Publication of the International Union Against Cancer Publication de l’Union Internationale Contre le Cancer Int. J. Cancer: 70, 330–334 (1997) r 1997 Wiley-Liss, Inc. EFFECTS OF THYMECTOMY AND TOLERANCE INDUCTION ON TUMOR IMMUNITY IN ADULT XENOPUS LAEVIS Jacques ROBERT1,2*, Chantal GUIET1, Nicholas COHEN2 and Louis DU PASQUIER1 1Basel Institute for Immunology, CH-4005 Basel, Switzerland 2University of Rochester Medical Center, Rochester, NY 14642, USA Major-histocompatibility-complex homozygous partially inbred adults of the ff strain of Xenopus reject transplants of tumor cells of ff strain origin; ff tadpoles do not. Thymectomy, performed 5 days after fertilization, abrogated the adult tumor-rejection response suggesting that in this system tumor rejection is immunologically mediated by T cells. Thymectomy later in larval life did not alter tumor rejection, but it did reduce T-cell numbers. Tolerance to minorhistocompatibility(H) antigens segregating within the ff family, which was induced by grafting adult skin to metamorphosing larvae, did not affect the tumor-rejection capacity of the tolerant adult hosts. This suggests that the ff-2 tumor expresses (a) tumor-specific antigen(s). Immunization of larvae with tumor cells did not induce tolerance to skin grafts transplanted during adult life. Indeed, such grafts were rejected in accelerated fashion, suggesting that memory cells generated in the larvae persist through metamorphosis. Int. J. Cancer, 70: 330–334, 1997. r 1997 Wiley-Liss, Inc. The concept that a tumor expresses antigens that can be recognized by the cognate immune system is supported principally by data from syngeneic murine model systems and from studies in humans (reviewed by Boon et al., 1994; Klein, 1991). The first clear-cut demonstration of tumor immunogenicity showed that immunization of recipient syngeneic mice or rats with irradiated cells from a chemically induced sarcoma of host strain origin protected the recipient against further challenge with live tumors (reviewed by Boon et al., 1994). The capacity to elicit protective immunity has been extended to other types of chemically or virally induced tumors (review by Klein, 1991) and, in a few studies, to spontaneous tumors (Srivastava and Old, 1988). The immune response against tumors is mainly T-cell-mediated, as shown by the impairment of immune protection after depletion of cytotoxic killer T cells and by the possibility of deriving autologous cytotoxic T-cell lines from tumor tissues that can specifically kill tumor-cell targets. Although some tumor antigens have been identified (Boon et al., 1994), the molecular nature of those antigens that elicit a cognate immune response is still largely unknown. By and large, experimental models for studying tumor immunity have been restricted to murine species. Recently, however, thymic tumors have been observed in individuals from a cloned Xenopus ‘‘strain’’ known as LG-15 (Kobel and Du Pasquier, 1975), and in Xenopus from a major-histocompatibility-complex(MHC) homozygous partially inbred strain known as ff (Du Pasquier and Chardonnens, 1975). Stable lymphoid cell lines derived from these tumors have been well-characterized and display a dual T/B-cell phenotype comparable to that seen in rare mammalian lymphocytic leukemias (Du Pasquier and Robert, 1992; Robert et al., 1994). These cell lines provide a unique opportunity to study tumor immunity in an ectothermic vertebrate. Cell lines derived either from LG-15 or ff tumors (e.g., 15/0 and ff-2 respectively) will grow after transplantation in larval recipients provided that the tumor and host are MHC-identical (Du Pasquier and Robert, 1992; Robert et al., 1994). Although LG-15 tumors will also grow in adult LG-15 recipients, tumors originating from ff-strain animals are rejected by adult members of that partially inbred strain. Tumor rejection can be abrogated by sub-lethal irradiation of the frog host (Robert et al., 1995). Rejection capacity developed gradually during the weeks following metamorphosis in parallel with (i) second histogenesis observed in the thymus, (ii) surface expression of MHC-class-II molecule by peripheral T-cells, and (iii) recovery of T-cell-effector function such as mixed leukocyte responses (MLR) (Robert et al., 1995). These observations are consistent with the possibility that tumor rejection is immunologically mediated. The precipitation of surface proteins expressed only by ff-2 cells with an ff adult anti-serum generated against ff-2 cells (Robert et al., 1995) indicates that ff-2 cells express unique determinants. However, given that the ff strain is not fully inbred, it is not easy to distinguish whether tumor rejection is evoked by tumor-specific antigens or by minor-histocompatibility(H) antigens that are still segregating within the ff family (Robert et al., 1995). The present study was designed to obtain basic information concerning the effector cells that mediate tumor rejection and the nature of antigens involved in immune responses elicited by ff-2 tumor cells in mature ff adults. Our first approach involved transplanting tumors to ff adults whose T-cell function (e.g., allograft rejection, MLR, cell-mediated lympholysis) had been ablated by early thymectomy (Horton and Manning, 1972). This procedure profoundly impaired tumor rejection by ff adult hosts, indicating that T cells are required to protect adults against the development of transplanted tumors. A second approach was based on the fact that perimetamorphic, but not adult, Xenopus invariably become tolerant of MHC-compatible minor-H-locus-disparate skin allografts (Du Pasquier and Chardonnens, 1975). Tolerance of ff recipients to minor-H-antigens segregating in the partially inbred ff family was induced by grafting adult ff skin to ff larvae during the metamorphic period. Since this allotolerant state failed to impair the ability to reject ff-2 tumors, the T-cell anti-tumor response appears to be directed against tumor-specific antigens. Both results are consistent with the idea that T-cell-mediated tumor immunity is a fundamental function of the immune system of vertebrates. Furthermore, this Xenopus system provides a new non-mammalian model of cancer immunology in which the ontogeny of tumor immunity can be studied. MATERIAL AND METHODS Animals The Xenopus ff partially inbred strain is MHC-homozygous (ff haplotype) and can retain fully viable ff-strain skin allografts for more than 40 days at 20°C. LG15 (MHC haplotype a/c) isogeneic clones have been described by Kobel and Du Pasquier (1975). Developmental stages were determined according to the normal tables of Nieuwkoop and Faber (1967). Flow cytometry Samples of 105 cells were stained with hybridoma supernatants, followed by fluorescein-labeled goat anti-mouse Ig, before being analyzed by flow cytometry on a FACScan. The technique and the monoclonal antibodies (MAbs) used (TB17, anti-MHC class I; AM20, anti-MHC class II; AM22, anti-CD8; XT-1, anti-pan T cell) *Correspondence to: Department of Microbiology & Immunology, Box 672, University of Rochester Medical Center, 601 Elmwood Avenue, Rochester, New York 14642, USA. Phone: (716) 275-5359. Fax: (716) 473-9573. E-mail: email@example.com. Received 29 July 1996; revised 23 October 1996. XENOPUS ANTI-TUMOR IMMUNITY are describe elswhere (Robert et al., 1995). The anti-CD5 MAb, 2B1, is described in Jürgens et al. (1995). Thymectomy Thymectomy was performed by microcautery (Horton and Manning, 1972) between 5 and 7 days of age (stage 47) or at the pro-metamorphic stage 58. The absence of the thymus was confirmed by external observation of larvae and by autopsy of adults. The immunological efficacy of thymectomy was determined by evaluating rejection of MHC-disparate skin grafts and by flow-cytometric analysis of peripheral-blood lymphocytes with anti-T-cell-specific antibodies; splenic lymphocytes were also analyzed flow-cytometrically at the time of autopsy. Skin grafting Skin from LG15 or ff donors was grafted to ff adults according to published methods (Du Pasquier and Bernard, 1980). Briefly, pieces of ventral skin (5 mm2) were inserted under the dorsal skin of recipient animals, and 48 hr later the overlying host skin was removed. Rejection was considered complete when all pigment cells in the donor graft were destroyed. Tumor-cell culture The cell lines ff-2 and 15/0 were derived from spontaneous thymic lymphoid tumors (Robert et al., 1994). Culture media and culture conditions described elsewhere (Du Pasquier and Robert, 1992) were used in the present study, except that the tumor cells were grown in Iscove DMEM basal medium (instead of L-15) supplemented with 0.25% Xenopus serum. Tumor-cell transplantation Adult ff frogs were challenged with 5 3 105 ff-2 tumor cells by s.c. injection in the dorsal lymph sac. At 20°C, solid tumors usually develop within one month close to the site of injection (Robert et al., 1995). To control for the viability and tumorigenicity of the ff-2 tumor cells at the time of injection, ff tadpoles were anesthetized in 0.01% tricaine methansulfonate (MS-222) and injected i.p. with 5 3 104 tumor cells aliquoted from the same population of cells that were injected into post-metamorphic animals. Peritoneal fluid samples were collected from these tadpoles after 2 to 3 weeks and tumor cells in the ascites fluid were counted. Larvae were immunized by an i.p. injection of 2 3 104 irradiated (50 Gy cobalt source) and washed (23 in amphibian PBS) tumor cells, and boosted one week later with 2 3 104 irradiated cells. All experiments were performed in accordance with Swiss and US regulations governing experimental animal welfare. RESULTS Effect of early larval thymectomy on adult tumor immunity ff-strain larvae were thymectomized 5 to 7 days after fertilization, before colonization of the organ by precursor lymphocytes. 331 Pre-metamorphic tadpoles were examined under a dissecting microscope and any larvae that had discernible thymic tissue were discarded from the study. Seven months after metamorphosis, thymectomized and control animals were examined for the presence of peripheral-blood T cells (Table I) and for their ability to reject MHC-disparate skin grafts. At this age, normal ff animals are resistant to tumor challenge (Robert et al., 1995). Skin allografts on 12 larvally thymectomized post-metamorphic ff frogs exhibited no signs of rejection for more than 2 months after transplantation, and peripheral blood from these animals had virtually no lymphocytes displaying T-cell markers (CD8, CD5). When these animals were killed later in the protocol, their spleens also lacked CD81 and CD51 T lymphocytes (Table I) as well as XT-11 lymphocytes (a pan-T-cell marker; data not shown). The CD5 homologue recognized by the 2B1 MAb is expressed by Xenopus by thymocytes and peripheral T cells, but not by sIgM1 B cells (Jürgens et al., 1995). The fraction of CD51 splenocytes that did not stain with anti-CD8 MAb presumably consisted of helper T cells (anti-Xenopus CD4 MAbs are not yet available). MHCrestricted cytotoxic and helper-T-cell responses have both been characterized in adult Xenopus (review in Flajnik et al., 1987). When these 12 effectively thymectomized ff animals and 15 intact sibling controls were 10 months old, they were each injected s.c. with 5 3 105 ff-2 tumor cells. One month later, tumor development was recorded at the site of injection in 7 of 12 thymectomized animals. In contrast, no tumors were detectable in any of the 15 controls (Table II). All controls and the 5 tumor-free thymectomized frogs were observed for another 6 months during which time no tumors developed. Since a second period of lymphocyte histogenesis occurs in the thymus just after metamorphosis, we questioned whether this event was important for generation of an anti-tumor response. Thus, we thymectomized ff tadpoles at the pro-metamorphic stage 57–58. None of these 20 late larvally thymectomized individuals developed a tumor during the 2-month period following their inoculation with 5 3 105 ff-2 cells tumor cells at 8 months of age (Table II). In all but one instance, a second injection of 5 3 105 ff-2 cells, followed 2 months later by a third challenge of 1 3 106 ff-2 cells, was similarly ineffective in producing tumors (one frog developed a tumor 2 months after the last challenge). To control for the viability and tumorigenicity of the ff-2 tumor cells at the time of injection, ff tadpoles were injected ip with 5 3 104 tumor cells aliquoted from the same population of cells that were injected into the post-metamorphic animals. Within one month after injection, all tadpoles (n 5 5 for each injection) developed ascites. Although late larval thymectomy was without a significant effect on tumor development or rejection of MHC- or minor-H-antigendisparate skin grafts, it did clearly effect a significant reduction in TABLE I – PERCENT (6SEM) OF VIABLE PERIPHERAL-BLOOD AND SPLENIC LYMPHOCYTES FROM CONTROL AND THYMECTOMIZED ff ADULTS THAT POSITIVELY STAINED AT THE CELL SURFACE WITH DIFFERENT MAbs (DETERMINED BY FLOW CYTOMETRY) Number AM22 (CD-8 like) of frogs Peripheral-blood lymphocytes ff adults (8 months old) Intact control Early thymectomy1 Late thymectomy1 Splenic lymphocytes ff adults (1 year-old) Intact control Early thymectomy1 Late thymectomy1 5 12 12 3.9 6 0.78 0.36 6 0.14 0.39 6 0.15 4 102 3 36.6 6 6.96 1.1 6 1.47 25.4 6 4.90 2B1 (CD5-like) AM20 (class II) 10A9 (IgM) 9.8 6 3.63 0.16 6 0.09 0.22 6 0.17 18.5 6 2.42 4.05 6 2.54 4.0 6 2.58 N.D. (not done) 62.3 6 12.95 93.2 6 1.80 21.7 6 4.16 0.35 6 0.92 59.6 6 22.25 35.6 6 16.05 27.75 6 17.27 82.0 6 17.27 48.64 6 12.69 1Early thymectomy was performed 5 to 7 days after fertilization, late thymectomy at pro-metamorphic stages 58–59.–2Two individuals displayed metastases and were not included, since ff-2 tumor cells strongly express surface CD8 and CD5 molecules. ROBERT ET AL. 332 TABLE II – TUMOR INCIDENCE AT THE SITE OF INJECTION AFTER ONE S.C. CHALLENGE1 ff adults 8 months old Tumor incidence after 6 months2 Intact controls Early thymectomy3 Late thymectomy3 0/15 (0%) 7/12 (58.3%) 1/20 (5.0%)2 1With 5 3 105 ff-2 tumor cells.–26 months after a third tumor challenge.–3Early thymectomy was performed 5 to 7 days after fertilization, late thymectomy at pro-metamorphic stage 58. TABLE III – TUMOR INCIDENCE AT THE SITE OF INJECTION AFTER 2 SUBCUTANEOUS CHALLENGES1 Treatment of ff hosts ff donor LG15 adult skin grafts given during metamorphosis Tolerance induced by ff adult skin grafts during metamorphosis4 Immunized with irradiated ff-2 cells before metamorphosis (st 54–55) Immunized with irradiated ff-2 cells during metamorphosis (st 58–59) Tumor incidence 6 months after 2nd tumor LG15 donor challenge3 Rejection of test grafts at 8 months of age2 chronic acute 0/20 no rejection acute 0/60 acute acute 0/20 acute acute 0/20 1Interval, 1 month, with 1 3 106 ff-2 tumor cells in ff recipient after tolerance induction or immunization at larval and metamorphic stages.– 2See Table IV for detailed results of graft rejection. Skin-graft rejection in less than 25 days.–3The viability and tumorigenicity of challenged tumor cells was controlled by injecting sensitive younger postmetamorphic ff froglets; all (5/5) developed tumors at the site of injection within one month.–4The extent of tolerance has been optimized by giving 2 groups of 20 ff animals skin grafts from adult ff donors of different progeny and a third group of 20 individuals both types of donor skin. the number of T lymphocytes in the peripheral blood and spleen, as determined by flow cytometry (Table I). Effect of tolerance induction Sharing of minor-H antigens between ff skin and ff-2 tumor cells was suggested by the accelerated rejection of ff skin grafts by adults that had been immunized with irradiated ff-2 tumor cells (Robert et al., 1995). To determine whether such minor-H antigens are the only determinants eliciting a tumor-rejection response by the adult, tolerance of adult ff skin grafts was induced in ff animals by grafting them with ff skin at metamorphosis. To optimize the diversity of minor-H antigens to which animals had become tolerant, 2 groups of 20 ff animals received grafts from adult ff donors derived from 2 different matings, and a third group was grafted with both types of donor skin. Tolerance was demonstrated not only by the long-term survival of these grafts but also by the inability of the hosts to reject a skin graft from a third-party ff donor when they were adults. This tolerance was specific, since the hosts were capable of acutely rejecting MHC-disparate grafts from an LG15 donor (Table IV). None of the 60 ff adults (8 months old) that were tolerant of ff skin developed a tumor within 6 months after receiving 2 injections of ff-2 tumor cells separated by one month (Table III). Since all (5/5) younger non-tolerant control ff animals that had been injected with an aliquot of the same ff-2 cells developed tumors within a month after inoculation, the preparation was tumorigenic. Furthermore, tumor cells injected in the 5 ff animals that were tolerant of 2 sets of adult ff skin grafts (Table IV), failed to produce tumors or to induce rejection of the tolerated ff skins within a 2-month period. TABLE IV – REJECTION OF SKIN GRAFTS FROM ADULT LG15 AND ff DONORS BY 2-YEAR-OLD ff ADULTS IN DIFFERENT TREATMENT GROUPS Individual mean survival times of skin grafts1 Treatment of adult ff hosts Untreated control Thymectomized at 5–7 days old Thymectomized at pro-metamorphic stage 58 Adult LG15 skin grafts at pro-metamorphic stage 58 Tolerance induced by adult ff skin grafts at pro-metamorphic stages 58–59 Immunized with irradiated ff-2 tumor cells before metamorphosis (stage 54) Immunized with irradiated ff-2 tumor cells at prometamorphic stage 58–59 1Days from LG15(a/c) donors from ff donors (MHC 1 minor-Hlocus-disparate) (minor-Hlocus-disparate) 18, 22, 24 (21.3 6 3.06) .60, .60, .60 (.60) 18, 18, 20 (18.7 6 1.15) 53, 53, 53, 55 (53.7 6 1.15) ND 20, 20, 20, 25 (21.25 6 2.50) 49, 53, 57, .60 (53 6 4.0) 18, 20, 24, 25 (21.8 6 2.86) .80, .80, .80, .80, (.80) 19, 20, 22, 25 (21.5 6 2.65) 22, 25, 25, 29 (25.25 6 2.87) 18, 20, 25 (19 6 1.41) 22, 25, 25 (24.25 6 1.5) 49, 49, 52 (50 6 1.73) at 22°C; mean 6 SD in parentheses; ND, not done. In another experiment, we attempted to induce tolerance to ff-2 tumor cells by priming pre-metamorphic stage-52 hosts or at peri-metamorphic stage-58-59 hosts with irradiated ff-2 cells. In neither instance did this priming affect the adults’ ability to reject subsequently transplanted ff-2 tumor cells. That is, none of the adult ff animals primed in larval life and challenged twice with ff-2 tumor cells had developed a tumor by 6 months after the second challenge. Furthermore, priming of ff recipients with irradiated ff-2 cells at either larval stage did not impair their capacity to reject ff skin grafts. Indeed, rejection of such skin grafts was sub-acute (25 days). This timing contrasts sharply with the much more chronic rejection of ff skin by adult ff animals that had not been primed with tumor cells during larval life. In other words, it appears that injecting ff larvae with ff tumor cells elicits a second-set rejection of skin grafts, whereas grafting larvae with skin induces long-lasting tolerance. DISCUSSION Nature of adult anti-tumor effector cells Data revealing the development of the ability of ff adults to reject ff-2 tumors during metamorphosis and the abrogation of this response by sub-lethal irradiation have suggested that tumor rejection was immunologically mediated (Robert et al., 1995). In the present study, thymectomy of very young tadpoles before initial colonization of the thymus by embryonic stem cells (Horton and Manning, 1972) provided a useful way to investigate the in vivo requirement for T cells in the rejection of ff-2 tumor cells by ff adults. Indeed, thymectomy markedly impaired host resistance to tumor development after ff-2-tumor-cell transplantation. Thymectomy of larval Xenopus abrogates T-cell functions in adult life, e.g., the IgY-antibody response, allograft rejection, and MLR and PHA responsiveness (review in Flajnik et al., 1987). Other immune functions such as IgM-antibody response against T-dependent or T-independent antigens, remain unimpaired (Flajnik et al., 1987). At the phenotypic level, T-cell populations expressing surface XENOPUS ANTI-TUMOR IMMUNITY pan-T-cell, CD8 or CD5 markers are virtually absent from cells in the peripheral blood and spleens of thymectomized animals (Jürgens et al., 1995; and Table I). The impairment of tumor rejection in ff adults thymectomized during the first week of life strongly suggests, therefore, that as in mammals, the immune response elicited against tumor cells is T-cell-mediated. Some thymectomized recipients of tumor cells did not develop tumors. Whether this reflects activity of residual T cells and/or natural-killer(NK) cells is unknown. On the other hand, the early education of T cells in the thymus in the absence of MHC-class-I expression (review in Flajnik et al., 1987; Flajnik and Du Pasquier, 1990) could be insufficient to allow for full protection in tadpoles; additional maturation of anti-tumor effectors must occur during metamorphosis. The thymus environment, however, does not appear to be required for this metamorphic maturation, since thymectomy performed at the beginning of metamorphosis failed to impair tumor resistance during adult life. This suggests that the second wave of stem-cell immigration in the thymus (Turpen and Smith, 1989) following the loss of about 90% of larval thymocytes (review in Flajnik et al., 1987) is not the major source of mature anti-tumor effectors. It has been shown in mice and humans that the generation of a potent cytotoxic response against tumors or virus depends on efficient antigen presentation by a macrophage sub-set (Suto and Srivastava, 1995) or by dendritic cells (Celluzzi et al., 1996), as well as on cytokine release. On the assumption of similar involvement in Xenopus, it may be speculated that some of these elements are still differentiating during metamorphosis. It is known, for example, that cells with characteristics of mammalian Langerhans cells first appear in Xenopus skin after metamorphosis (Flajnik and Du Pasquier, 1990). Such cells, or new macrophage populations that differentiate during metamorphosis, may promote a better immune response by more efficiently presenting antigen and/or by producing cytokines. Alternatively, other types of effectors, such as NK cells or gamma/delta T cells, may develop during metamorphosis and comprise a parallel anti-tumor defense system. There is some indirect evidence of NK-cell activity in Xenopus (Horton et al., 1996) but there is no information whatsoever about gamma/delta T cells in this species. Nature of antigens Since the ff family is only partially inbred (i.e., adults reject skin grafts chronically), rejection of ff-2 tumor cells could result from an immune response directed against minor-H antigens that are segregating in the family. Indeed, the accelerated rejection of skin grafts by ff adults that had been immunized with ff-2 tumor cells implies sharing of minor-H antigens between ff skin and ff-2 tumor cells (Robert et al., 1995). However, the fact that tumor development was never observed in any of the numerous ff adults transplanted with ff-2 tumor suggests rather rapid and efficient rejection, which is uncharacteristic of that elicited by minor-H antigens (Robert et al., 1994, 1995). In addition, a tumor-specific 333 product has been detected by immunoprecipitation of tumor cells with ff adult anti-serum generated against ff-2 cells (Robert et al., 1995), indicating that ff-2 tumor cells do express unique determinants. To further characterize the tumor antigens recognized by adults, we took advantage of the fact that an antigenically specific and long-lasting tolerance can be induced to minor-H antigens (Du Pasquier and Chardonnens, 1975) and sometimes to MHC antigens by skin grafts transplanted to larvae at metamorphic stages. Establishment of this tolerance is thymus-dependent (Barlow and Cohen, 1983). The induction of tolerance in ff animals toward ff minor-H-antigens did not affect the capacity of adults to reject ff tumors, suggesting that ff-2 tumor cells express additional antigens (i.e., tumor antigens) capable of eliciting an immune response. Immunization at pre- or pro-metamorphic stages with irradiated ff-2 cells did not impair tumor rejection by adults. Furthermore, it did not induce tolerance to minor antigens shared between ff skin and ff-2 tumor cells. On the contrary, ff animals that had been immunized with ff-2 tumor cells at pre- or pro-metamorphic stages acutely rejected ff skin grafts transplanted during adult life, suggesting that memory persisted through metamorphosis. The generation and persistence of tumor-specific memory through metamorphosis by immunization at larval stages was already suggested by the significant protection to tumor transplantation (a fraction of transplanted hosts did not develop tumors) obtained just after the end of metamorphosis, when very young frogs are usually sensitive to development of transplanted tumor (Robert et al., 1995). The amount and/or the preparation of immunogen (presence or absence of adjuvant) injected or the route by which it was introduced could have been inadequate to evoke tolerance, as has been recently demonstrated in tolerance induction of mouse neonates (Ridge et al., 1996). Nevertheless, memory detected after metamorphosis implies that larval lymphocytes, most likely T lymphocytes, can recognize tumor determinants; additional factors, however, must mature during metamorphosis to provide effective tumor immunity. In summary, current data support the proposition that Xenopus, like mice and rats, are capable of mounting a T-cell-mediated tumor-rejection response directed against tumor-associated antigens. Further analysis of each of these facets of tumor immunity are in progress. ACKNOWLEDGEMENTS We wish to thank Dr. E. Lord and Dr. C. Steinberg for helpful suggestions and critical reading of the manuscript. The Basel Institute for Immunology was founded and is supported by F. Hoffman La Roche, Basel, Switzerland. Research at the University of Rochester was supported by USPHS HD-07901. REFERENCES BARLOW, E.H. and COHEN, N., The thymus dependency of transplantation allotolerance in the metamorphosing frog Xenopus laevis. Transplantation, 35, 612–619 (1983). BOON, T., CEROTTINI, J.-C., VAN DEN EYNDE, B., VAN DER BRUGGEN, P. and VAN PEL, A., Tumor antigens recognized by T lymphocytes. Ann. Rev. Immunol., 12, 337–365 (1994). CELLUZZI, C.M., MAYORDOMO, J.I., STORKUS, W.J., LOTZE, M.T. and FALO, L.D., Peptide-pulsed dendritic cells induce antigen-specific CTL-mediated protective tumor immunity. J. exp. Med., 183, 283–287 (1996). DU PASQUIER, L. and BERNARD, C.C., Active suppression of the allogeneic histocompatibility reactions during the metamorphosis of the clawed toad Xenopus. Differentiation, 16, 1–7 (1980). DU PASQUIER, L. and CHARDONNENS, X., Genetic aspects of the tolerance to allografts induced at metamorphosis in the toad Xenopus laevis. Immunogenetics, 2, 431–440 (1975). DU PASQUIER, L. and ROBERT, J., In vitro growth of thymic tumor cell lines from Xenopus. Devel. Immunol., 2, 295–307 (1992). FLAJNIK, M.F. and DU PASQUIER, L., The major histocompatibility complex of frogs. Immunol. Rev., 113, 47–63 (1990). FLAJNIK, M.F., HSU, E., KAUFMAN, J. and DU PASQUIER, L., Changes in the immune system during metamorphosis of Xenopus. Immunol. Today, 8, 58–63 (1987). HORTON, J.D., HORTON, T.L. and RITCHIE, P., Immune system of Xenopus: T-cell biology. In: R.C. Tinsley and H.R. Kobel (eds.), The biology of Xenopus. Symposium of the Zoological Society of London, pp. 279–299, Oxford University Press, Oxford (1996). HORTON, J. and MANNING, M.J., Response to skin allografts in Xenopus laevis following thymectomy at early stages of lymphoid-organ maturation. Transplantation, 14, 141–154 (1972). JÜRGENS, J.B., GARTLAND, L.A., KISHIMOTO, T., DU PASQUIER, L., HORTON, 334 ROBERT ET AL. J.D. and COOPER, M.D., Identification of a candidate CD5 homologue in the amphibian Xenopus laevis. J. Immunol., 155, 4218–4223 (1995). KLEIN, G., Immunovirology of transforming viruses. Curr. Opin. Immunol., 3, 665–673 (1991). KOBEL, H.R. and DU PASQUIER, L., Production of large clones of histocompatible, fully identical clawed toads (Xenopus). Immunogenetics, 2, 7–91 (1975). NIEUWKOOP, P.D. and FABER, J., Normal table of Xenopus laevis (Daudin). North Holland, Amsterdam (1967). RIDGE, J.P., EPHRAIM, J.F. and MATZINGER, P., Neonatal tolerance revisited: turning on newborn T cells with dendritic cells. Science, 271, 1723–1726 (1996). ROBERT, J., GUIET, C. and DU PASQUIER, L., Lymphoid tumors of Xenopus laevis with different capacities for growth in larvae and adults. Dev. Immunol., 3, 297–307 (1994). ROBERT, J., GUIET, C. and DU PASQUIER, L., Ontogeny of the allo-immune response against transplanted tumor in Xenopus laevis. Differentiation, 59, 135–144 (1995). SRIVASTAVA, P.K. and OLD, L.J., Individually distinct transplantation antigens of chemically induced mouse tumors. Immunol. Today, 9, 78–83 (1988). SUTO, R. and SRIVASTAVA, P.K., A mechanism for the specific immunogenicity of heat-shock-protein-chaperoned peptides. Science, 269, 1585–1588 (1995). TURPEN, J.B. and SMITH, P.B., Precursor immigration and thymocyte succession during larval development and metamorphosis in Xenopus. J. Immunol., 142, 41–47 (1989).