ABSTRACT
The teratogenicity of rabbit anti-rat visceral yolk-sac antiserum, injected into pregnant rats at either 8 ·5 or 9 ·5 days of gestation, has been confirmed. Normal rabbit serum was found not to be teratogenic.
When conceptuses from 9 ·5-day pregnant rats were cultured for 48 h in heat-denatured homologous serum, to which antiserum was added for the final (or the penultimate) 6 h of culture, embryonic development was normal. The protein contents of embryos and yolk sacs (at harvesting) were however decreased.
When antiserum was present in cultures for the final 6 h, pinocytosis by the yolk sac, as measured by the uptake of 125l-labelled polyvinylpyrrolidone (PVP), was decreased to an extent related to the concentration of antiserum in the culture medium and to a minimum level of about 40%. The presence of antiserum in cultures for the penultimate 6 h only, with 125I-labelled PVP present for the final 6 h only, produced an identical result. No uptake of radioactivity into the embryo was observed, in either the absence or presence of antiserum.
When conceptuses were cultured for the final 6 h in vitamin- and glucose-supplemented dialysed homologous serum whose proteins were [3H]leucine-labelled, the presence of antiserum for either the final or penultimate 6 h again resulted in a decrease in the uptake of radioactivity by conceptuses. Uptake of radioactivity into yolk sac and embryo was decreased by the same amount, indicating that proteolysis in yolk-sac lysosomes was not inhibited.
In parallel control experiments in which normal rabbit serum replaced rabbit anti-rat visceral yolk-sac antiserum, no effects on embryonic development, on protein contents of yolk sacs and embryos at harvesting, or on the uptake of radioactivity by conceptuses were observed.
These results are interpreted as providing evidence that teratogenic antibodies decrease pinocytosis of protein by visceral yolk sac at the early organogenesis stage and consequentially decrease the availability of amino acids and thus protein synthesis in both yolk sac and embryo. It is proposed that this effect constitutes the mechanism of action of teratogenic antisera.
INTRODUCTION
In a series of papers Brent and colleagues have reported the teratogenic effects in rat of certain maternally administered antisera. The finding (Slotnick & Brent, 1966) that sheep anti-rat kidney antiserum, when administered to pregnant rats at 8 days of gestation, localized strongly in the visceral yolk sac, together with evidence that the yolk sac plays a nutritional role during early organogenesis (Beck, Lloyd & Griffiths, 1967a) led to a suggested mechanism of action. It was proposed (Brent, Johnson & Jensen, 1971 ; Lloyd, 1970) that the antisera might cause developmental anomaly by interfering with the uptake or digestion of macromolecular nutrients by the epithelial cells of the visceral yolk sac. The same authors drew attention to the parallels with teratogenesis by trypan blue, which had been attributed (Beck, Lloyd & Griffiths, 1967 b) to the same proposed mechanism. Further evidence that the visceral yolk sac was the primary target in antiserum-induced teratogenesis came from experiments by New and Brent (1972) using rat embryos cultured in vitro and by Jensen, Koszalka and Brent (1975), who showed that adsorption by visceral yolk sac, but not by Reichert’s membrane, destroys the teratogenic potency of antisera. Kobrin & Brent (1973) briefly reported that teratogenic antisera alter the distribution of radioactivity in rat conceptuses following an injection of free and bound [14C]leucine; they interpret their data as indicating that the antiserum interferes with intracellular digestion of protein and/or the trans-yolk-sac passage of small molecules or macromolecules.
Freeman, Beck & Lloyd (1981) have recently described experiments that provide direct evidence that thé 9 ·5- to 11 ·5-day rat embryo obtains amino acids from proteins that are pinocytosed by the yolk sac and digested within that tissue’s lysosomes. The present paper describes the effects of a teratogenic antiserum on this nutritional route in the early rat embryo.
MATERIALS AND METHODS
Rabbit anti-rat visceral sac antiserum
A New Zealand albino rabbit received four weekly injections each of 5 –6 mg of lyophilized 18-day rat visceral yolk sac, as described by Brent et al. (1971). γ-Globulin comprised 30 % of the final antiserum preparation.
Teratogenesis experiments
Pregnant Wistar rats received a single intraperitoneal injection of anti-rat visceral yolk-sac antiserum or normal rabbit serum. Each injection contained 100 mg serum protein per kg body weight. At 20 ·5 days the uterus was removed, and the numbers of resorption sites and live foetuses recorded. Each foetus was weighed and examined for evidence of external malformation. Two out of three foetuses were then fixed in Bouin’s fluid, and sectioned freehand with a razor blade to detect soft tissue malformations, each third foetus being cleared with KOH and stained with alizarin red-S to reveal skeletal anomalies, following the procedure of Wilson (1965).
Cultures of 9 ·5-day rat egg cylinders in the presence of radiolabelled macro-molecules
The techniques were as described by Freeman et al. (1981). The total culture period in all cases was 48 h, with antiserum or normal rabbit serum present either for the final 6 h or for the penultimate 6 h (i.e. from 36 to 42 h from the start of the culture).
(i) I-Lab el led polyvinylpyrrolidone (PVP) as substrate
The radiolabelled substrate (2 ·3 μg/ml) was present only for the final 6 h of culture. Where antiserum or normal rabbit serum was present for the final 6 h, serum and 125I-labelled PVP were simply added together at 42 h from the start of the culture. Where antiserum was present for the penultimate 6 h, it was added at 36 h from the start of the culture, and at 42 h the conceptuses were transferred to fresh rat serum (without antiserum) containing 125I-labelled PVP.
(ii) [3H]Leucine-labelled serum proteins as substrate
When antiserum or normal rabbit serum were present for the final 6 h of culture, conceptuses were cultured for 42 h in normal rat serum and then transferred to vitamin and glucose-supplemented dialysed [3H]leucine-labelled rat serum (Freeman et al. 1981) containing antiserum or normal rabbit serum. Where antiserum was present for the penultimate 6 h, it was added at 36 h from the start of a culture in normal rat serum, and at 42 h the conceptuses were transferred to vitamin- and glucose-supplemented [3H]leucine-labelled rat serum (without antiserum).
RESULTS
Teratogenicity of anti-rat visceral yolk-sac antiserum
Table 1 is a summary of the results obtained and Table 2 an analysis of the types of malformation detected.
Teratogenic effects of intraperitoneal injection of rabbit anti-rat yolk antiserum or normal rabbit serum Teratogenic antiserum-effect on rat embryo cultures

Abnormalities detected after rabbit anti-rat yolk-sac antiserum or normal rabbit serum treatment

Treatment with normal rabbit serum at either 8 ·5 or 9 ·5 days of gestation resulted in a very low incidence of foetal resorption and malformation. In contrast, anti-rat visceral yolk-sac antiserum caused high rates of foetal resorption and abnormality. Following injection at 8 ·5 days, anti-rat visceral yolk-sac antiserum caused malformations in all foetuses surviving to 20 ·5 days. The average weight of these foetuses was significantly lower than the average weight of foetuses surviving maternal injection of normal rabbit serum at 8 ·5 days. Treatment at 9 ·5 days with anti-rat yolk-sac antiserum was lethal to more than 70% of foetuses. Of the surviving foetuses, nearly all were malformed. Table 2 shows that the relative frequency of the different malformations produced by antiserum was similar whether treatment was at 8 ·5 or 9 ·5 days of gestation. Abnormalities of the eyes, the cerebral ventricular system, the urogenital system and the axial skeleton were particularly common.
Effects of teratogenic antiserum on uptake of radiolabelled macromolecules in vitro
The inclusion of antiserum or normal rabbit serum in the culture medium for either the final or penultimate 6 h of a 48 h culture did not decrease the yield of viable conceptuses or cause developmental deviations. However, the protein contents at harvesting of yolk sacs and embryos of conceptuses incubated for the final 6 h of culture in a medium containing antiserum were decreased (Table 3). Higher concentrations of antiserum were needed to effect a significant decrease in embryonic than in yolk-sac protein. The presence of normal rabbit serum had no effect on protein content at harvesting (Table 3).
(i) Uptake of125I-labelled PVP
Table 4 shows the effect of various concentrations of antiserum or normal rabbit serum on the uptake of 125I-labelled PVP by conceptuses. As previously found (Freeman et al. 1981), little or no radio-activity was detectable in the embryo. With antiserum present in the culture medium for the final 6 h of a 48 h culture, radioactivity in the yolk sac was 61 % of the control value, at an antiserum concentration of 167 μg/ml. At higher concentrations of antiserum, uptake was decreased further but less sharply,tending to a minimum value of about 40 % of the control. No effect was seen with normal rabbit serum.
Radioactivity associated with yolk sacs and embryos after incubation of 9 ·5-day rat conceptuses in vitro for 48 h with 125I-labelled PVP present for the final or the penultimate 6 h

For all subsequent experiments two concentrations only of antisera (83 and 333 μg/ml) were used, those inducing intermediate and near-maximal inhibition of uptake of 125I-labelled PVP. When conceptuses were exposed to these anti-serum concentrations for the penultimate 6 h period, the extent to which the uptake of radioactivity into the visceral yolk sac was inhibited (Table 4) was virtually identical to that seen when antiserum and radiolabelled substrate were present in the culture medium at the same time.
(ii) Uptake of [3H]leucine-labelled serum proteins
Freeman et al. (1981) showed that, if conceptuses are cultured in vitamin- and glucose-supplemented dialysed [3H] leucine-labelled rat serum, uptake by the visceral yolk sac is followed by intracellular digestion of labelled protein and the accumulation in both visceral yolk sac and embryo of proteins that have incorporated [3H] leucine. Figure 1 shows that antiserum (83 μg serum protein/ml) present in the culture medium for the final 6 h of a 48 h culture period did not affect either the total uptake of radioactivity or the distribution of radiolabel between the embryo and the visceral yolk sac. The percentage of tissue radioactivity that was soluble in 6 · 7% (w/v) trichloroacetic acid (TCA) was also unchanged (results not shown). However, at a concentration of 333 μg/ml, antiserum, but not normal rabbit serum, decreased the total amount of radioactivity taken up by conceptuses to a level of 62 % of the control value. The distribution in the conceptus of radiolabel, and the proportion that was TCA-soluble, was not altered. When conceptuses were exposed to antiserum for 6 h before being transferred to serum containing [3H] leucine-labelled serum proteins and cultured for a further 6 h, essentially identical data were obtained (Figure 2). The lower concentration (83/μg/ml) of antiserum had no effect on the incorporation of radiolabel into the tissues of conceptuses, whereas the higher concentration (333 μg/ml) decreased the total uptake of radioactivity by conceptuses to 55 % of the control value. As before, the distribution of radiolabel and the proportion that was TCA-soluble were not altered.
Radioactivity associated with yolk sac (hatched) and embryo (plain) after incubation for the final 6 h of culture in vitamin- and glucose-supplemented dialysed serum whose proteins were [3H]leucine-labelled, containing either anti-rat visceral yolk-sac anti-serum (1 A) or normal rabbit serum (IB). Each left-hand column expresses uptake as a function of yolk-sac protein content, and each right-hand column expresses uptake as a function of embryo protein content. Values are means (with S.D.) of at least six determinations.
Radioactivity associated with yolk sac (hatched) and embryo (plain) after incubation for the final 6 h of culture in vitamin- and glucose-supplemented dialysed serum whose proteins were [3H]leucine-labelled, containing either anti-rat visceral yolk-sac anti-serum (1 A) or normal rabbit serum (IB). Each left-hand column expresses uptake as a function of yolk-sac protein content, and each right-hand column expresses uptake as a function of embryo protein content. Values are means (with S.D.) of at least six determinations.
Radioactivity associated with yolk sac (hatched) and embryo (plain) after incubation for the penultimate 6 h of culture in the presence of anti-rat yolk-sac anti-serum, and for the final 6 h of culture in vitamin- and glucose-supplemented serum containing [3H]leucine-labelled proteins. Each left-hand column expresses uptake as a function of yolk-sac protein content and each right-hand column expresses uptake as a function of embryo protein content. Values are means (with S.D.) of at least eight determinations.
Radioactivity associated with yolk sac (hatched) and embryo (plain) after incubation for the penultimate 6 h of culture in the presence of anti-rat yolk-sac anti-serum, and for the final 6 h of culture in vitamin- and glucose-supplemented serum containing [3H]leucine-labelled proteins. Each left-hand column expresses uptake as a function of yolk-sac protein content and each right-hand column expresses uptake as a function of embryo protein content. Values are means (with S.D.) of at least eight determinations.
DISCUSSION
Our first concern was to demonstrate the teratogenic potency of anti-rat visceral yolk-sac antiserum in the strain of Wistar rat to be used for the sub-sequent experiments on rat embryos cultured in vitro. The observed incidence of foetal resorptions and malformations and the nature of the malformations induced are entirely consistent with earlier reports (Brent et al. 1971; Jensen et al. 1975).
When Cockroft (1979) showed that 9 ·5-day rat embryos could grow and develop in vitro in a medium comprising dialysed rat serum to which only glucose and certain vitamins had been added, it was apparent that the amino acids needed for synthesis of embryonic proteins must be derived from protein and not from free amino acids. Our recent experiments (Freeman et al. 1981) identify the source and supply route : exogenous serum proteins are captured by pinocytosis into the visceral yolk-sac epithelial cells, whereupon they are digested within the lysosomes to amino acids, which are used for protein synthesis in both the visceral yolk sac and the embryo.
The present experiments have shown that a teratogenic antiserum to visceral yolk sac decreases the pinocytic uptake of macromolecules by the visceral yolk sac of early-organogenesis-stage rat embryos cultured in vitro. When [3H] leucine-labelled protein was the macromolecule used, antiserum (at the higher concentration) decreased the extent of incorporation of [3H]leucine into both embryo and visceral yolk sac. These effects were not seen in conceptuses exposed to normal rabbit serum. The data afford convincing evidence in support of the hypothesis that teratogenic antibodies perturb normal embryogenesis by inhibiting uptake of macromolecular nutrients by the visceral yolk sac and in consequence decreasing the supply of nutrients to the embryo at a critical stage of its development. The precise target in the yolk sac appears to be the process of pinocytosis : the data indicate that the intralysosomal digestion of ingested macromolecules and the incorporation of digestion products into the embryo proceeds normally, although at a decreased rate owing to the inhibition of pinocytosis. The proposed mechanism of action precisely parallels that advanced to explain the teratogenic action of trypan blue (Williams et al. 1976).
It is of interest to compare the effective concentrations of antiserum in the present experiments with those reported to exert teratogenic effects in vivo (Brent et al. 1971 ; Jensen et al. 1975), and in vitro (New & Brent, 1972). In vivo, no malformations, embryonic death or growth retardation were observed in pregnant rats given less than 50 mg/kg of teratogenic yolk-sac antiserum. The concentration of teratogenic antiserum in the extracellular fluid when administered at 50 mg/kg (foetal LD0), 100 mg/kg (LD50) and 125 mg/kg (LD70) are 125 μg/ml, 250μg/ml and 312 μg/ml. In vitro, the no effect concentration was 33 μg/ml and the first concentration that produced embryonic growth retardation was 333 μg/ml of culture media. Table 5 summarizes these data and shows that there is good correspondence between concentrations of antiserum that are teratogenic and those that markedly decrease the incorporation of protein-derived [3H]leucine into embryonic protein in vitro.
There is an apparent inconsistency in the data from experiments using the lower concentration (83 μg protein/ml) of teratogenic antiserum. Although there was no effect on uptake of 3H-labelled serum proteins by the yolk sac or on the incorporation of radiolabel into the embryo, this concentration of antiserum inhibits pinocytosis of 125I-labelled PVP into the visceral yolk sac. At present we have no explanation for this anomaly.
It is interesting that the effectiveness of teratogenic antibody in inhibiting pinocytosis by the visceral yolk sac is as great if the antibody is present in the penultimate 6 h of culture as if it is present, together with the radiolabelled macromolecule, for the final 6 h. Antibody is ingested by the visceral yolk-sac cells, presumably by pinocytosis into the lysosomes, where it remains undigested for several days (Brent et al. 1971). The demonstration (Schneider, Tulkens, De Duve & Trouet, 1979) that membrane recycling between lysosomes and plasma membrane can carry with it anti-plasma membrane antibody probably explains how an intralysosomal material can continue to affect an event at the plasma membrane. It could also explain how antibody injected into rats at 2 ·5 or 4 ·5 days of gestation induces malformations whose nature indicates an effect on embryonic development several days later (Brent, 1966). The mechanism by which antibody inhibits pinocytosis remains to be resolved.
The antibody used in our experiments was prepared using whole rat visceral yolk sac as antigen. Experiments with antibody prepared against various sub-cellular fractions of the visceral yolk sac (M. Jensen, T. R. Koszalka & R. L. Brent, unpublished data) indicate a wide distribution of the antigen or antigens.
Acknowledgements
We thank the British Medical Research Council and the United States National Institutes of Health for grants in support of this work. We also thank Dr T. R. Koszalka and Mrs M. Jensen for their collaborative efforts on this project.