Al crest cells migrate from distant or closer ipsi- and contralateral sites and contribute to the shoulder girdle randomly. Even when excluding the participation of unlabelled neural crest we could not find GFP+ neural crest cells in the muscle attachment sites or any part of the shoulder girdle cartilage of 1.5? month-old juveniles (Fig. 3 e , i, j). In younger animals (2? weeks) GFP+ neural crest cells were seen as chains of migrating cells at the base of the forelimb bud (Fig. 3 h), which later occurred here only within the roots of spinal nerves. In both types of experiments we also did not find neural crest cells in the otic capsules or occipital bones of the skull (Fig. 4a), which are of mesodermal origin [7,17]. Furthermore, all GFP+ cells close to the shoulder girdle were well co-localized with immunostaining for a glial marker, myelin basic protein (Fig. 4 b?e). Since we found no difference between the two types of experiments in the labelling of the neural crest derivatives in the neck and shoulder girdle region of the trunk, we further used double sided, but shorter neural fold fragment transplantations. To examine the possibility that neural crest cells migrate into the shoulder girdle at later stages of development, we also Docosahexaenoyl ethanolamide web examined 2? year old adults (ossified bone) that had received double sided GFP+ neural fold transplants. We found that GFP+ signals were still restricted to nerve fibers of the brachial Madrasin cost plexus and to neuronal nets in the muscles, but were not present 1379592 in the shoulder girdle itself or in muscle attachment sites (Fig. 4 f ). Altogether, these results show that neural crest does not contribute to the endochondral shoulder girdle in the axolotl.DiscussionThe complete absence of neural crest cells in the endochondral shoulder girdle of the axolotl contrasts with the apparent contributions of the neural crest to the endochondral shoulder girdle and muscle attachment sites in the mouse [9,18], where neural crest cells were interpreted to be descendants of the neural crest-derived dermal portion of the ancestral shoulder girdle [9]. However, we have found that the shoulder girdle in axolotl did not retain neural crest derivatives, neither as neural crest-derived cell populations within the endochondral elements nor as connective tissue at muscle attachment sites. This may reflect a more common vertebrate trait where muscle attachment points to the endochondral shoulder girdle have no neural crest, but only a mesodermal contribution. It further suggests that neural crest contribution to endochondral elements of the shoulder girdle and the skull and corresponding attaching muscles is not necessary for normal connections between those parts of the skeleton. Observations in several other species have further implied that the neural crest does not make a significant contribution to muscle attachments of the endochondral shoulder girdle in those species either. For example, in quail/chick chimeras, neural crest erived cells were found in the dermal clavicle, but very few to none have been observed in the region near the endochondral scapula [4,19] (I McGonnell and R Huang, pers. comm., own unpublished observations). In turtles, neural crest marker gene expressionResultsTo determine whether neural crest cells contribute to the shoulder girdle in the axolotl, we first grafted the left neural fold (posterior cranial to anterior trunk region) including neural crest cells from a GFP+ donor to a white (d/d) host (Fig. 2a). O.Al crest cells migrate from distant or closer ipsi- and contralateral sites and contribute to the shoulder girdle randomly. Even when excluding the participation of unlabelled neural crest we could not find GFP+ neural crest cells in the muscle attachment sites or any part of the shoulder girdle cartilage of 1.5? month-old juveniles (Fig. 3 e , i, j). In younger animals (2? weeks) GFP+ neural crest cells were seen as chains of migrating cells at the base of the forelimb bud (Fig. 3 h), which later occurred here only within the roots of spinal nerves. In both types of experiments we also did not find neural crest cells in the otic capsules or occipital bones of the skull (Fig. 4a), which are of mesodermal origin [7,17]. Furthermore, all GFP+ cells close to the shoulder girdle were well co-localized with immunostaining for a glial marker, myelin basic protein (Fig. 4 b?e). Since we found no difference between the two types of experiments in the labelling of the neural crest derivatives in the neck and shoulder girdle region of the trunk, we further used double sided, but shorter neural fold fragment transplantations. To examine the possibility that neural crest cells migrate into the shoulder girdle at later stages of development, we also examined 2? year old adults (ossified bone) that had received double sided GFP+ neural fold transplants. We found that GFP+ signals were still restricted to nerve fibers of the brachial plexus and to neuronal nets in the muscles, but were not present 1379592 in the shoulder girdle itself or in muscle attachment sites (Fig. 4 f ). Altogether, these results show that neural crest does not contribute to the endochondral shoulder girdle in the axolotl.DiscussionThe complete absence of neural crest cells in the endochondral shoulder girdle of the axolotl contrasts with the apparent contributions of the neural crest to the endochondral shoulder girdle and muscle attachment sites in the mouse [9,18], where neural crest cells were interpreted to be descendants of the neural crest-derived dermal portion of the ancestral shoulder girdle [9]. However, we have found that the shoulder girdle in axolotl did not retain neural crest derivatives, neither as neural crest-derived cell populations within the endochondral elements nor as connective tissue at muscle attachment sites. This may reflect a more common vertebrate trait where muscle attachment points to the endochondral shoulder girdle have no neural crest, but only a mesodermal contribution. It further suggests that neural crest contribution to endochondral elements of the shoulder girdle and the skull and corresponding attaching muscles is not necessary for normal connections between those parts of the skeleton. Observations in several other species have further implied that the neural crest does not make a significant contribution to muscle attachments of the endochondral shoulder girdle in those species either. For example, in quail/chick chimeras, neural crest erived cells were found in the dermal clavicle, but very few to none have been observed in the region near the endochondral scapula [4,19] (I McGonnell and R Huang, pers. comm., own unpublished observations). In turtles, neural crest marker gene expressionResultsTo determine whether neural crest cells contribute to the shoulder girdle in the axolotl, we first grafted the left neural fold (posterior cranial to anterior trunk region) including neural crest cells from a GFP+ donor to a white (d/d) host (Fig. 2a). O.
