To investigate this question we examined pelvic muscle development in a range of species with the aim of answering two outstanding questions: (1) What was the developmental mechanism(s) at work in pelvic fins of fishes? (2) What changes resulted in the steps necessary for the evolution of these developmental mechanisms, which consequently allowed the generation of the different muscle morphologies necessary for the tetrapod transition. We postulated that three observations were possible. Either, (1) pelvic fin muscles were generated from a completely new mechanism or that (2) one of the existing mechanisms that had previously been characterized was altered/modified/re-utilized, or (3) the primitive mode found in cartilaginous fish orchestrated the formation of pelvic fin muscles. The zebrafish provided an opportunity to study these mechanisms in a genetically tractable model combining embryological manipulability with optical clarity of the early embryo and larvae, which would allow simple visualization of cell biological events directly in vivo.
In order to determine where pelvic fin muscle came from we performed orthotopic somite transplantation, to pinpoint the origin, intermediate and final position of the derivatives of the somite, between strains of transgenic zebrafish. This would allow us to determine if pelvic fin muscles have a somitic origin. The pelvic fins of zebrafish form at 3 weeks after fertilisation and significantly after the formation of pectoral fins which occurs by 48 hpf. (). We previously found that techniques such as lipophilic DiI labeling and uncaged fluorescein fate mapping13
were not robust enough and too short-lived for determining the origins of the pelvic fin muscle (5 weeks from somite label to fin muscle differentiation). Thus, we first needed to develop a long-term fate mapping strategy, which would allow us to examine the developmental origin of pelvic fin muscle precursors.
Figure 2. Pelvic fins, musculature and transplantation in zebrafish. (A) Lateral view of a 5 week old zebrafish. Scale bar= 1 mm. (B) Higher magnification view of area boxed in (A). (C and D) Lateral view (C) and ventral view (D), of green fluorescent (more ...)
When initially developing the technique we first transplanted donor somites expressing GFP from the α-actin GFP fish16
which expresses GFP in all skeletal muscle () into wild type hosts ().
Transplant surgery was performed on donor and host embryos at 15 somite stage (18 hpf). A single somite was carefully removed from the host while at the same time a single somite was removed from the donor and transferred into the host, and placed into the space where the host somite had been removed. The incision closed up over the next few hours and approximately 20% of the embryos survived through to adulthood. The transplanted tissue gave rise to normal developed structures and so we were able to observe the contribution of the donor somite to muscle growth of the host throughout development in real time in the living fish. We noticed that immediately ventral to, and separate from, each myotome was a muscle ( asterisks). Somites transplanted at any level would contribute to one of these blocks ventral to the corresponding myotome. The purpose of this muscle block is unclear but it may provide a rigid rod of muscle for the pelvic fins to contract against, rather like a rod running along the ventral surface of the fish. In the absence of any connection of the pelvic fin to the spine this may be required to prevent the fish “buckling” during contraction.
Transplantation could also provide the ability to visualize both the host and donor muscle in vivo. To do this we generated a zebrafish with RFP skeletal muscle (α actin-mCherrypc14
line). We next transplanted RFP expressing donor somites into GFP expressing hosts (). Specifically, we transplanted individual somites orthotypically from embryos carrying the muscle specific RFP transgene, into equivalent somite region of a stage matched host transgenic line of zebrafish that drives GFP expression from the same skeletal muscle-specific promoter16
(). Thus all the tissue generated in the host by the donor somite would be labeled red in a fish with green muscle allowing the clear visualization of the donor somite and its derivatives. Successfully operated embryos developed to adulthood and during this time were regularly documented to determine which musculature contains differentiated RFP positive cells.
Figure 3. Transgenic somite transplantation in D. rerio confirms pectoral fin muscle derive from somite 4 and pelvic fin muscles derive from myotomal extension of somite 10 and 11. (A and B) Cartoon detailing the method used for double transgenic (more ...)
In our final transplantation strategy we aimed to demonstrate that the donor somite contained only somitic tissue and that this only ever generated donor-derived muscle in the host. In order to do this performed a triple transgenic fluorescent transplant strategy. Two additional fish lines were generated, an α-actin BFP fish (blue skeletal muscle) and a ubiquitously expressed β-actin mCherry fish, which has mCherry in every cell. We transplanted donor GFP muscle/β-actin RFP into BFP muscle hosts. If the donor somite contributes to non-muscle tissue this tissue will be red in the host (see supplementary figure in Cole et al.17
To test our long-term fate mapping strategy we transplanted somite 4 from mCherry donor to GFP host. The 4th somite is known to give rise to the pectoral fin muscles which is formed by 48 hpf in zebrafish, as demonstrated by Neyt et al.13
using lipophilic DiI labeling. Similarly, when we transplanted somite 4, the pectoral fin musculature of the host consisted of functioning donor tissue. The donor tissue grew in the host for the entire life of the fish () and appeared to function normally such that no observable differences to the pectoral fin on the un-operated side were seen. This data confirmed that pectoral fin muscles do indeed form by the migration of precursors from somite 4 and that we had a workable long-term fate mapping strategy.
The origin of pelvic fin muscles was investigated by transplanting every somite from 8 to 14. The results clearly demonstrated that it is somites 10 and 11 that contribute to the pelvic fin muscles in zebrafish (). Somites anterior to 10 and posterior to 11 do not contribute precursors to pelvic fin muscles.17
We also found that if the host transplant was not initially placed ventrally enough in the host during the operation then the donor precursors would fail to contribute to the pelvic fin muscles even though the a large portion of the body wall muscle would be made from the donor tissue.17
It was clear the ventral tip of the extension was the source of the pelvic fin muscle precursors.
We then applied the more traditional methods of histology combined with immuno-histochemistry and in situ-hybridization to generate a complete developmental time series of pelvic fin muscle formation in the zebrafish. We analyzed expression of known players in muscle developmental mechanisms, including lbx and pax3, which mark the migrating muscle precursors. We found initially that a myotomal extension grew down toward the site of the future fin, and once in position, delamination and migration of lbx positive precursors occurred to deliver myoblasts to the fin. The data we obtained from examining pelvic fin muscle formation in zebrafish revealed that the muscles form in a process that appeared to be a combination of the primitive and derived mechanisms previously described.