Formation of ectopic neurepithelium in chick blastodermsAge-related capacities for induction and self-differentiation following transplantation of quail Hensen's nodes.код для вставкиСкачать
THE ANATOMICAL RECORD 229:437-448 (1990) Formation of Ectopic Neurepithelium in Chick Blastoderms: Age-Related Capacities for Induction and Self-Differentiation Following Transplantation of Quail Hensen’s Nodes MARK S.DIAS AND GARY C. SCHOENWOLF Department of Anatomy, University of Utah, School of Medicine, Salt Lake City, Utah 84132 ABSTRACT Hensen’s node, regarded as the avian and mammalian homologue of Spemann’s neural inducer (i.e., the amphibian dorsal blastoporal lip), has been transplanted in many previous studies to the germinal crescent of avian blastoderms to examine ectopic neural induction. All these studies have suffered from one or more major shortcomings, the most significant of which has been the lack of a reliable cell marker to determine the contributions of graft cells to ectopic embryos. In the absence of such marker, induced (i.e., derived from the host) and self-differentiated (i.e., derived from the graft) neurepithelium cannot be distinguished from one another with certainty. We have transplanted quail Hensen’s nodes to chick host blastoderms and have subsequently used the quail nucleolar heterochromatin marker to identify graft cells unequivocally. We systematically varied both donor and host ages (i.e., stages 3-8 and 3-5, respectively) to examine the effects of age on ectopic neural induction and self-differentiation. Our results demonstrate that the age of the donor is more critical than that of the host over the stages examined. With advancing donor age, the frequency of host induction decreases, while the frequency of graft self-differentiation increases. Previous studies not using cell markers have concluded that the craniocaudal level of the induced neuraxis is determined by the age of the donor, that is, young donors induce cranial neuraxial levels, whereas old donors induce caudal levels. By contrast, we found that with grafts from older donors, neurepithelium was more commonly self-differentiated rather than induced and that progressively more caudal levels of the neuraxis self-differentiated with advancing donor age. Induction of caudal neuraxial levels never occurred in the absence of induced cranial levels. The frequency of neural induction was inversely correlated with the age of the donor and directly correlated with the quantity of graft endodermal cells contributed to the ectopic embryo, supporting a previous assertion that in avian embryos, the earliest and principal source of neural inducer lies within the endoderm rather than mesoderm. From our results, we propose that the role of neural induction is to produce neurepithelium of unspecified regional character, and that the formation of regional character depends on subsequent morphogenetic events. One of the key processes in vertebrate morphogenesis is induction, wherein the fate of a particular tissue is determined through interaction with another, adjacent tissue. One of the most striking inductions-neural induction-occurs during gastrulation. As a result of neural induction, a region of the ectoderm becomes determined to form neurepithelium. In amphibians, the dorsal lip of the blastopore is responsible for induction of neurepithelium. To demonstrate this, Spemann and Mangold (1924) transplanted tissue from the dorsal blastoporal lip to the ventral region of a host embryo, which would not ordinarily become neurepithelium. An ectopic embryo resulted, which was composed 0 1990 WILEY-LISS, INC. of neurepithelium, notochord, endoderm, and (variably) somites and surrounding mesenchyme. Each of the tissues of the ectopic embryo could arise from the host, through induction; from the graft, through self-differentiation; or from both host and graft, through a combination of induction and self-dif- Received March 29, 1990; accepted June 18, 1990. Address reprint requests to Dr. Gary C. Schoenwolf, Department of Anatomy, University of Utah, School of Medicine, Salt Lake City, UT 84132. 438 M.S.DIAS AND G.C. SCHOENWOLF ferentiation. To distinguish among these, Spemann and Mangold transplanted grafts from Triton cristatus to embryos of T. taeniatus; these two species of newt differ in the amount of intracellular pigment granules. The use of such heteroplastic transplants revealed that the mesenchyme, notochord, somites and a portion of the endoderm were graft derived, having arisen through self-differentiation, whereas the neurepithelium was mostly host derived, having arisen chiefly through induction. These experiments led to the recognition of the dorsal blastoporal lip as the organizer of the amphibian embryo. Hensen’s node has been regarded as the avian and mammalian homologue of the amphibian dorsal blastoporal lip (Hara, 1978). Numerous investigators have examined neural induction in birds by transplanting Hensen’s node or other portions of the primitive streak beneath the epiblast of avian blastoderms (Waddington, 1932,1933; Waddington and Schmidt, 1933; Woodside, 1937; Grabowski, 1957,1962; Sher-Pu et al., 1963, 1965; Vakaet, 1964, 1965, 1981, 1984; Gallera and Ivanov, 1964; I-Lan et al., 1965; Shu-Dung et al., 1965; Gallera and Nicolet, 1969; Gallera, 1970a,b, 1971; McCallion and Shinde, 1973; Cuevas and Orts Llorca, 1974; Hornbruch et al., 1979; Sanders and Prasad, 1986). However, all these studies suffer from one or more major shortcomings. Earlier studies were neither consistent in what was transplanted nor thorough in analyzing the results of such transplants; in many (Waddington, 1932, 1933; Waddington and Schmidt, 1933; Woodside, 1937), donor tissues were obtained from various regions of the blastoderm, including a variety of primitive streak levels (with or without Hensen’s node) and portions of the head process region (containing neurepithelium, notochord, paraxial mesoderm, and underlying endoderm), or even from entire blastoderms. In these and other studies (Grabowski, 1957,1962), the ages of either donors or hosts, or both, either were not stated or their effects were not analyzed in a systematic fashion. A more significant shortcoming of many previous studies is that the origin of cells contributing to various tissues of ectopic embryos could not be determined with certainty because a reliable cell marker was not used (Waddington, 1932, 1933; Waddington and Schmidt, 1933; Woodside, 1937; Grabowski, 1957,1962; Sher-Pu et al., 1963, 1965; Vakaet, 1964, 1965; Gallera and Ivanov, 1964; I-Lan et al., 1965; Shu-Dung et al., 1965; Gallera, 1970a,b, 1971). Instead, assumptions were made regarding the origin of ectopic structures based exclusively on morphological criteria. For example, ectopic neurepithelium in continuity with or directly subjacent to the host ectoderm has been assumed to be of host origin, whereas that in continuity with or subjacent to the host endoderm has been assumed to be of graft origin. The use of such criteria may or may not permit valid interpretations. Some recent studies have used reliable cell markers to examine avian neural induction. Gallera and Nicolet (1969) used primitive streak grafts labelled with [3H]thymidine; however, only a single donor age was examined, and Hensen’s node was not grafted. Hornbruch et al. (1979), Sanders and Prasad (1986), McCallion and Shinde (1973), Cuevas and Orts Llorca (1974), and Vakaet (1981, 1984) have employed heteroplastic transplants of Hensen’s node, constructing quaillchick chimeras, to demonstrate that ectopic neurepithelium can form both by induction and by self-differentiation. However, these studies fall short of examining neural induction in a comprehensive fashion; all have used donors and hosts of similar age and have examined only one or two developmental stages, and none has examined the independent effects of donor and host age in a systematic manner. Deficiencies of previous studies on avian ectopic induction were addressed in the present study by transplanting tissue of constant type and quantity, by systematically analyzing the effects of both donor and host age on the frequency and regional character of induction, and by using a reliable cell marker to determine the host or donor origin of cells constituting the various tissues of ectopic embryos. Heteroplastic transplants of quail Hensen’s nodes were made to chick host blastoderms at several donor and host ages; the quail nucleolar heterochromatin marker was used subsequently to identify graft-derived cells in sections through ectopic embryos (Le Douarin, 1973). Hensen’s node was chosen for several reasons: (1)it is the structure that has been transplanted most frequently in the past and its transplantation reliably yields ectopic embryos containing neurepithelium (reviewed by Waddington, 1952; Gallera, 1971); (2) it is considered to be homologous to the neural inducer of amphibians (Hara, 1978); (3) it has well-defined boundaries and can be readily identified in donor blastoderms; and (4) it contains, a t various times, both presumptive endoderm and presumptive chordamesoderm (Nicolet, 1970, 1971), both of which have been strongly implicated in neural induction (Hara, 1961, 1978; Gallera and Nicolet, 1969). The use of the quail cell marker provided important new information. First, it demonstrated definitively the origin of ectopic tissues and suggested that assumptions of previous investigators regarding the origin of some of these may have been erroneous. For example, ectopic neurepithelium adjacent to host ectoderm did not always arise through induction as had been assumed; in some instances, it arose through self-differentiation. Second, the incorporation of graft cells into normal host tissues, such as the gut, the endodermal layer of the proamnion and the neural tube, could be recognized; this was previously impossible with homoplastic transplants. The present study addresses several questions regarding the interaction of tissues during the formation of ectopic embryos: 1. During which stages is the host ectoderm competent to form a neurepithelium under the inductive influence of grafted Hensen’s node? 2. During which stages is Hensen’s node capable of inducing neurepithelium from the host ectoderm? 3. During which stages is Hensen’s node capable of self-differentiating neurepithelium? 4. What effect, if any, do host tissues have on the graft’s ability to self-differentiate neurepithelium? 5. To what extent does induced ectopic neurepithelium undergo the characteristic morphogenetic events of normal neurulation? 6. Is the induced or self-differentiated neurepithelium sufficiently organized to permit identification of NEURAL INDUCTION AND SELF-DIFFERENTIATION specific craniocaudal levels of the neuraxis, and is either the donor or host capable of influencing which levels are formed? 7. Which graft-derived tissues are formed, and which of these are capable of inducing neurepithelium? Our results demonstrate that both neural induction and self-differentiation occur in ectopic embryos; the extent to which each occurs is dependent on the age of both the donor and host at transplantation. Induction of neurepithelium is most vigorous when grafts from young donors are transplanted to young hosts, whereas self-differentiation is most robust when grafts from older donors are transplanted to older hosts. Neither donor nor host age has any effect in specifying craniocaudal level of the induced neuraxis, but with advancing donor age, progressively more caudal regions of the neuraxis self-differentiate from the graft. These results provide a clearer understanding of Spemann’s organizer in avian embryos and establish baseline data for future studies of neural induction. MATERIALS AND METHODS To prepare host embryo cultures, White Leghorn chicken eggs were incubated in humidified incubators for 12-15 h; blastoderms, together with large portions of their adjacent vitelline membranes, were removed from the eggs and prepared for modified New culture (New, 1955). Each blastoderm, still attached to its vitelline membrane, was oriented ventral side up and placed in a 35-mm Petri dish on an agar-albumen substrate consisting of 0.6% Bactoagar (Difco Laboratories, Detroit) in 123 mM NaCl mixed 1 : 1with albumen obtained from fresh, fertilized eggs. A glass ring was then positioned around each blastoderm, and the vitelline membrane was draped over the ring. Such cultured chick blastoderms a t stages 3-5 (Hamburger and Hamilton, 1951) were used as hosts. Those at Hamburger and Hamilton’s stage 3 were further subdivided into four substages, designated 3a, 3b, 3c, and 3d, as described by Schoenwolf (1988). Quail blastoderms were used as donors. These were removed from their vitelline membranes and placed dorsal side up in Spratt culture (Spratt, 1947), using the culture medium just described. Hensen’s node grafts were obtained from these blastoderms at stages 3-8 (Hamburger and Hamilton, 1951), with subdivisions of stage 3 as described above. Each donor Hensen’s node was excised with a cactus needle mounted on a wooden handle, cutting along the cranial and lateral contours of the node. A small portion of the postnodal primitive streak (<lo0 pm in length) was included with the graft. The grafts so excised were approximately bullet shaped, facilitating identification of their cranial and caudal ends. Dorsal and ventral sides were also identifiable because the grafts typically curled ventrally. Each excised graft was transferred with a micropipette from the donor blastoderm and placed atop the host blastoderm. A small hole was made in the hypoblast of the host’s germinal crescent, and the graft was gingerly inserted between the hypoblast and ectoderm near the cranial margin of the area pellucida. Grafts were oriented with their cranial edge directed toward the caudal end of the host blastoderm. This allowed 439 regression of the grafted node toward the area opaca, so that growth of the ectopic embryo would be less encumbered by growth of the host embryo. Most grafts were placed with their ectodermal side adjacent to the host ectoderm, mimicking the dorsoventral orientation of the host. Some grafts were placed with their hypoblast side adjacent to the host ectoderm; that is, the grafts were placed upside down in relationship to the host. The results obtained with the latter grafts were similar to those obtained with the former; therefore, data were combined. Excess saline was removed from the hosts with a micropipette, and the cultures were returned to the incubator for an additional 24-30 h. Both donor and host blastoderms were videotaped during the grafting procedure to document the stages of each and to record the extent of graft excision and its position of placement in the host. All blastoderms were videotaped again when the cultures were terminated; selected ones were also photographed. Of 140 blastoderms with Hensen’s node grafts, 96 formed ectopic embryos; these were prepared for histological study. Each was fixed overnight in a mixture of glacial acetic acid, 37% formaldehyde solution, and 100% ethanol (1:2:7, v:v:v), processed for paraffin embedding and serially sectioned in the plane transverse to the longitudinal axis of the ectopic embryo. Sections were processed according to the Feulgen-Rossenbeck procedure described by Lillie (1965). Donors and hosts were grouped according to stage into one of three categories. For donors, the three categories encompassed stages 3a-3c (hereafter designated as young grafts), 3d-4 + (intermediate grafts) and 5-8 (old grafts). For hosts, the three categories encompassed stages 3b-3c (young hosts), 3d-4 (intermediate hosts), and 4 -5 (old hosts). Each transplantation thus involved a combination of graft and host of specified stages. Results were recorded in 9-quadrant grids, with each cell representing a combination of donor and host ages. We were therefore able to discern the independent effects of donor and host age on both neural induction and self-differentiation. + RESULTS Utility of the Quail Nucleolar Marker Ectopic embryos often had a gross appearance very similar to that of host embryos (Fig. 1). In serial sections through ectopic embryos, the quail marker was used to ascertain the origin of cells contained within various tissues (Figs. 2-4). Ectopic embryos were chimeric, consisting of some tissues derived exclusively from graft cells (e.g., notochord and somites), others derived exclusively from host cells (e.g., surface ectoderm) and still others derived from both graft and host cells (e.g., mesenchyme and endoderm). Ninety of the 96 ectopic embryos contained neurepithelium. In some cases, the neurepithelium arose by self-differentiation (Fig. 2a,b) and in others by induction (Fig. 3a,b). Graft cells incorporated with tissues of the host embryo proper; in all cases, donor cells incorporated with the host’s gut (Fig. 4a) and/or endodermal layer of the proamnion, and in one case, donor cells incorporated with the host’s neural tube (Fig. 4b). The contributions of donor cells to these structures would not have been recognized if a cell marker had not been used. 440 M.S. DIAS AND G.C. SCHOENWOLF Figs. 1-4. NEURAL INDUCTION AND SELF-DIFFERENTIATION 441 Fig. 5. Transverse sections from ectopic embryos collected approximately 24 h after transplantation of a quail Hensen’s node. a: Neurallike vesicle (arrow), which formed from quail cells (Hensen’s node transplanted at stage 4 to a stage 3c host). Note graft cells in the endodermal layer of the host’s proamnion (arrowheads). b Gutlike vesicle (arrow), which formed from quail cells (Hensen’s node transplanted at stage 4 to a stage 3c host). Arrowheads, induced neural plate (on right) continuous with host’s neural plate (on left). Asterisk (*) indicates the transition between the induced and host’s neural plate. a: ~ 3 0 0b; ~ 2 1 0 . Frequency of Neural Induction and Self-differentiation in Ectopic Embryos In addition, in 41 ectopic embryos, the grafts formed vesicles of indefinite character. On the basis of morphology and spatial relationships to other tissues, these vesicles were tentatively classified as neural tissue in 12 embryos (Fig. 5a) and as gut epithelium in the remaining 29 (Fig. 5b). To avoid error, these structures were excluded from the data in the analysis described below. Ectopic neurepithelium was induced most frequently (93% of cases) when grafts from young donors were transplanted to young hosts (Fig. 6a). Induction was common (frequency >50%) when young or intermediate grafts and hosts were combined, when young grafts and old hosts were combined or when old grafts and young hosts were combined. Induction was less frequent with both advancing graft and host age; it declined by an average of 36% when graft age was increased from young to old, and by an average of 41% when host age was increased from young to old (see Fig. 6a legend for details of averaging). These effects were additive; when old grafts were combined with old hosts, only 14%of the ectopic embryos formed induced neurepithelium. By contrast, neurepithelium self-differentiated least frequently when grafts from young or intermediate donors were transplanted to young or intermediate hosts (Fig. 6b). Self-differentiation occurred frequently when old grafts were combined with hosts of any age, or when old hosts were combined with grafts of any age. In ectopic embryos, neurepithelium formed solely by self-differentiation of graft cells in 38 (42%) of cases and solely by induction of host cells in 27 (30%);a further 25 ectopic embryos (28%of cases) each contained more than one portion of neurepithelium, with oneformed by induction and one or more formed by selfdifferentiation. In total, neurepithelium was induced in 63 embryos and it self-differentiated in 52. Fig. 1. Dorsal view of an ectopic embryo (asterisk) approximately 24 h after transplantation of a quail Hensen’s node to the germinal crescent. Note that its morphology is similar to that of the host’s. FB, forebrain. x 40. Fig. 2. Transverse section from an ectopic embryo collected approximately 24 h after transplantation of a stage 8 quail Hensen’s node to a stage 4 host. b is a n enlargement of part of a. Neural tube (hindbraidspinal cord; NT), notochord (N), somites (S), lateral plate (LP) and midline gut endodermal cells (E) have formed from the graft; surface ectoderm (SE) and lateral gut endodermal cells have formed from the host. Note the quail cell marker in b (indicated in the midline endoderm by arrows). a: X 215; b X 560. Fig. 3. Transverse section from an ectopic embryo collected approximately 24 h after transplantation of a stage 4 quail Hensen’s node to a stage 4 host. b is a n enlargement of part of a. Notochord, somites, lateral plate and gut endodermal cells have formed from the graft. Note the quail cell marker in b (examples indicated in the notochord by arrows). a: X 215; b X 710. Fig. 4. Transverse sections from ectopic embryos collected approximately 24 h after transplantation of quail Hensen’s nodes. a: Graft cells have incorporated with the floor (asterisk) of the host’s foregut (Hensen’s node transplanted at stage 3a to a stage 3d host). b: Graft cells (arrow) have incorporated with the host’s neural tube (forebrain/ midbrain) (Hensen’s node transplanted at stage 3c to a stage 3c host). x 275. Regional Character of the Ectopic Neuraxis Neurulation of induced neurepithelium in ectopic embryos closely mimicked that occurring in host embryos (Figs. 1, 7). The induced neural plate underwent characteristic shaping (i.e., apicobasal thickening, transverse narrowing, and longitudinal lengthening) 442 M.S. DIAS AND G.C. SCHOENWOLF FREOUENCY OF NEURAL INDUCTION HOST AGE Young Intermediate Old (3b - 3c) (3d - 41 (4+ - 5 ) 80% (12/15) 76% (13/17) 42% (5/12) 14% (1D) FREOUENCY OF NEURAL SELF-DIFFlERENTZATION HOST AGE Young Intermediate b I Intermediate 29% (4/14) 42% (5/12) 27% (4/15) 35% (6/17) I Old 80% (4/5) I I Fig. 6. Tables showing the frequency of neural induction (a)and self-differentiation (b)after transplantation of quail Hensen’s nodes. Numbers in parentheses indicate number of positive cases/total number of ectopic embryos. Tables are broken up into groups with similar results. To determine the percentage decrease of induction as a function of advancing donor age, the difference between the “young” and “old” percentages for each column were averaged. Similarly, to determine the percentage decrease of induction as a function of advancing host age, the difference between the “young”and “old”percentages for each row were averaged. and bending (i.e., formation of median and dorsolateral hinge points and elevation and convergence of neural folds in proper relationship to the hinge points) (for details of these processes in normal embryos, see Schoenwolf, 1982, 1985; Schoenwolf and Smith, 1990). Moreover, closure of the neural groove occurred on schedule at each craniocaudal level, and neural crest cells left the roof of the neural tube and began their migration. Finally, typical spatial relationships formed among the neural tube, notochord, somites, surface ectoderm, and gut endoderm, even though these structures were collectively derived from both graft and host cells. Embryos with induced ectopic neurepithelium were examined to determine whether the neuraxis had welldefined regional character (that is, typical forebrain, midbrain, hindbrain, and spinal cord levels). Regional character was identified using specific criteria. Forebrain was defined by its broad width and long neural folds, the presence of optic vesicles, and the absence of neural crest cells (Fig. 7a); the lack of an underlying notochord was not, by itself, an identifying feature of forebrain in the absence of other criteria. Midbrain was defined by its oval shape with abundant neural crest cells (Fig. 7b). Hindbrain was defined by its narrower shape and the presence of either otic placodes or somites (Fig. 7c). Spinal cord was defined by its characteristic tall, narrow shape with a small, slitlike central canal, and by the presence of somites (Fig. 7d). Neurepithelium that did not fit the above criteria was labelled only as neural tube/plate. Regional character developed in 45 of the 63 ectopic embryos (71%) with induced neurepithelium (Fig. 8a). Most frequently (40 cases) only rostra1 levels of the neuraxis (forebrain with or without midbrain) were induced. Less frequently (5 cases), the induced neuraxis began rostrally with forebrain and continued caudally to hindbrain or spinal cord levels. Caudal levels of the induced neuraxis were never formed in the absence of more cranial levels. In one case, a small neural structure was induced (Fig. 9). This structure might have been mistaken for a spinal cord in a few sections; however, serial sections revealed that it was a vesicle (<lo0 pm in length) rather than a tube. Therefore, it was classified as induced neuraxis lacking regional character. Recall that neural induction was most frequent with combinations of young donors and hosts (Fig. 6a); similarly, the neuraxis from such combinations frequently developed regional character (77% of cases) (Fig. 8a). Moreover, neuraxes with regional character were induced frequently (250%of cases) when young or intermediate grafts were combined with any age host. The percentage of induced neuraxes with regional character declined by an average of 71% when comparing young and intermediate grafts with old grafts, and regional identity developed in only 1 of 9 cases (4%) in which old grafts were used. By contrast, host age did not affect whether the induced neuraxis developed regional character. Self-differentiated neurepithelium developed regional character less frequently than did induced neurepithelium. Usually, self differentiated neurepithelium completely lacked regional character (Fig. 8b). Recall that neural self-differentiation was least frequent with combinations of young donors and hosts (Fig. 6b); similarly, neurepithelium from such combinations lacked regional character. By contrast to the lack of an effect of donor age on the level of the neuraxis that was induced, donor age had a pronounced effect on the level of the neuraxis that self-differentiated. With young grafts, only 1of 16 embryos (6%)had a self-differentiated neuraxis with regional character; it exhibited cranial (forebraidmidbrain) morphology. In contrast, with old nodes, 12 of 22 cases (55%)had a self-differentiated neuraxis with regional character; of these, all but one exhibited caudal (hindbraidspinal cord) morphology. With intermediate nodes, the response was in between that obtained with young and old nodes; 6 of 14 cases (43%) had a self-differentiated neuraxis with regional character: 4 exhibited cranial morphology and 2 exhibited caudal morphology. Host age had no consistent effect on the regional character of the self-differentiated neuraxis. Formation of Non-neural Structures in Ectopic Embryos Although neurepithelium was the most conspicuous structure formed in ectopic embryos, several nonneural structures also were present in many specimens. Surface ectoderm, otic placodes and, in part, en- NEURAL INDUCTION AND SELF-DIFFERENTIATION 443 Fig. 7. Transverse sections from ectopic embryos collected approximatley 24 h after transplantation of quail Hensen’s nodes. a: Induced forebrain (Hensen’s node transplanted a t stage 4 to a stage 4 host). Arrows, quail endodermal and mesodermal cells; arrowhead, induced lens placode. b Induced midbrain (Hensen’s node transplanted a t stage 4 to a stage 4 host). Arrow, quail notochord. c: Induced hind- brain (Hensen’s node transplanted a t stage 4 to a stage 4 host). Arrows, quail notochord and somite; arrowhead, induced otic placode. d, Induced spinal cord (Hensen’s node transplanted at stage 4 to a stage 3c host). Arrows, quail notochord (flexed) and lateral plate (somatic and splanchnic layers). x 270. doderdgut were formed from host tissues in ectopic embryos. Non-neural structures of graft origin included endoderm in 96 cases (100% of the embryos), mesenchyme and/or lateral plate in 83 cases (86%),notochords in 66 cases (69%), and somites in 26 cases (27%). Notochords formed least frequently with young grafts, and their frequency generally increased with advancing graft age (Fig. 10a). In a manner reminiscent of that seen with self-differentiated neural tubes (cf. Figs. 10a and 6b), notochords frequently developed following combinations of intermediate or old grafts with any age host, or of young grafts with old hosts. The frequency of notochord formation was unaffected by host age. Somites also formed least frequently with young grafts and generally increased in frequency with 444 M.S. DIAS AND G.C. SCHOENWOLF REGIONAL CHARACTER OF INDUCED NEURAXIS HOST AGE Young Intermediate Old 77% (10/13) 23% (3/13) 89% (8/9) 11% (1/9) 80% (4/5) 20% (1/5) I Wl I 50% (1/2) 0% 50% (1/2) 85% (11/13) 25% (3/12) 25% (3/12) 7.5% (1/13) 7.5% (1/13) I I I I I I 0% 100% (1/1) 20% (1/5) 80% (4/5) I 8a I I I REGIONAL CHARACTER OF SELF-DFFERENTIA’IED NEURAXIS HOST AGE F/M T Young Intermediate 0% 20% (1/5) I Intermediate 2 wc T 8 -8 100% (7/7) 16.5% (1/6) 16.5% (1/6) 67% (4/6) 50% (2/4) 0% 50% (2/4) ‘I r 9% (1/11) Old WC b I 1 1 33% (2/6) 67% (4/6) 55% (6/11) 36% (4/11) 60% (3/5) 40% (2/5) T I I I 25% (1/4) 25% (1/4) 50% (2/4) F/M 0% 80% (4/5) 100% (4/4) !?I Old J’ I Fig. 8. Tables showing the frequency with which regional subdivisions formed in induced (a)and self-differentiated (b)ectopic neuraxes. F/M, forebraidmidbrain; F/C, forebraidspinal cord; H/C, hindbraidspinal cord; TIP, neural tubelplate. advancing graft age (Fig. lob). Somites frequently developed with old grafts (average of 73% of cases). The frequency of somite formation was unaffected by host age. The quantity of graft-derived endoderm varied with the graft age. Young grafts typically gave rise to a large quantity of endoderm, some of which formed vesicles (see Fig. 3b), some of which incorporated with the host’s gut (see Fig. 4a) and/or endodermal layer of the proamnion and some of which appeared as lateral evaginations of the host’s gut (Fig. 11). By contrast, old grafts contributed only a few endodermal cells, which typically incorporated with the midline roof of the host’s gut (see Fig. 2b). DISCUSSION There are four major conclusions of this study. First, neural induction is optimal in combinations of younger grafts and hosts, whereas neural self-differentiation is optimal in combinations of older grafts and hosts. Second, in contrast to previous studies, our results show that irrespective of graft age, neural induction always results in formation of cranial neuraxis with or without more caudal neuraxis, provided that typical regional character is established in the induced neuraxis. Thus, graft age does not specify whether cranial or caudal neuraxis is induced. Third, neurulation of induced ectopic neurepithelium closely mimics that occurring during normal development. Therefore, this paradigm can serve as a useful model for studying events underlying neurulation. Fourth, there is a strong correlation between the quantity of graft endodermal cells incorporated with the host’s tissues and the vigor of the induction response. From this we infer that prospective endodermal cells of Hensen’s node likely have a paramount role in neural induction in avian embryos. Collectively, our results suggest that neural induction produces neurepithelium of unspecified regional character, and that the formation of regional character results from subsequent morphogenetic events. NEURAL INDUCTION AND SELF-DIFFERENTIATION Fig. 9. Transverse section from an ectopic embryo collected approximately 24 h after transplantation of a quail Hensen’s node (Hensen’s node transplanted at stage 5 to a stage 3c host). A small neural vesicle (asterisk) has been induced from host epiblast. Arrows, quail notochord, somites and endoderm. X 325. FORMATION OF ECTOPIC NOTOCHORDS 445 Fig. 11. Transverse section from an ectopic embryo collected approximately 24 h after transplantation of a quail Hensen’s node (Hensen’s node transplanted at stage 4 to a stage 3c host). Arrow, induced neural plate; arrowheads, host neural plate; double asterisk, quail gut vesicle; single asterisk, lateral extent of the host’s foregut. x 210. young hosts and decline with advancing host age (Woodside, 1937; Grabowski, 1962; Gallera and Ivanov, HOST AGE 1964), although none of these studies used a cell marker to differentiate between graft and host cells. Young Intermediate Old Woodside (1937) suggested that competence of the host 43% (6/14) 33% (4/12) ectoderm is greatest in stage 3 blastoderms, declines steadily with advancing age and is lost completely beyond stage 6. However, neither donor age nor graft composition was held constant in that study. Grabowski (1962) reported that induction occurred frequently (70% of cases) after transplantation of stage 4 Hensen’s nodes to stage 4 hosts, whereas it occurred infrequently (30% of cases) after their transplantation to stage 5 hosts. Gallera and Ivanov (1964) concluded that host ectoderm was competent to form ectopic neuFORMATION OF ECTOPIC SOMITES ral tubes from stages 2-4, but it could form only “neuroidal placodes” between stages 4 and 6, and it comHOST AGE pletely lost neural competence beyond stage 6. The frequency of induction of ectopic neurepithelium as a Young Intermediate Old function of donor age has been virtually ignored. There W Young 7%(1/14) 0% (0/12) 11% (1/9) is only one study suggesting that ectopic neural induc0 tion is more frequent with grafts from younger donors d and declines with advancing graft age (Gallera, 1970a). z Taken together with other results, our present data, 0 based on the use of a reliable cell marker, demonstrate 9 Old 80%(4/5) 83% (10/12) 57% (4/7) I I I I that the frequency of induction depends on both the b graft’s inductive ability (a function of donor age) and the competence of the host ectoderm (a function of host Fig. 10. Tables showing the frequency of notochord (a)and somite age) (Fig. 6a). Host competence is present a t the earli(b)formation in ectopic embryos. est stages examined and declines rapidly after stage 4. Frequency of Neural Induction Declines With Advancing Hensen’s nodes from young donors are the most capaHost and Donor Age ble of inducing neurepithelium from the overlying ecThere is general agreement that both the frequency toderm and this capability is diminished progressively and the quality of ectopic neural induction are best in with advancing graft age. The effects of advancing 1 446 M.S.DIAS AND G.C. SCHOENWOLF graft and host age are additive; although an old graft can still elicit a frequent inductive response with a young host, or an old host with a young graft, induction occurs rarely when both graft and host are old. The quality of the inductive response also declines a t least with advancing host age; well-formed neural tubes are induced when both graft and host are young, whereas nonspecific neuroidal structures are induced more frequently with advancing host age (Woodside, 1937; Gallera and Ivanov, 1964). The quality of induction also depends on the length of exposure to the inducing tissue; only neural plates are induced after 6 h of contact between Hensen’s node and the area opaca of a competent host epiblast, whereas well-formed neural tubes with regional character develop after 8Yz h of contact (reviewed by Gallera, 1971). These data are compatible with models (Saxen, 1980; Gurdon, 1987) based on interactions between an inducer substance and a receptor in the competent ectoderm: advancing donor andior host age could result in a loss of inductive interaction through a loss of inducer substance or of host receptors. Frequency of Neural Self-differentiation Increases With Advancing Host and Donor Age In contrast to neural induction, which declines with advancing graft and/or host age, neural self-differentiation increases with increasing graft or host age or both. Previous studies have lar ely ignored the effects of graft or host age on neural se f-differentiation. Only one has examined graft age (Veini and Hara, 1975); this showed that when older nodes were transplanted to the coelomic cavity, the frequency of neural selfdifferentiation increased. None has examined the effects of host age on neural self-differentiation. Why do graft and host age affect the frequency of neural self-differentiation? There are three likely possibilities: (1)there are fewer prospective neurepithelial cells in younger nodes than in older ones; (2) prospective neurepithelial cells are more likely to be lost from younger nodes than from older ones; and (3)the differentiation of prospective neurepithelial cells is favored with older nodes, and/or older hosts provide a more conducive environment for neural differentiation. The first possibility could explain only the effect of graft age on neural self-differentiation. Unfortunately, there is neither evidence for nor against it, so a conclusion cannot be drawn at present. The second possibility addresses the effect of both graft and host age. Prospective neurepithelial cells could be lost in two ways: by death, which might be greater after transplantation of younger nodes (but no evidence exists on this point); or by dispersion and incorporation with host tissues. In the latter case, prospective neurepithelial cell dispersion might be greater in younger nodes, containing principally prospective endodermal cells, than in older nodes, containing abundant prospective mesodermal cells (Nicolet, 1970; 1971; Veini and Hara, 1975). Perhaps such prospective mesodermal cells anchor the prospective neurepithelial cells and allow them to undergo neural differentiation and morphogenesis. The age of the host also could play a role in the amount of dispersion and incorporation that occur; namely, tissues of older hosts might inhibit graft cell dispersion and might be more refractory to the invasion of graft cells K than are tissues of younger hosts. The third possibility also could explain the effect of both graft and host age. It would be expected that with older grafts, prospective neurepithelial cells would have had longer exposure to neural inducer, or would have had more time to initiate differentiation, before their removal from the donor and their transplantation to the host. Moreover, with older hosts it would be expected that more (or perhaps a “better”) extracellular matrix might have been synthesized in the germinal crescent region, thereby potentially providing a more optimal site for neural differentiation. Obviously, our understanding of neural induction in avian embryos is far from complete and further studies will be required to ascertain which of the above possibilities play the most significant roles in neural self-differentiation following transplantation of Hensen’s node. In addition to the well-formed neural tubes that selfdifferentiated from a large percentage of the grafts, many vesicles of indefinite character self-differentiated; some of these were more characteristic of neurepithelium and others were more characteristic of gut epithelium. Both types of vesicles were present in similar locations (often underlying the induced neurepithelium) and many were connected with either the underlying endodermal layer of the proamnion or host gut. All but two of these vesicles self-differentiated from grafts encompassing stages 3b-4, when endoderm is known to be present within Hensen’s node (Nicolet, 1970,1971; Veini and Hara, 1975). Although these vesicles may have appeared to be either more neural or more endodermal in morphology, depending on the individual case, their tissue of origin remains uncertain in the absence of either an endodermal or neural-specific marker. Therefore, depending on how these vesicles were classified in previous studies, one could achieve a higher rate of either homeogenetic induction (if vesicles were regarded as neural) or nonhomeogenetic induction (if all vesicles were regarded as endodermal). Regional Identity of the Self-differentiated, But not Induced, Neuraxis Depends on Graft Age The concept of an embryonic organizer requires not only that the organizer be capable of inducing axial structures, such as neurepithelium, but that it also be capable of organizing the expression of regional character in these tissues (Grabowski, 1957). Taken one step further, the assumed homology of Hensen’s node with the amphibian organizer (i.e., the dorsal blastoporal lip) implies that Hensen’s node is capable of specifying the levels of neuraxis formed from induced tissues. Accordingly, it would be predicted that younger nodes would induce cranial levels of the neuraxis, whereas older nodes would induce caudal levels. The ability of Hensen’s node to control regional organization of the neuraxis was seemingly demonstrated in avian embryos by Vakaet (1965) and Gallera (1970a); only caudal levels of the neuraxis were reported to be induced following transplantation of older grafts, whereas either cranial levels (Gallera, 1970a) or both cranial and caudal levels (Vakaet, 1965) were reported t o be induced following transplantation of younger nodes. In contrast to previous studies (Vakaet, 1965; NEURAL INDUCTION AND SELF-DIFFERENTIATION Gallera, 1970a), our data do not support the view that Hensen’s node is responsible for specifying which craniocaudal levels of the neuraxis are induced. Older grafts in our series never induced caudal neuraxis in the absence of more cranial neuraxis. However, older grafts self-differentiated neurepithelium with greater frequency than they induced neurepithelium, and this self-differentiated neurepithelium typically had a caudal morphology. We suspect that these self-differentiated neural tubes were mistaken for induced neurepithelium in the previous studies, which did not use a cell marker. Alternatively, it is possible that neural vesicles (e.g., Fig. 9) were interpreted as spinal cords because of their narrowness. The fact that these structures are spherical rather than tubular argues against the validity of making such an interpretation. In addition, our study showed cranial levels of the neuraxis to be induced more frequently than caudal levels, regardless of the age of the graft. Moreover, there is a suggestion that in addition to cranial levels, caudal levels of the neuraxis were induced more frequently with both advancing graft age and declining host age (although the number of embryos with induction of both cranial and caudal levels of the neuraxis was small). The host effect can be explained by our present results, which clearly reveal that the host’s neural competence decreases with advancing age. However, the reason for a donor effect is less clear. Hensen’s node undergoes progression (i.e., moves cranially) during stages 3a-4 and regression (i.e., moves caudally) from stage 4 onward (e.g., Nicolet, 1971) and regional character of the neuraxis is manifested only as Hensen’s node regresses (e.g.,Schoenwolf,1982;Schoenwolf and Smith, 1990). Hence, older nodes may be able, by virtue of their stage in development, to organize induced neurepithelium into its various subdivisions sooner than would younger nodes. Because cultures were terminated at similar periods following node transplantation, regardless of node age, it would be expected that older nodes would organize more levels of neuraxis than would younger nodes over the same length of time. Our results just discussed are in general agreement with those of Gallera and Ivanov (1964), although they differ in that ectopic embryos with both cranial and caudal levels of the neuraxis formed less frequently in our study. This disparity may be owing to differences in the length of culture following transplantation, differences in the culture conditions or to species differences, as we used heteroplastic rather than homoplastic grafts. Ectopic-Induced Neurepithelium Undergoes Typical Neurulation Events Neurulation is a complex process involving forces generated by both the neurepithelium and surrounding tissues (e.g., Schoenwolf and Smith, 1990). Our results provide evidence that the induced ectopic neurepithelium undergoes the characteristic morphogenetic events of neurulation. Of particular importance is the observation that both median hinge point cells (neurepithelial cells overlying the notochord; MHP cells) and the more lateral neurepithelial cells (L cells) in induced ectopic neurepithelium are always derived exclusively from the host ectoderm. Previous experimen- 447 tal evidence suggests that MHP cells are induced by the subjacent notochord (van Straaten et al., 1988; Smith and Schoenwolf, 1989), which in our present study is always derived from graft (quail) cells. This observation would seemingly suggest that most (or perhaps all) notochordal and MHP cells are not derived from a common precursor population as has been proposed recently (Jessell et al., 1989). Endoderm Is the Most Likely Primary Source of Spemann’s Neural Inducer in Birds Graft cells contributed to a variety of non-neural structures, including notochord, somites, mesenchyme and endoderm. Although graft-derived notochords and somites would be identified easily in previous studies lacking a cell marker, graft contributions to host endoderm, mesenchyme and neural tube would not have been identified readily without such a marker. The quail nucleolar marker enabled us to make correlations between the extent to which graft cells contributed to these tissues and the induction of ectopic neurepithelium. No obvious correlations between the presence of graft-derived notochord, somites, neurepithelium or mesenchyme and the frequency of ectopic neural induction were found. By contrast, we found a striking correlation between the degree to which graft cells contributed to the host endoderm and the frequency of neural induction. These cells not only incorporated with the host’s endoderm but frequently extended laterally, subjacent to the induced neurepithelium. Ectopic neurepithelium was induced more frequently when young grafts (which contain more endoderm: Nicolet, 1970,1971; Veini and Hara, 1975) were used, and such grafts contributed a greater quantity of cells to the host endoderm. Graft cells in the host endoderm are not necessarily exclusively prospective endodermal cells. It is possible that some prospective ectodermal or mesodermal cells may have changed their fates following transplantation and incorporated with the host endoderm. Nevertheless, our results support the assertion of Gallera and Nicolet (1969) that the source of the neural inducer in birds is the endoderm and not the mesoderm (see also Hara, 1978, for a review). Further studies are underway to determine the precise role of the endoderm in neural induction in avian embryos. ACKNOWLEDGMENTS This research was supported by grant NS18112 from the National Institutes of Health. We wish to acknowledge the technical assistance of Fahima Rahman and secretarial assistance of Jennifer Parsons. LITERATURE CITED Cuevas, P., and F. Orts-Llorca 1974 Induction primaire chez le jeune blastoderme de poulet (Gallus gullus) par greffe du nceud de Hensen de caille (Coturniz coturnix juponica). Acta Anat., 89: 423-430. Gallera, J. 1970a Inductions cbrebrales et medullaires chez les Oiseaux. Experientia, 26t886-887. Gallera, J. 1970b Difference de reactivite a l’inducteur neurogene entre I’ectoblaste de l’aire opaque et celui de l’aire pellucide chez le Poulet. Experientia, 26:1353-1354. Gallera, J. 1971 Primary induction in birds. Adv. Morph., 9t149-180. Gallera, J., and I. Ivanov 1964 La competence neurogene du feuillet externe du blastoderme de Poulet en fonction du facteur “temps.” J. Embryol. Exp. Morph., 12r693-711. 448 M.S.DIAS AND G.C. SCHOENWOLF Gallera, J., and G. Nicolet 1969 Le pouvoir inducteur de l’endoblaste presomptif contenu dans la ligne primitive jeune de Poulet. J. Embryol. Exp. Morph., 21t105-118. Grabowski, C.T. 1957 The induction of secondary embryos in the early chick blastoderm by grafts of Hensen’s node. Am. J. Anat., 101: 101-133. Grabowski, C.T. 1962 Neural induction and notochord formation by mesoderm from the node area of the early chick blastoderm. J . Exp. Zool., 150t233-245. Gurdon, J.B. 1987 Embryonic induction-Molecular prospects. Development, 99r285-306. Hamburger, V., and H.L. Hamilton 1951 A series of normal stages in the development of the chick embryo. J. Morph., 88r49-92. Hara, K. 1961 Regional neural differentiation induced by prechordal and presumptive chordal mesoderm in the chick embryo. Ph.D. thesis, Utrecht. Hara, K. 1978 ‘Spemann’s organizer’ in birds. In: Organizer-A milestone of a half-century from Spemann. 0. Nakamura and s. Toivonen, eds. Elsevi&lNorth-Holland Biomedical Press, Amsterdam, pp. 221-265. Hornbruch, A., D. Summerbell, and L. Wolpert 1979 Somite formation in the early chick embryo following grafts of Hensen’s node. J. Embryol. Exp. Morph., 5 1 5 - 6 2 . I-Lan, N., T. Shu-Dung, and S. Sher-Pu 1965 Studies on the regional responses to inductor of the chick blastoderms of different ages. Acta Anat. Sin., 8r324-333. Jessell, T.M., P. Bovolenta, M. Placzek, M. Tessier-Lavigne, and J . Dodd 1989 Polarity and patterning in the neural tube: the origin and function of the floor plate. In: Cellular Basis of Morphogenesis. Ciba Foundation Symposium. John Wiley & Sons, Chichester, pp. 257-282. LeDouarin, N.M. 1973 A Feulgen-positive nucleolus. Exp. Cell Res., 77:459-468. Lillie, R.C. 1965 Histopathologic Technique and Practical Histochemistry. 3rd Ed. McGraw-Hill, New York. McCallion, D.J., and V.A. Shinde 1973 Induction in the chick by quail Hensen’s node. Experientia, 29.321-322. New, D.A.T. 1955 A new technique for the cultivation of the chick embryo in vitro. J . Embryol. Exp. Morphol., 3t326-331. Nicolet, G. 1970 Analyse autoradiographique de la localisation des differentes ebauches presomptives dans la ligne primitive de l’embryon de Poulet. J . Embryol. Exp. Morph., 23r79-108. Nicolet, G. 1971 Avian gastrulation. Adv. Morph., 9t231-262. Sanders, E.J., and S. Prasad 1986 Epithelial and basement membrane responses to chick embryo primitive streak grafts. Cell Diff., 18: 233-242. Saxen, L. 1980 Neural induction: Past, present, and future. In: Current Topics in Developmental Biology. R.K. Hunt, ed. Academic Press, New York, pp. 409-418. Schoenwolf, G.C. 1982 On the morphogenesis of the early rudiments of the developing central nervous system. Scann. Electron Microsc., 19821It289-308. Schoenwolf, G.C. 1985 Shaping and bending of the avian neuroepithelium: Morphometric analyses. Dev. Biol., 109t127-139. Schoenwolf, G.C. 1988 Microsurgical analyses of avian neurulation: Separation of medial and lateral tissues. J. Comp. Neurol., 276: 498-507. Schoenwolf, G.C., and J.L. Smith 1990 Mechanisms of neurulation: Traditional viewpoint and recent advances. Development 109: 243-270. Sher-Pu, S., N. I-Lan, and T. Shu-Dung 1963 Studies on the inductive action of the Hensen’s node following its transplantation in ovo to the early chick blastoderm. Acta Biol. Exp. Sin., 8.441-462. Sher-Pu, S., N. I-Lan, and T. Shu-Dung 1965 Experimental analysis on the reactive capacity of the epiblast of the chick blastoderm. Acta Anat. Sin., 8tl-10. Shu-Dung, T., N. I-Lan, and S. Sher-Pu 1965 Studies on the inductive action of the Hensen’s node following its transplantation in ovo to the early chick blastoderm. 11. Regionally specific induction of the node region of different ages. Acta Biol. Exp. Sin., lOt69-80. Smith, J.L., and G.C. Schoenwolf 1989 Notochordal induction of cell wedging in the chick neural plate and its role in neural tube formation. J. Exp. Zool., 250t49-62. Spemann, H., and H. Mangold 1924 Uber induktion von Embryonalanlagen durch Implantation artfremder Organisatoren. Roux Arch. Entwichlungsmech. Org., 100.599-638. Spratt, N.T., J r . 1947 Development in vitro of the early chick blastoderm explanted on yolk and albumen extract saline-agar subtrata. J. Exp. Zool., 106t345-365. Straaten, H.W.M. van, J.W.M. Hekking, E.J.L.M. Wiertz-Hoessels, F. Thors, and J. Drukker 1988 Effect of the notochord on the differentiation of a floor plate area in the neural tube of the chick embryo. Anat. Embryol., 177t317-324. Vakaet, L. 1964 Diversite fonctionnelle de la ligne primitive du blasBiol., 158~1964-1966. toderme de Poulet. C.R. SOC. Vakaet, L. 1965 Resultats de la greffe de naeuds de Hensen d’bge different sur le blastoderme de Poulet. C. R. SOC.Biol., 159:232233. Vakaet, L. 1981 Neurale inductie bij vogels. K. Acad. Gen. Belg., 43~79-103. Vakaet, L. 1984 Early development of birds. In: Chimeras in Developmental Biology. N. LeDouarin and A. McLaren, eds. Academic Press, London, pp. 71-88. Veini, M., and K. Hara 1975 Changes in the differentiation tendencies of the hypoblast-free Hensen’s node during “gastrulation” in the chick embryo. Roux Arch. Entwichlungsmech. Org., 177:89-100. Waddington, C.H. 1932 Experiments on the development of chick and duck embryos, cultivated in vitro. Philos. Trans. R. SOC.Lond. (Biol.). 221~179-230. Waddington, C.H. 1933 Induction by the primitive streak and its derivatives in the chick. J. Exp. Biol., 10t38-46. Waddington, C.H. 1952 The Epigenetics of Birds. Cambridge . - Univers i t y Press, Cambridge. Waddinpton. C.H., and G.A. Schmidt 1933 Induction bv heterodastic gra&sof the primitive streak in birds. Roux Arch. E n t w i c h n g s mech. Org., 128522-563. doodside, G.L. 1937 The influence of host age on induction in the chick blastoderm. J. Exp. Zool., 75t259-281.