Introductory Review Cellscience Reviews Vol 1 No.2 ISSN 1742-8130

Loss of Regenerative Capacity: A Trade-off for Immune Specificity?

A.L. Mescher & A.W. Neff

Introduction

Known since the beginnings of experimental biology, the ability of animals to regenerate parts of their bodies after loss or injury has long held a special fascination for naturalists and scientists interested in animal development. To actually regrow a lost organ or other structure rather than simply fill the void with scar tissue involves processes ranging from an injury response and wound healing to growth, patterning and differentiation of new tissues similar to that which occurred during embryonic development. With new evidence for the presence of stem cells in most if not all adult organs and their ability to participate in tissue repair, the field of regenerative biology has assumed much more widespread medical relevance (Stocum, 1995, 2004). An increasing number of medical scientists now look at striking examples of organ regeneration in invertebrates and lower vertebrates and ask why we and other mammals lack such an apparently advantageous trait.

If one looks at the phylogenetic distribution of regenerative ability in various organ systems, it appears that this capacity has been lost gradually in the course of animal evolution (Thouveny and Tassava, 1998; Sanchez Alvarado, 2000). General questions relating to the loss and sporadic distribution of regeneration during metazoan phylogeny have been recently considered by others (Alvarado, 2000; Brockes et al., 2001). We have focused on regeneration of amputated limbs in amphibians, one of the best-studied model systems and a useful paradigm for understanding many features of vertebrate organ regeneration. As an example of “epimorphic” regeneration, this system includes cellular dedifferentiation in the injured tissues of the limb stump and proliferation of these cells to form a distal blastema which undergoes patterning and growth to restore the missing limb structures. This review will provide a brief introduction to the various theories regarding the loss of regenerative ability in vertebrate limbs and indicate a promising new approach to this question revealed by recent analyses of gene expression during regeneration. The new data suggest that components of the immune response produce reactions to injury that inhibit tissue or organ regeneration, a hypothesis consistent with the evolution of the innate and adaptive immune systems and with the loss of regenerative capacity during mammalian ontogeny.

Why are amphibians the only tetrapods able to regenerate limbs?

The question of why limbs of phylogenetically advanced vertebrates fail to regenerate has been addressed by studies with limbs of anuran amphibians. Regeneration is excellent in the earliest stages of limb development in anurans (frogs and toads) but gradually diminishes as larvae approach metamorphosis (Dent, 1962). Limbs of adult anurans are incapable of complete regeneration. However, urodele amphibians (newts and salamanders) commonly regenerate limbs and often other organs such as tails, jaws, and parts of the eye throughout their lifetimes. Such regenerative phenomena are very rare among reptiles, birds, and mammals as adults, suggesting that the loss of regenerative capacity may have been an adaptive part of the evolutionary transition toward the more advanced tetrapods.

Studies comparing various aspects of regeneration-incompetent limbs, such as those of mammals, reptiles, or adult frogs, to regenerating larval frog or urodele limbs has led to several hypotheses for the loss of regenerative capacity. Early in the last century several investigators suggested that cells of differentiating muscle and other tissues in the anuran limb lose their ability to revert to the proliferative state and contribute to limb regrowth. Consistent with this view, regeneration and morphogenesis were found to be enhanced in limbs of adult frogs when tissue dissociation and cellular dedifferentiation were increased in stump tissues by additional trauma (Polezhaev, 1972). The plasticity of the differentiated state in regeneration-competent limbs and the potential of multinucleate muscle fibers to dedifferentiate and re-enter the cell cycle are currently active areas of investigation within the field of limb regeneration (Brockes et al., 2001; Brockes and Kumar, 2002).

Because regeneration requires epithelial-mesenchymal interactions at the distal limb stump like those that drive embryonic limb development, the changing nature of wound closure after amputation of anuran limbs during the transition from larvae to adults has also been studied. Closure of limb stumps in mammals involves contraction of full-thickness skin and in adult frogs involves rapid formation of connective tissue beneath the apical wound epidermis that initially covers the cut surface (Carlson, 1974). Tassava and Olsen (1982) suggested that the inability of higher vertebrates to form a functional wound epithelium explains the lost potential for regeneration. Interfering with distal scar formation in amputated limbs of mammals or adult frogs in order to elicit regeneration have at best been only marginally successful (see review by Stocum, 1996), but the importance of establishing proper conditions for the reciprocal interactions between the apical epithelium and the underlying mesodermal cells is clear if a limb is to regenerate.

In embryonic limbs signalling between the apical ectoderm and the adjacent mesodermal cells fibroblast growth factors (FGFs) and their receptors. Expression of specific FGFs such as FGF-10 in limbs tissues of the anuran Xenopus at regeneration-competent but not -incompetent stages has been documented (Christen and Slack, 1997; Yokoyama et al., 2000). Moreover, Yokayama et al. (2001) reported that treatment of Xenopus limb stumps with beads containing FGF-10 not only rapidly up-regulated FGF expression in both the apical epithelium and the underlying mesenchyme, but also was capable of enhancing regeneration at a stage when regenerative capacity is normally diminished. Slack et al. (2004) were unable to clearly replicate this result with FGF-10, perhaps because of the normal variability in regeneration at such intermediate stages (Dent, 1962), but overall the work to date affirms the importance of maintaining the functional signalling center between the mesenchyme and the apical wound epithelium for a limb stump to regenerate.

As an explanation for the failure of limbs of higher vertebrates to re-establish functional tissue interactions Galis et al. (2003) have suggested that limb regeneration is only possible when the limb develops as a semiautonomous module not dependent on interactions with transient structures such as somites. These authors point out that in reptiles, birds, and mammals limb development begins in the early embryo and involves signalling interactions with various temporary neighboring structures, while in amphibians limb development occurs much later in development and is not coupled to interactions with transient structures. This developmental difference in amphibians is seen by Galis et al. (2003) as important for their capacity to regenerate limbs as larvae, although it does not help explain the loss of this capacity in limbs of adult anurans but not urodeles. Whether this hypothesis will lead to productive experiments and new information about regeneration remains to be seen.

An adequate nerve supply in the limb stump is necessary for regenerative growth and Rzehak and Singer (1966) suggested nonregenerating species or stages no longer maintained a density of nerve fibers in limb tissues to sustain regenerative growth. This idea was consistent with earlier work by Singer (1954) that suggested improved regeneration in adult frog limbs with augmented nerve supplies. However subsequent analyses of nerve density in limbs from a range of regenerating and nonregenerating species showed no correlation between nerve quantity and regenerative ability (Scadding, 1982).

Changing hormonal levels, particularly levels of thyroid hormones and glucocorticoids, during metamorphosis in anuran present numerous additional opportunities to change regenerative potential of tissues and early studies reviewed by Schotté (1961) and Schmidt (1968) tested the dependence of regeneration on these and other endocrine factors. It became apparent from such studies that regenerating limbs have no special hormonal dependence and that systemic factors in general may play no role in regeneration apart from nonspecifically permitting growth. Rather, the endocrine studies and other work involving tissue grafting indicate clearly that regenerative ability is an intrinsic property of cells within the limb stump itself interacting with factors released in a paracrine fashion from nerves (Muneoka et al., 1989).

One aspect of regeneration that has attracted little attention from developmental biologists is the role of factors and cells from the immune system. Although studies relevant to this topic were reviewed by Sicard (1985), newer work in comparative immunology and the field of immunotolerance has prompted reconsideration of how the immune system might affect regenerative ability (Harty et al., 2003). This work has allowed new understanding of differences in the immune systems of urodeles and anurans, has demonstrated important immunological changes during the larval development in Xenopus, and shown the importance of local immune interactions for suppression of immune cell activity and tolerance of cells with “non-self” features that would otherwise be marked for removal. These aspects of immunology and their relevance to regenerative biology are reviewed in more detail elsewhere (Harty et al., 2003; Mescher and Neff, 2005).

Considerations of amphibian immunology and the well-known importance of specific immune cells for both tissue repair and the fibrotic activity that characterizes nonregenerating limb stumps suggest another hypothesis for regenerative failure in higher vertebrates. Development of adaptive immunity, which supplements more general and primitive innate immune mechanisms and allows an organism to acquire highly specific defense mechanisms against invading microorganisms, may have yielded immune cells and cytokines whose activity in traumatized tissue is inimical to cell dedifferentiation or the signalling required to initiate limb regeneration, so that the response to injury in the presence of such immunity is dominated by tissue repair and fibrosis rather than regeneration (Mescher and Neff, 2005).

Immunoregulatory genes expressed in regenerating Xenopus limbs

We have undertaken subtractive hybridization studies to identify genes expressed in blastemas from Xenopus hindlimbs at a regeneration-competent developmental stage and pseudoblastemas from older regeneration-incompetent limbs (King et al., 2003 and unpublished data). Of the thousands of genes sequenced and analyzed in these studies, many are involved in regulating adaptive immunity in mammalian systems. A sampling of these genes is discussed here.

Transforming growth factor-ß5 (TGF- ß5), a member of this signalling factor family found in amphibians, was expressed in regenerating blastemas subtracted against either pseudoblastemas or control limbs (King et al., 2003). Although TGF- ß5 may have various activities in the blastema, its expression and roles in regeneration have not yet been investigated. Another isoform, Xenopus TGF- ß2, was expressed in nonregenerating pseudoblastemas (King et al., 2003). All three mammalian isoforms of TGF- ß are released locally from various cells at sites of injury and are of central importance in the control of fibrosis and scarring during mammalian tissue repair. Studies by Ferguson have shown clearly that manipulation of specific TGF- ß isoforms is capable of producing scar-free healing of wounds in mice (Ferguson and O’Kane, 2004). Recent work has shown TGF-ß1 to be a potent immunoregulatory cytokine involved in suppression of inflammation and regulatory T cell activity, resulting in immune tolerance (Chen and Wahl, 2003). Functional comparisons between Xenopus and mammalian TGF- ß isoforms based on sequence analyses of the genes and their promoters have been inconclusive (Goswami et al., 2003), but immunosuppressive effects of Xenopus TGF- ß5 are predicted based on other work (Wieczorek et al., 1995). The work with wound healing and immunosuppression in mammals strongly suggests that differential activity of TGF- ß in regenerating amphibian limb stumps may be involved suppression of fibrosis and establishing conditions permissive for blastema formation.

Two forms of pro-opiomelanocortin (POMC), a 31kD precursor for a-melanocyte stimulating hormone (α-MSH), endorphins, and several other peptide hormones, were also expressed in regeneration blastemas (King et al., 2003). Expressed in skin as well as brain, pituitary, and other organs, POMC is a central importance in modulating immune activity within skin, primarily due to the activity of α-MSH (Luger et al., 1999). Paracrine release of α-MSH peptides exerts a potent immunomodulatory effect on immune cells. α-MSH inhibits all forms of inflammation against which it has been tested (Lipton et al., 1997) and localized production of α-MSH helps maintain “immune privilege” at specific cutaneous sites (Paus et al., 2003). Similar expression of α-MSH cells of the blastema would be expected to confer an anti-inflammatory effect potentially important for inhibiting fibrosis and regeneration.

Thymosin-ß4 is a 43 amino acid polypeptide originally described as a thymic maturation factor that has also been shown to promote angiogenesis, keratinocyte migration and wound healing (Malinda et al., 1999). Secreted by macrophages and T lymphocytes of skin, gut and other organs in addition to the thymus, thymosin-ß4 exerts potent anti-inflammatory activity in various assays (Young et al., 1999; Girardi et al., 2003). Expression of thymosin-ß4 is up-regulated in Xenopus pseudoblastemas (King et al., 2003) and regenerating blastemas (unpublished). Activities of thymosin-ß4 in tissues of amputated limbs may include immunomodulation of the inflammatory response in addition to stimulation of epithelial migration and other aspects of regeneration.

King et al., (2003) also reported that blastema minus pseudoblastema cDNA included sequences encoding the P2X7 receptor (P2X7R), a purinergic channel activated by high concentrations of extracellular ATP that plays a pivotal role in modulating several processes in immune and inflammatory responses (Labasi et al., 2002). P2X7R is expressed mainly by macrophages and lymphocytes and regulates secretion of the proinflammatory cytokine interleukin-1ß in response to locally elevated ATP levels released from damaged cells (Labasi et al., 2002). Another lymphocyte surface protein, called “regeneration and tolerance factor,” has recently been shown to modulate IL-ß1 secretion by regulating P2X7R (Derks and Beaman, 2004). Further studies of this receptor’s expression in regeneration blastema may thus yield insights into modulation of the inflammatory response in limbs capable of regeneration.

Evidence from in other regenerating systems

The hypothesis that the origin of adaptive immunity during evolution led to the restriction of regenerative ability is consistent with our knowledge of immune phylogeny (Flajnik et al., 2003). Invertebrates, which usually have well-developed capacities for regeneration, completely lack adaptive immunity. They rely instead on an array of defenses that constitute an extremely effective innate immune system. Mechanisms underlying acquired or adaptive immunity first appear in jawed vertebrates, becoming more efficient in various orders of fish and amphibians and highly developed in the homeotherms (Flajnik et al., 2003). Key changes are noted within the class Amphibia. Urodeles, which show anatomic features more primitive than those of anurans and which regenerate body parts well, are immunodeficient compared to anurans, having only an IgM-based humoral immunity with poor memory response and a very weak cellular immunity with slow allograft rejection (Flajnik et al., 2003; Robert and Cohen, 1998). It is now also known that anurans like Xenopus rely primarily on innate immunity through most of larval development when regenerative ability is highest, and only develop important components of the adaptive immune system, such as major histocompatibility complex proteins, in late larval stages as the approach metamorphosis. This new understanding of the ontogeny of adaptive immunity in Xenopus makes clear that this important system develops during the same period that regenerative ability in the hindlimbs is being lost.

Tsonis and others have shown that at least two components of the complement system, the highly conserved proteolytic cascade which is an important part of nonspecific innate immunity (Flajnik et al., 2003), are expressed in regenerating newt limbs and lenses, including the key protein C3 (Kimura et al., 2003). Binding of a C3 fragment to APCs has also recently been shown to be required for antigen-specific tolerance (Sohn et al., 2003). Moreover a major regulator of complement activity, the cell surface protein CD59, is also expressed by mesenchymal cells of newt limb blastemas with a differential proximal-distal distribution (Morais da Silva et al., 2002). This work indicates that complement may also contribute to a tolerant immune environment in the regeneration blastema and suggests that differential or patterned regulation of this activity may also be involved in the subsequent development of the regenerating organ.

Other work has shown a correlation between regenerative ability and defective immunity in mammalian models of regeneration. The ability of mammalian embryos to heal skin wounds in a scar-free manner is lost during the transition to the fetal stage, which corresponds to the period in which cells involved in adaptive immunity appear (Martin, 1997). Mice genetically deficient for macrophages and certain APCs retain the capacity for scarless wound healing (Martin et al., 2003). Scar-free wound healing and the capacity to regenerate other organs has also been reported for mice of the MRL strain which have immune defects and are used as a model for autoimmune diseases (Heber-Katz et al., 2004).

Summary and conclusions

As this brief review indicates there is considerable evidence from comparative immunology and from new work in several models of tissue and organ regeneration that phylogenetic development of the vertebrate immune system led to physiological responses to injury that blocked the progression of events required for regeneration. We have found that cells of blastemas express several immunosuppressive factors not expressed in nonregenerative pseudoblastemas, including TGF-ß5. Of these TGF-ß is the best-studied with regard to its ability to suppress adaptive immune responses in all species examined (Flajnik et al., 2003). Moreover, TGF- ß isoforms have central roles in fibroblast activation and collagen deposition during fibrosis (Wynn, 2004). Indeed studies of immune cell interactions within the TH1/ TH2 paradigm have clearly shown that specific cytokines released from various CD4+ T cell (T-helper cells) determine the nature of both the immune response to pathogens or cell injury, as well as the fibrotic or regenerative response of the injured tissues (Wynn, 2004).

New work in tissue repair and regeneration indicates that suppression of inflammation, regulation of the immune response, and generation of a tolerant or immune-privileged site may be key determinants of the quality or success of the regenerative response. Our studies of gene expression in Xenopus limb stumps, together with similar work by others in other amphibian systems, identify factors and signalling pathways whose activity is consistent with this hypothesis. How the failure to suppress components of immune reactions aborts regeneration is not clear. Potential mechanisms include allorecognition and removal of dedifferentiating cells by natural killer or T cells as well as cytokine stimulation of excessive fibroblast proliferation and collagen production, disrupting the activity of signalling centers required to form a normal blastema. Study of immune regulation in classic models of regeneration is likely to yield new insights into dedifferentiation and early phase of regeneration and a much greater understanding of the loss of regenerative capacity during phylogeny and ontogeny of higher vertebrates.

REFERENCES

Brockes JP, Kumar A (2002) Plasticity and reprogramming of differentiated cells in amphibian regeneration. Nature Rev. 3:566-574.

Brockes JP, Kumar A, Velloso CPO (2001) Regeneration as an evolutionary variable J. Anat. 199:3-11.

Carlson BM (1974) Factors controlling the initiation and cessation of early events in the regenerative process. In: Neoplasia and Cell Differentiation. Basel: Karger, pp. 60-105.

Chen W, Wahl SM (2003) TGF-ß: the missing link in CD4+CD25+ regulatory T cell-mediated immunosuppression. Cytokine Growth Fac. Rev. 14:85-89.

Christen B, Slack JMW (1997) FGF8 is associated with anteroposterior patterning and limb regeneration in Xenopus. Devel. Biol. 192:455-466.

Dent JN (1962) Limb regeneration in larvae and metamorphosing individuals of the South African clawed toad. J Morphol. 110:61-77.

Derks R, Beaman K (2004) Regeneration and tolerance factor modulates the effect of adenosine triphosphate-induced interleukin 1ß secretion in human macrophages. Human Immunol. 65:676-682.

Ferguson MWJ, O’Kane S (2004) Scar-free healing: from embryonic mechanisms to adult therapeutic intervention. Phil Trans. R Soc. Lon B 359:839-850.

Flajnik MF, Miler K, Du Pasquier L (2003) Evolution of the immune system. In: Paul WE ed. Fundamental Immunology (5th ed.) Philadelphia: Lippincott Williams and Wilkins, pp. 519-570.

Galis F, Wagner GP, Jockusch EL (2003) Why is limb regeneration possible in amphibians but not in reptiles, birds, or mammals? Evol. Develop. 5:208-220.

Girardi M, Sherling MA, Filler RB, Shires J, Theodoris E, Hayday AC, Tigelaar RE (2003) Anti-inflammatory effects in the skin of thymosin-ß4 splice-variants. Immunol 109:1-7.

Goswami MT, Desai KV, Kondaiah P (2003) Comparative functional analysis of rat TGF-ß1 and Xenopus laevis TGF-ß5 promoters suggest differential regulations. J. Mol. Evol. 57:44-51.

Harty M, Neff AW, King MW, Mescher AL (2003) Regeneration or scarring: An immunologic perspective. Devel. Dynam. 226:268-279.

Heber-Katz E, Leferovich JM, Bedelbaeva K, Gourevitch D. (2004) Spallazani’s mouse: a model of restoration and regeneration. Current Top. Microbiol. Immunol. 280:165-189.

Kimura Y, Madhavan M, Call MK, Santiago W, Tsonis PA, Lambris JD, Del Rio-Tsonis K (2003) J Immunol. 170:2331-2339.

Labasi JM, Petrushova N, Donovan C, McCurdy S, Lira P, Payette MM, Brissette W, Wicks, JR, Audoly L, Gabel CA (2002) Absence of P2X7 receptor alters leukocyte function and attenuates an inflammatory response. J Immunol. 168:6436-6445.

Lipton JM, Catania A Anti-inflammatory actions of the neuroimmunomodulator α-MSH. Immunol Today 18:140-145.

Luger TA, Paus R, Lipton JM, Slominski AT, eds. (1999) Cutaneous Neuroimmunomodulation: The Proopiomelanocortin System. NY Acad. Sci., Vol. 885.

Malinda KM, Sidhu GS, Mani H, Banaudha K, Maheshwari RK, Goldstein AL (1999) Thymosin-ß4 accelerates wound healing. J Invest Dermatol 113:364-368.

Martin P (1997) Wound healing – aiming for perfect skin regeneration. Science 276:75-81.

Martin P, D’Souza M, Martin J, Grose R, Maki R, McKercher SR. (2003) Wound healing in the PU.1 null mouse – tissue repair is not dependent on inflammatory cells. Current Biol. 13:1122-1128.

Mescher AL, Neff AW (2005) Regenerative capacity and the developing immune system. In Regenerative Medicine I – Theories, Models and Methods (Yannas I ed.), Adv Biochem Eng Biotechnol 93 (in press)

Morais da Silva S, Gates PB, Brockes JP (2002) The newt ortholog of CD59 is implicated in proximodistal identity during amphibian limb regeneration. Devel. Cell 3:547-555.

Muneoka K, Bryant SV, Gardiner DM (1989) Growth control in limb regeneration. In: Armstrong JB, Malacinski GM eds. Developmental Biology of the Axolotl. NY: Oxford Univ. Press, pp. 143-156.

Paus R, Ito N, Takigawa M, Ito T (2003) The hair follicle and immune privilege. J. Invest. Dermatol. Symp. Proc. 8:188-194.

Polezhaev LV (1972) Loss and Restoration of Regenerative Capacity in Tissues and Organs of Animals. Cambridge: Harvard Univ. Press.

Robert J, Cohen N (1998) Evolution of immune surveillance and tumor immunity: studies in Xenopus. Immunol. Rev. 166:231-243.

Rzehak K, Singer M (1966) Limb regeneration and nerve fiber number in Rana sylvatica and Xenopus laevis. J. Exp. Zool. 162:15-22.

Sánchez Alvarado A (2000) Regeneration in the metazoans: Why does it happen? Bioessays 22:578-590.

Scadding SR (1982) Can differences in regenerative ability between amphibian species be explained by differences in quantity of innervation.? J. Exp. Zool. 219:81-85.

Schmidt AJ (1968) Cellular Biology of Vertebrate Regeneration and Repair. Chicago: Univ. Chicago Press.

Schotté OE (1961) Systemic factors in initiation of regenerative processes in limbs of larval and adult amphibians. In: Rudnick D ed. Synthesis of Molecular and Cellular Structure. NY: Ronald Press, pp. 161-192.

Sicard RE (1985) Leukocytic and immunological influence on regeneration of amphibian hindlimbs. In: Sicard RE ed. Regulation of Vertebrate Limb Regeneration. NY: Oxford Univ. Press, pp. 128-145.

Slack JMW, Beck CW, Gargioli C, Christen B (2004) Cellular and molecular mechanisms of regeneration in Xenopus. Phil Trans R Soc Lon B 359:745-751.

Sohn JH, Bora PS, Suk HJ, Molina H, Kaplan HJ, Bora NS (2003) Tolerance is dependent on complement C3 fragment iC3b binding to antigen-presenting cells. Nature Med. 9:206-212.

Stocum DL (1995) Wound Repair, Regeneration, and Artificial Tissues. Austin TX: RG Landes Co.

Stocum DL (2004) Tissue restoration through regenerative biology and medicine. Advances Anat Embryol Cell Biol. 176:1-101

Tassava RA, Olsen CL (1982) Higher vertebrates do not regenerate digits and legs because the wound epidermis is not functional: A hypothesis. Differen. 22:151-155.

Thouveny Y, Tassava RA (1998) Regeneration through phylogenesis. In: Ferretti P, Géraudie, eds. Cellular and Molecular Basis of Regeneration: From Invertebrates to Humans. Chichester, UK: John Wiley & Sons, pp. 9-44.

Wieczorek Z, Slon J, Kluczyk A, Zbozien R, Stafanowitz P, Siemion IZ (1995) The immunomodulatory diversity of the proteins of the transforming growth factor ß (TGFß) family. Int J Pept Protein Res 46:113-118.

Wynn TA (2004) Fibrotic disease and the TH1/ TH2 paradigm. Nature Rev. Immunol. 4:583-594.

Yokoyama H, Yonei-Tamura S, Endo T, Belmonte JCI, Tamura K, Ide H (2000) Mesenchyme with fgf-10 expression is responsible for regenerative capacity in Xenopus limb buds. Devel. Biol. 219:18-29.

Yokoyama H, Ide H, Tamura K (2001) FGF-10 stimulates limb regeneration ability in Xenopus laevis. Devel. Biol. 233:72-79.

Young JD, Lawrence AJ, MacLean AG, Leung BP, McInmnes IB, Canas B, Pappin DJ, Stevenson RD (1999) Thymosin-ß4 sulfoxide is an anti-inflammatory agent generated by monocytes in the presence of glucocorticoids. Nature Med 5:1424-1427.