ABSTRACT
The work of Waddington and Waterman (1933) has shown that it is possible to cultivate young rabbit embryos in vitro, using the normal watch-glass technique and a medium composed of chicken embryo extract and either chick or rabbit plasma. Micro-surgical operations can be performed on such embryos before they are explanted into culture, and the mutual interactions of the various embryonic rudiments can be investigated in this way. It may eventually be possible to obtain as complete an account of the developmental mechanics of the mammalian embryo as Spemann and his school have provided in the Amphibia, but it is probable that operations on the mammalian embryo will always be attended with considerable difficulty. Apart from the laboriousness of the tissue culture technique with its concomitant necessity for absolute sterility, the rabbit embryo, if not every mammalian embryo, is very difficult to handle owing to its transparency, toughness and stickiness. The transparency of the blastocyst and embryonic shield in the primitive streak stage makes it difficult to distinguish exactly the various embryonic organs ; the toughness of the tissue hinders neat cutting and operating ; while the stickiness of the cells is probably the most annoying characteristic of all, since it results in any fragment of tissue, which may for example be intended for a graft, adhering to the operating knives with such tenacity that it frequently becomes entirely macerated during the attempts made to free it.
It is, however, quite possible to separate the endoderm and ectoderm and to place small fragments of tissue in the pocket so formed between these two layers. As in the chick and duck (Waddington, 1930, 1932) such grafts develop, at least in the best cases ; although in the rabbit development of the grafts, as of the complete embryo, is not so frequent nor so successful as it is in the birds. Grafts of the rabbit primitive streak have yielded neural tissue and notochord, and are thus apparently similar to grafts of chick primitive streak in their prospective significance and self differentiating capacity.
Owing to the small size of the embryonic shield in the primitive streak stage, and to the above-mentioned stickiness of rabbit tissue, it is difficult to make grafts of the rabbit primitive streak which lie within the shield. Usually, after the graft has been pushed into the cavity between the ectoderm and endoderm, it sticks to the knife point and is partly pulled away from the desired position as the knife is withdrawn: the grafts then usually lie in the trophoblastic area. Whether owing to this circumstance or not, such grafts do not usually succeed in producing inductions. Only two doubtful cases of induction by rabbit primitive streak have been obtained, and in both of them it is impossible to be certain whether the superfluous neural tissue is induced, i.e. originally host tissue, or whether it is derived from the graft.
Grafts have also been made of chick primitive streak into rabbit blastoderms and vice versa. The tissues as a rule develop fairly normally in the foreign surroundings and often show little sign of any species-specificity. Rabbit tissues grafted into chick blastoderms show a tendency towards the formation of many pycnotic nuclei, but this phenomenon is also seen in the unoperated rabbit embryos grown in vitro, and is probably due to some general deficiency of the culture conditions rather than to a specific action of the chick blastoderm. This deficiency might be due to the use of chick embryo extract and chick plasma, but Waddington & Waterman (1933) found that little improvement was obtained in the length of life of embryos grown in vitro when rat extract was substituted for chick extract, and recent experiments using mixtures of chick and rabbit plasma have also failed to give any better results. In both the specimens of chick grafts into rabbit embryos described here the grafted tissues have formed hollow sacs, but this behaviour is not by any means universal in such grafts. Hiraiwa (1927), Murphy (1913), and Nicholas & Rudnick (1931), using the chorio-allantoic technique, found no tissue incompatibility between birds and mammals till a considerably later stage of development.
Owing to its comparatively slight tendency to stick to the operating knives, chick tissue can fairly easily be grafted into the embryonic shield of the rabbit. In two cases, grafts of the chick primitive streak have succeeded in inducing the formation of a neural tube from the overlying rabbit ectoderm.
In one case, 33–174a, the rabbit blastoderm, which was aged days after copulation, had a long primitive streak. A fragment of chick primitive streak was grafted into the left side of the embryonic shield, and the shield together with some of the surrounding trophoblast was cultivated on a medium composed of two drops of chick embryo extract, two drops of rabbit plasma and four drops of chick plasma. The culture was fixed on the second day of cultivation. The chick graft can be found in the sections, lying close against the left side of the neural groove of the host : it has differentiated into notochord and some neural tissue, and a part of it has formed a hollow sac (Pl. II, fig. 11 b). Over the anterior part of the graft mass a subsidiary neural plate appears in the host ectoderm. For most (Pl. II, fig. 11b), but not all (Pl. II, fig. 11 a), of its length, this induced plate is united laterally with the host plate, but it is clearly to be regarded as a separate structure and not as a mere irregularity of the host plate.
In the other specimen, 33–201, the rabbit embryo had a long primitive streak or perhaps the beginning of a head process. The endoderm was torn away from the posterior part of the embryonic shield during the extraction of the egg from the uterus. Two grafts of chick primitive streak were pushed into pockets between the ectoderm and endoderm, one on each side, about opposite the middle of the primitive streak. In two days the rabbit embryo had developed a closed neural tube with optic vesicles. The graft on the right side is lost somewhere in the abundant mesoderm of the host, or in the thickened trophoblastic area. The other graft can be seen on the left side : it has formed a hollow sac, near which is an empty space with some degenerating cells. The host ectoderm above this is folded downwards, and from one side of this fold a strand of neural tissue has been induced (Pl. II, fig. 12). This strand of tissue is solid except in one place, where the sections are unfortunately not suitable for photography since the plasma part of the sections has partly obscured them. But in this region a lumen is present and the connection of the induced neural tissue with the ectoderm is more typical.
These two experiments show that in the rabbit, as in the chick and duck among birds, and in many species of Amphibia, the ectoderm in the gastrula stage is able to react to inducing stimuli by the formation of neural tissue. The suggestion that inducing agents are therefore present in the rabbit, and are localised in the primitive streak is perhaps logically unjustified but may nevertheless very likely prove to be correct.
These experiments are also a demonstration of the extraordinary lack of species specificity in embryonic organisers. Geinitz (1925) and Bytinski-Salz (1929) have shown that inductions are possible in xenoplastic transplantations between Urodeles and Anurans, and, after the completion of the experimental work on which this paper is based, Holtfreter (1933 b) showed that there are inducing agents in adult animal tissues and that these are very non-specific in their range of action. But the inductions between the birds and mammals reported here seem to cover the widest range of all those in which actual embryonic organisers have been used as the inducing agents1.
SUMMARY
The communication describes two cases of induction by pieces of the chick primitive streak grafted into the embryonic shield of the rabbit ( days after copulation, long primitive streak stage) and discusses the technical difficulties raised by such experiments. The results show that the ectoderm of the rabbit at this stage possesses the capacity of reacting to an inducing stimulus by the formation of neural plate (i.e. is “competent” (Waddington, 1932) to form neural plate), and thus make it probable that the embryonic development of the mammals is influenced by factors similar to those which have become familiar in the Amphibia and birds. They are also a demonstration of the lack of species specificity of the inducing agents involved in embryonic organisers.
REFERENCES
DESCRIPTION OF PLATES
Figs. 1–7 show inductions in the area opaca of the chick blastoderm; Figs. 8–10 inductions by coagulated organisers in the chick blastoderm; and Figs. 11 and 12 inductions performed in rabbit blastoderms by implanted fragments of chick primitive streak.
Plate I
Fig. 1. 33-3. Section through induced head region.
Figs. 2a, b. 33–76. Fig. 2a. Section through anterior (E) end of induced plate, lying within the area opaca vasculosa. Fig. 2b. Section through posterior end of induced plate, lying on a process projecting above the surface of the blastodes.
Figs.3a-c. 33–84. Fig. 3a. Photographs of whole mount. Fig. 3b. Section along line b in Fig. 3a.
Fig. 3 c. Section along line c in Fig. 3 a.
Fig. 4. 33–85. Section through induced neural plate.
Fig. 5. 33–86. Section through graft mass and induced plate.
Fig. 1. 33-3. Section through induced head region.
Figs. 2a, b. 33–76. Fig. 2a. Section through anterior (E) end of induced plate, lying within the area opaca vasculosa. Fig. 2b. Section through posterior end of induced plate, lying on a process projecting above the surface of the blastodes.
Figs.3a-c. 33–84. Fig. 3a. Photographs of whole mount. Fig. 3b. Section along line b in Fig. 3a.
Fig. 3 c. Section along line c in Fig. 3 a.
Fig. 4. 33–85. Section through induced neural plate.
Fig. 5. 33–86. Section through graft mass and induced plate.
Plate II
Fig. 6. 33–117. Section through induced axis along line 8 in Text-fig. 2.
Fig. 7. 33–121. Section through induced neural plate.
Fig. 8. 32–45. Section through graft enveloped in mesenchyme.
Fig. 9. 32–331. Section through host and induced neural plates.
Fig. 10. 33–17. Section through host and induced neural plates.
Figs 11 a, b. 33–174a. Sections through host and induced plates.
Fig. 12. 33–201. Section through region of induced neural tissue.
In all figures, the letters H, G and I stand for host, graft and induced structures respectively. co.gr. coagulated graft; emb. embryonic axis;fgt fore-gut; hf. head-fold; mes. mesoderm; n.p. neural plate; n notochord ; pro. process.
Since the above was written, Hatt (1933) has obtained inductions from fragments of chick primitive streak transplanted into the blastocoel cavity of the Urodele, Triton taeniatus.
