An Analysis of the Postgastrula Differentiation of the Hypomere: I. The Influence of tissue Mass and of Endoderm in Ambystoma punctatum

Since the publication of earlier papers (Finnegan, 1953, 1955) the investigation of the capacity of the salamander hypomeric mesoderm for histogenesis under a variety of experimental conditions has continued. It is perhaps prudent at this time to initiate a series of reports with results obtained from in vitro experiments which were designed to gain some insight into the roles of competence, tissue mass, and endodermal influence relative to hypomeric differentiation in Ambystoma. This portion of the mesoderm is destined to undergo its differentiation far removed from the dorsal axial influences of the chorda-mesoderm but with its inner (splanchnic) material in rather intimate association with the endoderm, a tissue known to be determined at an early age and metabolically active, two conditions which lead one to suspect it of inductor potentialities (Nieuwkoop, 1947; Copenhaver, 1955). The ‘synergistic’ or ‘additive’ role of the archenteron floor in heart development as demonstrated by Ekman (1925), Bacon (1945), and Jacobson (1960), and the possible role of the posterior endoderm in germcell development indicated by Nieuwkoop (1947, 1950) as well as the results of this investigation, are indicative of the validity of such a supposition. Recent work (see reviews of Grobstein, 1955, 1959) has emphasized the role of supra-cellular or tissue (mass) influences on the cytodifferentiation of experimentally isolated units. The attempt at a control of mass-size plus its systematic variation in this experimentation further evidences this fact.

This report then is concerned with the differentiational ability of a limited mass of Ambystoma neural-stage hypomere mesoderm when isolated in an ectodermal ball, and what occurs to this histogenesis when a small number of endoderm cells are added; when the mass of mesoderm is doubled; and, finally, when endodermal cells are added to the doubled mesodermal mass.

The A. punctatum embryos used in this investigation were maintained, after collection and prior to operation, in pond water at 7° C. The explants prepared from these donors were cultured in modified Holtfreter or Niu-Twitty solution at 18° C. Sterile techniques were utilized throughout. In preparing the explants for the various experimental series the following procedures were used.

The donor animal was placed in the operating medium on its side, and an incision was made in the trunk flank ectoderm in the vicinity of the ventral somite region (either actual or potential) with a finely sharpened steel needle. The cut was continued for a short distance both anteriorly and posteriorly and extreme care was taken not to disturb the underlying mesoderm. The ectoderm was then peeled ventrad from this point. Prior to the complete removal of the ectoderm from the donor’s flank, an incision was made in the hypomere a short distance ventral to the nephrogenic mesoderm. This cut was continued both anteriad and posteriad so that a piece approximately 0·5 mm. in length was separated and could be rolled ventrad on itself or in contact with the overlying ectoderm which had been lifted away from the flank. The sheet of hypomere separates neatly from the underlying endoderm. The distance the hypomere extended ventrally at the time of operation depended on the donor age and thus set the limit to this dimension of the explanted mass of hypomere. In practice this dimension appeared to approximate the anteroposterior dimension, exceeding it only in the older stages, particularly along the posterior edge of the explant. Considerable care was taken that no endoderm cells were clinging to the mesoderm. The superficial ectoderm was continued ventrad into the non-mesodermal area of the venter in order to provide a second ‘flap’ to approach and fuse with the dorsal one when the excised material balled-up in healing. In preparing the explants in all the experimental series care was taken to exclude the potential heart area as localized by Wilens (1955) for this species. Otherwise the tissue removed could be seen from later observation of the donors to reside in that portion of the hypomere which contributes material to the area lying ventral to the pronephros and posteriorly along the length of the liver; thus its somatic portion would probably include some of the limb field and peritoneum, while its splanchnic portion would have contributed to the posterior fore-gut and anterior mid-gut wall and associated ventral structures.

The operational procedure described in the above paragraph was repeated. After the removal of the explant from the donor, a small group of superficial endoderm cells taken from the lateral or ventral wall of the posterior fore-gut (pharynx) was placed on the hypomere prior to the closure of the ectodermal ball. While a definite count was not made on all such endodermal contributions, the cell counts made revealed that twelve to twenty endoderm cells were usually added to the explant.

To prepare each of these explants, two animals of identical age and egg-clutch were subjected to the operation described in I above, both contributions being removed from the same flank of the donors. A sandwich was constructed by juxtaposing the hypomeres centrally, the ectodermal layers remaining superficial. In such a construction, using elements from corresponding flanks, one must reverse either the dorso-ventral or the antero-posterior axes of the two pieces relative to one another. While in the event the healing frequently shifted the relationship of both these axes, the majority of explants were prepared initially with the antero-posterior axes aligned and the dorso-ventral axes reversed.

The experimental procedure outlined in III above was repeated and an endodermal contribution taken from each of the donors. The two pieces of endoderm were placed separately between the two layers of hypomere when the explants were brought together to construct the sandwich.

The explants remained in the operating dishes until they rounded up and any extraneous cells and debris were removed prior to transfer to the culture dishes. It should be noted that in Series I (M) and II there was a tendency to lose a small quantity of mesoderm and endoderm as the balls closed. Frequently, however, this material, in particular the endoderm, remained as an epithelial layer at the surface of the explant throughout the entire culture period, thereby contributing its influence to that of the endoderm which had remained internally.

The experimental results are based on observations on nearly ninety explants prepared from donors ranging from stage 13 to stage 21 inclusive, and approximately one-half of these were subsequently fixed in Michaelis’ Solution, Sectioned at 8–10 μ and stained with haematoxylin and eosin for histological examination. Macroscopic observations were made daily of the explants and illustrations prepared of their development. Individual cases were removed from culture at various intervals, typically at the end of the first week, in the middle and at the end of the second week, and in the third week, for fixation and microscopic observation. In order to facilitate comparison of the results the explants have been arranged into three major groups according to the age of the donor animals at the time of operation; stages 13–15 are considered as the stage 14 group, stages 16–18 as the stage 17 group, and stages 19–21 as the stage 20 group.

The experimental results concerned with histogenesis will be discussed within the framework of the operational series described in the preceding section and will be followed by an analysis of observations on the role of growth in these explants.

The stage 14 and the stage 17 groups developed into the second week (donor stage 35) after explantation, the former group remaining more solid while nearly one-half of the latter group became vesicular. In both groups, histological examination revealed that the more yolky cells of the splanchnic mesoderm were aggregated in the centre of the explant while the cells of the somatic mesoderm, lying more superficially, demonstrated a reduced yolk content. The latter did not, however, form a peritoneum beneath the overlying ectoderm, which remained as cuboidal cells at the surface. In both groups there appeared to be an attempt to form a haematopoietic structure in which haematoblasts could be distinguished and in the sinuses of which apparent erythroblasts were observed (Plate 1, fig. 1). The stage 17 group showed a tendency to produce a slightly thicker capsule around the splanchnic mesoderm with some separation between the somatic mesoderm and this capsule.

The stage 20 group explants remained vigorous through the second week (donor stage 41) and into the third in some cases, the majority becoming vesicular. Examination of these made evident the existence of involutions producing a cavity or, more often, several small cavities as the superficial somatopleure material (somatic mesoderm peritoneum plus squamous ectoderm) pushed into the interior. This would appear to be an expression of the antero-posterior extension of this material which must occur in the embryos as the length of the ventral axis increases during post tail-bud stages, and the embryo appears to straighten up from its earlier C-like silhouette. Apparently this migratory ability is present from around stage 18 onward but was not demonstrated in the stage 14 group unless the occasional wrinkles in the surface could be so considered. During the first week of culture the histogenesis of these explants included a definite cytological erythropoiesis in the splanchnic mesoderm with many late stages of this process in evidence while, at the end of the second week, there appeared to be more endothelial tissue and the erythropoietic areas now contained erythrocyte-like cells whose nuclei showed extremely condensed chromatin, to the point of being pycnotic. Some evidence of thromboplastic differentiation was found. In the third week these explants revealed the production of a fibrous connective tissue stroma in the splanchnic mesoderm and beneath this the establishment of a reticulo-endothelial tissue with possible signs of granulopoiesis being initiated (Plate 1, fig. 2; Plate 2, figs. 9,10). The erythropoietic area now consisted of large numbers of cells, erythrocytes and thromboplasts, with the aforementioned condensed nuclei. In no other area of these explants were these nearly solid nuclei observed, and the condensation far exceeded that observable in normal erythrocyte nuclei.

Macroscopic examination of all the explants in this series did not reveal that axial orientation played any part in the differentiation observed; no overall pattern was revealed sufficient to indicate response to an axial influence. Rather, the differentiations seemed to be local affairs with even the migratory responses showing no organization. From the foregoing it would appear that the hypomere involved in these explantations is capable, from immediately after gastrulation, of supporting the initial stages of erythropoiesis, and that little histogenesis beyond this appears until the stage 20 group is examined. Having initially demonstrated erythropoiesis in the splanchnic mesoderm this stage 20 group, in the late second to the third week of culture (donor stage 41–45) produces a stroma in which fibroblasts and subsequently a reticulo-endothelium are present and which may evidence some granulopoiesis. At the same time the erythrocytes seem to be degenerating. This histogenesis on the part of the hypomere, in particular the haematopoiesis of the splanchnic mesoderm, appears to be quite similar to the differentiation observed in the animal. Thus, after stage 18, the splanchnic portion of the hypomeric mesoderm responds to isolation in a small mass with a histogenesis reminiscent of its development in situ, whereas such histogenesis does not occur in the explants from earlier stages. It is interesting to note the separation of granulopoiesis and erythropoiesis in these postgastrula explants in the light of Jordon’s (1930, 1938) observation that such separation in the adult appears to be confined to the Urodeles. Copenhaver (1943) noted the appearance of a subcapsular granulopoiesis in Ambys toma embryos after removal of the potential liver; apparently this latter tissue is not required, at this later stage, for such histogenesis to occur.

The explants in the stage 14 group and the stage 17 group were rather similar in their development with 40 per cent. (4 out of 10) of the former and 75 per cent. (6 out of 8) of the latter becoming vesicular and showing involutions of the superficial somato-pleure. In the stage 14 group, the splanchnic mesoderm surrounded the yolky endoderm cells, was more visibly separated from the somatic mesoderm than in the previous series, and produced peritoneal epithelium and some endothelium. In the first week of culture this group produced a haematopoietic area in which erythropoiesis was suspected, though no criteria exist for differentiating between haematoblasts and megaloblasts. In the second week the yolk content of the splanchnic cells appeared much reduced from what it had been in the previous (M) series, further evidence of erythropoiesis was manifested, and slightly increased amounts of endothelial tissue were present. The condensed nucleated cells (erythrocytes) began to appear. In the third week of culture the splanchnic mesoderm in the vicinity of the endoderm produced fibroblasts; and the appearance of blood-cells with granular cytoplasm, the majority of which were eosinophilic with a small minority possibly showing basophilic granules (primitive eosinophils, Jordon & Speidel, 1930), gave evidence of granulopoiesis in this area.

The stage 17 group demonstrated an earlier and greater appearance of fibroblasts in the splanchnic mesoderm and usually developed some endothelial vessels. Toward the end of the second week of culture these explants contained the condensed nuclei in the erythropoietic area and had initiated elsewhere reticulo-endothelial tissue with its associated granulopoiesis.

In the stage 20 group, over 70 per cent. (10 out of 14) of the explants became vesicular with the inrolling confined to one or two areas and frequently with endoderm cells ‘collected’ at one lip. The splanchnic mesoderm appeared to have differentiated more rapidly in this group, producing the erythropoietic component with extremely condensed nuclei in erythrocytes and thrombocytes during the first week, and constructing, early in the second week of culture, a capsule in the endoderm area with reticulo-endothelium and some granulopoiesis. In the explants continued into the third week there appeared a large amount of reticulo-endothelium in what might be taken as a hepatic capsule with a considerable number of early eosinophils, lymphoid haemoblasts, and possibly a few basophils (the identity of the latter could not be assured). Elsewhere, monocytes and a few erythrocytes with pycnotic nuclei were found (Plate 1, fig. 3; Plate 2, figs. 11, 12).

In all of the explants of this series the haematopoietic area was observed, macroscopically, to be confined to one region of the explant. Some of the stage 17 and the stage 20 groups constructed an endothelial vessel ‘anteriad’ from this cell mass. The stage 20 group produced, early in the culture period, a reticulo-endothelial area just anterior to the initial haematopoietic region and granulopoiesis was found to be occurring therein. Careful examination of the stage 17 group revealed that the cases of such axiation were those whose donors had been at stage 18+. It would appear, therefore, that this axial organization was developed in those explants taken from donors at stage 19 or older and was only vaguely apparent in the stage 14 group. Further evidence of this is indicated in the observation that two of these explants (a stage 15 and a stage 19) produced small groups of pulsating endothelial cells (probably in response to the cardiac induction of the archenteron floor known in this species, Bacon, 1945) and the older explant (stage 19) located this group in the anterior region of the ball.

In comparing the previous series (M) with this one in which endoderm was present, it would appear that differentiation had been speeded-up in all ages, that the histogenetic spectrum had been increased in the younger ages (particularly in the splanchnic mesoderm), and that the older explants (after stage 18) demonstrated an axial response in their histogenesis which had not been evidenced when the hypomere was cultured alone. Thus, in both the stage 14 and the stage 17 groups the splanchnic mesoderm developed a granulopoietic area within a ‘capsule’ or ‘stroma’ of fibrous connective tissue (a histogenesis not observed for these groups in the previous series (M)), and the stage 20 group differentiated in a similar manner, but more rapidly than in the preceding series.

The appearance of vesicularity rose to 70 per cent. (5 out of 7) in the stage 14 group in response to the increased amount of somatopleure present in these explants. The large amount of splanchnic mesoderm was gathered in a mass and, after 2 weeks of culture, there were present some erythrocytes and thrombocytes, a number of erythroblasts, and a large population of haematoblasts. Monocytes, some containing phagocytized cell remnants, were also present in this area. In the main, these explants produced a large amount of peritoneum, an haematopoietic area, and some endothelial vessels in the areas removed from the haematopoietic ‘organ’. The amount of peritoneum and endothelium was in excess of that observed in any of the previous stage 14 groups, indicating a capacity not previously considered common in these stages. The haematopoiesis, in quality of differentiation, resembled the stage 14 group in the (M) series rather than the (M+E) series and no evidence of granulopoiesis was found.

All four explants of the stage 17 group became vesicular and the amount of peritoneum and endothelium exceeded that seen previously in this group. Here, also, the endothelial vessels were produced in the area away from the haematopoietic organ and the latter demonstrated some evidence of erythropoiesis in the first week and a number of monocytes plus condensed nucleated erythrocytes during the second week. Also, in the second week, these explants differentiated a large number of fibroblasts and much ground substance became apparent, particularly in the somatic hypomere. No reticulo-endothelial tissue appeared.

Only two specimens were prepared in the stage 20 group, one being fixed after the first week and the other late in the second week. Both became vesicular and demonstrated a large amount of peritoneum and endothelium with endothelium leading away from the haematopoietic area. In the latter area erythropoiesis was well advanced at the end of the first week, while in the second week the pycnotic nuclei appeared, though the near-by haematoblasts were frequently found in mitosis. Again, no reticulo-endothelium or granulopoiesis was observed and it seemed that all available splanchnic mesoderm had been utilized, leaving no reserve for other differentiation.

Macroscopically this group demonstrated better axial organization than had been seen in any of the previous experimental series. In 50 per cent. (2 out of 4) of the stage 17 group and in both cases of the stage 20 group there was produced an endothelium leading anteriad from the haematopoietic region and rhythmic pulsations were observed in the more anterior portion of this axial arrangement in both of the stage 20 cases and in one of the stage 17 group. Again, inclusion of the known heart area was avoided in these explants and no attempt at heart morphogenesis was observed under these conditions, whereas the culture of the known heart area from similar-aged donors under identical experimental conditions produced a pulsating, curved, heart-like structure in the explant (Plate 1, fig. 4). Apparently the heart field is sufficiently established in these late neurula stages for a small group of cells from its posterior extremity to differentiate functionally, cells which are external to the presumptive heart material and would not enter into normal heart development when the ventral migration of this material after stage 20 is completed (Wilens, 1955; Copenhaver, 1955). The rate of pulsation in the explants was always that described for the sinus venosus area (Copenhaver, 1955), but no organogenesis typical of that area occurred.

In general, this experimental series demonstrates that the increase in mass of mesoderm has resulted in the production of larger quantities of basic connective tissue (fibroblast) and of peritoneum and endothelium; a reduction in haematopoietic potential, though the initial erythropoiesis occurs; and an expression of axial pattern in the stage 17 and stage 20 groups more convincing than previously observed. Both the somatic and splanchnic hypomere are involved in these variations from the histogenesis seen in the first two types of experimental manipulation. The appearance of endothelium in these stage 14 group explants would indicate that cells capable of this differentiation exist in the early postgastrula splanchnic mesoderm, as suggested by Copenhaver (1955), but the absence of such amounts of this material from the previous series (M and M+E) would also indicate that these cells may be directed into some other histogenesis according to the local conditions in this mesoderm.

The stage 14 group explants became vesicular in over 50 per cent. (6 out of 11) of the cases and within the first week of culture erythropoiesis was obvious in the compact haematopoietic area produced. Blood-cell differentiation continued through the second week, thromboplasts and monocytes appeared, and the explants demonstrated normal erythrocytes into the third week of culture. Elsewhere, these older cultures indicated a larger production of intercellular ground substance by fibroblasts than had previously been observed in stage 14 explants, but no reticulo-endothelial stroma was produced (Plate 2, fig. 13).

Nearly 90 per cent. (8 out of 9) of the stage 17 group explants became vesicular and during the first week also demonstrated considerable erythropoietic activity in the splanchnic haematopoietic area. In addition, much endothelium differentiated in the splanchnic mesoderm leading away from the haematopoietic area (see Plate 1, fig. 5). This situation prevailed in the second week and, though the number of pycnotic nuclei increased in the erythropoietic area, the large number of normal erythrocytes and erythroblasts present indicated a continuing erythropoiesis (Plate 1, fig. 6). Again, the splanchnic mesoderm did not produce any reticulo-endothelium but the older cultures revealed localized increases in fibroblast differentiation and the accompanying appearance of quantities of ground substance.

Macroscopically both groups demonstrated axial organization; the stage 14 group in 36 per cent. (4 out of 11) of the cases, and the stage 17 group in 45 per cent. (4 out of 9) of the cases (Plate 2, figs. 7, 8). Thus, the stage 14 group produced this organization for the first time in these experiments while the stage 17 group retained the level previously demonstrated in the mesoderm doubled (2M) series. Both groups showed rhythmic pulsation in a small anterior area of endothelium in the same proportion that they demonstrated axial organization. It would appear that this axial organization in the early stage explants is facilitated by the presence of endoderm, probably through the latter functioning as an inductor of cardiac myogenesis (Bacon, 1945) with a subsequent organization within the mesoderm itself, rather than through the small amount of endoderm employed exerting regional influences along an axis. On the whole, these explants produced a better erythropoiesis than previously encountered (haemoglobin could be visibly identified in the haematopoietic area of many of the stage 17 group), and histological examination revealed the presence of endoderm cells in this area of the explant or in the surrounding epithelium.

The number of mitotic figures per ten sections was obtained for each of 51 explants and forms the basis for the information contained in Tables 1 and 2. In general it was found that the mitotic figures, which were found in all areas of the explants initially, came to be concentrated in the mesoderm after the first week, with the majority in the splanchnic mesoderm after the second week. This concentration was particularly noticeable in the haematopoietic area.

The values reported in these tables were obtained by calculating the mean mitotic count and the standard error for the cases within an experimental series. The mean values were statistically compared according to Student’s t test (Snedecor, 1952) and P values < 0·05 were considered to be significant. Grateful acknowledgement is made of the assistance and guidance of Dr. W. N. Holmes in this analysis of the data.

Comparison in Table 1 of the mitotic indexes for the hypomere explants and those for the explants to which some endoderm was added (columns I and II) or of which the mass of mesoderm was doubled (column III) reveals that the mitotic activity in the stage 14–17 groups has not been significantly affected by these treatments. The stage 20 cases also seem to retain their basic rate of celldivision after these treatments, only the 2M group (column III) showing a difference approaching significance. Comparison of the combined M+E group with the combined 2M group demonstrates a significant difference in the mitotic indexes (P < 0·05 and > 0·02). Apparently when the mesoderm mass is doubled (column III) sufficient tissue is present to stabilize the mitotic rate in some manner at a level significantly in excess of that obtained when endoderm is added to the single hypomere explant (column II). This situation would prevail if the endoderm were influencing the differentiation of the cells involved, in which case their mitotic ability would decrease, and examination of the resulting histogenesis substantiates this interpretation, particularly with regard to the splanchnic mesoderm. Thus, increasing the mass of mesoderm or adding endoderm to the explant appears to influence the differentiation of the constituent mesoderm rather than simply to stimulate an increase in cell number (growth). However, when endoderm is added to the increased mesodermal mass (column IV) in the stage 14–17 groups the mitotic rate is increased significantly over that of the M group (column I), and growth of the mesoderm as well as its differentiation appear to be assisted by the presence of the endoderm. This point is further emphasized when a comparison of the mitotic index in the 2M group and that of the 2M+E group (columns III and IV) indicates a significant difference (P < 0·05 and > 0·02) in the amount of cell-division when endoderm is added. Examination of the sectioned material herein demonstrates that the splanchnic portion of the explants, particularly in the haematopoietic areas, responds mitotically to the presence of the endoderm.

When Table 2 is examined, it can be seen that the mitotic rate of the hypomere explants (column I) remains rather constant throughout the culture period since comparison of the means for the second and third weeks (P < 0·3 and >0·2) demonstrate no significant difference in number of mitosis. When the mesoderm mass is increased (column III), the basic situation as regards celldivision is not disturbed, for the significant increment in this group in the second week does not appear to exceed that produced by adding the values for two similar but separate hypomere explants. The addition of endoderm, however, does appear to modify the basic situation. When it is added to the single-mesoderm explant (column II), the mitotic increase is postponed to the third week; these cultures differentiated during the first 10 to 12 days with little growth. Thus the mitotic index in the second week shows a significant repression below that of the M series, while no significance can be shown in the differences in mitotic indexes of these two groups (I and II) in the third week. On the other hand, the addition of endoderm to the doubled-mesoderm group (column IV) results in an immediate increase in mitotic rate during the first week which is sustained into the second week. Since the doubled-mesoderm group also shows an increase in the second week and the apparent difference between the 2M and 2M+E groups (columns III and IV) is not significant (P < 0·2, > 0·1), it would appear that the large shift in the first week can be attributed both to the larger mesoderm mass and to the endoderm. The impetus to haematopoiesis in these groups with endoderm added—considerable numbers of dividing haematoblasts were observed in the histological sections—would to a considerable extent account for the increase in number of mitotic figures.

Thus, the major role of the endoderm in these various experimental series would seem to be to exert a direct effect on the differentiation of the associated mesoderm, particularly on the splanchnic mesoderm, rather than an indirect one by influencing the growth of the mesoderm.

In their cogent analysis of progressive differentiation in amphibians, Holtfreter & Hamburger (1955) point out (p. 257) that the ‘… mesodermal districts do seem to require specific external stimuli for their normal differentiation to some extent’, and, in his chapter in the same compendium, Copenhaver (1955) notes (p. 453) that the evidence indicates that the endoderm ‘… has striking effects in the development of other tissues’. The results of the present in vitro study of hypomeric differentiation indicate that this mesoderm does respond to external stimuli emanating from the endoderm, but equally well demonstrates that the varying developmental propensities of the mesoderm involved play such a large role in the specific histogenesis manifested that caution is required in applying such an hypothesis to in vivo developmental events.

When the mass of hypomere is limited in size there results an attempt at erythropoiesis in explants from all neural stages, and a separate granulopoietic area appears in the splanchnic mesoderm of the post-stage 18 explants after sufficient culture time. When a few endoderm cells are added to these explants a similar histogenesis occurs but now the younger stages demonstrate the granulopoietic activity and the older stages do so more rapidly than in the above case. These results and the physical associations observed within the explants make it appear that the endoderm is influencing granulopoiesis in the splanchnic mesoderm. However, when the mass of hypomeric mesoderm is increased the granulopoietic differentiation does not appear and the addition of endoderm to such cultures does not restore that capacity to the splanchnic mesoderm. Examination of the sectioned material makes evident that the mesoderm responded to the increase in mass with the production of large amounts of endothelial and peritoneal tissue in addition to the omnipresent erythropoietic area. In the reduced mesodermal mass situation there usually remained a more or less compact mass of splanchnic mesoderm cells which apparently responded to the presence of the endoderm and established a ‘capsule’ within which reticulo-endothelium differentiated. When the mass of mesoderm was increased, however, these cells became involved in endothelial differentiation and were not available to respond to the associated endoderm. But the erythropoietic potential of this splanchnic mesoderm was enhanced by the proximity of the endoderm (Finnegan, 1953, 1955).

The presence of endoderm then does influence the differentiation of the splanchnic mesoderm, particularly as regards haematopoiesis, but does it play a role in the axial organization of this tissue? At the outset, it must be said that the indication of axial organization as reported in the results was based on events occurring in the splanchnic mesoderm for the most part, though notice was taken of the direction of ciliary beat in all explants (see Twitty, 1928).

In those cases where an axial pattern appeared in the splanchnic mesoderm its alignment was in agreement with the axis indicated by this ciliary beat. As the limited mesoderm mass developed as an explant it did not demonstrate any axial organization. When endoderm was added to this mass there appeared indications of such an alignment in the older stages (post-stage 18) but a more obvious demonstration of axiation in these older stages resulted from an increase in the mass of mesoderm. The increased mass resulted in the production of endothelial vessels by the splanchnic mesoderm leading away (i.e. anteriad) from the haematopoietic area and in the older neural stages this was often a single collection of endothelium, thereby resembling the axial circulatory system observable ventrally in vivo. It was also noted that the differentiation of this ‘system’ was initiated in the anterior end of the explant and continued to develop from the mass of mesoderm cells in a posterior direction, conforming to the axis previously determined by observation of the ciliary beat. These results would indicate that the antero-posterior axis is sufficiently well established in the splanchnic hypomere following stage 18 for its subsequent differentiation in vitro. Twitty (1928) pointed out a similar determination in the somatic hypomere as the probable cause of the ectodermal polarity following stage 18 in Ambystoma. Prior to that age this axis is not sufficiently well established for its expression in vitro even when the amount of splanchnic tissue present is increased. With the addition of endoderm to the increased mesodermal mass the younger neural stage explants demonstrate this axial alignment in the splanchnic mesoderm. It would appear that the endoderm influences two events in this development; cardiac myogenesis and erythropoiesis. In these explants these two histogenetic types are found at separate ends of the axial pattern (much in the manner of heart and spleen) as a small group of pulsating endothelial cells anteriorly and a large collection of erythopoietic cells toward the posterior end. The work of Bacon (1945) and Jacobson (1960) has indicated the role of endoderm in the former case and previous reports (Finnegan, 1953, 1955) have suggested an influence of endoderm in the latter differentiation. In both cases the endoderm would seem to play a synergistic or additive role in assisting the splanchnic mesoderm to realize its potential. Similarly, axial organization would seem to be present in the early postgastrula Ambystoma hypomere (stage 14 group), but the variation or heterogeneity (Grobstein, 1959) along the axis is insufficient in the splanchnic mesoderm at this time for its expression, even given a sufficient amount of material, while the association of endoderm with the endothelium and haematopoietic site and its influence thereon enhances this heterogeneity, enabling the early postgastrula splanchnic mesoderm to demonstrate its own axial organization in vitro. It must be emphasized that in this study a sufficient amount of hypomere tissue (2M) was necessary to ensure a histogenesis (endothelium) which would respond to the endodermal stimulus (cardiac myogenesis), indicating that the cytodifferentiation appearing under these conditions is, in the words of Grobstein (1959), ‘system-dependent’. The endoderm aids the fruition of such differentiation (be it granulopoiesis, erythropoiesis, or cardiac myogenesis) of the postgastrula splanchnic mesoderm as is initiated by conditions prevalent within that mesoderm itself.

1. Quand on cultive in vitro une petite quantité de mésoderme hypomérien à l’intérieur d’un manchon ectodermique (série M), la partie splanchnique se groupe au centre et il se produit une certaine hématopoièse (érythropoièse). Dans des explants réalisés à partir d’embryons ayant dépassé le stade 18, cultivés plus de deux semaines, on a observé la granulopoièse dans un réticulo-endothélium en cours de différenciation.

2. Quand on ajoute un groupe de cellules endodermiques à l’hypomère explanté (série les stades post-gastruléens les plus jeunes (stades 13–18) présentent une histogenèse semblable aux stades les plus avancés de la série M (ci-dessus), tandis que les explants des stades postérieurs au stade 18 subissent une histogenèse semblable, mais plus rapidement que dans la série précédente. Quelques-uns de ces explants de stades plus avancés ont montré une orientation axiale du mésoderme splanchnique.

3. Quand on a augmenté la masse de l’hypomère (série 2M), il s’est formé une grande quantité de péritoine et d’endothélium dans tous les explants, de concert avec le territoire érythropoiétique habituel. On n’a pas observé de réticulo-endothélium ou de granulopoièse. Les explants des donneurs plus âgés (stades 17–20) ont fréquemment présenté une structure axiale du mésoderme splanchnique.

4. Quand on a ajouté des groupes de cellules endodermiques à la masse accrue de l’hypomère (série 2M+E), les explants post-gastruléens précoces (stades 13–16) ont, pour la première fois, montré une répartition axiale de l’histogenèse endothéliale et hématopoiétique dans le mésoderme splanchnique. De nouveau, on n’a pas observé de réticulo-endothélium avec granulopoièse.

5. L’analyse des dénombrements mitotiques dans les divers explants a indiqué que l’endoderme influence la différenciation plutôt que la croissance (c.-à-d. la division cellulaire). De même, l’accroissement quantitatif de l’hypomère paraît influencer la différenciation et aussi stabiliser la croissance.

6. Ces résultats expérimentaux sont considérés comme une nouvelle indication du rôle synergique de l’endoderme dans l’histogenèse de l’hypomère splanchnique.

This investigation was supported in part by grants from the National Institutes of Health (U.S.) (RG–6178), the National Research Council of Canada, and the President’s Committee on Research of the University of British Columbia.

Abbreviations:

 
• A

  anterior

 
• ED

  endoderm

 
• EN

  endothelium

 
• ER

  haematopoietic (erythropoietic) area

 
• G

  granulopoietic area

 
• GS

  ground substance

 
• H

  haematoblast.

 
• M

  mitotic figure in haematoblast.

 
• P

  posterior so somatopleure.

 
• SP

  splanchnic mesoderm.