Talk:Paper - The morphological theory of monochorionic twins as illustrated by a series of supposed early twin embryos of the pig (1922)

The Morphological Theory Of Monochorionic Twins As Illustrated By A Series Of Supposed Early Twin Embryos Of The Pig

Corner GW. The morphological theory of monochorionic twins as illustrated by a series of supposed early twin embryos of the pig. (1922) Johns Hopkins Hospital Bulletin 33: .

By George W. Corner

(From, the Anatomical Laboratory of The Johns Hopkins University)

The subject of identical twins is of such iuterest from many biological viewpoints that conjecture as to the mechanism of their origin has far outrun our actual information. Almost the whole mass of discussion now current in the literature of embryologj- and genetics is pure hypothesis constructed by reasoning backward from the observed anatomy of late stages, chiefly from the dispo sition of the foetal membranes at birth in cases of twinning. Even the term "single-ovum twins" is an assumption, when applied to mammals.

So far as is knowTi to the present writer the extant specimens of monochorionic twin embryos, in stages early enough to be of value for morphogenetic study, are (1) the blastodermic vesicle of the sheep, about 7 days old, and a similar embryo of the ferret (Putorius), described by Assheton ('98), in each of which two inner cell masses are found in one blastocyst; (2) the human embryo "Mateer" described by Streeter ('19a), in which a similar arrangement is supposed to exist at a slightly later stage of development; (3) two human specimens, each containing twin embrj'os about 12 mm. long, recently described by Arey ('22) ; and (i) the complete series of enibn,'os of the nine-banded armadillo (a species in which singleovum quadruplets are produced at every birth) which we owe to the work of Newman and Patterson ('10, '13). With Asshetou's specimens and his own as the basis for hypothesis, Streeter ('lOb) has constructed a very plausible series of diagrammatic stages of human single-ovum twin-formation, in which this phenomenon is represented as due to the production of two embryonic areas of inner cell masses on the inside of a single blastocyst, before the formation of the amnion. Such a hypothesis fits in well with the usual disposition of the foetal membranes at birth ; and it would also readily account for the appearance of monochorionic twinning in any mammalian spe cies, since all mammalian embrj^os, whatever their later divergencies of form, pass through an essentially similar blastocyst stage. Such a view of twinning would furthermore accord with the fact of general occurrence of polyembryony in the animal kingdom, and would give added philosophical value to the experimental production of uniovular twins and double monsters in fishes and amphibians.

A different view has been proposed by Newman in his interesting book "The Biologj' of Twins" ('17), upon the basis of his studies of the armadillo. In this species development proceeds as in those mammals which undergo so-called "inversion of the germ-layers," until the amnion is formed ; then the embryonic area divides into four embryos, and each embrjo migrates down a meridian of the chorion, dragging with it the surrounding portion of the amnion, until finally a single chorion contains four embryos, each enclosed in a portion of the original amnion, now drawn out into four sacs with a common central connection. It is obvious that this form of polyembryony cannot take place except in mammals in which the formation of the embryo proper is delayed until after the appearance of the amniotic cavity; and since in man a similar proigenesis of the amnion occurs, Newman has been led to believe that in man also monochorionic twin-formation follows the same scheme as the polyembryony of the armadillo. Furthermore, if in mammals monochorionic twinning is indeed dependent upon proigenesis of tlie amnion, then it cannot occur in species in which the amnion is produced by rolling up of the chorion over the embryo. Newman has been willing to make this assumption, and in his book has ventured the prediction that single-ovum twins ^^ill probably never be found in ungulates.

The three specimens of earl}' enibiyos of the pig which are to be described in the following pages all came into the author's hands within the space of a few weeks, at the packing house of Joseph Stern and Company, in New York City.* It is remarkable that all three were found among about 500 pregnant swine examined in this one abattoir, whereas similar specimens have never been seen in many hundreds of sows at other places. The packinghouse in Baltimore which has furnished pig embryos to the Johns Hopkins laboratories for thirty years has never yielded, as far as the writer knows, anything resembling monochorionic twins except one double monster and one specimen to be mentioned later, in which it was not certain whether one or two chorions were present.

Tlie following descriptions and illustrations will be clear to the reader who is not familiar with the specific embryologj' of the pig, if he will recall that the blastocyst

The writer was at the time working under the auspices of the Department of Experimental Evolution of the Carnegie Institution of Washington, at Cold Spring Harbor, Long Island. To the Director, Dr. C. B. Davenport, thanks are due for providing support and assistance.

does not remain spherical, but is rapidly elongated to 30 or more centimeters, so that by the fifteenth day the embryo itself is a small thickening upon a very narrow but immensely elongated cylindrical chorion. The chorion does not develop villi, as does the human gestation sac, but remains smooth, and can at all times be readily peeled out of the uterus. The allantois reaches such size as to fill a large part of the cylindrical chorionic cavity, and ultimately becomes adherent to the inside of the chorion, giving rise to a difl'use allantoic placentation.

Specimen 1. — The uterus of sow C-15 5 contained 14 single embryos and the twins to be described. Eight of the single embr>'os were normal at the stage of 33-34 somites (i. e. of the 3rd week). Six showed abnormalities as follows: Two were retarded, without gross defects; one was retarded, with extreme flexure; one was slightly retarded, with undistended allantois. In another the embryo itself appeared normal, but its allantois was entirely cut off from the embryo by the amnion; the embryonic stump of the allantois was closed over, and the extra-amniotic part of the allantois which was adherent to the chorion had upon it a small highly vascularized vesicle. Finally, one embryo was represented merely by an almost amorphous chorionic mass. Since there were 16 corpora lutea, it seems that one of the embryos was totally missing.

In the twin specimen (Fig. 1), two embryos lay about 18 mm. apart within one continuous chorion of dimensions quite similar to those of the normal embryos in the same uterus. Fluid injected into the chorionic cavity dilated the whole space; that is. there was no septum between the embryos. When the cylindrical chorion was slit open, it was found that both embryos were abnormal as Illustrated. In one, the allantois was entirely cut off from the embryo, and was growing in an almost unexpanded conditiou upon the chorion adjacent to the embryo. In the other, the allantois had been converted into a cluster of small irregularly dilated vesicles. The yolk-sac vessels of this embryo had not grown, as normally, out along the yolk-stalk onto the sac, but remained as a convoluted nodule on the yolk-sac near the umbilicus. That portion of the yolk-sac (or sacs) lying between the embryos had become curiously twisted and vesiculated, as shown in Fig. 1, so that it was unfortunately Impossible to determine whether the two embryos had origina.lly possessed a common yolk-sac. The two amnions seemed perfectly normal; the enclosed embryos were slightly smaller than the normal single embryos in the same uterus.

Specimen 2. — The uterus of sow C-392 contained 10 normal embryos of 11 mm. crown-rump length, and one pair of twin embryos enclosed in a single chorion (Fig. 2). There were 14 corpora lutea; three ova were therefore missing, a proportion not much above the avei^age loss of ova in this species. The twin embryos appeared quite normal, but one of them occupied, with its allantois, about two-thirds of the available chorionic cavity, leaving but one-third for the other. The embryo with the larger allantois was 9.5 mm. crown-rump length, the embryo with the smaller allantois was 8. .5 mm.; there was no significant difference in the development as indicated by external form. Injection of fluid into the allantoic cavities showed that the two were not continuous. When the chorion was opened, it appeared that the yolk-sacs, which at this stage in the development of the pig are beginning to be obliterated by pressure of the allantois, were in the twin specimen very much twisted. It was thus again impossible to determine the original relation of the yolk-sacs, for although they were entirely separate in this specimen, the separation might conceivably have been produced by torsion, as the figure clearly shows.

Specimen 3. — The uterus of sow C-66 contained 7 normal embryos of 23 mm. crown-rump length and a pair of twin embryos. There were 15 corpora lutea. Fig. 3 shows that the arrangement was somewhat different from that in the other two specimens, for here the two embryos were not placed end to end in the chorionic cavity, but side by side. The allantoides were expanded so fully that they came entirely into contact with the inner chorionic surface, obliterating the chorionic cavity. Each amnion was now surrounded by its allantois, as usual at this stage, so that each embryo with its amnion appeared to occupy a separate cavity divided from the other by a septum. The cavity was obviously the allantoic cavity and the septum was formed by the apposed walls of the two allantoides. The areas of attachment to the chorionic wall were so near together that the two umbilical cords appeared to spring from almost the same point, in the line of reflection of the allantoic septum. A section taken through this line showed clearly (Fig. 4) that the chorion was continuous over the surface, not dipping between the allantoic walls; that is, there was a single chorion containing two allantoides. Obliteration of the yolk-sacs had proceeded so far that their relation to each other could not be made out; they seemed to be separate. The embryos occupied approximately equal parts of the chorionic cavity. One was 16 mm. crown-rump length and the other 15 mm. These embryos were sufficiently advanced to permit determination of their sex by the criterion described by Spaulding ('21), which (as Dr. Spaulding assures the author) can be applied to the pig as well as to human embryos. In both these twin embryos the urethral groove runs on to the glans portion of the phallus, and they are both therefore presumably of the male sex.

Discussion

The specimens which have been described and illustrated may obviously be explained by the assumption that each is an example of twins arising within a single blastocyst, or by the alternative supposition that a deceptive appearance of twinning arose bj' the fusion, in each case, of two originally separate blastocysts. This latter possibility seems unlikely ; one would expect to find, in addition to specimens like these, other and more numerous examples not so deceptively perfect. Such fusion is at least not common in early stages of development of the pig; the present writer has never seen it during the study of hundreds of embryos of the first three weeks. It is true that after mid-pregnancy the adjacent necrotic ends of two chorionic sacs occasionally adhere or invaginate in a more or less complete way, but in such cases the larger total dimension of the combined chorions,, the considerable distance between the foetuses, and the presence of a chorionic necrosis or invagination at the region of fusion gives sufficient evidence of the dual origin of the ftetus. In the two earlier specimens under consideration there was no evidence, to the eye or under the dissecting microscope, of any chorionic septum between the embryos, and in the third .specimen there was microscopic evidence that the chorionic space was single, i.e., that the septum was formed by the internal membranes alone, without participation of the chorion. On the whole, the theory that the specimens were monochorionic twins seems more

Pig. 5.

Explanation of Pig Twinning.

Single Twin

Fig. 6.

Hypothetical Diagrams of Human Monochorionic Twins

Duplication before Duplication after

formation of formation of

amniotic lavit.v amniotic cavity

probable than that they were examples of accidental fusion of separate blastocysts with obliteration of the intervening trophoblastic (chorionic) wall. It should be mentioned at this point that in a specimen from a local packing-house, studied through the courtesy of Dr. G. L Streeter, which appeared at first sight to be another case of monochorionic twinning, in mid-pregnancj' (fcetuses 130 mm. long), dissection revealed evidences hinting at a possible origin from two separate chorions, through the extensive invagination or penetration of one by the other. The impression gained from all of these specimens taken together is in accord with the recent views of Arey (1922) formulated after study of twin tubal pregnancies as recorded in the Carnegie Embryological Collection, that the appearance of monochorionic twinning may, at times be due to the fusion of two chorions, but that true polyembryony may also occur.

Assuming that our specimens are indeed monochorionic twins, it is possible to reconstruct the stages of their formation by means of a series of diagrams involving but one or two hypothetical steps. In Fig. 5, I to IV, such a series is placed in parallel with the actually known stages of development of the single pig embryo. Stage I in the twin series is borrowed from the sheep ol Assheton's description, stages II and III are hypothetical, and stage IV is specimen 1 of the author's collection. In this scheme it is assumed that twinning is first made evident by the appearance of two embryonic areas (inner cell-masses) on the inner surface of a single blastocyst. Following this stage we may suppose that an amnion is formed about each embrj'onic area exactly as in single embryos. Meanwhile the endoderm has been spreading around the inside of the blastocyst. At the same time the blastodermic vesicle has been undergoing the great elongation which is characteristic of the pig and allied ungulate species. Whether the endoderm would usually form a single inner lining of the blastocyst, and thus provide both embryos with a single common yolk-sac, or pinch off into two yolk-sacs, is a matter for conjecture, but in preparing the accompanying diagrams the author has preferred to follow his actual specimens and to represent the yolk-sacs as doubled structures. The result, as shown in Fig. 5. stage IV, (also. Fig. 1, specimen 1| would be a single chorionic vesicle bearing separate embryos, each enclosed in its own amnion, with a yolk-sac which, for all we know at present, might be either doubled or common to both embryos. Each embi-yo would thereafter produce its own allantois, and when the allantoides had expanded to a moderate degree we would have the stage represented by Fig. 2 (specimen 2). Next the allantoides would expand so voluminously as to surround the amniotic membranes completely, as in Fig. 3 (specimen 3), in which we have a single chorion containing two allantoic cavities with the amnions invaginating the allantoic walls and the yolk-sacs passing out through the necks of the invaginations to be between the allantoic walls and the chorion. Finally, as the expansion of the amniotic cavities brought about complete fusion of the three foetal membranes, the chorionic cavity would be lined at all points by both allantoic and amniotic tissue, and would be divided into two cavities by the abutting portions of the respective allantoic-amniotic membranes.

It thus seems apparent that monochorionic twinning in ungulates is not to be considered impossible merely because in mammals of this order the embryonic area is well differentiated before the amniotic cavity arises. Proigenesis of the amnion may no doubt influence the morphology of twins, but the occasional o^ irrence of monochorionic twins is not thereby limited to any special kinds of mammals.

In conclusion, we may permit ourselves to leave the more or less solid ground afforded by our three specimens, and to indulge in a brief speculation regarding the morphogenesis of human monochorionic twins as suggested by our studies of polyembryony in the pig. We have shown the probability that, in the pig, twinning must already be apparent before the formation of -the amnion, while in the amiadillo, as investigated by Newman and Paterson, the polyembryonic budding of the four embryos awaits the formation of an amniotic cavity, which is thus common to all the embryos of one ovum. It may also be pointed out that the human embryo, like all other mammalia, must first pass through a stage of simple blastocyst structure with an inner cell mass, before it proceeds to the splitting open of the amniotic cavity. It is true, as Newman points out, that at this latter stage the morphological relations are much like those at which in the armadillo multiplication of the embryonic areas begins ; but it is also true that at the earlier blastocyst stage there is opportunity for twinning as in the pig, by the appearance of two inner cell masses instead of one. Our hypothesis, illustrated diagrammatically in Fig. 6. is that human single-ovum twins may be of two types. One sort of twinning, (Fig. 6, A. B. C.) arising by duplication of the inner cell mass before formation of the amniotic cavity (pig type) would typically give ri.se to two embryos in a single chorion with two independent amniotic cavities. The "Mateer" embryo as interpreted by Streeter illustrates this relation, and it is this scheme of twin-formation which is propounded in Streeter's contribution of 1919. A second kind of twinning, occurring by duplication of the actual embryonic areas after formation of the amniotic cavity (armadillo type), would produce two embryos within a single amnion (Fig. 6, D. E. F.). There would probably be intermediate and unclassifiable stages, because the moment of separation of the embryos might be intennediate in time, (t. e., during the opening of the amniotic cavity) and also because in the first type the embryos might lie so near together that the separate amniotic cavities would fuse. Our theoiy thus covers, perhaps too easily, the known variations of arrangement of the foetal membranes of twins in pre-natal stages and at birth. (For collections of such data see O. Schulze, 1897, and Wilder, 1904). The conception that there may be a series of types of human single-ovum twins accords well with the fact that there is a rather wide variability in the degree of resemblance between so-called "identical twins;" but caution suggests that we leave these concluding notions to await the further test of actual specimens.

LEGENDS OF FIGURES

F*iG. 1. — ^Specimen 1. Embryos of the third week, showing the cylindrical chorion cut open to expose the embryos. Points indicated by the letters A-A were adjacent before the yolk-sac was cut in opening the chorion, as were points B-B. Note abnormalities of the allantoides and yolk-sacs as described in the text. (X 7.)

Fig. 2. — ^Specimen 2. Normal embryos 8.5 and 9.5 mm. long, enclosed in a single chorion, (x 2.5.1

Fig. 3. — Specimen 3. Semi-diagrammatic. The embryos lie side by side within a single chorion. In this stage the amnions have invaginated the allantoides. The yolk-sacs appear to be separate, as in the previous specimens, (x 2.5.)

Fig. 4. — Diagram of a section taken through the chorion and allantoic septum near the point of attachment of the embryos of Specimen 3, showing that the chorion does not participate in the formation of the septum.

Fig. 5. — Explanation of Pig Twinning. Single Twin

Fig. 6. — Hypothetical Diagrams of Human Monochorionic Twins.

Duplication before Duplication after

formation of formation of

amniotic cavity amniotic cavity

PAPERS CITED

Arey, L.: 1922 a. Direct proof of the monozygotic origin of human identical twins. Anat. Rec., XXIII, 245.

1922 b. Chorionic fusion and augmented twinning in the human tube. Anat. Rec, XXIII, 253.

Assheton, R.: 1S9S. An account of a blastodermic vesicle of the sheep of the seventh day with twin germinal areas. Jour. Anat. Physiol., XXXII, 362.

Newman, H. H.: 1917. The biologj- of twins. University of Chicago Press.

Newman, H. H. and Patterson, J. T. : 1910. The development of the nine-banded armadillo. Jour. Morph., XXI, 359.

Patterson, J. T.: 1913. Polyembryonic development in Tatusia novemcincta. Jour. Morph., XXIV, 559.

Schultze, O. : 1897. Grundriss der Entwickelungsgeschichte des Menschen und der Saugethiere. Leipzig.

Spaulding, M. H.: 1921. The development of the external genitalia in the human embryo. Contributions to Embryolog)% vol. 13, Carnegie Institution of Washington, Publication No. 276, p. 67.

Streeter, G. L.: 1919 a. A human embryo (Mateer) of the presomite period. Contributions to Embryology, vol. 9, Carnegie Institution of Washington, Publication No. 272, p. 389.

1919 b. Formation of single-ovum twins. Johns Hopkins Hosp. Bull., XXX, 235.

Wilder, H. H.: 1904. Duplicate twins and double monsters. Amer. Jour. Anat., Ill, 387.