Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Developmental Biology 221, 273–284 (2000) doi:10.1006/dbio.2000.9667, available online at http://www.idealibrary.com on REVIEW Regeneration in Vertebrates Panagiotis A. Tsonis Laboratory of Molecular Biology, Department of Biology, University of Dayton, Dayton, Ohio 45469-2320 One way or another, all species possess the ability to regenerate damaged tissues. The degree of regeneration, however, varies considerably among tissues within a body and among species, with urodeles being the most spectacular. Such differences in regenerative capacity are indicative of specific mechanisms that control the different types of regeneration. In this review the different types of regeneration in vertebrates and their basic characteristics are presented. The major cellular events, such as dedifferentiation and transdifferentiation, which allow complex organ and body part regeneration, are discussed and common molecular mechanisms are pinpointed. © 2000 Academic Press INTRODUCTION The fascination with regeneration of body parts has mythological origins in ancient times. Stories of regeneration are prominent in mythologies from many different cultures. In more recent times and especially in the 18th century, description of regenerative properties of animals formulated by Trombley, Reamur, Bonnet, and Spallanzani set the foundation for the birth of developmental biology and shaped the thought on reproduction (Dinsmore, 1991). During the past 25 years the field of limb regeneration provided the most concrete model for formulating concepts of pattern formation and positional information (Bryant et al., 1981). Regeneration, however, is a broad field with important idiosyncrasies. Not all animals are able to regenerate body parts, and not all tissues within a body can be equally repaired. Several important questions are in need of answers. What are the different types of regeneration and how do they differ? What determines the potential of tissue regeneration in an animal or among different species? Are there common features or mechanisms among the different types of regeneration? It is the goal of this review to address these questions and to present a useful synthesis for the students of regeneration. TYPES OF REGENERATION Not all tissues are equal in their regenerative potential or magnitude or in the mechanisms involved (Goss, 1969; 0012-1606/00 $35.00 Copyright © 2000 by Academic Press All rights of reproduction in any form reserved. Stocum, 1995). Having this as a starting point it is necessary to categorize the different types of regeneration and to pinpoint differences and similarities in the different types of regeneration. The reward is twofold. First, by defining the criteria, the mechanisms involved in the different types of regeneration can be more clearly defined. As a result, the exercise can provide insights concerning commonality. Having established such a relation between the different types of regeneration, it could create the necessary links and cross-talk between different disciplines and bring together scientists from different areas, which is most imperative for the success of regeneration research. Apart from wound healing (or wound repair), which is mostly closure of a wound by scar tissue, the degree of tissue renewal or regeneration in vertebrates varies in different tissues. In fact what is different is the complexity that is involved in the mechanisms and magnitude of regeneration. The simplest form of regeneration is the axonal outgrowth seen in severed nervous system. Regeneration by simple proliferation seen in organs, such as intestines, liver, or adrenal gland, is somewhat more complex. It involves proliferation of cells that compose the particular organ. Regeneration of other organs and tissues on the other hand can be channeled through the proliferation and differentiation of stem cells. More complex types of regeneration involve the process of dedifferentiation. In these cases (mostly seen in amphibia) cells at the damaged site dedifferentiate and then redifferentiate into their type. Regeneration of CNS (brain, spinal cord), intestine, and 273 274 Panagiotis A. Tsonis TABLE 1 Examples of Tissue and Organ Regeneration Organ Source of cells Axonal outgrowth CNS (spinal cord) CNS (auditory system) PNS Species Rats Experimentally induced (⫹FGF, NGF), neurotrophins, Nogo inhibition Inner hair cells Axonal outgrowth Rats Rats Cell proliferation Proliferation of liver cells Proliferation of glomerulosa cells Proliferation? Dedifferentiation? Mammals Mammals Rats Stem cells Mammals Stem cells Mammals Epidermis Liver Stem cells Stem cells Mammals Mouse Muscle Satellite cells Amphibia, mammals Proliferation Gastrointestinal mucosa Liver Adrenal Small intestine Stem cells Intestine (small and large) Cartilage/bone Dedifferentiation CNS (brain, spinal cord) Ependymal cells form blastema and back ependymal cells Intestine Serosal and smooth muscle produce blastema by dedifferentiation Heart Dedifferentiation of cardiomyocytes Dedifferentiation, transdifferentiation Pancreas Acinar or duct cells Remarks/factors PNS better than CNS (affinity of cells?) Rats Cheng et al., 1996; Oudege and Hagg, 1999; Grandpre et al., 2000; Chen et al., 2000 Ito et al., 1998 Pestronk et al., 1990; Daston and Ratner, 1991 TGF-␣ EGF Jones et al., 1999 Michalopoulos and DeFrances, 1997 Engeland et al., 1996 Experimentally induced (EGF, HGF) Kim et al., 1999 Inhibition of hepatocyte division, differentiation to oval cells Amphibia, fishes References Booth et al., 1999; Sattar et al., 1999 Prockop, 1997; Pittenger et al., 1999 Cotsarelis et al., 1999 Petersen et al., 1999 Cameron et al., 1986; Hansen-Smith and Carlson, 1979 EGF, PDGF, TGF- O’Hare et al., 1992 Amphibia O’Steen and Walker, 1962 Amphibia Oberpriller and Oberpriller, 1991 Mammals Dedifferentiation, transdifferentiation, pattern formation Eye (retina, lens) Pigment epithelium; Amphibia cornea Limb Blastema Amphibia Limb Blastema Mouse Transdifferentiation (Tgf-, Hox genes) Bouwens, 1998; Menke et al., 1999; Sharma et al., 1999 FGF Tsonis, 1999; Tsonis, 2000 Tsonis, 1996 Reginelli et al., 1995 Nerves, FGF Msx-1 Note. When “Mammals” appears, studies may include more than one species and even species lower than mammals. heart can be achieved by this mechanism of dedifferentiation. A more complex type of regeneration involves dedifferentiation and transdifferentiation and can be seen during pancreas regeneration. In this type, acinar and/or duct cells dedifferentiate and then transdifferentiate into insulinexpressing beta cells, thus reconstructing the lost part of Copyright © 2000 by Academic Press. All rights of reproduction in any form reserved. 275 Regeneration in Vertebrates the organ. The most complex type of regeneration, however, is seen in limb and eye regeneration (especially in amphibia) in which cells at the damaged or amputated site dedifferentiate and transdifferentiate, but they are also able to build back an exact replica of the lost part. This type involves the mechanisms of normal pattern formation as well. Historically, regeneration of amphibian limbs, and also regeneration of appendages or whole body parts in invertebrates, has been termed as epimorphic, while the rest of regeneration has been called tissue regeneration. Such a distinction, however, has not always been clear-cut and should be used as such only for the sake of convenience. Examples of regeneration grouped according to the increasing complexity are outlined in Table 1. In this table remarks on factors and mechanisms are also presented. It is obvious from Table 1 that overlapping mechanisms can control the different types of regeneration. For example, in both amphibian heart and limb regeneration muscle dedifferentiation is common, but limb regeneration is more complex because it involves the additional steps of transdifferentiation and pattern formation. These ideas could be very helpful in identifying common criteria and mechanisms in the different types of regeneration. As mentioned above, regeneration of amphibian lens, retina, and limb is the most complex and, therefore, the most spectacular of all. Such regeneration is possible by the recruitment of many different mechanisms involved in the process of dedifferentiation, transdifferentiation, and faithful pattern formation of the regenerated structure. Dedifferentiation is the most basic element of such form of regeneration. It entails terminally differentiated cells reentering the cell cycle and losing the typical characteristics of their origin. The only parallelism is when a normal cell is transformed to a cancer cell. Transdifferentiation allows a cell to change its identity and become a different cell. Regenerating an exact copy of a pattern that was lost involves all the marvels of embryogenesis where formation of tissues, organs, and whole body structures are controlled by a unique weaving of genetic programs. Having introduced the basic types of regeneration and outlined examples and mechanisms, the concepts of dedifferentiation, transdifferentiation, and pattern formation will be further explored in this review. For this I will concentrate on the paradigms of amphibian retina, lens, and limb regeneration, which best represent these concepts. Furthermore, as it will become apparent below, these regenerative processes might be controlled by similar mechanisms, and this might help elucidate the crucial molecular pathways that have en- dowed amphibia with such outstanding regenerative abilities. RETINA REGENERATION During eye development, the optic vesicle that originated from the anterior neural tube can give rise to either neural retina, a multilayered structure containing photoreceptors and neurons, or pigmented epithelium consisting of a single layer of nonneuronal, pigmented cuboidal cells. The anterior part of the optic vesicle that is in closer contact with the ectoderm becomes neural retina. Such an induction is likely to be controlled by fibroblast growth factor (FGF), which is found in abundance in the ectoderm and which has been shown to induce the neural phenotype in cultures of optic vesicle explants (Reh and Pittack, 1995). Retina regeneration takes place in a variety of animals, such as fishes, birds, amphibia, and mammals (Mitashov, 1996). The source of retina cells in fishes, frogs, and mammals seems to be precursor cells in the peripheral growth zone of the dorsal and ventral iris, the same cells that support normal growth throughout life. The most spectacular mode of retina regeneration, however, can be observed in chicks and amphibia, and it occurs by the transdifferentiation of the retinal pigment epithelial cells to all types of neural retina cells. In embryonic chicks (stage 22–24) such a transdifferentiation process can occur only if some neural retina is left behind. If all neural retina is removed, regeneration occurs only by treatment with -FGF (Park and Hollenberg, 1989). FGF, therefore, must be the most important factor in the induction of transdifferentiation and neural retina regeneration. The stages of these events are as follows. After dedifferentiation, the depigmented cells start dividing. One cell commits itself to the restoration of pigment epithelium and the other transdifferentiates to neural cells. First the photoreceptor cells are produced, then the cells of the amacrine layer and finally the ganglion cells (Fig. 1). In vitro systems for retina regeneration in which pigment epithelium explants can transdifferentiate to neural retina have been established. Using such in vitro system the effect of FGFs in inducing transdifferentiation has been verified (Reh and Pittack, 1995). Little is known about gene regulation during these events. Obviously FGF must down-regulate RPE cell-specific gene expression. Two transcriptional factors are known to control pigment epithelium identity. One is the product of Mitf (microphthalmia) and the other the product of pax-6 gene. Mutations in Mitf in chick embryos FIG. 1. The process of dedifferentiation during retina and lens regeneration in the adult newt. (a) A section through an adult newt retina showing the three layers of the neural retina, photoreceptors (p), amacrine (am), and ganglion cells (g). The dark pigment epithelium (arrows) lies over the photoreceptor layer. (b) The eye after retinectomy. The neural retina has been completely removed and the pigment epithelium (arrows) is intact. (c) Dedifferentiating pigment epithelium and transdifferentiation into neural retina cells (r) 14 days after retinectomy. Arrows denote pigment epithelium cells. (d) Transdifferentiation of the pigment epithelium from the dorsal iris (di) and formation of the regenerating lens vesicle (lv). Copyright © 2000 by Academic Press. All rights of reproduction in any form reserved. 276 Panagiotis A. Tsonis FIG. 2. The process of muscle dedifferentiation during limb regeneration in the adult newt. (a) The top of the picture is near the amputation level. The top half shows the dedifferentiation process in the muscle. The fibers have “melted down” and the characteristic striated nature of the muscle has been lost. The nuclei from the multinucleated muscle fibers are released to form mononucleated blastema cells and some are shown dividing (arrow). The loss of the muscle-specific characteristics is shown in b and c. (b) Scanning electron micrograph from a dedifferentiating muscle cell (such as the ones on the top of a) showing the loss of the muscle cell characteristics. (c) Scanning electron micrograph through a muscle cell from a normal muscle cell (such as the ones on the bottom of a) showing the normal structure of the cytoplasm with the hexagon-like shape of the myofibrils. The arrows show the sarcoplasmic reticulum. EM, extracellular matrix; n, nucleus. Copyright © 2000 by Academic Press. All rights of reproduction in any form reserved. 277 Regeneration in Vertebrates FIG. 3. The limb–lens connection. Lens formation in the regenerating limb. (a) A perfectly formed lens after dissociated pigment epithelial cells from the dorsal iris were implanted into the blastema. The lens possesses a normal appearance with normal posterior–anterior polarity (direction of the arrow), with the anterior (front) part facing the wound epithelium. In (c) the lens is magnified to show its normal appearance. The lens epithelium in the anterior (A) shows the characteristic cuboidal shape, while in the posterior (P) the cells become elongated and differentiate to lens fibers (lf). The AP polarity of lens does not coincide with the AP axis of the limb. (b) Failure to form a lens after transplantation of dissociated pigment epithelium cells from the ventral iris. The cells have not dedifferentiated and have remained pigmented. (d) Magnification of the pigment aggregate shown in b. (Reproduced by permission from Ito et al., 1999.) Copyright © 2000 by Academic Press. All rights of reproduction in any form reserved. 278 Panagiotis A. Tsonis result in loss of pigment epithelium and its transdifferentiation into neural retina (Mochii et al., 1997). These mutant animals have in fact developed with two neural retinas. Mutations in the pax-6 gene result in several abnormalities in the eye tissues, including in the pigment epithelium (Ton et al., 1991). FGF in fact down-regulates both genes. Although expression studies of pax-6 and Mitf during retina regeneration have not been performed, their down-regulation by FGF strongly supports their role in retina regeneration. LENS REGENERATION During development, the lens is induced by contact between the ectoderm and the optic vesicle. Likewise in retina development, FGFs seem to control such an induction (de Iong and McAvoy, 1993). Also, FGFs determine the polarity of the lens. It has been shown that FGF-1 is present as a gradient in the eyeball, with higher concentration needed for fiber differentiation in the posterior chamber and lower concentration in the anterior, where the lens epithelial cells are (Caruelle et al., 1989). Other important genes for such an embryonic induction are pax-6 and six-3 (Oliver et al., 1996; Altman et al., 1997). When the lens is removed, however, regeneration occurs by the dedifferentiation of the pigment epithelial (PE) cells of the dorsal iris (Fig. 1). These cells proliferate and produce a lens vesicle and later differentiate to form lens fibers and finally, 20 –25 days later, a normally polarized lens (see Tsonis, 1999, 2000, for review). As in retina regeneration, lens regeneration is restricted mostly to some urodeles. Some fishes can also regenerate their lens, while in chickens it is possible during a very narrow window during development. In vivo lens regeneration is possible only from the dorsal iris (never from the ventral). Nevertheless, pigment epithelial cells from anywhere in the eyecup are capable of transdifferentiating to lens in culture, and this ability has no species restrictions (Kodama and Eguchi, 1995). Even PE cells from aged humans have this capability. This suggests that in vivo, but not in vitro, there is spatial restriction to a specific position. Such a restriction implies that there must be specific regulation in the dorsal versus ventral iris. This is an excellent system in which to identify the components of the system that lead to inhibition except at the dorsal iris. Despite the difference between the induction during development and regeneration, it seems that similar regulatory events occur in both. Some of the factors that have been studied in detail are the FGFs and their receptors. During dedifferentiation of the PE from the dorsal iris, expression of FGF-1, FGFR-1, FGFR-2, and FGFR-3 is prominent in the dedifferentiating cells and in the subsequent regenerating lens vesicle and differentiating lens fibers (Del Rio-Tsonis et al., 1997). However, only FGFR-1 product seems to be present specifically in the dorsal iris during dedifferentiation. Its role in regulating lens regeneration has also been strengthened in experiments in which an inhibitor of FGFR-1 inhibited lens regeneration and lens fiber differentiation (Del Rio-Tsonis et al., 1998). In addition, as in cases of transgenic mice, exogenous FGFs were capable of inducing similar abnormalities in the regenerating lens (Del Rio-Tsonis et al., 1997). These abnormalities included vacuolated lens, double lens formation, and lenses with abnormal polarity. The FGF story can tell us that this particular signaling pathway is important for both retina and lens development and regeneration and thus conserved in both. In older experiments it has been shown that proteoglycans are lost from the dorsal iris during the processes of regeneration (Zalik and Scott, 1973). It is possible that this event is linked to FGFs since FGFs bind proteoglycans and affect availability of FGFs to activate the function of receptors. Such a line of research is very important for this field because it might provide clues about the key cell surface changes that govern the dedifferentiation process and thus lens regeneration. Spatial regulation along the dorsal–ventral axis may imply that genes such as Hox genes are involved. It has been well established that these genes are major players in cell fate determination, organogenesis, and pattern formation. It is, therefore, likely that Hox genes regulate the events of lens regeneration. Several Hox genes that are expressed in the newt eye have been studied, but two homeoboxcontaining genes have been pinpointed as key regulators so far. One is pax-6, a known lens-inducing gene. Pax-6 has been found to be expressed in the adult newt eye and during regeneration of the lens. However, its expression was not evident in the axolotl, a urodele unable to regenerate the lens (Del Rio-Tsonis et al., 1995). Prox-1, another Hox gene which seems to be important for lens development (Tomarev et al., 1996), has been found to be specifically expressed and regulated in the pigment epithelium of the adult newt dorsal iris and in the dorsal iris during lens regeneration (Del Rio-Tsonis et al., 1999). Such expression patterns suggest a role for these two genes in regenerationcompetent PE cells. The field of lens regeneration has benefited enormously from in vitro studies. As mentioned above, PE cells placed in culture undergo dedifferentiation and subsequent transdifferentiation to form lentoids. This has helped to unequivocally prove that PE cells indeed can transdifferentiate (the process has been observed even from a single cell; Kodama and Eguchi, 1995). In vitro experiments have shown the promoting effect of FGF, which is consistent with the role of FGFs and their receptors during in vivo regeneration (Hyuga et al., 1993). Finally, the ability of PE cells from other species to transdifferentiate in vitro opened new avenues in the field. We now know that PE cells from other species, including humans, do have the program to produce lens, and that knowledge could help in future applications (Kodama and Eguchi, 1995). Culturing has also been proven indispensable in the study of PE cells. For example, ventral iris or retina PE cells can transdifferentiate into lens cells in vitro. Ventral cells or explants from the ventral iris, cultured in vitro, are able to produce a lens Copyright © 2000 by Academic Press. All rights of reproduction in any form reserved. 279 Regeneration in Vertebrates upon transplantation into a lentectomized eye, but this depends on the duration of the culturing before transplantation. Obviously, some disruption in cell-to-cell communication and cell surface, which occurs when cells are cultured, is important for the initiation of dedifferentiation. Furthermore, Ito et al. (1999) have shown that ventral iris PE cells can in fact deprive the dorsal iris PE cells from dedifferentiating in vivo. When a number of dorsal iris PE cells (a number of cells able to produce a lens upon transplantation) is mixed with fivefold the number of PE cells from the ventral iris, their potential for lens transdifferentiation is inhibited. It is possible that the ventral PE cells produce a factor that inhibits their ability for dedifferentiation in vivo. Transplantation experiments have also revealed that the eye is not necessarily needed for dedifferentiation and transdifferentiation of PE cells into lens cells. When dissociated PE cells were implanted into the blastema of a regenerating limb a perfect lens can be formed with even a correct anterior–posterior orientation (see also below) (Ito et al., 1999). LIMB REGENERATION While the process of retina or lens regeneration is basically the transdifferentiation of one cell type to others, the regeneration of amphibian limbs (and tails) is more complex. After amputation, the wound is quickly covered by a specialized epithelium, the so-called wound epithelium. It is strongly believed that this epithelium provides the necessary signals for the underlying tissues to dedifferentiate, proliferate, and form the blastema. All the tissues at the stump undergo dedifferentiation, including muscle, bone, and other mesodermal tissues. The dedifferentiation process leads to cells that by proliferation form the blastema (Fig. 2). After a period of about 2 weeks the blastema redifferentiates to form an exact replica of the severed part (Tsonis, 1996). Due to its complexity the process of limb regeneration is regarded as more spectacular. Expression in the wound epithelium has been studied with the hope of identifying the factors that signal the initiation of the dedifferentiation process. While several factors have been found unique to the wound epithelium, FGFs and their receptors are also thought to be paramount for the signaling that leads to regeneration. In urodeles, FGF-1 and FGF-2 have been found in the wound epithelium. Interestingly, FGF-1 and FGF-2 have been implicated in the nerve dependency of limb regeneration (Mullen et al., 1996; Zenjari et al., 1997). FGFRs are also expressed in the newt wound epithelium. The KGFR variant of FGFR-2 is specifically located in the basal layer of the wound epithelium, while the bek variant is mesenchyme specific (Poulin et al., 1993). In another study using Xenopus limbs it was found that FGFR-1 and FGFR-2 are expressed in the wound epithelium of stumps at stages at which regeneration is permissive (premetamorphic), but they are absent from the wound epithelium of stumps that are unable to regenerate (postmetamorphic). More importantly, however, when premetamorphic amputated limbs were treated with FGFR inhibitors, regeneration was impaired, with formation of spikes reminiscent of the quality of regenerates seen in postmetamorphic frogs (D’Jamoos et al., 1998). The mechanism of dedifferentiation, which allows cells to reenter the cell cycle, must be a key for the ability to regenerate. Cells, such as muscle cells, lose their characteristic architecture (such as the actin–myosin arrangement); they become mononucleated and they start to divide (Fig. 2). Cell lines of dedifferentiated muscle cells, produced from muscle explants, have been useful in studying cellular and molecular events (Lo et al., 1993; Tsonis et al., 1995). When these cells were implanted into an amputated limb, in addition to muscle they were also able to produce cartilage, thus providing significant evidence for the process of transdifferentiation in limb regeneration as well (Lo et al., 1993; Tsonis et al., 1995). Also, a key molecular event associated with the process of dedifferentiation has been identified. When these cells are allowed to form myotubes in vitro and then induced to enter the cell cycle and become mononucleated (by serum stimulation) the product of the retinoblastoma (Rb) gene is specifically phosphorylated. Myotubes dominantly express the hypophosphorylated form of Rb, which inhibits the entry into the S phase. But as dedifferentiation and entering the cell cycle ensues, the hyperphosphorylated form of Rb (inactive) becomes dominant. This Rb phosphorylation event must be dedifferentiation-specific (Tanaka et al., 1997). Indeed, mouse myotubes are not able to enter the cell cycle (or undergo phosphorylation of Rb) after serum stimulation (Tam et al., 1995). The serum factor seems to be thrombin, which alone is able to stimulate newt myotubes to reenter the cell cycle (Tanaka et al., 1999). Only cells transfected by viral oncogenes are induced to enter the S phase by sequestering the Rb product (Cardoso et al., 1993). After the blastema is built it redifferentiates to reconstruct the lost part. This process is obviously very complex and key developmental genes are expressed to secure that the exact pattern will be regenerated. In this sense, it is expected that genes that control pattern formation during development will also be expressed and control patterning during regeneration. In the late 1970s it was shown that retinoic acid is able to respecify the positional memory of the blastema cells (Niazi and Saxena, 1978). Retinoic acid is able to proximalize the pattern when a limb is amputated distally, i.e., at the wrist level. Such a treated limb will regenerate as if it was amputated at the humerus level (for details on limb regeneration see Tsonis, 1996). This ability of retinoic acid suggested that control of limb morphogenesis might occur via genes that are affected by retinoic acid and provided the first hope for isolation of these genes. The excitement grew the next decade when it was found that the receptors for retinoic acid (RARs) are in fact transcriptional factors with very distinct properties (Evans, 1988). Several RARs have been isolated and found to be expressed in the newt blastema, among them, the RAR ␣ (both Copyright © 2000 by Academic Press. All rights of reproduction in any form reserved. 280 Panagiotis A. Tsonis isoforms ␣1 and ␣2), RAR , and RAR ␦ (isoforms ␦1 and ␦2). With a series of experiments that involved the creation of chimeric receptors and transfection of these receptors in the regenerating limb, Brockes and his colleagues were able to show that different receptors account for the different actions of retinoic acid. RAR ␣1 mediates the growth inhibition of blastema cells, one of the earliest effects of retinoic acid treatment (Schilthuis et al., 1993). RAR ␦1 is responsible for inducing wound epithelium-specific genes and RAR ␦2 is the one that mediates the ability of retinoic acid to proximalize the pattern (Pecorino et al., 1994, 1996). Similar to these RARs, attention has been paid to HOX genes as regulators of the morphogenetic events during limb regeneration (Gardiner and Bryant, 1996). Indeed some of the Hox genes have shown a proximodistal gradient of expression which is also regulated by retinoic acid (see Tsonis, 1996; Beauchemin et al., 1998). Hox genes (especially the 5⬘ genes from the A and D cluster) have been shown to be regulators of the developing limb pattern in chicks and mice. Consequently, it was thought that expression studies of Hox genes of the A and D cluster would provide important insights of how the pattern is formed in the regenerating limbs. HOXA-9 and A-13 expression in regenerating axolotl limbs showed some differences compared with expression during development. While in developing limbs A-13 was more distally expressed than A-9, during regeneration there was no difference. Moreover, A-13 was down-regulated with retinoic acid treatment during regeneration (Gardiner et al., 1995). HOXD-11 (but not D-8 or D-10) seemed to be specific for blastema formation. HOXD-11 appeared first in the early bud blastema at the posterior-distal part, when only the hand region is represented in the blastema (Torok et al., 1998). This might indicate that (unlike developing limb) during regeneration the distal tip is the first to be specified. Such expression patterns of HOX genes might explain the mechanisms of pattern formation during limb regeneration in line with theoretical models that have been established in the past (Bryant et al., 1981). Thus, the regenerating limb provides an invaluable system to study pattern formation and mechanisms of morphogenesis. Even though mammals are not endowed with such regenerative abilities, the tips of the toes in mice and even humans (young children) can be regenerated only when amputated distally to the last interphalangeal joint. An interesting association with this digit regeneration and the expression of the homeobox-containing gene msx-1 has been reported in mice. The domain of regeneration lies within the domain of msx-1. Msx-1 is believed to play a role in the maturation of limb tissues (Reginelli et al., 1995). This mammalian model can become a very important system in which similarities and differences with the amphibian counterpart can be studied. Related to these mammalian models, Goss has proposed that inhibition of blastema formation by dermal healing (as opposed to wound epithelium healing) seen in mammalian limb amputations could account for the loss of regenerative ability in mam- mals (Goss, 1980). These differences in healing could be explored in order to identify factors that are differentially present in the one healing mode versus the other. The identification of msx-1 in mouse (Reginelli et al., 1995) and of FGFRs in Xenopus limb regeneration (D’Jamoos et al., 1998) are good prospects toward this direction. THE RETINA–LENS–LIMB CONNECTION Despite the fact that the processes of retina, lens, and limb regeneration are quite different from each other and involve different tissues and structures, they all start with the dedifferentiation process. In this regard, we could hypothesize that what triggers dedifferentiation could be common in these regenerative processes. From what was mentioned above, some very interesting similarities are inescapably obvious. The most striking is the possible role of FGFs in signaling the initiation of all these events. As shown in Fig. 3, a perfect lens can be formed after transplantation of PE cells into the limb blastema. Even the anterior–posterior polarity of the lens is intact, with the anterior part of the lens (front) facing the wound epithelium (the reader should not confuse the AP axis of the limb with the polarity of the lens). This is interesting because FGF has been implicated in determining this polarity in the eye and because the wound epithelium is a rich source of FGF. It is tempting to speculate that FGF signaling does indeed play a more general role in inducing the dedifferentiation process in the different regenerative tissues in newt. FGFs have also been found to support spinal cord regeneration and to initiate liver development from the gut endoderm (Chen et al., 1996; Jung et al., 1999). Such similarities and common themes in regeneration can be studied further as more mechanisms of dedifferentiation are revealed (e.g., Rb phosphorylation). Identification of a common mechanism for dedifferentiation could unify concepts and prove paramount in expanding the study of regeneration in other animals as well. STEM CELLS IN REGENERATION Stem cells seem to be a prominent feature in regeneration of many organs and tissues such as intestine, cartilage, bone, epidermis, liver, and muscle. On the other hand, dedifferentiation of cells of local origin has been proven beyond any doubt in newt retina, lens, and limb regeneration. The search for cells other than local as a source of epimorphic regeneration has not been fruitful. However, some participation of stem cells, especially in limb regeneration, should not be ruled out. Mesenchymal stem cells (MSCs) from bone marrow have been shown to differentiate to bone, cartilage, adipocytes, or muscle and are thought to contribute to repair of these tissues (Prockop, 1997; Pittersen et al., 1999). But can such MSCs participate in the restoration of an amphibian limb? While the local origin of Copyright © 2000 by Academic Press. All rights of reproduction in any form reserved. 281 Regeneration in Vertebrates TABLE 2 Regeneration of Body Parts and Origin of the Blastema in the Animal Kingdom Species Part of the body Stentor (protozoa) Hydroids Flatworms Annelids (worms) Echinoderms Arthropoda Whole Whole Whole Whole Whole Extremities Fishes Amphibia Fins, retina, lens Limbs, retina, lens, jaw, tail Fingertip Mammals liver regeneration). Local dedifferentiation and production of a blastema could in fact preserve such memory and secure the faithful regeneration of a particular pattern. Mode of regeneration Recapitulates cell division No blastema Reserve cells Blastema Blastema Related to growth (molting); blastema? Blastema Blastema Blastema Note. Some hydroids and flatworms can regenerate by morphallaxis, in which the remaining tissues reorganize to replace the missing part with little dedifferentiation and proliferation. The rest (except in protozoa) can be considered epimorphic regeneration. For details see Goss (1969). blastema renders this unnecessary, such a case should not be impossible, especially when it is known that bone marrow contains MSCs that differentiate to cells that comprise the limb tissues. Recently it was shown that bone marrow stem cells were able to produce oval cells that contributed to liver regeneration when proliferation of hepatocytes was inhibited (Petersen et al., 1999). It seems that if a system is driven toward a particular direction, stem cells might take over. Due to lack of bone marrow markers and techniques, such experiments are lacking in the newt limb regeneration field. Since such experiments are feasible now and could be very informative, they should be pursued. Several issues could be addressed from studies with newt bone marrow MSCs. It will be very informative and productive to define similarities or differences between blastema cells and MSCs. Not much is known about specific gene expression in MSCs; however, some striking similarities do exist. Both MSCs and blastema cells are fibroblastic-like and they synthesize similar extracellular matrix rich in collagens and fibronectin. In a strict terminological sense a blastema cell that is differentiating to a cartilage cell should be similar to a MSC differentiating to cartilage. Consequently, understanding the biology of the blastema cells should be complementary in the study of stem cells and vice versa. Obviously, different strategies of regeneration have made use of either stem cells or locally produced dedifferentiated cells. Why one strategy is preferred over the other depending on the tissue or the mode of regeneration is not clear, but it will certainly be very important to answer this question. For example, an intriguing difference could be the positional memory of the cells. In cases of epimorphic regeneration the missing structure should be replicated with the exact pattern (something that requires positional information, but is not necessary, for example, in ORIGIN OF THE BLASTEMA IN VERTEBRATES In Table 2 the occurrence of body part regeneration in the animal kingdom is presented with respect to the origin of the blastema. Admittedly, invertebrates possess much more spectacular regenerative abilities and can readily regenerate whole body parts. In contrast, in vertebrates regenerative abilities have been scaled down and limited to appendages and eyes. Also, it is evident from Table 2 that as we climb from simple to more complex organisms, dedifferentiation or blastema formation is the only means for epimorphic regeneration and, of course, best represented in some fishes and amphibia. But even among urodeles, regenerative abilities are not, however, widespread nor do they follow a clear cladistic pattern. In closely related urodeles the capability of lens or limb regeneration is not universal and does not follow a cladistic pattern (Fig. 4). Is, therefore, the occurrence of blastema an evolutionary “remnant,” or atavistic feature, or a tangible physiological and developmental event that depends on specific molecular mechanisms? In order to answer the question of the origin of the blastema we should very carefully consider the molecular and cellular mechanisms that characterize its formation. The most seminal feature of the blastema is dedifferentia- FIG. 4. Relationship of salamander families as inferred using maximum parsimony analysis of aligned ribosomal RNA sequences (after Larson, 1991). Yes indicates presence and No absence of limb regeneration in adult salamanders (when known). No obvious cladistic pattern correlating with the capability of limb regeneration can be noted. Copyright © 2000 by Academic Press. All rights of reproduction in any form reserved. 282 Panagiotis A. Tsonis tion and the loss of tissue characteristics. In order for this to happen, terminally differentiated cells (such as muscle) should reenter the cell cycle. As mentioned above a very interesting molecular signature of the amphibian limb blastema is the phosphorylation patterns of the tumor suppressor Rb, which result in its inhibition. In mammals, this is seen only in cells allowed to enter the cell cycle after transfection with viral oncogenes (Cardoso et al., 1993). Another interesting and intriguing similarity between blastema and cancer cells is the identification of immortalizing sequences in a newt blastema cDNA expression library. Immortalizing sequences are usually oncogenes, genes that are paramount in the induction of cancer (Powell et al., 1998). Indeed, cancer cells do look like blastema cells; they are dedifferentiated and pluripotent. While we should exercise caution in equating blastema and cancer cells their similarities can be very informative and instructive. In the past it has been speculated that cancer and regeneration are inversely associated. In other words, an animal with regenerative capabilities is refractory to spontaneous or experimental cancer. This is especially true for the regenerationcompetent tissues of amphibia. It has been well documented that spontaneous tumors are difficult to find in amphibia (Tsonis and Del Rio-Tsonis, 1988). Also, carcinogen application in the regeneration-competent tissues of amphibia (such as limbs and eye) would result in normal morphogenesis and differentiation rather than tumor formation (Eguchi and Watanabe, 1973; Tsonis, 1983). These facts could suggest that cancer originated at the expense of regeneration. Perhaps as epimorphic regeneration ceased to exist cancer originated. The trade could be significant for the development of more advanced immune system and immunosurveillance mechanisms (Prehn, 1971). Reexpression of embryonic antigens in blastema requires a weak immune system, which in turn is a high cost to pay. Perhaps regeneration is not the result of a “trait” but a well-orchestrated developmental event governed by genes that control normal growth and differentiation. The synthesis presented in this review does not separate the different aspects of regeneration research, but rather leads to finding links that would increase collaboration of the different subfields of regeneration research and eventually will lead to a better understanding and to applications in medicine. It is my conviction that regeneration research will occupy a bright spot in the pantheon of biomedical achievements of the 21st century. ACKNOWLEDGMENTS This work was supported by NIH Grant EY10540. I thank Dr. M. Okamoto for providing Fig. 3. Due to space limitations it was not possible to include all references for such a broad topic. I apologize to authors whose work was not cited. I am also grateful to two anonymous reviewers for their helpful comments. REFERENCES Altman, C. R., Chow, R. L., Lang, R. A., and Hemmati-Brivanlou, A. (1997). Lens induction by pax-6 in Xenopus laevis. Dev. Biol. 185, 119 –123. Beauchemin, M., Del Rio-Tsonis, K., Tsonis, P. A., Tremblay, M., and Savard, P. (1998). Graded expression of Emx-2 in the adult newt limb and its corresponding regeneration blastema. J. Mol. Biol. 279, 501–511. Booth, C., O’Shea, J. A., and Potten, C. S. (1999). Maintenance of functional stem cells in isolated and cultured adult intestinal epithelium. Exp. Cell Res. 249, 359 –366. Bouwens, L. (1998). Transdifferentiation versus stem cell hypothesis for the regeneration of islet beta-cells in the pancreas. Microsc. Res. Tech. 43, 332–336. Bryant, S. V., French, V., and Bryant, P. J. (1981). Distal regeneration and symmetry. Science212, 993–1002. Cameron, S. A., Hilgers, A. R., and Hinterberger, T. J. (1986). Evidence that reserve cells are a source of regenerated adult newt muscle in vitro. Nature310, 607– 608. Cardoso, M. C., Leonhardt, H., and Nadal-Ginard, B. (1993). Reversal of terminal differentiation and control of DNA replication: Cyclin A and Cdk2 specifically localize at subnuclear sites of DNA replication. Cell 74, 979 –992. Chen, M. S., Huber, A. B., Van Der Haar, M. E., Frank, M., Schnell, L., Spillmann, A. A., Christ, F., and Schwab, M. E. (2000). Nogo-A is a myelin-associated neurite outgrowth inhibitor and an antigen for monoclonal antibody IN-1. Nature 403, 434 –399. Cheng, H., Cao, Y., and Olson, L. (1996). Spinal cord repair in adult paraplegic rats: Partial restoration of hind limb function. Science 273, 510 –513. Cotsarelis, G., Kaur, P., Dhouailly, D., Hengge, U., and Bickenbach, J. (1999). Epithelial stem cells in the skin: Definition, markers, localization and functions. Exp. Dermatol. 8, 80 – 88. Daston, M. M., and Katner, N. (1991). Expression of P30, a protein with adhesive properties in Schwann cells and neurons of the developing and regenerating peripheral nerve. J. Cell Biol. 112, 1229 –1239. De Iong, R., and McAvoy, J. W. (1993). Spatio-temporal distribution of acidic and basic FGF indicates a role of FGF in rat lens morphogenesis. Dev. Dyn. 198, 190 –202. Del Rio-Tsonis, K., Washabaugh, C. H., and Tsonis, P. A. (1995). Expression of pax-6 during urodele eye development and lens regeneration. Proc. Natl. Acad. Sci. USA 92, 5092–5096. Del Rio-Tsonis, K., Jung, J.-C., Chiu, I.-M., and Tsonis, P. A. (1997). Conservation of fibroblast growth factor function in lens regeneration. Proc. Natl. Acad. Sci. USA 94, 13701–13706. Del Rio-Tsonis, K., Trombley, M. T., McMahon, G., and Tsonis, P. A. (1998). Regulation of lens regeneration by fibroblast growth factor receptor 1. Dev. Dyn. 213, 140 –146. Del Rio-Tsonis, K., Tomarev, S. I., and Tsonis, P. A. (1999). Regulation of prox-1 during lens regeneration. Invest. Ophthalmol. Visual Sci. 40, 2039 –2045. Dinsmore, C. H., Ed. (1991). “A History of Regeneration Research.” Cambridge Univ. Press, New York. D’Jamoos, C., McMahon, G., and Tsonis, P. A. (1998). Fibroblast growth factor receptors regulate the ability for limb regeneration in Xenopus laevis. Wound Rep. Reg. 6, 388 –397. Eguchi, G., and Watanabe, K. (1973). Elicitation of lens formation from the ventral iris epithelium of the newt by a carcinogen, MNNG. J. Embryol. Exp. Morphol. 30, 63–71. Copyright © 2000 by Academic Press. All rights of reproduction in any form reserved. 283 Regeneration in Vertebrates Engeland, W. C., Gomez-Sanchez, C. E., Fitzgerald, D. A., Rogers, L. M., and Holzwarth, M. A. (1996). Phenotypic changes and proliferation of adrenocortical cells during adrenal regeneration in rats. Endocr. Res. 22, 395– 400. Evans, R. M. (1988). The steroid and thyroid hormone receptor superfamily. Science 240, 889 – 895. Gardiner, D. M., Blumberg, B., Komine, Y., and Bryant, S. V. (1995). Regulation of HoxA expression in developing and regenerating axolotl limbs. Development 121, 1731–1741. Gardiner, D. M., and Bryant, S. V. (1996). Molecular mechanisms in the control of limb regeneration: The role of homeobox genes. Int. J. Dev. Biol. 40, 797– 805. Goss, R. J. (1969). “Principles of Regeneration.” Academic Press, New York. Goss, R. J. (1980). Prospects of regeneration in man. Clin. Orthop. 151, 270 –282. Grandpre, T., Nakamura, F., Vartanian, T., and Strittmatter, S. M. (2000). Identification of the Nogo inhibitor of axon regeneration as a reticulon protein. Nature 403, 439 – 444. Hansen-Smith, F. M., and Carlson, B. M. (1979). Cellular responses to free grafting of the extensor digitorum longus muscle of the rat. J. Neurol. Sci. 41, 149 –173. Hyuga, M., Kodama, R., and Eguchi, G. (1993). Basic fibroblast growth factor as one of the essential factors regulating lens transdifferentiation of pigmented epithelial cells. Int. J. Dev. Biol. 37, 319 –326. Ito, J., Murata, M., and Kawaguchi, S. (1998). Spontaneous regeneration and recovery of hearing function of the central auditory pathway in young rats. Neurosci. Lett. 254, 173–176. Ito, M., Hayashi, T., Kuroiwa, A., and Okamoto, M. (1999). Lens formation by pigmented epithelial cell reaggregate from dorsal iris implanted into limb blastema in the adult newt. Dev. Growth Differ. 41, 429 – 440. Jones, M. K., Tomikawa, M., Mohajer, B., and Tarnawski, A. S. (1999). Gastrointestinal mucosa regeneration: Role of growth factors. Front. Biosci. 4, D303–309. Jung, J., Zheng, M., Goldfarb, M., and Zaret, K. S. (1999). Initiation of mammalian liver development from ectoderm by fibroblast growth factors. Science 284, 1998 –2002. Kim, S. S., Kaihara, S., Benvenuto, M., Choi, R. S., Kim, B.-S., Mooney, D. J., Taylor, G. A., and Vacavti, J. P. (1999). Regenerative signals for tissue-engineered small intestine. Transplant. Proc. 31, 657– 660. Kodama, R., and Eguchi, G. (1995). From lens regeneration in the newt to in vitro transdifferentiation of vertebrate pigmented epithelial cells. Semin. Cell Biol. 6, 143–149. Larson, A. (1991). A molecular perspective on the evolutionary relationships of salamander families. Evol. Biol. 25, 211–277. Lo, D. C., Allen, F., and Brockes, J. P. (1993). Reversal of muscle differentiation during urodele limb regeneration. Proc. Natl. Acad. Sci. USA 90, 7230 –7234. Menke, A., Geerling, I., Giehl, K., Vogelmann, R., Reinhagen, M., and Adler, G. (1999). Transforming growth factor beta-induced upregulation of transforming growth factor beta receptor in pancreatic regeneration. Biochim. Biophys. Acta 1449, 178 –185. Michalopoulos, G. K., and DeFrances, M. (1997). Liver regeneration. Science 276, 60 – 66. Mitashov, V. I. (1996). Mechanisms of retina regeneration in urodeles. Int. J. Dev. Biol. 40, 833– 844. Mochii, M., Mazaki, Y., Mizuno, N., Hayashi, H., and Eguchi, G. (1998). Role of Mitf in differentiation and transdifferentiation of chicken pigmented epithelial cells. Dev. Biol. 134, 201–205. Mullen, L. M., Bryant, S. V., Torok, M. A., Blumberg, B., and Gardiner, D. M. (1996). Nerve dependency of regeneration: The role of distal-less and FGF signaling in amphibia limb regeneration. Development 122, 3487–3497. Niazi, I. A., and Saxena, S. (1978). Abnormal hindlimb regeneration in tadpoles of the toad, Bufo andersoni, exposed to excess vitamin A. Folia Biol. (Krakow) 26, 3–11. Oberpriller, J. O., and Oberpriller, J. C. (1991). Cell division in adult newt cardiac myocytes. In “The Developmental and Regenerative Potential of Cardiac Muscle” (J. O. Oberpriller, J. C. Oberpriller, and A. Mauro, Eds.), pp. 293–311. Harwood, New York. O’Hara, C. M., Egar, M., and Chernoff, E. A. G. (1992). Reorganization of the ependyma during axolotl spinal cord regeneration. Changes in intermediate filaments and fibronectin expression. Dev. Dyn. 193, 103–115. O’Steen, W. K., and Walker, B. E. (1962). Radioautographic studies of regeneration in the common newt. III. Regeneration and repair of the intestine. Anat. Rec. 142, 179 –188. Oliver, G., Loosli, F., Koster, R., Wittbrodt, J., and Gruss, P. (1996). Ectopic lens induction in fish in response to the murine homeobox gene six-3. Mech. Dev. 60, 233–239. Oudega, M., and Hagg, T. (1999). Neurotrophins promote regeneration of sensory axons in the adult rat spinal cord. Brain Res. 818, 431– 438. Park, C. M., and Hollenberg, M. J. (1989). Basic fibroblast growth factor induces retina regeneration in vivo. Dev. Biol. 134, 201–205. Pecorino, L. T., Lo, D. C., and Brockes, J. B. (1994). Isoform-specific induction of a retinoid-specific antigen after biolistic transfection of chimeric retinoic acid/thyroid hormone receptors into a regenerating limb. Development 120, 325–333. Pecorino, L. T., Entwistle, A., and Brockes, J. P. (1996). Activation of a single retinoic acid receptor isoform mediates proximodistal respecification. Curr. Biol. 6, 563–569. Pestronk, A., Watson, D. F., and Yuan, C. M. (1990). Neurofilament phosphorylation in peripheral nerve: Changes with axonal length and growth state. J. Neurochem. 54, 977–982. Petersen, B. E., Bowen, W. C., Patrene, K. D., Mars, W. M., Sullivan, A. K., Murase, N., Boggs, S. S., Greenberger, J. S., and Goff, J. B. (1999). Bone marrow as a potential source of hepatic oval cells. Science 284, 1168 –1170. Pittersen, M. F., Mackay, A. M., Beck, S. C., Jaiswal, R. K., Douglas, R., Mosca, J. D., Moorman, M. A., Simonetti, D. W., Craig, S., and Marshak, D. R. (1999). Multilineage potential of adult mesenchymal cells. Science 284, 143–147. Poulin, M. L., Patrie, K. M., Bethelo, M. J., Tassava, R. A., and Chiu, I-M. (1993). Heterogeneity in the expression of fibroblast growth factor receptor 2 during limb regeneration in the newt (Notophthalmus viridescens). Development 119, 353–361. Powell, A. J., Gates, P. B., Wylie, D., Velloso, C. P., Brockes, J. P., and Jat, P. S. (1998). Immortalization of rat embryo fibroblasts by a 3⬘-untranslated region. Exp. Cell Res. 240, 252–262. Prehn, R. T. (1971). Immunosurveillance, regeneration and cancer. Prog. Exp. Tumor Res. 14, 1–24. Prockop, D. J. (1997). Marrow stromal cells as stem cells for nonhematopoietic tissues. Science 276, 71–74. Reginelli, A. D., Wang, Y.-Q., Sassoon, D., and Muneoka, K. (1995). Digit tip regeneration correlates with regions of Msx1 (Hox 7) expression in fetal and adult mice. Development 121, 1065–1076. Reh, T. A., and Pittack, C. (1995). Transdifferentiation and retinal regeneration. Semin. Cell Biol. 6, 137–142. Sattar, A., Robson, S. C., Patel, H. R., Angus, B., and Campell, F. C. (1999). Expression of growth regulatory genes in a SCID mouse– Copyright © 2000 by Academic Press. All rights of reproduction in any form reserved. 284 Panagiotis A. Tsonis human model of intestinal epithelial regeneration. J. Pathol. 187, 229 –236. Sharma, A., Zangen, D. H., Reitz, P., Taneja, M., Lissauer, M. E., Miller, C. P., Weir, G. C., Habener, J. F., and Bonner-Weir, S. (1999). The homeodomain protein IDX-1 increases after an early burst of proliferation during pancreatic regeneration. Diabetes 48, 507–513. Schilthuis, J. G., Gann, A., and Brockes, J. P. (1993). Chimeric retinoic acid/thyroid hormone receptors implicate RAR-alpha1 as mediating growth inhibition by retinoic acid. EMBO J. 12, 3459 –3466. Stocum, D. L. (1995). “Wound Repair, Regeneration and Artificial Tissues.” Springer-Verlag, Heidelberg. Tam, S. K., Gu, W., Mahdavi, V., and Nadal-Ginard, B. (1995). Cardiac myocyte terminal differentiation. Potential for cardiac regeneration. Ann. N. Y. Acad. Sci. 725, 72–79. Tanaka, E. M., Gann, A. A. F., Gates, P. B., and Brockes, J. P. (1997). Newt myotubes reenter the cell cycle by phosphorylation of the retinoblastoma protein. J. Cell Biol. 136, 155–165. Tanaka, E. M., Drechel, D. N., and Brockes, J. P. (1999). Thrombin regulates S-phase re-entry by cultured newt myotubes. Curr. Biol. 9, 792–799. Tomarev, S. I., Sundin, O., Banerjee-Basu, S., Duncan, M. K., Yang, J. M., and Piatigorsky, J. (1996). Chicken homeobox gene Prox 1 related to Drosophila prospero is expressed in the developing lens and retina. Dev. Dyn. 206, 354 –377. Torok, M. A., Gardiner, D. M., Shubin, N. H., and Bryant, S. V. (1998). Expression of HoxD genes in developing and regenerating axolotl limbs. Dev. Biol. 200, 225–233. Ton, C. C. T., et al. (1991). Positional cloning and characterization of a paired box and homeobox-containing gene from the aniridia region. Cell 67, 1059 –1074. Tsonis, P. A. (1983). Effects of carcinogens on regenerating and non-regenerating limbs in amphibia. Anticancer Res. 3, 195–202. Tsonis, P. A. (1996). “Limb Regeneration.” Cambridge Univ. Press, Cambridge, UK. Tsonis, P. A. (1999). Regeneration of the vertebrate lens and other eye structures. In “Embryonic Encyclopedia of Life Sciences.” Nature Publ., London. [www.els.net] Tsonis, P. A. (2000). Regeneration of the lens in amphibia. In “Vertebrate Eye Development” (E. Fini, Ed.). Springer-Verlag, Heidelberg, in press. Tsonis, P. A., and Del Rio-Tsonis, K. (1988). Spontaneous neoplasms in amphibia. Tumor Biol. 9, 221–224. Tsonis, P. A., Washabaugh, C. H., and Del Rio-Tsonis, K. (1995). Transdifferentiation as a basis for limb regeneration. Semin. Cell Biol. 6, 127–135. Zalik, S. E., and Scott, V. (1973). Sequential disappearance of cell surface components during lens dedifferentiation in lens regeneration. Nat. New Biol. 244, 212–214. Zenjari, C., Boilly, B., Hondermarck, H., and Boilly-Marer, Y. (1997). Nerve– blastema interactions induce fibroblast growth factor-1 release during limb regeneration in Pleurodeles waltl. Dev. Growth Differ. 39, 15–22. Received for publication September 22, 1999 Revised February 12, 2000 Accepted February 12, 2000 Copyright © 2000 by Academic Press. All rights of reproduction in any form reserved.