46
Chapter II
3
Chapter II: Structural effects of destabilizing peptide-
MHC
II interactions and implications for the peptide
exchange mechanism of HLA-DM
3.1
Introduction
Elucidating the molecular basis for the generation of ligands presented to CD4+ T
cells requires a deeper understanding of how MHC II molecules are loaded with
antigenic peptides catalyzed by DM. Besides susceptibility to proteolytic cleavage
resistance to DM catalyzed peptide release is a major factor in determining whether or
not a particular peptide is presented on the cell surface. Therefore, understanding the
mechanism of DM mediated peptide loading and release would promote efforts to
predict immunogenicity of known and emerging pathogens. However, the molecular
mechanism of DM catalysis is not well understood.
The co-crystal structure of the DR/DM complex would expose crucial interacting
residues and also reveal the overall conformation both proteins adapt in the complex,
thus shedding light on the mechanism of how DM catalyzes peptide exchange of
MHC II molecules. DM catalyzes exchange of peptides with various sequences (Weber
et al., 1996) suggesting that DM might partially target MHC II-peptide interactions that
are peptide sequence-independent. Recent studies have shown
that release of the peptide
N-terminus of the peptide-binding groove of MHC II molecules is necessary in order
for DM to bind DR molecules and catalyze peptide exchange (Anders et al., 2011). In
particular, disruption of the four N-terminal conserved hydrogen bonds and release of
the peptide anchor residue from the P1 pocket seems to be necessary for DM binding.
Taking this new information into account, a new approach was designed to crystallize
the DR/DM complex using a modified DR molecule missing the three N-terminal
peptide residues which exhibited higher affinity for DM than DR complexes with full
length peptide. As the complex was not covalently linked the affinity between both
molecules had to be optimized to favor crystallization of the complex versus the
individual components. Therefore, surface plasmon resonance experiments were carried
out to determine the dissociation constant of the complex at different pH to direct
crystallization screens.
47
Chapter II
Furthermore, the structure of the DR1 molecule carrying an N-terminally truncated
peptide was pursued, as the implications of the partial peptide release for the DR1
conformation were unknown. The structure of the modified DR1/peptide complex could
indicate whether conformational changes occur in the DR region around the missing
peptide N-terminus or whether the DR conformation is stable, even if the N-terminal
peptide residues are missing and several hydrogen bonds are not formed. Of particular
interest is the conformation of the extended strand of the DRα chain around residues
Serα53 and Pheα51, which is adjacent to the peptide N-terminus and forms an anti-
parallel β-sheet with the peptide. This region includes residues shown to be involved in
DM activity (Anders et al., 2011; Doebele et al., 2000). The DR1 structure, with a
partially filled peptide-binding groove, might reveal a DM receptive DR conformation
giving further insight into the process of peptide exchange by MHC II molecules. To
prepare DR1 complex with the very low-affinity HA peptide variant, the peptide is
covalently linked to the DRα chain in the peptide-binding groove via a disulfide bond.
However, the yield of complex formation was very low. Since large amounts of
DR1/peptide complex are needed for crystallization experiments, the protocol for the
peptide loading and covalent linkage reaction had to first be optimized.
3.2
Materials and Methods
3.2.1
Preparation of biotinylated HLA-DM and HLA-DM mutant
Soluble DM and DM mutant (DMα R98A-R194A, DM mut), which shows no
binding to DR, were expressed in
Sf9 insect cells. The cells were infected with
recombinant Baculovirus (MOI > 5) and proteins purified after 3 days. Both proteins
carried a C-terminal BirA tag (GLNDIFEAQKIEWHE) at the α-chain and a protein C
tag at the ß-chain. Two N-linked glycosylation sites (DMα N165D and DMß N92D)
were removed to diminish heterogeneity leaving the DMα Asn15 glycosylation site.
Proteins were purified by anti-Protein C affinity chromatography (Roche Applied
Science) and aggregates were removed by gel filtration (Superose 6, GE Healthcare).
BirA was added at a molar ratio of 1:20 (BirA:protein) for site-specific biotinylation
of the BirA tag. The reaction was incubated for 2 h at 30 ºC at
a protein concentration of
2-3 mg/mL in the presence of biotin (100 μM), ATP (10 mM) and protease inhibitors.