LMP2A does not require palmitoylation to localize to buoyant complexes or for function

Abstract

Epstein-Barr virus (EBV) latent membrane protein 2A (LMP2A) is expressed constitutively in lipid rafts in latently infected B lymphocytes. Lipid rafts are membrane microdomains enriched in cholesterol and sphingolipids selective for specific protein association. Lipid rafts have been shown to be necessary for B-cell receptor (BCR) signal transduction. LMP2A prevents BCR recruitment to lipid rafts, thereby abrogating BCR function. As LMP2A is palmitoylated, whether this fatty acid modification is necessary for LMP2A to localize to lipid rafts and for protein function was investigated. LMP2A palmitoylation was confirmed in latently infected B cells. LMP2A was found to be palmitoylated on multiple cysteines only by S acylation. An LMP2A mutant that was not palmitoylated was identified and functioned similar to wild-type LMP2A; unmodified LMP2A localized to lipid rafts, was tyrosine phosphorylated, was associated with LMP2A-associated proteins, was ubiquitinated, and was able to block calcium mobilization following BCR cross-linking. Therefore, palmitoylation of LMP2A is not required for LMP2A targeting to buoyant complexes or for function.

FIG. 1.

LMP2A is fatty acid modified by palmitic acid in LCLs. LCLs ES1 (LMP2A negative [−]) and LCL1 (LMP2A positive [+]) were incubated with [³H]palmitic acid (³H-Palm) or with [³⁵S]methionine (³⁵S-Met) for 4 h at 37°C while rocking and then lysed. Lysates were immunoprecipitated for LMP2A with a monoclonal rat antibody and then prepared for SDS-PAGE. Gels were treated for fluorography before exposing them to film. Fluorography of half of the [³H]palmitic acid-labeled immunoprecipitated proteins (A). [³⁵S]methionine-immunoprecipitated proteins (B). Half of the [³H]palmitic acid-labeled immunoprecipitated proteins were analyzed on a parallel gel, transferred to Immobilon membrane, and immunoblotted for LMP2A with a biotin-conjugated monoclonal rat antibody (αLMP2A) (C). Representative data of three independent experiments are shown.

FIG. 2.

LMP2A deletion mutants are palmitoylated. EBV-negative BJAB B cells were transiently transfected with full-length and LMP2A deletion constructs (shown in panel A). After 4 h, the cells were incubated with [³H]palmitic acid (³H-Palm) for an additional 4 h and then lysed. Lysates were immunoprecipitated for metabolically labeled LMP2A followed by SDS-PAGE and autoradiography (B, top). Western blot analysis was performed in parallel with antibodies to LMP2A (αLMP2A) (B, bottom). Each deletion mutant maintained palmitoylation like full-length LMP2A, whereas no bands were detected in the pSG5 vector control lane. The data shown represent at least three independent experiments.

FIG. 3.

LMP2A cysteine-to-alanine point mutants retain palmitoylation. The putative structure of LMP2A showing where the cysteine residues of LMP2A are located is shown in panel A. The position in the sequence of LMP2A of the cysteines mutated to alanine is indicated in the diagram, and all of the point mutants constructed are listed. EBV-negative BJAB B cells were transiently transfected with full-length and mutant LMP2A constructs (B and C). After 4 h, the cells were incubated with [³H]palmitic acid (³H-Palm) for an additional 4 h and then lysed. Lysates were immunoprecipitated for metabolically labeled LMP2A followed by SDS-PAGE and autoradiography. Western blot analysis was performed in parallel using antibodies to LMP2A (αLMP2A). Relative palmitoylation (Rel Palm) was calculated following densitometry by dividing the ratio of tritiated protein to protein detected by Western blotting by the ratio determined for wild-type LMP2A. The ratio for wild-type LMP2A was thus set to 1.0. Representative data from four independent experiments are shown.

FIG. 4.

LMP2A is modified by S acylation. [³H]palmitic acid (³H-Palm)-labeled immunoprecipitates of transiently transfected BJAB cells with the indicated constructs were divided into three parts for SDS-PAGE. After fixing the gels, the first gel was treated with 2 M Tris-HCl, pH 7.0, overnight (A), whereas the second gel was treated with 2 M hydroxylamine in 2 M Tris-HCl, pH 7.0, overnight prior to fluorography (B). The third gel was transferred to Immobilon membrane for Western blotting with antibodies to LMP2A (αLMP2A) (C). Shown are representative data from two independent experiments. The positions of the 66 and 45 molecular size markers in kilodaltons are shown to the left of each gel.

FIG. 5.

Palmitoylation is not required for LMP2A buoyant complex localization. EBV-negative B cells were transiently cotransfected with empty vector, LMP2A, or mutant LMP2A constructs, and with L₁₀-GFP, lysed with 1% Triton X-100 lysis buffer after 15 h, and subjected to density gradient ultracentrifugation. Samples from each fraction were analyzed by immunoblotting for wild-type or mutant LMP2A. L₁₀-GFP cofractionates with Lyn in lower-density fractions representing DRMs (A, fractions 3 to 6), but not with CD45, which remains in the high-density soluble protein fractions (A, fractions 10 to 12). Wild-type LMP2A cofractionates with L₁₀-GFP in DRM fractions (B). Representative data of the LMP2A cysteine-to-alanine point mutants (C and D). Each mutant assayed was found in DRM fractions. The blots shown are representative of two independent experiments.

FIG. 6.

Palmitoylation is not required for LMP2A protein association, tyrosine phosphorylation, and ubiquitination. Wild-type LMP2A or mutant LMP2A was transiently transfected into BJAB cells. Fifteen hours after transfection, the cells were lysed. Lysates were immunoprecipitated (IP) for LMP2A, and immunoprecipitates were analyzed by SDS-PAGE, transferred to Immobilon membrane, and probed for either LMP2A (A and B; LMP2A), phosphorylated tyrosine residues (B; APT), or AIP-4 coassociation (B; AIP-4). Lyn immunoprecipitates were analyzed by Western blotting for Lyn and for LMP2A coassociation (A; Lyn and LMP2A). BJAB cells were also cotransfected with HA-Ub (C). Lysates were immunoprecipitated for LMP2A, and membranes were probed with an antibody for the HA tag to detect higher-molecular-weight LMP2A species representing ubiquitinated LMP2A (C, arrows). Empty vector was included as a control. Representative immunoblots of two independent experiments are shown.

FIG. 7.

Palmitoylation is not required for LMP2A to block Ca mobilization. BJAB cells were cotransfected with RFP and the indicated LMP2A expression construct or vector control. Ten hours after transfection, the cells were loaded with the calcium-sensitive dye indo-1 1.5 h prior to analysis by flow cytometry. The cells were stimulated with goat anti-human Ig about 40 s after observing baseline calcium levels (indicated by a break in the histogram). Events were collected over 5 min. The x axis represents time, and the y axis represents the ratio of fluorescence at 395 nm to that at 525 nm (ratio of calcium-bound indo-1 to unbound indo-1). Transfected cells were determined by gating on RFP-positive cells. RFP-negative cells represent untransfected cells and serve as an internal control. LMP2B lacks the cytoplasmic amino terminus of LMP2A, and Y74/85F is defective for Syk binding; therefore, both proteins are unable to block BCR signaling. Calcium fluxes of 5-min time courses show representative data of at least three independent experiments.