The V3 Loop of HIV-1 Env Determines Viral Susceptibility to IFITM3 Impairment of Viral Infectivity.

AD8-1 Env confers resistance to IFITM3 inhibition.

We next asked which viral protein renders the AD8-1 virus resistant to IFITM-mediated inhibition. We started with an HIV-1 proviral DNA clone called NL(AD8), which has the Env sequence of AD8-1 substituted for that of the NL4-3 virus in the context of the NL4-3 proviral DNA (34). We performed cotransfection experiments in HEK293T cells with NL(AD8) DNA and IFITM1, IFITM2, or IFITM3 plasmid DNA. The results showed that the NL(AD8) virus was as resistant to IFITM inhibition of viral infectivity as the AD8-1 virus (Fig. 2A), suggesting that the AD8-1 Env protein is sufficient to confer resistance to IFITM inhibition.

In order to determine which domain of AD8-1 Env was responsible for this resistance phenotype, we changed each of the V1, V2, and V3 loop sequences of NL4-3 Env to those of AD8-1. The V1 and V2 substitutions generated noninfectious viral particles (data not shown). This incompatibility of different variable loops between different HIV-1 Env proteins has also been reported by others (35). In contrast, V3-substituted NL(AD8V3) produced viruses that were almost as infectious as wild-type NL4-3 and showed resistance to IFITM3 inhibition (Fig. 4A to ​toC).C). We further tested the V3 loops of HIV-1 strains YU-2, WITO, THRO, and RHPA by inserting these V3 sequences into NL4-3. These V3 loop sequences conferred different levels of resistance to IFITM3, with the WITO V3 loop generating as much resistance to IFITM3 as the AD8-1 V3 loop (Fig. 4A to ​toC).C). Since all of these V3 loop-substituted viruses use CCR5 as their coreceptor and have shown different levels of sensitivity to IFITM3 inhibition, coreceptor usage does not determine virus resistance to IFITM3.

 

Open in a separate window

FIG 4

The V3 loop sequences differentially modulate IFITM3 inhibition of HIV-1. (A to C) Viruses were produced by cotransfecting individual chimeric viral DNAs together with either the empty vector pQCXIP or IFITM3 DNA (100 ng) into HEK293T cells. The viruses were quantified by measuring their viral RT activity. Viral infectivity was then determined by infecting the TZM-bl cells using viruses with the same amounts of RT activity. (A) Results of one representative infection experiment. RLU, relative light units. Effects of IFITM3 on the infectivity of each virus were determined by calculating the ratio of luciferase activities of the IFITM3-free (pQCXIP) viruses to those of the IFITM3-bearing viruses (fold inhibition). (B) Averages of data from three independent transfection experiments. (C) Sequences of the V3 loop in Env from different HIV-1 strains. (D and E) Effects of IFITM3 on HIV-1 gp160 processing. Virus particles of different HIV-1 strains were produced by transfecting HEK293T cells in the absence or presence of IFITM3. Levels of virus particle-associated Gag, IFITM3-FLAG, and Env (gp160 and gp120) were assessed by Western blotting. The intensities of gp160 and gp120 signals for each virus were quantified by using the NIH ImageJ program. The ratios of gp120/(gp120 plus gp160) were calculated for each virus, and the results from four independent transfection experiments are summarized (E). (F) Correlation analysis of the effect of IFITM3 on gp160 processing and IFITM3 inhibition of HIV-1 infectivity (data from panel B). (G to I) Stability of Env proteins from different HIV-1 strains. HEK293K cells were transfected with proviral DNA of NL4-3, AD8-1, NL(AD8), or NL(AD8V3) and then labeled with [³⁵S]methionine-cysteine. HIV-1 Env in the cell lysates as well as in the supernatants was immunoprecipitated and analyzed by SDS-PAGE followed by PhosphorImager analysis. (G) Results of one representative experiment from 3 independent experiments. (H and I) The processing index and the association index values of HIV-1 Env were calculated from three independent experiments, and the results are summarized. The values for NL4-3 were arbitrarily set to 1. ***, P < 0.001.

We previously observed that IFITM3 interferes with the processing of the viral Env precursor gp160 (25). Indeed, when Env proteins in virus particles that were produced from HEK293T cells were examined by Western blotting, the IFITM3-bearing NL4-3 virus contained less gp120 concurrently with the accumulated gp160 precursor than did the IFITM3-free NL4-3 virus (Fig. 4D and ​andE).E). In contrast, the IFITM3-resistant AD8-1 virus had similar levels of gp120 and gp160 regardless of virus incorporation of the IFITM3 protein. The NL(AD8), NL(AD8V3), and NL(WITOV3) viruses, which are also resistant to IFITM3, had nearly 50% of the Env protein processed into gp120 when IFITM3 was coexpressed. In contrast, upon virus incorporation of IFITM3, the NL(YU-2V3) and NL(THROV3) viruses had 14.9% and 23% of Env processed into gp120, respectively (Fig. 4D and ​andE).E). There is a significant correlation between IFITM3 inhibition of HIV-1 infectivity and gp160 processing (Fig. 4F). Further studies are warranted to investigate whether impaired gp160 processing directly diminishes HIV-1 infectivity. We also noted that there was no correlation between the inhibition of viral infectivity and the incorporation of IFITM3 into HIV-1 virions, similar to data in a previous report (25).

The results of Western blotting of virus samples also revealed lower levels of total Env proteins for NL4-3 and NL(YU-2V3) than for the rest of the viruses that are resistant to IFITM3. We therefore performed metabolic labeling of viral Env using [³⁵S]methionine-cysteine and assessed the efficiency of Env gp160 processing and the stability of the gp120/gp41 complex by calculating the processing index (PI) and association index (AI) values. The results showed that all viruses tested had similar PI values but that the IFITM3-sensitive NL4-3 virus had a lower AI value than did the IFITM3-resistant viruses, including AD8-1, NL(AD8), and NL(AD8V3) (Fig. 4G to ​toI).I). These data suggest that the V3 loop sequences from AD8-1 and WITO confer resistance to IFITM3 by maintaining the stability of the trimeric Env complex.

IFITM3-bearing HIV-1 has been shown to be defective in entry (23, 25). Indeed, the results of a beta lactamase (BlaM)-Vpr virion fusion assay showed that the entry of IFITM3-bearing NL4-3 was reduced by 5-fold (Fig. 5). In contrast, the entry of IFITM3-resistant strains AD8-1, NL(AD8V3), and NL(WITOV3) was slightly affected by IFITM3 (Fig. 5). The entry of the NL(RHPAV3), NL(THROV3), and NL(YU-2V3) viruses was diminished by 2- to 4-fold, which correlates with their levels of resistance to IFITM3. Together, these data demonstrate that the V3 loop sequences are able to overcome IFITM3 inhibition of viral infectivity by stabilizing Env and therefore rescuing virus entry.

 

Open in a separate window

FIG 5

Effect of IFITM3 on the entry of V3 loop chimeric viruses. (A) The BlaM-Vpr-containing viruses were produced by cotransfecting HEK293T cells with BlaM-Vpr DNA, each HIV-1 DNA, and either the empty vector or IFITM3 DNA. Viruses with the same viral RT activity were used to infect C8166-R5 cells. Viral entry was determined by measuring the cleavage of CCF2 by BlaM. The results of one representative entry assay are shown. (B) The effects of IFITM3 on virus entry were determined by calculating the ratios of the entry efficiency of each virus in the absence of IFITM3 to that of the same virus in the presence of IFITM3. The results are presented as fold inhibition values and are the averages of data from four independent assays. *, P < 0.05; ***, P < 0.001.

Plasmid DNA.

pNL4-3 (catalog number 114), p89.6 (catalog number 3552), pYU-2 (catalog number 1350), pWITO.c/2474 (catalog number 11739), pCH040.c/2625 (catalog number 11740), pCH058.c/2960 (catalog number 11856), pCH077.t/2627 (catalog number 11742), pCH106.c/2633 (catalog number 11743), pRHPA.c/2635 (catalog number11744), pTHRO.c/2626 (catalog number11745), pREJO.c/2864 (catalog number 11746), pZM246F_10 (catalog number 11828), and pZM247Fv2 (catalog number 11829) were obtained from the NIH AIDS Reagent Program. pAD8-1 and pNL(AD8) were kindly provided by Eric O. Freed (34). N-terminally FLAG-labeled IFITM1, -2, or -3 was cloned into pQCXIP as previously described (3). pQCXIP (catalog number 631516) was purchased from Clontech. NL4-3 env sequences that contain the V3 loop sequence of AD8-1, YU-2, WITO, THRO, or RHPA were synthesized by Invitrogen. The synthesized sequences were cloned into pNL4-3 via restriction enzyme digestion by SalI and NheI.

Virus production.

HEK293T cells were seeded into a 6-well plate at 0.6 × 10⁶ cells/well 20 h prior to transfection. A total of 500 ng of a proviral DNA plasmid and 0, 25, 50, 100, and 200 ng of IFITM1, -2, or -3 plasmids were cotransfected into HEK293T cells by using polyethylenimine (PEI). The medium was changed at 6 h posttransfection. The supernatant and cells were harvested at 48 h posttransfection. The collected supernatant was centrifuged at 3,000 rpm (CS-6R; Beckman Coulter) for 20 min to remove the cell debris. The virus was then separated into 1-ml aliquots and stored in a −80°C freezer. The amount of virus was quantified by measuring viral RT activity.

Virus infectivity.

A total of 5.0 × 10⁴ TZM-bl cells were seeded into each well of a 24-well plate 20 h prior to infection. Viruses with equal amounts of RT activity were used to infect TZM-bl cells. At 40 h postinfection, TZM-bl cells were washed once with cold Dulbecco's phosphate-buffered saline (PBS) (DPBS) (catalog number 14190-144; Gibco by Life Technology) and lysed in 100 μl of 1× passive lysis buffer (catalog number E1941; Promega). Ten microliters of the lysate from each sample was mixed with 40 μl of the luciferase substrate (catalog number E4530; Promega). Luciferase activity was measured by using a Glomax 20/20 luminometer.

Ultracentrifugation to harvest virus particles.

In order to produce enough HIV-1 particles for ultracentrifugation and Western blotting, transfection was performed with HEK293T cells that were seeded into a 10-cm cell culture plate at 4 × 10⁶ cells/plate. Twenty hours after seeding, 5 μg of HIV-1 proviral DNA, 1 μg of IFITM3 DNA, or 1 μg of pQCXIP DNA was cotransfected into each dish by using PEI. The ratio of HIV-1 DNA to IFITM3 DNA (5 μg–1 μg) was kept the same as the one that was used in the transfection experiments to measure IFITM3 inhibition of HIV-1 (500 ng–100 ng). The medium was changed at 6 h posttransfection, and the viruses were harvested 48 h later. The harvested viruses were filtered through a 0.2-μm sterile polyethersulfone membrane syringe filter (catalog number 28145-501; VWR International) to remove the cell debris. Viruses were slowly transferred to 14- by 89-mm ultracentrifugation tubes (catalog number 344059; Beckman Coulter) that were preloaded with 2 ml of a 20% sucrose cushion. Virus particles were pelleted at 35,000 rpm for 1 h at 4°C by ultracentrifugation (Optima L-100XP ultracentrifuge; Beckman Coulter). The pelleted viruses were resuspended in 100 μl of DMEM. Five microliters of the pelleted viruses was used for measuring viral RT activity.

BlaM-Vpr viral entry assay.

A BlaM-Vpr assay was performed to measure HIV-1 entry (65, 66). The BlaM-Vpr-containing viruses were produced by the cotransfection of 5 μg of HIV-1 proviral DNA, 1 μg of pCMV-BlaM-Vpr DNA, and 1 μg of IFITM3 or the pQCXIP vector into HEK293T cells. The collected supernatant was filtered through a 0.2-μm filter and pelleted at 35,000 rpm for 1 h at 4°C. The viruses were resuspended in RPMI 1640 medium. Five microliters of each virus was used for RT activity quantification. The rest of the viruses were aliquoted and stored in a −80°C freezer. C8166-R5 cells were used for the viral entry assay. A total of 1 × 10⁶ C8166-R5 cells were plated with 400 μl of 10% FBS–RPMI 1640 medium into each well of a 24-well plate. The volumes of the viruses were adjusted according to their RT measurements. Fresh medium was added to achieve a final volume of 100 μl for each virus. A total of 5 μg/ml of Polybrene (catalog number 107689; Sigma-Aldrich) was included in the medium, and a 45-min spinoculation at 1,800 rpm (CS-6R; Beckman Coulter) was carried out, followed by 2 h of incubation in a humidified environment at 37°C with 5% CO₂. The cells were then transferred to Eppendorf tubes (EP tubes), and the viruses that did not succeed in entry were washed off by using CO₂-independent medium (catalog number 18045-088; Gibco by Life Technology). The cells were loaded with CCF2-AM (catalog number K1032; Invitrogen) and incubated in the dark at room temperature for 1 h. The cells were washed with developing solution, transferred into a V-bottom 96-well cell culture plate (catalog number 651180; Greiner Bio-One International), and bathed in developing solution overnight (16 h) at room temperature in the dark. The next day, the cells were washed twice with cold DPBS containing 2% FBS, fixed with 2% paraformaldehyde (catalog number PAR070; BioShop) dissolved in 2% FBS–DPBS, and analyzed by flow cytometry (BD LSR Fortessa analyzer; BD Biosciences). The results obtained by flow cytometry were analyzed by using FlowJo software.

Neutralization assay.

Monoclonal neutralizing antibodies VRC03 (catalog number 12032), 17b (catalog number 4091), 7H6 (catalog number 12295), 10E8 (catalog number 1294), 447-52D (catalog number 4020), 10-1074 (catalog number 12477), and PG16 (catalog number 12150), together with sCD4 (catalog number 4615), were obtained from the NIH AIDS Reagent Program. The volumes of viruses were adjusted according to their RT activity, and 10% FBS–RPMI 1640 was used to adjust the volume to 100 μl. The viruses were then coincubated with each of the antibodies or sCD4 at various concentrations for 1 h in a cell culture incubator at 37°C with 5% CO₂. After incubation, the viruses were used to infect TZM-bl cells that were preseeded in a 24-well plate 20 h prior to infection. Four hundred microliters of additional fresh medium was added to each well to achieve a final volume of 1 ml, and the cells were incubated in a cell culture incubator. The cells were harvested at 40 h postinfection, and the luciferase activity was measured.

Western blotting.

Harvested cells were lysed in Cytobuster protein extraction reagent (catalog number 71009; EMD Millipore Novagen) containing a protease inhibitor (catalog number 11836153001; Roche) on ice for 1 h. The cell debris were removed by centrifugation at 13,200 rpm (Microfuge 16; Beckman Coulter) for 20 min at 4°C. The lysates were denatured by the addition of 4× protein loading buffer followed by 5 min of boiling. The protein samples were loaded onto a 1% sodium dodecyl sulfate (SDS) (catalog number SDS001; BioShop)–12% polyacrylamide (catalog number ACR009; BioShop) gel for the detection of tubulin, Gag, and FLAG-IFITM3 and onto a 1% SDS–8% polyacrylamide gel for the detection of Env. The proteins were separated by electrophoresis at 100 V for ∼2 to 3 h and then transferred onto polyvinylidene difluoride (PVDF) membranes (catalog number 3010040; Roche). The membranes were blocked in 5% skim milk dissolved in 1× PBS with 0.1% Tween 20 (catalog number TWN510; BioShop) (PBST) for 1 h, followed by incubation with a 1:5,000 dilution of primary antibodies for 2 h at room temperature. After washing with PBST, the membranes were further incubated with a 1:10,000 dilution of secondary antibodies for 1 h at room temperature. The primary antibodies included monoclonal mouse antitubulin antibody (B-5-1-2) (catalog number sc-23948; Santa Cruz Biotechnology), monoclonal mouse anti-FLAG antibody (catalog number F1804-1MG; Sigma-Aldrich), polyclonal rabbit anti-p24 antibody (catalog number SAB3500946; Sigma-Aldrich), monoclonal anti-gp41 antibody (catalog number 526; NIH AIDS Reagent Program), and sheep anti-gp120 single-chain antibody (catalog number 11710; NIH AIDS Reagent Program). The secondary antibodies included enhanced chemiluminescence (ECL) rabbit IgG horseradish peroxidase (HRP)-linked whole antibody from donkey (catalog number NA934V; GE Health Care Life Science), Amersham ECL mouse HRP-linked IgG whole Ab from sheep (catalog number NA931; GE Health Care Life Science), and HRP-rabbit anti-sheep IgG (catalog number 618620; Invitrogen). The membranes were treated with the Western Lightening Plus-ECL substrate (catalog number NEL105001EA; PerkinElmer), and the chemiluminescent signals were detected by exposure to X-ray films (catalog number 6041768; Carestream).

Assay of HIV-1 inhibition by AMD3100 and maraviroc.

TZM-bl cells were incubated with 0, 5.14 × 10⁻⁵, 2.67 × 10⁻⁴, 1.28 × 10⁻³, 5.14 × 10⁻³, and 0.01 μg/ml of the CXCR4 inhibitor AMD3100 or the CCR5 inhibitor maraviroc for 1 h in a humidified incubator at 37°C with 5% CO₂. The cells were then infected by the same amount of viruses. Infected cells were harvested at 40 h postinfection, and the luciferase activity was measured. Both AMD3100 (catalog number 8128) and maraviroc (catalog number 11580) were obtained from the NIH AIDS Reagent Program.

HIV-1 Env stability assay.

HEK293T cells were transfected with proviral DNA (67). Twenty-four hours after transfection, the cells were metabolically labeled with 100 μCi/ml of [³⁵S]methionine-cysteine (catalog number NEG772007MC; PerkinElmer) dissolved in 5% dialyzed FBS-supplemented methionine- and cysteine-negative DMEM for 16 h. The cells were then lysed in radioimmunoprecipitation assay (RIPA) buffer containing 140 mM NaCl, 8 mM Na₂HPO₄, 2 mM NaH₂PO₄, 1% NP-40, and 0.05% SDS. The radiolabeled Env proteins in the cell lysates or supernatants were precipitated by using serum from HIV-1 patients for 1 h at 37°C. The immunoprecipitated samples were separated by electrophoresis in polyacrylamide gels and analyzed by using a PhosphorImager (Molecular Dynamics). The association index measures the relative ability of the gp120 domain to stay on the Env trimer on virus-producing cells. The association index values thus describe the intrinsic stability of Env (i.e., the ability of gp120 to remain associated with gp41). A low association index is a good indicator of decreased levels of gp120 in virus particles (68, 69). The association index of NL4-3 was arbitrarily set to a value of 1, and the values for other viruses were calculated relative to the value for NL4-3 {association index = [(target gp120)_{cell lysate} × (NL4-3 gp120)_supernatant]/[(target gp120)_supernatant × (NL4-3 gp120)_{cell lysate}]}. The processing index measures the relative efficiency of Env maturation from the gp160 precursor to gp120. The processing index of NL4-3 was arbitrarily set to a value of 1, and the values for other viruses were calculated relative to the value for NL4-3 {processing index = [(total gp120)_target × (gp160)_NL4-3]/[(gp160)_target × (total gp120)_NL4-3]}.

We thank Eric O. Freed for providing the pAD8-1 and pNL(AD8) proviral DNA clones.

This work was supported by funding from the Canadian Institutes of Health Research to C.L. and from the National Institutes of Health to S.-L.L. (R01 AI112381). A.F. is the recipient of a Canada research chair on retroviral entry. Y.W. is a recipient of a Canada graduate scholarship (master's) from the Natural Sciences and Engineering Research Council of Canada. S.D. is a recipient of a CRCHUM postdoctoral fellowship.