Tolerance of Skin Grafts produced by Various Adult Cells, Soluble Extracts, and Embryonic Cells

At the present time there are two principal experimental means by which animals can be induced to accept homografts. First, tolerance to homografts can be induced by injection of cells from the future homograft donor into embryonic or neonatal animals (Billingham, Brent, & Medawar, 1956; Billingham & Brent, 1957). Second, successful homografts can be made in a fair percentage of animals which have been rescued from an otherwise lethal irradiation by the injection of bone-marrow or splenic cells from the homograft donor (Main & Prehn, 1955, 1957; Trentin, 1956,1957). Recently, it has been shown that in both these types of experiments the injected cells often react immunologically against the host, so causing a homograft reaction in reverse. This ‘graft against host’ reaction in embryonic animals has been studied by Billingham & Brent (1957, 1959), Billingham (1958), Simonsen (1957), and Cock & Simonsen (1958). The ‘secondary disease’ and delayed death following the injection of homologous bone marrow into x-irradiated recipients have been ascribed in a large measure to a graft against host reaction by various workers (Loutit, 1955; Trentin, 1956, 1957, 1958; Uphoff, 1958; Ilbery, Koller, & Loutit, 1958).

In view of the presence of an underlying graft against host reaction in both of these experimental means by which homografts have been caused to be successful, it is of some importance to know whether it is possible to obtain tolerance to homografts without complications of this kind. The use of embryonic cells which are immunologically immature, and which may themselves become tolerant to the host, has yielded significant numbers of tolerant animals without evidence of other than accidental mortality (Billingham et al., 1956; Hasek, 1954; Cannon, Terasaki, & Longmire, 1957). Correspondingly, in the repair of radiation injuries the use of fetal liver tissue has to a large extent prevented the occurrence of ‘homologous disease’ (Uphoff, 1958).

A second method of inducing tolerance without runt disease is applicable only when inbred animals are available, viz. the induction of tolerance in animals of strain X or Y by hybrid XIY tissues (Billingham & Brent, 1959). Under these circumstances the hybrid cells cannot react against the host.

Experiments are reported here in which a third method has been tested, viz. the induction of tolerance by adult cells which are immunologically (as opposed to genetically) incapable of reacting against the host. As shown in an earlier paper (Terasaki, 1959), two such cells in the chicken are the monocyte and the thymocyte. The power of these cells to produce tolerance has been compared with that of lymphoid cells, which do react against the host. As a fourth possible approach, attempts were made to see whether tolerance could be induced by cell-free antigenic preparations.

The ‘graft versus host’ interpretation of runt disease and splenic enlargement clearly depends upon the assumption that iso-antigens are present in the embryonic host which are capable of eliciting a reaction from the adult lymphoid cells injected into it. Some importance therefore attaches to the determination of the stage of embryonic life at which iso-antigens become effectively mature. This problem has been studied by examining the competence of chick embryonic tissues of various ages to produce tolerance after injection into newly hatched chicks of a different breed.

The breeds of chickens used in these studies were the inbred I and W lines,¹ the Rhode Island Red × White Leghorn cross (R × L), and the Light Sussex (LS).

Diluted adult blood, blood lymphocytes, monocytes, thymocytes, and splenic cells were obtained in the manner described by Terasaki (1959). Table 1 shows the dosages of cells, routes of injection, breed combinations, and ages at which the cells were injected. In most cases the cells were injected late in embryonic development to avoid complications attributable to a severe graft against host reaction.

All chicks were tested for tolerance by transplantation of skin grafts from the original tissue donor or from a chicken isologous with the original tissue donor. The grafts were observed up to 30 days after grafting.

The technique of isolation of antigens was essentially that described by Billingham, Brent, & Medawar (1958). Briefly, the method consisted of first lysing splenic cells by washing successively in 0·15 M NaCl, 0·05 M NaCl, and water. The sediment was resuspended in water with a piston-and-cylinder type of blendor and exposed to mild ultrasonic vibrations for 60 seconds. The precipitate which formed upon the restoration of the NaCl concentration to 0·15 M was discarded and the supernatant centrifuged for 60 minutes at 25,000 g. The supernatant was discarded, and the antigenically active sediment resuspended in Ringer-phosphate. Intravenous injection of this extract into embryos or into newly hatched chicks led to immediate death. It was therefore injected either intraperitoneally or subcutaneously, as indicated in Table 1. All the newly hatched chicks so injected were skin-grafted on the same day with skin isologous or autologous with the original donor of the splenic extract. The dosage per chick of antigenic extract in wet weight equivalent of spleen (allowing for a 15 per cent, loss in preparation) was as follows: experiment No. 1, 500, 250, or 125 mg.; experiment No. 2, 210 mg.; experiment No. 3, 500 mg.; experiment No. 4, 720 mg. and 720 mg. after storage for 24 hours at 4°C.; experiment No. 5, 1,674 mg. followed 4 days later by 2,300 mg. In experiment No. 5 the extract was prepared from the spleens of 9 W line chickens for the first injection and 8 W line chickens for the second injection.

Cell suspensions were prepared from the spleens or kidneys, hearts, and lungs of 5 19-day- and 14 13-day-old embryos. The tissues were broken up in a small glass piston blender, the clumps allowed to settle, and the cells taken from the supernatant fluid. Preparations of cell suspensions from 22 6-day- and 39 4-day-old embryos were made from the whole body of the embryos. The embryos were broken up by pipetting in Ringer-bicarbonate and digested for 30 minutes at 37° C. with 0·5 per cent, of ‘Difco’ trypsin 1 : 250’. About 0·1 mgm. of DNA-ase was then added to break up the nucleoprotein gel and, after 30 minutes incubation at 37° C., the cells were washed in heparinized Ringer-phosphate (heparin: 25 u./ml.). In order to obtain a single cell suspension, the supernatant fluid was cropped four times after very light centrifugation for 5 minutes at about 300 r.p.m.

These suspensions were injected intravenously into newly hatched chicks in the dosages and strain combinations indicated in Table 2. All the injected chicks were skin-grafted from an isologous adult chicken on the same day.

Table 1 shows the incidence of tolerance to skin homografts produced by injections of various cells. Adult whole blood diluted up to 1:27 induced tolerance (the grafts of 4/5 subjects showing some degree of survival at 30 days), whereas blood diluted 1:81 and beyond did not (0/18). Blood lymphocytes produced a marked degree of tolerance (18/20), and a measurable degree of tolerance was induced by monocytes (6/14), thymocytes (6/8) and splenic cells after intraperitoneal injection (5/11). The duration of tolerance was not tested longer than 30 days, but in the few chicks allowed to go further the skin grafts were sloughed in the second month.

As is evident from Table 1, soluble extracts prepared from spleens produced uniformly negative results, whether after single or repeated injections (0/39). At no stage did animals which received splenic extract differ significantly from the control animals which received no treatment at all (0 / 27 surviving homografts at 30 days).

In the experiments designed to determine the earliest age at which embryonic tissues possess antigens capable of inducing tolerance to homografts, even 4-day chick embryo tissues were shown to induce some degree of tolerance (Table 2). Thus, at 20 days, when there were no surviving grafts in 17 controls, 8/12 chicks injected with 4-day embryo tissues still retained surviving grafts. Tissues from older embryos induced a degree of tolerance which increased with the age of the tissue: taking survival at 20 days as the criterion for tolerance, 6-day embryo tissues induced tolerance in 10/15,13-day tissues in 12/13, and 19-day tissue in 7/8 chicks.

In the preceding paper (Terasaki, 1959) blood lymphoid cells were shown to cause a splenic enlargement attributable to a graft against host reaction, whereas monocytes and thymocytes caused no such enlargement. In the results described here, monocytes and thymocytes induced tolerance to homografts almost as effectively as blood lymphoid cells. Thus it is clear that cells capable of inducing tolerance to homografts can be distinguished from cells responsible for the graft against host reaction. The graft against host reaction can apparently be avoided by use of tissues free of lymphoid cells.

Lymphoid cells themselves, although capable of causing a graft against host reaction, can elicit tolerance to homografts without overt or clinical signs of a graft against host reaction. But this is only true if the lymphocytes are injected late in the development of the chick, i.e. on the threshold of that stage at which the chick changes over to becoming reactive against homografts. It seems probable that the lymphocytes get into the animal early enough to induce tolerance, but too late to damage the host fatally. Injections of lymphocytes earlier in life, or in higher dosages, are known to have a deleterious or fatal effect.

Previous attempts to induce tolerance to skin grafts in mice with cell-free antigenic extracts have been unsuccessful (Billingham, Brent, & Medawar, personal communication). The chick seemed to be an ideal animal to reinvestigate the problem since in the chick it is possible both to inject soluble extracts and to graft skin on the day of birth. Thus a minimum of time elapses between first exposure to the antigenic stimulus and the transplantation of the skin graft that is used to test for tolerance. In spite of this advantage, no detectable degree of tolerance was found in five separate trials—even with a second injection of antigenic extract 4 days after the first. Unfortunately, the extract could not be injected intravenously because this led to almost immediate death of chicks and chick embryos. Intraperitoneal injection, however, should have been at least partially effective, since it was shown that intraperitoneal injection of spleen cells induced tolerance. Whether the inability of the extracts to induce tolerance is due to the lack of certain antigens in the preparation, or to the antigens being in unsuitable form, or to other causes, remains to be determined.

Although much work has been done on the question of whether fetal homografts will ‘take’ better than adult grafts (e.g. Baxter et al., 1958; Toolan, 1958; Kooreman & Gaillard, 1950; May, 1953; Terasaki, Longmire, & Cannon, 1957) little systematic study has been made of the age at which adult transplantation antigens appear during development. Antigenic analysis of embryonic tissue by the ordinary immunity test has been hindered by the difficulty that the embryonic cells would grow and presumably mature during the 1–2 week period in which they induce immunity to their antigens. This difficulty was overcome by testing the power of embryonic cells to induce tolerance in newly hatched chicks. Such chicks are known to be near the end of the period in which they can be rendered tolerant (Billingham, Brent, & Medawar, 1956). The injected cells must therefore exert their effect within a few days, and any maturation of the cells after that time will not be effective in inducing tolerance. With such tests, embryonic chick tissues from as early as the 4th day of development produced a detectable degree of tolerance. Adult antigens are therefore present very early in development. Although older embryos have induced a greater degree of tolerance, it is difficult to conclude from this that there is a gradual development of adult antigens during development, as it was not technically feasible to compare the effect of any one kind of tissue from embryos of different ages.

The author wishes to express his deep gratitude to Professor P. B. Medawar for his guidance and encouragement throughout all phases of this work and for the preparation of the antigenic extracts. He also wishes to acknowledge the valuable advice and kind hospitality of Dr. L. Brent. This work was supported in part by the United States Public Health Service Fellowship of the National Cancer Institute, and in part by the Department of Zoology, University College, London.