from cells expressing host cell restriction factors of the serine incorporator (SERINC) family and, in particular, SERINC5, which potently reduces virion infectivity and is counteracted by Nef (29–31).
A steadily increasing amount of interactions with host cell ligands is being identified for Nef, and the sum of these interactions is thought to explain the multitude of functions of the versatile pathogenicity factor Nef (32, 33). For example, interactions with endocytic sorting motifs in a C-terminal flexible loop allow Nef to engage the cellular trafficking machinery, such as the AP-2 adaptor complex, and to relocalize cargo in infected cells (34–36). Alternatively, as in the case of MHC-I and Lck, Nef alters host cell trafficking via its SH3 binding motif in its folded core domain (14, 37). Finally, effects of Nef on host cell actin dynamics depend on its association with the cellular p21-activated kinase (PAK2) as well as the exocyst complex via interactions of the Src homology 3 (SH3) domain binding motif and a hydrophobic patch involving a critical phenylalanine in the Nef C terminus (20, 38–42). All these interactions are conserved in lentiviral evolution and occur in relevant infected target cells. This, however, does not entirely account for the plethora of Nef activities.
In this context, another protein interaction surface at the N terminus of HIV-1 Nef was recently implicated in a wide range of Nef activities. Residues 12 to 39, located in the largely unfolded anchor domain of HIV-1 Nef, were initially described to recruit the Nef-associated kinase complex (NAKC) to mediate phosphorylation of Nef at serine 6 (43). Subsequent studies identified protein kinase C θ (PKC-θ) and PKC-δ as the kinases responsible for Nef phosphorylation; defined Lck, polycomb protein EED, and hnRNPK as additional NAKC components; and suggested that Nef, via the formation of the NAKC, facilitates HIV replication in primary target cells (44–47). In addition, the NAKC was shown to enhance the secretion of extracellular vesicles (EVs) containing Nef but also proinflammatory cytokines such as tumor necrosis factor alpha (TNF-α) (48–50). In addition, deletions of this protein interaction site in HIV-1 Nef were shown to prevent downregulation of cell surface CD4 and MHC-I, antagonism of SERINC5, as well as disruption of CD4 T cell polarity and homing to lymph nodes (51–53). Beyond the demonstration of its overall relevance, which is based on a Nef mutant in which the entire stretch of residues from positions 12 to 39 are deleted (Δ12-39), this multifunctional N-terminal interaction site in HIV-1 Nef is less characterized, and we therefore set forth to define specific residues that mediate different activities of this region. Our results reveal that this region of HIV-1 Nef serves as a multifunctional adaptor that contains at least three independent motifs that govern specific Nef effector functions.
RESULTS
MHC-I downmodulation by HIV-1SF2Nef depends on M20 but not the entire N-terminal interaction site.With downmodulation of cell surface CD4 and MHC-I, antagonism of SERINC5, disruption of T cell polarity and lymph node homing, as well as Nef phosphorylation and EV production, the N-terminal interaction site (residues 12 to 39) (Fig. 1A) in HIV-1 Nef was identified to be involved in a wide range of different activities. Since these findings were made by using a Nef mutant carrying a deletion encompassing all residues from positions 12 to 39, we wondered if the relevance of this stretch of amino acids reflects one general or several independent protein interactions of Nef. Among the Nef activities that depend on this region, a specific molecular determinant was defined previously for MHC-I downregulation: methionine 20 (M20) is required for the recruitment of AP-1 and MHC-I for the reduction of MHC-I density at the cell surface (54–56). We therefore tested if the role of M20 is specific for MHC-I downregulation by Nef from HIV-1 strain SF2 (HIV-1SF2Nef) or is also involved in other effector functions of the N-terminal interaction site. Since ectopic expression of green fluorescent protein (GFP) fused with Nef (Nef.GFP) in human cell lines provides a good correlate for the activity of nonfusion Nef in HIV-1-infected primary cells (21, 51), functional characterization was performed using cells transiently expressing GFP or Nef.GFP. In a first series of experiments, wild-type (WT) Nef (from HIV-1 strain SF2) was compared to a mutant lacking the entire N-terminal interaction site (Δ12-39) and a mutant in which M20 was replaced by alanine (M20A) (Fig. 1B). Immunoisolation of these proteins via anti-GFP beads and a subsequent in vitro kinase assay (IVKA) revealed robust phosphorylation of the WT (Fig. 1C, top). This Nef phosphorylation was markedly reduced for the Δ12-39 mutant (see Fig. 1D for quantification), reflecting the lack of interaction with the NAKC containing the Nef-phosphorylating kinase PKC-θ (45). The extent of phosphorylation of M20A was comparable to that of the WT, indicating that the M20 residue is dispensable for interactions of Nef with the NAKC. Expectedly (51, 54), MHC-I cell surface levels were significantly less reduced by Δ12-39 or M20A than by the WT in A3.01 T lymphocytes (Fig. 1E and F). In contrast, Nef-mediated downregulation of cell surface CD4 was impaired by the deletion of residues 12 to 39 but not the M20A exchange (Fig. 1G and H), a result that matches the ability of these Nef mutants to antagonize virion infectivity restriction by SERINC5 (52). Finally, the M20A mutant disrupted CD4 T cell polarity as efficiently as the WT, while deletion of the region (Δ12-39) abrogated this Nef activity (Fig. 1I and J [asterisks indicate GFP-positive cells]). Thus, within the N-terminal interaction module, the M20 residue is a specific determinant for MHC-I downregulation by Nef that is not involved in other described effector functions of this protein interaction module.
MHC-I downmodulation by HIV-1 Nef depends on M20 but not the entire N-terminal interaction region. (A) Schematic overview of functional motifs in HIV-1SF2Nef. Residues with previously known host cell interaction partners are highlighted, with a focus on the N-terminal interaction site (residues 12 to 39). (B) Schematic representation of the Δ12-39 and M20A Nef.GFP mutants analyzed. eGFP, enhanced green fluorescent protein. (C and D) Analysis of Nef phosphorylation by the NAKC. COS7 cells transiently expressing the indicated Nef.GFP constructs or a GFP control together with PKC-θ.HA were lysed, followed by anti-GFP immunoprecipitation (IP) and a subsequent in vitro kinase assay (IVKA) with [γ-32P]ATP. (C) Samples were subjected to SDS-PAGE and Western blotting (WB) and then analyzed by autoradiography (top) or antibody detection (bottom). (D) Densitometric quantification of p-Nef levels relative to total amounts of GFP/Nef.GFP present in immunoprecipitates, normalized to the value for WT Nef.GFP, which was arbitrarily set to 1 (means ± standard deviations [SD] from three or more independent experiments). (E to H) Downregulation of cell surface MHC-I (E and F) or CD4 (G and H). A3.01 T lymphocytes transiently expressing the indicated GFP or Nef.GFP constructs were harvested at 24 h posttransfection, stained for surface receptor expression with allophycocyanin (APC)-conjugated antibodies against HLA-ABC or CD4, and analyzed by flow cytometry. (E and G) Representative flow cytometry dot plots and median fluorescence intensity (MFI) quantification. (F and H) Total cell surface expression calculated as the ratio of the MFI of medium- to high-GFP-expressing cells (right gate) to the MFI of untransfected cells (left gate). Shown are mean values ± SD relative to the value for GFP, which was arbitrarily set to 100%, from three independent experiments. (I and J) T cell polarization. (I) Representative confocal images of A3.01 T cells transiently expressing the indicated GFP/Nef.GFP constructs. Cells were seeded onto fibronectin-coated coverslips at 24 h posttransfection and allowed to polarize for 2 h at 37°C. Cells were then fixed with 8% paraformaldehyde, permeabilized with 0.1% Triton X-100, and stained with an antibody against the T cell polarization marker CD44. Cell polarity was scored by observed asymmetry and accumulation of CD44 in uropods (indicated by arrows) in GFP-positive cells (indicated by *). (J) Polarization frequency of the cells analyzed in panel I (mean values ± SD from at least three independent experiments with at least 100 cells counted per condition and experiment). Statistical significance was calculated using one-way ANOVA (n.s., nonsignificant; **, P < 0.01; ***, P < 0.001). Scale bar, 10 μm.
The N-terminal interaction site is dispensable for perinuclear membrane retargeting of Lck and PAK2-dependent inhibition of actin remodeling.We next sought to assess the relevance of this interaction site in Nef activities, such as vesicular trafficking of the Src family kinase Lck and remodeling of host cell actin cytoskeleton (14, 15, 20). To this end, we analyzed the effect of Nef on the subcellular localization of endogenous Lck in Jurkat T antigen (TAg) T lymphocytes (Fig. 2A and B). Lck was distributed between the plasma membrane, a soluble cytoplasmic pool, and a slight enrichment at an intracellular, perinuclear membrane compartment in GFP-expressing control cells (15.0% ± 5.7% of cells with a predominant perinuclear Lck localization), while the majority of cells expressing WT Nef.GFP (75.3% ± 7.1% of cells) displayed a strong enrichment in perinuclear Lck, which reflects TGN retargeting of the kinase by Nef (14, 15). While disruption of the SH3 domain binding motif in Nef (AxxA; P76,79A) abrogated this effect, deletion of the N-terminal interaction site did not impair the membrane-retargeting activity of Nef (14.3% ± 4.1% and 56.4% ± 16.7% of cells with pronounced perinuclear Lck localization, respectively). Similar results were obtained when staining for phosphorylation of the actin-severing factor cofilin at position 6, which is increased in Nef-expressing cells as a consequence of functional Nef-PAK2 association (20, 41, 57). In this assay, the enrichment of phosphorylated inactive cofilin is indicated by high-intensity fluorescence that is homogenously distributed in all cellular compartments and thus is distinguishable from cells with basal background levels. Expectedly (20, 57), one-quarter of GFP-expressing control cells (24.0% ± 6.2%) spontaneously displayed such high levels of phosphorylated cofilin (p-cofilin), and expression of the WT markedly increased this (70.7% ± 5.4%) (Fig. 2C and D). Induction of cofilin phosphorylation was prevented by disruption of Nef’s association with PAK2 (F195A), but deletion of the N-terminal region (Δ12-39) did not significantly impair this Nef activity (34.3% ± 8.3% and 57.6% ± 1.5% of cells with elevated p-cofilin levels, respectively). We next analyzed the ability of Nef to impair chemotaxis of Jurkat (CCR7) T lymphocytes toward stromal cell-derived factor 1alpha (SDF-1α) using a well-established transwell migration assay (18–20). Analysis of cells expressing high levels of Nef (high-GFP-expressing cells) reveals effects of Nef as they are observed in HIV-1-infected cells (18, 21, 58, 59). Cells expressing high levels of WT and Δ12-39 Nef showed a strong inhibition of T cell chemotaxis (12.2% ± 3.9% and 18.7% ± 10.2% of migrating cells, respectively), while the F195A mutation significantly reduced the ability of Nef to interfere with this process, as reported previously (Fig. 2E and F) (20). Analysis of all GFP-positive cells yields similar results but with a reduced magnitude of Nef-mediated inhibition (data not shown). Together, these results reveal that the multifunctional N-terminal region is dispensable for effects of Nef on the intracellular transport of Lck as well as for effects on host actin organization and migration that are mediated by the association of Nef with PAK2.
The N-terminal interaction site is dispensable for membrane retargeting of Lck and PAK2-dependent inhibition of actin remodeling. (A and B) Nef-mediated retargeting of Lck. (A) Representative confocal images showing the subcellular localization of endogenous Lck in Jurkat (TAg) T lymphocytes transiently expressing GFP/Nef.GFP constructs. Cells were seeded onto poly-l-lysine-coated coverslips at 24 h posttransfection, fixed, permeabilized, and stained with primary anti-Lck antibody, followed by Alexa Fluor 568 secondary antibody. GFP-positive cells are indicated by an asterisk. Bar, 10 μm. (B) Frequency of GFP-expressing cells with perinuclear Lck accumulation (means ± SD from three independent experiments with at least 100 cells counted per condition and experiment). (C and D) Nef-mediated accumulation of phosphorylated cofilin (p-cofilin). (C) Representative confocal images of Jurkat (CCR7) T lymphocytes transiently expressing the indicated GFP/Nef.GFP constructs. At 24 h posttransfection, cells were fixed, permeabilized, and stained with primary anti-p-cofilin antibody, followed by a secondary Alexa Fluor 568 antibody. GFP-expressing cells are indicated by an arrow. Scale bar, 10 μm. (D) Frequency of GFP-positive cells expressing high p-cofilin levels. Values shown are means ± SD from three independent experiments, with at least 100 cells counted per condition and experiment. (E and F) Chemotaxis toward SDF-1α. (E) Representative dot plots showing migration frequencies of Jurkat (CCR7) T lymphocytes transiently expressing the indicated GFP/Nef.GFP constructs. Cells were harvested at 24 h posttransfection, starved for 2 h in 0.5% FCS-containing medium, and seeded in the upper chamber of a transwell system (5-μm pore size), with medium containing 10 ng/ml SDF-1α in the lower chamber. The percentages of GFP-expressing cells (gating) in the input and after 2 h of migration were quantified by flow cytometry. (F) Quantification of the percentage of migrated cells, calculated as the ratio between the percentages of GFP-expressing cells before and after migration. Biological duplicates were analyzed per experiment. Shown are means ± standard errors of the means (SEM) from three independent experiments plotted relative to the value for GFP, which was set to 100%. Statistical significance was calculated using one-way ANOVA (n.s., nonsignificant; **, P < 0.01; ***, P < 0.001).
Alanine replacements within the N-terminal interaction site do not alter expression levels or subcellular localization of Nef.For a more detailed dissection of molecular determinants within the multifunctional N-terminal interaction site, we generated a panel of mutant Nef.GFP expression constructs in which two or three consecutive amino acid residues were replaced by alanine within residues 12 to 39 of HIV-1SF2Nef (Fig. 3A). All mutant Nef.GFP constructs were expressed as full-length proteins to comparable levels, except for Δ12-39, which displayed lower levels of expression, presumably due to reduced rates of synthesis and/or stability of the protein (Fig. 3B). We also analyzed the overall subcellular localization of the mutant Nef.GFP constructs to test whether any of the mutations caused a marked redistribution of the viral protein. All Nef proteins showed a distribution characterized by a strong enrichment in the perinuclear region, which includes the trans-Golgi network, and additional pools of protein localized to the inner leaflet of the plasma membrane and diffusely to the cytoplasm and thus matched the typical distribution of HIV-1 Nef in CD4 T cells (15, 16, 37) (Fig. 3C). This initial characterization of the alanine replacements reveals that the protein expression and subcellular localization of these mutant Nef proteins are not exceedingly altered in comparison to the WT, even though subtle differences in distributions are not revealed, and are suitable for further functional characterization.
Alanine replacements within the N-terminal interaction site do not alter protein expression levels or subcellular localization of Nef. (A) Schematic representation of the panel of alanine mutants generated within the N-terminal interaction site of HIV-1SF2Nef. (B) Protein expression in Jurkat (TAg) lymphocytes transiently expressing the GFP control or the indicated Nef.GFP constructs. Cells were harvested at 24 h posttransfection in lysis buffer and lysed by ultrasonication. Samples were then subjected to SDS-PAGE, Western blotting, and antibody detection using anti-GFP and anti-GAPDH. (C) Subcellular localization of alanine mutants. Shown are confocal images of the GFP control or the indicated Nef.GFP constructs expressed in Jurkat (TAg) cells. Cells were seeded onto poly-l-lysine-coated coverslips, fixed, and mounted for immunofluorescence analysis. Scale bar, 10 μm.
Residues 32 to 39 are critical for Nef-mediated CD4 downmodulation and SERINC5 antagonism.When analyzed for the ability to downmodulate cell surface CD4, most alanine mutants displayed an activity similar to that of WT Nef (data not shown) (Table 1), including the Δ16-22 deletion mutant that was used in the original study to disrupt Nef-NAKC interactions (43) (Fig. 4A and B). Replacing residues 32 and 33 (A32,33), 34 and 35 (A34,35), or 37 to 39 (A37-39) at the C terminus of this region, however, either partially or completely abrogated Nef’s ability to downmodulate cell surface CD4 (Fig. 4A and B). Since the molecular determinants in Nef for CD4 downmodulation and antagonism of the virion infectivity restriction by SERINC5 largely overlap (29, 30, 52), the panel of alanine replacements were expressed as nonfusion proteins from an HIV-1 provirus in the presence of SERINC5, and the relative infectivity of particles released into the cell culture supernatant was determined in single-round infections of TZM-bl reporter cells. Under these conditions, SERINC5 reduced virion infectivity of an HIV-1 variant lacking Nef expression (ΔNef) by at least 10-fold, and the expression of Nef fully counteracted the negative effect of SERINC5 on virion infectivity (Fig. 4C). While mutations in the first half (N terminus) of the interaction site or the deletion of residues 16 to 22 (Δ16-22) had no effect on Nef counteraction of SERINC5 (Table 1), the deletion of residues 12 to 39 significantly reduced the ability of Nef to counteract SERINC5. Importantly, the extent of impairment in downregulating cell surface CD4 of the alanine replacement mutants in the second half (C terminus) of the interaction site, A32,33, A34,35, or A37-39, was paralleled by a similar reduction in SERINC5 antagonism (Fig. 4C). Consistent with the idea that virion exclusion of SERINC5 is one of the mechanisms by which Nef increases the infectivity of HIV particles produced in the presence of the restriction factor (29, 30, 52), virion-associated SERINC5 was undetectable in particles produced in the presence of WT Nef or antagonism-competent mutant Δ16-22, and the lack of SERINC5 antagonism coincided with increased virion incorporation of SERINC5 (Fig. 4D). These results reveal that the requirement of the N-terminal interaction site for Nef-mediated downregulation of cell surface CD4 and antagonism of SERINC5 reflects the involvement of a shared and specific amino acid motif in the second half of the N-terminal region. Residues 32 to 39 in HIV-1SF2Nef are henceforth referred to as the SERINC5 antagonism motif (S5AM).
Nef activity of HIV-1SF2Nef mutantsa
Residues 32 to 39 are critical for Nef-mediated CD4 downmodulation and SERINC5 antagonism. (A and B) Cell surface expression of CD4. A3.01 T lymphocytes transiently expressing the indicated GFP or Nef.GFP constructs were harvested at 24 h posttransfection, stained for cell surface CD4 with an antibody coupled to APC, and analyzed by flow cytometry. (A) Representative flow cytometry dot plots showing MFI values of untransfected (left gate) and transfected GFP-expressing (right gate) cells. (B) Total cell surface CD4 expression calculated as the ratio of the MFI of GFP-expressing cells to the MFI of untransfected cells. Shown are mean values ± SD relative to the value for GFP, which was arbitrarily set to 100%, from four independent experiments. (C) Normalized infectivity of HIV-1 particles. 293T cells were transfected with pBJ5-SERINC5.intHA or a control vector along with the proviral HIV-1 Nef WT or the indicated mutants. Infectivity was measured by infection of TZM-bl reporter cells with a virus-containing supernatant and is normalized to virus release. Depicted are mean values ± SD from three independent experiments. (D) Representative Western blot analysis of producer cell lysates and virions showing antibody detection of SERINC5.intHA, Nef, and HIV-1 Gag (p24) proteins. Statistical significance was calculated using one-way ANOVA (n.s., nonsignificant; **, P < 0.01; ***, P < 0.001).
The N-terminal region is dispensable for Nef incorporation into extracellular vesicles.We next assessed the role of the NAKC interaction for the incorporation of Nef into extracellular vesicles (EVs). Nef.GFP was expressed in 293T cells (48) alone or together with the NAKC components Lck, PKC-θ, and EED (Fig. 5A), and EVs were purified from the cell culture supernatant (Fig. 5B). While we observed a robust incorporation of WT Nef into EVs, the presence of Nef did not affect the overall amount of EVs produced, as determined by quantification of the total protein content of the EV fraction (Fig. 5C). EV incorporation of Nef was not increased upon coexpression of NAKC components and was unaffected by deletion of the entire N-terminal interaction site (Fig. 5B). Consistent with a report by Ali et al. (60), disruption of the membrane association of Nef by removal of the myristoyl acceptor and basic residues in the Nef N terminus (G2AKR Nef mutant) (61) did not impair EV incorporation of Nef. These results suggest that in 293T cells, the N-terminal interaction site as well as NAKC components are dispensable for the incorporation of Nef into EVs.
NAKC binding is mediated by EP repeats but is dispensable for Nef incorporation into extracellular vesicles. (A to C) Nef incorporation into extracellular vesicles (EVs). (A and B) Western blot analysis of the cell lysate and the purified EV fraction from 293T cells cotransfected with the indicated GFP/Nef.GFP constructs along with or without NAKC factors (EED.HA, Lck.RFP, and PKC-θ.HA). At 24 h posttransfection, cells were harvested for lysis, and the supernatant was used for further centrifugation, precipitation, and purification of exosomes. A total of 200 μg of the cell lysate and 50 μg of the EV protein lysate were loaded per lane for Western blot analysis. Primary antibodies against epitope and fluorescent tags (HA, RFP, and GFP) and the exosome marker (CD63) were used. Shown are representative blots from three independent experiments. (C) Total protein quantification (micrograms per microliter) of EVs produced from either GFP- or WT Nef.GFP-expressing cells in the presence or absence of NAKC factors. Values shown are means ± SD from at least three or more independent experiments. (D to F) Analysis of Nef phosphorylation in COS7 cells transiently expressing the indicated GFP/Nef.GFP constructs. (D and E) Representative autoradiography (top) or antibody detection (bottom) of immunoprecipitates from cells transiently expressing the indicated proteins incubated with [γ-32P]ATP. (F) Densitometric quantification of p-Nef levels relative to total amounts of GFP/Nef.GFP present in immunoprecipitates, normalized to the value for WT Nef.GFP, which was arbitrarily set to 1 (means ± SD from five or more independent experiments). (G) Surface expression of CD4 in A3.01 T lymphocytes expressing the indicated GFP or Nef.GFP constructs relative to value for the GFP control, which was arbitrarily set to 100% (means ± SD from three independent experiments). (H and I) Nef antagonism of SERINC5. (H) Normalized relative infectivity of HIV-1 virions of the indicated variants in the presence or absence of ectopic SERINC5 expression (means ± SD from three independent experiments). (I) Representative Western blot analysis of cell lysates and virions showing antibody detection of SERINC5.intHA, Nef, and HIV-1 Gag (p24) proteins. Statistical significance was calculated using one-way ANOVA (n.s., nonsignificant; *, P < 0.05; ***, P < 0.001).
Interactions of Nef with the NAKC are mediated by an EP repeat.We next sought to identify the precise interaction motif for the NAKC within the N-terminal interaction site and asked if it overlaps the S5AM spanning residues 32 to 39 (Fig. 5D to F). We used COS7 cells since they were shown to best reveal the phosphorylation of Nef at serine 6 (43, 46). Phosphorylation of the A32-39 Nef mutant by the NAKC was variable but overall occurred with an efficiency that was not statistically different from that of the WT, suggesting that the S5AM is dispensable for the recruitment of the NAKC (Fig. 5D and F). Screening of our panel of alanine replacement mutants for phosphorylation by the NAKC revealed a trend toward a reduction in phosphorylation by the NAKC for the Nef mutant A28-29 (not shown), which, however, was not statistically significant (Table 1). Since this region in HIV-1SF2Nef contains two glutamate, proline (EP) repeats that were not mutated simultaneously by these alanine replacements, we created a mutant in which both EP repeats between amino acids (aa) 24 and 29 were replaced by alanine (Nef mutant 2EPAA). 2EPAA was as defective for phosphorylation by the NAKC as the Nef mutant Δ12-39 (Fig. 5E and F). As established above, the S5AM was essential for CD4 downregulation (Fig. 5G), SERINC5 antagonism (Fig. 5H), and virion exclusion of SERINC5 (Fig. 5I). In contrast, the EP repeat was dispensable for all of these activities. Hence, the EP residues at positions 24/25 and 28/29 of HIV-1SF2Nef represent the minimal interaction site for the NAKC. However, the role of the Nef N-terminal interaction domain in CD4 downregulation and SERINC5 antagonism is thus mediated exclusively by the S5AM, and NAKC binding is not required for these activities.
Inhibition of T cell polarization requires the entire N-terminal interaction site of HIV-1SF2Nef.We next tested our alanine replacement mutant panel to define which determinants in the interaction site are involved in the disruption of T cell polarization by Nef. The ability of these Nef mutants to disrupt T cell polarization was not significantly impaired (Fig. 6A and B). In addition, disruption of the independent determinants for interaction with the NAKC and SERINC5 antagonism/CD4 downregulation also did not impair the ability of Nef to interfere with T cell polarization (Fig. 6C; see also Fig. 6D for the polarization frequencies of untransfected neighboring cells). We conclude that the disruption of T cell polarization by Nef relies on the overall integrity of the N-terminal interaction site.
Inhibition of T cell polarization requires the entire N-terminal interaction site of HIV-1SF2Nef. (A) Representative confocal images of A3.01 T cells transiently expressing the indicated GFP/Nef.GFP constructs. Cell polarity was scored by observed asymmetry and accumulation of CD44 in uropods (indicated by arrows) in GFP-expressing cells (indicated by *). Scale bar, 10 μm. (B and C) Polarization frequency of cells expressing the indicated GFP/Nef.GFP constructs. Values shown are means ± SD from three independent experiments with at least 60 cells counted per condition and experiment. (D) Polarization frequency of untransfected cells from the samples shown in panel C. Shown are mean values ± SD from three independent experiments. Statistical significance was calculated using one-way ANOVA (n.s., nonsignificant; **, P < 0.01; ***, P < 0.001).
The S5AM drives HIV-1 replication in primary CD4 T cells.Finally, we sought to determine which motif in the interaction site mediates the reported enhancement of HIV-1 replication in primary human cells (45, 47). To compare the three independent molecular determinants within the N-terminal interaction site, HIV-1 variants encoding WT Nef and the Δ12-39, M20A, 2EPAA, and A32-39 (S5AM) Nef mutants were produced in 293T cells. During virus production, these Nef proteins were expressed to similar levels, with the exception of Δ12-39, which was detected with less efficiency, probably due to reduced recognition by the anti-Nef antibody used and/or reduced expression levels and protein stability (51) (Fig. 7A). These different viruses were produced in comparable amounts of viral particles (quantified by SYBR green I-based product-enhanced reverse transcriptase [SG-PERT] assay) (Fig. 7B) as well as in their relative infectivity (quantified by single-round infection of TZM-bl reporter cells) (Fig. 7C). These stocks were used to infect primary CD4+ T cells that had been activated via CD3/CD28 stimulation 72 h previously at a multiplicity of infection (MOI) of 0.001 (1,000 infectious units/106 cells). Postinfection, cells were kept in the absence of activating stimuli, and the quantity of productively infected cells was determined by intracellular fluorescence-activated cell sorter (FACS) analysis for p24CA. Initial experiments confirmed previous reports (59–62) that under these conditions, WT HIV-1 rapidly spreads in these cultures, with peak numbers of infected cells at days 5 to 7 postinfection, while replication of HIV-1ΔNef was markedly delayed and did not reach high levels of productively infected cells (Fig. 7D). HIV-1 expressing the Nef mutant lacking the interaction site (Δ12-39) was indistinguishable from HIV-1ΔNef, emphasizing the role of this protein interaction module for HIV-1 spread. Analysis of the Nef mutants with a specific disruption of individual motifs within the N-terminal region revealed that the M20A and 2EPAA mutants supported virus replication as efficiently as WT Nef or displayed a slight reduction in peak numbers of infected cells in cells from some but not all donors (Fig. 7E and F). In sharp contrast, replication of the A32-39 Nef mutant virus was as impaired as those of HIV-1ΔNef and HIV-1 encoding the Nef mutant Δ12-39. These results identify the S5AM as a critical determinant of HIV-1 replication in primary CD4+ T cells, while determinants of MHC-I downregulation and NAKC interactions do not have an impact on HIV-1 spread.
HIV-1 replication in primary CD4+ T cells is dependent on S5AM. (A) pNL4.3SF2Nef mutants and ΔNef proviruses were transfected into 293T cells, and the supernatant was harvested after 48 h for virus concentration. Cell lysates of 293T cells show expression of Nef and p24.CA by Western blotting. (B and C) Characterization of viruses used for infection in panels D to F. (B) Quantification of viral particles from the supernatant after concentration on a 20% sucrose solution by SG-PERT, shown as reverse transcriptase (RT) activity in pico units per microliter (pU/μl). (C) Relative infectivity of viral particles displayed as a ratio between relative luciferase units generated by single-round infection on TZM-bl reporter cells and the total number of viral particles. (D) HIV-1 replication in primary activated CD4+ T cells. At 72 h postactivation, cells were infected with ΔNef, WT, and Δ12-39 viruses at an MOI of 0.001, and HIV-1 spread was quantified by intracellular p24 staining using FACS analysis between days 3 and 12 postinfection. Shown are mean values ± SD for four donors. (E and F) Replication in activated primary CD4+ T cells infected with ΔNef, WT, and Nef mutant viruses from two different donors.
DISCUSSION
The N-terminal region of the HIV-1 pathogenesis factor Nef recently emerged as a central element by which the viral protein subverts host cell processes, including cell surface exposure of receptors, antagonism of the restriction factor SERINC5, morphology and trafficking of CD4 T lymphocytes, and EV release. The detailed mapping presented here reveals that these activities represent independent functions of the viral protein that rely on distinct residues within the N-terminal region and thus likely reflect interactions with different host cell ligands (Fig. 8A and Table 1). With M20, the S5AM, and the NAKC recruiting EP repeats, three nonoverlapping determinants for Nef function were identified within the N-terminal interaction site of HIV-1SF2Nef that independently mediate cardinal activities, such as evasion of the host immune system or optimization of HIV-1 spread. Consistent with the high level of conservation of the downregulation of cell surface MHC-I and CD4 as well as antagonism of SERINC5 (13, 62–66) among Nef variants from HIV-1 strains of all subtypes, the determinants for these activities are highly conserved (Fig. 8B): the M20I replacement observed commonly in HIV-1 subtype C Nefs does not interfere with MHC-I downregulation (13, 67), and a VGAxS core motif within the S5AM is present in Nef proteins of all HIV-1 subtypes. In contrast, the EP repeat required for NAKC binding in HIV-1SF2Nef is poorly conserved. The N-terminal region of HIV-1 Nef thus emerges as a versatile protein interaction platform that mediates independent central activities of the viral pathogenesis factor via the M20 and S5AM protein interaction motifs.
Conservation of residues in the N-terminal region between HIV-1 and HIV-2/SIV Nef alleles and model. (A) Schematic model of interactions of the HIV-1 Nef N terminus with host cell ligands and their effector functions: M20 (MHC-I downregulation), EP repeat (residues 24/25 and 28/29) (NAKC association), and S5AM (residues 32 to 39) (CD4 downregulation, SERINC5 antagonism, HIV-1 replication), and disruption of T cell polarity, which requires the entire region. (B) Multiple-sequence alignment of residues in the N-terminal regions of different HIV-1, HIV-2, and SIV Nef alleles highlighting the motifs identified for different Nef functions, such as MHC-I downmodulation, NAKC binding, disruption of T cell polarity, CD4 downregulation, SERINC5 antagonism, and HIV-1 replication, in comparison to the reference sequence (SF2 Nef [in red]). Identical residues compared to the reference sequence are indicated by dots, and gaps are indicated by —. * indicates a position with fully conserved residues.
Among the functional motifs present in the HIV-1 N-terminal interaction site, residue M20 is well established to facilitate downmodulation of MHC-I molecules from the cell surface by providing a physical link to AP-1 complexes, resulting in the rerouting of MHC-I molecules, which contributes to the evasion of lysis of HIV-infected cells by cytotoxic T cells (12, 55). Of note, none of the additional Nef activities that involve this region was affected by the M20A mutation, suggesting that the AP-1-dependent mechanisms governing MHC-I downregulation by Nef are not involved in these additional activities. This includes association with the NAKC and the resulting effector functions, which was the first interaction partner identified for the Nef N-terminal region (43). Using Nef phosphorylation by the NAKC as a readout (45), our study narrows down the initial mapping of the NAKC interaction region of HIV-1SF2Nef to the two EP repeats at positions 24/25 and 28/29. Functional characterization of a Nef mutant with a disruption in this NAKC-interacting motif suggests that the association with the NAKC is dispensable for Nef functions in cell surface receptor modulation, SERINC5 antagonism, and virus replication in primary CD4 T cells. However, since the NAKC is a large multiprotein complex that may have multiple contacts with the N-terminal region, and the EP repeat identified as an NAKC binding determinant in HIV-1SF2Nef is not well conserved among HIV-1 Nef variants, these results do not exclude that NAKC subcomplexes lacking Nef-phosphorylating kinase activity can still associate with the Nef 2EPAA mutant or that divergent Nef proteins employ different interaction surfaces to associate with the NAKC. This may be particularly relevant in the context of EVs. While our study suggests that Nef-phosphorylating NAKC activity is dispensable for the recruitment of Nef into EVs, future studies will need to address the specific molecular determinants by which HIV-1 Nef modulates the content of EVs in HIV patients (48, 68, 69). Finally, the specific determinant of the Nef N-terminal region for antagonism of virion infectivity restriction by SERINC5 was mapped to residues 32 to 39, forming the S5AM. Since mutations in the S5AM did not affect the stability or subcellular localization of Nef, this motif likely mediates interactions with yet-to-be-determined host cell ligands. This is consistent with the available structures of the HIV-1 Nef N-terminal region, which is unstructured and flexible in the absence of a specific ligand but could rapidly fold into a protein interaction module upon contact with a specific ligand (32, 70). The identification of relevant S5AM ligands will be an important goal of future studies.
Importantly, the S5AM was also required for CD4 downregulation and the promotion of HIV-1 replication in primary CD4 T cells by Nef. This result further underscores that Nef may antagonize SERINC5 and reduce cell surface exposure of CD4 by related mechanisms (29, 30). Similar to the effects of Nef on CD4 (71, 72), internalization of SERINC5 from the cell surface and its subsequent degradation were suggested to be instrumental for reducing virion incorporation of the restriction factor and, thus, for antagonism by Nef (73). However, we previously demonstrated that HIV-1 Nef can antagonize SERINC5 in the absence of cell surface downregulation, intracellular accumulation, and degradation (52). Since mutations in the S5AM were associated with elevated virion incorporation of SERINC5, it seems more plausible that the action of this motif in preventing virion incorporation of the restriction factor is by affecting its anterograde transport or lateral recruitment to HIV budding sites. In addition, the S5AM may also facilitate the inactivation of virion-associated SERINC5 by Nef (52).
How the role of the S5AM in SERINC5 antagonism and CD4 downregulation relates to its requirement for efficient HIV-1 replication in primary human CD4 T lymphocytes remains to be established. Importantly, the inability of S5AM mutant Nef to boost HIV-1 replication in this cell system matches previous results for Nef mutants that fail to downregulate CD4 and antagonize SERINC5 due to the disruption of a dileucine motif in Nef’s C-terminal flexible loop (74, 75). The molecular determinants for CD4 downregulation/SERINC5 antagonism in Nef are thus essential for efficient HIV-1 spread in primary human CD4 T lymphocytes. In principle, this may reflect adverse effects of CD4 on virion infectivity and production. However, these effects are also counteracted by HIV-1 Vpu and Env or require unphysiologically high levels of CD4 and are thus unlikely to account for the effect of Nef on HIV-1 spread in primary CD4 T cells (76–78). Moreover, two recent studies suggest that the ability of Nef to counteract the restriction to virion infectivity in a single round of infection by SERINC5 also does not translate into improved virus spread in these cells over multiple rounds of infection (75, 79). As recently suggested for the MOLT3 T cell line, the Nef-mediated enhancement of HIV-1 spread in such cultures may thus reflect the antagonism of a yet-to-be-identified host cell restriction factor other than SERINC5, which may be counteracted by Nef via a S5AM-dependent mechanism.
While the M20 and S5AM interaction modules are largely conserved among Nef variants from HIV-1 strains of different subtypes, the N-terminal region of Nef proteins from HIV-2 and SIV strains is not homologous to that from HIV-1 and does not contain these protein interaction determinants (Fig. 8B). However, cardinal Nef activities, such as downregulation of cell surface MHC-I and CD4 as well as SERINC5 antagonism, are highly conserved among all lentiviral Nef proteins (13, 62–66). Similar to the antagonism of the host cell restriction factor tetherin by Nef in SIV and HIV-1 group O strains and Vpu in most HIV-1 strains (80, 81), different molecular solutions for the same Nef activity may have been selected during lentiviral evolution, and it will be of interest in future studies to delineate which molecular determinants exert S5AM function in the HIV-2/SIV lineage.
Together, data from this study define three independent protein interaction motifs in the N-terminal region of HIV-1SF2Nef that independently mediate cardinal functions of the pathogenesis factor. Future efforts to develop strategies for therapeutic intervention with Nef function should therefore aim at interfering with the activity of this multivalent protein interaction module.
MATERIALS AND METHODS
Cell culture and plasmids.Jurkat T antigen (TAg) (expressing the simian virus 40 large T antigen) and A3.01 T lymphocytes were cultivated in RPMI 1640 plus GlutaMAX-I supplemented with 10% fetal calf serum (FCS) and 1% penicillin-streptomycin (all from Invitrogen). Jurkat (CCR7) T lymphocytes were cultivated in the same supplemented medium as the one described above, along with 1× nonessential amino acids, 1× sodium pyruvate, 10 mM HEPES (pH 7.4), and 45.8 μM β-mercaptoethanol (standard concentrations of up to 55 μM can be used). 293T and COS7 cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% FCS and 1% penicillin-streptomycin. The expression constructs for GFP fusions of HIV-1SF2Nef wild-type (WT) and mutant proteins were described previously (4, 26, 40, 82). The alanine mutants used in the study were generated using a Q5 site-directed mutagenesis (SDM) kit, and primers were designed using NEBaseChanger (New England Biolabs). The proviral chimera pHIV-1NL4-3 SF2 Nef WT (HIV-1 Nef WT) was generated as described previously (51, 52). The alanine mutants in this backbone were generated for infection-based experiments using the same primers and SDM kit as the ones described above for the expression constructs. pCMVHA EED wt was purchased from Addgene (plasmid 24231). pPKC-θ.HA and pBJ5-SERINC5.intHA were kind gifts from Andreas Baur and Heinrich Göttlinger, respectively.
Flow cytometry.A total of 1 × 107 A3.01 cells were electroporated with 30 μg plasmid DNA for each sample (GenePulser, 850 μF and 300 V; Bio-Rad). Cells were harvested at 24 h posttransfection, washed with 1× phosphate-buffered saline (PBS), and incubated with allophycocyanin (APC)-conjugated antibodies against CD4 (1:100) (RPA-T4; BioLegend) and HLA-ABC (1:50) (G46-2.6; BioLegend) for 30 min in 1× PBS. Samples were measured by flow cytometry using the BD FACSVerse system with BD FACSuite software. Within samples, the surface receptor levels (geometric mean of the median fluorescence intensity [MFI]) of medium- to high-GFP-expressing cells were compared to surface receptor levels of non-GFP-expressing cells as described previously (4, 8). Gating and MFI analysis were done using FlowJo software 10.4.2, and data were processed with Microsoft Office Excel 2016 and GraphPad Prism 6.0 software.
Immunofluorescence.T cell polarization was monitored by changes in cell morphology and the formation of uropods, as described previously (53). In brief, stimulatory coverslips (CSs) were prepared by coating with fibronectin (45 μg/ml; Sigma-Aldrich) for 1.5 h. A total of 1 × 107 A3.01 T cells per sample were electroporated as described above, harvested at 24 h posttransfection, resuspended in 100 μl supplemented RPMI, seeded onto CSs, and allowed to polarize for 2 h at 37°C in 5% CO2. Cells were then fixed with 8% paraformaldehyde (PFA), permeabilized with 0.1% Triton X-100, and stained with rat anti-CD44 (1:800) (clone IM7; BioLegend), followed by Alexa Fluor 568 goat anti-rat antibody (Life Technologies). Lck accumulation and phosphorylated cofilin (p-cofilin) assays were performed as described previously (17, 41). Briefly, Jurkat (TAg) or Jurkat (CCR7) T cells were electroporated with 20 to 50 μg of total plasmid DNA per 1 × 107 cells (GenePulser, 250 V and 950 and 850 μF, respectively; Bio-Rad), and at 24 h postelectroporation, cells were seeded onto 0.01% poly-l-lysine (PLL; Sigma-Aldrich)-coated CSs, fixed with 3% PFA, permeabilized, and stained with rabbit anti-Lck (1:50) (clone 3A5; Santa Cruz Biotechnology) or rabbit anti-phospho-cofilin (1:50) (clone 77G2; Cell Signaling) and secondary Alexa Fluor 568 goat anti-rabbit antibody (Life Technologies). Coverslips were mounted with Mowiol 4-88 (Merck Millipore) and analyzed by epifluorescence microscopy (IX81 SIF-3 microscope and Xcellence Pro software; Olympus) and confocal microscopy (TCS SP5 microscope and LAS AF software; Leica). Images were processed using Adobe Photoshop CC 2017 and ImageJ. For quantification of phenotype frequencies, 50 to 100 transfected cells were counted per experiment, and their phenotype was judged in comparison to that of untransfected neighboring cells (15).
T cell chemotaxis.Analysis of SDF-1α-mediated T lymphocyte chemotaxis was performed as described previously (18, 20, 83). Briefly, 10 × 106 Jurkat T (CCR7) lymphocytes were electroporated with 50 μg GFP or Nef.GFP plasmid DNA and harvested after 24 h. Cells were starved for 2 h in starvation medium (RPMI supplemented with 0.5% FCS). A total of 1 × 106 cells resuspended in 100 μl were collected as the input sample for FACS analysis, 100 μl of 1 × 106 cells was placed on the upper chamber of a transwell system (Costar 3421, 5-mm pores, 24-well plates; Corning, Kaiserslautern, Germany), and 450 μl starvation medium with or without 10 ng/ml SDF-1α was added to the lower chamber. Cells were allowed to migrate for 2 h at 37°C in 5% CO2 and then collected from the lower chamber for FACS analysis of the total percentage of GFP-expressing cells. As a measure for chemotaxis, the ratio of the percentage of high-GFP-expressing cells that migrated to that present in the input was calculated. All samples were measured for 1 min at a constant flow rate.
Immunoprecipitation and IVKAs.Immunoprecipitation and in vitro kinase assays (IVKAs) were performed as described previously, with minor modifications (41, 43). In brief, 3 × 105 COS7 cells were seeded per well 24 h prior to transfection in 6-well plates. Cells were transfected with 2 μg GFP/Nef.GFP or mutant expression constructs along with 2 μg of a PKC-θ–hemagglutinin (HA) fusion protein (PKC-θ.HA) using polyethylenimine (PEI) at a 1:3 (DNA-to-PEI) ratio. At 24 h posttransfection, cells were collected and lysed in lysis buffer (50 mM Tris-HCl [pH 7.4], 75 mM NaCl, 1 mM EDTA, 1 mM NaF, and 0.5% NP-40) with a freshly added protease inhibitor cocktail and sodium vanadate and subjected to ultrasonication (Bioruptor Plus; Diagenode). Cell lysates were then mixed with 25 μl GFP-TrapA beads (Chromotek) and incubated at 4°C with rotation for 2 h. The beads were washed three times with lysis buffer and incubated with 50 μl kinase assay buffer (50 mM HEPES [pH 8.0], 150 mM NaCl, 5 mM EDTA, 0.02% Triton X-100) containing freshly added 10 mM MnCl2 and 10 μCi of [γ-32P]ATP (Hartmann Analytic) for 20 min at room temperature (RT). The beads were again washed three times with lysis buffer, resuspended in 50 μl 6× sample buffer (10% sucrose, 0.1% bromophenol blue, 5 mM EDTA [pH 8.0], 200 mM Tris [pH 8.8]), and boiled at 95°C for 15 min. Bound proteins were separated by SDS-PAGE and subjected to autoradiography. Membranes were also incubated with monoclonal rat anti-GFP (1:1,000) (clone 3H9; Chromotek) and developed by chemiluminescence. Densitometric analyses were performed using ImageJ 1.50i (NIH, USA), and data were analyzed using Microsoft Office Excel 2016.
Exosome release assay.A total of 5 × 106 293T cells were seeded in 150-mm culture dishes with 20 ml DMEM supplemented with 10% exosome-depleted FCS (ultracentrifugation at 26,000 rpm for 16 h) and 1% penicillin-streptomycin. At 24 h postseeding, each culture dish was cotransfected using PEI with the following expression vectors: 20 μg GFP or Nef.GFP, 5 μg EED.WT (Addgene), 10 μg hnRNPK.myc.His (Addgene), PKC-θ.HA, and an Lck-red fluorescent protein (RFP) fusion (Lck.RFP). At 24 or 48 h posttransfection, cells were harvested for lysis as mentioned above, and the supernatant was collected and processed for exosome purification using the Exo-Spin kit (Cell Guidance Systems) according to the manufacturer’s protocol. Briefly, 40 ml of the supernatant was spun at 300 × g for 10 min to remove cell debris. The supernatant was then transferred to a new tube, spun at 16,000 × g for 30 min to remove multivesicular bodies, and further passed through a 0.22-μm filter. A half-volume of Exo-Spin buffer was added, and the mixture was incubated overnight at 4°C. The samples were spun at 16,000 × g for 1 h, and the precipitates were collected in 300 μl PBS and passed through purification columns. The purified exosome fraction and cell lysates were then used for total protein estimation (Micro bicinchoninic acid [BCA] protein assay kit; Thermo Fisher). A total of 200 μg of the cell lysate and 50 μg of the exosome lysate were loaded per lane on a 10% SDS gel, subjected to PAGE, blotted onto a nitrocellulose membrane, and incubated overnight at 4°C with the following primary antibodies in 5% bovine serum albumin (BSA)–PBS–Tween 20: rabbit anti-glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (14C10; Cell Signaling), mouse anti-HA (HA.11, clone 16B11; BioLegend), rat anti-GFP (Chromotek), rabbit anti-RFP (Novus Biologicals), and mouse anti-CD63 (Merck Millipore). The membranes were incubated with horseradish peroxidase (HRP)-conjugated secondary antibodies at RT for 1 h and developed by chemiluminescence.
Virus production, infectivity measurements, and infection assays.Virus production from 293T cells, infectivity measurements using SG-PERT for total viral particles and a TZM-bl luciferase reporter assay, and Western blotting for protein expression in cell lysates and virions were performed as described previously (52, 84, 85). Primary CD4+ T cells were isolated from buffy coats obtained from the Heidelberg University Hospital Blood Bank using a RosetteSep human CD4+ T cell enrichment kit (StemCell Technologies, Canada) according to the manufacturer’s protocol. Cells were activated using 60 μl CD3/CD28 Dynabeads (Thermo Fisher Scientific, Germany) per 1 × 107 cells for 72 h in complete medium (RPMI supplemented with 1% penicillin-streptomycin, 20% FCS, and 10 ng/ml interleukin-2 [IL-2]). The beads were removed, and cells were infected at an MOI of 0.001 for 4 h. The virus was washed away, and cells were maintained in medium containing IL-2. Cell pellets were collected every 2 to 3 days, fixed for 90 min with 3% paraformaldehyde, and stained with p24 Kc57-fluorescein isothiocyanate (FITC) (Beckman Coulter, USA) antibody in 0.1% Triton X-100 for 20 min for FACS analysis. Cells were gated for the live population, followed by FITC for p24 expression.
Statistical analysis.Statistical analysis of data sets was carried out using Prism version 6.0 (GraphPad). Statistical significance was calculated using one-way analysis of variance (ANOVA).
ACKNOWLEDGMENTS
We are grateful to Heinrich Göttlinger and Andreas Baur for the gift of reagents, Ina Ambiel for technical assistance, and Virginia Pierini for expertise and protocols on the analysis of SERINC restriction and antagonism.
This project is supported by the Deutsche Forschungsgemeinschaft (DFG) (German Research Foundation) (SPP1923, Projektnummer 240245660, and SFB 1129), TTU HIV of the German Centre for Infection Research (DZIF) (TTU 04.810, an Integrated Approach for HIV-Cure “Shock and Kill” Strategies, to O.T.F.; TI 07.003, FP2016 fellowship, to B.O.; and TTU 04.704, infrastructural measure tenure track professorship, to M.L.), Gilead Sciences GmbH funding to M.L., and the Humboldt Foundation (scholarship reference 3.3-ITA-1193954-HFST-P) to I.L.S. O.T.F. and M.L. are members of the CellNetworks cluster of excellence (EXC81).
FOOTNOTES
- Received 20 August 2019.
- Accepted 1 October 2019.
- Accepted manuscript posted online 9 October 2019.
- Copyright © 2019 American Society for Microbiology.
REFERENCES
- 1.↵
- 2.↵
- 3.↵
- 4.↵
- 5.↵
- 6.↵
- 7.↵
- 8.↵
- 9.↵
- 10.↵
- 11.↵
- 12.↵
- 13.↵
- 14.↵
- 15.↵
- 16.↵
- 17.↵
- 18.↵
- 19.↵
- 20.↵
- 21.↵
- 22.↵
- 23.↵
- 24.↵
- 25.↵
- 26.↵
- 27.↵
- 28.↵
- 29.↵
- 30.↵
- 31.↵
- 32.↵
- 33.↵
- 34.↵
- 35.↵
- 36.↵
- 37.↵
- 38.↵
- 39.↵
- 40.↵
- 41.↵
- 42.↵
- 43.↵
- 44.↵
- 45.↵
- 46.↵
- 47.↵
- 48.↵
- 49.↵
- 50.↵
- 51.↵
- 52.↵
- 53.↵
- 54.↵
- 55.↵
- 56.↵
- 57.↵
- 58.↵
- 59.↵
- 60.↵
- 61.↵
- 62.↵
- 63.↵
- 64.↵
- 65.↵
- 66.↵
- 67.↵
- 68.↵
- 69.↵
- 70.↵
- 71.↵
- 72.↵
- 73.↵
- 74.↵
- 75.↵
- 76.↵
- 77.↵
- 78.↵
- 79.↵
- 80.↵
- 81.↵
- 82.↵
- 83.↵
- 84.↵
- 85.↵
