ABSTRACT
Leakage of chromium-labelled erythrocytes from the circulation of pregnant mice into the foetal circulation could not be detected, either following X-irradiation at various dose-levels and gestational stages or in untreated pregnancies. Two exceptions were encountered out of a total of 279 foetuses examined. This was taken as an estimate of the spontaneous incidence of transplacental leakage of maternal erythrocytes and contrasted with the relatively high incidence encountered in the human species.
INTRODUCTION
Exposure of new-born mammals and birds to foreign cells can induce a specific immunological non-reactivity (‘tolerance’) towards grafts of donor tissues later in life. In the case of mammals, if large numbers of maternal cells were to leak into the foetal circulation during pregnancy, one possible immunological consequence would be that the offspring would be tolerant of the mother’s tissues in later life. In the human species there is evidence that leakage of maternal cells can occur on a sufficient scale to induce immunological tolerance of rhesus blood-group antigens, detectable in some rhesus-negative offspring of rhesus-positive mothers (Owen, 1957). In the mouse, however, Billingham, Brent, & Medawar (1956) cite the results of skin-grafting from mother to young as evidence that spontaneous leakage occurs very rarely, if at all.
Mitchison (1953) showed in mice, and Woodruff (1957) in rats and rabbits, that the experimental induction of a state of immunity in the pregnant female against her own young is without detectable effect upon the latter’s survival. A reasonable inference is that, at least in the case of immunologically competent cells, massive mother-to-foetus transmission does not occur.
Lengerová (1957) was able to produce tolerance in rats by irradiating the mother’s uterus with 200 r. on day 15 of pregnancy. The maternal cells responsible for inducing tolerance in this case might be derived from the maternal component of the damaged placenta itself, or alternatively they might be circulating cells passing into the foetus through placental defects caused by the treatment.
Moulton, Stimpfling, & Storer (1960) attempted, without success, to repeat, ascertain whether tolerance could be induced by orthodox procedures in the donor-recipient combination which they used. It is therefore not easy to evaluate the results of their experiment.
If irradiation damages the placenta so that cells pass from the maternal into the foetal circulation, the leak should be demonstrable by injecting labelled cells into the mother’s blood and recovering them in the blood of the young. In the work to be described, mice were exposed to X-rays in varying doses at varying times during pregnancy (see below). Whole blood labelled with radioactive chromium was transfused into pregnant animals within 2 days of their irradiation. Under the conditions employed, the overwhelming majority of the chromium is fixed by the erythrocytes in the inoculum (Mollison & Veall, 1955). The radioactivity of maternal blood, of whole new-borns, and of the blood of certain new-borns was compared on the day of delivery.
MATERIALS AND METHODS
Pregnant mice from a random-bred stock were irradiated at the times and doses stated below. Only the abdomen was exposed to the beam, the remainder of the body being shielded by of an inch of lead. Röntgen values shown are calculated at the anterior abdominal wall and include the expected contribution of back-scatter. A Westinghouse radiotherapy unit was operated at 230 kV., 15 mA., at a distance of 25 cm., with 0·5 mm. Cu and 1 mm. Al filters, giving a half-value layer of 1·2 mm. Cu.
In most cases 5·4–8·0 μc. of radiochromium activity were delivered either intravenously or intraperitoneally in 0·15–0·20 ml. of whole blood. For labelling, 4 volumes of whole blood were incubated for 1 hour with 1 volume of an isotonic solution of sodium chromate containing 27 μg. Cr, of approximately 200 μc. activity, per ml. Judged by the activity of subsequent blood samples, intraperitoneal injection was about one-third as effective as intravenous injection. In a few cases 0·2 ml. of a solution of Na251CrO4, having 8–10 –c. activity, was injected directly into a vein, but blood samples proved to be weakly labelled by this method.
At all dose levels above 400 r., mice were given 2·5 mg. of progesterone in oil subcutaneously on day 18 to ensure maintenance of pregnancy until the following day, permitting (a) the anticipated effects of radiation on the placenta to operate for a standard time, and (b) the offspring to be delivered by Caesarian section so that the uterus could be examined for dead foetuses.
Radioactivity was measured in a Panax well-type scintillation counter at settings of 1,000 E.H.T. volts, 10 discriminator volts, and 100 seconds. Counts shown have been corrected for background radiation. Whole new-borns, or blood samples, were placed in the bottom of a test-tube within the well of the counter. Blood samples ranged from 0·040 to 0·065 ml., but were equal for each mother and her offspring. Combination of blood from two new-borns was sometimes necessary to make 0 04 ml. Blood was collected from adults by tapping the retro-orbital venous plexus with a thick-walled capillary pipette, and from new-born mice by decapitation.
RESULTS
Twenty-nine pregnant mice were irradiated in single doses which ranged from 184 to 1,000 r. Two hundred and five live and 74 dead young were counted in all. The general outline of the experiment is shown in Table 1.
An unexpected observation was that almost all young, when counted as whole animals, showed weak activity, ranging up to 1,300/100 sec. with an average of 355.
However, when the bloods of the seemingly most active young were counted, no significant activity was present, with two exceptions to be described below. The source of the radioactivity found has not been determined, although skin, stomach with milk content, liver, and spleen have been found to be inactive.
Two exceptional results occurred with mice receiving 400 r. on day 13 and day 15. In the latter case, the maternal blood gave 6,041/100 sec. counts, one of seven whole new-borns gave 5,994, and its blood 2,657. The counts of the other six live whole new-borns in that litter ranged from 343 to 526, and the blood of one of them gave 16 counts. No activity was detected among the six new-borns of another litter treated with 400 r. on day 15, or in any of the off spring of mothers receiving the same or smaller doses on days 16 and 17 of pregnancy.
The mouse which was treated on day 13 was unfortunately poorly labelled and her blood on day 19 gave only 500/100 sec. counts. Two members of the litter of 14 were alive, and their pooled blood gave 148 counts. Since the standard error of this figure is 37, the excess of activity over background must be accepted as real. But because of the unsatisfactory labelling of the mother’s blood, and the negative results obtained with higher radiation doses administered on the same and following day of pregnancy, we are hesitant to regard it as an effect of the treatment. We prefer to ascribe both of these cases to spontaneous leakage of maternal cells into the foetal blood. On this basis we may estimate from our data that the frequency of this as a natural phenomenon is in the region of 1 per cent, of foetuses.
DISCUSSION
The idea that, in the mouse, the placental barrier to circulating erythrocytes can be breached by X-irradiation is not supported by our results. It also appears that spontaneous breakdown of the barrier sometimes occurs, but only with extreme rarity. This presents a striking contrast with the situation in the human species, where the transmission of erythrocytes from mother to foetus is relatively common. Many investigators, including Hedenstedt & Naeslund (1946), Naeslund & Nylin (1946), Naeslund (1951), Mengert et al. (1955), and Macris, Hellman, & Watson (1958), have been able to recover maternal erythrocytes labelled in various ways from the blood of the new-born infant. The estimates of frequency given by these authors show considerable heterogeneity, but by pooling their figures we can arrive at an estimate which approaches 50 per cent, of all infants tested. This may be contrasted with the estimate of about 1 per cent, of foetuses tested in our material.
The degree of mixing found in the two positive cases can be estimated as 44 per cent, and 30 per cent, respectively. These values are surprisingly high. On the other hand, they fall near the centre of the range of values found in human material by the authors previously cited. Even if all estimates based on the use of radioactive tracers are ruled out of court as being conceivably subject to difficulties of interpretation, we are still left with the results obtained by Macris et al. who exploited hereditary sickle trait as an erythrocyte marker. Two of their three positive cases betrayed a degree of admixture in the region of 20 per cent., confirming that transplacental leakage, when it occurs, is by no means always a minor or marginal phenomenon.
RÉSUMÉ
Expériences sur la barrière placentaire chez la souris: I. Test pour la transmission des érythrocytes maternels à travers le placenta de la souris, après irradiation aux rayons X
Le passage d’érythrocytes marqués au chrome, de la circulation de souris gestantes vers la circulation fœtale n’a pu être détecté, d’une part après irradiation par des doses variées de rayons X et à différents stades de la gestation, d’autre part chez des souris non traitées. Deux exceptions ont été enregistrées sur un total de 279 fœtus examinés. Ce résultat est considéré comme une évaluation de l’incidence spontanée du passage transplacentaire des érythrocytes maternels et contraste avec l’incidence relativement élevée enregistrée dans l’espèce humaine.
ACKNOWLEDGEMENTS
We wish to acknowledge the interest taken in this work by Professor M. F. A. Woodruff. One of us (M. F.) was supported as a visting research worker in the Department of Surgical Science, University of Edinburgh, by the University of Rochester School of Medicine.