My primary interest is the molecular role of signaling molecules in the control of skeletal development. Understanding
the mechanism of action of molecular signals involved in skeletogenesis may provide a scientific basis for novel
tissue engineering approaches for regeneration of skeletal tissue following damage or disease.
I have maintained an interest in cartilage and bone biology throughout my research career, initially working in
the cartilage matrix field and more recently on factors involved in cartilage and bone development. My research
program is aligned in two major directions encompassing the study of a sub-group of the TGF-? superfamily and a
newly described protein with Wnt-binding activity.
The CDMPs are members of the TGF-? superfamily of signaling molecules and are most closely related to the bone
morphogenetic proteins (BMPs). BMPs induce osteogenesis, whereas in vivo studies on the CDMPs have shown that they
preferentially stimulate chondrogenesis. During limb development CDMP-1 is specifically expressed in the cartilage
condensations and in the interzone region of future joint spaces. Indeed, CDMP-1 is recognized as the first molecular
marker of joint morphogenesis. From the study of three different inherited human chondrodysplasias I demonstrated
that CDMP-1 plays a critical role in the development of the appendicular skeleton. Firstly, disruption of CDMP-1
function results in severe limb abnormalities and joint dysplasia {Thomas et al (1996), Nature Genet. 12, 315-317}.
Secondly, by studying the underlying mechanism of action of the causative mutations I discovered that the mutated
CDMP-1 protein acts in dominant-negative manner by inhibiting the secretion of other BMPs. This, in turn led to
the first in vivo evidence for the existence of heterodimers between different BMP family members {Thomas et al
(1997), Nature Genet. 17, 58-64}. The chondrogenic activity of CDMP-1 and the severe skeletal abnormalities which
occur in its absence have made it the focus of much optimism as a potential therapeutic agent to combat degenerative
joint disease.
2. Frzb, a secretable modulator of Wnt activity.
Frzb (pronounced Frisbee) is a secreted protein initially isolated from purified cartilage extracts and shares
homology to the cysteine rich domain (CRD) of frizzled, a Wnt receptor. Wnt proteins are secreted signaling molecules
having numerous developmental functions, including skeletal development, as well as dysfunction in oncogenesis
It was recently shown that Frzb can bind to and inactivate Wnt activity {Wang et al. (1997) Cell, 88, 757-766},
leading to speculation for a potential therapeutic use of such activity in modifying Wnt induced developmental
and oncogenic events.
Frzb is temporally and spatially expressed during skeletal and craniofacial development. It localizes to the ventral
mesenchyme of the developing limb bud, a pattern complementary to Wnt 7a, the predominant Wnt gene involved in
dorsal/ventral limb patterning. In the face, Frzb localizes to the mesenchyme of the first and second branchial
arches and to the ectomesenchyme of developing tooth buds{Hoang et al (1998), Dev. Dynam. 212, 364-372}. My current
interest is to determine the role of Frzb during skeletogenesis using a gene knockout approach. However, a classical
gene knockout may not provide information on Frzb function in the limbs and face. This is because Frzb expression
is not restricted to the skeleton and is also present early in development, during gastrulation and in early heart
and brain formation. Consequently, a null Frzb mutation may result in early embryonic lethality, precluding analysis
of its function at later stages of development. Therefore, I am conducting a conditional gene knockout experiment,
using the Cre-LoxP recombination system.
