what-when-how
In Depth Tutorials and Information
FKBP65 might act as such a chaperone,
83
but FKBP65
deficiency does not seem to delay procollagen folding
in vivo
.
85
For additional discussion of potential roles of
different ER proteins in procollagen folding, we refer the
reader to the preceding chapter in this topic.
The response of cells to procollagen misfolding is also
distinct. C-propeptide misfolding causes sequestration
and upregulation of BiP, activates UPR signaling, and
leads to ER associated degradation (ERAD) of misfolded
chains by proteasomes.
86-88
Yet, triple-helix misfolding
does not trigger these conventional responses.
86,89-91
Instead, misfolded procollagen forms insoluble aggre-
gates, which are delivered to lysosomes by autophagy
and degraded there.
91
Similar features were observed
in so-called ER overload response to misfolding and
aggregation of some proteins from the serine protease
inhibitor (serpin) family,
92
but the ER stress response
to procollagen triple-helix misfolding is still poorly
understood and might not involve the same signaling
pathways.
91,93
Substitutions of obligatory Gly, X-position Pro,
Y-position 4-Hyp, and Y-position Arg residues make
procollagen folding more difficult, since they further
destabilize the triple helix. For instance, Gly substitu-
tions reduce the apparent triple-helix
T
m
by Δ
T
m
~2°C
on average; and some reduce it as much as ~ 5°C.
60
Cells manage to fold these molecules anyway, prob-
ably by upregulating the corresponding chaperones.
For instance, HSP47 upregulation was reported for Gly
substitutions.
94
Still, triple-helix destabilization increases
procollagen misfolding, results in retention of misfolded
molecules in the ER, increases ER stress, and thereby
affects the function of collagen-producing cells.
ER stress response to procollagen misfolding might
have particularly severe consequences for osteoblasts,
since these cells produce massive amounts of type I
collagen. Hence, ER stress has been proposed to be a
major factor determining the severity of OI.
80,89,90,95,96
However, not only the mechanism of this response, but
also the extent of the underlying ER stress is still poorly
characterized, even for the most common and best stud-
ied OI cases caused by Gly substitutions. The complex-
ity of the relationship between such substitutions and
the resulting procollagen misfolding may be illustrated
on the example of Δ
T
m
, which is a measure of the colla-
gen triple-helix destabilization by the mutation.
One is intuitively tempted to use Δ
T
m
as a proxy
for the extent of procollagen misfolding, but such an
assumption turns out to be wrong. There is actually
no simple relationship or correlation between the Δ
T
m
value and the extent to which procollagen folding is
disrupted by the substitution. For instance, Gly sub-
stitutions with Arg, Glu, Asp or Val cause much larger
disruptions in the triple-helix folding and structure near
the mutation site than substitutions with Ala or Ser.
97,98
As a result, Arg, Glu, Asp and Val substitutions should
cause longer folding delays, more extensive misfold-
ing and accumulation of procollagen in the ER, as well
as more severe ER stress than Ala and Ser substitutions.
However, all of these mutations produce similar Δ
T
m
when they occur at the same location within the triple
helix.
60,99
This seeming discrepancy may be explained as
follows. The value of Δ
T
m
depends on the contribution
of a region surrounding the mutation site to the stabil-
ity of the whole triple helix.
60
In contrast, the extent of
procollagen folding disruption depends on the abil-
ity of the triple-helix folding process to accommodate
the substitution and proceed through the mutation
site.
98,100
When a Gly substitution is encountered, the
zipper-like folding process pauses until the triple helix is
re-nucleated on the N-terminal side of the mutation.
37,101
The re-nucleation efficiency is affected by the location of
Gly-X-Y triplets with high triple-helix propensity and
stability and HSP47 binding sites on the N-terminal side
of the mutation as well as by the ability of the helix to
accommodate the mutation. In particular, a Gly substitu-
tion on the C-terminal side of a large region with low tri-
ple-helix stability that has few or no HSP47 binding sites
is likely to cause a major disruption of procollagen fold-
ing and accumulation of misfolded molecules in the ER.
At the same time, such a mutation might cause smaller
than average Δ
T
m
because of the minimal contribution
of this region to the stability of the whole triple helix.
For additional discussion of Gly substitution effects on
the triple-helix structure and folding, see other chapters
in this topic.
Despite these complexities, the expected severity
of ER stress caused by procollagen misfolding might
explain at least some of the observed trends in the sever-
ity of OI caused by Gly substitutions, as illustrated in
Figure 7.2
. Indeed, panels C and D in
Figure 7.2
show
that Gly substitutions tend to be more severe in the
α1(I) chain compared to the α2(I) chain, which might be
explained by a higher fraction of molecules containing
mutant α1(I) chains (heterozygous mutations in the α1(I)
and α2(I) chains are expected to produce 75 and 50%
molecules with mutant chains, respectively, since chain
association at the C-propeptide should not be affected
by triple-helix mutations). More severe OI caused by
Gly substitutions with Arg, Glu, Asp or Val compared
with Ala or Ser might be explained by larger disrup-
tions in the triple-helix folding and structure introduced
by the former.
97,102
The lack of apparent correlations of
OI severity with Δ
T
m
62
might be explained by the lack
of a simple relationship between Δ
T
m
and the disrup-
tion of procollagen folding.
60
Milder OI phenotypes of
substitutions near the N-terminal end of the triple helix
might be explained by weaker ER stress response to
delayed folding of a shorter part of the triple helix, since
such molecules are likely to have lower probability of
