Search

Studies conducted upon inbred strains of mice, hamsters and rats have shown that following the interstrain matings the now familiar covert reactivity of pregnant females to the alloantigens of their conceptuses may benefit the latter in two ways; firstly, it exerts a significant influence upon placental weight, and indirectly upon the birth weight of the fetus-allogeneic placentas tending to be heavier than syngeneic placentas, and mothers specifically presensitized against alien paternal tissue antigens gestate fetuses with heavier placentas than normal females. Specifically tolerant mothers, on the other hand, produce smaller, F1 hybrid, fetoplacental (fp) units. The classic notion that the disparity between the birth weights of F1 hybrid and homozygous offspring is due to hybrid vigor has been challenged by the finding that DA and (DA times F1)F1 hybrid blastocysts transferred to the uteri of genetically tolerant (DA times F1)F1 hybrid rats produce fp units of similar weight Maternal immunological reactivity against the fetus qua allograft may make a significant contribution here. Additional support for the premise that maternal reactivity against fetal alloantigens in some way promotes the growth of the fp unit was afforded by the finding that excision of the para-aortic lymph nodes (which drain the uterine horns) from females before interstrain matings resulted in smaller fp units than in females subjected to sham operations. The finding with one rat strain combination that passive immunization of females with serum against their F1 hybrid conceptuses promoted the growth of the latter suggests that a humoral rather than a cellular immunity may be involved. Secondly, in the three species studied, it was observed that genetic disparity between a conceptus and its mother significantly improved its chances of implantation and development to term.

Human milk lymphocytes (ML) can be partially purified and propagated in vitro as a means of assessing their immunological function. When exposed to a variety of stimuli known to activate T lymphocytes, ML respond in a unique manner that indicates a selected population of immunocompetent cells. ML are hyporesponsive to to nonspecific mitogens and respond in a reduced manner to histocompatibility antigens on allogeneic cells. In most cases, they are completely unresponsive to C. albicans although blood lymphocytes from the same patients respond to the antigen. The Kl capsular antigen of E. coli induces significant proliferation in lymphocytes obtained from milk, but fails to stimulate blood lymphocytes. This dichotomy of reactivity does not appear to result from suppressive factors or cells in milk or insufficient adherent cell function. Rather it appears to reflect the accumulation of particular lymphocyte clones in the breast and the local nature of mammary tissue immunity at the T-lymphocyte level.

Using guinea pigs of strains 2 and 13 and their F1 hybrids as experimental subjects, various lines of evidence have been obtained that in this species, as in all others tested, the only significant cellular antigens with which donor lymphocytes engage when normal and immune lymphocyte reactions are incited are radiosensitive leukocytes. Constitutive cells of the skin are unimportant. (a) The intensities of these reactions in irradiated subjects are dependent upon the peripheral leukocyte concentration. When this falls below a certain threshold no reactions are incitable. (b) Highly leukopenic animals are capable of developing immune lymphocyte transfer (ILT) reactions if normal lymphoid cells of their own genetic constitution are mixed with the putative attacking donor cells, as "supplementing antigen," before inoculation. (c) Radiation-chimeric strain 13 animals having F1 hybrid leukocytes in their bloodstream give typical ILT reactions when challenged intradermally with strain 13 anti-2 node cells. Exposure of strain 2 animals to 600 R does not prevent their becoming actively immunized if, 24 hr later, they are injected intradermally with strain 13 lymphocytes. However, this sensitization, revealed by the host's capacity to give delayed hypersensitivity reactions, wanes as leukopenia progresses. On the basis of this and other findings it is argued that the flare-up stage of the NLT reaction in preirradiated hosts is mainly an expression of host sensitivity against the transferred alien cells. Two unexpected observations have been made in the course of this study: (a) F1 hybrid animals developed what appeared to be a strong delayed hypersensitivity after intradermal inoculation with parental strain lymphoid cells or antigenic extracts prepared from them. (b) If strain 13 guinea pigs which had been sensitized against strain 2 tissue antigens by intradermal injection of lymphocytes 7 days beforehand were inoculated intravenously with strain 2 antigenic extract a significant proportion of the animals developed severe delayed necrotizing reactions, recall flares, at some or all of the healed skin inoculation sites.

Rats have been shown to be capable of displaying the various kinds of delayed cutaneous hypersensitivity reactions attributable to transplantation immunity previously described in guinea pigs, hamsters, rabbits, dogs, and man. The rat is unlike most other species in that much larger numbers of lymphoid cells are needed to incite these cutaneous reactions. With direct, transfer, or normal lymphocyte transfer reactions, the cutaneous responses were greater when donor and recipient differed at the Ag-B histocompatibility locus than when donor and recipient shared the same Ag-B alleles. An experiment was performed in which adult rats of a genetically defined backcross population, resulting from matings between DA and (DA x Lewis) F1 hybrid rats, were inoculated intradermally with lymph node cells from DA rats sensitized against tissues from Lewis-strain donors. Some of the R2 animals gave a biphasic transfer reaction with peak reactivities occurring first at 48 hr and recurring at 96–120 hr, while the others lacked this second component. Hemagglutination tests revealed that the R2 rats giving the biphasic response possessed the Ag-B,1 antigen, which is also present in Lewis rats, whereas rats which gave monophasic reactions were homozygous for the Ag-B,4 antigenic determinant which is present in the DA strain. This suggested that the recall flare at 96–120 hr reflects proliferative activity on the part of the inoculated cells confronted by the disparate Ag-B isoantigen in the host's dermis. Skin homografts from R1 animals bearing the Ag-B,1 antigen were uniformly rejected by DA hosts in 11 days or less, while grafts from backcross animals homozygous for the Ag-B,4 antigen usually lived longer, being rejected in 9 to 27 days. Evidence is also presented which suggests that specific isoantibodies may act synergistically with immune lymphocytes to bring about cutaneous inflammatory reactions in the rat.

The intracutaneous inoculation of lymph node cell suspensions from normal hamsters into normal, homologous hosts causes the development of delayed inflammatory reactions, normal lymphocyte transfer (NLT) reactions, the intensities of which reflect the antigenic disparity between donor and recipient. The immunogenetic situations in which they occur indicate that these reactions are due to sensitization in situ on the part of inoculated immunologically competent cells against alien antigens of the host. They are graft versus host reactions. If hamsters are sensitized by cellular or solid tissue homografts and then challenged intracutaneously with lymphoid or epidermal cells from the donor strain, direct reactions are provoked. A positive correlation exists between the development of transplantation immunity and the capacity to give direct reactions. The sensitivity responsible for direct reactions can be suppressed by irradiation or administration of cortisone, and it is transferable by means of viable lymphoid tissues or cells but not by means of serum. Intracutaneous inoculation of viable node or splenic cells from specifically sensitized hamsters into hosts of the donor strain incites transfer reactions, the intensities of which depend upon the number of cells transferred and the level of sensitivity in the animal that provided them. These reactions are not incitable by thymocytes, by killed lymphoid cells, or by serum. Transfer reactions differ from NLT reactions only on a quantitative and chronologic basis. However, a sharp experimental distinction can be drawn between them. For example, if normal MHA node cells and node cells from MHA hamsters sensitized against CB antigens are injected into the skins of MHA's tolerant of CB tissue, only the sensitized cells incite reactions. Both direct and transfer reactions are highly specific immunologically. The inability of node cells to incite NLT or transfer reactions in heavily irradiated hosts and other findings sustain the thesis that it is host cells of hematologic origin, rather than indigenous skin cells, that contribute the antigenic stimulus required for these reactions. All the findings presented are consistent with the hypothesis that the various delayed inflammatory reactions described are manifestations of immunological responses on the part of immunologically competent cells against alien transplantation antigens.

Melanocytes may be regarded as unicellular, pigment-secreting glands, largely though not entirely confined to the epidermis, of which they are a constant though often un-recognized cellular component. Their numerical incidence, and their branching form, with numerous dichotomizing processes, are such as to form a reticular system within the epidermis. Evidence is evaluated that they are a race or lineage of cells sui generis and form a self-maintaining system within the epidermis. Their product-melanin granules-is secreted directly into the cytoplasm of neighboring Malpighian cells upon which the processes of the melanocytes terminate in the form of small swellings or end-caps. The several types of branched cells demonstrable within the epidermis by various techniques are discussed in relation to their identity as living melanocytes, or as effete cells of this lineage which, having lost or discharged their pigment, participate in the general out-ward movement of epidermal cells to be lost from the skin su...

The experiments reported herein provide ample evidence that mice, like most other mammalian species, are capable of displaying readily observable and reproducible delayed cutaneous hypersensitivity reactions indicative of transplantation immunity. By employing a variety of genetically defined strains, it has been shown that a genetic requirement for the development of a positive normal lymphocyte transfer reaction in mice is a difference between host and cell donor at the H-2 locus. By contrast, the immune lymphocyte transfer reaction consistently reflected the full range of histoincompatibility, both inclusive and exclusive of the H-2. It was incidentally discovered that erythema regularly accompanied delayed cutaneous reactions in the skins of female mice, whereas no local redness accompanied their counterparts in male skins. The influence of cutaneous erythema on the scoring of delayed skin reactions is discussed.

Free tail skin grafts or suspensions of viable epidermal cells have been placed in the atraumatized uterus of isologous rat hosts and allowed to "implant" of their own accord to study the possible uniqueness of this site for other than nature's transplants, i.e. conceptuses, and its response to unnatural grafts. Despite the presence of an intact endometrial epithelium, free skin grafts heal-in rapidly, provided that a state of estrogen excess is established at the time of transplantation. In the absence of estrogen most of the grafts failed to implant. Once established, the grafts survive indefinitely without further estrogen. However, if at any stage a state of continual estrus is established, skin epidermis migrates centrifugally from the graft perimeter invasively replacing the native uterine epithelium. The results of an analysis of the modus operandi of this estrogen-facilitated epidermal migration in utero sustain the view that the hormone acts upon the uterine stroma rather than upon the epidermal cells. When grafts of lingual mucosa or vaginal "skin" were placed in the uteri of rats maintained in chronic estrus, migration of epidermis took place even more vigorously than from tail skin. These epithelia conserved their distinctive histologic specificities indefinitely when growing as heterotypic recombinants on the alien mesenchymal stroma of the uterus. Monodisperse suspensions of epidermal cells appear to "implant" and establish small epidermal plaques in the uterus only at sites predestined to accept conceptuses. That the endocrinologic parameters for the establishment of skin grafts in the uterus are similar to those for blastocysts is suggested by the finding that both kinds of graft can become established in the same uterine horn in the absence of exogenous hormones.

F1 hybrid hamsters derived from genetically disparate strains develop a severe and often lethal cutaneous disorder when inoculated intracutaneously with immunologically competent lymphoid cells from either parental strain, The disease is characterized clinically by extensive epidermal necrolysis, and histologically by a complete dissolution of the dermal-epidermal junction. The requisites for elicitation of this syndrome were determined to be: (a) the parental strains must differ from each other at a major histocompatibility locus, and (b) the donor inoculum must contain immunologically competent parental strain cells. In addition it was found that specifically sensitized cells surpassed normal unsensitized ones in their ability to elicit the disease, and that the disease can be transferred adoptively from affected to normal F1 hosts by means of lymphoid cells. On the basis of these observations, it was concluded that the disease was immunologic in nature, and graft-versus-host in type. However, a series of critical studies failed to demonstrate that the epidermolysis had an immunogenetically specific basis, thus invalidating the provisional assumption that this lesion resulted from a direct immunologic attack upon parenchymal cells of the epidermis and dermis. With the aid of radiation chimeras, it was clearly established that typical epidermolysis could be induced in skin of the same genetic constitution as the attacking donor lymphoid cells. This paradox was taken into account by the possibility that, amid the intense local cutaneous graft-versus-host reactions, "skin-specific" antigenic determinants are bared which incite a quasi autoimmune response that in turn is responsible for the epidermolytic lesions.

The epidermolytic syndrome that can be obtained at will in F1 hybrid hamsters by the cutaneous inoculation of adequate doses of parental strain lymphoid cells has been investigated to determine whether the cutaneous lesions are due to an autoimmune process arising from the severe, initial GVH reactions in the skin. It was amply demonstrated that inoculation of donor cells into the skin was of crucial importance to the development of epidermolysis. Parental strain lymphoid cells in similar doses delivered by any other route into normal F1 hybrids failed absolutely to incite the acute syndrome. If "immune lymphocyte transfer" reactions incited by donor cells in the host's skin were surgically removed at timed intervals after inoculation, only complete excision within 24 hr prevented the appearance of epidermolysis in F1 hybrid hosts, indicating that inoculated donor cells must remain within the confines of the skin for approximately 24 hr in order to evoke the disease, persistence for longer periods of time being unnecessary for the subsequent course of the disease. However, reconstitution experiments involving the intramuscular inoculation of suspensions containing mixtures of donor cells and host lymphoid cells, in the presence or absence of epidermal cells, unequivocally indicated that no intimate exposure of lymphoid cells to putative skin-specific antigens was essential. Similarly, the elicitation of generalized epidermolysis in F1 hybrids irradiated with 300 r and then inoculated intravenously with donor cells casts further doubt on the pathogenic importance of the skin as a source of tissue-specific antigen. The results of subsequent experiments indicated that host leukocytes, rather than parenchymal cells of the dermis or the epidermis, were important contributors of the transplantation antigenic stimulus. Moreover, a series of experiments, using (CB x MHA)F1 hybrid hosts that had been lethally irradiated and reconstituted with bone marrow cells from ALS-treated MHA donors, indicated that from 6 to 10 wk after reconstitution—when direct and immune lymphocyte transfer reactions showed a virtual absence of native F1 leukocytes from the circulation—donor cells obtained from specifically sensitized MHA donors were completely ineffective in inducing epidermolysis, while equivalent lymphoid cell inocula derived from CB donors evoked the cutaneous disease irrespective of the time elapsed since reconstitution. To explain these findings it is postulated that in hamsters, the primary targets in graft-versus-host disease incited by the intracutaneous inoculation of donor cells are leukocytes originating in bone marrow or lymph node, or both.

Using rats of the Lewis and BN (Ag-B locus incompatible) isogenic strains, a comparative study has been made of the capacity to prevent or mitigate the development of runt disease with: (a) lymph node cell suspensions from normal adult BN rats, (b) node cells, or (c) serum from donors sensitized against Lewis tissue antigens, or (d) heterologous anti-lymphocyte serum (ALS) raised in rabbits against rat thymocytes. Following a standard intravenous or intraperitoneal inoculation of 20 x 106 Lewis node cells into neonatal BN hosts, there are cutaneous manifestations of runt disease within 11–15 days and death invariably takes place within 20 days. However, complete protection is afforded by administration of a similar number of normal BN node cells via a different vein, or admixed with the otherwise harmful Lewis node cells. However, timing of the administration was crucially important—precedence or delay by as little as 4 hr resulted in a great impairment of protection. When the inoculations of the two cell suspensions were separated by 24 hr, no protection was afforded. These and other observations suggested that a necessary condition for protection of the hosts by unsensitized isologous cells requires that they establish a prompt and intimate confrontation with the homologous target cells. At the same dosage level, suspensions of node cells from sensitized isologous donors were much more effective therapeutically, saving the lives of 92% of treated subjects when administered after a delay of 3 days, and of 19% when the delay was 4 or 5 days. Of the various immunotherapeutic agents studied, daily injections of 0.2 ml of isoantiserum gave the best results, and could totally reverse the course of the disease even when initiated at age 10–13 days and subjects already presented symptoms. ALS, although inferior to isoantiserum at the dosage levels tested, proved to be superior to sensitized isologous cells as a protective agent, since the initiation of daily injections after delays of 6 or 8 days were still effective. The observations that delayed treatments of infant rats with isoantisera or ALS resulted in complete recoveries sustain the thesis that the lesions responsible for the fatal outcome of runt diseases are not inflicted at a very early stage. The efficacy of both isoantisera and ALS as a means of inhibiting the progression of homologous disease also suggests that they may have therapeutic value in situations where this condition is encountered.

Pigment spread is the natural or experimentally procured (through grafting) progressive encroachment of pigmentation from black or red skin areas into juxtaposed white skin areas, or from black skin areas into red skin areas in spotted guinea pigs and other mammals. So far as spread from black into white skin is concerned, it had previously been shown that migration of epidermal melanocytes into skin lacking homologues of these cells was responsible. However, since red skin already has its own complement of phenotypically "red" melanocytes, the intriguing possibility remained that when black pigment encroaches upon red, rather than melanocyte migration being responsible, phaeomelanin (red)-producing melanocytes are transformed into eumelanin (black)-producing cells by some kind of serially transmissible factor derived from contiguous eumelanotic melanocytes. By utilizing two isogenic strains (Nos. 2 and 13) of spotted guinea pigs and their F1 hybrids, the mechanism underlying the spread of pigment from black into red skin has been analyzed, employing cellular transplantation antigens as melanocyte "markers." The findings demonstrate unequivocally that a physical migration of pigment cells is responsible. By comparing the extents of pigment spread from black ear skin grafts, or from epidermal cell suspensions prepared therefrom, from parental strain or from F1 hybrid donors in white host skin areas of F1 hybrid guinea pigs, it has been possible to evaluate the influence of the intimate contact of melanocytes with alien transplantation antigens on their survival and migratory behavior. No evidence was forthcoming that pigment spread takes place less readily when the cells responsible are confronted by epidermal cells bearing foreign antigens than when they are confronted by cells of their own antigenic constitution. These findings are contrary to expectation if the phenomena of allogeneic inhibition or contact-induced cytotoxicity apply to normal cells in in vivo situations.

IN mice and rats there is one exception to the rule that grafts are always accepted between members of an isogenic population. This applies when females are challenged with skin and other tissues from male donors1–6. The ability of females to reject these grafts varies from strain to strain. For example, while virtually 100 per cent of C57BL/6 (hereafter C57) females and about 75 per cent of A strain females reject male skin isografts, CBA and AU strain females usually accept such grafts. This interstrain diversity occurs despite the fact that the specificity of the Y antigen or “Y factor” is apparently the same in male mice of all stocks7. The basis for this variability stems from two factors: the first is the genotype of the female which determines her capacity to react against male skin and the second is the genetic background of the male which influences the expression of the Y antigen8. Thus the complete penetrance of the Y factor in C57 mice is due to the females of this strain being relatively strong reactors against male skin isografts, and is not because the antigen is strongest in this strain. Indeed, that the antigen is weaker in C57 males than in CBA males follows from the fact that (CBA × C57)F1 hybrid females reject grafts from CBA males more frequently than from C57 males8.

A polyoma virus-induced CB strain hamster tumour has been defined with regard to its immunogenetic specificity (§ 3.1) and evaluated as a convenient source of cells from which a crude but biologically active, strain-specific 9transplantation antigen’ can easily be pre­pared. This antigen is resistant to repeated freezing and thawing, lyophilization and sub­sequent storage at 5°C (§ 3.3). Although the extracted antigen elicited ‘ direct hypersensitivity ’ reactions when inoculated intracutaneously into MHA strain hamsters specifically presensitized against CB tissue antigens by means of skin homografts or lymphoid cellular inocula, it was incapable of inciting transplantation immunity in normal MHA hosts. However, when emulsified in complete Freund’s adjuvant and injected into the footpads of MHA recipients, the material was immunogenic as evidenced by ( a ) their accelerated rejection of subsequent test tumour or skin homografts, ( b ) their prompt rejection of skin homografts hitherto sustained in a privileged site, and ( c ) their capacity to give direct reactions when challenged intracutaneously with cells or antigen extract from CB strain donors (§ 3.4). Exposure of normal MHA lymphoid cells to CB antigen in vitro altered their behaviour in such a way as to suggest a prompt transformation to an ‘immune’ or specifically activated status—such cells, after careful washing, were capable of inciting ‘immune lymphocyte transfer’ reactions in the skins of MHA and (MHA x CB) F 1 hosts. However, various findings leave open the possibility that the sensitization actually took place in the skin of the host animal into which the cells were transferred after their exposure to the antigen in vitro (§ 3.7). Chronic treatment, over a 90-day period, of adult MHA hamsters with CB antigenic material administered intracutaneously and intraperitoneally, without adjuvant, induced a high degree of tolerance in respect of subsequent homografts of CB skin (§ 3.8).

Hamster cheek pouch skin, transplanted to the side of an isogenic host's chest wall, retains its immunologically privileged status as evidenced by the prolonged survival of inlaid homografts of ordinary skin. Various findings sustain the premise that exemption from rejection by otherwise susceptible homografts in both intact pouch tissue and in established pouch skin isografts is due to an impediment in the afferent pathway of the immunologic reflex, i.e., to deficient lymphatic drainage. Although lymphatics were not apparent when dye was injected into pouch skin grafts or into grafts of ordinary skin sustained by them, lymph vessels were readily and consistently revealed by dye injected into intact trunk skin or established isografts of trunk skin. When suspensions of viable lymph node cells from specifically sensitized parental strain donors were injected superficially into either the intact skin or established grafts of normal skin on F1 hybrid hamsters, a striking hypertrophy of the regional lymph nodes occurred, due to graft-versus-host reactivity. However, similar cell suspensions inoculated into intact pouch tissue or into pouch skin grafts on F1 hamsters incited no regional lymphadenopathy, indicating the lack of appropriate pathways to the nodes. When skin homografts were inlaid eccentrically into pouch skin isografts, so that they were in contact with host skin at one edge, rejection occurred. Furthermore, rejection of long-established intrapouch skin homografts resulted if the hosts received: (a) small homografts of ordinary skin transplanted to conventional beds; (b) suspensions of donor strain pouch skin epidermal cells, injected intracutaneously; (c) lymph node cells from specifically sensitized donors of the same strain, i.e. adoptive immunization; or, (d) if a portion of the target homograft's perimeter was surgically approximated to body skin. Treatment of normal hamsters with two closely spaced pulses of ALS, although only marginally effective in prolonging the lives of homografts of trunk skin, enabled pouch skin homografts to survive for very long periods. The influence of this brief treatment with immunosuppressant was still demonstrable if challenge of hosts with the weakly immunogenic pouch skin homografts was delayed for 100 days.

ALTHOUGH there has been a considerable number of reports dealing with the susceptibility of dissociated cellular homografts of many tissues, notably those derived from the lymphohaematopoietic system, to humoral isoantibodies1, the majority of attempts to demonstrate that humoral factors play any significant part in the destruction of skin and other solid tissue homografts have failed2. Indeed, most of the available evidence sustains the premise that a ‘cellular’ immunity is involved and that the demise of a homograft is causally related to its characteristic infiltration by mononuclear cells.

Without altering the structural integrity of the placenta by irradiation or drugs, we have shown that it is possible to immunize females both adoptively and actively against the paternally inherited transplantation antigens of their fetuses. Such immunization causes a high incidence of runt disease among the litters. Although the putative chimeric status of the affected offspring has yet to be confirmed, the results of our experiments support the thesis that runt disease is caused by the activities of "unwanted" immigrant lymphocytes from the maternal circulation. Our results suggest that immunologically activated cells are more likely to cross the placenta than normal cells and that this greater mobility may not be related to the immunologic specificity of the activated cells. Two factors may have contributed to the apparent failure of numerous previous attempts to demonstrate the capacity of transplantation immunity to affect the well-being of a fetus or, more correctly, its placenta, in the way that might be expected of a homograft. (i) Investigators were preoccupied with obtaining a classic type of rejection, in utero, analogous to the rejection of an orthotopic skin homograft. The birth of consistently healthy-looking litters, interpreted as a failure of the experiment, convinced the investigators of the efficacy of nature's solution of the homograft problem and there was no reason for them to suspect its possible limitations. Observation of the litters for several weeks might have uncovered the phenomenon of maternally induced runt disease. (ii) Most investigators resorted to hyperimmunization of the mothers. This would have facilitated the synthesis of protective isoantibodies capable of interfering with the expression of the potentially harmful cellular immune response (6). Ever since the abnormalities of runt disease were first described they have repeatedly been compared to those observed in patients with certain lymphomas (17). Various theories have been propounded as to how maternally transmitted graft-versus-host reactivity might lead to the development of these tumors. In mice it has been established that graft-versus-host reactivity may result in a high incidence of lymphomas (18). Recent analysis indicates that this graft-versus-host reactivity unmasks and activates normally latent and undemonstrable oncogenic viruses (19). The work we describe in this article may have some relevance to the possible clinical significance of transplacental cellular mobility in man. We suggest that the relatively high incidence of lymphomas in children might also be, in part at least, due to unmasking of oncogenic viruses by subclinical graft-versus-host reactivity mediated by immunocompetent cells of maternal origin. The statistical evidence that male infants are at greater risk than females (20) is concordant with our observation that maternally induced runts include a significantly higher proportion of males than females (10).

The epidermolytic syndrome that can be obtained at will in F(1) hybrid hamsters by the cutaneous inoculation of adequate doses of parental strain lymphoid cells has been investigated to determine whether the cutaneous lesions are due to an autoimmune process arising from the severe, initial GVH reactions in the skin. It was amply demonstrated that inoculation of donor cells into the skin was of crucial importance to the development of epidermolysis. Parental strain lymphoid cells in similar doses delivered by any other route into normal F(1) hybrids failed absolutely to incite the acute syndrome. If "immune lymphocyte transfer" reactions incited by donor cells in the host's skin were surgically removed at timed intervals after inoculation, only complete excision within 24 hr prevented the appearance of epidermolysis in F(1) hybrid hosts, indicating that inoculated donor cells must remain within the confines of the skin for approximately 24 hr in order to evoke the disease, persistence for longer periods of time being unnecessary for the subsequent course of the disease. However, reconstitution experiments involving the intramuscular inoculation of suspensions containing mixtures of donor cells and host lymphoid cells, in the presence or absence of epidermal cells, unequivocally indicated that no intimate exposure of lymphoid cells to putative skin-specific antigens was essential. Similarly, the elicitation of generalized epidermolysis in F(1) hybrids irradiated with 300 r and then inoculated intravenously with donor cells casts further doubt on the pathogenic importance of the skin as a source of tissue-specific antigen. The results of subsequent experiments indicated that host leukocytes, rather than parenchymal cells of the dermis or the epidermis, were important contributors of the transplantation antigenic stimulus. Moreover, a series of experiments, using (CB x MHA)F(1) hybrid hosts that had been lethally irradiated and reconstituted with bone marrow cells from ALS-treated MHA donors, indicated that from 6 to 10 wk after reconstitution-when direct and immune lymphocyte transfer reactions showed a virtual absence of native F(1) leukocytes from the circulation-donor cells obtained from specifically sensitized MHA donors were completely ineffective in inducing epidermolysis, while equivalent lymphoid cell inocula derived from CB donors evoked the cutaneous disease irrespective of the time elapsed since reconstitution. To explain these findings it is postulated that in hamsters, the primary targets in graft-versus-host disease incited by the intracutaneous inoculation of donor cells are leukocytes originating in bone marrow or lymph node, or both.

TISSUES and organs are not all equally vulnerable to transplantation immunity. Renal homografts sometimes succeed in conditions in which homografts of skin would have failed1 and there is suggestive evidence that liver grafts may be less exacting in their genetical requirements than those of kidney, especially in pigs2. A few reports during the early days of human cardiac transplantation prompted speculation that this organ might be superior to kidney with regard to its capacity to override histoincompatibility barriers. It was suggested that: (1) because the heart is essentially a muscular pump, without complex metabolic functions, it might be less vulnerable to immunological insult than kidney; and (2) being composed chiefly of muscle and devoid of epithelial components, hearts might be “less antigenic” than organs with large amounts of epitheial tissue such as kidneys.

White skin from a spotted black and white guinea pig may be caused to turn black by grafting it to a pigmented area. Conversely, black skin grafted to a white area slowly blackens the white skin that surrounds it1,2. Recent experiments by Billingham and Medawar2 have set aside the possibility that the blackening reaction is the outcome of a mere diffusive process, or of a mass replacement of white epidermal epithelium by black. We have now examined the possibility, for various reasons unfairly belittled in our earlier communication, that the blackening reaction is due to the differential migration of melanoblasts from pigmented into unpigmented skin.