VI. Binding to MHC/Peptide Ligands

In the structure of the 2C TCR complexed with its class I MHC(H-2K^b)-peptide(dEV8) ligand, several hallmarks of T cell antigen recognition are evident. The basis for MHC restriction is elucidated, as the TCR's primary contact is made with the alpha helices of the H-2K^b molecule, which shield the dEV8 peptide in the binding groove. The relatively high incidence of TCR alloreactivity and the ability of the TCR to bind multiple ligands (a prerequisite for thymic selection) are similarly explained by the structural plasticity and unoptimized shape complementarity of the binding site. Finally, the orientation of the six CDR loops reveals the TCR's ability to "scan" several MHC molecules for optimal interaction with bound peptide.

Each CDR loop of the TCR performs a specific function in binding the H-2K^b-dEV8 ligand. CDR2α and 2β lie directly above the α2 and α1 helices of H-2K^b, respectively, and interact exclusively with the MHC framework. The two CDR1 loops are positioned to bind both peptide and MHC residues, with CDR1α docking above the amino-ternimus of the bound peptide and CDR1β docking above the carboxy-terminus. The two CDR3 loops extend into the peptide-binding groove and interact directly with the antigenic peptide, making minimal contact with the MHC framework. The fine specificity of antigen binding is thus determined primarily by CDR3α and CDR3β.

While CDR3α and β finely discriminate among bound peptides, the remaining CDR loops provide a general binding orientation to the MHC ligand. A standard docking engagement allows the TCR to superficially interact with several different MHC-peptide complexes, inducing the activation signal only upon binding a strongly antigenic ligand. This interaction of CDRs 1 and 2 with H-2K^b predominantly occurs at conserved residues on the MHC helices, and forms the basis for the class I or class II restriction of the TCR. Ser27α(CDR1) and Ser51α(CDR2) make critical hydrogen bonds with the conserved H-2K^b residues Glu58 and Glu166 to help define the binding orientation and MHC restriction of the 2C TCR.

When compared with the TCR's strong interaction with H-2K^b, the complementarity of CDR3α and b with the dEV8 peptide is surprisingly weak. (dEV8 is a self peptide derived from the murine mitochondrial respiratory protein complex and has the sequence EQYKFYSV.) The peptide's most prominent structural features are two fully extended residues at Lys4 and Tyr6 which seem ideally positioned to engage the TCR. The lack of complementary residues on the CDR3 loops, however, prevents this interaction. CDR3α forms only three weak polar associations with the charged nitrogen atom of Lys4, all of which are derived from main chain carbonyl groups. CDR3β accomplishes only one hydrophobic contact with Tyr6, and the prominent cleft between the two CDRs remains empty. The large proportion of empty space within the binding pocket contributes to the low affinity of the complex, and suggests that the dEV8 peptide may act as a generic antigen for several TCRs during thymic development.

The weak association between the 2C TCR and the self MHC-self peptide complex of H-2K^b-dEV8 acts to positively select thymocytes expressing the 2C TCR. The fact that this same interaction cannot elicit a strong activation signal in the periphery highlights the dual nature of intracellular signaling through the TCR. A much stronger ligand for the 2C TCR is the H-2K^b complex with the synthetic peptide SIYRYYGL, which causes negative selection in the thymus as well as robust activation in the periphery. Presumably, the substitution of upward-facing residues in the center of the peptide allows for better shape complementarity with the CDR3 loops, resulting in activation through the TCR complex.

In a similar manner, alteration of the MHC ligand can lead to increased signaling through the TCR. When the dEV8 peptide is bound to H-2K^bm3(a 2 amino acid mutant of H-2K^b), thymocytes bearing the 2C TCR are negatively selected while mature T cells in the periphery are induced to proliferate. The structurally important mutation in this case occurs at residue 77 of H-2K^b (D77S), where the loss of an aspartate residue disrupts a critical hydrogen bond with a main chain carbonyl of the peptide. This results in altered presentation to the receptor and greatly increased binding affinity with the 2C TCR.

The examples above reveal how changes in the structure of either the MHC or peptide ligand can increase the activation signal delivered through the TCR complex. It is not clear, however, how alterations in the binding affinity of MHC-peptide ligands are translated into intracellular activation signals. Because a crystallographic structure of the TCR bound to a strongly agonistic ligand has not yet been solved, predictions must be made from the present complex of 2C-H-2K^b-dEV8. Firstly, the complex exists as a monomer, which contradicts both the observation that class II MHC molecules crystallize as dimers and the prediction that TCR signaling occurs through multimerization of the TCR-MHC-peptide complex. Secondly, the four domains of the 2C TCR have maintained the same orientations and internal structure as found in the unliganded receptor, including the energetically-unstable Ca domain. This supports the model of a structurally rigid TCR, and challenges the notion of a conformational change to induce intracellular activation. Minor alterations in the structure of the TCR variable domains, however, do occur upon antigen binding, perhaps indicating greater conformational rearrangement upon true activation. Most notably, CDR1α is angled away from the peptide binding groove, and CDR3α is displaced to accommodate the extended side chain of peptide residue Lys4. The movement of CDR3α is quite dramatic (6 angstroms), and is accompanied by internal rotation of extended amino acid side chains such as Phe100α.

Until more structures of liganded T cell receptors are available, the exact nature of TCR antigen binding and intracellular activation will remain a mystery. From the 2C structure, however, it appears that the TCR is designed to recognize antigen-presenting MHC molecules from a pre-determined orientation. This may allow the TCR to effectively "scan" several antigens until a precise complementarity of fit is obtained with the bound peptide. Higher affinity binding may result in either a conformational change or the aggregation of the receptor, or perhaps subtle conformational changes allow the TCR to form multimers when bound to its MHC-peptide ligands. In any case, the T cell receptor is remarkably adept at recognizing foreign peptides in the context of self MHC molecules, while the TCR complex as a whole initiates a variable level of intracellular activation for thymic selection or clonal expansion.