what-when-how
In Depth Tutorials and Information
FIGURE 7.2
Collagen triple-helix folding and OI severity. (A) HSP47 binding sites and large regions with low (flex regions) and high (anchor
regions) triple-helix stability
/
propensity in human type I collagen. HSP47 binding sites were mapped from the sequences reported for triple-
helical peptides.
61
The map of low and high stability regions updates the previously published one
60
based on Δ
T
m
measurements for additional
Gly substitutions. Less stable regions with fewer HSP47 binding sites are the most likely bottlenecks for the procollagen folding process. (B)
Severity scores for Gly-to-Ser substitutions in the α1(I) chain based on patient phenotypes reported in the online OI variant database
54
prior to
May 1, 2012. Individual substitution scores (filled circles) were assigned as follows: mild mutations reported as type I OI were given the score 1,
type I
/
IV OI - score 1.5, type IV OI - score 2, type III
/
IV OI - score 2.5, type III OI - score 3, type II
/
III OI - score 3.5, and most severe type II OI -
score 4. Averaged scores for substitution sites (open circles) were calculated by averaging the scores for all substitutions at the same site and then
averaging the resulting values for sites located within five Gly-X-Y triplets. The five triplet size of the “running average” window was based
on the correlation length of fluctuations in the triple-helix structure estimated from H-D exchange experiments.
60
Similar scores were obtained
for running average window sizes from three to eight triplets. (C) Averaged severity scores for Gly substitutions in the α1(I) chain. Open circles
show Gly-to-Ser substitutions (same as in panel B), the bold line shows all substitutions pooled together without distinguishing the identity of
the substituting residue. Comparison of the two plots reveals three regions of extremely severe
/
lethal mutations: (1) from ~250 to ~300; (2) from
~700 to ~800; and (3) from ~900 to ~950. On the N-terminal side, these regions border
/
overlap with N-flex, M-flex2, and C-flex regions. Reduced
triple-helix stability and low density of HSP47 binding sites within the latter regions should hinder re-nucleation of triple-helix folding past
the mutation sites (cf.,
Figure 7.1
), potentially explaining the increased mutation severity. (D) Averaged severity scores for Gly substitutions
in the α2(I) chain. The peaks with increased severity correspond to previously described clusters of lethal mutations.
62
However, these peaks
might result from clustering of more severe Gly-to-Asp, Glu, Arg, and Val substitutions and not from effects of the substitution location, since
less severe Gly-to-Ser substitutions do not show the same pattern as all Gly substitutions pooled together. Overall, regional severity variations
for Gly substitutions in the α2(I) chain appear to be different from the α1(I) chain, but the reported set of mutations does not appear to be large
enough to establish a reliable pattern.
conformation is more favorable than the folded one.
BiP and GRP94 appear to be required for procolla-
gen C-propeptide folding,
70
but their binding to the
triple-helical region of procollagen chains would be
counterproductive, making folding of this region
even more unfavorable.
80
One of the chaperones that
make the triple-helix folding favorable appears to be
HSP47.
61,64,81,82
Other potential candidates include
FKBP65
83
and a complex of CRTAP, prolyl 3-hydrohylase
1 (P3H1) and a peptidyl-prolyl isomerase cyclophilin
B (CYPB),
84
but their role in procollagen folding is less
clear.
80
For instance,
in vitro
studies suggested that
