what-when-how
In Depth Tutorials and Information
each new mutation to a new specific OI phenotype. This
concern was considered by the Nosology Group of the
International Skeletal Dysplasia Society which met in
August 2010.
16
In addressing the extensive phenotypic
variability arising from single loci, it was considered not
possible to maintain close correlation between “Sillence
types” and their molecular etiology. It was stated that the
proliferation of OI types to reflect each gene separately
was confusing and that the responsible genes should be
listed separately.
16
For similar reasons, van Dijk et al. have
proposed continued use of the Sillence criteria I, II-A, II-B,
II-C, III and IV for clinical and radiological classification
of OI with addition of types V and VI because of their
distinguishing clinical/radiological and/or histological
features.
13
However, type VI OI, which is associated with
moderately severe skeletal disease, is due to mutations in
the gene SEREPINF1 which encodes the pigment-derived
epithelial factor (see Chapter 17). This phenotype is some-
what anomalous as an OI phenotype as the basic histo-
logic feature is marked osteomalacia, consistent with a
mineralization defect, which is not characteristic of other
phenotypes.
17,18
However, current literature includes limited reference
to types IX, X and XI OI, each defined based on a DNA
mutation in a specific protein. As listed by Forlino et al.,
these involve mutations in, respectively, PPIB encoding
cytophyllin B, SERPINH1 encoding HSP 47 and FBK10
encoding FKB 65.
19,20
The editors currently suggest lim-
iting the defined OI phenotypes to numbers I-VIII. This
encompasses Sillence types I-IV and adds the recently
described recessive phenoytypes VI, VII and VIII.
OI is characterized by marked phenotypic variation
in the expression of the type I collagen-related genes
(COL1A1 and COL1A2), with genes responsible for the
synthesis of chaperone proteins, and non-collagen-related
genes (BMP1, Osterix/SP7, IFITMIT5, TEM38B and
Wnt1) known to cause an OI phenotype.
21,22
The various
phenotypes are the result of different mutations including
point substitutions, deletions, insertions and splice site
mutations. These are discussed in Chapters 18 and 19.
Ninety-eight percent of the mutations associated with
OI involve the pro-alpha-1 and pro-alpha-2 polypep-
tide chains of type I collagen, the genes encoding sev-
eral proteins which modify type I collagen propeptide
chains during chain assembly in the rough endoplasmic
reticulum or recently reported genes which alter the
Wnt pathway or receptors related to intracellular cal-
cium transport (Chapter 58). The clinically mild forms of
OI are usually caused by mutations leading to haploin-
sufficiency for the COL1A1 gene, while the severe and
lethal phenotypes classified as either Sillence type II OI
or OI types VII and VIII result from dominant/negative
mutations in COL1A1 or COL1A2 or mutations in the
CRTAP/LEPRE1/cytophyllin complex. There are now
several instances where the OI phenotype has resulted
from mutations involving non-COLIA1 or COL1A2 pro-
teins. Examples include the recent association of Wnt
protein Osterix/SP7, Wnt1 and TEM38B, mutations
involving bone morphogenetic protein 1 (BMP1) with
recessive OI, mutations in the SERPINF1 encoding pig-
ment epithelium-derived factor (PEDF) in type VI OI
and involvement of the osteoblast development protein,
ITMIT5 (interferon-induced transmembrane protein 5,
OMIM # 614757) in type V OI. These broaden, but do
not limit, the numbers of genes that may lead to a reces-
sive phenotype. Interestingly, the Bruck syndrome phe-
notype (arthrogryphosis and bone fragility) has now
been associated with mutations in two genes, PLOD 2
and FKBP10. There is debate as to whether the Bruck
syndrome should be classified with OI.
From a clinically predictive standpoint, meaning-
ful genotype/phenotype relationships have not been
defined. However, recent studies have provided con-
siderable understanding of the manner in which certain
mutations alter bone matrix synthesis. Type I collagen
is a heterotrimer containing two pro-alpha-1(I) and one
pro-alpha-2(I) chains. COLA1 (17q21.33) and COL1A2
(7q21.3) are large genes containing 51 and 52 exons,
respectively. Although mutations related to OI have
previously been limited to coding regions of the gene,
type V OI has recently been associated with a muta-
tion in the 5′ untranslated region of IFITM5 (interferon-
induced transmembrane protein 5) gene which creates
an in-frame start codon by which five amino acids
(Met-Ala-Leu-Glu-Pro) are added to the N-terminus of
the protein.
23,24
Following the association of this muta-
tion with type V OI, it became apparent that there was
marked variability in the clinical expression of the muta-
tion.
25,26
The reason for this variability is not understood;
however, this observation raises important questions
about factors modifying gene expression.
The role of the C- and N-terminal domains has
been emphasized and the scope of OI phenotypes has
recently been expanded. Mutations involving the C- and
N-terminal domains are relatively uncommon. These
lead to phenotypes that differ in certain respects from
the phenotype associated with mutations in the heli-
cal regions: first, by association of bone fragility with
joint laxity which is more pronounced than is seen in
most OI patients, and second, by association with matrix
hypermineralization which is recognized at the tissue
level. Hypermineralization may be associated as rela-
tively higher bone density measurement (DXA) when
compared to other OI patients.
27,28
The N- and C-terminal collagen propeptides are
cleaved in the extracellular matrix by specific propep-
tidases. Initial interest in cleavage of the N-terminal
propeptide developed following the discovery that the
condition of fragile skin or dermatosparaxis in sheep was
due to mutations involving the N-terminal propeptidase.
29
