5
Introduction
1.3
Structure and function of MHC class II/peptide complex and
comparison with MHC class I/peptide complex
Structure of MHC II/peptide complex
MHC II molecules are heterodimers consisting of an α and a ß chain which associate
non-covalently and are similar in size with about 190 residues each. Both chains are
type I transmembrane proteins which span the membrane once and have a C-terminal
cytosolic domain and an N-terminal extracellular domain. The external part of each
chain contains two domains - α1, α2 and β1, β2, respectively (figure 1.3, B). The
membrane-proximal domains (α2, β2) show structural similarities with immunoglobulin
domains whereas the membrane-distal domains (α1, β1) pack closely together and form
the peptide-binding groove. This extended cleft where a peptide can bind is composed
of a ‘platform’ shaped by an anti-parallel ß sheet consisting of eight strands (four from
the α1 domain and four from the ß1 domain) and two α helices (one from each domain)
flanking the groove (figure 1.3, D).
Comparison with MHCI/peptide complexes
The three-dimensional structure of MHC II molecules is very similar to the
appearance of MHC I molecules although the subunit structure is distinct as MHC I
molecules consist of an chain subdivided into 1, 2 and 3 with 1 and 2 forming
the peptide-binding groove and 3 an Ig-like domain (figure 1.3, A). The other Ig-like
domain is formed by
2
-microglobulin. But the major structural and functional
difference between MHC I and II molecules lies in the ends of the peptide-binding cleft
(figure 1.3, C, D). In MHC I molecules the termini of the peptide are tightly bound at
either end of the cleft by several contacts with conserved residue clusters and the typical
peptide length is 8-10 residues. In comparison, peptides bound to MHC II molecules
extend beyond the cleft and are usually 13 amino acids long or longer. The length of
bound peptides is not constrained and partially unfolded proteins have been shown to
bind to MHC II molecules through the peptide-binding groove (Runnels et al., 1997).
However, it is likely that proteins and longer peptides are trimmed by proteases to a
length of 13-17 residues.