Cell News 3/2013
19
actions between cartilage fibrils and the extrafibrillar matrix.
Furthermore, these studies indicated that a correct macromole-
cular organisation of the cartilage matrix is essential for chon-
drocyte function. Targeted inactivation of the mouse matrilin-3
gene in two independent approaches led to contradictory results.
In one set of studies, matrilin-3 was found to be dispensable for
skeletal growth and development (Ko et al., 2004; Nicolae et
al., 2007). In the second study matrilin-3 null mice were descri-
bed as being more predisposed to develop severe OA than wild
type littermates (van der Weyden et al., 2006). Based on these
observations, it was assumed that the reason for the mild phe-
notype in of matrilin-3 deficient animals might be a functional
redundancy between structurally related perfibrillar proteins (Ko
et al., 2004). This hypothesis was further corroborated in mat-
rilin-1/-3 double deficient animals. Here, an increased collagen
fibril volume density was observed, indicating that matrilins in-
deed influence matrix structure and assembly in vivo (Nicolae et
al., 2007). COMP deficient mice have a completely normal skele-
tal development and the expression of all other members of the
thrombospondin family was unchanged (Svensson et al., 2002).
Therefore it was speculated that, in this case, the lack of COMP
might be compensated for by other extracellular matrix proteins
that are structurally unrelated but share similar functions. As
said, human mutations in either collagen IX or matrilin-3 result
in chondrodysplasia types that are radiologically indistinguisha-
ble from COMP-associated diseases. This led to the assumpti-
on that collagen IX and matrilin-3 are potential candidates to
compensate for the lack of COMP. Therefore, mouse lines double
deficient for COMP and collagen IX or for COMP and matrilin-3
were generated and indeed a more pronounced phenotype than
in single deficient mice was observed (Blumbach et al., 2008,
2009; Groma et al., 2012). In the future, it will be interesting
to study the skeletal development in other combinatorial knock
out lines. In addition, the importance of COMP is underlined by
the observation that in mouse the expression of COMP variants
carrying mutations cause phenotypic changes that mimic human
chondrodysplasia (Pirog-Garcia et al., 2007; Posey et al., 2009;
Schmitz et al., 2008). Due to disrupted growth plate organization,
including impaired chondrocyte alignment in an altered collagen
matrix, the mice develop defects ranging from slight growth re-
tardation to short limb dwarfism. In case of one mutation, dege-
nerative joint disease was observed in adulthood. The fact that
the expression of a mutated protein causes a phenotype could
be explained by an intracellular retention leading to an unfolded
protein response and ER stress also in vivo (Nundlall et al 2010;
Suleman et al., 2012). Interestingly, a dominant negative effect
on matrix assembly was also described. Another mouse model
was generated by knocking-in a chondrodysplasia-causing mat-
rilin-3 mutation (Leighton et al., 2007). Homozygous matrilin-3
mutant mice developed short-limbed dwarfism, the mutated
protein was retained in the ER and a reduced proliferation and
a spatially dysregulated apoptosis of chondrocytes was obser-
ved. Recently, a proteomic analysis of different chondrodysplasia
mouse models was performed and revealed common pathogenic
pathways but also distinct disease signatures specific for each
mutation (Bell et al., 2013). However, so far it remains unclear
why, in contrast to the autosomal dominant human disease, both
copies of the mutant allele were required for the mice to develop
a detectable chondrodysplasia.
In summary, the mouse models so far available point towards a
functional redundancy among perifibrillar proteins and to the
hypothesis that structurally unrelated proteins can compensa-
te for each other. Thus, the generation of compound knockouts
together with the analysis of transgenic lines will be necessary
to unravel the in vivo function of certain perifibrillar proteins. In
combination with an in depth biochemical characterization of
wildtype and mutated proteins these analysis will give insight
into mechanisms involved in both common and rare cartilage
diseases.
The role of the extracellular matrix in osteoarthritis
In contrast to the rare forms of chondrodysplasia, OA is a com-
mon degenerative disease. It affects all structures of a joint but
it remains unclear how the disease is initiated and what factors
trigger the disease process. Inflammation might also be an early
event that could impair the ability of the joint tissue to handle
mechanical loads that do not only have a central role in the
initiation and progression of OA but also in cartilage breakdown
(Heinegard & Saxne, 2011). Regardless of the nature of factors
that initiate the disease, the pathological progression of cartilage
degeneration follows a typical pattern. The extracellular matrix
of articular cartilage is the primary target of osteoarthritic carti-
lage degeneration and, as mentioned earlier, both the absence of
or mutations in cartilage matrix proteins can predispose for OA.
The progressive destruction of cartilage involves the degradation
of ECM components by proteases. One of the earliest changes
at the molecular level is the loss of proteoglycans mainly by the
so-called aggrecanases ADAMTS-4 and -5 (Stanton et al., 2005;
Glasson et al., 2005), followed by the degradation of matrix pro-
teins such as COMP, fibronectin, CILP-1 and many others. Finally,
collagens are cleaved by matrix metalloproteases (Heinegard &
Saxne, 2011).
On the one hand, the release of protein fragments provides op-
portunities to monitor the disease process and its progression.
COMP is already widely used as a diagnostic and prognostic
marker for knee OA (Mobasheri, 2012) and recently, increased
levels of matrilin-3 were found in serum and synovial fluid of OA
patients (Vincourt et al., 2008, 2012). On the other hand, it has
been demonstrated that matrix molecules, like e.g. collagen II
and matrilin-3, as well as fragments thereof, have the potential
to induce pro-inflammatory cytokines and matrix metalloprote-
ases and to activate the expression of osteoarthritis-associated
genes in a feed forward mechanism (Klatt et al., 2009a, b). This
means that also the chondrocyte itself plays a pivotal role, as it
is mainly responsible for the anabolic/catabolic balance requi-
red for matrix maintenance and tissue function. The similarities
in the pathologic progression of the disease indicate that there
may be a common sequence of molecular events underlying pro-
gression.
