Within the Class II receptors the
Within the Class II receptors, the Type I IFN receptor subunit, IFNAR-1, is exceptional in having four tandem FNIII domains, denoted here D1–D4. This structure appears to have arisen as a tandem ochratoxin a mg of the basic D1/D2 structure; thus, D1 and D3 of IFNAR-1 are more closely related, as are D2 and D4 (see also ).
Before focusing on the extracellular ligand-binding domain, it is important to note a functional typology among the intracellular (IC) domains, even though they lack the sequence and pattern similarities found in the extracellular domains. Within each heterodimeric receptor, the intracellular domain of one subunit is considerably larger than that of its partner (diagrammed in Fig. 1). In most receptors, one subunit also has substantially higher affinity for ligand than the other. The subunit with higher ligand affinity also has the larger intracellular domain; this subunit can be denoted the “R1” subunit. The “R2” subclass of receptor subunits have smaller intracellular domains and generally have lower intrinsic affinity for ligand. Thus, each functional receptor is composed of an R1 and R2 subunit, with larger and smaller intracellular domains, respectively (see Table 1 and Fig. 1). To trigger intracellular signals, both R1 and R2 subunits associate with Jak family tyrosine kinases; in most cases examined thus far, it is the R1 type subunit that has been found to be phosphorylated on Tyr residues following ligand binding. The phosphorylation of R1 promotes the recruitment of other signaling molecules to its intracellular domain, thereby contributing heavily to the specificity of cytokine signaling. While the R2 subunit recruits an additional tyrosine kinase to the receptor complex, it does not appear in most cases to determine the specificity of signaling, i.e., the major outcome of downstream signaling events. The role of the IC domain in conferring specificity was partially determined using chimeric cytokine receptors engineered with the IC domains of one receptor fused to the ECD of another. These chimeric receptors have also been important for studying ligand specificity and intracellular signaling , , , .
Into the future Structural (crystallographic or NMR) data or structure/function mutagenesis data on ligand–receptor complexes are only available for 4 of 12 CRF2 proteins. Nevertheless, mapping the receptor contact residues from crystal structures and from mutagenesis results onto the sequence alignments provides a fairly consistent picture of the use of various parts of the receptors in ligand binding (Fig. 3, Fig. 5). As with several Class I receptors, many of the ligand binding interactions are dominated by hydrophobic receptor residues, particularly the aromatics, often flanked by charged residues , . While these central residues may change affinity by >5–10 fold, accurate biophysical measurements, such as BIAcore or reflectometric interference spectroscopy, with soluble extracellular domains are needed to confidently determine 2–3-fold changes in affinity which may also be important. However, the general dominance of large hydrophobic receptor residues does not hold in all cases, and cannot be assumed. Furthermore, the importance of identifying and understanding residues in addition to the strongest contributors to binding is perhaps illustrated by bovine IFNAR-1. In this case, the hydrophobic residues implicated as key to ligand binding are conserved between species; thus, neither the large differences between the bovine and human orthologs nor the discrimination of different ligands is resolved by the identification of these residues. In other cases, such as IL-20R1, IL-20R2 and IL-22R1, where a single receptor subunit interacts with different ligands, the detailed interactions with each ligand will also be of particular interest. The importance of having both structural studies and mutagenesis results, emphasized previously by Wells and others , is also clear in these studies. Structures of ligand–receptor derived from X-ray crystallography or NMR provide beautiful views of the interacting surfaces, which, for these complexes, are generally extensive, involving about 1000Å2 of surface area and 15–20 residues on each side of the interface. However, the structural studies do not demonstrate the relative importance of any residue for overall interaction (although “guesstimates” can be made from buried surface calculations, or more extensive calculations can be done of the energetics of interactions). Mutagenesis is critical to examine the roles of the individual residues. Conversely, in the absence of high-resolution structural data, mutagenesis can be combined with homology modeling to provide important data and testable hypotheses regarding the specific mechanism of ligand binding and, in the case of receptors with multiple ligands, ligand discrimination.