Joint Bone Spine 78 (2011) 252–258

Review

Genetics of bone diseases: Paget’s disease, fibrous dysplasia, osteopetrosis,

and osteogenesis imperfecta
Laetitia Michou ∗ , Jacques P. Brown
Service de rhumatologie du CHUQ-CHUL-H1365, département de médecine, centre de recherche du CHUQ-CHUL, université Laval,
2705, boulevard Laurier, G1V 4G2, Québec, Canada

a r t i c l e i n f o a b s t r a c t

Article history: Over the last few years, research into the genetics of bone diseases has produced new insights into
Accepted 23 July 2010 the pathophysiology of bone remodeling. The identification of SQSTM1 mutations in Paget’s disease of
Available online 19 September 2010 bone established that osteoclast activation involved both binding to ubiquitin and the proteasome path-
way. However, murine models fail to replicate the full phenotype, and somatic SQSTM1 mutations have
Keywords: been identified, suggesting a role for complex mechanisms. In patients with fibrous dysplasia of bone,
Paget’s disease of bone postzygotic somatic mutations in the GNAS gene are now well documented. Technological advances
Fibrous dysplasia of bone
have improved the detection of somatic mutations in peripheral blood cells. Osteopetrosis is character-
Osteopetrosis
Osteogenesis imperfecta
ized by increased bone density due to deficient osteoclastic bone resorption. Most of the genes involved
Mutation in the various clinical patterns of osteopetrosis have been identified. The identification of LRP5 gain-of-
function mutations in autosomal dominant osteopetrosis type I prompted a revision of the classification
scheme, and this form is now being included among the high-bone-mass diseases. Osteogenesis imper-
fecta is characterized by an inherited abnormality in bone formation that manifests as osteopenia with
increased bone fragility. Mutations in the COL1A1 and COL1A2 genes are found in over 90% of patients.
The recent identification of mutations in the CRTAP, LEPRE1, and PPIB genes in recessive forms has
radically changed the classification of osteogenesis imperfecta and generated new pathophysiological
hypotheses.
© 2010 Société française de rhumatologie. Published by Elsevier Masson SAS. All rights reserved.

1. Introduction the affected bones. Radiographs typically disclose sharply demar-

Recent advances in the genetics of bone diseases have shed light
on the pathogenesis of several bone diseases and on the pathophys-
iology of bone remodeling. Here, we discuss the genetics of four
bone diseases that are relatively common in rheumatology prac-
tice. Considerable progress in understanding the pathophysiology
and improving the classification of these four diseases has been
made in recent years.
2. Paget’s disease of bone
2.1. Clinical and epidemiological background
Paget’s disease of bone (PDB) is characterized by the delayed
onset of symptoms related to foci of increased bone remodeling
containing large overactive osteoclasts [1]. The clinical features
may include bone hypertrophy; deformities, predominantly affect-
ing the long weight-bearing bones; and increased skin warmth over
∗ Corresponding author. Tel.: +1 418 654 2178; fax: +1 418 654 2142.
E-mail address: laetitia.michou@crchul.ulaval.ca (L. Michou).
cated foci of sclerosis, osteolysis, thickening of the cortex, loss
of corticomedullary differentiation, a fibrillar bone structure, and
bone hypertrophy or deformities [2]. Although PDB is often asymp-
tomatic, about 30% of patients experience complications such as
bone pain, osteoarthritis, difficulty walking due to limb shorten-
ing and deformities, fissures and/or fractures, headaches, hearing
loss, cranial nerve palsies, and hydrocephalus [2]. Sarcomatous
transformation occurs in about 0.3% of patients. PDB is the second
most common bone disease after osteoporosis. The prevalence is
low before 40 years of age then increases with advancing age to
reach 3% of individuals older than 55 years. Epidemiological stud-
ies done in geographic regions characterized by a high prevalence
of PDB have documented decreases in the prevalence and sever-
ity of the disease over time. These decreases remain unexplained
but may involve exposure to an environmental factor and the
massive immigration of individuals from low-prevalence countries
such as the Indian subcontinent and South-East Asia [3]. During
the same period, the incidence of PDB remained stable in the US
and in Italy, where the disease phenotype became more severe
[3]. In about one-third of cases, PDB is an inherited disease that
is transmitted on an autosomal dominant basis with incomplete
penetrance.

1297-319X/$ – see front matter © 2010 Société française de rhumatologie. Published by Elsevier Masson SAS. All rights reserved.
doi:10.1016/j.jbspin.2010.07.010
 L. Michou, J.P. Brown / Joint Bone Spine 78 (2011) 252–258
2.2. Susceptibility locus
PDB is a genetically heterogeneous condition that has been
linked to at least eight different chromosome loci [4]. Some of these
loci, however, may be false-positive sites, such as PDB1 in the HLA
region, PDB5 on the long arm of chromosome 2, and PDB7 near
the RANK gene. PDB2, which contains RANK, has been associated
with some forms of early-onset PDB and PDB2 polymorphisms have
been linked to the typical form of PDB. PDB3, at 5q35-qter, contains
Sequestosome 1 (SQSTM1), of which several mutations have been
reported in PDB. PDB4 on the long arm of chromosome 5 and PDB6
on the short arm of chromosome 10 are being investigated. Finally,
polymorphisms affecting the TNFRSF11B gene for osteoprotegerin
at 8q24 have been linked with Paget’s disease in women but not in
men [5].
2.3. SQSTM1 gene mutations
The first SQSTM1 mutation (P392L) identified in relation with
PDB was found in large affected families in Quebec. This recurrent
mutation is at present the most commonly found mutation in fam-
ilies with PDB [6]. To date, over 20 other SQSTM1 mutations have
been reported. Most of the SQSTM1 mutations are located in or near
the region encoding the ubiquitin-binding domain [7]. The main
mechanisms reported to date are missense mutations, stop muta-
tions, and splice-site mutations. In a study of French patients with
PDB, routine testing identified SQSTM1 mutations in 13% of cases
[8]. The P392L mutation was the most common finding. Two new
mutations were identified, L413F and A381V, and two patients each
had two different mutations in the SQSTM1 gene, with no major
impact on the phenotype [8].
SQSTM1 mutations are believed to explain about 37% of famil-
ial forms and 8% of sporadic forms of PDB [6]. Although no clear
genotype–phenotype correlations have been identified to date,
stop mutations have been reported to be associated with a younger
age at diagnosis and with greater severity of polyostotic involve-
ment [7,8]. The focal nature of PDB and low prevalence of SQSTM1
mutations among patients with sporadic disease prompted studies
to look for somatic SQSTM1 mutations. One study used laser capture
microdissection to obtain cell populations from pagetic bone and
pagetic osteosarcoma in patients with sporadic PDB [9]. Sequencing
of nuclear DNA identified SQSTM1 mutations in these samples but
not in peripheral blood cells. In the patients with osteosarcoma, the
P392L mutation was found in the tumor but not in the healthy adja-
cent bone. Although these findings constitute evidence of somatic
SQSTM1 mutations, a study of osteoblasts and bone marrow cells
from 23 patients with sporadic PDB found an SQSTM1 mutation in
a single patient, who was heterozygous for the P392L mutation in
both the bone marrow and peripheral blood cells [10].
2.4. Functional impact of mutations
SQSTM1 mutations in the region encoding the ubiquitin-binding
domain result in an SQSTM1/p62 protein that is unable to bind to
ubiquitin, which probably causes aberrant RANK-NF-kB signaling
via failure of the proteasome to degrade SQSTM1/p62 [7]. How-
ever, the protein produced in patients having SQSTM1 mutations
outside the ubiquitin-binding domain has normal affinity for ubiq-
uitin, suggesting that more subtle mechanisms may be at work [7].
The identification of SQSTM1 mutations in PDB established that the
SQSTM1/p62 protein played a key role in osteoclast signaling [11].
Osteoclasts derived from peripheral blood monocytes from patients
with PDB were found in higher numbers, contained more nuclei,
and exhibited greater bone-resorption capacity, compared to cells
from healthy donors [11]. The presence of the P392L mutation had
little effect on the osteoclast phenotype in the PDB patients but
253
was associated with an intermediate phenotype in the healthy con-
trols [11]. An abnormality in the regulation of osteoclast autophagy
has been suggested to explain the impact of SQSTM1 mutations
in PDB [1,7,12]. Mice transfected with the P392L mutation fail to
replicate the full PDB phenotype [13]. Thus, the P392L mutation
may predispose to PDB by sensitizing the osteoclast progenitors
to osteoclastogenic cytokines and/or by enhancing the osteoclas-
togenic potential of the bone microenvironment [14].
2.5. Practical implications
In several European countries, a genetic test is available for
detecting mutations in exons 7 and 8 of the SQSTM1 gene [4]. This
test can be prescribed on a routine basis and performed at refer-
ence laboratories. When a mutation is found in an index patient,
the family members can be screened. In autosomal dominant PDB,
first-degree relatives have a 50% likelihood of being affected. It must
be borne in mind that genetic screening in asymptomatic patients,
known as presymptomatic diagnosis, must be performed by an
experienced multidisciplinary team [4].
In patients with PDB, identifying a mutation does not benefit the
individual patient, as the genotype is not known to correlate with
the phenotype and has no impact on treatment decisions [4]. How-
ever, detection of a mutation in an index patient allows screening of
the family members before the development of complications from
undiagnosed PDB. When a mutation is found in an asymptomatic
relative, radionuclide bone scanning shows the foci of pagetic bone,
thus ensuring the identification of high-risk foci that may have
therapeutic implications. If the bone scan is normal, there is no
indication for preventive treatment in an asymptomatic individual
who is found to carry a mutation.
The test for SQSTM1 mutations should not be used to confirm
a diagnosis of PDB. Furthermore, absence of a detected mutation
in a patient with PDB neither calls the diagnosis into question nor
challenges the genetic nature of PDB, as the disease is genetically
heterogeneous.
3. Fibrous dysplasia and McCune-Albright syndrome
3.1. Clinical and epidemiological background
Fibrous dysplasia is a focal bone disease that affects a single bone
in 60% of cases. The lesions are composed of immature osteoblast
progenitors derived from the mesenchymatous layer, spicules of
immature woven bone and, in some cases, nests of hyaline carti-
lage [15]. In most cases, fibrous dysplasia lesions progress from the
medullary cavity to the neighboring cortical bone and contain a rim
of osteoclasts. Although fibrous dysplasia is often asymptomatic
and discovered fortuitously, some patients present with bone pain,
deformities, most notably in craniofacial forms, or pathological
fractures, which often constitute the first clinical manifestation
[16]. In 80% of patients, the first symptoms occur before 15 years
of age [16]. In addition to fractures, complications of fibrous dys-
plasia include exophthalmos, dental abnormalities, and leontiasis
ossea. Fewer than 5% of patients with fibrous dysplasia also have
precocious puberty and café au lait spots with tortuous contours, a
combination known as McCune-Albright syndrome (MAS). Fibrous
dysplasia is found in 99% of patients with MAS. Endocrine system
abnormalities in MAS may consist of hyperthyroidism, excessive
growth hormone production, hypercorticism, and hyperprolactine-
mia. Renal dysfunction is a feature in about 50% of patients with
MAS. Non-endocrine abnormalities reported in MAS include intra-
muscular myxoma, cholestasis, cardiac involvement, myelofibrosis,
and gastrointestinal disorders [15]. Sarcomatous transformation
has been reported [17]. Although the prevalence of fibrous dys-
254 L. Michou, J.P. Brown / Joint Bone Spine 78 (2011) 252–258
plasia is unknown, the disease is estimated to account for 7% of all
benign bone tumors [17] and to affect about 1 in 30,000 individuals
[16]. MAS is rarer, with less than one case per 10,000 individuals
[17]. Males and females are equally affected.
3.2. Mutations in the GNAS gene
Fibrous dysplasia of bone and MAS are caused by dominant
somatic mutations in the GNAS gene (GNAS complex locus) that occur
at a very early stage of embryonic development, just after formation
of the zygote [15]. The GNAS gene encodes the ␣ subunit of the Gs
protein. The mutation usually leads to substitution of a histidine or
cysteine for the arginine at position 201 of the protein. Less often,
the 201 arginine is replaced by a glycine or leucine [18]. Mutations
on codon 227 are also known to induce activating effects but have
not been identified in patients with MAS [15]. In affected patients,
the cells harboring the mutation coexist with the normal cells in
a mosaic. The cells having the mutation are present in increased
numbers in involved tissues (endocrine glands, skin, or bone tissue,
depending on the manifestations) but may also be found in tissues
that are usually unaffected such as the peripheral leukocytes, liver,
heart, thymus, and gastrointestinal tract [19].
The phenotype may depend on the extent and distribution of the
mutated cells, which in turn depend on the time at which the muta-
tion occurred during embryonic development. A somatic mutation
that occurs later during development may lead to a less severe
phenotype (e.g., monostotic fibrous dysplasia) [15]. Although the
somatic mutations are easier to identify in cells from involved
tissues, currently available sensitive amplification techniques can
detect them in DNA from peripheral blood cells in some patients
[15]. These techniques use peptide nucleic acid primers that block
synthesis from the normal allele, thus selectively amplifying the
mutant allele and considerably improving the detection of muta-
tions, even those present in small numbers of cells [20]. Finally,
an important point is that the GNAS locus is subjected to com-
plex paternal imprinting of both paternal and maternal alleles via
the involvement of various promoter types and alternative splicing
mechanisms [21]. Thus, the Gs␣ protein is expressed biallelically
in all tissues except the proximal renal tubules, thyroid gland,
gonads, and pituitary gland, where the maternal allele is selectively
expressed [15]. On the other hand, a Gs␣ variant is expressed only
from the paternal allele, chiefly within neuroendocrine tissues and
the nervous system. However, in MAS, GNAS mutations may be car-
ried only by the maternal allele in patients with pituitary gland
tumors and increased growth hormone release, whereas no allelic
bias has been reported in other situations [15].
3.3. Functional impact of the mutations
In fibrous dysplasia and MAS, GNAS mutations induce gain-
of-function of the Gs␣ protein. The amino acid substitutions due
to the mutations occur at sites that are critical to the activity of
the enzyme GTPase, which normally deactivates the Gs␣ protein.
The result is prolonged Gs␣ protein activation with overproduc-
tion of cyclic AMP (cAMP) via adenyl cyclase, which stimulates
endocrine gland growth and hormone secretion even in the absence
of stimulating hormone [15]. Activation of Gs␣/cAMP also leads to
increased pigment production by the melanocytes via the induction
of tyrosinase expression and to cardiac hypertrophy via activa-
tion of the MAP kinase pathway [15]. In bone, activation of the
Gs␣/protein kinase A signaling pathway represses the transcrip-
tion factor Runx2 and may cause the abnormalities in osteoblast
differentiation. Gs␣/protein activation increases the secretion of
interleukin-6, which may cause the osteolytic lesions [15].
3.4. Practical implications
The GNAS gene mutations are well documented, and a genetic
test for fibrous dysplasia and MAS should therefore be within reach.
However, no test is available for routine practice, although sev-
eral laboratories in the world carry out genetic tests for research
projects. A negative test in one tissue does not rule out the diag-
nosis, since both fibrous dysplasia and MAS are associated with
somatic mosaicism. In most cases, genetic testing makes very lit-
tle contribution to the diagnosis and no contribution at all to the
management, as no genotype–phenotype correlations have been
reported [17]. Genetic counselors can reassure the patients that
there is no risk of transmission to their offspring, no known associ-
ation with environmental factors, and no selective involvement of a
specific ethnic group [17]. The nearly complete absence of familial
forms suggests that the activating mutations may be viable only in
mosaic form and that their presence in the gametes might lead to
death during embryonic development [15,19]. Finally, as GNAS gene
mutations occur shortly after formation of the zygote, monozygotic
twins may be discordant for fibrous dysplasia or MAS [22].
4. Osteopetrosis
4.1. Clinical and epidemiological background
Osteopetrosis is a heterogeneous group of disorders character-
ized by abnormal bone remodeling with increased bone density
due primarily to defective osteoclastic resorption [23]. In forms of
osteopetrosis with normal or increased osteoclast counts, the bone
resorption defect usually results from failure to produce a ruffled
border [24]. In forms with decreased osteoclast counts, an abnor-
mality in the molecular pathways involved in osteoclastogenesis
has been suggested [25]. The severity spectrum is broad, rang-
ing from no symptoms at all to death in infancy (Table 1). Severe
forms are often inherited on an autosomal recessive basis, whereas
autosomal dominant transmission is the rule in moderate forms
diagnosed in adulthood (Figs. 1 and 2) [26]. Classical manifesta-
tions of malignant osteopetrosis consist of fractures, short stature,
compression of nervous structures, and pancytopenia.
Distinctive clinical patterns of osteopetrosis are character-
ized by extraosseous manifestations such as neurodegenerative
disorders, mental retardation, skin abnormalities, immune dys-
function, and renal tubular acidosis (Table 1). The occurrence in
late childhood or adolescence of fractures and/or osteomyelitis of
the mandible suggests an autosomal dominant form. Although the
overall incidence of osteopetrosis is difficult to evaluate, malignant
autosomal recessive osteopetrosis is believed to affect one child in
250,000 [26]. Malignant osteopetrosis is far more common in Costa
Rica, with 3.4 cases per 100,000 births, as a result of a founder effect
[23]. Autosomal dominant osteopetrosis is estimated to occur in
five of 100,000 births [26].
4.2. Mutations
Mutations in the gene encoding the A3 subunit of the H+ /ATPase
proton pump are the most common causes of osteopetrosis and
explain 50% of the malignant forms [27]. About 10% of malignant
forms are related to mutations in the chloride channel 7 gene
(CLCN7), which encodes a chloride channel found only in osteo-
clasts. The other mutations (Table 1) are considerably less common.
A study of patients with autosomal recessive osteopetrosis and no
osteoclasts in bone biopsies showed mutations in the RANKL gene
(also known as TNFSF11 for tumor necrosis factor (ligand) super-
family, member 11), opening up new pathogenic and therapeutic
hypotheses [25].
 L. Michou, J.P. Brown / Joint Bone Spine 78 (2011) 252–258 255

Table 1
Classification of the main forms of human osteopetrosis.

Type Phenotype Geneb

Autosomal recessive osteopetrosis

1 Malignant neonatal or infantile form, bone sclerosis, fractures, TCIRG1
pancytopenia, infections, hepatosplenomegaly, neurological
abnormalities
2 Intermediate form, low osteoclast numbers, short stature, RANKL
fractures
3 Intermediate form associated with renal tubular acidosis, short CAII
stature, mental retardation
4 Malignant infantile form, fractures, bone marrow CLCN7
involvement; or intermediate form
5 Malignant infantile form, fractures, bone marrow involvement OSTM1
6 Variable severity, often intermediate PLEKHM1
7 Severe form with low osteoclast numbers associated with RANK
hypogammaglobulinemia
Autosomal dominant osteopetrosis
1a Diffuse bone sclerosis predominating at the cranial vault, often LRP5
asymptomatic, pain, hearing impairment, no fractures
2 Sandwich vertebras (Fig. 1), bone-within-bone images (Fig. 2), CLCN7
fractures, dental abscesses
Other forms of osteopetrosis
Autosomal recessive osteopetrosis with infantile neuroaxonal dystrophy Unknown
X-linked recessive osteopetrosis, anhidrotic ectodermal dysplasia, immune deficiency, lymphedema, dental abnormalities, infections NEMO
a
This type is no longer classified as a form of osteopetrosis, because it is due to increased bone formation and not to decreased bone resorption.
b
TCIRG1, T-cell, immune regulator 1, ATPase, H+ transporting, lysosomal V0 subunit A3; RANKL or TNFSF11, tumor necrosis factor (ligand) superfamily, member 11; CAII,
carbonic anhydrase II; CLCN7, chloride channel 7; OSTM1, osteopetrosis-associated transmembrane protein 1; PLEKHM1, pleckstrin homology domain containing, family M
(with RUN domain) member 1; RANK or TNFRSF11A, tumor necrosis factor receptor superfamily, member 11a, NFKB activator; LRP5, low-density lipoprotein receptor-related
protein 5; NEMO or IkBkG, inhibitor of kappa light polypeptide gene enhancer in B-cells, kinase gamma.

Studies of autosomal dominant osteopetrosis Type 1 identified
gain-of-function mutations in the LRP5 gene encoding low-density
lipoprotein receptor-related protein 5. These mutations cause
increased bone formation. Consequently, autosomal dominant
osteopetrosis Type 1 is now classified among the high-bone-mass
disorders [24,28]. In autosomal dominant osteopetrosis related
to CLCN7 mutations, there is considerable phenotypic variability
across patients having the same mutation and across members of
a same family, suggesting a role for as yet unidentified modifying
genes [29].

4.3. Functional impact of the mutations

Most of the genes that are mutated in osteopetrosis encode

proteins involved in regulating the intra- and extracellular pH of
osteoclasts. These proteins play a pivotal role in acidifying the

Fig. 2. Anteroposterior radiograph of the pelvis showing bone-within-bone images

in the iliac wings and fixation material used to treat a bone-deficiency fracture in
Fig. 1. Lateral radiograph of the thoracic spine showing the typical sandwich ver- the proximal left femur, a classic and common complication of autosomal dominant
tebra appearance in a patient with autosomal dominant type II osteopetrosis. type 2 osteopetrosis.
256 L. Michou, J.P. Brown / Joint Bone Spine 78 (2011) 252–258
Table 2
Expanded classification scheme for osteogenesis imperfecta.
Type Phenotype
I Moderately severe, blue sclerae, no dentinogenesis imperfecta,
joint laxity
II Perinatal lethal, dark sclerae, multiples fractures of ribs and
long bones, respiratory distress, limb deformities
III Severe, short stature, multiple fractures, triangular face,
scoliosis, gray sclerae, dentinogenesis imperfecta
IV Moderately severe, normal or gray sclerae, dentinogenesis
imperfecta
V Moderately severe, radial head dislocation, ossification of the
interosseous membrane, hypertrophic calluses, white sclerae,
no dentinogenesis imperfecta
VI Moderately severe to severe, accumulation of osteoid bone on
histological specimens, fish-scale appearance of lamellar bone,
white sclerae, no dentinogenesis imperfecta
VII Moderately severe, short femurs and humeri, coxa vara, white
sclerae, no dentinogenesis imperfecta
VIII Severe to perinatal lethal, short long bones, severely retarded
growth, white sclerae, bulbous metaphyses
IX Severe to perinatal lethal, multiple fractures of long bones,
gray sclerae, no dentinogenesis imperfecta
a
There have been reports of autosomal recessive forms related to mutations in CRTAP, LEPRE1, or PPIB and characterized by a phenotype consistent with types IIB/III in the
Sillence classification scheme.
b
COL1A1, collagen, type I, alpha 1; COL1A2, collagen, type I, alpha 2; CRTAP, cartilage associated protein; LEPRE1, leucine proline-enriched proteoglycan (leprecan) 1; PPIB,
peptidylprolyl isomerase B (cyclophilin B).
resorption lacunae. CAII encodes carbonic anhydrase isoenzyme
type II, the enzyme catalyzing the reaction of CO2 and water to
form carbonic acid (H2 CO3 ), which then releases protons. TCIRG1
encodes an enzyme and CLCN7, a chloride channel located in the
ruffled border. OSTM1 is associated with the function of the chlo-
ride channel encoded by CLCN7 and, finally, PLEHKM1 encodes a
protein involved in vesicle trafficking and acidification [24]. These
advances in knowledge regarding the genes involved in osteopet-
rosis and their effects on the osteoclast phenotype have shed light
on the pivotal role for the osteoclast in this disease, on osteoclast
biology, and on the coupling between bone formation and bone
resorption [30].
4.4. Practical implications
Depending on the country, genetic tests for most of the genes
involved in osteopetrosis may be available for routine use or only
for research projects. However, the causative genetic abnormal-
ity remains unknown in 30% of patients with osteopetrosis [26].
Genetic testing may prove useful for confirming the diagnosis or
differentiating the types of osteopetrosis, with the goal of improv-
ing treatment decisions, outcome prediction, and estimation of
the risk of recurrence [26]. Antenatal or preimplantation genetic
testing may be in order in families known to carry mutations
for malignant osteopetrosis. In families with autosomal reces-
sive malignant osteopetrosis but no known mutation, antenatal
radiographs may be obtained to plan for hematopoietic stem cell
transplantation before 3 months of age should the child be affected
[26]. In autosomal recessive forms, the risk of recurrence is 25% at
each pregnancy and two-thirds of unaffected siblings are expected
to harbor the mutation. Given the low incidence of osteopetro-
sis in the general population, the risk of having an affected child
is low among affected individuals and their healthy siblings [26].
In autosomal dominant forms, affected individuals have a 50%
risk of transmitting the disease to each child. Sporadic cases in
children of parents who have normal clinical and radiographic
findings may indicate gonadal mosaicism with a low but non-
negligible risk of recurrence in subsequent pregnancies [26]. In
X-linked recessive forms, when the mother of the affected child
carries the mutation, 50% of male offspring are expected to be
affected and 50% of female offspring to be carriers. If the mother
Geneb Transmission
COL1A1 Autosomal dominant
COL1A2
COL1A1 Autosomal dominanta
COL1A2
COL1A1 Autosomal dominant
COL1A2
COL1A1 Autosomal dominant
COL1A2
Unknown Autosomal dominant
Unknown Autosomal recessive
CRTAP Autosomal recessive
LEPRE1 Autosomal recessive
PPIB Autosomal recessive
does not carry a mutation, gonadal mosaicism should be considered
[26].
5. Osteogenesis imperfecta
5.1. Clinical and epidemiological background
Osteogenesis imperfecta (OI) is an inherited disorder of bone
formation characterized by osteopenia and bone fragility with frac-
tures [31,32]. Severity and age at onset vary widely, from severe
forms that are fatal in the perinatal period to mild forms diagnosed
much later in life. The manifestations of OI consist of skeletal abnor-
malities such as fractures, deformities, and joint laxity; combined
with extraskeletal features, which are highly distinctive but incon-
sistent, such as blue sclerae, dentinogenesis imperfecta, hearing
loss, and blood vessel fragility [33]. Table 2 lists the currently rec-
ognized forms of OI with their modes of transmission. As shown in
this table, major changes have been made to the original Sillence
classification in four types. The heterogeneity of former Type IV
disease led to the differentiation of three new types based on clin-
ical, histological, and genetic criteria [34]. Then the identification
of two new mutations prompted the individualization of the last
two types, which are transmitted on an autosomal recessive basis
and range from severe to lethal [35,36]. However, this expanded
classification scheme is generating debate because the phenotypes
associated with the newly identified mutations are consistent with
types IIB/III in the Sillence classification, although transmission
occurs on an autosomal recessive basis [31]. The prevalence of OI
may be about 6–7 per 100,000 individuals [31]. The overall inci-
dence of OI has been estimated at about 1/10,000 births [32] but
varies considerably according to the criteria used to define OI.
5.2. Mutations
Mutations in the collagen genes COL1A1 and COL1A2 explain
over 90% of cases of OI. These mutations are usually private, i.e.,
found in a single family. Over 850 different mutations have already
been identified. Most of them are point mutations in one of the
two type I collagen genes (COL1A1 and COL1A2) that cause glycine
substitutions in the collagen polypeptide [32]. However, recurrent
mutations have also been identified, most notably in CpG islands
 L. Michou, J.P. Brown / Joint Bone Spine 78 (2011) 252–258
[35]. The severity of the phenotype varies with the type of mutation,
position of the amino acid substitution on the triple helix of the col-
lagen protein, and type of amino acid substitution [33]. Although no
reliable genotype-phenotype correlations have been found, COL1A1
mutations produce a premature stop codon often associated with
Type I OI [34], whereas mutations involving the ␣2 chain may be
involved in the most severe OI phenotypes [33]. Finally, studies of
patients with severe OI recently identified homozygous mutations
including splice-site mutations, point mutations, and deletions in
CRTAP and LEPRE1; recently, point and deletion mutations produc-
ing a stop codon in PPIB were described [36].
5.3. Functional impact of the mutations
Mutations affecting COL1A1 or COL1A2, which encode the ␣1
and ␣2 chains of Type I collagen, respectively, may produce the OI
phenotype via two mechanisms, chain exclusion and chain nonex-
clusion. Type I collagen is composed of three polypeptide chains
(two ␣1 and one ␣2) intertwined into a triple helix, which is main-
tained by the presence of glycine residues at regular intervals [33].
In moderately severe forms of OI (Type I), the chain encoded by the
mutated allele is unstable and is consequently destroyed without
being incorporated into the collagen molecule. Therefore, all col-
lagen chains are produced by the normal allele, and the collagen
molecule is qualitatively normal but quantitatively deficient [32].
The second mechanism involves incorporation of abnormal colla-
gen chains into the collagen molecule, which is therefore abnormal
[34]. The dominant negative effect of the mutated allele on the nor-
mal allele results in the most severe OI phenotypes (types II, III, and
IV) [35]. Alpha-hydroxylation of the collagen molecule is depen-
dent on three proteins that form a protein complex located in the
endoplasmic reticulum: cartilage-associated protein (CRTAP), lep-
recan, and cyclophilin B [36–38]. Mutations affecting the genes for
these three proteins (Table 2) were recently identified in patients
with recessive forms of OI. These mutations may lead to dysreg-
ulation of posttranslational hydroxylation of the proline residue
at position 986 of the ␣1 chains of type I collagen [35,36]. These
structural abnormalities in the type I collagen molecule lead to
structural bone abnormalities including trabecular bone deminer-
alization and cortical thinning, which explain the abnormal bone
fragility.
5.4. Practical implications
In selected cases, genetic testing can be performed in a reference
laboratory. Testing involves DNA sequencing and/or evaluation of
type I procollagen production [39]. The DNA can be taken from
peripheral blood or saliva. All exons and exon-intron junctions in
the COL1A1 and COL1A2 genes must be sequenced. Other genes can
be studied also, most notably in families with documented autoso-
mal recessive transmission of the disease. In severe forms, genetic
testing can be performed during pregnancy or before implantation.
Evaluation of type I procollagen requires collection of a skin biopsy
followed by fibroblast culturing to look for a decrease or abnormal-
ity in the type I procollagen chains. It is important to note that a
negative test does not rule out OI, particularly for types V to IX but
also for mutations in the type I collagen genes. Currently available
techniques identify only 90% of the mutations affecting COL1A1 and
COL1A2 [39]. In difficult cases, a bone biopsy may be valuable for
confirming the diagnosis of OI.
In autosomal dominant forms of OI, affected individuals have a
50% risk of transmitting the disease to each child. In sporadic forms,
the risk is more difficult to evaluate, as gonadal mosaicism exists
in 6 to 7% of these cases [33]. In addition, the phenotype may be
considerably more severe at the second generation [33].
257
6. Conclusion
In conclusion, recent genetic advances have improved our
understanding of the pathophysiology of PDB and fibrous dysplasia,
most notably by improving our ability to detect somatic mutations.
Identification of the genes involved in the main forms of human
OI has produced new insights into the biology and differentia-
tion of osteoclasts. The discovery of new mutations responsible for
osteopetrosis or OI has radically changed the clinical and radiolog-
ical classification of these diseases.
Conflict of interest statement
The authors declare no conflict of interest.
References
[1] Helfrich MH, Hocking LJ. Genetics and aetiology of pagetic disorders of bone.
Arch Biochem Biophys 2008;473:172–82.
[2] Rousiere M, Michou L, Cornelis F, et al. Paget’s disease of bone. Best Pract Res
Clin Rheumatol 2003;17:1019–41.
[3] Ralston SH, Langston AL, Reid IR. Pathogenesis and management of Paget’s
disease of bone. Lancet 2008;372:155–63.
[4] Michou L, Collet C, Laplanche JL, et al. Genetics of Paget’s disease of bone. Joint
Bone Spine 2006;73:243–8.
[5] Beyens G, Daroszewska A, de Freitas F, et al. Identification of sex-specific asso-
ciations between polymorphisms of the osteoprotegerin gene, TNFRSF11B, and
Paget’s disease of bone. J Bone Miner Res 2007;22:1062–71.
[6] Morissette J, Laurin N, Brown JP. Sequestosome 1: mutation frequencies, hap-
lotypes, and phenotypes in familial Paget’s disease of bone. J Bone Miner Res
2006;21(Suppl. 2):P38–44.
[7] Goode A, Layfield R. Recent advances in understanding the molecular basis of
Paget’s disease of bone. J Clin Pathol 2010;63:199–203.
[8] Collet C, Michou L, Audran M, et al. Paget’s disease of bone in the French popu-
lation: novel SQSTM1 mutations, functional analysis, and genotype-phenotype
correlations. J Bone Miner Res 2007;22:310–7.
[9] Merchant A, Smielewska M, Patel N, et al. Somatic mutations in SQSTM1
detected in affected tissues from patients with sporadic Paget’s disease of bone.
J Bone Miner Res 2009;24:484–94.
[10] Matthews BG, Naot D, Bava U, et al. Absence of somatic SQSTM1 mutations in
Paget’s disease of bone. J Clin Endocrinol Metab 2009;94:691–4.
[11] Chamoux E, Couture J, Bisson M, et al. The p62 P392L mutation linked
to Paget’s disease induces activation of human osteoclasts. Mol Endocrinol
2009;23:1668–80.
[12] Nakamura K, Kimple AJ, Siderovski DP, et al. PB1 domain interaction of
p62/sequestosome 1 and MEKK3 regulates NF-kappaB activation. J Biol Chem
2010;285:2077–89.
[13] Kurihara N, Hiruma Y, Zhou H, et al. Mutation of the sequestosome 1 (p62) gene
increases osteoclastogenesis but does not induce Paget disease. J Clin Invest
2007;117:133–42.
[14] Hiruma Y, Kurihara N, Subler MA, et al. A SQSTM1/p62 mutation linked to
Paget’s disease increases the osteoclastogenic potential of the bone microen-
vironment. Hum Mol Genet 2008;17:3708–19.
[15] Weinstein LS. G(s)alpha mutations in fibrous dysplasia and McCune-Albright
syndrome. J Bone Miner Res 2006;21(Suppl. 2):P120–4.
[16] Chapurlat RD, Orcel P. Fibrous dysplasia of bone and McCune-Albright syn-
drome. Best Pract Res Clin Rheumatol 2008;22:55–69.
[17] Dumitrescu CE, Collins MT. McCune-Albright syndrome. Orphanet J Rare Dis
2008;3:12.
[18] Lumbroso S, Paris F, Sultan C. Activating Gsalpha mutations: analysis of 113
patients with signs of McCune-Albright syndrome–a European Collaborative
Study. J Clin Endocrinol Metab 2004;89:2107–13.
[19] Ringel MD, Schwindinger WF, Levine MA. Clinical implications of genetic
defects in G proteins. The molecular basis of McCune-Albright syndrome and
Albright hereditary osteodystrophy. Medicine (Baltimore) 1996;75:171–84.
[20] Lietman SA, Ding C, Levine MA. A highly sensitive polymerase chain reaction
method detects activating mutations of the GNAS gene in peripheral blood cells
in McCune-Albright syndrome or isolated fibrous dysplasia. J Bone Joint Surg
Am 2005;87:2489–94.
[21] Hayward BE, Moran V, Strain L, et al. Bidirectional imprinting of a single gene:
GNAS1 encodes maternally, paternally, and biallelically derived proteins. Proc
Natl Acad Sci U S A 1998;95:15475–80.
[22] Peleg R, Luba A, Eliakim A, et al. McCune-Albright syndrome in a discordant
monozygotic twin. Isr Med Assoc J 2009;11:343–7.
[23] Balemans W, Van Wesenbeeck L, Van Hul W. A clinical and molecular overview
of the human osteopetroses. Calcif Tissue Int 2005;77:263–4.
[24] Del Fattore A, Cappariello A, Teti A. Genetics, pathogenesis and complications
of osteopetrosis. Bone 2008;42:19–29.
[25] Sobacchi C, Frattini A, Guerrini MM, et al. Osteoclast-poor human osteopetrosis
due to mutations in the gene encoding RANKL. Nat Genet 2007;39:960–2.
[26] Stark Z, Savarirayan R. Osteopetrosis. Orphanet J Rare Dis 2009;4:5.
258 L. Michou, J.P. Brown / Joint Bone Spine 78 (2011) 252–258
[27] Tolar J, Teitelbaum SL, Orchard PJ. Osteopetrosis. N Engl J Med
2004;351:2839–49.
[28] Van Wesenbeeck L, Cleiren E, Gram J, et al. Six novel missense mutations in
the LDL receptor-related protein 5 (LRP5) gene in different conditions with an
increased bone density. Am J Hum Genet 2003;72:763–71.
[29] Chu K, Koller DL, Snyder R, et al. Analysis of variation in expression of auto-
somal dominant osteopetrosis type 2: searching for modifier genes. Bone
2005;37:655–61.
[30] Segovia-Silvestre T, Neutzsky-Wulff AV, Sorensen MG, et al. Advances in osteo-
clast biology resulting from the study of osteopetrotic mutations. Hum Genet
2009;124:561–77.
[31] Van Dijk FS, Pals G, Van Rijn RR, et al. Classification of osteogenesis imperfecta
revisited. Eur J Med Genet 2009;53:1–5.
[32] Glorieux FH. Osteogenesis imperfecta. Best Pract Res Clin Rheumatol
2008;22:85–100.
[33] Baujat G, Lebre AS, Cormier-Daire V, et al. [Ostéogenèse imparfaite, annonce du
diagnostic (classification clinique et génétique)]. Arch Pediatr 2008;15:789–91.
[34] Rauch F, Glorieux FH. Osteogenesis imperfecta. Lancet 2004;363:1377–85.
[35] Basel D, Steiner RD. Osteogenesis imperfecta: recent findings shed new light
on this once well-understood condition. Genet Med 2009;11:375–85.
[36] van Dijk FS, Nesbitt IM, Zwikstra EH, et al. PPIB mutations cause severe osteo-
genesis imperfecta. Am J Hum Genet 2009;85:521–7.
[37] Morello R, Bertin TK, Chen Y, et al. CRTAP is required for prolyl 3-
hydroxylation and mutations cause recessive osteogenesis imperfecta. Cell
2006;127:291–304.
[38] Cabral WA, Chang W, Barnes AM, et al. Prolyl 3-hydroxylase 1 deficiency causes
a recessive metabolic bone disorder resembling lethal/severe osteogenesis
imperfecta. Nat Genet 2007;39:359–65.
[39] Cheung MS, Glorieux FH. Osteogenesis imperfecta: update on presentation and
management. Rev Endocr Metab Disord 2008;9:153–60.