Control of Limb Regeneration

1. Nerves: Since the pioneering experiments of Todd (1), it has been known that a denervated limb will not regenerate. The incredibly detailed series of experiments by Singer and colleagues in the 1940s (16) in which he partially dener-vated limbs, counted the remaining nerve fibre numbers at various limb levels and then recorded the resulting frequency of regeneration, led to the hypothesis that a threshold number (between 30 and 50% of normal) was required for regen-

Fig. 2. (continued) At the bottom of the micrograph are typical small, darkly staining chondrocytes. In the middle of the micrograph, the nuclei of the chondrocytes can be seen to be enlarging and staining less intensely. At the cut tip of the cartilage, cells are released into the blastema as the cartilage matrix has been degraded. (D) Apical cap. As dedifferentiation progresses and an early bud blastema begins to accumulate, the wound epithelium piles up thickly, and the apical cap (ac) forms at the tip of the stump. Cartilage dedifferentiation can be clearly seen from the cut ends of the radius and ulna. (E) Blastema. Dedifferentiation followed by proliferation of the released cells generates a blastema consisting of a mass of rapidly dividing, embryonic cells covered by an epithelium. This is at the medium bud blastema stage. (F) Redifferentiation. After the blastema has reached a certain size, redifferentiation begins proximally and spreads distally until all the elements that were removed by amputation are replaced. Here the solid line marks the amputation plane, and the new distal radius can be seen to be fused perfectly with the remaining proximal radius in the stump. The first two digits can clearly be seen as well as a large cartilage mass between the radius and the digits, which will form the wrist elements. This is at the early digit stage.

eration to occur. Either nerve type, motor or sensory, will suffice provided they are present in sufficient quantity. It was subsequently assumed that the nerves provide a neurotrophic factor and that it was concentration of this factor that Singer had been quantitating. The neurotrophic factor is responsible for stimulating blastema cell division. A strong candidate for the neurotrophic factor is glial growth factor (17), although no one has managed to replace completely the function of the nerves by a defined compound. This is an extremely difficult experiment to perform because of the problem of administering minute quantities of a test substance over a prolonged period of many weeks and the problem of keeping the limb denervated as amphibian nerves readily regrow after crushing or severing.

This well-founded neurotrophic theory also encompasses a fascinating paradox. Limbs that have never been innervated, so-called aneurogenic limbs, regenerate perfectly in the absence of nerves. When nerves are allowed to enter the aneurogenic limb, they gradually become dependent on innervation—the blastemas cells are thought to have become "addicted" to the neurotrophic factor. More recent experiments have cast some light on these rather vague concepts by showing that the nerve controls the molecular phenotype of blastemal cells (18).

2. Hormones: It is generally taken for granted that the appropriate "hormonal mileau" is required for regeneration. This mileau includes the adrenal corticosterioids, somatotrophin, thyroxine, insulin, and prolactin. However, it has been particularly difficult to demonstrate specific requirements for several reasons. First, removal of the gland under consideration is usually so severe an operation either physically or physiologically that the animals do not survive. Second, the successful removal of an organ, such as the pituitary, has such a profound effect on many other hormonal systems that it is impossible to dissect out individual requirements for regeneration. With the advent of cloned products, however, specific requirements are being demonstrated, e.g., for growth hormone (19). Third, removal of a gland followed by replacement therapy has either involved mammalian preparations whose similarity of action in amphibia is unknown, or impure preparations have been used. Nevertheless, we would expect that hormones, such as growth hormone or prolactin, which have such an important role in the control of basic cell metabolism, should be involved in regeneration, but not nesessarily in a controlling capacity.

3. Origin of blastemal cells: The cells that form the blastema arise from within 1-2 mm of the amputation plane by the process of dedifferentiation of the mesoder-mal tissues as described above. The epidermis cannot contribute to any internal tissues, but only forms the apical cap. Meticulous studies recording cell and mitotic counts have concluded that all mesodermal tissues contribute to the blastema (dermis, muscle, connective tissue, periosteum, bone and even Schwann cells) roughly in proportion to the number of cells in each tissue in a cross-section of the stump (20, although see 21). There is a tendency to assume that the majority of cells revert back to their former differentiated state, but this is not necesarily so, as two recent studies have conclusively demonstrated. In the first, cells were marked by grafting between diploid and triploid animals followed by meticulous cell counting. It transpired that there was an overrepresentative contribution of dermis from the stump and an underrepresentative contribution from the cartilage (21). In the second study, cultured myotubes were labeled with lysinated dextran, implanted into blastemas, and when redifferentiation began, labeled cells were occasionally seen in the cartilage of the regenerate (22). Thus, a proportion of cells seem to undergo a metaplastic transformation during normal regeneration.

Another type of experiment in which cells are forced to do more than they naturally would has also demonstrated the metaplastic potential of blastemal cells. In these experiments, limbs are irradiated with X-rays, which permanently prevents cell division, and then a graft of unirradiated tissue is provided, which supplies new cells with regenerative potential. If the graft is from a white axolotl and the host is a black axolotl, then the regenerate will be white confirming the origin of the tissues of the regenerate. In these experiments, it was shown that grafts of muscle could provide all the tissues of the regenerate, including cartilage (as revealed in the labeling experiment described above). However, grafts of dermis and grafts of cartilage could provide all the tissues of the regenerate, except muscle. Thus, it seems that most metaplastic transformations are possible, except that only myoblasts can generate new myoblasts.

This type of tissue transformation is readily demonstrable in a simple experiment where the cartilage of the stump is removed prior to amputation. Even though there is no cartilage at the amputation plane and no chondrocytes dedif-ferentiate to supply the blastema, perfect cartilage elements are produced distal to the amputation plane (Fig. 3A).

4. Regional and axial determination: That individual blastemal cells remember the region of the body from which they come is the general conclusion from a long history of grafting studies. Tail tissue grafted to limbs or vice versa or forelimb tissues grafted to hindlimbs or vice versa results in the regeneration of organs specific to the graft type, not the host type. Until recently, it had never been possible to change the organ specificity of cells, but this has now been done. The regenerating tail blastema of frogs can be homeotically transformed into tails by treatment with retinoids prior to metamorphosis (23,24). The same general conclusion is also true of axial determination. From the very beginning, blastemal cells carry a knowledge of their axial position. This has been demonstrated many times in experiments in which blastemas are cut off the stump and then either rotated 180° and put back on or grafted from left to right (or vice versa). The former manipulation reverses both anteroposterior and dorsoventral axes and results in the appearance of supernumerary limbs (Fig. 3B). The same is true if either the anteroposterior or dorsoventral axis is reversed (Fig. 3C). Thus, axes cannot be respecified. The interaction between cells whose axes conflict results in the generation of extra tissue to resolve the conflict.

In the proximodistal axis level-specific information is similarly present within the blastemal cells. Clearly, it must be or the limb would not know how much of itself to regenerate. If a proximal blastema is grafted onto a distal amputation stump, then the result will be a limb that has serially duplicated elements (Fig.

Fig. 3. Victoria blue-stained limb regenerates to show the cartilage patterns after various treatments. (A) The structure of the regenerate after removing all the cartilage and bone from a forelimb and then amputating through the lower arm level. A solid line marks the amputation plane. Proximal to the amputation plane, there is only muscle, which does not stain with Victoria blue. Distal to the amputation plane, the ends of the radius and ulna, wrist elements, and digits have regenerated perfectly despite the absence of cartilage in the stump from which the dedifferentiated tissues of the blastema were derived. This is a simple demonstration of tissue metaplasia. (B) The result of cutting off a blastema from a regenerating hindlimb, rotating it 180° and putting it back on the stump. Two supernumerary limbs (S1 and S2) have been produced in addition to the original limb. (C) The result of putting a left forelimb blastema on a right stump to reverse the anteroposterior axis. As in B, two supernumerary limbs are generated (S1 and S2) in addition to the original limb. (D) The result of grafting a proximal blastema (from the shoulder level) onto a distal amputation stump (through the ends of the radius and ulna). The result is two limbs in tandem. A solid line marks the amputation plane. (E) The result of grafting a distal blastema (from the wrist level) onto a proximal stump (the midupper arm level). The dotted line marks the level of the grafted distal blastema and the solid line marks the amputation plane through the mid upper arm. The result is a

3D). In this case, there seems to be no interaction between stump and blastema in reaction to the disparity in level-specific information. When the converse experiment is performed, a distal blastema grafted onto a proximal stump, then interaction does take place because the disparity is recognized and the gap filled in to generate a normal limb (Fig. 3E). The gap is filled in by proximal cells, not by distal cells, and these observations led to the formulation of the "law of distal transformation" which only allows cells within the limb field to become more distal, never more proximal. Significantly, just like the respecification of regional determination by retinoids, the only instance where the law of distal transformation has been broken is when distal blastemas are treated with retinoids. In this case, a complete limb, including proximal elements, can be regenerated from a distal level amputation after retinoid treatment (Fig. 3F) (25).

Was this article helpful?

0 0

Post a comment