what-when-how
In Depth Tutorials and Information
residues.
11
Remarkably, the ratio of glucosylgalactosyl-
hydroxylysines in null to wild-type animals was the high-
est in bone (2.40 fold, 1.49 in skin and 1.28 in tendon). The
number of hydroxylysines was increased 1.42 times in
bone null type I collagen, 1.32 times in skin and 1.08 times
in tendon.
11
The mouse data are in good correlation with
the data obtained from patients with mutations in P3H1.
Type I collagen produced by patient fibroblasts had on
average 1.4 times more hydroxylysines compared to con-
trol.
10,17
There was no correlation between the amount of
the residual 3-hydroxylation of Pro986 and the degree of
overmodification of collagen.
we think that different mutations might impose different
effects on collagen modifications. However, the conclu-
sion that type I collagen folds normally in a CypB patient
is solely based on gel migration.
63
An alternative expla-
nation of these results is that collagen folding is slow in
these cells, but the lack of CypB reduces the activity of
LH1 and the combination leads to normally modified
collagen chains. Unpublished data from our laboratory
show that type I collagen extracted from skin of CypB
null mice contains less hydroxylysine, consistent with the
data obtained from the HC horse.
71
Secretion Rates of Type I Collagen in Mice and
Patients with Mutations in the Complex
An important consideration for understanding the
leading mechanism in OI is the rate of collagen secretion.
The secretion rate of properly folded collagen is a direct
measure of the quality and efficiency of collagen folding.
Misfolded collagen is either degraded or stored in the ER
causing ER stress.
80
In general, the longer the collagen
chains remain in the unfolded state the larger the extent
of overmodifications (the modifying enzymes only rec-
ognize unfolded collagen as a substrate).
81
Deficiencies in CRTAP Lead to Overmodified
Type I Collagen
A similar picture is seen in cases with CRTAP muta-
tions. Although the amino acid analysis has not been
done for the CRTAP null mouse collagen, it was per-
formed for collagen synthesized by patients' fibroblasts.
The results demonstrated a 28 to 42% increase in the
amount of hydroxylysine depending on the mutation.
62
In other cases, the ratio of fibroblast-synthesized collagen
hydroxylysines in affected individuals to the normal con-
trols averaged at around 1.4 times.
10
Overmodiication
of collagen was also suggested qualitatively based on
the gel shift in cases when amino acid analysis was not
performed.
7
P3H1
Our group has demonstrated a slower secretion rate
for the type I collagen in the P3H1 null mouse together
with lower amount of secreted collagen.
11
Patients with
P3H1 null mutations were also shown to have moder-
ately delayed collagen secretion even though the amount
of the secreted collagen was increased.
17
Therefore type
I collagen folding is slower in P3H1 deficiencies, but the
overall amount of the secreted collagen can differ.
The Mystery of CypB and Collagen Modification
The situation with overmodification of collagen in
cases of CypB mutations is somewhat controversial. Two
reports showed overmodified type I collagen secreted
by fibroblasts of CypB knockout mouse
15
and human
patients.
13
The assessment in these publications was
done based on the gel shifts and no amino acid analysis
to determine the amount of hydroxylysines or
/
and sug-
ars was performed. However, one report demonstrated
the lack of overmodification in type I collagen in a human
CypB mutation providing data for a normal hydroxyly-
sine content.
63
It has to be noted that the assessment of
the mutation in the work by Barnes et al. was done using
a
PPIB
(gene for CypB) sequence different from the cur-
rent reference sequence.
79
Barnes and coauthors, how-
ever, believe that the sequence they used is the correct
one (response to the comment
79
). A recent study from our
group found that another mutation in CypB, which does
not cause OI but causes hyperelastosis cutis (HC) in the
American Quarter Horse, leads to fewer hydroxylysines
and sugars in type I collagen.
71
In this case the muta-
tion might disrupt the interaction of cyclophilin B with
the P domain of calreticulin and with lysyl hydroxylase
1. Considering the versatile functions assigned to CypB
and its presence in several different protein complexes,
68
CRTAP
In the case of CRTAP alterations, the type I colla-
gen secretion rate was determined for the null mouse
18
but not for patients. In contrast to the data obtained for
P3H1 cases, fibroblasts from the CRTAP null mouse had
an increased secretion rate of type I collagen compared
to the control. However, no quantification of the gel was
presented. In the published gel, collagen was present
at every time point. The data presented definitely dem-
onstrate an increase in the amount of secreted collagen
(which is consistent with P3H1-deficient patients
17
) but
not necessarily in the secretion rate.
CypB
A recent report has shown delayed collagen fold-
ing in the fibroblasts derived from an individual with
