resistance of fetal pvg rats to induction of allograft tolerance

Eur. J. Immunol. 1989.19: 787-793 Resistance of fetal PVG rats to induction of allograft tolerance 787

Peter McCullagh Resistance of fetal PVG rats to induction of allograft Department of Immunology, John tolerance Curtin School of Medical Research, Australian National University, Canberra Artificially induced immunological tolerance is often envisaged as experimental

mimicry of events which occur naturally during the acquisition of self tolerance. Thus, it has been inferred that similar circumstances will facilitate the induction of both types of unresponsiveness. Since tolerance to most self determinants appears to have been established by the time of birth and allograft tolerance can be experimentally induced more readily, if not exclusively, in very young animals, it is commonly assumed that susceptibility to induction of tolerance to foreign antigens will continue to increase as progressively younger animals are tested. The experiments reported in this study extend the range of circumstances under which younger animals may be found to be less susceptible, or even completely resistant, to tolerance induction. While neonatal PVG rats were highly susceptible to the induction of tolerance to (PVG x DA)F1 hybrid determinants, fetal PVG rats were only partially susceptible at 19 days and were completely resistant at 17 and 18 days of gestation. Furthermore, only 39% of neonatal PVG rats that had been inoculated with F1 hybrid bone marrow cells at 18 days of gestation could be rendered tolerant by procedures which were effective in 100% of previously untreated neonates. These observations were inter- preted to indicate that a certain stage of maturity must be attained before fetal rats become susceptible to tolerance induction. Asymmetry in susceptibility to tolerance induction in response to the inoculation of F1 hybrid bone marrow cells was observed when PVG and D A fetuses, but not neonates, were compared. This difference in the response of the two strains recalls the divergence in immune reactivity of offspring, previously reported when PVG and DA embryos were transferred to surrogate mothers of the other strain. It is also in agree- ment with earlier reports of substantial differences in the ease of tolerance induction between different strain combinations. The asymmetry between PVG and D A rats in the present experiments is speculatively attributed to differences in rates of immunological maturation of the two strains.

1 Introduction

An inevitable outcome of the short interval between the prediction that unreactivity to foreign determinants could be induced by exposure of the immature immune system to them [ 11 and the experimental fulfillment of that prediction [2] was that tolerance to self and to foreign determinants was assumed to have much in common. For instance, it was inferred that, as non-self tolerance was more readily inducible in young animals, an increasing susceptibility to its induction would be observed as progressively younger animals were examined.

Direct comparison of the relative susceptibility of fetal and neonatal animals to tolerance induction has been rare. Experi- ments in which neonatal animals were exposed to antigen have usually been posited in such a way that induction of tolerance would be regarded as a positive result. Consequently, instances in which early exposure to an antigen did not interfere with subsequent responsiveness to it, as in the case of chick embryos challenged with influenza virus [3] were likely to be dismissed as failures [2].

[I 65321

Correspondence: Peter McCullagh, Department of Immunology, John Curtin School of Medical Research, Australian National University, GPO Box 334, Canberra, A.C.T. 2601, Australia

Abbreviations: BMC: Bone marrow cell LN: Lymph node LNC: LN cell

Interpretation of experiments intended to demonstrate increased fetal susceptibility to tolerance induction is not necessarily clear cut as, if antigen remains available over an extended period, the actual age at which tolerance induction has occurred remains uncertain [4]. The complexity of the interaction between immunological maturity and antigen dosage was emphasized by the finding that weight-adjusted doses were much more effective than a fixed dose of allogeneic spleen cells in inducing tolerance in mice during the 1st week after birth [5]. Another indication that tolerance susceptibility is not simply a function of an animal’s age was provided by the demonstration that either tolerance or immunity of allogeneic cells could be induced in neonatal mice by appropriate adjust- ment of cell dose [6].

In contrast with the predicted outcome of an inverse relation- ship between maturation and tolerance susceptibility, both the successful immunization of very young animals and an increase in susceptibility to tolerance induction with increasing age of pennatal animals have been reported. In addition to immunization of chick embryos against influenza virus [3], immunization against this antigen has been achieved in 14-day fetal mice [7]. Similarly, Rees and Garbutt [8] successfully immunized 15-day fetal mice against Bacillus Calmette-GuCrin (BCG). The subsequent development of immunity to Border Disease and Akabane viruses can occur despite the presence of micro- organisms in the tissues of fetal lambs from a very early age [9, 101. Waters et al. [ l l ] noted that susceptibility of fetal mice to induction of tolerance to protein antigens failed to increase with immaturity, animals at 13 days of gestation displaying a

0 VCH Verlagsgesellschaft mbH, D-6940 Weinheim, 1989 001 4-2980/89/0505-0787$02.50/0

788 P. McCullagh

selectivity in induction of tolerance to protein antigens similar to that of adult animals.

Eur. J . Immunol. 1989.19: 787-793

Apart from poor correlation between immaturity and susceptibility to tolerance induction, the latter has been observed to increase with age of the immature animals in some situations. Yolk sac inoculation of the 9-day-old chick embryo with Sal- monella pullorum did not induce tolerance whereas i.v. exposure of 15-day chicks markedly curtailed their postnatal responsiveness [12]. High doses of S. paratyphi B induced par- tial paralysis of specific antibody response in 5-day-old, but not in neonatal, rabbits [13]. Failure of susceptibility to tolerance induction to match immaturity was highlighted by some instances in which weight-adjusted dosage thresholds for induction of tolerance of neonates to pneumococcal polysac- charide were higher than for adults [14].

The present study presents data on the induction of allograft tolerance in fetal rats which suggests that the attainment of a certain stage of immunological maturation is a pre-requisite for this event.

2 Materials and methods

2.1 Animals

Rats of the inbred PVG (RTI') and DA (RTIa") strains, together with (PVG x DA) and [PVG x JC (Lewis)]F1 hybrids, were used in these experiments [The JC (Lewis) strain is RT-1'1. Fetal rats of the required ages were produced by timed matings. The day on which a vaginal plug was observed was counted as day 0. Both PVG and DA strains have a gestation period of 22 days.

2.2 Preparation of cell suspensions

Bone marrow cell (BMC) suspensions were prepared from F1 hybrid donors by syringing out femoral and tibia1 marrow cavities with heparinized Hanks' saline. Thymus cell suspensions were prepared by disrupting tissue from newborn rats on a stainless steel grid. Lymph node cell (LNC) suspensions for testing were prepared in a similar manner from putatively tolerant rats.

2.3 Inoculation of fetal and neonatal rats with cell suspensions

When fetal rats were to be inoculated with cell suspensions, the pregnant rat was anesthetized with ether and the uterus was exposed at laparotomy. The cells to be injected were introduced into the peritoneal cavity of each fetus in a volume of 50 p1. While accurate placement of inocula in the cavity could be readily verified visually in the case of 17- and 18-day fetuses, this was more difficult at 15 and 16 days. Newborn rats were inoculated, either i.v. or i.p. as indicated in the text, with 150 pi cell suspension.

2.4 Assessment of immunological tolerance

Assessment of immunological tolerance was undertaken by skin grafting of the subject followed by examination of the reactivity of its LNC in the popliteal LN assay of Ford et al.

[15]. Observation of skin grafts commenced with removal of the dressing early in the 2nd week after p1acement.Unless specifically indicated otherwise in the text, grafts that were being accepted had completely healed in at the end of 1 month. Rats were retained for several months after skin grafting, and for a longer period if acceptance remained in doubt. When using the popliteal LN assay, inocula of 5 x lo6 viable LNC from each donor were injected S.C. into each hind paw of two (PVG x DA)F1 or (PVG x JC)Fl hybrid rats of from 4 to 8 weeks of age. In the absence of any footpad chal- lenge, the mean weight of 14 popliteal LN from (PVG X DA)FI hybrid rats was 5.6 k 0.1 mg, while 17 LN examined 7 days after the injection of 0.25 ml cell-free saline weighed 7.9 k 0.4 mg.

Statistical significance of differences in the frequency of induction of allograft tolerance following different forms of treatment of rats was assessed by means of the x2 test.

3 Results

3.1 Induction of tolerance in fetal DA rats by the injection of (PVG X DA)FI hybrid BMC

A series of experiments was undertaken in which the ability of inocula of (PVG x DA)F1 and (PVG X JC)F1 hybrid BMC to induce allograft tolerance in fetal DA and PVG rats of different gestational ages was examined.

Ninety-three fetal DA rats, of between 15 and 18 days of gestation, were each inoculated i.p. with 5 X lo7 (PVG x DA)F1 hybrid BMC. This dose of BMC was arbitrarily selected on the grounds that it was not only equivalent to, but in excess of (in weight-adjusted terms), the dose (10' cells) which regularly induces tolerance after i.v. transfer to DA and PVG neonates. It was also a survivable inoculum. All members of some of the inoculated litters, amounting to 37 rats in total, were lost before birth. Litters in which complete loss occurred had all been inoculated at 15 or 16 days of gestation.

Table 1 summarizes the immunological status, as assessed by skin grafting, of the 35 liveborn DA rats. When graft rejection

Table 1. Rate of acceptance of (PVG X DA)F, hybrid skin grafts by DA rats following inoculation with (PVG X DA)F, hybrid BMC in utero

No. of rats Gestational No. of No. of age when fetuses .inoculated accepting F, inoculated"' inoculated rats born hybrid skin (days) grafts

15/16 46b' 6 2 17 26 17 11 18 21 12 11

Fetuses received an i.p. inoculum of 5 x lo7 F, hybrid BMC at the indicated age. All survivors were grafted with F, hybrid skin, after testing the popliteal LN reactivity of their lymphocytes, in the 2nd or 3rd month of postnatal life. The popliteal LN reactivity of their lymphocytes is summarized in Fig. 1. Of the DA fetuses inoculated at 15 or 16 days of gestation, loss of entire litters accounted for 37 rats. The survivors were derived from 9 rats in two litters inoculated at 15 days.


In Utero Inoculation Neonatal Inoculation Normal

A

AA

AA

A

A

I /UW 6 7 9 10

Age at Assay (weeks)

Figure 1. Popliteal LN responses of lymphocytes from normal rats and from DA rats inoculated with (PVG X DA)FI hybrid BMC either between 15 and 18 days of gestation in ufero or on the day of birth. Each point represents the mean of four observations of popliteal LN response to 5 x lo6 lymphocytes. Rats which had been inoculated in utero were each examined at a single time between 6 and 10 weeks of age as indicated on the horizontal axis. Normal rats and rats that had been inoculated as neonates were examined as adults. (+) Rat inoculated in utero, rejected F1 hybrid skin graft; (0) rat inoculated in utero, accepted F1 hybrid skin graft; (W) rat inoculated as neonate, all of this group accepted skin grafts; (A) normal rats, not skin grafted.

by these rats occurred, this was invariably indicated by failure of healing 3 weeks after placement. Contraction of the graft bed was sometimes delayed for a further 1 or 2 weeks but, once started, was rapidly completed. The frequency of tolerance induction increased with gestational age of recipients at the time of BMC inoculation. The anti-PVG reactivity of LNC from these rats was examined between 6 and 10 weeks of age, before skin graft placement, using popliteal LN assay in (PVG x DA)Fl hybrids. Lymphocyte reactivity determined in this way was almost invariably in accord with the alloreactivity of each rat as indicated by its response to a skin graft. The mean popliteal LN response to a dose of 5 x lo6 lymphocytes was 6.5 f 0.4 mg in the case of the 24 D A rats that accepted skin grafts, whereas that for the 11 rats which rejected grafts was 14.0 f 2.0 mg. Popliteal LN assay results for rats inoculated in utero are summarized and compared with those of normal and “conventionally tolerant” D A rats in Fig. 1. A small groups of D A fetuses was inoculated with (PVG x DA)F, hybrid BMC at 19 days of gestation. Of nine survivors, three accepted F, hybrid skin grafts as adults.

3.2 Resistance of fetal PVG rats to the induction of tolerance by injection of F1 hybrid BMC

The outcome of inoculation of fetal PVG rats of from 15 to 18 days gestational age with F, hybrid BMC was completely different from that described above for D A rats. The administration of an identical dose of 5 x lo7 cells or, in the case of 50 fetuses, of 8 x lo7 (PVG x DA)Fl hybrid BMC failed to produce a single tolerant rat. Apart from the two rats described below, graft rejection occurred at the same tempo as in nontolerant D A rats that had been inoculated in utero. The frequency of tolerance induction in PVG fetuses differed significantly from that observed in DA fetuses in the preceding experiment (p < 0.001). Of 119 PVG fetuses inoculated in the first experiment, 37 were subsequently liveborn (Table 2). All of the liveborn rats had been inoculated at either 17 or 18 days

of gestation with 5 x lo7 or 8 X lo7 BMC (none of 39 fetuses treated at 15 or 16 days of gestation survived).

The popliteal LN reactivity of lymphocytes from the 37 surviv- ing PVG rats inoculated in utero is summarized in Fig. 2 and compared with that of normal PVG rats and of PVG rats rendered tolerant to F, hybrid tissues by the “conventional” approach of inoculation as neonates. In two instances, the popliteal LN reactivity of rats inoculated in utero was reduced to the range of rats with neonatally induced tolerance. How- ever, both of these in utero inoculated rats rejected (PVG x DA)FI hybrid skin grafts, albeit at a slightly slower tempo than did the remaining 35 rats in this group. While regrowth of fur had occurred by the end of the first month, alopecia and progressive breakdown of the surface took place in the 3rd month leading to complete rejection in the 4th.

In view of the inability to induce tolerance to (PVG x DA)F, hybrid tissues in any of the PVG fetuses inoculated at 17 or 18 days of gestation, and of the susceptibility of PVG neonates to tolerance induction, an attempt was made to measure the stage of development at which tolerance could first be induced. The handling required to exteriorize the uterus and inject a cell suspension was invariably followed by loss of 20- day fetuses. However, 19-day fetal rats could survive to term after inoculation of a weight-adjusted inoculum of 4 x lo7 BMC. The doses of F1 hybrid BMC administered to 17- and 18-day fetuses in the original experiment had been substantially in excess of the weight-adjusted equivalent to the dose of 10’ cells that is regularly effective in induction of tolerance in neonates. Table 3 indicates the basis for calculation, of weight-adjusted doses administered to 17- to 19-day fetuses.

The mean popliteal LN assay result produced by 5 x lo6 LNC from ten adult PVG rats which had been inoculated with (PVG x DA)Fl hybrid cells at 19 days of gestation was 57.9 f 8.0 mg. After placement of F, hybrid skin grafts, five rats rejected these at the rate to be expected of normal rats. FI

P. McCullagh

PVG Rats:

t t t

t

t +

t t t t +t

t

h Utero hoculatii

t

t * t + +

t t

t +

t

t t

5 6 7 9 / I

Age at Assay (weeks)

A

Eur. J. Immunol. 1989.29: 787-793

t Neonatal lnoculation Normal

Figure 2. Popliteal LN responses of lymphocytes from normal PVG rats and from PVG rats inoculated with (PVG x DA)F, hybrid BMC either at 17 or 18 days of gestation in utero or on the day of birth. Each point represents the mean of four observations of popliteal LN response to 5 X lo6 lymphocytes. Rats which had been inoculated in ufero were each examined at a single time between 5 and 9 weeks of age as indicated on the horizontal axis. Normal rats and rats that had been inoculated as neonates were examined as adults. (+) Rat inoculated in ufero, rejected F, hybrid skin graft; (*+) rat inoculated in utero, rejected F, hybrid skin graft 2 weeks later than remainder of group; (W) rat inoculated as neonate, all of this group accepted skin grafts; (A) normal rat, not skin grafted.

hybrid skin grafts on two of the remaining five rats did not produce fur but appeared to be healing well until 25 days after placement, when rejection abruptly became evident. The remaining three grafts had acquired new fur by 25 days and remained healthy after 53 days. However, two of these grafts were then rejected over the following 3 weeks. The tenth graft remained very healthy 80 days after placement even though lymphocytes from this rat retained reactivity equivalent to that of cells from a normal PVG rat (47.2 mg) in the popliteal LN assay.

To determine the extent to which the resistance of fetal PVG rats to the induction of tolerance to (PVG x DA)FI hybrid

Table 2. Survival of PVG rats following inoculation with (PVG x DA)FI hybrid BMC in uteroa)

determinants could reflect a peculiarity of this strain combination, a similar experiment was performed to investigate the responsiveness of fetal PVG rats to (PVG X JC)Fl hybrid cells. Of seven PVG rats receiving lo8 (PVG X JC)F1 hybrid BMC i.v. on the day of birth, six accepted F1 hybrid skin grafts as adults. The mean popliteal LN response of (PVG x JC)Fl hybrid recipients to a dose of 5 x lo6 LNC from these tolerant donors was 3.1 f 0.5 mg. In contrast, 18 out of 20 adult PVG rats that had been inoculated with 5 x lo7 ((PVG x JC)F1

Table 3. Body weights of PVG fetuses and neonates

Age No. of rats Mean Dose of F, examined’) weight k SEM hybrid BMCb)

(mg)

Gestational age No. of fetuses No. of when inoculated’) inoculated inoculated rats (days) born

Fetus, 17 days 21 0.66 k0.03 2x107 Fetus, 18 days 11 1.22 kO.04 3 X 10’ Fetus, 19 days 10 1.894 k 0.05 4 x lo7 Newborn 50 5.035 k 0.05 1 x loB

15/16 17 18

39 48 32

0 22 15

a) Each fetus was inoculated i.p. with 5 X lo7 or 8 x lo7 F, hybrid BMC at the indicated times. None of the survivors accepted skin grafts. The popliteal LN reactivity, in F, hybrid recipients, of lymphocytes from the survivors is summarized in Fig. 2.

a) Examination of fetal DA rats revealed mean weights of 0.67k 0.06 mg (five rats at 17 days), 1.19 k 0.08 mg (three rats at 18 days) and 2.05 k 0.08 mg (nine rats at 19 days of gestation).

b) Cell doses indicate the number of cells administered when a weight-adjusted dose was used. The exact equivalent number of cells (to a neonatal dose of 108) would be 1.31 X lo’, 2.42 X lo7 and 3.76 X lo7 for average-sized fetuses of 17, 18 and 19 days, respec- tively.


hybrid BMC at 18 days of gestation rejected F1 hybrid skin grafts at the same rate as rats that had received (PVG x DA)Fl hybrid cells (mean popliteal LN response of rats rejecting skin grafts was 20.7 k 2.8 mg and that of those accepting grafts was 3.9 k 0.2 mg). The administration of 5 x lo7 F1 hybrid BMC to PVG fetuses at 19 days of gestation resulted in acceptance of skin grafts by three out of seven as adults. The effects on immunological reactivity of exposure of fetal PVG rats to (PVG x JC)FI and (PVG x DA)F1 hybrid BMC thus appeared to be similar.

The results of the preceding attempts to induce allograft tolerance in PVG fetuses implied that susceptibility to tolerance induction required the attainment of a certain stage of maturity of the immune system. The possibility that supplementa- tion of the fetal immune system with cells from more mature, syngeneic donors might assist tolerance induction was tested. Accordingly, a group of 17-day, fetal PVG rats were given mixed inocula of 2 X lo7 (PVG x DA)FI hybrid BMC together with lo7 thymus cells from newborn PVG donors. The survival rate of these fetuses was even lower than that of rats injected with BMC alone. All fetuses in six litters were lost, while only 6 of the 19 rats in the two remaining litters attained adult life. Lymphocytes from these 6 rats were highly reactive in the popliteal LN assay (76.0 k 14.4 mg in response to 5 x lo6 cells). All rats rejected skin grafts at a normal rate.

3.3 Efficacy of the i.p. administration of BMC for the induction of tolerance in newborn PVG rats

Apart from the age of the inoculated rats, the other factor which was considered as a possible reason for failure of an adequate cell dose to induce tolerance in PVG fetuses before 19 days of gestation was the route of cell administration. Neonatal rats have invariably been inoculated i.v. to induce tolerance, whereas fetal rats in the present experiment received BMC i.p. To clarify this point, the relative efficacy of

8 8 - 8

8

the i.p. and i.v. routes of administration of BMC to newborn rats was examined. Following i.p. inoculation with lo8 (PVG X DA)FI hybrid BMC on the day of birth, 28 out of 41 PVG rats were subsequently assessed as tolerant on the basis of indefinite acceptance of skin grafts and greatly reduced lymphocyte reactivity in popliteal LN assay. Fig. 3 summarizes the results of popliteal LN assays performed with lymphocytes from these rats. This rate of tolerance induction compares with a success rate of 100% in 21 PVG rats inoculated i.v. with 10' cells during the same month (p < O.Ol).The relative efficacy of i.v. and i.p. routes of inoculation of newborn DA rats with (PVG x DA)FI hybrid BMC in tolerance induction was similar to that observed in PVG rats. Thus, all 16 rats that had received F1 hybrid cells via the i.v. route accepted skin grafts while only 14 out of 22 rats that had been inoculated i.p. did so (p c 0.01). Over an extended period of years, success rates with the i.v. injection of lo8 BMC have always exceeded 90%. Consequently, substitution of the i.p. for the i.v. route of BMC administration reduced the efficiency of tolerance induction in neontal PVG and DA rats by approximately one third.

3.4 In utero administration of BMC modifies, but does not abolish, susceptibility to tolerance induction as neonates

The absolute resistance of PVG fetuses, exposed before 19 days of gestation, to induction of tolerance to (PVG x DA)Fl hybrid tissues cannot be attributed to inade- quacies of either the BMC dose or the route of administration selected. To examine the nature of this resistance, susceptibility to tolerance induction of neonates that had already been exposed to semi-allogeneic BMC in urero was examined. A group of newborn PVG rats which had received F1 hybrid cells in urero were reinoculated. This second procedure, when undertaken in isolation on the day of birth, reliably produces

I Figure 3. Popliteal LN responses of lymphocytes from normal PVG rats and from PVG rats inoculated on the day of birth with (PVG X DA)F, hybrid BMC using i.p. or i.v. routes df administration. Each point represents the mean of four observations of popliteal LN response to 5 X 106 lymphocytes. All rats were examined as adults. (*) Rat inoculated i.p. (graft accepted); ( -. e) rat inoculated i.p. (graft rejected); (A) rat inoculated i .v.; (W) normal rats.

Figure 4. Popliteal LN responses of lymphocytes from rats inoculated with (PVG x DA)FI hybrid BMC in utero, at 18 days of gestation, and again on the day of birth. Each point represents the mean of four observations of popliteal LN response to 5 x lo6 lymphocytes. (-+) Rat rejected graft; (h) rat died after LNC collected but before placement of skin graft. (Remaining rats accepted grafts.)

792 P.McCullagh

allograft tolerance. Given the reproducible failure to establish tolerance in PVG rats by inoculation at 17 and 18 days of gestational age, any rat found to be tolerant after receiving both in utero and neonatal inocula could have become SO only in consequence of the latter.

A total of 27 PVG rats attained adult life following the i.p. injection of 3 X lo7 (PVG X DA)F1 hybrid BMC at 18 days of gestation and the i.v. injection of lo8 cells on the day of birth. Popliteal LN assays were undertaken using LNC from all of these rats. Two thirds of these tests gave the low level of response typical of tolerant rats (Fig. 4). (PVG x DA)Fl hybrid skin grafts were then placed on all rats. Four of the rats died accidentally and, of the 23 remaining, 9 accepted grafts. LNC from 6 of the 14 rats that rejected skin grafts had mani- fested low reactivity in popliteal LN assays (6.2-10.3 mg). (In the case of 2 of these 6 rats with reduced popliteal LN reactivity, graft status had been doubtful 4 weeks after placement but the rapid rejection of second grafts placed at this time, together with the original grafts, facilitated their classifica- tion.) On retesting after graft rejection, cells from 3 of the 6 rats that rejected grafts despite low levels of reactivity retained the latter (6.8-7.3 mg) while the other 3 were clearly no longer tolerant (19.3, 27.2, 29.1 mg). Both of the rats which had rejected two successive skin grafts remained in the poorly reactive group. Whereas in utero exposure to semi-allogeneic BMC did not prevent the subsequent induction of tolerance in neonates, it reduced the frequency of this occurrence from 21 out of 21 to 9 out of 23 (p < 0.001).

Eur. J. Imrnunol. 1989.19: 787-793

had been a selective survival of nontolerant rats. If the volume and placement of inoculum required to induce tolerance was lethal for fetal rats, only those fetuses in receipt of treatment inadequate for tolerance induction would have survived to be tested for immunological reactivity. However, given the contrast between the approximate comparability of survival rates of DA (61%) and PVG (47%) fetuses inoculated at 17 and 18 days but the very dissimilar frequencies of nontolerance of 24% (DA) and 100% (PVG) among survivors, the complete absence of tolerant PVG rats cannot be explained as a selective loss. Cell doses were carefully controlled and the accura- cy and ease of inoculation of PVG and DA fetuses did not differ. Consequently, the explanation that tolerance of (PVG x DA)Fl hybrid cells was inducible in 17- and 18-day DA, but not PVG, fetuses seems inescapable. The experiment in which PVG fetuses were inoculated with (PVG x JC)Fl hybrid cells implied that the resistance of these rats to tolerance induction was not confined to DA-derived antigens. Once again there was a very low rate of tolerance induction to cells of this second F1 hybrid genotype in 17- and 18-day PVG fetuses in comparison with neonates.

4 Discussion

Allograft tolerance can be induced in DA and PVG rat neonates with equal facility by injection of (PVG X DA)Fl hybrid BMC. However, whereas the injection of (PVG x DA)Fl hybrid cells at 17 to 18 days of gestation successfully induced tolerance in 22 of 29 fetal DA rats, identical treatment failed to induce tolerance in any of 37 fetal PVG rats. Furthermore, exposure of 18-day fetal PVG rats to F1 hybrid BMC reduced their susceptibility to induction of tolerance following further exposure to these cells as neonates. Fetal PVG rats appeared to be similarly unresponsive to attempts to induce tolerance to (PVG X JC)FI hybrid tissues when compared with newborn rats. Inoculation of BMC of either F1 hybrid genotype induced incomplete tolerance of varying duration in a minority of 19-day PVG fetuses. The doses of BMC which were not tolerogenic when injected into fetal rats substantially exceeded, on a weight-adjusted basis, doses that were adequate for PVG neonates. Although the i.p. route of BMC administration that was used in fetal rats was found to be less efficient for inducing tolerance in neonates then the i.v. route, it nevertheless proved effective in more than two thirds of these.

Two issues raised by these results relate specifically to the protocol while others have wider implications for the mechanisms operative in tolerance. The specific issues concern the extent to which “technical” factors could explain resistance of PVG fetuses to tolerance induction before 19 days of gestation and the basis for divergence in responsiveness of DA and PVG fetuses.

The possibility was considered that tolerance had been induced in some PVG fetuses before 19 days,but that there

The possibility that fetal PVG rats were resistant to the induction of tolerance because of the use of the i.p. route of administration of F1 hybrid cells was considered. While the i.p. route was not as efficient in tolerance induction, as the i.v. route, the difference was certainly not sufficient to explain the failure to induce tolerance in PVG fetuses. Furthermore, tolerance was successfully induced in DA fetuses using the i.p. route.

Another question peculiar to the specific experimental protocol concerned the asymmetry in tolerance susceptibility of PVG and DA fetuses. Asymmetry in the facility of tolerance induction between inbred strains was described in one of the earliest reviews of the experimental induction of allograft tolerance [16]. The peculiar aspect of the asymmetry in the present experiments was that it was manifest only when induction of tolerance was attempted in fetal rats. Neonatal PVG and DA rats are both readily susceptible to induction of tolerance to the other genotype. It seems more likely that asymmetry in tolerance induction that is restricted to fetal rats reflects different rates of immunological maturation with fetal DA rats attaining the state required for tolerance induction several days earlier.

Resistance of fetal PVG rats to tolerance induction may have been predicted from experiments in which rat embryos were transferred to allogeneic surrogate mothers [17]. PVG rats born to DA surrogates always expressed significantly greater anti-DA reactivity than did normal PVG rats, implying that they may have been sensitized before birth. In those experiments, 5 x lo6 lymphocytes from normal PVG donors evoked a popliteal LN response of 32.7 k 1.2 mg in (PVG x DA)Fl hybrid recipients. However, a similar number of cells from PVG rats born to DA surrogates produced a response of 51.0-e 1.4 mg. In contrast, DA rats which had been transferred to PVG surrogates possessed normal reactivity. Popliteal LN reactivity of cells from PVG rats that had been inoculated in utero in the present experiments was within the normal range implying that sensitization had not occurred.

As it was not possible to produce large numbers of rats which had been inoculated with F1 hybrid BMC before 17 days of gestation, the distinction between variation in timing of immune development and absolute differences between PVG

bur. J. Immunol. 1Y89.19: 787-793 Resistance of fetal PVG rats to induction of allograft tolerance 793

and D A fetuses as explanations for differences in susceptibility to tolerance could not be drawn. Interstrain variations in rate of immunological maturation could imply that DA fetuses would become increasingly insusceptible to tolerance induction before 17 days.

The general relevance of observations which are not equally applicable to all strains of a species may justifiably be ques- tioned. However, it is equally justifiable to point out that conventional beliefs about allograft tolerance are substantially based on results observed only in some strain combinations and their acceptance as generalizations requires that some conflicting observations be ignored. For example, in first reporting induction of allograft tolerance in immature mice, Billingham et al. [2] acknowledged their fortunate and acci- dental choice of strain combination as tolerance could not be so readily induced in other combinations. It is also worth not- ing that some of the observations which were most influential in establishing the concept of heightened susceptibility to tolerance induction in utero were misinterpreted. For example, the inability to detect anti-viral activity in mice infected with lymphocytic choriomeningitis in utero [18] was taken to indicate that tolerance had been induced. The subsequent demonstration that mice exposed to virus in fetal life were actually responding vigorously [19] did not occasion any ques- tioning of the established belief that fetal animals were espe- cially susceptible to tolerance induction.

The resistance of PVG fetuses to induction of tolerance pro- vides a further example of the necessity of attaining a certain stage of immunological development before allograft tolerance can be induced. Previous instances of such a requirement have been cited above [12-141, together with situations in which embryonic and fetal animals proved resistant to tolerance induction [3 ,7 , 8, 111. The reduction in rate of tolerance induction to 40% in neonatally inoculated PVG rats which had previously been inoculated as fetuses suggests that some animals may have remained unaffected by the first exposure but that the remainder had sustained an otherwise undetected modification in immunological development. The occurrence of these two forms of response within the group may reflect the entry of 18-day fetal PVG rats into a transition phase.

The present results carry implications for hypotheses that inte- grate tolerance to self and foreign determinants within a common framework. If self tolerance is dependent on exposure of the developing immune system of each self antigen, as appears to be the case with thyroid gland determinants in fetal rats [20], exposure to self antigen before a certain stage of development of the immune system may not induce tolerance. If common mechanisms were responsible for tolerance to self and

foreign determinants, this would be predicted. Alternatively, if self tolerance could be shown to be inducible earlier than allograft tolerance, this would suggest that different mechanisms were operating. As regards self antigens that appear rela- tively late in development, it could be speculated that tolerance to them might depend o n mechanisms shared with tolerance to foreign antigens which differ from those responsible for tolerance to earlier-appearing self antigens. When origi- nally describing tolerance, Billingham, Brent and Medawar [2] envisaged the existence of a “null” period that intervened between the stage at which a young animal was susceptible to tolerance induction and its acquisition of the capacity to be immunized. This concept was subsequently abandoned when the importance of the interplay between age and antigen dosage became apparent [6] . It may be that the present experiments indicate that some form of null period precedes susceptibility to tolerance rather than following it.

Received December 1, 1987; in final revised form January 19, 1989.

5 References

1 Burnet, F. M. and Fenner, F., Theproduction of antibodies, Mac-

2 Billingham, R. E. , Brent, L. and Medawar, P. B., Nature 1953.

3 Burnet, F. M., Stone, J . D. and Edney, M., Aust. J . Exp. Biol.

4 Pike, B. L., Kay, T. W. and Nossal, G. J. V., J . Exp. Med. 1980.

5 Brent, L. and Gowland, G . , Nature 1961. 192: 1265. 6 Howard, J . G. and Michie, D., Transplant. Bull. 1962. 29: 1. 7 Nossal, G. J . V. , Aust. J . Exp. Biol. Med. Sci. 1957. 35: 549. 8 Rees, R. J. W. and Garbutt, E. W . , Immunology 1961. 4: 88. 9 Barlow, R. M., Brain Res. 1983. 59: 255.

Millan, Melbourne 1949, p. 103.

172: 603.

Med. Sci. 1950. 28: 291.

152: 1407.

10 Hashiguchi, Y . , Naba, K. and Kumagai, T., Nutl. Inst. Animal Health Q. 1979. 19: 1.

11 Waters, C. A. , Diener, E., Singh, B . , Hraba, T. and HaSek, M. (Eds.), Cellular and Molecular Mechanisms of Immunological Tolerance, Marcel Dekker, New York 1981, p. 483.

12 Buxton, A. , J . Gen. Microbiol. 1954. 10: 398. 13 Sterzyl, J . and Trnka, Z., J. Hyg. Epidemiol. Microbiol. lmmunol.

14 Howard, J . G. and Hale, C., Eur. J . lmmunol. 1976. 6: 486. 15 Ford, W. L. , Burr, W. and Simonsen, M., Transplantation 1954.

16 Billingham, R. E. and Brent, L., Philos. Trans. R . SOC. London

17 McCullagh, P., Aust. J . Exp. Biol. Med. Sci. 1985. 63: 645. 18 Traub, E., J. Exp. Med. 1939. 69: 801. 19 Oldstone, M. A. and Dkon, F. J., Science 1967. 158: 1193. 20 Eishi, Y. and McCullagh, P., Immunology 1988. 64: 319.

1957. 1: 292.

10: 258.

[B] 1959.242: 439.

resistance of fetal pvg rats to induction of allograft tolerance

Documents