760 Nature Vol. 286 21 August 1980 The source of cells for regeneration from J. M. W. Slack A solution to the long standing problem of Planaria can regenerate from either an whose somatic tissues are of an unusual the source of cells for regeneration has been anterior or a posterior cut surface and ploidy so that metaplasia between different brought nearer by recent studies on contain scattered small basophilic cells somatic tissues can be looked for. planaria and the amphibian limb. The which have been called neoblasts by many Many species of urodele amphibia will experiments use two techniques: X authors. Early experiments 1 showed that regenerate limbs after amputation. The radiation doses which can inhibit cell regeneration could be prevented by limb contains no small basophilic cells and division and regeneration without killing X-irradiation, but if the worm contained the histology of the blastema suggests the tissues, and markers which allow the an unirradiated portion (either a shielded formation by dedifferentiation of muscle, fate of cell populations to be followed. region or an unirradiated graft) then it cartilage and connective tissue . Both planaria and the limb undergo could regenerate fully, even when the Amputation through a small X-irradiated 'epimorphic' regeneration in which a healthy portion was some distance from region does not result in regeneration even blastema consisting of cells which are the level of amputation. The time taken for when the remainder of the lim b and the rest visibly undifferentiated and regeneration to begin increases with this of the animal is unirradiated 7 so, in indistinguishable from one another is distance and correlates with migration of contrast to the planaria, it seems that a formed at the wound site. The blastema the neoblasts through the intervening blastema cannot be formed without the cells first proliferate and then differentiate tissue. participation of tissues local to the wound . to reform the missing body part containing This showed that a blastema can be Recently studies on metaplasia have several different cell types. formed from migratory cells and is capable been carried out using axolotl tissues The key questions are:- (1) Is the of forming all the tissues of the regenerate. labelled by tritiated thymidine and blastema formed (a) from cells drawn from However, it did not exclude the possibility triploidy8-10 or pigmentation and allogenic all parts of the body, or (b) from tissues that the migratory cells were formed by differences leading to graft rejection 11 . 12 local to the wound? (2) Is the regenerate dedifferentiation, or that in un irradiated Grafts of a particular marked tissue into an formed (a) by the de novo differentiation worms differentiated tissues local to the X-irradiated limb followed by amputation of reserve cells, or (b) by dedifferentiation wound can also contribute to the blastema. of the limb through the graft leads to the and redifferentiation of functional tissue Indeed, histological and electron formation of a regenerate wholly cells? (3) Is the blastema (a) a mixture of microscopic studies suggest that local composed of graft-derived cells, confirmed cells pre-committed to differentiate into dedifferentiation does occur2. by the retention of the marker in the particular cell types, or (b) composed of Now studies have been published3- 6 on various cell types. The irradiated similar pluripotent cells able to become any the regeneration of a strain of planaria which background tissue is necessary because of the cell types in the regenerate? are naturally occurring mosaics: the although it does not contribute any cells to Although the questions are logically inde­ somatic cells are triploid, the cells of the the regenerate it does contribute pattern pendent the discussion usually becomes pola­ male germ line are diploid and the cells of information without which few structures rised into an argument of 'neoblasts versus the female germ line are hexaploid. are formed from a small graft 13 • metaplasia'. 'Neoblast' is a term often used Regeneration from the gonad less regions The results are summarised in the Table to indicate small basophilic cells present in gives rise to worms whose somatic tissues below. It is clear that there is extensive meta­ planaria and other organisms, or consist wholly of triploid cells, whilst plasia between the different internal tissues, sometimes in a special sense to indicate regeneration from a cut surface through which are all derived from the mesoderm of mobile pluripotential cells reserved for the gonadal region gives rise to blastemas the embryo. However, epidermis never regeneration. 'Metaplasia' is the and regenererates containing some diploid contributes to the internal tissues and vice dedifferentiation of functiona: cells and/or hexaploid cells. Of particular versa, indicating the existence of a lineage followed by redifferentiation of the cells or significance is that myotubes with diploid restriction between tissues of different their progeny into a different histological nuclei were found in the muscle of the embryological origin. type - for example transformation from regerated pharynx6• If germ cells and their So although there may be some myotube to chondrocyte. The 'neoblast precursors are accepted as bona fide restrictions on the possible inter­ versus metaplasia' argument thus differentiated cells then metaplasia has conversions, there is now good evidence corresponds to options I a, 2a and 3b versus clearly occurred by the sequence: germ from both planariaand the amphibian limb options I b, 2b and 3b. Although the cell ..... blastema cell- muscle cell. that some functional cell types can controversy has become fixed in this However, it could be argued that the germ dedifferentiate and that their progeny can dichotomy, it must be remembered that cells are a special case because in other appear in the regenerate as different other combinations of answers to the three circumstances they may exhibit histological types. These results should be questions are also possible. pluripotency without passing through a taken seriously by those who advocate J.M. W. Slack is in the Imperial Cancer sexual cycle, for example in mammalian theories of cellular differentiation based on Research Fund Laboratories, Mill Hill, teratocarcinoma. This objection might be irreversible events such as the chemical London. overcome by building up stocks of worms modification or rearrangement of DNA. 0

Cell types found in regenerates formed from grafts of particular tissues. 1. Wolff, E. in Rexefleralion (ed. Rudnick , D.) 5.1 ·84 Grafted Tissue (Ronald Press Co ., New York. 1962). 2. Hay, E. D. in The Stability of fhl' Differentialed Slafe ectodermally mesodermally (Sprin ger~ Vcria g. Berlin/ Heidelberg, 1978). derived derived 3. Gregmigni. V. & Puccinelli. I. J. Exp. 7001. 199, 57 (1977). cell types in r------~~------~ 4. Gregmigni. Y. • Miceli , C. &. Pm;(.;indli . 1..1. "'mbrynl regenerate: / . \ Exp. Morph. 55, 53 (1980). epidermis dermiS muscle cartilage nerve (sheath) 5. Gregmign i. V" Miceli, C. & Pi t ano. E. J. t ;mhryo/. Exp. epidermal + Morph. 55,65 (1980). 6. Gregmigni. V. & Mi celi, C . Wdhelm ROllx·... Archives myotube + + ? + 188.107 (1980) . chondrocyte 7. Butler, E.G . & O ' Brien. J.P. Anal. Rec. 84.407 (19411 . + + + + 8. Steen, T.P . J. Exp. Zoot. 167,49 (1968). 9. Namewirth . M. Dev. Bioi. 41,42 (19741. 10. Dunis. D.A . & Namenwirth, M. Dev. Biul. 36. f.)7 (1977 1. II. Wallace. H. J. Embryol. Exp. Morph. 2X. 419 (19 72) + marked cells are present, - marked cells absent,? conflict of results between laboratories. 12. Maden . M. & \Vallance H. Acfa Fmbryol Exp. 2. 77 (1975). 13. Maden, M. J. Embryol. Exp. Morph. 50. 325 (1979) .

0028·0836/ 80/ 340760·01 $01.00 V t 980 Macmillan Journals Ltd