The origin and movement of nephrogenic cells in the chick embryo as determined by radioautographic mapping

The radioautographic mapping of the primitive-streak to head-process stage chick embryo (Rosenquist, 1966) demonstrated that the intermediate (presumptive nephrogenic) mesoderm originates in the epiblast layer lateral to the primitive streak (Fig. 1, left side), migrates medially to the streak, and is invaginated at a position about one-third of the distance from the anterior to the posterior end of the streak. Considerable mixing of the premesodermal cells takes place at the streak, resulting in the crossover of some cells into the mesoderm layer opposite the side of their origin. After entering the mesoderm, the cells destined for intermediate mesoderm migrate anteriorly and laterally from the streak into the isthmus which joins the paraxial accumulation of mesoderm to the somatic and splanchnic leaflets of the lateral plate (Fig. 1, right side). Since the previous mapping study did not include embryos older than the 17-somite stage, the differentiation of the nephros from the intermediate mesoderm was not investigated.

Using radioautographic analysis, the present investigation traces the movements of tritiated thymidine-labelled transplants from their original positions in the nephrogenic regions of epiblast and streak of recipient embryos, into the ducts, tubules, and undifferentiated intermediate mesoderm of the host embryos.

Embryos were removed from white leghorn eggs incubated 10–26 h. The vitelline membranes to which the embryos remained attached were secured to a glass ring and suspended over a pool of albumen (New, 1955). Embryos were staged by the criteria below (Fig. 2; Rosenquist, 1970).

There is a slightly pear-shaped area pellucida, with a moderately broad but well-formed primitive streak which may extend 50–60% of the length of the area pellucida (Hamburger & Hamilton, 1951, p. 54, stage 3; plate 1, stage 3 + ).

There is a slightly pear-shaped area pellucida, with the anterior end wider than the posterior end. The streak is approximately 65 % of the length of the area pellucida, and the length of area pellucida is greater than at the previous stage (Hamburger & Hamilton, 1951, p. 54, stage 3).

The length of the streak is 70-80 % of the length of the area pellucida. The anterior end of the streak is compact and rounded, and is more dense in appearance than at the late medium-streak stage. Hypoblast folds at anterior margins of area pellucida are well developed. The streak splits easily at the midline into lateral halves (Willier & Rawles, 1931, DPS stage; Rudnick, 1932, DPS stage; Spratt, 1942, DS stage; Hamburger & Hamilton, 1951, plate 2, stage 4).

A short straight head process extends anteriorly from the anterior end of streak, which is 65–80% of the length of the area pellucida, depending upon the amount of increase in size of the latter. There is no evidence of amniocardiac vesicles upon microscopic investigation. The streak easily splits at the midline into lateral halves (Willier & Rawles, 1931, HP stage; Rudnick, 1932, HP stage; Spratt, 1948, HP stage; Hamburger & Hamilton, 1951, plate 2, stage 5).

There is a long head process with anterior accumulation of cells in a shallow depression (future site of the foregut) which has begun to fold, forming an anterior intestinal portal (Rudnick, 1932, HF stage; Spratt, 1942, HF stage; Hamburger & Hamilton, 1951, plate 2, stage 6).

Stages after the head-fold stage are determined by the number of pairs of somites present, e.g. stage 4S refers to an embryo with four pairs of somites (Willier & Rawles, 1931, stage 4S; Hamburger & Hamilton, 1951, plate 2, stage 8).

In the two embryos which had progressed to the early limb-bud stage, the number of somites could not be determined accurately. In both cases the head was slightly stunted in growth, and although the heart was beating at the time the embryos were fixed, circulation appeared to have stopped. (The early limb-bud stage is equivalent to Hamburger & Hamilton, 1951, plate 6, stages 17–19.)

Donor embryos were labelled by placing 1 μCi [³H]thymidine in 0 037 ml physiologic saline solution on the hypoblast surface of the explanted embryo. The donors were then reincubated until they reached the stages of the recipients (1–5 h), and washed six times with 1–2 ml saline at room temperature. This procedure labels 50–70% of the cells, which is sufficient to trace the movements of the grafts with radioautography (Rosenquist, 1966). Grafts were excised from the donor embryos and transplanted to identical sites in the epiblast, streak or endoderm-mesoderm (hypoblast) layer of similarly staged recipient embryos, and the graft position was recorded by photographs or sketches of the recipient embryos (Rosenquist, 1966).

The recipient embryos were incubated in sealed moisturized cans containing atmospheric air for 4–38 h following the grafting operation. Seventeen of the embryos were then re-photographed or sketched, fixed in 10 % formol in physiologic saline, embedded in paraffin wax, sectioned serially at 10 μ radioautographed, stained and reconstructed after study (Rosenquist, 1966). The two embryos which survived to limb-bud stages, and two which reached the 23 + somite stage, were selected when they had developed good extra-embryonic circulation (16- to 19-somite stage); they were turned upside down on a physiologic agar (Howard, 1953) and reincubated in sealed cans containing a moisturized mixture of 95 % oxygen and 5 % carbon dioxide. Additional chick albumen was added to the exposed vitelline membrane from time to time. These four embryos had a total incubation time of 41–63 h after the grafting procedure, and they were re-photographed, fixed, etc., in the same manner as were the first seventeen embryos.

Of the twenty-one embryos to be reported, none which received grafts at the medium-streak stage developed all the way to the early limb-bud stage. Therefore, to trace the movements of the nephrogenic cells from their origin in epiblast at the medium-streak stage, to a position in intermediate mesoderm at somite stages, it was necessary to combine the pathways followed by two or three labelled grafts. For example, a graft transplanted to the epiblast about half-way between the margin of the area pellucida and the streak (embryo 3, Fig. 2B) migrated to the centre of the streak (embryo 3*, Fig. 2C); a graft placed at the centre of the streak (embryo 11, Fig. 2C) moved from that position into intermediate mesoderm (embryo 11 *, Fig. 2H).

Although part of each of the twenty-one transplants illustrated in Fig. 2 lay in the nephrogenic region, the number of embryos investigated was relatively small, and each transplant contained cells other than nephrogenic cells (as indicated in Table 1). Therefore the positions of the transplants at each stage in Fig. 2 suggest the location of the nephrogenic region at that stage, but do not define it precisely. The mapping of the nephrogenic region is based upon the following assumptions: (1) that previous studies have established the general position of the intermediate mesoderm in the epiblast, streak and mesoderm layers of the chick blastoderm without mapping every part of these layers at each stage (Rosenquist, 1966). Consequently, a small number of transplants, carefully placed, can demonstrate the position of more specific portions of the intermediate mesoderm, such as the kidney. (2) That graft positions in different embryos at the same stage are homologous even if the embryos were incubated for different lengths of time, and that the pathways followed by more than one accurately placed graft can be used to follow movements of a group of cells through several stages of development. (3) That maps of presumptive organ-forming regions of the embryo are valid even if structures other than nephros in the recipient embryos contain labelled cells.

Throughout the text and figures, an asterisk (*) after the embryo number indicates that the position shown is that of the graft after its migration in the host embryo.

The presumptive nephrogenic cells of the chick blastoderm are divided into two categories. The first, the anterior migrating nephros (AMN), includes the entire pronephros and the anterior part of the mesonephros as they were described by Abdel-Malek (1950). The AMN is so named because its definitive position, after it migrates from the epiblast through the streak into mesoderm, is anterior to that of the second category of presumptive nephrogenic cells, the posterior migrating nephros (PMN). The latter category includes the major portion of the mesonephros, and the metanephros (Abdel-Malek, 1950). In the following description only the migration of the nephrogenic cells is described. It must be remembered that cells from each graft contributed to other tissue as well, as noted in Table 1.

This is demonstrated by transplants contributing to the intermediate mesoderm lateral to the 1st–16th somites. AMN is represented in Fig. 2 as striped areas.

At the medium-streak stage, the AMN was found in the epiblast about halfway between the primitive streak and the lateral margin of the area pellucida, as illustrated by the migrating grafts in embryos 1 and 2 (Table 1 and Fig. 2A). At the late medium-streak stage, the AMN had moved closer to the primitive streak (embryos 1* and 3–10, Table 1 and Fig. 2B). At the definitive-streak stage, it had reached the primitive streak at a position which was approximately one-third to one-half the distance from the anterior to the posterior end of the streak (embryos 3*, 11 and 12, Table 1 and Fig. 2C). At the head-process stage, the AMN had begun to migrate into mesoderm adjacent to the streak (embryos 4* and 13, Table 1 and Fig. 2D). The route of the later migration from the streak was anterior and lateral; the AMN was parallel to the mesoderm cells destined for the lateral plate (which were located posterior and lateral to the AMN, Fig. 1, right side) and also to those destined for paraxial mesoderm (which were located anterior and medial to the AMN, Fig. 1, right side). At the head-fold to 1-somite stage, the AMN had moved farther into the mesoderm layer (embryos 5* and 6*, Table 1 and Fig. 2E), although some cells were still in the primitive streak (embryos 14–16, Table 1 and Fig. 2E).

By the 4- to 6-somite stage, some of the AMN had reached its definitive position lateral to the first 4-6 somites (embryos 2* and 7*, Table 1 and Fig. 2F). At the 8- to 9-somite stage, additional AMN had reached the definitive position in intermediate mesoderm (embryos 8* and 9*, Table 1 and Fig. 2G). At the 17- to 19- somite stage, a ridge had formed along the dorsal surface of the intermediate mesoderm, consisting of cells which had emerged from the 4th–16th nephro-meres (Abdel-Malek, 1950), and which had fused to form the rudiments of the Wolffian ducts (embryos 11 *–13*, Table 1 and Fig. 2H). As yet none of the labelled cells in this ridge appeared posterior to the labelled cells in the strip of undifferentiated intermediate mesoderm which separated the somites and the leaves of the lateral plate. By the 22- to 23-somite stage, parts of the AMN forming the Wolffian duct had begun to extend posteriorly through non-labelled intermediate mesoderm, toward the hind gut (embryos 14*–16*, Table 1, Figs. 2I and 3B). At the early limb-bud stage, the Wolffian duct had reached the future position of the cloaca (embryo 10*, Table 1 and Fig. 2J). In embryos 10 and 14–16, occasional labelled cells clung to this posteriorly extended labelled cord without being a part of its relatively compact mass; it is not clear whether these cells were pulled passively toward the cloaca as the duct migrated posteriorly, or whether they migrated from the cord into the non-labelled intermediate mesoderm surrounding it.

This is demonstrated by transplants contributing to the intermediate mesoderm surrounding the Wolffian duct, posterior to the 16th somite. PMN is represented in Fig. 2 as shaded areas.

At the medium-streak stage, the PMN was located in the epiblast near the lateral margins of the area pellucida, at about the level of the anterior end of the streak (embryos 17–19, Table 1 and Fig. 2A). By the definitive-streak stage, the PMN was still near the area pellucida margin, but it was more posteriorly placed in relation to the streak (embryo 20, Table 1 and Fig. 2C); this apparent posterior movement was probably related to the anterior growth of the streak in relation to the anterolateral margin of the area pellucida (Rosenquist, 1966) and was thus not an actual migration of cells. At the head-process stage, the PMN had migrated toward the streak (embryo 20*, Table 1 and Fig. 2D), and parts of it may have entered the centre of the streak at approximately the same position at which the AMN entered (one-third to one-half the distance from the anterior to the posterior end of the streak). The remainder of the PMN at this stage extended laterally and anteriorly from the streak as a narrow strip of labelled epiblast, demonstrating how a square-shaped graft transplanted to epiblast elongates during its migration toward the streak. By the 4- to 6-somite stage, the PMN had moved through the streak into the migration pathway which leads to the intermediate mesoderm (embryos 17*–19* and 21, Table 1 and Fig. 2F). Fig. 3A shows labelled cells in the streak of embryo 18. At the time of fixing, part of the grafted tissue in this embryo had not yet entered the streak, part was in the streak, and part had already left the streak for intermediate mesoderm. By the early limb-bud stage, the PMN had migrated from the streak into a column of intermediate mesoderm posterior to the 23rd somite, where it had begun to differentiate into nephric ducts and tubules (embryo 21*, Table 1, Fig. 2J, 3C). It did not contribute to the Wolffian duct. Although no embryos were available which showed the position of the PMN between the 4- to 6-somite stage and the early limb-bud stage, it may be assumed that the PMN continued in the route it was following at the 4- to 6-somite stage (embryos 17*-19*, Fig. 2F) until it entered intermediate mesoderm and formed mesonephric tubules (embryo 21*, Fig. 2J).

The first successful attempts to locate the nephrogenic cells in the chick blastoderm utilized a method of culturing blastoderm fragments on the chorioallantoic membrane of host embryos until they differentiated. Hoadley (1926) divided embryos of 4–6 h incubation into five full-thickness (epiblast and hypoblast) horizontal strips (Fig. 4, first row), and transplanted each one to the chorioallantoic membrane of a host embryo. After 7–9 days incubation in this culture medium, convoluted tubules lined with columnar epithelium differentiated from the strip of embryo which included the anterior part of the streak. Since Hoadley described the donor embryos as having a broad primitive streak, they are probably comparable in staging to the medium-streak-stage recipient embryos of the present investigation. The findings of the present study are consistent with those of Hoadley. The present study indicates that at the mediumstreak stage the nephrogenic cells are in the epiblast layer between the lateral margin of the area pellucida and the streak; this site was present in the donor strip of Hoadley which included the anterior part of the streak (compare Figs. 1 and 4).

Similar convoluted tubules differentiated in chorioallantoic membrane cultures from fragments cut from definitive-streak-stage embryos (Hoadley, 1926; Hunt, 1931; Rudnick, 1932), head-process-stage embryos (Hoadley, 1926; Rudnick, 1932; Rawles, 1936), and head-fold-stage embryos (Hoadley, 1926; Hunt, 1931). Rudnick and Rawles made longitudinal cuts in addition to the horizontal incisions (Fig. 4, third and fourth rows), dividing each horizontal strip into right, left, and medial fragments. Since nephrogenic material was found in all three, it was concluded that at the definitive-streak to head-process stages, nephrogenic cells are present near the anterior end of the streak, both in the midline and on each side of the streak. This is also in agreement with the present investigation (Figs. 1, 2C) because by the definitive-streak stage, some of the nephrogenic cells have reached the streak, while others are in a position in the epiblast lateral to the streak (Fig. 2). By the head-process stage, nephrogenic cells are distributed to the mesoderm layer in addition to the epiblast and streak (Fig. 2D). This investigation indicates therefore that although the chorioallantoic membrane grafts were useful in the general localization of the nephrogenic areas of the chick blastoderm, they were not satisfactory for tracing morphogenetic movements. First, they did not distinguish between nephrogenic cells in the epiblast and those in the hypoblast, and second, they were too large to show the location of the nephrogenic cells as precisely as does the present method of grafting.

Neither the present nor previous investigations provide definite information about the extreme posterior end of the intermediate mesoderm. It is presumed that some of each type of mesoderm (i.e. lateral plate, intermediate and paraxial) continues to be invaginated as long as the streak is active. In embryo 21, the streak was still visible at the time of fixing, although it was quite short; it is not clear whether gastrulation was completed at this time. It is not clear whether the labelled cells seen in the epiblast layer next to the streak in this embryo represent the end of the intermediate mesoderm, i.e. part of the metanephros which has not yet been invaginated, or whether they are ectodermal elements destined for epidermis or neural tube.

Neither the present nor a previous radioautographic mapping study (Rosenquist, 1966) indicates the location of the migrating nephros prior to the mediumstreak stage. Its peripheral position in epiblast near the lateral margin of the area pellucida, at the medium-streak stage (Fig. 2A), suggests that at earlier stages it may be even farther from the streak, perhaps beyond the area pellucida margin. This is only speculation, however, since the critical early embryos which would settle this point are not available for study at this time.

The Wolffian (pronephric) duct has long been considered to be formed by the fused ends of the pronephric cords or tubules, and therefore to be derived from cells originating in the 4th to 16th nephromeres. The posterior growth of this pronephric duct in relation to the rest of the intermediate mesoderm may be traced in serial sections of embryos of various stages; it has been confirmed experimentally by interrupting the posterior growth of this structure with electrocautery or mechanical impediments (Boyden, 1924, 1927) and by microsurgery (Gruenwald, 1937; Waddington, 1938). In some of Boyden’s embryos the mesonephros became functional, but since the Wolffian duct did not connect with the cloaca, there was accumulation of fluid proximal to the site of cauterization, and the growth of the allantois and cloaca was stunted. Gruenwald and Waddington cut transversely through the intermediate mesoderm posterior to the last somite of embryos which had not yet developed a mesonephros; the posterior growing tip of the pronephric duct was in each case unable to cross the incision. Since neither Wolffian duct nor mesonephros was found posterior to the incision in embryos incubated to later stages, it was concluded that the mesonephros and metanephros needed contact with the Wolffian duct to differentiate.

The present investigation indicates that the Wolffian duct begins its posterior growth relative to the rest of the intermediate mesoderm some time between the 19- and 22-somite stages (Figs. 2H, I). In Table 1 (embryos 10 and 14–16) the position of this posteriorly growing duct is shown in relation to the position of other labelled cells in the same embryo. There are numerous examples of labelled cells in lateral plate, aorta, or endoderm, which are just as posteriorly placed as the labelled cells in the Wolffian duct and which, like the intermediate mesoderm, originated in the labelled graft. The posterior position of these cells is due probably to the fact that they migrated more slowly than did the pronephric cells, and thus did not reach their definitive position (anterior to the Wolffian duct) before the embryos were fixed. It is evident from the distribution of the cells that they did not grow posteriorly in relation to the rest of the intermediate mesoderm as did the pronephric duct.

L’origine des cellules néphrogènes présomptives dans l’épiblaste de l’embryon de Poulet est déterminée par une analyse radioautographique des mouvements de greffons marqués par la thimidine tritiée, excisés des embryons aux stades ligne primitive à 5 somites et transplantés dans l’épiblaste, la ligne primitive et la couche endoderme-mésoderme d’embryons hôtes des mêmes stades.

Les cellules néphrogènes ont une origine proche du bord de l’aire pellucide de l’embryon au stade ligne primitive, elles migrent vers la ligne primitive, et sont invaginées environ du tiers à la moitié de la distance entre les extrémités antérieure et postérieure de la ligne, entre les stades ligne primitive et 1 à 4 somites. Leur itinéraire dans le mésoderme poursuit une voie relativement étroite entre les cellules migrant vers le mésoderme paraxial ou présomitique d’une part et les cellules destinées aux membres proximaux de la plaque latérale d’autre part.

Les cellules qui formeront la partie antérieure du mésoderme intermédiaire sont les cellules les plus proches du plan médian dans l’épiblaste, elles atteignent la ligne à un stade de développement plus précoce et sont les premières cellules néphrogènes à migrer dans le mésoderme. Après le stade d’environ 17 à 19 somites, les cellules de ce groupe qui ont formé le cordon ou le canal pronéphriques commencent à se mouvoir postérieurement par rapport au reste du mésoderme intermédiaire, vers le futur cloaque.

Les dernières cellules néphrogènes à quitter l’épiblaste pour pénétrer dans la ligne et le mésoderme sont celles destinées à l’extrémité postérieure du mésoderme intermédiaire. Ce groupe de cellules entoure le canal pronéphrétique (canal de Wolff) qui migre postérieurement, et se différencie en mésonéphros.

This investigation was supported by USPHS research grants HE 10191 and KE HE 20074 from the National Heart Institute. The author wishes to thank James D. Ebert for his continued interest in this research, and Soame D. Christianson for help in the preparation of the manuscript.