Biology Reference
In-Depth Information
evolution toward extant human protein C. The persistence of the anionic poten-
tial of this area through evolutionary time, together with the report of a type II
(dysfunctional) substitution (R352W, Reitsma
et al
., 1995) in this region, implies
a functional role in both ancient and extant proteins.
10.2.7 Comparative geometry of the active sites of early mammalian and
extant protein C
Extant human protein C and early mammalian protein C exhibit 83.6% sequence
identity at the amino acid sequence level, and exhibit a backbone RMS diver-
gence of 2.6 Å (computed against a model of the serine protease domain of extant
human protein C; Wacey
et al
., 1993). Recently, a relatively low resolution crystal
structure of human APC (des Gla domain) has been solved (Mather
et al
., 1996).
Comparison of our model with this structure served to demonstrate that the
majority of functionally important residues identified in extant protein C were
already present in its early mammalian predecessor. The Ca
2+
binding loop
(residues 70-80 with the exception of N78), the active site loop (residues 146-152)
and the insertion helix (residue 129) are all to be found intact in early mammalian
protein C. The asymmetric surface charge distribution of extant protein C is also
apparent in early mammalian protein C. Finally, residues E192 and Y143 (the
major determinants of substrate specificity) which span the active site, are present
in both proteins.
Since the regions close to the catalytic triad of extant human and early mam-
malian protein C are likely to possess a similar functional architecture (owing to
the minimal divergence in the template structures underlying the two models), the
active site geometries of the two proteins may be reliably contrasted at relatively
high resolution. Within the catalytic site pocket of extant human protein C, the S2
residue (Ser 198) is predicted to have arisen by substitution of a Phe residue pre-
sent in early mammalian protein C. This would imply that an initially tightly fit-
ting catalytic pocket became more capacious during evolution. Why, however, the
substrate specificity of extant human protein C is restricted to factors Va and VIIIa,
although its active site pocket can theoretically accommodate larger substrates,
remains unclear. With this exception, the amino acid sequence of the active site
pocket of early mammalian protein C is identical to that of extant human protein
C. Thus the structural features noted in the crystal structure of human APC would
probably also have been found in the early mammalian protein.
A prominent hydrophobic and solvent-accessible region is found on the surface
of the serine protease domain of early mammalian protein C. This region has been
shown to bind the second (C-terminal) EGF-like domain of the light chain in fac-
tors Xa, IXa, and protein C (residues 100-116 and 45-50 respectively). Early
mammalian protein C may therefore have possessed a light chain, an hypothesis
consistent with Patthy's (1985) view of serine protease evolution.
In summary, the application of homology modeling to a reconstructed amino
acid sequence allowed Wacey
et al
. (1997) to trace the evolution of specific
structural features in protein C. This approach provided new insights into the
evolution of protein C, allowed an assessment of the nature of the minimal throm-
bomodulin binding site and permitted inferences to be made as to the possible
