A ligand for CD5 is CD5

Abstract

Recognition by scavenger receptor cysteine-rich domains on membrane proteins regulates innate and adaptive immune responses. Two receptors expressed primarily on T cells, CD5 and CD6, are linked genetically and are structurally similar, both containing three scavenger receptor cysteine-rich domains in their extracellular regions. A specific cell surface interaction for CD5 has been difficult to define at the molecular level because of the susceptibility of CD5 protein to denaturation. By using soluble CD5 purified at neutral pH to preserve biological activity, we show that CD5 mediates species-specific homophilic interactions. CD5 domain 1 only is involved in the interaction. CD5 mAbs that have functional effects in humans, rats, and mice block homophilic binding. Ag-specific responses by mouse T cells in vitro were increased when engagement of human CD5 domain 1 was inhibited by mutation or by IgG or Fab fragment from a CD5 mAb. This showed that homophilic binding results in productive engagement. Enhancement of polyclonal immune responses of rat lymph node cells by a Fab fragment from a CD5 mAb shown to block homophilic interactions provided evidence that the extracellular region of CD5 regulates inhibition in normal cells. These biochemical and in vitro functional assays provide evidence that the extracellular region of CD5 regulates immunity through species-specific homophilic interactions.

Figure 1. Species specific homophilic binding of CD5

(a, b) hCD5-His, rCD5-His, mCD5-His and hrCD5-His migrated as monomers in SDS-PAGE analysis under nonreducing or reducing (not shown) consistent with noncovalent dimer formation in gel filtration (not shown). (c) Monomeric hCD5-His (~0.3 μM) and rCD5-His (not shown) bound immobilised CD5 domain 1 (LEU1) and 3 (CD5-48) mAbs. Over time slow binding of presumably aggregated material to the rabbit anti- mouse Fc was observed. (d) BIAcore sensorgram traces of dimeric hCD5-His (2 μM) over hCD5CD4 (1474 RU), rCD5CD4 (1941 RU), mCD5CD4 (2790 RU) and control IgSF protein rCD4d3+4 (1582 RU) immobilised on streptavidin coated flowcells. The horizontal line indicates injection period. (e, f) Equilibrium binding data at 37°C from injections of increasing concentrations of monomeric hCD5-His or hCD5CD4 over hCD5CD4 (1434 RU), hCD6CD4 (1931 RU), mCD5CD4 (1477 RU) and control IgSF protein hTREM1CD4 (1439 RU) immobilised on streptavidin coated flowcells. At the end of the experiment, species specific CD5 mAb binding were injected over all flowcells to confirm the identity of immobilised proteins.

Figure 2. Homophilic interactions between CD5 domain 1

(a-d) Chimaeric proteins contain rCD4 domains 3 and 4 (circles) hrCD5CD4, rhCD5CD4, hCD5CD4 and rCD5CD4 were immobilised (~2,000-3,000 RU) on streptavidin coated flowcells. Increasing concentrations of (a) hCD5-His, (b) rCD5-His, (c) hrCD5-His or (d) mCD5-His were injected over all flowcells at 37°C. Equilibrium binding data, relative to responses over rCD5CD4 (a, c) or hCD5CD4 (b, d) are plotted. The identity and equivalent levels of immobilised proteins were confirmed with mAb specific for hCD5 domain 1, (LEU1), rCD5 domain 1, (OX19) and hCD5 domain 3 (not shown). (c) Binding before (closed symbols) and after (open symbols) saturation of immobilised proteins with LEU1 (hCD5CD4, hrCD5CD4) or OX19 (rCD5CD4 and rhCD5CD4) is shown. (e) Recombinant proteins are shown schematically. SRCR domains (squares) of human, rat and mouse CD5 are depicted, domain 1 uppermost. (e) mCD5CD4 was saturated with 53.7 or control rCD4 mAb, OX68 mAb and mCD5-His reinjected.

Figure 3. The V88D/V97K CD5 domain 1 mutant does not bind homophilically

(a, b) hCD5CD4, L26KCD4, M124QCD4, V88D/V97KCD4 and rCD5CD4 (~3,000 RU) were immobilised on streptavidin coated flowcells. (b) Increasing concentrations of hCD5-His were injected over all flowcells at 37°C (closed symbols). (b) hCD5-His was injected over hCD5CD4 saturated with UCHT2 mAb, hCD5CD4 and V88D/V97KCD4 (closed symbols). V88D/V97KCD4 was then saturated with UCHT2 mAb and hCD5-His reinjected (open symbols). Binding to hCD5CD4 saturated with UCHT2 was the same in the two sets of injections. Equilibrium binding data, relative to responses over rCD5CD4 are plotted.

Figure 4. Homophilic interactions of CD5 regulate an immune response

(a) Mouse 2B4 T hybridoma cells were analysed by flow cytometry for expression of transfected mCD2, mCD3, mCD5 and transduced hCD5 with hCD5 domain 1 (LEU1 and UCHT2) and domain 3 (hCD5d3) mAbs. (b-c) Antigen specific IL-2 production by 2B4 hybridoma cells in response to (b) CHO IE^K or CHO IE^K expressing mCD48 (c) and (b) different concentrations or (c) 1 μM mcc peptide were enhanced by expression of V88D/V97K compared with wild type hCD5 (b) or blocking with a hCD5 domain 1 (LEU1) or Fab (LEU1) and not with a control mAb (OX21) or Fab (T12) (c). The difference between LEU1 and control is significant for both IgG and Fab (p<0.001) (d) Proliferation of polyclonally activated rat lymph node cells was enhanced by Fab fragments of rat CD5 mAb (OX19) and not control Fab (OX21) (* p<0.01). The data shown are representative of ≥ 3 independent experiments.