The Genetics of Pagets Disease of the Bone
Robin J. Leach,
Frederick R. Singer and
G. David Roodman
Departments of Cellular and Structural Biology (R.J.L.), Pediatrics
(R.J.L.), and Medicine (G.D.R.), The University of Texas Health Science
Center at San Antonio, San Antonio, Texas 78229-3900; Department of
Medicine (G.D.R.), The Veterans Administration Hospital, San Antonio,
Texas 78284; and John Wayne Cancer Institute at St. Johns Hospital
and Health Center (F.R.S.), Santa Monica, California 90404
Address correspondence and requests for reprints to: G. D. Roodman, M.D., Ph.D., Department of Medicine/Hematology (7880), University of Texas Health Science Center at San Antonio, 7703 Floyd Curl Drive, San Antonio, Texas 78229-3900. E-mail:
roodman{at}uthscsa.edu
 |
Introduction
|
---|
In 1877, Sir James Paget first described the
localized skeletal disease that is now known as Pagets disease of the
bone (PDB) (1). The disease is primarily a disorder of the
osteoclast with marked increase in bone resorption, followed by
abundant new bone formation. In this report, we present a brief
overview of the genetics of PDB; a complete review was recently
published by Singer and Leach (2).
PDB has a familial tendency (3, 4, 5), suggesting that there
is a genetic predisposition. In a study of the frequency and
characteristics of the familial aggregation of PDB in Spain,
Morales-Piga et al. (6) found that 40% of
their index cases had at least one first-degree relative affected with
PDB. In the pedigrees they reported, PDB seemed to be transmitted
through either parent, suggesting an autosomal dominant mode of
inheritance.
Siris (7) conducted an epidemiological study of PDB in the
United States, using questionnaires completed by 864 patients with
physician-diagnosed PDB, and compared these results to 500 controls of
similar age. A history of PDB was noted in a first-degree relative in
12% of the patients, compared with only 2% of controls. The risk of
first-degree relatives of a pagetic patient developing PDB was seven
times greater than for an individual without an affected relative. The
cumulative risk for developing PDB up to age 90 for a first-degree
relative of a patient was 9% compared with 2% in individuals with
unaffected relatives. This is good evidence that a gene, or genes,
plays a role in the acquisition of the disease.
 |
Histocompatibility leukocyte antigen (HLA) association
studies
|
---|
The role of the major HLA loci on chromosome 6 in many human
diseases makes it a logical starting point for studying the genetics of
PDB. Several studies since 1975 have examined the possibility that
there is an association between HLA and PDB (8, 9, 10, 11, 12).
Although HLA typing of class I antigens revealed no significant
associations, significant associations were observed between class II
antigens and PDB in two studies (11, 12). An increased
incidence of HLA-DQW1 and HLA-DR2 was found in a preliminary report of
53 patients in Los Angeles (11). A second study of 25
Ashkenazi Jews in Israel revealed an increased incidence of HLA-DR2 in
this population (12).
 |
Family studies
|
---|
Although multiple families have been reported with PDB, the
average number of affected individuals in these families is three.
Thus, these families are not sufficiently large for performing
classical linkage analysis. However, there have been a few larger
kindreds that have been used for linkage analysis. As reviewed below,
analysis has identified two loci, one on chromosome 6 and one on
chromosome 18, as susceptibility loci for PDB. Both loci seem to play a
role in PDB, and there is evidence from other studies that there are
other loci yet to be identified.
 |
HLA linkage studies
|
---|
The first linkage studies used the HLA loci because of their
highly polymorphic nature. In a study conducted in New York City
(13), three families with 29 informative children (all
over the age of 45) were used. These kindreds were typed at the HLA-A,
-B, and -C loci. Using haplotype data, the investigators obtained a
maximum LOD (the logarithm of the odds ratio of a particular locus
being linked or not linked to the disease locus) score of 2.44 with
11% recombination. Because a LOD score of 3 is considered linkage,
these data were only deemed "suggestive." In a second study
performed in New Zealand (14), two additional families
were identified that seemed to segregate the disease in an autosomal
dominant fashion. Each family had four affected members. These families
were genotyped for the HLA loci. Although significant linkage was not
obtained with this study, the combined linkage analysis from both
studies resulted in a maximum LOD score of 3.69 with 10%
recombination. The combined data were considered sufficient for
establishing a predisposition locus for PDB on chromosome 6 near the
HLA loci. Two other family studies have been published that explored
the linkage of the HLA loci with PDB (15, 16). Both
studies reported no linkage between the HLA loci and a PDB
predisposition gene. However, neither family was large enough to
exclude linkage, and the relevance of these observations is
unclear.
 |
Chromosome 18 linkage studies
|
---|
Familial expansile osteolysis (FEO) is a rare bone dysplasia,
which is transmitted as an autosomal dominant trait (17)
in a large kindred in Northern Ireland and a kindred in the United
States. FEO shares some features with PDB. The bone lesions appear
similar to early pagetic osteolytic lesions, although they occur at a
much earlier age; also, these lesions never become sclerotic.
Interestingly, osteoclasts from these patients contain
paramyxoviral-like nuclear inclusions, which have also been reported in
PDB (18). These results supported the hypothesis that FEO
may be an allelic variant of PDB.
Hughes et al. (19) have used genetic linkage
analysis to localize FEO to chromosome 18q. The disease shows a tight
linkage with several polymorphic markers on chromosome 18q, with a
maximum LOD score of 11.53 and no recombination. To further test the
hypothesis that FEO is an allelic variant of PDB, linkage analysis has
been performed between chromosome 18 markers and PDB disease
kindreds.
The first published linkage study with chromosome 18 markers was
performed by Cody et al. (20), using one large
pedigree. The pedigree is presented in Fig. 1
. In this study, 16 DNA samples were
evaluated with a total of 12 markers from 18q. The maximum
LOD score of 3.44 was obtained with genetic marker D18S42 with no
recombination. This marker lies in the same region of chromosome 18q
that is linked to FEO. Thus, this study supports the hypothesis that
FEO and PDB result from mutations in the same locus or tightly linked
loci.

View larger version (27K):
[in this window]
[in a new window]
|
Figure 1. Pedigree of kindred with PDB that is linked
to the loci on chromosome 18q. Black symbols represent
affected individuals. An asterisk indicates those
individuals for whom the phenotype is unknown. The disease haplotype is
indicated by a rectangle. [Adapted from Cody et
al. (20 ).]
|
|
The following year, Haslam et al. (21) reported
a study of eight families with familial PDB. Using seven polymorphic
markers, they obtained a maximum LOD score of 2.97 with marker D18S42
with 5% recombination. Statistically, it seemed that the families were
genetically heterogeneous. Only five of the eight families had positive
LOD scores in this region, whereas the remainder seemed to be linked to
another locus. There is also a preliminary report of a French pedigree
that is clearly linked to 18q (22). This pedigree was
analyzed with 12 chromosome 18 markers and gave a maximum LOD score of
3.10 with marker D18S68 at 1% recombination.
There have been several recent reports of families with PDB that do not
seem to be linked to chromosome 18 (21, 23, 24). The most
recent family was sufficiently large to exclude linkage across the
entire interval (23). (A LOD score of -2 is needed to
exclude linkage.) Unfortunately, the majority of the nonchromosome
18-linked families have not been evaluated with markers from chromosome
6. Until positive LOD scores are obtained with these "unlinked"
families, it is difficult to estimate the number of predisposition loci
for PDB.
 |
Role of receptor activator of nuclear factor B
(RANK) in FEO
|
---|
RANK ligand (RANKL) is a newly described member of the tumor
necrosis factor (TNF) family that has been identified as a critical
osteoclastogenic factor (25, 26, 27). RANKL is expressed on
marrow stromal cells and osteoblasts and seems to mediate the effects
of most osteoclastogenic factors. In vivo studies have
demonstrated that mice lacking RANKL develop severe osteopetrosis
(28). Factors such as
1,25-(OH)2D3, interleukin
1, interleukin 11, and prostaglandin E2 seem to
induce osteoclast formation indirectly by up-regulating RANKL
expression on marrow stromal cells (29). These data
suggest that RANKL may be the common mediator for the effects of most
osteotropic factors on osteoclast formation.
The receptor for RANKL, RANK, is a member of the TNF receptor family
and interacts with TNF receptor-associated factor 2 and translocates it
to the nucleus to induce nuclear factor
B signaling. RANK is
expressed on osteoclast precursors and osteoclasts, and overexpression
of RANK can induce nuclear factor kB signaling in the absence of RANKL
(30). Furthermore, deletion of RANK by homologous
recombination results in osteopetrosis (31). These data
confirm the critical role of RANK/RANKL in osteoclastogenesis.
Earlier this year, Hughes et al. (32)
identified the FEO gene on chromosome 18. Using mutational analysis,
they identified gene mutations in three different FEO families.
Interestingly, all three families carried the identical mutation. The
mutation was an 18-bp in-frame insertion in the signal peptide
sequence of the RANK gene (also known as TNFRSF11A). The 18-bp
insertion was the result of a tandem duplication of bases 84 through
101 in exon 1. The mutation resulted in stabilization of the RANK
protein, which in turn increased RANK signaling (32). As
noted above, this increased RANK signaling could result in increased
osteoclast formation, although this has not been proven.
 |
Role of RANK in PDB
|
---|
With the identification of the mutated gene in FEO on chromosome
18, it is now possible to test the hypothesis that FEO and PDB are the
result of mutations in the same gene. Hughes et al.
(32) screened 90 sporadic PDB patients and found none with
this 18-bp insertion mutation. They also performed mutation screening
in members of four PDB families who had evidence of linkage to 18q. All
four of these families had been used in the linkage study by Haslam
et al. (21). In one of these families, a
slightly larger duplication involving bases 75 through 101 was observed
in exon 1 (32). The PDB disease family with the mutation
in RANK was of Japanese ancestry. This mutation segregated with the
disease in this family. No other mutations were observed in the other
three PDB families.
In the PDB family with the RANK mutation, most of the affected
individuals presented in their teens and early twenties with bone pain
or deformity. In addition, all the patients were described as having
dental problems and several had hearing impairment. This clinical
description begs the question whether this family has PDB
vs. a mild version of FEO. It is uncommon in PDB to have
tooth loss, but this seems to be almost universal for FEO
(33). In addition, hearing loss in PDB usually occurs
later in life and is associated with the thickening of the skull,
whereas hearing loss associated with FEO occurs early
(33). Unfortunately, clinical information was not provided
in the publication by Hughes et al. (32); thus,
it is difficult to conclude that RANK mutations are associated with
familial PDB, although it does not seem to play a major role in
sporadic PDB.
To date, there have been two large families that demonstrated
significant linkage to chromosome 18 (20, 22). It will be
important to thoroughly evaluate the RANK gene in these families before
making a conclusion concerning the role of RANK in familial PDB. The
fact that the other families first described by Haslam et
al. (21) did not show mutations in the RANK gene may
be explained two different ways. First, the families may actually be
linked to another locus. Clearly, there is more than one PDB
predisposition gene, and these families are relatively small. Thus,
linkage using these families can only identify a large region that
carries the predisposition gene. It is possible that these families are
truly linked to chromosome 18 and that the responsible gene is near the
RANK locus, or that they are linked to another chromosome.
Alternatively, the "mutation" in the RANK gene could be a promoter
"mutation" that slightly increases the gene expression. Such a
"mutation" may be only a polymorphism, and it may be difficult to
demonstrate functional significance. Based on the data for the
insertion mutation and the role of RANK in osteoclast differentiation,
one would expect that any alterations in the RANK gene causing PDB
would demonstrate increased RANK signaling.
 |
PDB and osteosarcoma
|
---|
In the majority of patients, PDB is asymptomatic
(34). The 510% of patients with symptoms have bone pain
with a wide range of complications, including increased fractures,
deafness, and neurological findings (35). The most
devastating complication of PDB is malignant transformation of the
bone. Although these transformations are rare, occurring in less than
1% of patients, they contribute significantly to the morbidity and
mortality of the disease. There are various forms of malignant
transformation associated with PDB, but the most frequent is
osteosarcoma. Osteosarcoma, like most cancers, is believed to result
from a series of genetic alterations that transform the osteoblast to a
malignant state. There is interesting evidence to imply that there is a
molecular connection between osteosarcoma and PDB
(36).
Loss of heterozygosity (LOH) mapping is a method used to identify tumor
suppressor loci in cancer (37). Yamaguchi et
al. (38) have used LOH mapping to examine
osteosarcomas. They found high-frequency LOH on 3q, 13q, 17p, and 18q,
suggesting that these chromosome arms contain tumor suppressor genes
important in the development of osteosarcoma. Nellisery et
al. (36) have further refined the location of this
tumor suppressor gene between genetic markers D18S60 and D18S42 (see
Fig. 2
). This is the same region that
codes the FEO and the familial PDB genes. This has led us to the
hypothesis that there is a link between the predisposition gene for PDB
and the tumor suppressor gene for osteosarcoma on chromosome 18q.
Interestingly, seven of seven osteosarcomas from PDB patients showed
LOH in this region (36). However, it is unclear, at
present, whether the PDB predisposition gene and the osteosarcoma tumor
suppressor gene are one and the same gene or two tightly linked genes
on chromosome 18q.

View larger version (17K):
[in this window]
[in a new window]
|
Figure 2. Map of the minimal region of LOH on
chromosome 18q in osteosarcoma tumors. The polymorphic loci and their
estimated physical distances are shown across the top.
Note that D18S51, D18S55, and D18S68 were not ordered in the contig.
Below the map is the information gained from the LOH studies of each
tumor. , No LOH; , LOH. Uninformative markers have a
dash. The arrows depict the region and
the direction of LOH for each tumor. The minimal region of LOH in the
sporadic osteosarcomas is denoted by the large shaded
rectangle. [Adapted from Nellissery et al.
(36 ).]
|
|
Molecular analysis by Hughes et al. (32)
localized the RANK gene between markers D18S64 and D18S51, which was
the same interval known to contain the FEO locus. Further mapping in
our laboratory has demonstrated that the RANK gene lies between markers
D18S60 and D18S42 (R. J. Leach, unpublished data), which is the
critical region for the putative osteosarcoma tumor suppressor gene
(see Fig. 2
). These data suggest that the role of RANK in osteosarcoma
needs to be thoroughly analyzed.
 |
PDB and paramyxovirus
|
---|
In addition to a genetic predisposition for PDB, many studies have
suggested a potential viral etiology for PDB as well (39),
because pagetic osteoclasts contain paramyxoviral-like nuclear
inclusions. Although the presence of a virus in pagetic osteoclasts is
still controversial, the majority of studies have supported either
measles virus or canine distemper virus as the paramyxovirus present in
pagetic osteoclasts and their precursors. It is possible that the
combination of a genetic predisposition to PDB and chronic infection of
osteoclast precursors by paramyxovirus may be required to express the
pagetic phenotype in cells of the osteoclast lineage. However, until a
virus is isolated or a paramyxoviral gene can be shown to induce PDB in
an animal model, the role of the virus in the pathogenesis of PDB is
still unclear.
 |
Conclusion
|
---|
PDB disease clearly has a hereditary component. Family studies
demonstrate that there is more than one predisposition gene for PDB.
The mapping of osteosarcoma tumor suppressor and a PDB predisposition
locus on 18q makes this region of the genome of particular interest,
and a region of future study. Until more extensive linkage analyses are
performed on larger kindreds with classic PDB, the other susceptibility
loci for PDB will remain elusive. The identification of PDB
predisposition genes will increase our understanding of the
pathogenesis of this disorder and could lead to alternate treatment
strategies. In addition, the identification of these genes will be
useful for understanding bone biology and the molecular basis of
the alterations that give rise to PDB.
Received June 19, 2000.
Revised September 13, 2000.
 |
References
|
---|
-
Paget J. 1877 On a form of chronic
inflammation of bones (osteitis deformans). Med Chir Trans. 60:3764.
-
Singer FR, Leach RJ. 2000 Genetics of Pagets
disease of bone. In: Econs MJ, ed. The genetics of osteoporosis and
metabolic bone disease. Totowa, NJ: Humana Press; 309318.
-
Jacobs TP, Michelsen J, Polay JS, DAdamo AC, Canfield
RE. 1979 Giant cell tumors in Paget disease of bone: familial and
geographic clustering. Cancer. 44:742747.[Medline]
-
Sofaer JA, Holloway SM, Emery A. 1983 A family
study of Paget disease of bone. J Epidemiol Community Health. 37:226231.[Abstract]
-
Mirra JM. 1986 CPC case: Paget disease and giant
cell tumor. N Engl J Med. 314:105113.[Medline]
-
Morales-Piga AA, Rey-Rey JS, Corres-Gonzales J,
Garcia-Sagredo JM, Lopez-Abente G. 1995 Frequency and
characteristics of familial aggregation in Paget disease of bone. J Bone Miner Res. 10:663670.[Medline]
-
Siris ES. 1994 Epidemiological aspects of Paget
disease: family history and relationship to other medical conditions. Semin Arthritis Rheum. 23:222225.[Medline]
-
Roux H, Mercier P, Maestracci D, Eisinger J, Recordier
AM. 1975 HL-A et maladie de Paget. Rev Rhum Mal Osteoartic. 42:661662.[Medline]
-
Cullen P, Russel GG, Walton RJ, Whiteley J. 1976 Frequencies of HLA-A and HLA-B histocompatibility antigens in Pagets
disease of bone. Tissue Antigens. 7:5556.[Medline]
-
Gordon MT, Cartwright EJ, Mercer S, Anderson DC, Sharpe
PT. 1994 HLA polymorphisms in Pagets disease of bone. Semin
Arthritis Rheum. 23:229.[Medline]
-
Singer RF, Mills BG, Park MS, Takemura S, Terasaki
PI. 1985 Increased HLA-DQWI antigen pattern in Pagets disease of
bone. Clin Res. 33:574A.
-
Foldes J, Shamir S, Brautbor C, Schermann L, Menczel
J. 1991 HLA-D antigens and Pagets disease of bone. Clin Orthop. 266:301303.[Medline]
-
Fotino M, Haymovits A, Falk CT. 1977 Evidence for
linkage between HLA and Pagets disease. Transplant Proc. 9:18671868.[Medline]
-
Tilyard MW, Gardner RJM, Milligan L, Cleary TA, Stewart
RDH. 1982 A probable linkage between familial Pagets disease and
HLA loci. Aust NZ J Med. 12:498500.[Medline]
-
Moore SB, Hoffman DL. 1988 Absence of HLA linkage
in a family with osteitis deformans (Pagets disease of bone). Tissue
Antigens. 31:6970.[Medline]
-
Kim GS, Kim SH, Cho JK, et al. 1997 Paget bone
disease involving young adults in 3 generations of a Korean family. Medicine. 76:157169.[CrossRef][Medline]
-
Wallace RG, Barr RJ, Osterberg PH, Mollan RA. 1989 Familial expansile osteolysis. Clin Orthop. 248:265277.[Medline]
-
Dickson GR, Shirodira PV, Kanis JA, Beneton MN, Carr KE,
Mollan RAB. 1991 Familial expansile osteolysis: a morphological,
histomorphometric and serological study. Bone. 12:331338.[Medline]
-
Hughes AE, Shearman AM, Weber JL, et al. 1994 Genetic linkage of familial expansile osteolysis to chromosome 18q. Hum
Mol Genet. 3:359361.[Abstract]
-
Cody JD, Singer FR, Roodman GD, et al. 1997 Genetic
linkage of Paget disease of the bone to chromosome 18q. Am J Hum
Genet. 61:11171112.[CrossRef][Medline]
-
Haslam SI, Van Hul W, Morales-Piga A, et al. 1998 Pagets disease of bone: evidence for a susceptibility locus on
chromosome 18q and for genetic heterogeneity. J Bone Miner Res. 13:911917.[Medline]
-
Lucotte G. 1999 Genetic linkage of Pagets disease
to chromosome 18q in a French pedigree (Abstract). Bone. 24:31S.
-
Nance MA, Nuttall FQ, Econs MJ, et al. 2000 Heterogeneity of Paget disease of the bone. Am J Med Genet. 92:303307.[CrossRef][Medline]
-
Hocking L, Slee F, Haslam SI, et al. 2000 Familial
Pagets disease of bone: Pattern of inheritance and frequency of
linkage to chromosome 18q. Bone. 26:577580.[CrossRef][Medline]
-
Lacey DL, Timms E, Tan HL, et al. 1998 Osteoprotegerin ligand is a cytokine that regulates osteoclast
differentiation and activation. Cell. 93:165176.[Medline]
-
Yasuda H, Shima N, Nakagawa N, et al. 1998 Osteoclast differentiation factor is a ligand for
osteoprotegerin/osteoclastogenesis-inhibitory factor and is identical
to TRANCE/RANKL. Proc Natl Acad Sci USA. 95:35973602.[Abstract/Free Full Text]
-
Hofbauer LC, Heufelder AE. 1998 Osteoprotegerin and
its cognate ligand: a new paradigm of osteoclastogenesis. Eur J
Endocrinol. 139:152154.[Medline]
-
Kong YY, Yoshida H, Sarosi I, et al. 1999 OPGL is a
key regulator of osteoclastogenesis, lymphocyte development and
lymph-node organogenesis. Nature. 397:315323.[CrossRef][Medline]
-
Yasuda H, Shima N, Nakagawa N, et al. 1998 Identity
of osteoclastogenesis inhibitory factor (OCIF) and osteoprotegerin
(OPG): a mechanism by which OPG/OCIF inhibits osteoclastogenesis in
vitro. Endocrinology. 139:13291337.[Abstract/Free Full Text]
-
Anderson DM, Maraskovsky E, Billingsley WL, et al. 1997 A homologue of the TNF receptor and its ligand enhance T-cell
growth and dendritic-cell function. Nature. 390:175179.[CrossRef][Medline]
-
Dougall WC, Glaccum M, Charrier K, et al. 1999 RANK
is essential for osteoclast and lymph node development. Genes Dev. 13:24122424.[Abstract/Free Full Text]
-
Hughes AE, Ralston ST, Marken J, et al. 2000 Mutations in TNFRSF11A affecting the signal peptide of RANK, cause
familial expansile osteolysis. Nat Genet. 24:4549.[CrossRef][Medline]
-
Hughes AE, Barr RJ. 1996 Familial expansile
osteolysis: a genetic model of Pagets disease. In: Sharpe PT, ed. The
molecular biology of Pagets disease. Austin, TX: R.G. Landes Company;
179200.
-
Dalinka MK, Aronchick JM, Haddack Jr JG. 1983 Pagets disease. Orthop Clin North Am. 14:319.[Medline]
-
Cawley MI. 1983 Complication of Pagets disease of
bone. Gerontology. 29:276287.[Medline]
-
Nellissery MJ, Padalecki SS, Brkanac Z, et al. 1998 Evidence for a novel osteosarcoma tumor-suppressor gene in the
chromosome 18 region genetically linked with Paget disease of bone. Am J Hum Genet. 63:817824.[CrossRef][Medline]
-
Hansen MF, Cavenee WK. 1988 Tumor suppressors:
recessive mutations that lead to cancer. Cell. 53:173174.[Medline]
-
Yamaguchi T, Toguchida J, Yamamuro T, et al. 1992 Allelotype analysis in osteosarcomas: frequent allele loss on 3q, 13q,
17p, and 18q. Cancer Res. 52:24192423.[Abstract]
-
Singer FR, Roodman GD. 1996 Pagets disease. In:
Bilezikian J, Raisz L, Rodan G, eds. Principles of bone biology. San
Diego: Academic Press; 969977.