Our results reveal a spatio-temporal window in which Tbx4
are required for patterning the soft tissues (muscles and tendons) of the musculoskeletal system. Tbx4
exclusively regulate muscle and tendon patterning while having no apparent effect on the generation, proliferation or migration of the progenitors of these tissues, strongly suggesting that they regulate a distinct patterning signal(s), which our results indicate are dependent on proper organisation of MCT. Regulation of the Tcf4-expressing connective tissue can account for the independent patterning activity Tbx5 and Tbx4 have on both muscles and tendons since Tcf4 is also expressed in domains where tendon progenitors arise (Kardon et al., 2003
). We propose a model in which Tbx4/Tbx5
expressed in the MCT positively regulate expression of N-Cadherin and β-Catenin that are required for the proper integrity and organisation of this tissue that in turn is critical for correct patterning of the adjacent muscles () and tendons. The loss of Tbx4/Tbx5
, leads to a downregulation of N-Cadherin and β-Catenin, disorganisation of the MCT, resulting in mispatterning of muscles (). Consistent with this model, we show that deletion of β-Catenin results in similar phenotypes.
N-Cadherin has been previously implicated in limb myoblast pathfinding (Brand-Saberi et al., 1996
). In addition, Cadherins and β-Catenin expressed in craniofacial connective tissue have been suggested to play a role in patterning adjacent head muscles (Rinon et al., 2007
). Interestingly, β-Cateninlox/lox; Prx1Cre(98)
mice also exhibit some cranial muscle mispatterning (not shown) since this Cre-deleter is also expressed in the ventral part of the 1st
branchial arch (Logan et al., 2002
). Together these data suggest that Cadherin/β-Catenin activity in connective tissue could be a general mechanism regulating vertebrate muscle patterning. Recent data suggests that like β-Catenin, N-Cadherin does not function solely in cell adhesion, but has many other roles, such as in cell signalling and transcriptional regulation (Halbleib and Nelson, 2006
). A major challenge now will be to precisely determine how the N-Cadherin/β-Catenin complex functions in connective tissue to regulate muscle and tendon morphogenesis.
is expressed in muscle connective tissue but not myoblasts themselves and has also been implicated in muscle patterning and formation. Tcf4
continues to be expressed in Tbx5
mutants suggesting it either acts in a parallel pathway or potentially upstream. Although the Tbx5
-dependent β-Catenin reduction we observe could, in principle, effect Tcf4/Wnt signalling in the MCT, all our data suggests this is not the case. First, following the deletion of β-Catenin protein levels of its associated membranal cofactor N-Cadherin are reduced in a Wnt-independent manner (Cali et al., 2007
). Second, blocking the Wnt pathway using dominant negative Tcf4 (Tcf4-EN) affects myotube differentiation, leading to some muscles failing to form while the patterning of other muscle is affected variably (Kardon et al., 2003
). In contrast, in the Tbx4
-deleted limbs all muscles are uniformly affected and there is no effect on muscle differentiation. Finally, Tcf4
mRNA was shown to be responsive to Wnt signalling (Kardon et al., 2003
) however, we do not detect any changes in Tcf4
transcript levels following the deletion of Tbx5
(). Collectively, these differences suggest that the Tbx4/Tbx5
-dependent β-Catenin loss in the MCT affects muscle patterning in a mechanism that is distinct from the Wnt-dependent Tcf4 pathway.
have equivalent roles in initiating limb outgrowth during a narrow, early time-window at around E9.0 (Hasson et al., 2007
; Naiche and Papaioannou, 2007b
) Our current results demonstrate that at later stages of limb development when the genes are no longer required to initiate limb outgrowth, both genes have a role patterning the limb muscles and tendons. This second pulse of activity lasts for 24-48 hours. Limb muscles are formed from a subpopulation of the hypaxial myoblasts that migrate into the limb buds and it is once they have entered this environment that these cells receive instructive cues that dictate ultimate muscle morphology (Buckingham et al., 2003
). Our data suggests that Tbx4
have been co-opted to pattern limb muscles by regulating a general Cadherin/β-Catenin-dependent muscle patterning “cassette” after myoblast migration has terminated and co-incident with the onset of terminal differentiation. Little is known about the tissue interactions that occur during tendon patterning and the deletion approaches we have taken do not distinguish whether these T-box
genes are acting autonomously or non-autonomously on tendon progenitors.
Together, our results point to MCT organisation and integrity being critical for normal patterning of soft tissues. Accordingly, we suggest that disruption of MCT development, and specifically the Cadherin/β-Catenin complex, play a role in human soft tissue pathologies. In humans, HOS patients can present with soft tissue abnormalities that are not associated with skeletal defects (Newbury-Ecob, ; Newbury-Ecob et al., 1996
; Spranger et al., 1997
) consistent with the observations that, despite the widespread soft tissue defects produced in our mouse models following deletion of Tbx5/4
, the skeleton could be unaffected. We propose that defects in muscle connective tissue integrity should be explored as an explanation for soft tissue abnormalities and the influence of connective tissue considered in developing strategies for musculoskeletal tissue regeneration therapies.