ASCT2 inhibits HIV-1 infectivity by promoting the incorporation of a gp160/ASCT2 complex into virions

ABSTRACT

Restriction factors are type I interferon (IFN)-inducible proteins that provide an early line of defense against HIV-1. Our current findings propose the amino acid transporter ASCT2 as a new host plasma membrane transporter that restricts the infectivity of HIV-1 viral particles. Our results show that ASCT2 is constitutively expressed in CD4⁺ T cells and macrophages, but its expression can be upregulated by the antiviral cytokine IFNα. We also found that ASCT2 is downmodulated from the cell surface of HIV-1-infected CD4⁺ T cells and macrophages. Our experiments suggest that ASCT2 interacts with the Env precursor gp160 in the endoplasmic reticulum (ER) and forms a complex of uncleaved gp160 and immature ASCT2. This complex can be incorporated into nascent viral particles and may lead to lower infectivity of progeny HIV-1. However, the restriction activity of ASCT2 against HIV-1 is counteracted by the accessory protein Vpu. We found that Vpu interacts with ASCT2 in the ER and prevents ASCT2 maturation through the N-glycosylation pathway, resulting in the downmodulation of ASCT2 from the plasma membrane. In summary, our results suggest that the amino acid transporter ASCT2 functions as an IFNα-inducible restriction factor against HIV-1 by promoting the incorporation of the complex formed by the inactive precursor of Env gp160 and immature ASCT2 into the viral particles, which is counteracted by the HIV-1 accessory protein Vpu.

IMPORTANCE

Globally, there are about 39 million people living with HIV (PLWH). Antiretroviral treatment (ART) has transformed HIV infection into a chronic condition. However, effective ART does not cure HIV since the virus establishes long-lasting reservoirs that are resistant to currently available ART. Nowadays, the choice of the ART regimen depends on several factors, such as drug resistance, health conditions of PLWH, and possible side effects of medications. To address the complexity of the ART regimen, drug toxicities, and drug resistance, the research on novel formulations continues. Currently, new antiviral therapies are being developed that target important interactions between viral accessory proteins and host restriction factors, such as the interaction between Vpu and BST-2 and between Vif and APOBEC3. Therefore, our current findings about ASCT2 and Vpu not only provide novel insights about HIV-1 virus-host restriction factor interaction but also propose a novel and promising strategy to develop a new treatment against HIV.

INTRODUCTION

The interplay between HIV-1 and various host plasma membrane proteins allows HIV-1 to avoid antiviral immune responses and establish a systemic and persistent infection. It has been previously reported that amino acid transporters can modulate HIV-1 replication through different mechanisms, such as the serine amino acid transporters SERINC3 and SERINC5 (1). Conversely, HIV-1 expression was shown to affect the alanine amino acid transporter SNAT1, with subsequent downstream effects on T cell activation (2). Alanine, serine, cysteine transporter 2 (ASCT2), also known as SLC1A5, is a 50 to 80 kDa multi-pass transmembrane glycoprotein, which is responsible for amino acid transport across the plasma membrane, with glutamine being the preferred substrate (3). Because of the key physiological role of ASCT2, this transporter is widely distributed at the plasma membrane of a broad range of cells (4, 5). In fact, ASCT2 is a major regulator of glutamine transport in HIV-1 primary target cells, such as activated CD4⁺ T cells (5–7) and differentiated macrophages (8, 9). Moreover, it has been reported that infection of HIV-1 leads to considerable changes in the cellular glutamine metabolism of CD4⁺ T cells (10, 11) and macrophages (12, 13). Interestingly, it has been shown that ASCT2 is downregulated from the cell surface of CEM-T4 T cells infected with VSVg-pseudotyped HIV-1 NL43 virus (2).

HIV-1 encodes two membrane-associated accessory proteins, called negative factor (Nef) and viral protein U (Vpu), that ensure viral persistence, replication, dissemination, and transmission by altering the expression of cell surface proteins that counteract or restrict HIV-1 replication. Vpu is a 17 kDa type I transmembrane protein expressed coordinately with the viral Env protein during late stages of the HIV-1 replication cycle (14–16). Vpu is distributed throughout the cellular membrane trafficking system, including the ER and Golgi apparatus (17, 18), where it can act as a nonenzymatic adapter between target proteins and cellular complexes involved in degradation pathways. Vpu targets several host cell surface proteins that interfere with HIV-1 replication, including the HIV receptor CD4 (19), BST-2 (20, 21), CD1d (22, 23), CD28 (24), HLA-C (25, 26), NTB-A (27, 28), PVR (29, 30), and CD99 (31) immunoreceptors. Other factors reported as targets of Vpu are CCR7 (32), CD62L (33), ICAM-1 (34, 35), homing molecules (36), tetraspanins (37), and the amino acid membrane transporter SNAT1 (2).

Given the reported effects of HIV-1 infection on cellular glutamine metabolism, we decided to study the functional relationship of HIV-1 with ASCT2. We found that ASCT2 is constitutively expressed in CD4⁺ T cells and macrophages, but its expression can be upregulated by the antiviral cytokine IFNα. Interestingly, the expression of endogenous ASCT2 is downregulated at transcriptional and post-transcriptional levels in HIV-infected CD4⁺ T cells and macrophages. We found that ASCT2 interacts with the Env precursor gp160 and forms a complex comprising the uncleaved gp160 and immature molecules of ASCT2. This complex is packaged into the viral particles, leading to the inhibition of viral infectivity. However, our results also show that Vpu counteracts the effect of ASCT2 on HIV-1 infectivity. Indeed, Vpu interacts with ASCT2, which inhibited ASCT2 maturation through the N-glycosylation pathway. Our data suggest that Vpu delays the exit of ASCT2 from the ER, thereby delaying ASCT2 maturation and resulting in reduced expression of ASCT2 at the plasma membrane.

RESULTS

HIV-1 infection of MDM and CEM-SS cells promotes the Vpu-dependent downmodulation of endogenous ASCT2

ASCT2 is the main glutamine transporter in mammalian cells (3), including the primary target cells of HIV-1. A cell surface proteomic study of CEM-T4 cells suggested that HIV-1 infection of these cells led to Vpu-dependent cell surface downregulation of ASCT2 (2). To further explore this phenomenon, we examined the expression of endogenous ASCT2 in human monocyte-derived macrophages (MDM), CEM-SS cells, and primary peripheral blood lymphocytes (PBLs) following infection with wild-type HIV-1 or its Vpu-deficient variant. To maximize the efficiency of infection in MDM and CEM-SS cells, we employed VSVg pseudotyped virus stocks. MDM were infected using the R5-tropic AD8 isolate (Fig. 1A, lanes 2 and 3), and CEM-SS cells were infected using the X4 tropic NL43 isolate (Fig. 1A, lanes 5 and 6). Uninfected cells were included as negative controls (Fig. 1A, lanes 1 and 4, respectively). For infection of PBL, we did not use VSVg pseudotyped virus stocks in order to limit infection to CD4⁺ cells (Fig. 1B). Infections were allowed to proceed for 7 days (MDM), 2 days (CEM-SS), or 14 days (PBLs). MDM and CEM-SS cells were processed for immunoblotting using the anti-human ASCT2 (hASCT2), anti-Env, anti-HIV-IG, anti-Vpu, and anti-GAPDH antibodies (Fig. 1A). PBLs were analyzed by FACS analysis using anti-SLC1A5-FITC to detect ASCT2, anti-human CD4-APC to discriminate CD4⁺ T cells from CD4-negative cell types (e.g., CD8⁺ T cells, NK, and B cells), and anti-p24-Gag-PE to discriminate between infected and uninfected cells (Fig. 1B). Interestingly, levels of endogenous ASCT2 were reduced in MDM and CEM-SS cells infected with wild-type virus relative to uninfected cells (Fig. 1A, compare lanes 2 to 1 and 5 to 4). Indeed, quantitation of ASCT2 protein levels from at least six biological replicates revealed that the downmodulation of ASCT2, especially in MDM, is Vpu-dependent (Fig. 1C; compare blue and red symbols). Of note, Vpu-deficient HIV-1 AD8 virus failed to significantly downmodulate ASCT2 when compared to AD8 WT. In CEM-SS cells, Vpu-deficient virus only partially lost the ability to downmodulate ASCT2, which could be an indication for the involvement of additional viral factor(s) in the control of ASCT2 expression or could be a virus- and/or cell type-specific effect (Fig. 1C). FACS analysis of endogenous ASCT2 and CD4 levels in p24⁺ PBLs showed that endogenous ASCT2 levels as well as CD4 levels were significantly reduced in infected primary CD4⁺ T cells when compared to uninfected cells (Fig. 1B). In contrast to MDM and CEM-SS cells, infection of PBL with Vpu-deficient NL43 did not abolish the ability to downmodulate CD4 or ASCT2, suggesting the involvement of an additional viral factor(s), such as Nef, in the control of ASCT2 and CD4 in these cells (Fig. 1B).

Fig 1.

Infection of various cell types with HIV-1 correlates with the downmodulation of endogenous ASCT2 in a Vpu-dependent manner. (A) VSVg pseudotyped wild-type or Vpu-defective AD8 (lanes 2 and 3) or NL43 viruses (lanes 5 and 6) were used to infect MDM or CEM-SS cells, respectively. Uninfected cells were included as a control (lanes 1 and 4). Cells were collected 7 days (MDM) or 2 days (CEM-SS cells) post-infection and processed for immunoblot analysis using antibodies to human ASCT2 (hASCT2), Env, HIV-IG, Vpu, and GAPDH. Digital images were acquired using BioRad Image Lab software. A representative result from three independent experiments is shown. (B) PBLs from two healthy donors were analyzed by FACS analysis. Cells were stained for 1 hour using anti-SLC1A5-FITC (ASCT2), anti-human CD4-APC, and anti-p24-Gag-PE antibodies. Cells were then fixed with 1% formaldehyde fixation solution and analyzed with a BD Flow Cytometer (BD Accuri C6 Plus). Results were graphed using GraphPad software. The error bars represent the standard error of the mean (SEM) from two experiments. The statistical significance was determined using two-way ANOVA with a Dunnett multiple comparison test. P: *, ≤0.05; **, ≤0.01. (C) Pixel intensity of ASCT2 and GAPDH bands from panel A was quantified using GraphPad software. The error bars represent the standard error of the mean (SEM) from three experiments. The statistical significance was determined using one-way ANOVA with a Tukey multiple comparison test. P: *, ≤0.05; **, ≤0.01; ***, ≤0.001; ****, ≤0.0001; ns, no significance. (D) Total RNA was extracted from MDM 6 days after infection and from CEM-SS cells 2 days after infection, using the Direct-zol RNA miniprep kit (Zymo Research, Irvine, CA). RT-qPCR was performed with 50 ng of total RNA and specific primer sets for ASCT2 and GAPDH genes, using the iTaq Universal SYBR Green one-step kit. The housekeeping gene GAPDH was used as an internal control, and relative mRNA levels were determined using the threshold cycle (ΔΔCT) quantification method. mRNA levels from three independent experiments were graphed and analyzed using GraphPad software. Error bars represent the standard error of the mean (SEM). The statistical significance was determined using one-way ANOVA with a Tukey multiple comparison test. P: *, ≤0.05; ns, no significance. (E) Cryo-EM studies previously revealed the domain structure of ASCT2. Based on this structural model, we constructed a variant of ASCT2 carrying an HA epitope tag in the extracellular loop 2b (ECL2b) of ASCT2 to allow detection of ASCT2 at the cell surface. The arrows point to the HA-tag insertion site as well as the approximate location of two N-linked carbohydrates. (F) HEK293T cells were transiently transfected with 0.5 µg of pCMV-ASCT2-Flag and increasing amounts of pcDNA-Vphu WT (0, 0.5, 1, 2.5, and 4.5 µg). Total amounts of transfected DNA were adjusted to 5.0 µg for each sample using empty vector DNA, as needed. Mock-transfected cells were included as a negative control. After 24 hours of transfection, cells were collected and processed for FACS analysis. Cells were stained with anti-HA-APC and anti-TfR-APC antibodies for 1 hour at RT. Cells were then fixed with 1% formaldehyde fixation solution and analyzed with a BD Flow Cytometer (BD Accuri C6 Plus). Results were graphed using GraphPad software. The error bars represent the standard error of the mean (SEM). The statistical significance was determined using a one-way ANOVA with a Dunnett multiple comparison test. P: **, ≤0.01; ***, ≤0.001; ns, no significance. (G) CEM-SS cells were infected with VSVg pseudotyped NL43 WT and NL43 Udel viruses. Supernatants were processed to determine the concentration of glutamine using the Glutamine/Glutamate Glow Assay kit (Promega Corp., Madison, WI; Cat# J8201), according to the manufacturer’s instructions. The values were graphed using GraphPad software. The error bars represent the SEM of three independent experiments. The statistical significance was determined using two-way ANOVA with a Tukey multiple comparison test. P: *, ≤0.05; ****, ≤0.0001.

To further test how HIV-1 affects endogenous ASCT2 expression, we analyzed ASCT2 mRNA expression in uninfected or HIV-infected MDM and CEM-SS cells (Fig. 1D). Infections were done as described for Fig. 1A. We found that HIV-1 WT was able to downregulate the expression of ASCT2 mRNA in CEM-SS cells but not in MDM. Interestingly, deletion of Vpu did not affect the ability of HIV-1 to inhibit ASCT2 mRNA expression in MDM and CEM-SS cells. These results indicate that regulation of ASCT2 in HIV-1-infected cells is a multifactorial process, part of which may be virus dependent and part of which may be cell type specific. Overall, however, we can conclude that HIV-1 infection of MDM and CEM-SS cells leads to the dysregulation of ASCT2 expression.

Vpu has been implicated in the downmodulation of multiple cellular membrane proteins from the cell surface (for review, see 38–40). Because of that, we wondered if Vpu might similarly affect the cell surface expression of ASCT2. HEK293T cells were transfected with plasmids expressing an ASCT2 variant containing an HA-tag in one of its ectodomains (see Fig. 1E) and increasing amounts of Vpu. After transfection, cell samples were processed for FACS analysis using the anti-HA-APC and anti-TfR-APC antibodies. We found that plasma membrane expression levels of ASCT2 were reduced by Vpu in a dose-dependent manner, whereas surface expression of endogenous transferrin receptor (TfR), which served as a negative control, was unaffected (Fig. 1F). Thus, the expression of Vpu results in the specific downmodulation of ASCT2 from the plasma membrane.

ASCT2 is the major transporter of the essential amino acid glutamine in macrophages and T CD4⁺ T cells. We, therefore ,tested if ASCT2 downmodulation from the plasma membrane correlates with a change in glutamine uptake from the culture medium. For that purpose, CEM-SS cells were infected with VSVg pseudotyped NL43 WT or the Vpu-deficient variant. Two days after infection, a Glutamine/Glutamate-Glow assay, which is a sensitive method to measure glutamine levels in the culture medium, was performed according to the manufacturer’s instructions (Fig. 1G). The assay measures the conversion of glutamine to glutamate by a glutaminase enzyme. Glutamate oxidation and NADH production are coupled with a bioluminescent NADH detection system. In this assay, the amount of light produced is proportional to the amount of glutamine. Thus, inhibition of ASCT2 activity in the culture supernatant is inversely correlated with levels of light production. The assay calculates glutamine concentration based on the relative light units (RLU) by using a glutamine titration curve. Indeed, we found that glutamine concentrations were higher in supernatants of CEM-SS cells infected with NL43 WT virus when compared to the medium of uninfected cells or cells infected with Vpu-deficient HIV-1 (Fig. 1G). These results suggest that ASCT2 expression at the plasma membrane correlates with a higher glutamine uptake and a subsequently lower concentration of glutamine in the culture supernatants. We therefore conclude that Vpu dysregulates the uptake of the essential amino acid glutamine through the downmodulation of ASCT2 from the plasma membrane of infected cells.

The expression of ASCT2 is IFN-inducible in MDM and primary CD4⁺ T cells

Interferon (IFN)-inducible proteins play a crucial role in restricting HIV-1 replication and are a critical part of the host’s innate immune response. Expression of these proteins, which is triggered or enhanced by IFN signaling, can interfere with various stages of the HIV-1 replication cycle (41). We tested the effect of IFN treatment on ASCT2 expression in primary MDM and CD4⁺ T lymphocytes. MDM from three healthy donors and PBLs from two healthy donors were treated for 24 hours with increasing doses of IFN-α2A. After treatment, MDM were prepared for immunoblot analysis using antibodies to human ASCT2 (ASCT2), human BST-2, and tubulin (Fig. 2A). As expected, levels of tubulin remained constant, while levels of BST-2 increased with increasing levels of IFN. Quantitation of ASCT2 bands also revealed a statistically significant dose-dependent increase of ASCT2 levels (Fig. 2B). PBLs were prepared for FACS analysis using antibodies to ASCT2 (anti-SLC1A5-FITC) and human CD4 (anti-CD4-APC) (Fig. 2C). Taken together, these results suggest that ASCT2 is an IFN-inducible protein in MDM and CD4⁺ T cells.

Fig 2.

The expression of ASCT2 is IFN-inducible in MDM and primary CD4⁺ T cells. (A) Human monocytes were maintained in six well-plates (2 × 10⁶ cells/well) in DMEM supplemented with 4.5 g/L D-glucose, 10% pooled human serum, 100 U/mL penicillin, 100 µg/mL streptomycin, 2 mM L-glutamine, and 1 mM sodium pyruvate (differentiation medium). Cells were cultured for 7 days to allow differentiation into macrophages (MD). MDM from three healthy donors were treated with increasing amounts of IFN-α2A (0, 0.01, 0.1, 1, 10, and 100 ng/mL) for 24 hours. After treatment, whole-cell extracts were prepared and processed for immunoblotting using anti-human ASCT2, anti-BST-2, and anti-tubulin (Tub) antibodies. Digital images were acquired using BioRad Image Lab software. A representative immunoblot image from three independent experiments is shown. (B) ASCT2 and tubulin from panel A were quantified using GraphPad software. The error bars represent the standard error of the mean (SEM). The statistical significance was determined using one-way ANOVA with a Dunnett multiple comparison test. P: *, 0.05 ≤ 0.01; **, ≤0.01. (C) PBLs from two healthy donors were stimulated with 10% FCS RPMI in the presence of concanavalin A (20 µg/ml) for 20 hours. After stimulation, the medium containing concanavalin A was replaced with 10% FCS RPMI in the presence of human IL-2 (20 U/ml). Cells were plated in wells of a 24-well plate (2 million cells/well) and incubated with increasing amounts of IFN-α2A (0, 0.01, 0.1, 1, 10, and 100 ng/mL) for 24 hours. After treatment, cells were processed for FACS analysis using anti-ASCT2-FITC and anti-CD4-APC antibodies. The values were graphed using GraphPad software. The error bars represent the SEM of two different experiments. The statistical significance was determined using one-way ANOVA with a Tukey multiple comparison test. P: *, ≤0.05 ≤ 0.01; **, ≤0.01; ***, ≤0.001.

ASCT2 interacts with Vpu

Given the observed effects of Vpu on the expression of ASCT2, it seemed likely that the two proteins interact either directly or indirectly. We used two approaches to assess the physical interaction of Vpu and ASCT2. First, we studied the colocalization of the two proteins by confocal microscopy (Fig. 3A). HeLa cells were used for this experiment because of their more homogeneous morphology when compared to HEK293T cells. Cells were grown on cover slips and transfected with pCMV-ASCT2-Flag and either empty vector (Fig. 3A, Ctrl) or pcDNA-Vphu [Fig. 3A, Vpu (+)]. After 24 hours, cells were fixed with PFA and stained with anti-Flag (ASCT2) or anti-Vpu antibodies. ASCT2 staining is shown in red, Vpu staining appears in green, and the overlap of ASCT2 and Vpu staining is depicted in yellow. The nuclei were visualized by DAPI staining and appear in blue. As expected, Vpu was found concentrated in the perinuclear region, presumably the Golgi apparatus, as well as in a reticular network, presumably the ER (Fig. 3A, panel F). In the absence of Vpu, ASCT2 was found distributed throughout the cell and at the plasma membrane (Fig. 3A, panel C). In contrast, in the presence of Vpu, ASCT2 was found to colocalize with Vpu in the perinuclear region of the cells (see inset in panel H of Fig. 3A). To measure the linear correlation of the fluorescence signal emitted by ASCT2-Flag and Vpu, the Pearson correlation coefficient (PCC) mean value of 50 different cells from three independent experiments was calculated using the FIJI imaging software. The calculated PCC value was 0.604, which indicates that there is a strong overlap of ASCT2-Flag and Vpu staining. These results suggest that ASCT2 colocalizes with Vpu at the perinuclear region of infected cells, presumably the ER and Golgi.

Fig 3.

ASCT2 interacts with Vpu. (A) HeLa cells were transiently transfected with 0.6 µg of pcDNA-Vphu WT with or without 0.2 µg of pCMV-ASCT2-Flag. Total amounts of transfected plasmid DNA were adjusted to 0.8 µg for each sample using empty vector DNA. After 24 hours of transfection, cells were fixed with 4% PFA, permeabilized with 1% Triton X-100, and stained with anti-Flag (mouse) and anti-Vpu (rabbit) primary antibodies and Alexa Fluor 488 (rabbit) and Alexa Fluor 594 (mouse) secondary antibodies. The images were acquired on a Leica TCS SP8 X White Light confocal microscope and analyzed with FIJI/Image J software. The PCC mean value was calculated from 50 different cells from three different experiments. A representative IFI image from three independent experiments is shown. ASCT2 is shown in red; Vpu in green; and areas of colocalization appear in yellow. PFA: paraformaldehyde. PCC: Pearson correlation coefficient. (B) HEK293T cells were transiently transfected with 0.5 µg pCMV-ASCT2-Flag and 4.5 µg of pcDNA-Vphu. Total amounts of transfected plasmid DNA were adjusted to 5.0 µg for each sample using empty vector DNA, as needed. Cells were harvested 24 hours post-transfection and processed for immunoprecipitation as described in Materials and Methods. Cell lysates were incubated with anti-Flag beads for 1 hour at 4°C. Whole-cell extracts (input control) and immunoprecipitated samples (IP: α-Flag) were subjected to SDS-PAGE and analyzed by immunoblotting using antibodies to the Flag epitope tag in ASCT2, Vpu, or tubulin (Tub). Digital images were acquired using BioRad Image Lab software. A representative immunoblot image from three independent experiments is shown.

We next assessed the interaction of ASCT2 and Vpu by co-immunoprecipitation (co-IP). HEK293T cells were either mock transfected (Fig. 3B, lane 1) or transfected with pCMV-ASCT2-Flag (Fig. 3B, lane 2), pcDNA-Vphu (Fig. 3B, lane 3), or pCMV-ASCT2-Flag with pcDNA-Vphu (Fig. 3B, lane 4). After 24 hours of transfection, cells were lysed in co-IP buffer. A fraction of the lysate (10%) was used as input control. The remaining lysates were immunoprecipitated with anti-Flag beads. Input lysates and immunoprecipitated samples were separated by SDS-PAGE and processed for immunoblotting using antibodies to the Flag epitope in ASCT2, or antibodies to Vpu, or tubulin. As expected, immunoblot analysis of the input samples revealed the presence of ASCT2 in lanes 2 and 4 and Vpu in lanes 3 and 4. The loading control tubulin was constant in all samples. In the immunoprecipitated samples, ASCT2 was heavily enriched in lanes 2 and 4, as expected. Importantly, Vpu was detected only in the sample containing both Vpu and ASCT2 but not in the sample lacking ASCT2 (Fig. 3B, compare lanes 3 and 4). These results suggest that Vpu specifically interacts with ASCT2. Thus, results from confocal analysis in panel A and co-IP in panel B support the conclusion that ASCT2 physically interacts with Vpu.

Vpu affects the ASCT2 expression pattern

In Fig. 1, we showed that Vpu triggers the downmodulation of ASCT2 from the plasma membrane. To elucidate the underlying molecular mechanism, we next studied the effect of viral production on the expression of exogenous ASCT2. To that end, constant amounts of pNL43 WT (Fig. 4A, lanes 4–6), pAD8 WT (Fig. 4B, lanes 4–6), or empty vector (Fig. 4A and B, lanes 1–3) were transfected together with increasing amounts of pCMV-ASCT2-Flag. After 24 hours, whole-cell extracts were processed for immunoblotting. As expected, tubulin, which served as a loading control, remained constant in all samples. Similarly, ASCT2 had no effect on the expression of NL43 and AD8 Vpu or Gag proteins (Fig. 4A and B, lanes 4–6). Of note, expression of a 73 kDa form of ASCT2 increased in a dose-dependent manner in both NL43 and AD8 virus-producing and non-virus-expressing cells, with its overall intensity comparable to that of virus-producing and non-producing cells. However, in virus-producing cells, a shorter 46 kDa form of ASCT2 appeared with increasing amounts of ASCT2 (Fig. 4A and B, lanes 5 and 6). This 46 kDa form of ASCT2 was not obvious in non-virus-producing cells. Considering both the 46 kDa and 73 kDa forms of ASCT2, overall levels of total ASCT2 were dramatically increased in virus-producing cells, which can be attributed mainly to the 46 kDa form of ASCT2.

Fig 4.

Vpu affects the ASCT2 expression pattern. (A) HEK293T cells were transiently transfected with increasing amounts (0.00, 0.02, and 0.05 µg) of pCMV-ASCT2-Flag together with 5.0 µg of the empty vector or 5.0 µg of pNL43 WT molecular clone DNA. Total amounts of transfected plasmid DNA were kept constant using empty vector DNA, as needed. After 24 hours of transfection, cells were collected, and whole-cell extracts were processed for immunoblot analysis using antibodies to Flag, HIV-IG, Vpu, and tubulin (Tub). A representative immunoblot image from three independent experiments is shown. (B) HEK293T cells were transiently transfected with increasing amounts (0.00, 0.02, and 0.05 µg) of pCMV-ASCT2-Flag together with 5.0 µg of the empty vector or 5.0 µg of pAD8 WT. Total amounts of transfected plasmid DNA were kept constant using empty vector DNA, as needed. After 24 hours of transfection, cells were collected, and whole-cell extracts were processed for immunoblot analysis using antibodies to Flag, HIV-IG, Vpu, and tubulin (Tub). A representative immunoblot image from two independent experiments is shown. (C) HEK293T cells were transiently transfected with increasing amounts (2.0 and 5.0 µg, respectively) of pNL43 WT or pNL43 Udel together with 0.02 µg of pCMV-ASCT2-Flag. Total amounts of transfected DNA plasmid were adjusted to 5.0 µg for each sample using empty vector DNA, as needed. A sample without viral vector DNA was included as a negative control. After 24 hours of transfection, whole-cell extracts were processed for immunoblot analysis using antibodies to Flag, HIV-IG, Vpu, and tubulin (Tub). A representative immunoblot image from three independent experiments is shown. (D) HEK293T cells were transiently transfected with increasing amounts (2.0 and 5.0 µg, respectively) of pAD8 WT or pAD8 Udel together with 0.02 µg of pCMV-ASCT2-Flag. Total amounts of transfected DNA plasmid were adjusted to 5.0 µg for each sample using empty vector DNA, as needed. A sample without viral vector DNA was included as a negative control. After 24 hours of transfection, whole-cell extracts were processed for immunoblot analysis using antibodies to the Flag epitope in ASCT2, HIV-IG, Vpu, and tubulin (Tub). Digital images were acquired using BioRad Image Lab software. A representative immunoblot image from two independent experiments is shown.

 To see if the appearance of the 46 kDa form of ASCT2 was dependent on Vpu, we modified the above experiment such that constant amounts of ASCT2 were analyzed against increasing amounts of either NL43 WT (Fig. 4C, lanes 2 and 3) or NL43 Udel (Fig. 4C, lanes 4 and 5) or pAD8 WT (Fig. 4D, lanes 2 and 3) or pAD8 Udel (Fig. 4D, lanes 4 and 5) virus. Expression of ASCT2 in the absence of viral proteins served as control (Fig. 4C and D, lane 1). Results for NL43 and AD8 viruseswere very similar. As expected, the expression of viral proteins increased with increasing amounts of transfected DNA. Expression of the 73 kDa form of ASCT2 was relatively constant and only slightly higher expressed at the highest concentration of NL43 WT (Fig. 4C, lane 3) or AD8 WT (Fig. 4D, lane 3). In contrast, increased expression of NL43 WT and AD8 WT correlated with the appearance of large amounts of the 46 kDa form of ASCT2 (Fig. 4C and D, lanes 2 and 3). Interestingly, the expression of NL43 Udel and AD8 Udel resulted in only a modest increase in 46 kDa ASCT2 (Fig. 4C and D, lanes 4 and 5). These results suggest that the appearance of the 46 kDa form of ASCT2 is linked to the expression of Vpu.

Vpu inhibits the maturation of ASCT2

ASCT2 is a 541-residue multi-span integral membrane protein encoding two possible motifs for N-linked glycosylation (N163 and N212) located in the large extracellular loops (ECL) 2 a and 2b (see Fig. 1G) (42, 43). Since protein glycosylation can significantly affect the electrophoretic mobility of a protein, it seemed likely that the two forms of ASCT2 observed in Fig. 4 represent differentially glycosylated forms of ASCT2. To directly assess the effect of Vpu on the maturation of ASCT2, we performed a pulse-chase experiment in which we followed the fate of newly synthesized metabolically labeled ASCT2 over time. HEK293T cells were transfected with pCMV-ASCT2-Flag and either empty vector (Fig. 5A, lanes 1–4) or pcDNA-Vphu (Fig. 5A, lanes 5–8). After 24 hours, cells were collected, and 70% of samples were processed for pulse-chase analysis. The remaining 30% were processed for immunoblotting (Fig. 5C). For pulse-chase analysis, cells were then pulse-labeled for 10 minutes in medium containing [³⁵S]methionine and [³⁵S]cysteine and chased for up to 2 hours in the presence of excess unlabeled amino acids, as indicated. Cells were then lysed in co-IP buffer, immunoprecipitated with anti-Flag beads, and proteins bound to the Flag beads were visualized by fluorography. We observed two protein bands corresponding to the 46 kDa and 73 kDa forms of ASCT2. In the absence of Vpu, the 43 kDa form of ASCT2 gradually changed into the 73 kDa form, a process that is indicative of post-translational maturation and that was largely complete within 60 minutes (Fig. 5A, lanes 1–4). In contrast, in the presence of Vpu, the conversion of the 43 kDa form of ASCT2 to the 73 kDa form was much slower (Fig. 5A, lanes 5–8). Intensities of protein bands from three independent experiments were quantified by phosphoimage analysis, and the proportion of 73 kDa ASCT2 relative to total (46 kDa plus 73 kDa) ASCT2 was calculated and plotted as a function of time (Fig. 5B). In the absence of Vpu, ASCT2 rapidly converted to the 73 kDa form, which was 80% complete after 60 minutes of chase (Fig. 5B, green graph). In the presence of Vpu, however, conversion of the 46 kDa ASCT2 to the 73 kDa form was delayed and reached a plateau of about 40% conversion after 60 minutes, with the remaining 60% of ASCT2 never reaching maturity (Fig. 5B, red graph). To analyze the steady-state expression of ASCT2 and Vpu before the pulse-chase analysis, the remaining 30% of whole-cell extracts were prepared and processed for immunoblotting using antibodies to Flag (ASCT2), Vpu, and tubulin. Consistent with our previous results, in the absence of Vpu, we observed only one protein band, belonging to the 73 kDa form of ASCT2. However, in the presence of Vpu, we observed two protein bands corresponding to the 46 kDa and 73 kDa forms of ASCT2 (Fig. 5C, lane 2). These results suggest that Vpu prevents post-translational maturation of ASCT2. Of note, Vpu expressed from the full-length proviral plasmid was sufficient to affect ASCT2 maturation, indicating that the accumulation of the 46 kDa form of ASCT2 is not a consequence of Vpu over-expression.

Fig 5.

Vpu inhibits the maturation of ASCT2. (A) HEK293T cells were transiently transfected with 0.1 of pCMV-ASCT2-Flag and 4.90 µg of empty vector or pCDNA-Vphu WT. After 24 hours post-transfection, two-thirds of the transfected HEK293T cells were pulse-labeled with [³⁵S]-Expre³⁵S³⁵S-label for 10 minutes, chased for 0, 30, 60, and 120 minutes, and processed for immunoprecipitation using Flag-beads (EZview Red ANTI-FLAG Affinity gel [Sigma Aldrich]), as described in Materials and Methods. Immunoprecipitated samples were separated by SDS-PAGE and visualized by photofluorography using a Typhoon FLA 9500 Phosphorimager. A representative image from three independent experiments is shown. (B) Protein bands corresponding to Flag-tagged ASCT2 in panel A were quantified using GraphPad Prism software and plotted as a function of time. Error bars represent the standard error of the mean (SEM) from three independent experiments. The statistical significance was determined using a two-way ANOVA with a Tukey multiple comparison test. P: ***, ≤0.001; ****, ≤0.0001. (C) Cells from panel A not used for the pulse/chase were used to prepare whole-cell extracts, which were used for immunoblot analysis using antibodies to tubulin (Tub), the Flag epitope in ASCT2 (Flag), and Vpu. Lane 1 corresponds to lanes 1–4 in panel A; lane 2 represents lanes 5–8 in panel A. Digital images were acquired using BioRad Image Lab software. A representative immunoblot image from three independent experiments is shown. (D) HEK293T cells were transfected with 0.1 of pCMV-ASCT2-Flag and 4.90 µg of pCDNA-Vphu WT. After 24 hours post-transfection, cells were pulse-labeled with [³⁵S]-Expre³⁵S³⁵S-label for 10 minutes and chased for 60 minutes. Cells were lysed in IP buffer, divided into three equal aliquots, and immunoprecipitated using Flag-beads as in panel A. Immunoprecipitated samples were either left untreated or treated as follows: Immunoprecipitated samples were first fully denatured with denaturing buffer for 10 minutes at 65°C according to the manufacturer’s instructions and then treated with 0.2 U of Endo H or 7 U of PNGase F for 3 hours at 37°C. Samples were then subjected to SDS-PAGE and analyzed by fluorography. A representative result from three independent experiments is shown. (E) HEK293T cells were transiently transfected with 0.1 of pCMV-ASCT2-Flag and 4.90 µg of empty vector (Ctrl), pCDNA-Vphu WT (WT), pCDNA-Vphu 2/6 (2/6), or pCDNA-Vphu RD (RD). After 24 hours post-transfection, two-thirds of the transfected HEK293T cells were pulse-labeled with [³⁵S]-Expre³⁵S³⁵S-label for 10 minutes, chased for 0, 10, 30, and 60 minutes, and processed for immunoprecipitation using Flag-beads (EZview Red ANTI-FLAG Affinity gel [Sigma Aldrich]), as described in Materials and Methods. Immunoprecipitated samples were separated by SDS-PAGE and visualized by photofluorography using a Typhoon FLA 9500 Phosphorimager. A representative image from two independent experiments is shown. (F) Cells from panel E not used for the pulse/chase were used to prepare whole-cell extracts, which were used for immunoblot analysis using antibodies to tubulin (Tub), the Flag epitope in ASCT2 (Flag), and Vpu. Digital images were acquired using BioRad Image Lab software. A representative immunoblot image from three independent experiments is shown.

To further confirm that Vpu prevents the maturation of newly synthesized core glycosylated ASCT2, we performed endoglycosidase analysis using endoglycosidase H (Endo H) and peptide: N glycosidase F (PNGase F) enzymes. Core-glycosylated newly synthesized proteins remain sensitive to Endo H, while all proteins are sensitive to the PNGase F activity (44). EndoH and PNGaseF analysis revealed that both forms of ASCT2 are PNGaseF-sensitive and thus glycosylated (Fig. 5D, lane 3), but only the 46 kDa form was sensitive to EndoH treatment (Fig. 5D, lane 2) and thus carried high mannose modified carbohydrates. We conclude that ASCT2 is co-translationally modified by N-linked glycosylation and undergoes a maturation process that includes modifications of the carbohydrate side chains. This presumably happens as the protein travels from the ER, via the Golgi network, to the plasma membrane. In the presence of Vpu, maturation of ASCT2 is severely delayed, and much of the protein remains in an EndoH-sensitive high mannose state. This suggests that Vpu inhibits the normal trafficking of ASCT2 from the ER to the cell surface.

Vpu consists of a single transmembrane α-helical domain (TMD) and a cytoplasmic domain formed by two α-helices connected by an acidic linker region, which contains two serines (S52 and S56) that are constitutively phosphorylated (45). The Vpu TMD and/or the phosphoserine-acidic cluster (or PSAC) of the cytoplasmic domain supports the interaction of Vpu with distinct cellular complexes involved in protein degradation (reviewed in reference 40). To further elucidate the relevance of the Vpu TMD domain and the PSAC motif on ASCT2 maturation, we performed another pulse-chase experiment of two well-characterized TMD and PSAC Vpu mutants, called Vpu RD and Vpu S2/6A, respectively. The Vpu RD mutant stands for "randomized domain," in which the primary amino acid sequence of the TM domain is scrambled while maintaining the same overall amino acid composition as the wild-type Vpu protein (46). On the other hand, the Vpu S2/6A mutant is a variant where serine residues 52 and 56 in the PSAC motif are mutated to alanine residues (47). HEK293T cells were transfected with pCMV-ASCT2-Flag and either empty vector, pcDNA-Vphu WT, pcDNA-Vphu S2/6A, or pcDNA-Vphu RD mutants. After 24 hours, cells were collected and processed for pulse-chase analysis, as described for Fig. 5A. As expected, in the absence of Vpu, the 43 kDa form of ASCT2 gradually changed into the 73 kDa form (Fig. 5E, Ctrl panel). However, in the presence of Vpu, the conversion of the 43 kDa to the 73 kDa form was much slower, consistent with the results from Fig. 5A (Fig. 5E, WT panel). Interestingly, expression of the Vpu S2/6A mutant also delayed the conversion of the 43 kDa form of ASCT2 to the 73 kDa form (Fig. 5E, U2/6 panel). In contrast, the Vpu RD mutant only partially allowed the conversion of the 43 kDa to the 73 kDa form (Fig. 5E, Urd panel), similar to what was observed in the Ctrl panel. Immunoblot analysis was done as in Fig. 5C and confirmed the inhibitory effect of Vpu WT and Vpu S2/6A on ASCT2 maturation. The Vpu RD mutant, on the other hand, looked more like the Vpu-negative control sample (Fig. 5F, compare lanes 1 and 4). These results suggest that the structural integrity of the Vpu TM domain, but not the PSAC region, is critical for inhibition of ASCT2 maturation in the ER and Golgi and subsequent trafficking to the plasma membrane.

ASCT2 interacts with the Env precursor gp160

Previous studies found that ASCT2 interacts with Env proteins from a variety of retroviruses that cause infectious immunodeficiencies, such as human endogenous virus type W (HERV-W) (48–51). We used immunofluorescence and co-IP analyses to test if ASCT2 also interacts with HIV-1 Env. For immunofluorescence studies, HeLa cells were transfected with ASCT2-Flag together with pNLA1 for the expression of Env. After 24 hours, cells were fixed and incubated with antibodies to Flag (ASCT2) or Env. Confocal microscopy was done, as in Fig. 3A. ASCT2 staining is shown in red, Env staining appears in green, and the overlap of ASCT2 and Vpu staining is depicted in yellow. The nuclei were visualized by DAPI staining and appear in blue (Fig. 6A). HIV-1 Env protein, as well as exogenously expressed ASCT2, exhibited cytoplasmic fluorescence with accumulation in a perinuclear compartment, which most likely represents the Golgi complex. Indeed, the merged panel of Fig. 6A shows an overlap of the ASCT2 and Env signals in the perinuclear region of the cells, similar to the colocalization of ASCT2 and Vpu in Fig. 3A. This suggests that ASCT2 and HIV-1 Env are in close proximity and may interact in the perinuclear region, presumably the Golgi complex, of transfected HeLa cells.

Fig 6.

ASCT2 interacts with the Env precursor gp160. (A) HeLa cells were transiently transfected with 0.2 µg of pCMV-ASCT2-Flag and 0.6 µg of pNLA1 DNA expressing Env and all accessory proteins, including Vpu, but not Gag or Pol genes. After 24 hours of transfection, cells were fixed with 4% PFA, permeabilized with 1% Triton X-100, and stained with anti-Flag (mouse) and anti-Env (rabbit) primary antibodies, followed by staining with Alexa Fluor 488 (rabbit; green) and Alexa Fluor 594 (mouse; red) secondary antibodies. A representative IFI image is shown. PFA: paraformaldehyde. Regions of overlap appear as yellow. (B) HEK293T cells were transiently transfected with 0.5 µg of pCMV-ASCT2-Flag and 4.5 µg of pNLA1 or pcDNA-Vphu, as indicated at the top. Total amounts of transfected DNA plasmid were adjusted to 5.0 µg for each sample using empty vector DNA as needed. Mock-transfected cells were included as a negative control. After 24 hours of transfection, cells were lysed in X-100 lysis buffer. An aliquot (10%) of the lysate was used as input control (Input). The remaining extracts were immunoprecipitated using anti-Flag beads for 1 hour at 4°C. Cell extracts (Input) and immunoprecipitated samples (IP: α-Flag) were separated by SDS-PAGE and analyzed by immunoblotting using anti-Flag, anti-Env, anti-Vpu, and anti-GAPDH antibodies, as indicated on the right. Digital images were acquired using BioRad Image Lab software. A representative immunoblot image from three independent experiments is shown. (C) HEK293T cells were transfected with 0.5 µg pCMV-ASCT2-Flag together with 3.0 µg of pcDNA-Vphu (Vpu), and increasing amounts of pNLA1 Udel (0.0, 1.5, and 3.0 µg) (Env). Total amounts of transfected plasmid DNA were adjusted to 6.5 µg for each sample using empty vector DNA as needed. Mock-transfected cells were included as a negative control. After 24 hours of transfection, cells were lysed in X-100 lysis buffer. An aliquot (10%) of the lysate was used as input control (Input). The remaining extracts were immunoprecipitated using anti-Flag beads for 1 hour at 4°C. Cell extracts (Input) and immunoprecipitated samples (IP: α-Flag) were separated by SDS-PAGE and analyzed by immunoblotting using anti-Flag, anti-Env, anti-Vpu, and anti-GAPDH antibodies, as indicated on the right. Digital images were acquired using BioRad Image Lab software. A representative immunoblot image from three independent experiments is shown.

To study the interaction of ASCT2 and Env by co-IP, HEK293T cells were transfected with pCMV-ASCT2-Flag together with either pNLA1 (expressing Vpu and Env) (Fig. 6B lanes 3 and 5), pcDNA-Vphu (expressing Vpu only; Fig. 6B, lanes 4 and 6), or empty vector (Fig. 6B, lane 2). A mock-transfected sample was included as a control (Fig. 6B, lane 1). HEK293T cells were processed for immunoprecipitation with Flag antibodies, as in Fig. 3B. A fraction of the cell lysates was used as the input control. Cell extracts and immunoprecipitated proteins were separated by SDS PAGE, and membranes were probed with antibodies to Flag (ASCT2), Env (gp160, gp120), Vpu, and GAPDH (input sample only), as indicated. Env expression from pNLA1 was efficient, while levels of Vpu were too low for quantitative analysis (Fig. 6B, lanes 3 and 5). However, Vpu was efficiently expressed from codon-optimized pcDNA-Vphu (Fig. 6B, lanes 4 and 6). Interestingly, both gp160 (expressed from pNLA1) (Fig. 6B, lane 5) and Vpu (expressed from pcDNA-Vphu) (Fig. 6B, lane 6) specifically co-precipitated with ASCT2, verifying the physical interaction of ASCT2 with both Vpu and gp160. Thus, results from confocal analysis in panel A and co-IP in panel B support the conclusion that ASCT2 physically interacts with the Env precursor gp160.

To study if Vpu competes for the interaction of ASCT2 and Env, HEK293T cells were transfected with pCMV-ASCT2-Flag together with pcDNA-Vphu (expressing Vpu only: Fig. 6C, lanes 2 and 7–9), increasing amounts of pNLA1 Udel (expressing Env but not Vpu; Fig. 6C, lanes 5 and 6 and 8 and 9), or empty vector (Fig. 6C, lane 3). A mock-transfected sample was included as a control (Fig. 6C, lane 1). HEK293T cells were processed for immunoprecipitation with anti-Flag beads, as described for Fig. 6B. A fraction of the cell lysates was used as the input control. Cell extracts and immunoprecipitated proteins were separated by SDS PAGE, and membranes were probed with antibodies to Flag (ASCT2), Env (gp160 and gp120), Vpu, and GAPDH (input sample only), as indicated. Interestingly, both gp160 (expressed from pNLA1 Udel) (Fig. 6C, lane 9) and Vpu (expressed from pcDNA-Vphu) (Fig. 6C, lanes 7–9) co-precipitated with ASCT2 when the three proteins are co-expressed together. Furthermore, the expression of higher amounts of Env increased the expression of ASCT2 by a mechanism that is not yet fully understood (Fig. 6C, lane 9). We conclude that Vpu and Env interact with ASCT2 in a non-competitive manner. In fact, it is possible that Env, Vpu, and ASCT2 form a ternary complex.

ASTC2 promotes the incorporation of uncleaved gp160 Env precursor into viral particles

Inhibiting the normal trafficking of ASCT2 could explain the Vpu-induced cell surface downmodulation of ASCT2 observed in Fig. 1F. A possible benefit of Vpu-mediated depletion of ASCT2 from the cell surface could be that amounts of ASCT2 available for packaging into progeny virions would be limited. To test this possibility, we performed OptiPrep gradient centrifugation of cell-free viral supernatants (Fig. 7). For that purpose, ASCT2 was expressed in HEK293T cells in the absence of virus production or together with pNL43 WT or pNL43 Udel. Virus-containing supernatants were collected 24 hours after transfection and concentrated by ultracentrifugation, as detailed in the Methods section. Pelleted virus was subjected to OptiPrep gradient centrifugation. Thirteen equal fractions were collected from the top of the gradient and analyzed by immunoblotting (Fig. 7A). Under the conditions of this experiment, virus-associated capsid (CA) and Env proteins were enriched in fractions 5–8 (Fig. 7A, red box). ASCT2 was identified in concentrated supernatants as well, even in the absence of virus production (Fig. 7A, no virus). Both the mature and immature forms of ASCT2 were observed in the concentrated supernatants. However, the immature form of ASCT2 primarily accumulated in the viral fractions (fractions 5–8), while the mature form was enriched in lighter fractions (fractions 2–6). This suggests that ASCT2 is secreted from transfected HEK293T cells irrespective of whether they produce virus or not. For better comparison, fractions 5–8 of the three gradients were reanalyzed on a single gel (Fig. 7B). Quantitation of ASCT2 in fractions 5–8 of NL43 WT and Udel virus relative to CA signals confirmed increased levels of primarily immature ASCT2 in Udel virus when compared to the WT virus (Fig. 7C). Furthermore, levels of virus-associated gp160 relative to gp120 were higher in NL43 Udel virus relative to WT virus (Fig. 7D). These results suggest that NL43 Udel virus contains higher levels of gp160 compared to the WT virus, while levels of gp120 are comparable between both viruses. Taken together, we conclude that in Vpu-deficient viruses, the expression of ASCT2 results in increased packaging of both gp160 and immature ASCT2.

Fig 7.

ASCT2 promotes the incorporation of uncleaved gp160 Env precursor into viral particles. HEK293T cells were transiently transfected with 0.15 µg of pCMV-ASCT2-Flag and either 4.85 µg of empty vector DNA (No Virus), NL43 WT, or NL43 Udel. Empty vector DNA was used to adjust the total amount of transfected DNA plasmid to 5.0 µg, as needed. After 24 hours of transfection, culture supernatants were collected, filtered, and concentrated by ultracentrifugation. The concentrated virus was subjected to OptiPrep density gradient centrifugation as described in Materials and Methods, and 13 equal fractions were collected from the top of the gradient. Fractions were analyzed by immunoblotting using antibodies to Env, the Flag epitope in ASCT2, as well as HIV-Ig. Digital images were acquired using BioRad Image Lab software. A representative immunoblot image from three independent experiments is shown. Fractions enriched in the viral capsid protein are boxed in red. (B) To allow direct comparison of relative protein profiles between the gradients, fractions 5 to 8 of the three gradients from panel A were run on a single gel. Proteins were transferred to a PVDF membrane, which was subsequently cut into three sections to allow probing with antibodies to HIV-1 Env (top section; >90 kDa), Flag (middle section; 35 to 90 kDa), and HIV-Ig (bottom section; <35 kDa). A representative gel of three independent experiments is shown. Digital images were acquired using Image Lab software (BioRad). A representative immunoblot image from three independent experiments is shown. (C) ASCT2 and p24 specific signals from all four fractions of the NL43 WT and NL43 Udel samples in panel B were combined and quantified using GraphPad Prism software. To visualize the expression of ASCT2 relative to CA, the ratio of ASCT2 over CA signals was calculated. The graph integrates results from three independent OptiPrep gradients. (D) The relative levels of gp160 and gp120 in the viral fractions were similarly quantified by calculating the ratio of gp160 over gp120. The graph integrates results from three independent OptiPrep gradients. The statistical significance was determined using a one-tailed paired t-test. P: * ≤0.05.

ASCT2 interacts with HIV-1 Env in cell-free viruses

To test if the packaging of ASCT2 into viral particles involves an interaction with viral Env protein, we performed a co-IP analysis of concentrated supernatants from transfected HEK293T cells. Cells were transfected with ASCT2 either alone (Fig. 8A, lane 2) or together with pNL43 WT (Fig. 8A, lane 3) or pNL43 Udel (Fig. 8A, lane 4). An aliquot (10%) of each sample was analyzed directly by immunoblotting (Fig. 8A); the remaining samples were immunoprecipitated using an Env-specific polyclonal antibody (Fig. 8B). Immunoprecipitation of supernatants with the Env-specific antibody led to the enrichment of both gp120 and uncleaved gp160 from the virus preparations (Fig. 8B, lanes 7 and 8). Importantly, ASCT2 was immunoprecipitated by the Env antibody in the samples containing Env protein (Fig. 8B, lanes 7 and 8), but not in the sample lacking Env (Fig. 8B, lane 6). These results indicate that ASCT2 is present in virus particles in a complex with HIV-1 Env. The presence of increased levels of gp160 and ASCT2 in samples lacking Vpu (Fig. 8B, compare lanes 7 and 8) could suggest that Vpu functions to limit the packaging of gp160 and ASCT2 into virus particles.

Fig 8.

ASCT2 interacts with Env in cell-free viruses. (A) HEK293T cells were transiently transfected with 0.15 µg of pCMV-ASCT2-Flag and 4.85 µg of pNL43 WT or NL43. Empty vector DNA was used to adjust the total amount of transfected DNA plasmid to 5.0 µg, as needed. Mock-transfected cells were used as a negative control. After 24 hours of transfection, cells and virus-containing supernatants were collected. A fraction (10%) of the concentrated viruses was analyzed by immunoblotting to verify the expression of HIV-1 Env (gp160/120), ASCT2, Gag (Pr55, CA), and Vpu. A tubulin (tub) blot was included to demonstrate the absence of contaminating cell-associated proteins in the cell extracts. (B) The remaining viral samples were dissolved in NP-40 lysis buffer and immunoprecipitated with Env-specific antibodies, followed by immunoblot analysis of the precipitates with antibodies to ASCT2 (WB:Flag) and Env (WB:Env). Digital images were acquired using BioRad Image Lab software. A representative immunoblot image from three independent experiments is shown.

ASCT2 affects HIV-1 infectivity

So far, it has been shown that ASCT2 promotes the incorporation of the Env precursor gp160 into the viral particles when Vpu is not expressed. To evaluate if the incorporation of the uncleaved gp160 together with ASCT2 affects the infectivity of viral particles, we performed an infectivity assay in which the effect of increasing amounts of ASCT2 on viral infectivity was tested. For that, HEK293T cells were transfected with constant amounts of NL43 WT or pNL43 Udel, together with increasing amounts of pCMV-ASCT2-Flag. Cell extracts and concentrated supernatants were subjected to immunoblot analysis to assure comparable protein expression (Fig. 9A). Cell-free, filtered supernatants were then used to infect reporter TZM-bl cells, and the Tat-dependent induction of luciferase expression was determined in a standard luciferase assay. Results from triplicate infections of six independent experiments (i.e., 18 samples total) were plotted as a function of the ASCT2 concentration (Fig. 9B). Interestingly, ASCT2 significantly inhibited the infectivity of Vpu-defective virus in a dose-dependent manner (Fig. 7B, red circles). At the highest level of ASCT2, the difference in infectivity of WT and Vpu-defective virus was almost twofold and highly statistically significant.

Fig 9.

ASCT2 affects HIV-1 viral infectivity. (A) HEK293T cells were transiently transfected with increasing amounts of pCMV-ASCT2-Flag (0.00, 0.05, 0.15, and 0.25 µg) and 4.5 µg of pNL43 WT or pNL43 Udel. Total amounts of transfected DNA plasmid were adjusted to 5.0 µg for each sample using empty vector DNA, as needed. After 24 hours of transfection, whole-cell extracts and pelleted virions were subjected to SDS-PAGE and analyzed by immunoblotting using antibodies to the Flag epitope tag in ASCT2, as well as to HIV-1 Env, HIV-IG, Vpu, or tubulin (Tub). Digital images were acquired using BioRad Image Lab software. A representative immunoblot from three independent experiments is shown. (B) TZM-bl indicator cells (CD4⁺, CCR5⁺, and CXCR4⁺) were plated in 24-well plates (150,000 cells/well) and then infected in triplicate with 150 µL each of unconcentrated filtered supernatants from panel A. Input virus was quantified by RT assay. After 48 hours of infection, TZM-bl cells were lysed in 250 µL of Reporter Lysis Buffer (Promega). Luciferase activity was determined by combining 15 µL of the cell lysate with 50 µL of Steady Glo Luciferase Substrate (Promega). The RLU of each sample was measured using a GloMax Microplate Reader (Promega). Results were corrected for differences in the input viral RT activity of each input virus. The statistical significance was determined using two-way ANOVA with a Šidák multiple comparison test (GraphPad Prism software). P: *, ≤0.05; ****, ≤0.0001. RLU: Relative Light Unit, cpm: counts per minute. (C) HEK293T cells were transiently transfected with increasing amounts of pCMV-ASCT2-Flag (0.00, 0.05, 0.15, and 0.25 µg) and 4.5 µg of pNL43 Udel. One set of samples was processed in the absence of VSVg, while the second set was cotransfected with 0.5 µg of pCMV-VSVg DNA. Immunoblot analyses were done as in panel A. (D) Infectivity analyses of TZM-bl indicator cells were done as in panel B. The statistical significance was determined using two-way ANOVA with a Tukey multiple comparison test (GraphPad Prism software). P: **, ≤0.01; ****, ≤0.0001; ns, no significance. RLU: relative light unit.

If ASCT2 affects the biological function of HIV-1 Env, we should be able to restore viral infectivity by pseudotyping with VSVg. To this end, we produced VSVg pseudotyped viruses by transfecting HEK293T cells with constant amounts of pNL43 Udel and pVSVg plasmids, together with increasing amounts of pCMV-ASCT2-Flag. Whole-cell extracts and concentrated viral supernatants were assessed for protein expression (Fig. 9C). Filtered, cell-free supernatants were used to infect reporter TZM-bl cells to determine viral infectivity. Results from triplicate infections of three independent experiments (i.e., nine samples total) were plotted as a function of the ASCT2 concentration (Fig. 9D). As expected, co-expression of VSVg did not affect the expression and release of viral Gag proteins. However, VSVg expression was associated with a reduction of virus-associated ASCT2 as well as virus-associated gp160. Importantly, co-expression of VSVg restored virus infectivity. These results suggest that inhibition of viral infectivity in ASCT2-expressing cells is either due to the increased levels of gp160 or the presence of ASCT2 in cell-free virions.

HIV-1 transmitted/founder (T/F) viruses expressing different Vpu variants exhibit different sensitivity to ASCT2

To see if the effect of ASCT2 on viral infectivity is specific for the laboratory-adapted NL43 isolate, we tested a panel of five transmitted/founder viruses obtained through the NIH HIV Reagent Program. Individual clones were sequenced across the vpu region, and all clones were found to have intact Vpu ORFs with significant variations observed between T/F isolates (Fig. 10A). To assess the effect of ASCT2 on viral protein production and viral infectivity, HEK293T cells were transfected with pCMV-ASCT2-Flag and individual T/F clones. After 24 hours of transfection, cell extracts and concentrated supernatants were subjected to immunoblot analysis to assure comparable protein expression (Fig. 10B). All clones expressed comparable levels of Gag proteins, except for T/F #7, which expressed reduced amounts of Gag and Env proteins. (Fig. 10B, lanes 9 and 10). Of note, T/F clones #4 and #6 did not express detectable levels of Vpu, despite the presence of intact Vpu ORFs (Fig. 10A and B lanes 3, 4, 7, and 8). It is not clear whether the apparent absence of Vpu is caused by poor antibody recognition or is caused by mutations in the viral genome outside the Vpu ORF that affect Vpu expression. Virus-containing supernatants were used to infect reporter TZMbl cells to determine viral infectivity (Fig. 10C). Relative infectivity values obtained in the absence of ASCT2 were defined as 100% for each T/F virus. Consistent with the results from Fig. 9B, ASCT2 expression slightly increased the infectivity of NL43 WT. Similar results were observed for T/F #4, T/F #7, and T/F #9. Of note, T/F #5 and T/F #6 exhibited significantly lower infectivity when ASCT2 was expressed. In particular, the infectivity of T/F #6, which does not exhibit Vpu expression, was reduced about fivefold. The Vpu protein expressed by T/F #5 exhibited altered mobility, which may be indicative of a functional defect. Taken together, our results demonstrated that different HIV-1 isolates encoding different Vpu variants can exhibit differential sensitivity to ASCT2.

Fig 10.

HIV-1 transmitted/founder (T/F) viruses expressing different Vpu variants exhibit different sensitivity to ASCT2. (A) Sequence of the vpu genes of the T/F viruses #4, #5, #6, #7, and #9. All clones were found to have intact Vpu ORFs. (B) HEK293T cells were transiently transfected with 0.01 µg of pCMV-ASCT2-Flag and 5.0 µg of pNL43 WT, T/F #4, T/F #5, T/F #6, T/F #7, and T/F #9. Total amounts of transfected DNA plasmid were adjusted to 5.0 µg for each sample using empty vector DNA as needed. After 24 hours of transfection, whole-cell extracts and pelleted virions were subjected to SDS-PAGE and analyzed by immunoblotting using antibodies to the Flag epitope tag in ASCT2, as well as to HIV-1 Env, HIV-IG, Vpu, or tubulin (Tub). A representative immunoblot from three independent experiments is shown. (C) TZM-bl indicator cells (CD4⁺, CCR5⁺, CXCR4⁺) were plated in 24 well plates (150.000 cells/well) and then infected in triplicate with 75 µL each of unconcentrated filtered supernatants from panel B. Input virus was quantified by RT assay. After 48 hours of infection, TZM-bl cells were lysed in 250 µL of Reporter Lysis Buffer (Promega). Luciferase activity was determined by combining 15 µL of the cell lysate with 50 µL of Steady Glo Luciferase Substrate (Promega) and incubation for 5 minutes BST at RT. The RLU of each sample was measured using a GloMax Microplate Reader (Promega). Results were corrected for differences in input viral RT activity of each input virus. The statistical significance was determined using one-way ANOVA with a Šidák multiple comparison test (GraphPad Prism software). P: **, ≤0.01; ****, ≤0.0001; ns: no significance. RLU: relative light unit.

DISCUSSION

Complex organisms have evolved an innate immune system that is made up of a series of restriction factors that are either constitutively expressed or induced by IFN after infection (reviewed in references 52–55). It has been previously shown that IFN-inducible proteins represent a critical part of the host’s innate immune response by interfering with various stages of the HIV-1 replication cycle (41). Furthermore, it has been previously shown that amino acid transporters can modify HIV-1 replication through different mechanisms, such as the serine amino acid transporters SERINC3 and SERINC5 (1) and the alanine amino acid transporter SNAT1 (2). Results from our present study identify the neutral amino acid transporter ASCT2 as an IFN-inducible plasma membrane protein that inhibits HIV-1 viral infectivity. How does ASCT2 affect viral fitness? Overall, our data support the conclusion that the expression of ASCT2 in cells expressing HIV-1 results in the accumulation of immature gp160 Env and immature ASCT2, which can be packaged into virions as a gp160/ASCT2 complex, which presumably reduces the biological activity of Env. As such, the mechanism of action of ASCT2 against HIV-1 appears to be similar to the one exploited by the GTPases GBP5 and GBP2 (56–59), as well as M2BP or 90K (60) proteins, all of which promote the incorporation of non-functional gp160 into HIV-1 viral particles.

As with other host restriction factors, HIV-1 has found a way to antagonize the inhibitory effect of ASCT2. Indeed, we identified Vpu as a viral factor in charge of ASCT2 antagonism. Of note, Vpu was already reported to target another amino acid transporter, SNAT1, by promoting its downmodulation through the β-TrCP proteasomal degradation pathway (2). In the case of ASCT2, we now report that Vpu affects the intracellular trafficking of ASCT2, resulting in the accumulation of immature protein in the ER and the concomitant downmodulation of ASCT2 from the cell surface.

Interestingly, in the case of PBL, both wild-type and Vpu-defective viruses were able to reduce ASCT2 expression from the surface of CD4⁺ T cells (Fig. 1B). It is conceivable that the ability to reduce the surface expression of ASCT2 in these cells is redundant and can be exerted by additional viral factors. This is reminiscent of the ability of HIV-1 Vpu and Nef to independently downmodulate the cell surface expression of CD4. Similarly, the antiviral activity of BST-2 can be antagonized by Vpu, Nef, and even Env in a virus- and host-specific manner (reviewed in reference 61).

As noted above, ASCT2 is responsible for amino acid transport across the plasma membrane, with glutamine being the preferred substrate (3). In addition, ASCT2 is a major regulator of glutamine transport in primary HIV-1 target cells, such as activated CD4⁺ T cells (5–7). Our results suggest that Vpu modifies glutamine metabolism in actively dividing CD4⁺ T cells by targeting the glutamine transporter ASCT2. Our data show that ASCT2 plasma membrane expression correlates with a higher glutamine uptake. Consequently, Vpu-dependent downmodulation of ASCT2 dysregulates the uptake of the essential amino acid glutamine. The biological impact of ASCT2 downmodulation and subsequent dysregulation of glutamine uptake remains to be investigated. Similarly, Vpu downmodulates the expression of the alanine transporter SNAT1 to antagonize the alanine SNAT1-dependent transport, which modulates T cell activation (2). Therefore, downmodulation of ASCT2 and SNAT1 defines a new paradigm of HIV-1 interference with immunometabolism mediated by the accessory protein Vpu.

How does Vpu downmodulate ASCT2? To address that question, we studied the processing and trafficking of ASCT2 within cells expressing Vpu. Our Co-IP, IFA, pulse-chase, and endoglycosidase analyses suggest that Vpu physically interacts with ASCT2, presumably in the ER and Golgi, inhibiting the core-N-glycosylation and maturation of newly synthesized ASCT2 molecules and subsequent trafficking of the transporter to the plasma membrane. Preliminary studies of the Vpu TM domain and the phosphoserine-acidic cluster (or PSAC) in the cytoplasmic domain showed that the structural integrity of the Vpu TM domain, but not the PSAC region, is critical for inhibition of ASCT2 maturation in the ER and Golgi. Taken together, our data suggest that Vpu might be exploiting a mechanism resembling the downmodulation of CD4 by Vpu. Experiments are ongoing to define in more detail the exact regions in Vpu and ASCT2 involved in their physical interaction.

Our analysis of T/F viruses revealed that the sensitivity to ASCT2 is not limited to the lab-adapted NL43 isolate. There is significant variation in the Vpu sequences of individual T/F strains, although all isolates have full-length intact ORFs. Maybe not surprisingly, there is also significant variation in the sensitivity of T/F isolates to ASCT2. T/F viruses #4, #7, and #9 behave like NL43 WT, and viral infectivity is slightly upregulated in the presence of ASCT2, consistent with the results presented in Fig. 9B. However, not all T/F variants express detectable Vpu proteins. Particularly, T/F #6 shows no Vpu expression at all and behaves functionally like the Vpu-defective NL43 variant, as shown in Fig. 9B. Of note, the relative sensitivity of T/F #6 to ASCT2 was significantly higher than that of NL43 WT or even the Vpu-deficient NL43. Thus, the sensitivity of HIV-1 to ASCT2 is not exclusive to the laboratory-adapted NL43 isolate but instead is a common feature of other HIV-1 strains, even though the degree of inhibition is dependent on the virus isolate.

Taken together, we propose a model, as outlined in Fig. 11. The model assumes two different scenarios, one in which Vpu is absent and one in which Vpu is present. In the absence of Vpu, ASCT2 is N-glycosylated in the ER and Golgi and matures while trafficking through the N-glycosylation pathway to reach the plasma membrane (a). In that scenario, a fraction of the HIV-1 Env precursor gp160 exits the ER and traffics toward the cell surface. On the way, it is cleaved by furin at the Golgi complex into the active gp120/gp41 complex. After reaching the cell surface, the mature gp120/gp41 complex can be packaged into nascent virions (b). In parallel, molecules of immature ASCT2 can interact with the precursor gp160 in the ER. This complex of uncleaved gp160 and immature ASCT2 travels through the secretory pathway and reaches the virus assembly sites at the plasma membrane. The gp160/ASCT2 complex is incorporated into viral particles, thereby reducing viral infectivity (c).

Fig 11.

Proposed model. The details of the proposed model are explained in Discussion. The figure was created using the software BioRender.

In the second scenario, Vpu is synthesized alongside ASCT2 and Env at the ER and can interact with de novo synthesized immature ASCT2. This may have two major consequences. First, binding of Vpu to ASCT2 prevents or delays the exit of ASCT2 from the ER, thereby inhibiting the maturation of ASCT2. Second, the interaction between Vpu and immature ASCT2 interferes with the association of gp160 and ASCT2, allowing gp160 to be cleaved into the active complex gp120/gp41 by furin. The active complex of Env can reach the assembly site at the plasma membrane, where it is incorporated into viral particles to produce infectious virions (d).

Currently, new antiviral therapies are being developed that target important interactions between viral and host proteins, such as the interaction between Vpu and BST-2 (62) and Vif and APOBEC3 (63). Therefore, the interaction between ASCT2 and Vpu represents a promising strategy to develop a novel treatment for HIV. And, as mentioned before, the downmodulation of ASCT2 defines a new paradigm of HIV-1 interference with immunometabolism mediated by the accessory protein Vpu.

MATERIALS AND METHODS

Cell culture

HEK293T, HeLa, and TZM-bl cell lines were maintained in DMEM (Dulbecco’s modified Eagle’s medium) with 4.5 g/L D-glucose (ThermoFisher Scientific, Waltham, MA), supplemented with 10% heat-inactivated fetal bovine serum (Phoenix Scientific, Laguna Niguel, CA), 100 U/mL penicillin, 100 µg/mL streptomycin, and 2 mM L-glutamine, at 37°C in a humidified 5% CO₂ atmosphere. CEM-SS cells were grown in RPMI 1640 (Roswell Park Memorial Institute) supplemented with 10% heat-inactivated fetal bovine serum, 100 U/mL penicillin, and 100 µg/mL streptomycin, at 37°C in a humidified 5% CO₂ atmosphere. Elutriated human monocytes from multiple anonymous healthy donors were obtained from the NIH blood bank under protocol 99 CC-0168: “Collection and Distribution of Blood Components from Healthy Donors for In Vitro Research Use.” Human monocytes were maintained in six well-plates (2 × 10⁶ cells/well) with DMEM with 4.5 g/L D-glucose, supplemented with 10% pooled human AB serum (Gemini Bio-Products, West Sacramento, CA), 100 U/mL penicillin, 100 µg/mL streptomycin, 2 mM L-glutamine, and 1 mM sodium pyruvate (ThermoFisher Scientific, Waltham, MA), for 7 days at 37°C in a humidified 5% CO₂ atmosphere, to allow differentiation into macrophages (MDM).

Plasmids and viral vectors

The pCMV6-SLC1A5-Myc-DDK DNA plasmid encoding ASCT2 (referred to here as pCMV-ASCT2-Flag) was purchased from OriGene (Origene Technologies, Rockville, MD). A variant of pCMV6-SLC1A5-Myc-DDK carrying an HA epitope tag (YPYDVPDYA) in the protruding loop between TM3 and TM4a was created by PCR amplification of pCMV6-SLC1A5-Myc-DDK with the 5′ primer 5′-GCCATCAACG CCTCCGTGGG AGCCTACCC ATACGATGTT CCAGATTACG CTGCGGGCAG TGCCGAAAAT GC-3′, and the 3’ primer 5′- GCATTTTCGG CACTGCCCGC AGCGTAATCT GGAACATCGT ATGGGTAGGC TCCCACGGAG GCGTTGATGG C-3′. Successful insertion of the HA epitope into ASCT2 was verified by sequence analysis. For the transient expression of Vpu, the codon-optimized vector pcDNA-Vphu was employed (64). The full-length infectious HIV-1 molecular clones pNL4-3 WT (65), the Vpu deleted mutant of pNL4-3 called pNL4-3 Udel (66), pAD8 (67), and the Vpu deleted mutant of pAD8 called pAD8 Udel (68) were previously reported. The subviral vector pNLA1 encoding the NL43 Env WT is a derivative of pNL43, lacking the gag and pol genes but expressing all other viral genes (69). The subviral vector pNLA1 Udel encoding the NL43 Env WT is a derivative of pNL43, lacking the gag, pol, and vpu genes but expressing all other viral genes (69). The env-defective pNLenv1 (also referred to as pNL43△Env) carries a 1,264 bp out-of-frame deletion in the env gene (70). For the transient expression of the glycoprotein vesicular stomatitis virus G (VSVg), the vector pCMV-VSVg was used (71). A panel of five transmitted/founder viruses, called T/F #4, T/F #5, T/F #6, T/F #7, and T/F #9, was obtained from Dr. John Kappes through the NIH HIV Reagent Program (cat# HRP-11919; https://www.beiresources.org).

Antibodies. Human ASCT2 (hASCT2) was identified using a rabbit polyclonal antibody directed to hASCT2 (Abcam Inc., Cambridge, MA; Cat# AB187692). Transiently expressed ASCT2-Myc-DDK (FLAG) was detected using a mouse monoclonal anti-FLAG M2-Peroxidase (HRP)-conjugated antibody (Millipore Sigma, St. Louis, MO; Cat# T9026). Polyclonal anti-Vpu rabbit serum directed against the cytoplasmic domain of Vpu expressed in Escherichia coli was used for the detection of Vpu. This in-house antibody is freely available through the NIH AIDS Research and Reference Reagent Program (Cat# 969). HIV-1 Env was detected using an in-house rabbit polyclonal antibody directed against purified recombinant gp120 expressed in CHO cells. HIV-1 Gag was identified using pooled HIV IG (NIH AIDS Research and Reference Reagent Program; Cat# 3957). Polyclonal rabbit serum directed against the extracellular portion of BST-2 expressed in Escherichia coli was used to detect BST-2. This in-house antibody is freely available through the NIH AIDS Research and Reference Reagent Program (Cat# 11721; https://www.beiresources.org). Tubulin was detected using a mouse monoclonal antibody to alpha-tubulin (Millipore Sigma, St. Louis, MO; Cat# A8592). Mouse monoclonal anti-HA antibody conjugated to APC (Miltenyi Biotec, Gaithersburg, MD; Cat# 130-123-553), anti-CD71 antibody conjugated to APC (Miltenyi Biotec, Gaithersburg, MD; Cat# 130-123-788), rabbit polyclonal anti-SLC1A5 antibody conjugated to FITC (Bioss Inc., Woburn, MA; Cat# bs-0473R-FITC), mouse monoclonal anti-human CD4 antibody conjugated to APC (BD Biosciences, Franklin Lakes, NJ; Cat# 551980), and anti-p24-Gag antibody (clone KC57 RD1) conjugated to PE (Beckman Coulter, Brea, CA; Cat# 6604667) were used to detect the HA epitope tagged to ASCT2, transferrin receptor (TfR) (CD71), human ASCT2, human CD4, and HIV-1 core antigens (55, 39, 33, and 24 kDa proteins), respectively, by fluorescence-activated cell sorter (FACS) analysis.

Transient transfection of HEK293T cells

For transient transfection of HEK293T cells, 3 × 10⁶ cells were plated in a 25 cm² flask and grown overnight to about 80% confluency. Cells were transfected using Lipofectamine PLUS (Invitrogen/Thermo Fisher Scientific Inc., Waltham, MA) according to the manufacturer’s instructions. Generally, total amounts of plasmid DNA were adjusted to 5 µg, using empty DNA vector as needed. Unless noted otherwise, transfected cells were harvested the following day (approximately 20 hours after transfection) and processed as described here.

Transient transfection of HeLa cells for IFI and confocal microscopy

For transient transfection of HeLa cells, 1 × 10⁶ cells were plated in 12 well-plates containing cover slips and grown overnight to about 50% confluency. Cells were transfected on the cover slips using Lipofectamine PLUS (Invitrogen/Thermo Fisher Scientific Inc., Waltham, MA) according to the manufacturer’s instructions. Total amounts of plasmid DNA were adjusted to 0.8 µg, using empty DNA vector as needed. After 24 hours of transfection, HeLa cells were fixed with 4% paraformaldehyde (PFA) for 20 minutes at room temperature (RT) and washed twice with PBS. Cells were permeabilized with 1% Triton X-100 in PBS for 2 minutes at RT and then washed twice with PBS. For antibody staining, coverslips were incubated with the appropriate primary antibodies in 1% BSA in PBS overnight at 4°C in a humid chamber. Coverslips were then washed twice with PBS and incubated with the appropriate secondary antibodies in 1% BSA in PBS for 1 hour at RT in a dark humid chamber. Coverslips were again washed twice with PBS and incubated with a 300 nM DAPI solution for 5 minutes at RT in a dark chamber, followed by two final washes with PBS. Coverslips were mounted onto microscope slides with ProLong Glass Antifade Mountant (Invitrogen/ThermoFisher Scientific Inc., Waltham, MA). Images were acquired with a Leica TCS SP8 X White Light Laser or an Olympus Fv 1000 confocal microscope. Images were analyzed using FIJI/Image J software.

Co-immunoprecipitation analysis

HEK293T cells were harvested 20 hours post-transfection and washed twice with cold PBS. Cells were lysed in 1 mL of co-IP lysis buffer (25 mM Tris-HCl [pH 7.4], 150 mM NaCl, 1% NP-40, and 5% glycerol) or 1 mL of Triton X-100 lysis buffer (50  mM Tris-HCl [pH 7.5], 150  mM NaCl, 1% Triton X-100, 10% glycerol) for 30 minutes at 4°C on a rotating platform, as described in the text. Insoluble debris was removed by centrifugation (10 min, 15,000 rpm at 4°C) in a refrigerated centrifuge. Supernatants were collected for subsequent immunoprecipitation. Ten to fifteen percent of each lysate was used as an input control. The remaining cell lysates were used for immunoprecipitation. Cell lysates were incubated with EZview Red ANTI-FLAG M2 Affinity Gel (Millipore Sigma, St. Louis, MO; Cat# SLCP8789) for 1 hour at 4°C or with GammaBind Plus Sepharose agarose beads (Millipore Sigma, St. Louis, MO; Cat# GE17-0886-02) conjugated with Env antibody overnight at 4°C on a rotating platform. After immunoprecipitation, beads were washed three times with co-IP lysis buffer or Triton X-100 lysis buffer.

Metabolic labeling and pulse-chase analysis

For pulse-chase analysis, transfected HEK293T cells were harvested by scraping. Cells were pelleted by centrifugation at 1,500 rpm for 5 minutes at RT and suspended in 5 mL of the labeling medium (methionine- and cysteine-free RPMI; MP Biomedicals, Santa Ana, CA). Cells were incubated for 20 minutes at 37°C to deplete the internal methionine and cysteine pools. Cells were then labeled in 200  µL of the labeling medium supplemented with 30  µL (300 µCi) of [³⁵S]-Expre³⁵S³⁵S-label (PerkinElmer, Waltham, MA; Cat# NEG072) for 10 minutes at 37°C. After the labeling period, cells were pelleted, and the unincorporated isotope was removed. Cell pellets were suspended in 1 mL of 10% FCS DMEM (chase medium), and equal aliquots were distributed into 1.5 mL tubes containing 1  mL of the prewarmed chase medium and chased in suspension for the indicated times. The cell aliquot constituting time 0 was distributed into the cold chase medium to stop metabolic activity. Cells were collected at the appropriate times and pelleted. Supernatants were aspirated, and cell pellets were stored on dry ice until all samples had been collected. For immunoprecipitation, cells were lysed in 1 mL of Triton X-100 lysis buffer for 30 minutes at 4°C on a rotating platform. Insoluble material was removed by centrifugation (15,000 rpm for 10 minutes), and supernatants were collected for subsequent immunoprecipitation. Ten to fifteen percent of each lysate was used as input control. The remaining samples were used for immunoprecipitation, as described above. After immunoprecipitation, samples were washed three times with Triton X-100 lysis buffer, eluted with 100 µL of sample buffer (4% sodium dodecyl sulfate, 125 mM Tris-HCl [pH 6.8], 10% 2-mercaptoethanol, 10% glycerol, and 0.002% bromophenol blue) for 10 minutes at 65°C, and subjected to SDS-PAGE. Gels were soaked in 1 M Na-salicylate for 20 min, dried, and exposed to Kodak XMR film to visualize proteins via fluorography. For quantification of signals, the dried gels were also exposed to imaging plates and visualized using a Typhoon FLA 9500 Phosphoimager. FujiFilm Multi Gauge software was used for the data quantification.

Endoglycosidase H (Endo H) and peptide-N-glycosidase F (PNGase F) analysis

For enzyme treatments, transfected HEK293T cells were processed for immunoprecipitation, as described above. Ten to fifteen percent of each lysate was used as input control. The remaining cell lysates were used for immunoprecipitation, as described above. After immunoprecipitation, beads were washed three times with Triton X-100 lysis buffer, then washed once at room temperature with denaturing buffer (0.1% SDS, 50 mM βME), and eluted and denatured with 40 µL of denaturing buffer for 10 minutes at 65°C. Samples were incubated with 5× assay buffer (0.25 M sodium acetate, pH 5.5) and 0.2 U of Endo H enzyme or 5× assay buffer (0.25 M Tris-HCl, pH 8.8) and 7 U of PNGase enzyme for 3 hours at 37°C. Samples were eluted with 100 µL of sample buffer for 10 minutes at 65°C and subjected to SDS-PAGE and immunoblotting.

Immunoblotting

For immunoblot analysis of cell-and virus-containing supernatant proteins, transfected HEK293T cells were harvested by scraping, pelleted by centrifugation at 1,500 rpm for 5 minutes at RT, and washed twice with PBS. Cells were lysed in 300 µL Triton X-100 lysis buffer for 30 minutes at 4°C on a rotating platform. Virus-containing supernatants were filtered with a 0.45 µm syringe filter and pelleted (1.3 mL) in a refrigerated Eppendorf minicentrifuge at 14,000  rpm for 90  minutes at 4°C. Cell samples were eluted with 300 µL of sample buffer for 10 minutes at 65°C, and pelleted viral particles were solubilized with 60 µL of sample buffer for 10 minutes at 65°C. All samples were subjected to SDS-PAGE and transferred to polyvinylidene difluoride (PVDF) membranes. Membranes were incubated for 30 minutes with blocking buffer (5% non-fat milk solution in 1× TNT buffer [10 mM Tris {pH 7.4}, 150 mM NaCl, 0.3% Tween-20]) and subsequently were washed once with TNT-N buffer (TNT buffer, 0.05% IGEPAL CA-630), followed by one wash with TNT buffer for 5 minutes at RT. Afterward, membranes were incubated with the appropriate primary antibodies in TNT buffer for 1 hour at RT on a rocker platform. Unbound antibodies were removed by washing the membranes once each with TNT-N and TNT buffers for 5 minutes at RT, followed by incubation with horseradish peroxidase-conjugated secondary antibodies (GE Healthcare, Piscataway, NJ) in TNT buffer for 1 hour at RT. Unbound antibodies were removed by washing the membranes twice with TNT-N buffer for 5 minutes each at RT, followed byan additional two washes with TNT buffer for 5 minutes each at RT. Membranes were incubated with Clarity Western ECL substrate (Bio-Rad Laboratories, Hercules, CA; Cat# 170-5061) for 5 minutes at RT. Images were acquired using a ChemiDoc imaging system (Bio-Rad Laboratories, Hercules, CA), and the data were analyzed using the Image Lab 6.0 software.

Quantitation of extracellular virus by reverse transcriptase (RT) assay

Virus-containing supernatants were filtered with a 0.45 µm syringe filter. Extracellular virus was quantified by measuring the amounts of virus-associated reverse transcriptase (RT) using a ³²P-based assay, as previously described (72). Briefly, a reaction mixture was prepared by combining 10  µL of the virus-containing supernatant with 50  µL of an RT reaction cocktail containing poly(A) (5  µg/mL), oligo(dT) [oligo(dT)_12-18; 0.16  µg/mL] in 50  mM Tris (pH 7.8), 7.5  mM KCl, 2  mM dithiothreitol, 5  mM MgCl₂, 0.05% NP-40, and 1 µCi per mL of cocktail of [³²P]dTTP (800  Ci/mmol) and incubated for 90 minutes at 37°C. After incubation, 10  µL of the reaction mixture was spotted onto DEAE ion exchange paper (Whatman) and washed three times with 2× SSC buffer (0.3 M NaCl and 0.03 M sodium citrate) to remove unincorporated [³²P]dTTP. Spots were counted with a scintillation counter or subjected to phosphorimage analysis.

Viral infectivity assay using TZM-b1 cells

Virus-containing supernatants were filtered with a 0.45 µm syringe filter. Total amounts of virus in the supernatants were quantified by RT assay, as described above. TZM-bl indicator cells (CD4⁺, CCR5⁺, and CXCR4⁺) were plated in a 24-well plate (5 × 10⁴ cells/well) in 1 mL of DMEM and infected with 50–150 µL of viral supernatants. Typically, infections were performed in triplicate. After 48 hours of infection, TZM-bl cells were lysed with 250 µL of Reporter Lysis Buffer (Promega Corp., Madison, WI; Cat# E4030). Luciferase activity was determined by combining 15 µL of each lysate with 50 µL of luciferase substrate (Steady-Glo; Promega Corp., Madison, WI). The light emission was measured using a GloMax microplate reader (Promega Corp., Madison, WI). Results were corrected for differences in input virus based on the RT assay.

Viral infection of MDM

Stocks of macrophage tropic viruses were prepared as follows. HEK293T cells were transfected using Lipofectamine PLUS according to the manufacturer’s instructions. For VSVg pseudotyping, HEK293T cells were cotransfected with 0.5 µg of pCMV-VSVg and 5 µg of pAD8 WT, pAD8 Udel, pNLMO WT, pNLMO Udel, pNLMO S2/6A, or pNLMO RD plasmid DNA. After 48 hours of transfection, HEK293T cells and virus-containing supernatants were harvested by scraping and pelleted by centrifugation at 1,500 rpm for 5 minutes at RT. Virus-containing supernatants were filtered with 0.45 µm syringe filters. Amounts of virus in the supernatants were quantified by RT assay, as described above. Human monocytes were maintained in six well-plates (2 × 10⁶ cells/well) in DMEM supplemented with 4.5 g/L D-glucose, 10% pooled human serum, 100 U/mL penicillin, 100 µg/mL streptomycin, 2 mM L-glutamine, and 1 mM sodium pyruvate (differentiation medium). Cells were cultured for 7 days to allow differentiation into macrophages (MDM). MDM were then exposed to 200 µL of viral supernatants in 500 µL of differentiation medium for 3 hours at 37°C in a humidified 5% CO₂ atmosphere to allow viral adsorption. After 3 hours, 1.5 mL of differentiation medium was added to each well, and MDM were incubated at 37°C. On day 1 and every 2–3 days thereafter, 1 mL of the culture supernatant was collected and replaced by 1 mL of the fresh medium. MDM were harvested by scraping after 7 days post-infection and washed twice with cold PBS. MDM were lysed in 1 mL of Triton X-100 lysis buffer for 30 minutes at 4°C on a rotating platform. Insoluble material was removed by centrifugation at 15,000 rpm for 10 minutes at 4°C. Soluble fractions were collected for subsequent immunoblot analysis.

Viral infection of CEM-SS cells

Stocks of CD4⁺ T cell tropic viruses were prepared as described for the macrophage tropic virus using 0.5 µg of pCMV-VSVg and 5 µg of pNL43 WT or pNL43 Udel plasmids DNA. Total amounts of virus in the supernatants were quantified by RT assay, as described above. CEM-SS cells were plated in 24-well plates (1 × 10⁶ cells/well), and 150 µL of virus-containing supernatants was added and incubated at 4°C for 30 minutes to allow viral adsorption. CEM-SS cells were then shifted to 37^oC and incubated with the virus-containing supernatants for 48 hours in a humidified 5% CO₂ atmosphere. CEM-SS cells were harvested after 48 hours post-infection and washed twice with cold PBS. CEM-SS cells were lysed in 1 mL of Triton X-100 lysis buffer for 30 minutes at 4°C and on a rotating platform. Insoluble material was removed by centrifugation at 15,000 rpm for 10 minutes at 4°C. Supernatants were collected for subsequent SDS-PAGE and immunoblotting.

Viral infection of PBLs

Human peripheral blood lymphocyte (PBL) cells were obtained from the NIH blood bank under protocol 99 CC-0168: “Collection and Distribution of Blood Components from Healthy Donors for In Vitro Research Use.” Cells were stimulated with 10% FCS RPMI in the presence of Concanavalin A (20 µg/mL) for 20 hours. After stimulation, the medium containing concanavalin A was replaced with 10% FCS RPMI in the presence of human IL-2 (20 U/mL). Stimulated PBL cells from two healthy donors (10 million cells each) were spin-inoculated with supernatants containing no virus (mock) or containing NL43 WT or NL43 Udel viruses at 1,200 g, for 1 hour at RT. Infections were allowed to proceed for 13 days. After infection, PBL cells were collected and processed for FACS analysis.

OptiPrep density gradient

Supernatants from transfected HEK293T cells (11 mL) were pelleted in an ultracentrifuge (Beckman, SW41 rotor) at 35,000 rpm, for 75 minutes at 4°C. The pellets were suspended in 500 µL of DMEM and subjected to a 6–13% OptiPrep Density Gradient, as previously described (73). Nine OptiPrep dilutions (30% to 6% in 3% increments) were prepared by diluting the 60% OptiPrep Density Gradient Medium (Millipore Sigma, St. Louis, MO; Cat # D1556) in PBS. Starting with the 30% solution, 445 µL of each dilution was sequentially added to a SW55 tube (total volume of 4,005 µL). The gradient was overlaid with 500 µL of the concentrated virus and subjected to ultracentrifugation (Beckman, SW55Ti rotor) at 45,000 rpm for 90 minutes at 4°C. Twelve fractions (380 µL each) were collected from the top of the gradient and mixed with 150 µL of 4× sample buffer (8% SDS, 250 mM Tris [pH 6.8, 20% 2-mercaptoethanol, 20% glycerol, and 0.002% bromophenol blue). The samples were heated at 65°C for 10 minutes and analyzed by immunoblotting.

RNA isolation and RT-qPCR

Total RNA was extracted from MDM and CEM-SS cells using the Direct-zol RNA miniprep kit (Zymo Research, Irvine, CA) according to the manufacturer’s instruction. RNA concentrations were determined using a NanoDrop reader (ThermoFisher Scientific, Waltham, MA). RT-qPCR was performed with 50  ng of total RNA and specific primer sets for ASCT2 and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) using iTaq Universal SYBR Green one-step kit (Bio-Rad Laboratories, Hercules, CA). Primers for RT-qPCR were as follows: ASCT2 sense, 5′-TCCTCTTCAC CCGCAAAAAC CC-3′; ASCT2 antisense, 5′-CCACGCCATT ATTCTCCTCC AC-3′; GAPDH sense, 5′-AAGGTCGGAG TCAACGGATT-3′, and GAPDH antisense, 5′-CTCCTGGAAG ATGGTGATGG-3′. RT-qPCR was carried out using the CFX96 Touch real-time PCR detection system (Bio-Rad Laboratories, Hercules, CA). The housekeeping gene GAPDH was used as an internal control. Relative mRNA levels were determined using the threshold cycle (ΔΔCT) quantification method.

Flow cytometry

For cell surface staining of ASCT2, HEK293T cells were transfected with 0.5 µg of the vector expressing HA-tagged ASCT2 and increasing amounts (0, 0.5, 1, 2.5, and 5 µg) of pCDNA-Vphu WT. Total amounts of transfected plasmid DNA were adjusted to 5.5 µg for each sample using empty vector DNA as needed. Mock-transfected cells were included as a negative control. After 24 hours of transfection, HEK293T cells were processed for flow cytometry. Cells were incubated with anti-HA APC and, as a control, anti-TfR (CD71) APC antibodies for 40 minutes at 4°C and fixed with 1% formaldehyde in PBS solution. Samples were analyzed with a BD Flow Cytometer (BD Biosciences, Franklin Lakes, NJ).

For the detection of endogenous ASCT2 in HIV-1-infected CD4 +T-lymphocytes, PBL cells from two healthy donors were infected with NL43 WT, NL43 Udel, or were left uninfected. After 13 days of infection, PBL cells were processed for flow cytometry. Cells were permeabilized and incubated with anti-ASCT2 FITC, anti-CD4 APC, and p-24-Gag PE antibodies for 40 minutes at 4°C and fixed with 1% formaldehyde in PBS solution. Samples were analyzed with a BD Flow Cytometer.

Glutamine/Glutamate Glow assay

CEM-SS cells were plated in 24-well plates (1 × 10 ⁶ cells/well) and incubated with 150 µL of VSVg pseudotyped NL43 WT or NL43 Udel virus at 4°C for 30 minutes to allow viral adsorption. Cells were then shifted to 37^oC and incubated for 48 hours in a humidified 5% CO₂ atmosphere. Supernatants were then used to determine glutamate and glutamine concentrations by using the commercial bioluminescent kit Glutamine/Glutamate Glow Assay (Promega Corp., Madison, WI; Cat# J8201), according to the manufacturer’s instructions. Resulting light emission was measured using a GloMax microplate reader (Promega Corp.)

IFN-α treatment and stimulation of PBL

Human monocytes were maintained in six well-plates (2 × 10⁶ cells/well) in DMEM supplemented with 4.5 g/L D-glucose, 10% pooled human serum, 100 U/mL penicillin, 100 µg/mL streptomycin, 2 mM L-glutamine, and 1 mM sodium pyruvate (differentiation medium). Cells were cultured for 7 days to allow differentiation into MDM. MDM were incubated with 20 increasing amounts of IFN-α2A (0, 0.01, 0.1, 1, 10, and 100 ng/mL) (Millipore Sigma, St. Louis, MO; Cat # GF416) for 24 hours. After treatment, 21 whole-cell extracts were prepared and processed for immunoblotting using anti-ASCT2, anti-BST-2, and anti-tubulin antibodies.

PBL cells were kept in 10% FCS RPMI in the presence of concanavalin A (20 µg/mL) for 20 hours. After stimulation, the medium containing concanavalin A was replaced with 10% FCS RPMI in the presence of human IL-2 (20 U/mL). Stimulated PBL cells were plated in wells of a 24-well plate (2 × 10⁶ cells/well) and incubated with increasing amounts of IFN-α2A (0, 0.01, 0.1, 1, 10, and 100 ng/mL) for 24 hours. After treatment, whole-cell extracts were prepared and processed for FACS analysis using anti-ASCT2-FITC and anti-CD4-APC antibodies.

Statistical analysis

Statistical analyses were performed using GraphPad Prism 10. The data are presented as means, with error bars indicating the standard errors of the mean (SEM), from three or more independent experiments, as indicated in the text. All statistical analyses were conducted using Student’s t-test, one-way, or two-way ANOVA. Statistical significance was defined as a P value of 0.05, using the appropriate statistical test method. Details regarding statistics, number of biological replicates, and sample sizes are reported in each figure legend.