in biology, the process by which some organisms replace or restore lost or amputated body parts.

Organisms differ markedly in their ability to regenerate parts. Some grow a new structure on the stump of the old one. By such regeneration whole organisms may dramatically replace substantial portions of themselves when they have been cut in two, or may grow organs or appendages that have been lost. Not all living things regenerate parts in this manner, however. The stump of an amputated structure may simply heal over without replacement. This wound healing is itself a kind of regeneration at the tissue level of organization: a cut surface heals over, a bone fracture knits, and cells replace themselves as the need arises.

Regeneration, as one aspect of the general process of growth, is a primary attribute of all living systems. Without it there could be no life, for the very maintenance of an organism depends upon the incessant turnover by which all tissues and organs constantly renew themselves. In some cases rather substantial quantities of tissues are replaced from time to time, as in the successive production of follicles in the ovary or the molting and replacement of hairs and feathers. More commonly, the turnover (mitosis) is expressed at the cellular level. In mammalian skin the epidermal cells produced in the basal layer may take several weeks to reach the outer surface and be sloughed off. In the lining of the intestines, the life span of an individual epithelial cell may be only a few days.

The motile, hairlike cilia and flagella of single-celled organisms are capable of regenerating themselves within an hour or two after amputation. Even in nerve cells, which cannot divide, there is an endless flow of cytoplasm from the cell body out into the nerve fibres themselves. New molecules are continuously being generated and degraded with turnover times measured in minutes or hours in the case of some enzymes, or several weeks as in the case of muscle proteins. (Evidently, the only molecule exempt from this inexorable turnover is deoxyribonucleic acid 【DNA】 which ultimately governs all life processes.)

There is a close correlation between regeneration and generation. The methods by which organisms reproduce themselves have much in common with regenerative processes. Vegetative reproduction, which occurs commonly in plants and occasionally in lower animals, is a process by which whole new organisms may be produced from fractions of parent organisms; e.g., when a new plant develops from a cut portion of another plant, or when certain worms reproduce by splitting in two, each half then growing what was left behind. More commonly, of course, reproduction is achieved sexually by the union of an egg and sperm. Here is a case in which an entire organism develops from a single cell, the fertilized egg, or zygote. This remarkable event, which occurs in all organisms that reproduce sexually, testifies to the universality of regenerative processes. During the course of evolution the regenerative potential has not changed, but only the levels of organization at which it is expressed. If regeneration is an adaptive trait, it would be expected to occur more commonly among organisms that appear to have the greatest need of such a capability, either because the hazard of injury is great or the benefit to be gained is great. The actual distribution of regeneration among living things, however, seems at first glance to be a rather fortuitous one. It is difficult indeed to understand why some flatworms (flatworm) are able to regenerate heads and tails from any level of amputation, while other species can regenerate in only one direction or are unable to regenerate at all. Why do leeches fail to regenerate, while their close relatives, the earthworms, are so facile at replacing lost parts? Certain species of insects regularly grow back missing legs, but many others are totally lacking in this capacity. Virtually all modern bony fishes can regenerate amputated fins, but the cartilaginous fishes (including the sharks and rays) are unable to do so. Among the amphibians, salamanders regularly regenerate their legs, which are not very useful for movement in their aquatic environment, while frogs and toads, which are so much more dependent on their legs, are nevertheless unable to replace them. If natural selection operates on the principle of efficiency, then it is difficult to explain these many inconsistencies.

Some cases are so clearly adaptive that there have evolved not only mechanisms for regeneration, but mechanisms for self-amputation, as if to exploit the regenerative capability. The process of losing a body part spontaneously is called autotomy. The division of a protozoan into two cells and the splitting of a worm into two halves may be regarded as cases of autotomy. Some colonial marine animals called hydroids shed their upper portions periodically. Many insects and crustaceans will spontaneously drop a leg or claw if it is pinched or injured. Lizards are famous for their ability to release their tails. Even the shedding of antlers by deer may be classified as an example of autotomy. In all these cases autotomy occurs at a predetermined point of breakage. It would seem that wherever nature contrives to lose a part voluntarily, it provides the capacity for replacement.

Sometimes, when part of a given tissue or organ is removed, no attempt is made to regenerate the lost structures. Instead, that which remains behind grows larger. Like regeneration, this phenomenon—known as compensatory hypertrophy—can take place only if some portion of the original structure is left to react to the loss. If three-quarters of the human liver is removed, for example, the remaining fraction enlarges to a mass equivalent to the original organ. The missing lobes of the liver are not themselves replaced, but the residual ones grow as large as necessary in order to restore the original function of the organ. Other mammalian organs exhibit similar reactions. The kidney, pancreas, thyroid, adrenal glands, gonads, and lungs compensate in varying degrees for reductions in mass by enlargement of the remaining parts.

It is not invariably necessary for the regenerating tissue to be derived from a remnant of the original tissue. Through a process called metaplasia, one tissue can be converted to another. In the case of lens regeneration in certain amphibians, in response to the loss of the original lens from the eye, a new lens develops from the tissues at the edge of the iris on the upper margin of the pupil. These cells of the iris, which normally contain pigment granules, lose their colour, proliferate rapidly, and collect into a spherical mass which differentiates into a new lens.

Not all organisms regenerate in the same way. In plants and in coelenterates such as the hydra and jellyfishes, missing parts are replaced by reorganization of preexisting ones. The wound is healed, and the neighbouring tissues reorganize themselves into whatever parts may have been cut off. This process of reorganization, called morphallaxis, is the most efficient way for simple organisms to regenerate. Higher animals, with more complex bodies, regenerate parts differently, usually by the production of a specialized bud, or blastema, at the site of amputation. The blastema, made up of cells that look very much alike despite their often diverse origins, made its first appearance evolutionarily in flatworms and is encountered in the regenerative processes of all higher animals. It provides the tissue that will form the regenerated part.

Sometimes the part that grows back is not the same as that which was lost, and, occasionally, regeneration may be induced without having lost anything at all. It is not uncommon for a regenerated part to be incomplete. Earthworms, for example, usually regenerate only five segments in the anterior direction even if more than that number have been amputated. Many insects regenerate abnormally small legs from which some segments may be missing. Tadpole tails when amputated grow back to about only half their original length. These and other cases testify to the fact that a little regeneration is often good enough—that it is not necessary in every case to reproduce a flawless copy of the original.

Sometimes that which is regenerated is very different from the original. Among the arthropods there are cases in which the stump of an antenna grows a leg, while a cut eyestalk regenerates an antenna. More commonly, the regenerated part may be a reasonable facsimile of the original but will differ in details. A regenerated lizard tail contains an unsegmented cartilaginous tube instead of a series of vertebrae as did the original tail. The spinal cord lacks segmented ganglia, and the scales in the skin differ in character from the original ones. A regenerated tail, therefore, is easily distinguished from an original one yet appears sufficient to serve the purpose. Another interesting case is that of jaw regeneration in salamanders. If the lower jaw is amputated a new one will grow back, but it is often smaller than the original. It contains teeth and a mandible, but lacks a new tongue. Furthermore, the new mandible is a cartilaginous model of the original, and is not known to convert into bone.

Sometimes more of a part grows back than has been removed by amputation. A limb stump, for example, can occasionally give rise to hands with extra digits. Lobsters have been known to regenerate double structures, in which case the new parts are mirror images of each other.

Following amputation, an appendage capable of regeneration develops a blastema from tissues in the stump just behind the level of amputation (see photograph). These tissues undergo drastic changes. Their cells, once specialized as muscle, bone, or cartilage, lose the characteristics by which they are normally identified (dedifferentiation); they then begin to migrate toward, and accumulate beneath, the wound epidermis, forming a rounded bud (blastema) that bulges out from the stump. Cells nearest the tip of the bud continue to multiply, while those situated closest to the old tissues of the stump differentiate into muscle or cartilage, depending upon their location. Development continues until the final structures at the tip of the regenerated appendage are differentiated, and all the proliferating cells are used up in the process.

The blastema cells seem to differentiate into the same kind of cells they were before, or into closely related types. Cells may perhaps change their roles under certain conditions, but apparently rarely do so. If a limb blastema is transplanted to the back of the same animal, it may continue its development into a limb. Similarly, a tail blastema transplanted elsewhere on the body will become a tail. Thus, the cells of a blastema seem to bear the indelible stamp of the appendage from which they were produced and into which they are destined to develop. If a tail blastema is transplanted to the stump of a limb, however, the structure that regenerates will be a composite of the two appendages.

Each living thing exhibits polarity, one example of which is the differentiation of an organism into a head, or forward part, and a tail, or hind part. Regenerating parts are no exception; they exhibit polarity by always growing in a distal direction (away from the main part of the body). Among the lower invertebrates, however, the distinction between proximal (near, or toward the body) and distal is not always clear cut. It is not difficult, for example, to reverse the polarity of “stems” in colonial hydroids. Normally a piece of the stem will grow a head end, or hydranth, at its free, or distal, end; if that is tied off, however, it regenerates a hydranth at the end that was originally proximal. The polarity in this system is apparently determined by an activity gradient in such a way that a hydranth regenerates wherever the metabolic rate (metabolism) is highest. Once a hydranth has begun to develop, it inhibits the production of others proximal to it by the diffusion of an inhibitory substance downward along the stem.

When planarian flatworms are cut in half, each piece grows back the end that is missing. Cells in essentially identical regions of the body where the cut was made form blastemas, which, in one case gives rise to a head and in the other becomes a tail. What each blastema regenerates depends entirely on whether it is on a front piece or a hind piece of flatworm: the real difference between the two pieces may be established by metabolic differentials. If a transverse piece of a flatworm is cut very thin—too narrow for an effective metabolic gradient to be set up—it may regenerate two heads, one at either end. If the metabolic activity at the anterior end of a flatworm is artificially reduced by exposure to certain drugs, then the former posterior end of the worm may develop a head.

Appendage regeneration poses a different problem from that of whole organisms. The fin of a fish and the limb of a salamander have proximal and distal ends. By various manipulations, it is possible to make them regenerate in a proximal direction, however. If a square hole is cut in the fin of a fish, regeneration takes place as expected from the inner margin, but may also occur from the distal edge. In the latter case, the regenerating fin is actually a distal structure except that it happens to be growing in a proximal direction.

Amphibian limbs react in a similar manner. It is possible to graft the hand of a newt to the nearby body wall, and once a sufficient blood flow has been established, to sever the arm between the shoulder and elbow. This creates two stumps, a short one consisting of part of the upper arm, and a longer one made up of the rest of the arm protruding in the wrong direction from the side of the animal. Both stumps regenerate the same thing, namely, everything normally lying distal to the level of amputation, regardless of which way the stump was facing. The reversed arm therefore regenerates a mirror image of itself.

Clearly, when a structure regenerates it can only produce parts that normally lie distal to the level of amputation. The participating cells contain information needed to develop everything “downstream,” but can never become more proximal structures. Regeneration, like embryonic development, occurs in a definite sequence.

There are certain prerequisites without which regeneration cannot occur. First and foremost, there must be a wound, although the original appendage need not have been lost in the process. Second, there must be a source of blastema cells derived from remnants of the original structure or an associated one. Finally, regeneration must be stimulated by some external force. The stimuli often involve the nervous system. An adequate nerve supply is required for the regeneration of fish fins, taste barbels, and amphibian limbs. In the case of many tail regenerations, the spinal cord provides the necessary stimulus. Lens regeneration in salamander eyes depends upon the presence of a retina. Arthropod appendages regenerate in the presence of molting hormones. Protozoan regeneration requires the presence of a nucleus. In case after case, regeneration depends on more than a healed wound and a source of blastema cells. It is often triggered by some physiological stimulus originating elsewhere in the body, a stimulus invariably associated with the very function of the structure to be regenerated. The conclusion is inescapable that regeneration is primarily the recovery of deficient functions rather than simply the replacement of lost structures.

The imperative of need is of further importance in suppressing excess regeneration. To be able to regenerate is to run the risk of regenerating too much or too often. If regeneration did not depend upon a physiological stimulus, such as those mediated by nerves or hormones, there would be no reason why simple wounds should not sprout whole new appendages.

It is not known why regeneration fails to occur in many cases, as in the legs of frogs or the limbs and tails of mammals. The nerve supply might be inadequate, for when the number of nerves is artificially increased, regeneration is sometimes induced. This cannot be the whole answer, however, because not all appendages depend on nerves for their regeneration; newt jaws, salamander gills, and deer antlers do not require nerves to regenerate.

Possibly the failure to regenerate relates to the ways in which wounds heal. In higher vertebrates there is a tendency to form thick scar tissue in healing wounds, which may act as a barrier between the epidermis and the underlying tissues of the stump. In the absence of direct contact between these two tissues, the stump may not be able to give rise to the blastema cells required for regeneration.

Virtually no group of organisms lacks the ability to regenerate something. This process, however, is developed to a remarkable degree in lower organisms, such as protists and plants, and even in many invertebrate animals such as earthworms and starfishes. Regeneration is much more restricted in higher organisms such as mammals, in which it is probably incompatible with the evolution of other body features of greater survival value to these complex animals.

One of the most outstanding feats of regeneration occurs in the single-celled green alga Acetabularia. This plant-like protist of shallow tropical water consists of a group of short rootlike appendages; a long thin “stem,” up to several centimetres in length; and an umbrella-like cap at the top. The entire organism is one cell, with its single nucleus situated at the base in one of the “roots.” If the cap is cut off, a new one regenerates from the healed over stump of the amputated stem. The nucleus is necessary for this kind of regeneration, presumably because it provides the information needed to direct the development of the new cap. Once this information has been produced by the nucleus, however, the nucleus can be removed and regeneration continues unabated.

If the nucleus from one species of Acetabularia is added to a cell-body of another species, and the cap of the recipient cell is amputated, the new cap that regenerates will be a hybrid because each nucleus exerts its own morphogenetic influences. On the other hand, if the nucleus from one species is substituted for that in another, regeneration reflects the properties of the new nucleus.

Most single-celled, animal-like protists regenerate very well. If part of the cell fluid, or cytoplasm, is removed from Amoeba, it is readily replaced. A similar process occurs in other protozoans, such as flagellates and ciliates. In each case, however, regeneration occurs only from that fragment of the cell containing the nucleus. Amputated parts that lack a nucleus cannot survive. In some ciliates, such as Blepharisma or Stentor, the nucleus may be elongated or shaped like a string of beads. If either of these organisms is cut in two so that each fragment retains part of the elongated nucleus, each half proceeds to grow back what it lacks, giving rise to a complete organism in less than six hours. The way in which such a bisected protozoan regenerates is almost identical with the way it reproduces by ordinary division. Even a very tiny fragment of the whole organism can regenerate itself, provided it contains some nuclear material to determine what is supposed to be regenerated.

The mechanisms by which vascular plants grow have much in common with regeneration. Their roots and shoots elongate by virtue of the cells in their meristems, the conical growth buds at the tip of each branch. These meristems are capable of indefinite growth, especially in perennial plants. If they are amputated they are not replaced, but other meristems along the stem, normally held in abeyance, begin to sprout into new branches that more than compensate for the loss of the original one. Such a process is called restitution.

Plants are also capable of producing callus tissue wherever they may be injured. This callus is proliferated from cambial cells, which lie beneath the surface of branches and are responsible for their increase in width. When a callus forms, some of its cells may organize into growing points, some of which in turn give rise to roots while others produce stems and leaves.

The vast majority of research on coelenterates has been focussed on hydras (Hydra) and some of the colonial hydroids. If a hydra is cut in half, the head end reconstitutes a new foot, while the basal portion regenerates a new hydranth with mouth and tentacles. This seemingly straightforward process is deceptively simple. From tiny fragments of the organism whole animals can be reconstituted. Even if a hydra is minced and the pieces scrambled, the fragments grow together and reorganize themselves into a complete whole. The indestructibility of the hydra may well be attributed to the fact that even the intact animal is constantly regenerating itself. Just below the mouth is a growth zone from which cells migrate into the tentacles and to the foot where they eventually die. Hence, the hydra is in a ceaseless state of turnover, with the loss of cells at the foot and at the tips of the tentacles being balanced by the production of new ones in the growth zone. If such an animal is X-rayed, the proliferation of new cells is inhibited and the hydra gradually shrinks and eventually dies owing to the inexorable demise of cells and the inability to replace them.

In colonial hydroids, such as Tubularia, there is a series of branching stems, each of which bears a hydranth on its end. If these hydranths are amputated they grow back within a few days. In fact, the organism normally sheds its hydranths from time to time and regenerates new ones naturally.

Planarian flatworms are well-known for their ability to regenerate heads and tails from cut ends. In the case of head regeneration, some blastema cells become brain tissues, others develop into the eyes, and still others differentiate as muscle or intestine. In a week or so, the new head functions almost as well as the original.

The blastema that normally gives rise to a single head is, under certain circumstances, even capable of becoming two heads if the stump of a decapitated flatworm is divided in two by a longitudinal cut. Each of the two halves then gives rise to a complete head. Thus, each blastema develops into an entire structure regardless of its size or position in relation to the rest of the animal.

In the case of flatworms there is still considerable disagreement concerning the origins of the blastema. Some investigators contend that it is derived from neoblasts, undifferentiated reserve cells scattered throughout the body. Others claim that there are no such reserve cells and that the blastema develops from formerly specialized cells near the wound that dedifferentiate to give rise to the blastema cells. Whatever their source, the cells of the blastema are capable of becoming many different things depending upon their location.

Regeneration in flatworms occurs in a stepwise fashion. The first tissue to differentiate is the brain, which induces the development of eyes. Once the head has formed, it in turn stimulates the production of the pharynx. The latter then induces the development of reproductive organs farther back. Thus, each part is necessary for the successful development of those to come after it; conversely, each part inhibits the production of more of itself. If decapitated flatworms are exposed to extracts of heads, the regeneration of their own heads is prevented. Such a complex interplay of stimulators and inhibitors is responsible for the successful regeneration of an integrated morphological structure.

The segmented worms exhibit variable degrees of regeneration. The leeches, as already noted, are wholly lacking in the ability to replace lost segments, whereas the earthworms and various marine annelids (polychaetes) can often regenerate forward and backward. The expression of such regenerative capacities depends very much on the level of amputation. Anteriorly directed regeneration usually occurs best from cuts made through the front end of the worm, with little or no growth taking place from progressively more posterior bisections. Posteriorly directed regeneration is generally more common and extensive. Some species of worms replace the same number of segments as were lost. Hypomeric regeneration, in which fewer segments are produced than were removed, is more common, however.

Anterior regeneration depends upon the presence of the central nerve cord. If this is cut or deflected from the wound surface, little or no forward regeneration may take place. Posterior regeneration requires the presence of the intestine, removal of which precludes the formation of hind segments. Thus, it would seem that no head will regenerate without a central nervous system, nor a tail without an opening.

Many insects (insect) and crustaceans (crustacean) regenerate legs, claws, or antennas with apparent ease. When insect legs regenerate, the new growth is not visible externally because it develops within the next proximal segment in the stump. Not until the following molt is it released from its confinement to unfold as a fully developed leg only slightly smaller than the original. In the case of crabs, regenerating legs bulge outward from the amputation stump. They are curled up within a cuticular sheath, not to be extended until the sheath is molted. Lobsters and crayfish regenerate claws and legs in a straightforward manner as direct outgrowths from the stumps. As in other crustaceans, however, these regenerates lie immobile within an enveloping cuticle and do not become functional until their sheath is shed at the next molt.

In all arthropods regeneration is associated with molting, and therefore takes place only during larval or young stages. Most insects do not initiate leg regeneration unless there remains ample time prior to the next scheduled molt for the new leg to complete its development. If amputation is performed too late in the intermolt period, the onset of regeneration is delayed until after shedding; the regenerate then does not appear until the second molt. Metamorphosis into the adult stage marks the end of molting in insects, and adults accordingly do not regenerate amputated appendages.

Crustaceans often tend to molt and grow throughout life. They therefore never lose the ability to grow back missing appendages. When a leg is lost, a new outgrowth appears even if the animal is not destined to molt for many months. Following a period of basal growth, during which a diminutive limb is produced, the regenerated part eventually ceases to elongate. Not until a few weeks before the next molt does it resume growth and complete its development, triggered by the hormones that induce molting.

Many different parts of the fish's body will grow back. Plucked scales are promptly replaced by new ones, and amputated gill filaments can regenerate easily. The “whiskers,” or taste barbels, of the catfish grow back as perfect replicas of the originals. The most conspicuous regenerating structures in fishes, however, are the fins. When any of these are amputated, new fins grow out from the stumps and soon restore everything that was missing. Even the coloured stripes or spots that adorn some fins are reconstituted by new pigment cells that repopulate the regenerated part. Fin regeneration depends on an adequate nerve supply. If the nerves are cut leading into the fin, regeneration of neither the amputated fin nor excised pieces of the bony fin rays can take place.

Salamanders are remarkable for their ability to regenerate limbs. Larval frogs, or tadpoles (tadpole), also possess this ability, but usually lose it when they become frogs. It is not known why frog legs do not regenerate, and under appropriate stimuli they can be induced to do so.

Tadpoles and salamanders can replace amputated tails. Tadpole tails have a stiff rod called the notochord for support, whereas salamanders possess a backbone, composed of vertebrae. Both tails contain a spinal cord. When the salamander regenerates its tail, the spinal cord grows back and segmental nerve-cell clusters (ganglia) differentiate. Tadpoles also regenerate their spinal cords, but not the associated ganglia. If the spinal cord is removed or destroyed in the salamander, no tail regeneration occurs; if it is removed from the tadpole tail, however, regeneration can proceed without it.

Lizards (lizard) also regenerate their tails, especially in those species that have evolved a mechanism for breaking off the original tail when it is grasped by an enemy. When the lizard tail regenerates, however, it does not replace the segmented vertebrae. Instead, there develops a long tapering cartilaginous tube within which the spinal cord is located and outside of which are segmented muscles. The spinal cord of the lizard tail is necessary for regeneration, but the regenerated tail does not reproduce the ganglia that are normally associated with it. Occasionally, a side tail may be produced if the original tail is broken but not lost.

Regeneration of amputated appendages in birds is not known to occur; however, they do replace their feathers as a matter of course. While most species shed and regenerate feathers one at a time so as not to be grounded, flightless birds, such as penguins, may molt them all at once. Male puffins cast off their colorful beaks after the mating season, but grow new ones the following year. In like manner, the dorsal keel on the upper beaks of male pelicans is shed and replaced annually.

Although mammals are incapable of regenerating limbs and tails, there are a few exceptional cases in which lost tissues are in fact regenerated. Not the least of these cases is the annual replacement of antlers in deer. These remarkable structures, which normally grow on the heads of male deer, consist of an inner core of bone enveloped by a layer of skin and nourished by a copious blood supply. During the growing season the antlers elongate by the proliferation of tissues at their growing tips. The rate of growth in some of the larger species may surpass one centimetre (0.39 inch) per day; the maximum rate of growth recorded for the elk is 2.75 centimetres (1.05 inches) per day. When the antlers have reached their full extent, the blood supply is constricted, and the skin, or velvet, peels off, thus revealing the hard, dead, bony antlers produced by the male deer in time for the autumn mating season. The regeneration of elk antlers spans about seven months. The following spring, the old antlers are shed and new ones grow to replace them.

Still another example of mammalian regeneration occurs in the case of the rabbit's ear. When a hole is punched through the external ear of the rabbit, tissue grows in from around the edges until the original opening is reduced or obliterated altogether. This regeneration is achieved by the production of new skin and cartilage from the margins of the original hole. A similar phenomenon occurs in the case of the bat's wing membrane.