CD43 interaction with ezrin-radixin-moesin (ERM) proteins regulates T-cell trafficking and CD43 phosphorylation

Abstract

Cell polarization is a key feature of cell motility, driving cell migration to tissues. CD43 is an abundantly expressed molecule on the T-cell surface that shows distinct localization to the migrating T-cell uropod and the distal pole complex (DPC) opposite the immunological synapse via association with the ezrin-radixin-moesin (ERM) family of actin regulatory proteins. CD43 regulates multiple T-cell functions, including T-cell activation, proliferation, apoptosis, and migration. We recently demonstrated that CD43 regulates T-cell trafficking through a phosphorylation site at Ser-76 (S76) within its cytoplasmic tail. Using a phosphorylation-specific antibody, we now find that CD43 phosphorylation at S76 is enhanced by migration signals. We further show that CD43 phosphorylation and normal T-cell trafficking depend on CD43 association with ERM proteins. Interestingly, mutation of S76 to mimic phosphorylation enhances T-cell migration and CD43 movement to the DPC while blocking ERM association, showing that CD43 movement can occur in the absence of ERM binding. We also find that protein kinase CΘ can phosphorylate CD43. These results show that while CD43 binding to ERM proteins is crucial for S76 phosphorylation, CD43 movement and regulation of T-cell migration can occur through an ERM-independent, phosphorylation-dependent mechanism.

FIGURE 1:

CD43 S76 phosphorylation is modulated by PMA, chemokine, and CD43 cross-linking. (A–C) T lymphoblasts were activated by PMA, CCL21, or α-CD43 (R260) for indicated times; lysed; analyzed on SDS–PAGE; and blotted for phospho-S76 and total CD43 on the same membrane. The molecular weight markers are visible only in one channel of the dual channel detection on the Li-Cor Odyssey. (B) Quantitation of S76 phosphorylation from (A) normalized to total CD43. (C) Quantitation of S76 phosphorylation in response to PMA and CCL21 from an average of three independent experiments.

FIGURE 2:

Failure to bind ERM proteins decreases CD43 S76 phosphorylation. (A and B) T-cells from Tg mice expressing CD43-FL or CD43-NGG were purified; activated with PMA, CCL21, or anti-CD43 (R260) for indicated times; lysed; and blotted for phospho-S76 and total CD43 (S11). (B) Quantitation of S76 phosphorylation shown in (A) normalized to total CD43.

FIGURE 3:

Ezrin/moesin double-deficient T-cells show defective CD43 S76 phosphorylation in response to chemokine signaling. (A and B) Ezrin^+/+ or ezrin^−/− T-cells were transfected with siRNA duplexes to murine moesin or control siRNA as described in Materials and Methods. Ez^+/+Mo^siC or Ez^−/−Mo^siM cells were either unstimulated or stimulated with PMA or CCL21 as indicated, lysed, and analyzed by SDS–PAGE. (A) Samples were blotted for ERM and actin to determine the level of moesin knockdown and actin blotted as a loading control (top). Cells were then analyzed for CD43 S76 phosphorylation, and level of S76 phosphorylation and total CD43 are shown from one representative experiment (middle and bottom). (B) Quantitation of S76 phosphorylation in response to PMA and CCL21 normalized to total CD43 (average of three independent experiments). ***p < 0.001 using two-way analysis of variance comparing CCL21-induced pS76 phosphorylation in T-cells expressing Ez^+/+Mo^siC and Ez^−/−Mo^siM at all time points.

FIGURE 4:

CD43 binding to ERM proteins regulates T-cell trafficking. (A and B) DO.CD43^−/− T-cells were transduced with CD43-FL or CD43-NGG mutants, harvested, purified, and labeled as described in the Materials and Methods. (A) Equal numbers of FL and NGG T-cells were mixed and added to the top of a 3-μm Costar Transwell apparatus. The ratio of FL and NGG T-cell migration was assayed after addition of 300 ng/ml CCL21 to the bottom of the transwell for 1 h. ***p < 0.001. (B) CD43-FL T-cells were mixed with equal numbers of CD43-NGG T-cells and injected into recipients. Blood, spleen, and lymph nodes were harvested and analyzed for ratio of CD4⁺ FL:NGG cells within the dyed population by flow cytometry. Horizontal bars represent the mean and the SD shown.

FIGURE 5:

Constitutive phosphorylation at S76 partially reconstitutes T-cell migration in the absence of ERM protein binding. DO.CD43^−/− T-cells were transduced with CD43-FL, CD43-S76D, CD43-NGGS76D, or CD43-NGG mutants; harvested; and labeled. (A) CD43-S76D– and CD43-FL– or (B) CD43-NGGS76D– and CD43-NGG–expressing T-cells were added to a 3-μm Costar Transwell apparatus and migration was assayed as described in Figure 3. ***p < 0.001.

FIGURE 6:

Constitutive phosphorylation at S76 does not reconstitute CD43 movement to the T-cell uropod. DO.CD43^−/− T-cells were transduced with CD43-FL, CD43-NGGS76D, or CD43-NGG mutants; harvested; and labeled. Purified retrovirally transduced T-cells were treated with SDF1α and processed for immunofluorescence. Cells were scored for CD43 localization to the uropod. (A) Representative T-cells expressing CD43-FL, CD43-NGG, and CD43-NGGS76D mutants stained for tubulin and CD43. Scale bar indicates 20 μm. (B) Average of the quantitation of CD43 localized to the uropod in at least 50 cells in three independent experiments. Percentage of localization was normalized to CD43-FL as 100%.

FIGURE 7:

Constitutive phosphorylation at S76 partially reconstitutes CD43 movement to the distal pole complex. DO.CD43^−/− T-cells were transduced with CD43-FL, CD43-NGGS76D, or CD43-NGG mutants; harvested; and labeled. Purified retrovirally transduced T-cells were conjugated to OVA peptide–loaded, CellTracker™ Blue (7-amino-4-chloromethylcoumarin) (CMAC)–stained A20 B-cells and processed for immunofluorescence. (A) Representative T-cell–B-cell conjugates are shown with the B-cell labeled with CMAC in blue, CD43 in red, and PKCθ in green. Scale bar indicates 10 μm. (B) Cells were scored for CD43 localization to the DPC as previously described. Average of the quantitation of CD43 localized to the DPC in at least 50 T–B-cell conjugates in four independent experiments. Percent localization was normalized to CD43-FL as 100%. **p < 0.01.

FIGURE 8:

CD43 phosphorylation does not enhance ERM binding. (A and B) 293T cells transfected with Ez-FERM-GFP were lysed and lysates incubated with the indicated GST CD43 fusion proteins confirmed for equal loading by Coomassie stain. Immunoprecipitates were processed by SDS–PAGE and blotted with anti-GFP to detect ezrin. (B) Quantitation of ezrin binding phosphorylation shown in (A) normalized to GST signal in three independent experiments. *p < 0.05; **p < 0.01.

FIGURE 9:

PKCθ can phosphorylate CD43 at S76. (A) T lymphoblasts were incubated in the presence of indicated inhibitors of specified kinases (in parentheses) for 2 h, cells were lysed, and lysates were analyzed on SDS–PAGE and blotted for phospho-S76 and actin. (B) GST, GST-CD43, and GST-CD43 mutant fusion constructs were incubated with or without purified PKCθ, analyzed by SDS–PAGE, detected for anti–phospho-S76, and stained with Coomassie stain.