Regionalisation of the Mouse Embryonic Ectoderm: Allocation of Prospective Ectodermal Tissues During Gastrulation

Development 107, 55-67 (1989) 55 Printed in Great Britain © The Company of Biologists Limited 1989

Regionalisation of the mouse embryonic ectoderm: allocation of prospective ectodermal tissues during gastrulation

PATRICK P. L. TAM

Department of Anatomy, Faculty of Medicine, The Chinese University of Hong Kong, Shatin, NT, Hong Kong

Summary

The regionalisation of cell fate in the embryonic ecto- grafted to the anterior midline are found in the oral derm was studied by analyzing the distribution of graft- ectodermal lining, whereas cells from the archenteron derived cells in the chimaeric embryo following grafting are found in the notochord. With respect to mesodermal of wheat germ agglutinin-gold-labelled cells and cul- tissues, ectoderm at the archenteron and the distal- turing primitive-streak-stage mouse embryos. Embry- lateral region of the egg cylinder gives rise to rhomben- onic ectoderm in the anterior region of the egg cylinder cephalic somitomeres, and the embryonic ectoderm contributes to the neuroectoderm of the prosencephalon adjacent to the primitive streak contributes to the and mesencephalon. Cells in the distal lateral region give somitic mesoderm and the lateral mesoderm. Based rise to the neuroectoderm of the rhombencephalon and upon results of this and other grafting studies, a map of the spinal cord. Embryonic ectoderm at the archenteron prospective ectodermal tissues in the embryonic ecto- and adjacent to the middle region of the primitive streak derm of the full-streak-stage mouse embryo is con- contributes to the neuroepithelium of the spinal cord. structed. The proximal-lateral ectoderm and the ectodermal cells adjacent to the posterior region of the primitive streak produce the surface ectoderm, the epidermal placodes Key words: embryonic ectoderm, primitive-streak stage, and the cranial neural crest cells. Some labelled cells mouse embryo, microsurgical grafting, lectin-gold labelling.

Introduction ectoderm is the elevation of proliferative activity in a small group of cells in the anterior region of the The embryonic ectoderm (epiblast) of the gastrulating embryonic ectoderm near the rostral end of the primi- mouse embryo is believed to be the sole precursor of all tive streak of the gastrulating embryo (Snow, 1977). definitive tissues in the fetus. The evidence supporting The developmental fate of the progeny of this mitoti- this notion is provided by the extensive range of cally active population is unknown though neuroecto- embryonic and adult tissues produced during the differ- derm has been suggested (Snow & Bennett, 1978). entiation of the whole epiblast or fragments of it after Morphological studies in the mouse embryo suggest transplanting to ectopic sites (Diwan & Stevens, 1976; that the embryonic ectoderm in the anterior region of Beddington, 1983; Svajger etal. 1986) and the multi- the egg cylinder gives rise to major segments of the tude of tissue types colonized by the epiblast cells brain on the grounds that this part of the embryonic grafted orthotopically in the primitive-streak-stage em- ectoderm is topographically related to the first three to bryo (Beddington, 1981, 1982). It is therefore conceiv- four somitomeres normally underlying the forebrain to able that, in order to generate a embryonic body plan upper hindbrain (Tarn & Meier, 1982; Meier & Tarn, from the histologically homogeneous embryonic ecto- 1982; Jacobson & Tarn, 1982). Orthotopic grafting of derm (Reinius, 1965; Batten & Haar, 1979), an orderly embryonic ectoderm cells to the most anterior region allocation of cells to various tissue types in specific and the distal tip (the node, the head process or regions of the body needs to be accomplished during the archenteron - Theiler, 1972; Poelmann, 1981) of the gastrulation. The process of pattern generation would primitive-streak-stage mouse embryo results in the be further facilitated if cells with diversified develop- colonization of, respectively, the head and trunk neur- mental fate are strategically located within the embry- ectoderm by the graft-derived cells but the precise onic ectoderm so that tissues of different lineages but segmental distribution of these cells in the cephalic belonging to specific parts of the body are properly neural tube is not known (Beddington, 1981). Details of juxtaposed in preparation for ingression through the the normal fate of cells in other anterior and lateral regions of the embryonic ectoderm are incomplete primitive streak. because previous mapping studies tended to focus An example of early regionalization of the embryonic 56 P. P. L. Tarn mostly on specific groups of cells adjacent to and within In vitro culture of embryos the primitive streak of the gastrulating embryo (Bedd- Primitive-streak-stage embryos were cultured in rotating ington, 1981, 1982; Copp etal. 1986; Tarn & Bedd- (30 rev min"') 50 ml glass bottle (Wheaton) containing 3-4 ml ington, 1987). An in vitro study on the morphogenesis of culture medium. To culture embryos for 22-24 h until they of isolated fragments of primitive-streak-stage embryo reached the early-somite stage, a 1:1 (v/v) mixture of rat seems to suggest that, within the embryonic ectoderm, serum and Dulbecco's modified Eagle's medium (DMEM, cell populations destined for specific brain parts are Gibco) was used as the culture medium. For culturing embryos for 44-46h until the forelimb-bud stage, a mixture of already spatially delineated (Snow, 1981). The present mouse serum, rat serum and DMEM (1:2:1 by volume) was grafting study was carried out to examine the regionali- used and the embryos were transferred to fresh medium after sation of prospective ectodermal tissues in the anterior 24 h of culture (Hunter et al. 1988). The culture was gassed and lateral regions of the embryonic ectoderm. Special with 5% CO2, 5% O2 and 90% N2 during the first 24 h of attention is given to (1) the segmental distribution of development to the early-somite stage and then with 5 % CO2 the epiblast-derived cells in the neural tube and (2) the in air for further culture. location of cells destined for the surface ectoderm, the placodes and cephalic neural crest cells of the mouse Labelling of embryonic ectoderm and preparation of embryo. grafts Wheat germ agglutinin (WGA)-gold conjugate used for Materials and methods labelling the embryonic ectoderm was prepared as described by Tarn & Beddington (1987). About 2-5 nl of the colloidal WGA-gold label was injected into the amniotic cavity of the Recovery of embryos primitive-streak-stage embryo, which was then cultured in rat Primitive-streak-stage embryos were obtained from a close- serum-DMEM for 3-4 h. After culture, the embryo was bred colony of ICR strain mice. At 7-5 days p.c., embryos transferred to PB1 medium and rinsed twice in the same were dissected from the uterus in PB1 medium containing medium. Fig. 2A shows the orientation of embryonic axes of either 4mgml~' bovine serum albumin (Miles Lab) or 10% the egg cylinder with the posterior aspect of the embryo fetal calf serum (FCS, Gibco). The parietal yolk sac was indicated by the allantoic rudiment and the primitive streak removed microsurgically with a pair of fine glass needles and and the proximal border by the amnion. Using a pair of fine the embryos were washed in several changes of fresh PB1 glass needles, the egg cylinder was transected at the level of medium. Only late-primitive-streak-stage embryos (Fig. 1) the amnion to remove the extraembryonic region. A longi- showing expanded exocoelom and amniotic cavity, a com- tudinal cut was made down the anterior midline (Fig. 2B) to plete amnion and clearly discernible embryonic ectoderm unfold the embryonic portion of the egg cylinder. The were used for labelling and grafting. Other features character- dissected embryo was then left in PB1 medium on a warm istic of embryos at this stage are the early allantoic rudiment stage (30-35°C) for about 5-10min. Often this was sufficient and the anterior-posterior gradient of tissue opacity due to to cause the spontaneous separation of the more turgid the presence of the spreading mesodermal wings. embryonic ectoderm from the loose mesoderm and the thin endodermal layer, which started to curl back from the free edges. Further separation of tissue layers was achieved by inserting a fine needle underneath the embryonic ectoderm, which was colored deep red by the gold label, to slice away the mesoderm and endoderm. A longitudinal cut was then made on each side of the primitive streak (Fig. 2B) to isolate two half-fragments of the embryonic ectoderm without the primitive streak (Fig. 2C). The embryonic ectoderm fragments were transferred to fresh PB1 medium and divided into four major fragments: two smaller anterior quadrants and two larger posterior quadrants (fragments A-D; Fig. 2C) from which smaller clumps of 20-30 cells were isolated for grafting. Usually 5 to 6 egg cylinders would provide enough tissues for grafting of 16-20 embryos.

Grafting experiments The strategy was to graft WGA-gold-labelled cells isolated from the four major fragments of the embryonic ectoderm to different sites in the primitive-streak-stage mouse embryos. Donor cells were grafted to sites in the same quadrant from which they were isolated. Exact orthotopic grafting, which was technically much more complicated, was not attempted. The embryos were then cultured until they developed to the early-somite stage (22-24 h HI vitro) or to the stage of formation of forelimb bud and closure of anterior neuropore Fig. 1. Scanning electron micrograph of a bisected 7-5-day (stage 14 - Thieler, 1972; 44-46 h in vitro). The pattern of primitive-streak-stage embryo showing the exocoelom tissue colonization by the graft-derived cells was studied in the completely separated from the amniotic cavity by the chimaeric embryos following histological preparation and amnion (am), al, allantoic rudiment; ch, chorion; ps, silver enhancement of the colloidal gold particles in the primitive streak; ect, embryonic ectoderm. Bar= 100/.im. labelled cells. The manipulation of the egg cylinder and the Tissue fate of embryonic ectoderm 57

EXT

Al LAT fr- PROXIMAL ,PS -I POST POST EMB ANT

LAT DISTAL

(a) Ib)

Fig. 2. (A) Anterior (ANT)-posterior (POST) and proximal-distal embryonic axes of the 7-5-day egg cylinder. The primitive streak (PS) and allantois (AL) mark the posterior side of the egg cylinder and the proximal border is defined by the amnion attaching to the junction between the extraembryonic (EXT) and the embryonic (EMB) parts of the egg cylinder. (B) The position of cut I along the anterior midline of the egg cylinder and cuts II on the two sides of the primitive streak (PS) producing the fragments shown in fig. 2c. LAT, lateral aspect of the egg cylinder, ECT, embryonic ectoderm; AC, amniotic cavity. (C) The dissection of the half-embryo obtained in B to yield fragments A,B,C and D by cuts III and IV. grafting of labelled cells to the embryonic ectoderm followed the egg cylinder as the reference (Fig. 3). These regions were that described by Beddington (1987). Briefly, grafting was separated from one other by a distance of 80-100 ^m along done by microinjecting clumps of labelled cells to the embry- the orthogonal axes which could be discerned under a onic ectoderm after pushing the injection needle through the dissecting microscope at 40x magnification. The following primitive endoderm and the mesoderm of the egg cylinder. grafts were made: The graft was placed well within the ectodermal epithelium I Labelled cells from proximal-anterior quadrant (frag- but it was inevitable that, because of the wound produced by ment A - Fig. 2C) were grafted to the injection needle, some labelled cells might be lodged in a (1) the anterior midline position near the attachment mesodermal or even endodermal position after grafting. The of the amnion to the embryonic ectoderm exact location of the donor cells was examined in 40 embryos (anterior-marginal: Ant-Mar). which were fixed for histological study of WGA-gold-labelled tissues at 4-5 h after grafting. Labelled donor cells were grafted to seven different sites of the embryonic ectoderm. Because of the variation in the size of the 7-5-day embryos, it was technically not feasible to obtain uniform results from different embryos by positioning the graft on the basis of absolute physical distance from any morphological landmark. Instead, the seven locations were so chosen that they represented the Cartesian points of an imaginary rectangular grid mapped on the lateral aspect of the embryo using the allantois, the amnion and the distal tip of

Fig. 3. A schematic diagram showing the position in the embryonic ectoderm where microsurgical grafting of WGA-gold-labelled embryonic ectoderm cells were made. Three graftings were made to the midline position (ANT- MAR, MID-ANT and ARCH) and others to the lateral areas with two (PL-PS and ML-PS) adjacent to the primitive streak (PS) and two midway between the anterior and posterior midline of the embryo (P-LATand D-LAT). PL-PS graft was made close to the allantoic rudiment (AL) and P-LAT close to amnion (AM). MID-ANT, D-LAT and ML-PS was halfway down the proximal-distal distance in the embryonic part of the egg cylinder. ARCH was at tip of the egg cylinder. 58 P. P. L. Tarn

11 Labelled cells from the distal-anterior quadrant (fragment B - Fig. 2C) were grafted to (2) the region halfway down the anterior midline of the egg cylinder (mid-anterior: Mid-Ant). III Labelled cells from the proximal-posterior quadrant (fragment C - Fig. 2C) were grafted to: (3) the region lateral to the posterior end of the primitive streak (posterior-lateral primitive streak: PL-PS), and (4) the proximal area near to the attachment of the amnion on the lateral flank of the egg cylinder (proximal-lateral: P-Lat). IV Labelled cells from distal-posterior quadrant (fragment D - Fig. 2C) were grafted to: (5) the region halfway down the posterior side of the egg cylinder and lateral to the middle portion of the I primitive streak (mid-lateral primitive streak: ML- \ PS), (6) the area halfway down on the lateral flank of the egg ect cylinder (distal-lateral: D-Lat), and (7) the distal tip of the egg cylinder, which corresponded to the site of the archenteron and the head process (archenteron: Arch). Embryos were inspected under the dissection microscope immediately after grafting for the location of the graft. The labelled cells could easily be discerned by their red coloration. Embryos containing grafts in the mesodermal or endodermal Fig. 4. A primitive-streak-stage embryo examined 4h after germ layers were discarded. grafting, snowing the proper incorporation of WGA-gold- labelled cells (arrowheads) in the pseudostratified Examination of embryos after culture epithelium of the embryonic ectoderm, ect, embryonic Embryos were harvested at 22-24 or 44-46 h of culture and ectoderm. Silver-enhanced and Fast Green. Bar = 20,um. examined for vitelline circulation and heart activity. The fetal membranes were then removed for a closer inspection of developmental features such as the appearance of neural folds study, since they might represent artefacts of improper and closure of cephalic neural tube, formation of somites and graftings. forelimb bud and axis rotation. Embryos were fixed with Carnoy fluid, embedded in paraffin wax and processed for Embryonic development in vitro light microscopy. To visualize the colloidal gold marker, After 22-24 h of culture, over 50% of experimental histological sections were treated with a silver developer and embryos developed an actively beating heart, and over counterstained with Fast Green (Tarn & Beddington, 1987). 85 % of them formed early neural folds. About 4-8 The location of labelled cells in different embryonic tissues pairs of somites were formed, which was similar to the was identified in serial sections of the cultured embryos. The number of somites found in the intact embryos cultured segmental position of the labelled cells in the neural tube, the for the same duration (Table 2). When examined after surface ectoderm, the cranial mesenchyme, the somitic meso- 44-46h of culture, experimental embryos of the Ant- derm and the lateral plate mesoderm was determined with Mar, Mid-Ant, P-Lat and D-Lat groups developed to respect to the primary brain parts and the branchial arches in the cephalic region and to the somite in the trunk region. the same extent as the intact embryo with respect to axis rotation, cephalic neurulation, limb bud formation and somite number (Table 2). Embryos of Arch, ML-PS Results and PL-PS groups were retarded when compared to the

The initial location of grafted cells Table 1. The location of Forty embryos receiving grafts at different sites of the WGA-gold-labelled cells in embryos examined at embryonic ectoderm were examined at 4-5 h after 4-5 h after grafting grafting. Results in Table 1 show that about 68% Number of embryos having labelled cells in (57-80%) of the experimental embryos have donor cells confined to the embryonic ectoderm (Fig. 4) and Total Ectoderm & Mesoderm Group Ectoderm mesoderm & endoderm about 20 % have donor cells in both the ectoderm and the adjacent mesoderm. About 20-30 WGA-gold- Ant-Mar 14 8 4 2 Mid-Ant 5 4 0 1 labelled cells were found in these embryos. In about Arch 6 4 1 1 12 % of cases, labelled cells were found only in the Lat 15 11 3 1 mesoderm and endoderm of the host embryo. Based 40 27(68%) 8(20%) 5(12%) upon this observation, 18 chimaeric embryos (of a total of 132) that showed tissue colonization by donor cells Experimental groups: Ant-Mar = anterior margin; Mid- only in mesodermal and endodermal tissues and not in Ant = mid-anterior midline; Arch = archenteron; Lat = lateral embryonic ectoderm. any ectodermal derivatives were excluded from this Tissue fate of embryonic ectoderm 59 Table 2. Development of embryos receiving grafts of labelled embryonic ectoderm at the egg cylinder stage and cultured for 22-24 or 44-46 h in vitro 44-46h 22-241 No. of embryos (%) showing No. of embryos (%) showing \Ipnrnl oomiic No. of Neural Beating number No. of Axis tube Yolk sac Forelimb number Groups embryos folds Somites heart (n) embryos rotation closure circulation bud (n) Intact embryos 23 19 (83) 19 (83) 11 (48) 5-0 ±0-3 (19) 14 12 (86) 10 (71) 12 (86) 7(50) 16-2 ±0-6 (14) Embryos with grafts at: Ant-Mar 32 30 (94) 30 (94) 19 (59) 5-5 ± 0-3 (30) 16 12 (75) 10 (63) 15 (93) 9(56) 17-3 ±0-6 (15) Mid-Ant 20 19 (95) 19 (95) 14(60) 6-4 ±0-2 (18)" 24 15 (63) 13 (54) 22 (91) 19 (79) 17-1 ±0-4 (21) Arch 35 31(89) 31 (89) 18 (51) 5-8 ± 0-4 (30) 17 7 (41)e 5 (29)c 15 (88) 9(53) 14-4 ± 0-8 (14)b ML-PS 10 9(90) 9(90) 6(60) 6-0 ±0-5 (9) 16 9(56) 6(38) 15 (94) 4(25) 13-5 ± 0-6 (13)b PL-PS 6 6(100) 6(100) 3(50) 4-5 ±0-7 (6) 18 8(44)c 6 (33)c 15 (83) 6(33) 13-3 ±0-5 (17)b P-Lat 25 25 (100) 24 (96) 14 (56) 5-7 ± 0-3 (24) 9 6(67) 4(44) 7(78) 4(44) 18-2 + 0-8(6) D-Lat 26 22 (85) 22 (85) 14 (54) 5-4 ± 0-4 (24) 25 14(64) 11 (44) 17 (68) 15(60) 16-6 ±0-7 (21) Experimental groups: Ant-Mar = anterior margin; Mid-Ant = mid-anterior midline; Arch = archenteron; ML-PS = lateral to the mid-region of the primitive streak; PL-PS = lateral to the posterior region of the primitive streak; P-Lat = proximal area of lateral embryonic ectoderm; D- Lat = distal area of lateral embryonic ectoderm. Somite numbers; a = groups with more somites and b = groups with fewer somites than intact embryos (Duncan's multiple range comparison at P<0-05). Embryonic development: c = significantly different from intact embryos at f<005 by Chi-squared test.

intact embryo developing under similar culture con- more frequently in embryos receiving grafts along the ditions. Axis rotation and cephalic neural tube closure anterior midline ectoderm (Ant-Mar, Mid-Ant and were delayed and fewer somites were formed (Table 2). Arch) and the lateral ectoderm (P-Lat and D-Lat), in contrast to embryos with grafts made to embryonic Tissue distribution of labelled cells ectoderm adjacent to the primitive streak (Table 4). The number of WGA-gold-labelled cells in different The segmental distribution of labelled cells in the tissues of 36 chimaeric embryos was scored (Table 3). neural tube is summarized in Table 5. Most embryos The total number of labelled cells varied from 28 to 154 receiving grafts at Ant-Mar ectoderm had labelled cells per embryo. In embryos of the Ant-Mar, Mid-Ant, in the prosencephalon, mainly on the floor of the Arch, P-Lat and D-Lat groups, the majority of labelled diencephalon near the invaginating Rathke's pouch cells was found in ectodermal derivatives such as the (Fig. 5), the lamina terminalis and the optic evagi- neural tube and epithelia of the oral cavity, body nation. Some labelled cells were also found at the surface and cephalic placodes of chimaeric embryos. junction between the forebrain and the midbrain. Colonization of the notochord by graft-derived cells Labelled cells of the Mid-Ant graft were found mostly occurred only in the Arch group embryos. More in the mesencephalic floor and at the junction between labelled cells were found in the paraxial and lateral the midbrain and adjacent brain segments. The rhom- mesoderm of the PL-PS, ML-PS and D-Lat embryos bencephalon and the neural tube at the level of the first than in the other groups (P<0-01, Mann-Whitney three (occipital) somites (Fig. 6) were colonized mainly test). About 12 % of the labelled population was found by labelled cells grafted to the D-Lat ectoderm, with a in the cranial mesenchyme and heart mesoderm of the minor contribution from the P-Lat and PL-PS ecto- Ant-Mar, Mid-Ant and Arch embryos, compared to derm. Colonization of the neural tube at more caudal 33% mesodermal colonization in embryos receiving somitic levels was observed in embryos with Arch and graft in the lateral regions. Colonization of the endo- ML-PS grafts. derm was found in only 3 chimaeras, with about 10 % of the labelled cells (2-10 per embryo) present in the gut (ii) Surface ectoderm and placode epithelium. Colonization of the surface ectoderm was observed in 25-43 % of embryos receiving grafts in the lateral Patterns of tissue colonization ectoderm and adjacent to the primitive streak Altogether 93 chimaeric embryos, including 15 from the (Table 4). The surface ectoderm over the branchial cell counting study, were analysed in detail for the arches was colonized by cells grafted to lateral ecto- spatial distribution of labelled cells in various embry- derm but those over the hindbrain and somitic levels onic tissues (Tables 4-6). were derived from grafts in the lateral ectoderm as well as adjacent to the primitive streak (Table 5). Labelled (i) Neural tube cells of the P-Lat and PL-PS grafts were frequently Colonization of the neural tube and placodes occurred found in the columnar placode epithelium on the lateral 60 P. P. L. Tarn Table 3. The distribution of the labelled cell population in the chimaeric embryos examined at 22-24 h after grafting of WGA-gold labelled embryonic ectoderm cells Number of labelled cells in

Oral Surface Placode Neural Neural Gut Grafts Total tube crest ectoderm Mesoderm endoderm Notochord Ant-Mar 33 25 5 50 44 6 80 67 ll+2ht 31 14 11 6ht 27 5 18 2+2ht Mid-Ant 32 30 2 23 21 2 35 29 4 21 19 2 42 42 Arch 28 24 4 28 6 2 20 39 3 8 24 134 81 22 31 44 25 19 25 25 5 64 55 5 4 PL-PS 72 10 5 57 42 55 4 77 11 33 33 ML-PS 102 26 66 10 59 25 34 154 122 32 121 97 24 P-Lat 49 2 27 20 59 21 17 21 37 13 24 80 35 6 39 110 62 48 60 8 46 6 D-Lat 51 25 17 9 105 95 10 49 41 8 124 106 18 79 40 39 84 64 20

Tissues containing more than 50 % of the labelled cell population are in italic. For experimental groups, refer to footnote of Table 2, ht = heart mesoderm.

Table 4. Tissue colonization by WGA-gold-labelled cells in chimaeric embryos No. of embryos (%) showing colonization in

Neural Cranial Neural Lateral No. of tube & Surface Oral mesen- crest Somitic plate Groups embryos placode ectoderm ectoderm chyme cells mesoderm mesoderm Heart Gut Notochord Embryos with grafts at: Ant-Mar 20 18 (90) 1(5) 8(40) 1(5) 0 1(5) 0 2(10) 3(15) 0 Mid-Ant 13 13 (100) 0 2(15) 6(46) 0 1(7) 0 0 0 1(7) Arch 22 14(64) 0 0 2(9) 0 8(36) 0 0 4(18) 14(64) ML-PS 8 4(50) 3(38) 0 1(12) 0 4(50) 3(38) 0 2(25) 0 PL-PS 8 3(38) 3(38) 0 2(25) 2(25) 1(13) 5(63) 0 0 0 P-Lat 14 10 (71) 6(43) 0 4(29) 4(29) 1(7) 1(7) 2(14) 0 0 D-Lat 12 11 (92) 3(25) 0 6(50) 0 2(16) 5(41) 1(8) 0 0

For experimental groups, refer to footnote of Table 2. Table 5. The distribution of WGA-gold-labelled cells in the neural tube, placode and surface ectoderm of the chimaeric embryos No. of embryos with labelled cells in

Spinal cord Surface ectoderm Forebrain Midbrain Hindbrain Jen Jen Segmental level by somites Segmental level N LT R OP W F F-M R W F M-H Met Mye OT 1 2 3 4 5 6 7 8 9 10 11 12 P BA1 BA2 Mye 12 3 4 5 Ant-Mar 20 15 Mid-Ant 13 1 2 8 1 Arch 15 4 6 6 6 2 3 1 ML-PS 8 2 2 11 1 6 1111 2 3 3 2 1 PL-PS 8 1 2 11 1 P-Lat 10 1 1 11111 D-Lat 12 5 5 4 2 1

Abbreviations For experimental groups, refer to footnote of Table 2. Embryonic parts: BA1, BA2 = first and second branchial arch; F = floor of brain vesicle; Jen = junction between forebrain and midbrain (F-M) and of midbrain and hindbrain (M-H); LT= lamina terminalis; Met = metencephalon; Mye = myelencephalon; OP = optic evagination; OT = otic placode or vesicle; P = posterior neuropore and adjacent areas; R = roof of brain vesicles.

Table 6. The distribution of WGA-gold-labelled cells in the cranial mesenchyme, somitic mesoderm and lateral plate mesoderm of the chimaeric embryos No. of embryos with labelled cells in

Cranial mesenchyme Somitic mesoderm Lateral mesoderm K N I II BA1 MBNC III BA2 HBNC IV V VI VII 1 8 AP MP PP My re

Ant-Mar 1 1 1 re* Mid-Ant 6 3 Arch 9 1 1 2 3 4 2 4 2 1 1 re MP-PS 5 1 1 1 2 1 2 3 3 3 1 1 1 1 2 1 PL-PS 6 2 1 2 1 1 1 1 3 3 P-Lat 3 3 3 1 2 1 1 1 1 D-Lat 9 1 1 1 1 1 3 2 2 1 1 1 o 3 Abbreviations For experimental groups, refer to footnote of Table 2. 8 Embryonic parts: I—VII = cranial somitomeres; AP, MP, PP = anterior, middle and posterior portions of presomitic mesoderm; BA1, BA2 = first and second branchial arch; F = floor of brain vesicle; HBNC = neural crest cells in the lateral mesenchymal region of the hindbrain; MBNC = neural crest cells in the lateral mesenchymal region of the midbrain; My = myelencephalon.

o\ 62 P. P. L. Tarn

aspect side of the head folds of the early-somite embryo and otic placodes after grafting to the PL-PS and P-Lat (Fig. 7). The labelled epithelium was located close to ectoderm. the root of the first two branchial arches and at the myelencephalic level of the chimaeric PL-PS and P-Lat (iii) Cranial mesenchyme and neural crest cells embryos. Following further in vitro development, Labelled cells were found in the cranial mesenchyme of labelled cells were found in the trigeminal ganglia and all groups of experimental embryos especially in those the otic capsule of the advanced embryos, suggesting with grafts in the Mid-Ant and lateral ectoderm that the donor cells might have colonized the trigeminal (Table 4). When cells were grafted to Mid-Ant ecto-

Fig. 5. A sagittal section of the diencephalon (di) of an embryo cultured for 44 h after Ant-Mar grafting showing the presence of labelled cells in the neurohypophyseal diverticulum meeting the invaginating Rathke's pouch (rp) on the roof of the oral cavity. Fig. 6. A sagittal section of an embryo cultured for 24 h after Mid-Ant grafting showing WGA-gold-labelled cells (arrowheads) in the rhombencephalon (rh) and in the mesenchyme adjacent to the truncus of the heart, fg, foregut portal; ht, heart tube. Fig. 7. A frontal section of an embryo cultured for 24 h after P-LAT grafting resulting in the colonization of placodal epithelium at the upper hindbrain level (arrowheads), fg, foregut; nt, neural tube. Fig. 8. A transverse section through the hindbrain level of an embryo cultured for 24 h after P-Lat grafting showing the presence of labelled cells (arrowheads) in the neuroepithelium (ne) and at the lateral subectodermal position along the putative migratory path for neural crest cells. Fig. 9. A section along the long axis of the first branchial arch (I) of an embryo cultured for 44 h after D-Lat grafting showing labelled cells (arrowheads) in the mesenchyme of the branchial arch, p, first pharnyngeal pouch; II, second ' -' scl branchial arch. Fig. 10. A sagittal section through the trunk of an embryo cultured for 44 h after ML-PS graft showing the colonization of somites by WGA-gold- labelled cells (arrowheads), se, surface dm ectoderm; scl, sclerotome; dm, dermamyotome. Bars= 100/im. Silver- 9 10 enhanced and Fast Green. Tissue fate of embryonic ectoderm 63 derm, 46% of the embryos showed colonization of found in the epimyocardium and pericardium of 4 somitomeres II and III (Table 6). Labelled cells grafted embryos in the Ant-Mar and P-Lat groups and in 1 to lateral areas (P-Lat and D-Lat) of the egg cylinder D-Lat embryo (Table 4). and to regions adjacent to the primitive streak (ML-PS and PL-PS) colonized predominantly the rhombencephalic somitomeres (Table 6). Only 3 examples of Discussion colonization of the cranial somitomeres were encountered in 42 embryos receiving grafts at Ant-Mar and The developmental fate of cells located in the anterior Arch regions. and lateral regions of the embryonic ectoderm of intact In embryos receiving grafts at P-Lat and PL-PS late-primitive-streak-stage embryos has been exam- regions, some labelled cells were found subjacent to the ined. Experimental evidence has been obtained for surface ectoderm on the lateral aspect of the cranial regionalisation in the deployment of embryonic ectomesenchyme (Fig. 8). These were probably neural crest derm cells to specific segments of the neural tube and cells since they were localized along the migratory path other ectodermal tissues during gastrulation. The study where such cells are expected (Chan & Tarn, 1988). of the mesodermal fate of embryonic ectoderm was, Mesencephalic neural crest cells were derived from however, fraught with technical problems and proved grafts in the P-Lat ectoderm and rhombencephalic less conclusive. neural crest cells came from grafts in the PL-PS ectoderm (Table 6). Technical consideration In older embryos receiving P-Lat and PL-PS grafts, In the present study, labelled cells were isolated from WGA-gold-labelled cells were found in cellular clus- four major quadrants of the embryonic ectoderm and ters at the base of the branchial arches, which are were grafted back to various sites in their quadrant of reminiscent of primordia of trigeminal and acoustico- origin. Grafting of cells to exactly orthotopic sites was facial ganglia. Although these cells were likely to be not attempted because of technical difficulties. There is, neural crest cells, a placodal contribution cannot be however, no significant variation in the pattern of tissue ruled out. In embryos with P-Lat and D-Lat grafts, colonization among chimaeric embryos of the same labelled cells were sometimes found in the mesenchy- grafting groups. Indeed, it has been shown that when mal core of the branchial arches (Fig. 9) but again it was embryonic ectoderm is grafted to heterotopic sites, the not possible in this study to identify their tissue of origin cells may change their normal fate to conform with the as either neural crest or cranial somitomere. pattern of tissue differentiation of the new locations in the embryonic ectoderm (Beddington, 1982). The one (iv) Somitic mesoderm and lateral plate mesoderm exception is the anterior marginal ectoderm which is Colonization of the somitic mesoderm (Fig. 10) was predisposed to produce ectodermal derivatives. Since in observed predominantly in embryos receiving grafts at the present study, donor cells were always grafted to Arch and ML-PS sites and to a lesser extent at D-Lat quasi-orthotopic sites, the differentiation of the graft- sites in the embryonic ectoderm (Table 4). Labelled derived cells is likely to reflect the normal fate of cells at cells grafted at the Arch and ML-PS ectoderm col- the selected sites in the embryonic ectoderm. onized the first 8-9 somites and the presomitic meso- One major drawback of the microsurgical grafting derm. Labelled cells derived from the lateral ectoderm technique is the possibility that, despite great caution were allocated to the first three somites (Table 6). taken to place the graft within the embryonic ectoderm, Colonization of the lateral plate mesoderm was some grafted ectodermal tissue may accidentally be observed primarily after grafts were made to the D-Lat grafted to the subjacent germ layers. The passage of the embryonic ectoderm and to the ectoderm adjacent to injection micropipette through the tissue layers might the primitive streak (Table 4). The graft-derived cells also create artefactual channels through which grafted colonized both the somatopleure and the splanchno- cells in the ectoderm could readily migrate to other pleure at the level of the lower hindbrain to first 4 germ layers. A survey of 40 embryos examined at 4-5 h somites (Table 6). after grafting did reveal that, in about 20% of cases, grafted cells were distributed to both the ectoderm and (v) Other embryonic tissues mesoderm which might result in the colonization of Extensive colonization of the notochord was found in tissues derived from both germ layers in these chimaeric embryos receiving grafts at the archenteron (Table 4). embryos. In another 12% of cases, the grafted cells In 5 out of the 14 cases, labelled cells were distributed in were not found in the embryonic ectoderm and there- the notochord over a length equivalent to 5-6 somites. fore colonization by graft-derived cells would probably Donor cells colonized the ectodermal lining of the be confined to mesodermal and endodermal tissues. In stomadeum, the buccopharyngeal membrane and the both cases, it would be difficult to draw any conclusion foregut endoderm of embryos receiving Ant-Mar grafts about the normal fate of embryonic ectodermal cells (Table 4). Two of the 13 Mid-Ant chimaeras also with respect to mesodermal and endodermal develop- contained labelled cells in the oral ectoderm, but not in ment. In the present study, chimaeric embryos that the foregut endoderm. Colonization of the endoderm of showed no colonization of ectodermal derivatives by mid- and hind-gut occurred in 6 cases following grafting WGA-gold-labelled cells were excluded on the as- to Arch and ML-PS sites (Table 4). Labelled cells were sumption that they were the result of misintegration of 64 P. P. L. Tarn

NTube SEct Ecl-M«3 PxMeso MiscMes GITract Others

1 Pros OrEct 2 n Di-Me Me Sme II -IV 3 m Me-RhRhSpC TrSEct BAr Sme IV- VI LPM 4 m SpC TiSEct Sme V- VII Som LPM 5 m SpC Som LPM 6 H s*c Som MGN HGN Noto 7 D Rh CrSEct PI BAr NCC Sme IV, V 8 nm CrSEct PI NCC LPM 9 SmeV-VH Son TB PS

10 D LPM PS

11 E3 LPM Ext Emb PS PGC

Fig. 11. The normal fate of cells in the embryonic ectoderm of the primitive-streak-stage mouse embryo studied by orthotopic graftings of labelled cells (Data from the present study: grafts 1-4 and 6-8; Tarn & Beddington, 1987: grafts 5 and 9-11; Beddington, 1981, 1982 and Copp et al. 1986: grafts 1,6 and 11). The dotted lines mark the boundary of the various brain parts and the spinal cord. For results obtained in the present study, only data showing colonization of the tissue in over 20% of the chimaeras were used for constructing this map. Neural tube (NTube): Pros, prosencephalon; Di~Me, diencephalic-mesencephalic junction; Me, mesencephalon; Me~Rh, mesencephalic-rhombencephalic junction; Rh, rhombencephalon; SpC, spinal cord. Surface ectoderm (SEct): CrSEct, head surface ectoderm; TrSEct, trunk surface ectoderm; PI, trigeminal placode or otic placode/vesicle. Ectomesenchyme (Ect~Mes): BAr, branchial arch mesenchyme; NCC, presumptive neural crest cells. Paraxial mesoderm (PxMeso): Sme, cranial somitomeres I-VII; Som, somites and presomitic mesoderm. Other mesodermal tissue (MiscMes): LPM, lateral plate mesoderm including somatopleure and splanchnopleure; TB, caudal mesenchyme; ExtEmb, extraembryonic mesoderm of the amnion, yolk sac and allantois. Gastrointestinal tract (GITract): OrEct, ectodermal lining of the oral cavity and buccopharyngeal membrane; MGN, midgut endoderm; HGN, hindgut endoderm. Others: Noto, notochord; PS, primitive streak; PGC, primordial germ cells.

grafts. Only embryos containing labelled cells in the whereas cells grafted to more lateral positions end up ectodermal tissues were studied, with specific emphasis predominantly on the lateral wall of the neural tube. It on external tissues such as the neural tube and the seems therefore that not only the craniocaudal pattern surface ectoderm. The results for internal (mesodermal but also the dorsoventral orientation of the neural tube and endodermal) tissues must be viewed with caution is established within the embryonic ectoderm at this since they might be an artefact of grafting procedure. stage of gastrulation. Other ectodermal tissues besides the neuroectoderm The ectodermal derivatives are also derived from cells of the embryonic ectoderm. Results of the present grafting experiments demon- Cells in the more proximal areas of the lateral embry- strate clearly that cells in the anterior regions of the onic ectoderm contribute to the cranial surface ecto- embryonic ectoderm give rise to the neuroectoderm of derm, the epidermal placode and the neural crest cells. the prosencephalon and mesencephalon. Cells destined A transplantation study in the mid- to late-primitive- for the rhombencephalon come from the ectoderm in streak-stage chick embryo shows that a crescent-shaped the distal-lateral regions flanking the rostral end of the zone in the anterior and lateral aspects of the epiblast is primitive streak. The neuroectoderm of the spinal cord destined for neural crest cells and peripheral to this is derived from the ectoderm overlying the archenteron zone are those cells destined for the epidermis (Rosen- (alias the node region) and lateral to the anterior and quist, 1981). If such a neural crest cell zone could be the middle regions of the primitive streak (sites 4 and 5, taken to demarcate the boundary of the neural primor- Fig. 11). The various segments of the brain and the dium then the neural plate of the mouse embryo is trunk neural tube are represented in the correct cranio- occupying a major portion of the anterior and lateral caudal order along the anterior-posterior axis of the regions of the embryonic ectoderm leaving a small area embryonic ectoderm of the primitive-streak-stage em- in the proximal-lateral embryonic ectoderm for the bryo. Labelled cells grafted to the midline of the non-neural ectodermal tissues. Fig. 11 shows the sub- embryo colonize the floor plate and the basal plate division of the neural primordium into domains for the Tissue fate of embryonic ectoderm 65 brain segments and the spinal cord, based on the scribed in the chick epiblast (Vakaet, 1984). Such cell distribution of graft-derived cells in major neural tube movement is probably linked to the anisotropic growth segments and particularly at junctions between seg- of the egg cylinder in the posterior and distal direction ments. The forebrain and the midbrain occupy a (Tarn & Meier, 1982) and could be achieved by the relatively small area making up about one third of the active proliferation of cells in the lateral and distal anterior embryonic ectoderm. The hindbrain covers the embryonic ectoderm (Snow, 1977; Poelmann, 1980). distal area of the anterior embryonic ectoderm and The expansion in tissue areas can also be achieved by most of the lateral region. The trunk neural tube the changes in cell size and shape as exemplified by the occupies the node areas and extends posteriorly and attenuation of the epithelium of the surface ectoderm proximally to embryonic ectoderm lateral to the primi- accompanying neurulation in the rat embryo (Morriss- tive streak. Taking into account the cup-shaped con- Kay, 1981). figuration of the primitive-streak-stage mouse embryo, there is a remarkable similarity in the spatial pattern of Mesodermal derivatives brain segments in the embryonic ectoderm when com- Results of tissue colonization by donor cells in this study pared to that in the epiblast of the stage 4-5 chick suggested that precursor tissues for the paraxial meso- embryo. In the chick, the neural primordium is mapped derm and the lateral plate mesoderm are largely con- in a series of wedges stretching from the prenodal to the fined to the distal-lateral and posterior regions of the postnodal- areas in the epiblast adjacent to the cranial embryonic ectoderm close to the primitive streak. end of the primitive streak (Nicolet, 1971; Rosenquist, Colonization of the first three cranial somitomeres 1981; Packard, 1986). occurs in 1 out of 20 Ant-Mar and 6 out of 13 Ant-Mid Labelled cells derived from grafts in the anterior- chimaeric embryos. A minor contribution to 'loose marginal embryonic ectoderm also colonize the ecto- head mesoderm' and even heart mesoderm by grafts of dermal lining of the oral cavity, the buccopharyngeal anterior ectoderm has previously been reported (Bedd- membrane and the Rathke's pouch, in addition to ington, 1981, 1982). As previously discussed, it is lamina terminalis and the diencephalic (neurohypophy- doubtful whether this actually reflects the normal fate seal) diverticulum of the forebrain. It is interesting to of cells in this regions of the embryonic ectoderm. The draw a comparison with the prosencephalic plate of the translation of the anterior and lateral ectoderm into the early-somite-stage avian embryo. Using chick-quail cephalic neural tube necessitates a forward displace- chimaeras, the anterior neural primordium of the avian ment of a coherent epithelial tissue and a concomitant embryo is found to give rise to the typical neural expansion of the posterior ectoderm adjacent to the structures such as the telencephalon and diencephalon streak to generate the spinal cord and the paraxial and also to non-neural tissue including the lining of the mesoderm. If the colonization of the rhombencephalic nasal and oral cavities, the prosencephalic neural crest mesoderm by Ant-Mid graft is real then the generation cells and the hypophysis (Couly & Le Douarin, 1985, of such mesodermal tissues from an anterior sites in the 1987; and the morphological studies by Takor Takor & epithelial sheet would pose a difficult mechanistic Pearse, 1975). problem. A specific group of prospective mesodermal cells will have to move towards the streak in a direction The relative size of different parts of the neural tube opposite to that of tissue sheet expansion and then to as mapped onto the embryonic ectoderm of the primi- reverse their course after invagination to reach their tive-streak-stage embryo is not in proportion to their final segmental position. Alternatively, the colonization ultimate size at subsequent stages of development. of the mesoderm by labelled cells could readily be Although the forebrain and midbrain occupy a small explained by a misplaced graft or a local delamination portion of the anterior embryonic ectoderm, by the of the grafted cells from the embryonic ectoderm. early-somite stage they become the most prominent brain segments and constitute over half the tissue More extensive colonization of the paraxial meso- volume of the head folds (Morriss-Kay, 1981; Jacobson derm begins with the 4th cranial somitomeres and & Tarn, 1982). This is probably the result of differential extended to all subsequent somites in embryos receiv- tissue growth in the neural tube. Indeed, it has been ing grafts in the lateral and posterior embryonic ecto- shown in the rat embryo that the most active tissue derm adjacent to the primitive streak. Previous grafting growth is encountered in the developing forebrain, experiments carried out for cells in the primitive streak which may account for the rostral and lateral expansion and adjacent to the streak (Fig. 11: sites 5, 9-11; of this part of the brain during neurulation (Tuckett & Beddington, 1981; Tarn & Beddington, 1987) have Morriss-Kay, 1985). Concomitant to the morphogenesis demonstrated a significant contribution to the paraxial of the neural tube, there is an enormous expansion in mesoderm and lateral mesoderm during gastrulation of the area of the proximal-lateral embryonic ectoderm the mouse embryo. That colonization by graft-derived leading to the separation of the hindbrain and the spinal cells began with the rhombencephalic somitomeres also cord which are originally juxtaposed in the embryonic agrees with the morphological finding that the first ectoderm of the primitive-streak-stage embryo. Analy- three somitomeres are already established in the meso- sis of the movement of cells in the proximal regions of derm at the mid- to late-primitive-streak stage (Tarn & the embryonic ectoderm has revealed a relocation of Meier, 1982) and with the result of grafting studies that cells converging towards the primitive streak during newly recruited paraxial mesoderm from the embryonic gastrulation (Lawson etal. 1987) similar to that de- ectoderm and the primitive streak is allocated to more 66 P. P. L. Tam caudal somitomeres (Tarn & Beddington, 1986, 1987). of germ layers in the mouse at the time of mesoderm formation. Similar to the situation of the ectodermal derivatives, Anat. Rec. 194, 125-142. BEDDINGTON, R. S. P. (1981). An autoradiographic analysis of the there is again striking homology in the mesodermal fate potency of embryonic ectoderm in the 8th day postimplantation of cells in the mouse embryonic ectoderm and that of mouse embryo. J. Embryol. exp. Morph. 64, 87-104. the chick epiblast at the primitive streak stage of BEDDINGTON, R. S. P. (1982). An autoradiographic analysis of development. In both cases, the paraxial mesoderm is tissue potency in different regions of the embryonic ectoderm located lateral to the anterior region of the primitive during gastrulation in the mouse. J. Embryol. exp. Morph. 69, 265-285. streak and posterior to the Hensen's node. The lateral BEDDINGTON, R. S. P. (1983). Histogenetic and neoplastic potential plate mesoderm is found adjacent to and within the of different regions of the embryonic egg cylinder. J. Embryol. middle region of the primitive streak, whereas the exp. Morph. 75, 189-204. extraembryonic mesoderm is associated with the pos- BEDDINGTON, R. S. P. (1987). Isolation, culture and manipulation terior part of the streak (Nicolet, 1971; Meier & of post-implantation mouse embryos. In Mammalian Jacobson, 1982; Packard, 1986; Copp etal. 1986; Tarn & Development. A Practical Approach (ed. M. Monk), IRL Press, Oxford, pp. 43-70. Beddington, 1987). CHAN, W. Y. & TAM, P. P. L. (1988). A morphological and experimental study of the mesencephalic neural crest cells in the Regionalisation of the embryonic ectoderm mouse embryo using wheat germ agglutinin-gold conjugate as Sufficient information is now available to build a map of the cell marker. Development 102, 427-442. prospective ectodermal and mesodermal tissues for the COPP, A. J., ROBERTS, H. M. & POLANI, P. E. (1986). Chimaerism embryonic ectoderm of the primitive-streak-stage of primordial germ cells in the early postimplantation mouse embryo following microsurgical grafting of posterior primitive mouse embryo. Fig. 11 shows the grafts of embryonic streak cells in vitro. J. Embryol. exp. Morph. 95, 95-115. ectoderm that have been investigated: sites 1-4 and 6-8 COULY, G. F. & LE DOUARIN, N. M. (1985). Mapping of the early from the present study and sites 5 and 9-11 from study neural primordium in quail-chick chimaeras. I. Developmental by Tarn & Beddington (1987). Additional information relationships between placodes, facial ectoderm and for sites 1, 6 and 11 is provided by similar grafting prosencephalon. Devi Biol. 110, 422-439. COULY, G. F. & LE DOUARIN, N. M. (1987). Mapping of the early studies by Beddington (1981, 1982) and Copp et al. neural primordium in quail-chick chimaeras. II. The (1986). The neural tube occupies most of the anterior prosencephalic neural plate and neural fold: implications for the and lateral embryonic ectoderm. The proximal areas of genesis of cephalic human congenital abnormalities. Devi Biol. the embryonic ectoderm contain the cells destined for 120, 198-214. the surface ectoderm in the head and oral regions, the DIWAN, S. B. & STEVENS, L. C. (1976). Development of teratomas epidermal placodes and the neural crest cells. Paraxial from ectoderm of mouse egg cylinders. J. natn. Cancer lnst. 57, 937-942. mesoderm (rhombencephalic somitomeres and HUNTER, E. S., BALKAN, W. & SADLER, T. W. (1988). Improved somites) is mapped to the embryonic ectoderm in the growth and development of presomitic mouse embryos in whole distal-lateral regions and adjacent to the primitive embryo culture. J. exp. Zool. 245, 264-269. streak. Within the primitive streak, distinctive regional JACOBSON, A. G. & TAM, P. P. L. (1982). Cephalic neurulation in diversity of tissue fate is observed: the anterior region is the mouse embryo analyzed by SEM and morphometry. Anat. Rec. 203, 375-396. associated with the trunk neural tube and paraxial LAWSON, K. A., MENESES, J. J. & PEDERSON, R. A. (1987). Germ mesoderm, the middle region with the lateral plate layer formation in the mouse embryo studied with an mesoderm and caudal mesoderm and the posterior intracellular lineage tracer. Abstract of Bath Symposium on region with the extraembryonic mesoderm and primor- Craniofacial Development, British Society of Developmental dial germ cells (Copp et al. 1986; Tarn & Beddington, Biology. MEIER, S. & JACOBSON, A. G. (1982). Experimental studies of the 1987). The present fate map, which is based upon the origin and expression of metameric pattern in the chick embryo. result of grafting experiments, is very similar to the map J. exp. Zool. 219, 217-232. constructed by Snow (1981) using an entirely different MEIER, S. & TAM, P. P. L. (1982). Metameric pattern development experimental approach: the prospective fate of the in the embryonic axis of the mouse. I. Differentiation of the embryonic ectoderm is deduced from the types of cranial segments. Differentiation 21, 95-108. MORRISS-KAY, G. M. (1981). Growth and development of pattern embryonic tissues formed in embryo fragments contain- in the cranial neural epithelium of rat embryos during ing a full complement of germ layers. Although a spatial gastrulation. J. Embryol. exp. Morph. 65 Supplement, 225-241. pattern of tissue diversification has been demonstrated NICOLET, G. (1971). Avian gastrulation. Adv. Morphog. 9, 231-262. within the embryonic ectoderm by these studies, it does PACKARD, D. S., JR (1986). The epiblast origin of avian somite not imply that the cells are precommitted to specific cells. In Somites in Developing Embryos, (ed. R. Bellairs, D. A. Ede & J. W. Lash), New York: Plenum Press, pp. 37-45. lineages prior to gastrulation. Although the consensus POELMANN, R. E. (1980). Differential mitosis and degeneration of studies using heterotopic grafts and ectopic tera- patterns in relation to the alterations in the shape of the tomas indicates that there is little regional restriction in embryonic ectoderm of early post-implantation mouse embryos. the developmental potency of embryonic ectoderm cells J. Embryol. exp. Morph. 55, 33—51. (Beddington, 1983; Svajger et al. 1986), definitive POELMANN, R. E. (1981). The head process and the formation of the definitive endoderm in the mouse embryo. Anat. Embryol. information on cell potency has to be obtained by a 162, 41-49. proper clonal analysis of individual cells of the embry- REINIUS, S. (1965). Morphology of the mesoderm, from the time of onic ectoderm (Lawson et al. 1987). implantation to mesoderm formation. Z. Zellforsch. mikrosk. Anat. 68, 711-723. References ROSENQUIST, G. C. (1981). Epiblast origin and early migration of neural crest cells in the chick embryo. Devi Biol. 87, 201-211. BATTEN, B. E. & HAAR, J. L. (1979). Fine structural differentiation SNOW, M. H. L. (1977). Gastrulation in the mouse: growth and Tissue fate of embryonic ectoderm 67

regionalisation of the epiblast. J. Embryol. exp. Morph, 42, in the mouse embryo. In Somites in Developing Embryos, (ed. 293-303. R. Bellairs, D. A. Ede & J. W. Lash), New York: Plenum Press. SNOW, M. H. L. (1981). Autonomous development of parts isolated pp. 17-36. from primitive-streak-stage mouse embryo. Is development TAM, P. P. L. & BEDDINGTON, R. S. P. (1987). The formation of clonal? J. Embryol. exp. Morph. 65 Supplement, 269-287. mesodermal tissues in the mouse embryo during gastrulation and SNOW, M. H. L. & BENNETT, D. (1978). Gastrulation in the mouse: early organogenesis. Development 99, 109-126. assessment of cell populations in the epiblast of tw'8/tw'8 TAM, P. P. L. & MEIER, S. (1982). The establishment of a embryos. J. Embryol. exp. Morph. 47, 39-52. somitomeric pattern in the mesoderm of the gasatrulating mouse SVAJGER, A., LEVAK-SVAJGER, B. & SKREB, N. (1986). Rat embryo. Am. J. Anat. 164, 209-225. embryonic ectoderm as renal isograft. J. Embryol. exp. Morph. THEILER, K. (1972). The House Mouse. Springer-Verlag,'Berlin. 94, 1-27. TUCKETT, F. & MORRISS-KAY, G. M. (1985). The kinetic behaviour TAKOR TAKOR, T. & PEARSE, A. G. E. (1975). Neuroectodermal of the cranial neural epithelium during neurulation in the rat origin of avian hypothalamo-hypophyseal complex: the role of embryo. J. Embryol. exp. Morph. 85, 111-119. the ventral neural ridges. J. Embryol. exp. Morph. 34, 311-325. VAKAET, L. (1984). Early development of birds. In Chimaeras in TAM, P. P. L. & BEDDINGTON, R. S. P. (1986). The metameric Development, (ed. N. M. Le Douarin & A. McLaren). Academic organization of the presomitic mesoderm and somite specification Press, London, pp. 71-88.