Formation of syncytia is repressed by tetraspanins in human immunodeficiency virus type 1-producing cells

Abstract

In vitro propagation studies have established that human immunodeficiency virus type 1 (HIV-1) is most efficiently transmitted at the virological synapse that forms between producer and target cells. Despite the presence of the viral envelope glycoprotein (Env) and CD4 and chemokine receptors at the respective surfaces, producer and target cells usually do not fuse with each other but disengage after the viral particles have been delivered, consistent with the idea that syncytia, at least in vitro, are not required for HIV-1 spread. Here, we tested whether tetraspanins, which are well known regulators of cellular membrane fusion processes that are enriched at HIV-1 exit sites, regulate syncytium formation. We found that overexpression of tetraspanins in producer cells leads to reduced syncytium formation, while downregulation has the opposite effect. Further, we document that repression of Env-induced cell-cell fusion by tetraspanins depends on the presence of viral Gag, and we demonstrate that fusion repression requires the recruitment of Env by Gag to tetraspanin-enriched microdomains (TEMs). However, sensitivity to fusion repression by tetraspanins varied for different viral strains, despite comparable recruitment of their Envs to TEMs. Overall, these data establish tetraspanins as negative regulators of HIV-1-induced cell-cell fusion, and they start delineating the requirements for this regulation.

FIG. 1.

Tetraspanins repress HIV-1 Env-induced cell-cell fusion. (A) Fusion assays. Left panel, fluorescence microscopy-based cell-cell fusion assay. Target Jurkat cells labeled with Cell Tracker Blue (CMAC) were mixed with HeLa cells expressing NL4-3-iGFP. Syncytia were visualized 4 h after addition of target cells. Right panel, reporter gene-based cell-cell fusion assay. TZM-bl cells were used as target cells, β-Galactosidase activity was measured to indicate fusion. (B) For tetraspanin overexpression, HeLa cells were cotransfected with pNL4-3 and tetraspanin expressor plasmids. For tetraspanin knockdown, HeLa cells transduced with shRNA against the tetraspanin CD9, CD63, or CD81, or all three (combined), were transfected with pNL4-3. Reporter gene-based fusion assays were performed as in for panel A. The upper panel shows quantification of tetraspanin overexpression and knockdowns by flow cytometry. The lower panel shows relative fusion activity as measured upon cocultivation of transfected HeLa cells and Cell Tracker-labeled Jurkat cells. (C) In overexpression experiments, Jurkat T cells were coinfected with NL4-3 virus and lentiviral vectors encoding CD9, CD63, or empty vector. In knockdown experiments, Jurkat T cells were coinfected with NL4-3 virus and lentiviral vectors encoding shRNA against CD9, CD63, and/or CD81. Four days later, cells were mixed with target Jurkat cells labeled with Cell Tracker Blue (CMAC) for 16 h, and syncytia were visualized and quantified. (D) HeLa cells were transfected with pNL4-3 and cotransfected with different amounts of CD9 and CD63 expressor plasmids (the x axis indicates the tetraspanin expressors; the value is the amount relative to the standard conditions used for panel B) to achieve increased tetraspanin levels, and the reporter gene-based cell-cell fusion assay was performed. Data shown represent the means and standard errors of the means, based on at least three independent experimental repeats. *, value differs significantly from mock (P < 0.05).

FIG. 2.

Repression of Env-mediated cell-cell fusion by tetraspanins depends on Gag coexpression. HeLa cells expressing Env only or coexpressing Env and Gag were cocultured with target Jurkat cells, and syncytium assays were carried out as described for Fig. 1A. Tetraspanin levels were either augmented by overexpression (A) or downregulated by siRNA (B). The cell surface levels of Env with or without Gag cotransfection were measured by flow cytometry. Transfection conditions were adapted to achieve equal surface Env levels (mean fluorescence intensity [MFI]) (C). The data shown represent the means and standard errors of the means, based on at least three independent experimental repeats. *, value differs significantly from mock (P < 0.05).

FIG. 3.

Clustering of Env by Gag, and thus Env-CD63 colocalization, is required for cell-cell fusion repression by tetraspanins. (A) Microscopy-based fusion assays were performed with cells expressing either wild-type (WT), MA-deleted (ΔMA), or Env CT-deleted (ΔCT) NL4-3 and coexpressing CD9, CD63, or empty vector. (B) HeLa cells expressing either WT, ΔMA, or ΔCT NL4-3 were surface labeled for Env and CD63 at 4°C and then fixed and imaged by fluorescence microscopy. Bottom sections are shown. Average colocalization percentages are displayed in the merge panels. Quantification of colocalization was done with Volocity software (Improvision) as described in Materials and Methods; at least 20 fields were selected for quantification in each condition. Data shown for fusion activity represent the means and standard errors of the means, based on at least three independent experimental repeats. *, value differs significantly from mock (P < 0.05).

FIG. 4.

Strain-specific sensitivity to fusion repression by CD63. (A) Fusion repression by CD63 overexpression was assayed using cells expressing pNL4-3, pJRCSF, or pJRFL. Reporter gene-based cell-cell fusion assays were performed as described in Materials and Methods. (B) Intrinsic fusion activities of JRCSF and JRFL relative to that of NL4-3. (C) HeLa cells expressing the indicated viruses were surface labeled for Env and CD63 at 4°C and then fixed and imaged by fluorescence microscopy. Bottom sections are shown. Average colocalization percentages are displayed in the merge panels. Quantification was done as described in Materials and Methods.

FIG. 5.

Fusion repression by tetraspanins in both virus producer and target cells. HeLa cells transfected with pNL4-3 and tetraspanin expressor plasmids were used as producer cells, and TZM-bl cells transfected with tetraspanin expressor plasmids served as target cells. Fusion assays were carried out using the reporter gene-based assay. Data shown represent the means and standard errors of the means, based on at least three independent experimental repeats. *, value differs significantly from mock (P < 0.05).