The mesodermal cell layer is created by ingression and migration of the cells from the primitive streak region in mouse embryos on day 7 of pregnancy. In order to study the mechanisms of mesodermal cell migration during development, the mesodermal cells isolated from the primitive streak were cultured on various substrata, and cell behaviour and motility were analysed with a time-lapse video system. The mesodermal cells on the surface of extracellular matrix (ECM)-coated dishes (ECM produced by bovine corneal endothelial cells) showed extensive migration at a mean rate of approx. 50μmh-1. They also showed frequent cell division and exhibited contact paralysis of lamellipodia and contact inhibition of movement. On plastic or glass surfaces, however, the mesodermal cells became more flattened and less motile (approx. 20–30μm h-1). Cell shape and mean rate of movement on the ECM were very similar to those in situ, as investigated in a previous study (Nakatsuji, Snow & Wylie, 1986). Therefore, this culture condition could provide a useful experimental system for analysing the cellular basis of normal and abnormal morphogenetic movements in mouse embryos.

Employing such a culture system, we studied motility of the mesodermal cells from embryos homozygous for Brachyury (T) mutation, which are lethal at the midgestation stage in utero. Histological observations have suggested that anomalous morphogenesis of the T/T embryos may be brought about by defects in migration of the mesodermal cells derived from the primitive streak. When mesodermal cells from the primitive streak of the T/T mutant embryos on days 8–9 were cultured on the ECM substratum, mean rate of cell migration was significantly reduced compared to cells from normal embryos. Results support the idea of retarded migration by the mutant mesodermal cells as an important factor causing abnormalities in morphogenesis.

There has been a rapid increase in our knowledge of the locomotion of tissue cells in vitro (Porter & Fitzsimons, 1973; Trinkaus, 1976; Bellairs, Curtis & Dunn, 1982). Although most tissue cells in adult animals retain a potential ability for active migration, such motility is mainly displayed during embryonic development, where directional movement of cells takes place to build the complex organization of tissues and organs (Trinkaus, 1984; Armstrong, 1985). However, we know relatively little about the movement of embryonic cells involved in morphogenetic movements, because of our poor understanding of the nature of microenvironmental cues to which cells are responsive during the course of their movements (Trinkaus, 1976).

The study of events occurring inside multicellular masses is much more difficult than the study of isolated individual cells in vitro. In the case of opaque embryos such as those of amphibians and birds, it is not possible to look at moving cells in vivo. Morphological studies of fixed samples have provided information about the static features of cells presumed to be moving. On the other hand, attempts to observe cell movements in vivo have been successful in the case of the relatively transparent embryos such as those of sea urchin (Dan & Okazaki, 1956; Gustafson & Kinnander, 1956), nematode (Deppe et al. 1978) and fishes (Trinkaus, 1973, 1985; Eisen, Myers & Westerfield, 1986). Even in such successful studies, however, it has been difficult to manipulate the cells’ environment in an attempt to learn more about the control of morphogenetic cell movements. Therefore, as an alternative, we have examined behaviour of cells isolated from embryos in order to obtain additional information on cell movement. Such in vitro study allows a detailed examination of individual embryonic cells and the interactions among them under precisely controlled environmental conditions. In these studies, it is most important to find culture conditions that permit embryonic cells to exhibit a similar motility in vitro to that in vivo (Nakatsuji & Johnson, 1982).

In recent years, there have been important advances in the analysis of the various morphogenetic cell migrations, such as the neural crest cells (Le Dourain, 1984; Thiery, Duband & Tucker, 1985a; Thiery, Boucaut & Yamada, 19856), primordial germ cells (Heasman et al. 1981; Heasman & Wylie, 1983; England, 1983), amphibian mesodermal cells (Nakatsuji, 1984; Thiery et al. 1985b) or sea urchin primary mesenchymal cells (Katow, 1986). From these studies, it has become apparent that extracellular matrix (ECM), such as the basal lamina, plays a key role in control of embryonic cell migration by providing substrata for cell attachment and movement.

We still know little about gastrulation movement in mammals (for a review see Bellairs, 1986). Although early mouse embryos are relatively transparent because of the small amount of yolk, they are not accessible in situ due to the enveloping uterine and decidua tissues. Based on examination of fixed materials, two conflicting hypotheses on the formation of the mesodermal cell layer have been introduced. One hypothesis is that the mesodermal cells migrate actively away from the primitive streak in both anterior and antimesometrial directions (Batten & Haar, 1979; Tam & Meier, 1982; Franke et al. 1982; Franke, Grund, Jackson & Illmensee, 1983; for a review see Beddington, 1983). The other is that the mesodermal cells do not migrate away from the primitive streak, but that they only appear to do so because of a shift in the relative positions of the embryonic ectoderm layer and of the primitive streak during rapid growth of the embryo (Poelmann, 1981). Recently, Nakatsuji et al. (1986) attempted to look directly at the migration of the mesodermal cells inside the whole embryo using whole embryo culture and time-lapse cinemicrography with Nomarski differential interference contrast (DIC) optics. They showed that the mesodermal cells are motile and migrate actively away fom the primitive streak in both anterior and antimesometrial directions at a mean rate of approx. 46μmh-1.

Mutations that cause developmental defects in early embryogenesis, such as T/t complex on chromosome 17 of the mouse, provide useful tools in the search for mechanisms of mammalian morphogenesis (Gluecksohn-Waelsch & Erickson, 1970; Bennett, 1975; Sherman & Wudl, 1977). Brachyury (T) mutation is lethal at the midgestation stage in homozygotes, where it severely affects the axial organization of the posterior trunk up to, but not including, the forelimb buds on day 9 (Chesley, 1935). It also retards extension of the allantois, inhibiting development of the complete umbilical blood circulation (Gluecksohn-Schoenheimer, 1944). Histologically, the chordamesoderm in the posterior region of the T/T embryo develops extensive tubular structures instead of the notochord and somites by day 9 (Spiegelman, 1976; Fujimoto & Yanagisawa, 1979). Grüneberg (1958) proposed that anomalies of the notochord may be involved in the massive abnormalities of the posterior trunk region in the T/T embryos as a part of a more general disturbance of the primitive streak. Indeed, the head process, derived from the cephalad extremity of the primitive streak, is defective and the prenotochordal cells are fewer in the archenteron area (Fujimoto & Yanagisawa, 1983). It is known that T/T embryos have a bulky primitive streak at the middle of day 8 or the later stages (Spiegelman, 1976). Morphometric analyses of histological sections of day-6 or -7 embryos suggest that cells derived from the epiblast layer of T/T embryos are stuck in the primitive streak region (Yanagisawa, Fujimoto & Urushibara, 1981). These morphological studies have suggested that anomalous morphogenesis of T/T embryos is caused by lesions in the migration of cells invaginated from the primitive streak. Due to these effects in morphogenesis, the T/T mutant serves as an excellent probe to study mechanisms of mesodermal cell migration during development.

In the present study, we examined the behaviour of mesodermal cells isolated from the primitive-streakstage mouse embryos. We have compared three kinds of substrata for cell culture in order to find culture conditions that allow mesodermal cells to move in a similar fashion to that in vivo. We found that dishes coated with ECM produced by endothelial cells provide substrata on which the mesodermal cells exhibit behaviour most similar to that inside embryos in terms of cell shape, rate of movement and cell division. Employing such culture conditions, we examined the effect of T mutation on the motility of mesodermal cells by comparing the behaviour of individual cells isolated from the T/T embryos with that of those from normal embryos. Results directly prove the T/T mutant mesodermal cells to be retarded in their migration ability.

Mice

Randomly bred ICR (Crj:CD-1) strain mice were purchased from Charles River in Japan and kept at a normal (7 a.m. –7 p.m. light) or reversed (4p.m. –4 a.m. light) light/dark cycle. Heterozygous mice (T/+) were originally supplied by Dr D. Bennett at the Sloan-Kettering Institute for Cancer Research. Mice homozygous for the Robertsonian translocation Rb (4,17) 13 Lub (hereafter designated Rb13) were originally supplied by the late Dr A. Gropp, Medizinische Hochschule, Lübeck. They were in the 25th and the 8-10th inbreeding generation, respectively, at the time of the present study. The T/+ males were mated to Rb13/Rb13 females to produce short-tailed F1 progeny carrying the T mutation and the Rb13 translocation (designated T/Rb13). These F1 mice were maintained under a reversed light cycle.

Embryos

A male was mated with three to five females during the dark period. When the vaginal plug was detected at the beginning of the light period, the midpoint of the preceding dark period was designated as day 0 of pregnancy. Embryos used in this study were obtained from ICR females of day 6·5,7, 7·5, 8 or 9 of pregnancy and T/Rb13 females of day 8 or 9 of pregnancy mated with T/Rb13 males.

Pregnant females were killed by cervical dislocation. Uteri were dissected out and transferred to Petri dishes containing PB-1 solution (Whittingham & Wales, 1969). Embryos were freed from the decidua tissue using fine forceps under a binocular stereomicroscope.

T/T embryos derived from T/Rb13 intercrosses were distinguished from their normal littermates according to the criteria of morphological abnormalities in the posterior body and the allantois. The normal littermates, T/Rb13 and Rb13/Rb13, were used as controls. Effectiveness of the morphological identification of T/T embryos was confirmed by karyotypic analysis of the cultured embryonic cells; cells derived from morphologically identified T/T homozygotes contained no Robertsonian translocated chromosomes and those from the normal embryos contained one or two translocated chromosomes.

Isolation of the mesoderm cell layer

The mesoderm cell layer was isolated from the primitivestreak-stage embryo (egg-cylinder stage) obtained from females of day 6·5 to 7·5 of pregnancy with sterile microsurgery and enzyme treatment, as described by Snow (1978a). Briefly, the embryonic portion was dissected from the egg cylinder and a single cut was made along the primitive streak with a tungsten needle. The tissue fragment was transferred into a mixture of 0·5% trypsin (Difco 1:250) and 2·5% pancreatin (Sigma) in Ca2+,Mg2+-free phosphate-buffered saline (PBS-) for 5 to 15 min at 4°C. Enzyme digestion was stopped by rinsing twice with PBS— supplemented with 2 % calf serum. Afterwards, the fragment was gently agitated with a flame-polished micropipette. The mesoderm cell layer was carefully separated from both the embryonic ectoderm and the primitive endoderm layers with a pair of sharpened tungsten needles. Using a stereomicroscope at higher magnifications, we could distinguish the mesoderm layer, which appeared as a loosely connected thin cell sheet with spaces present between cells, from the much thicker cohesive cell sheet of the embryonic ectoderm and the primitive endoderm, which quickly shrank into a rounded-up cell mass. After rinsing twice with culture medium, fragments of the mesodermal cell sheets were torn into small pieces and explanted in culture dishes.

In the case of 8- to 9-day-old embryos, the mesodermal cells were isolated from the primitive streak region occupying the caudal portion of the embryo, using the same microsurgery and enzyme digestion.

Culture conditions

Three kinds of substrata were used in this study. Extracellular matrix (ECM)-coated dishes were purchased from International Bio-Technologies Ltd (Jerusalem, Israel). The dish surface had been coated with basement membrane ECM produced by endothelial cells. The others were Falcon plastic tissue culture dishes (Falcon 3001) and glass coverslips placed in Falcon plastic Petri dishes (Falcon 1008). Small pieces of the mesoderm cell layer isolated from each embryo were transferred into one of these dishes containing Dulbecco’s modified minimum essential medium (DMEM) supplemented with glucose (final, 4·5 g l-1) and 20 % fetal calf serum (Gibco). They were incubated at 37°C with an atmosphere of 5 % CO2 and 95 % air.

Time-lapse video recording and analysis

Cell movement was recorded using a time-lapse video system (Professional Editing Recorder BR-8600, JVC controlled by Time-lapse Video Controller SIV, Sankei, Tokyo) equipped with a video camera (C1965, Hamamatsu Photonics K. K., or CTC-2100, Ikegami Tsushin Co. Ltd), which was connected to an inverted microscope (Nikon Diaphoto-TMD or Olympus IMT-2) equipped with a heated (37°C) box using ×10 or ×20 phase-contrast objective lenses. Single recording for 0·1s was repeated at 30s intervals.

Video recording during the period from 12 to 24 h after the start of culture was analysed with a video digitizer (For.A, Tokyo, Model FVW-300) connected to a personal computer (NEC PC-9801). Rate of cell migration was obtained by tracing the centre of the nucleus of each cell every 30 min for the total of 5 h. When a cell entered mitotic division during tracing, one of the daughter cells was chosen for the succeeding tracing. When a cell left the field, its tracing was stopped and discarded from the data. Movement of approximately 25 to 100 cells was analysed from one video recording.

Cell behaviour

Explants of the mesodermal cell sheet attached to the bottom of an ECM-coated dish within 1 h of culture. Attached cell masses continued to change their shapes due to cell dislocation and cell division accompanied by vigorous blebbing within the cell mass. Peripheral cells protruded cell processes onto the dish surface. Shortly, marginal cells began to migrate individually away from the periphery of the cell mass (Fig. 1A). The whole cell masses became gradually flattened as outgrowth proceeded. At the beginning of outgrowth, individual cells migrated radially away from the cell mass. Once fully dispersed, they moved about apparently randomly on the substratum. The outgrowth was completed within 12h of culture, leaving no cell aggregate at the centre (Fig. 1B).

Fig. 1.

Phase-contrast micrographs of the mesoderm (A–C) or embryonic ectoderm (D) explants on the surface of ECM-coated dishes. (A) Start of outgrowth after a few hours of culture. (B) Cells have completed their outgrowth and dispersion by 12 h of culture. (C) A higher magnification view of the mesodermal cells during their extensive migration. (D) The embryonic ectoderm cells remain as a cohesive cell sheet even after one day of culture. Marginal cells frequently extend very thin large lamella (arrows). Bars, 100μm.

Fig. 1.

Phase-contrast micrographs of the mesoderm (A–C) or embryonic ectoderm (D) explants on the surface of ECM-coated dishes. (A) Start of outgrowth after a few hours of culture. (B) Cells have completed their outgrowth and dispersion by 12 h of culture. (C) A higher magnification view of the mesodermal cells during their extensive migration. (D) The embryonic ectoderm cells remain as a cohesive cell sheet even after one day of culture. Marginal cells frequently extend very thin large lamella (arrows). Bars, 100μm.

Dispersed mesodermal cells were mostly bipolar or unipolar, although some multipolar cells were observed (Fig. 1C). Each cell migrated in the direction of an active lamellipodium. When lamellipodia of migrating cells made contact with each other, the lamellipodium stopped advancing immediately and retreated locally at the site of contact, thus exhibiting contact paralysis and contact inhibition of movement. The lamellipodium at the opposite end of the cell next became active and eventually the two cells moved away from each other. Mesodermal cells obtained from the primitive-streak-stage embryos and more advanced embryos (day 8–9) exhibited almost the same cell behaviour on the ECM-coated surface.

In the case of plastic tissue-culture dishes or glass coverslips, attachment of the mesodermal cell mass to the substratum required much more time than attachment to the ECM-coated dishes. Moreover, only a minority of the explants attached. Although attached cell masses became gradually flattened, few cells migrated away from the periphery of the cell masses. Most cells kept close contact with each other and remained as a cell sheet (Fig. 2). Mesodermal cells on plastic or glass surfaces took more circular and flattened shapes than on the ECM-coated surface. Each cell formed a large ruffling lamellipodium and moved slowly toward that direction. Nuclei showed frequent movement within the cytoplasm. This movement was more conspicuous in the cells on a glass surface than on a plastic surface and much less so on the ECM-coated surface. The cells showed contact paralysis on plastic or glass surfaces as well.

Fig. 2.

(A) A phase-contrast micrograph showing mesodermal explants cultured for one day on the surface of a plastic culture dish. (B) A higher magnification view. Bars, 100μm.

Fig. 2.

(A) A phase-contrast micrograph showing mesodermal explants cultured for one day on the surface of a plastic culture dish. (B) A higher magnification view. Bars, 100μm.

Explants of the primitive endoderm layer did not attach to any substrata. Those of the embryonic ectoderm attached to the ECM-coated surface and less frequently to the plastic or glass surfaces. The explants flattened to become a relatively thick epithelial cell sheet on the ECM-coated surface, but individual cells almost never moved away from the explant, even after one day of culture (Fig. 1D).

Cell motility

Table 1 shows the mean rate and standard deviation (s.D.) of movement by mesodermal cells cultured on the ECM-coated, plastic-culture-dish or glass surfaces. On the ECM-coated surface, mesodermal cells obtained from the primitive-streak-stage embryos moved at a mean rate of approx. 50 μm h . Mesodermal cells from the 8- or 9-day-old embryos were found to move at the same rate as younger cells. On the other hand, the mean rate on plastic or glass surfaces was uniformly lower (approx. 20–30μm h-1), regardless of the age of embryos from which mesodermal cells were isolated. The difference in the rate of movement between the ECM-coated surface and plastic or glass surfaces was found to be significant at the level of P = 0· 005 by Student’s t-test. However, there is no significant difference in the rate between the plastic and glass surface.

Table 1.

Motility (mean rate and s.D.) (μmh-1 ) of the mesodermal cells isolated from ICR (CD-1) mouse embryos on ECM, plastic or glass surfaces

Motility (mean rate and s.D.) (μmh-1 ) of the mesodermal cells isolated from ICR (CD-1) mouse embryos on ECM, plastic or glass surfaces
Motility (mean rate and s.D.) (μmh-1 ) of the mesodermal cells isolated from ICR (CD-1) mouse embryos on ECM, plastic or glass surfaces

Fig. 3 shows histograms illustrating distribution of the rates of cell movement on the ECM-coated or plastic surfaces. One remarkable feature is that the distribution on the ECM-coated surface is broader than that on the plastic surface. In most cases, approx. 10% of cells moved faster than 80μmh-1, and approx. 30 % of cells moved faster than 60μmh-1 on the ECM-coated surface. The distribution on the plastic surface is shifted to the lower rate range and there are no cells moving faster than 60μmh-1.

Fig. 3.

Histograms showing distribution of the rates of cell movement on the ECM-coated (A) or plastic (B) surfaces. Total cell number is 49 (A) and 47 (B). The mean values are shown by arrows.

Fig. 3.

Histograms showing distribution of the rates of cell movement on the ECM-coated (A) or plastic (B) surfaces. Total cell number is 49 (A) and 47 (B). The mean values are shown by arrows.

Fig. 3 shows histograms illustrating distribution of the rates of cell movement on the ECM-coated or plastic surfaces. One remarkable feature is that the distribution on the ECM-coated surface is broader than that on the plastic surface. In most cases, approx. 10% of cells moved faster than 80μmh-1, and approx. 30 % of cells moved faster than 60μmh-1 on the ECM-coated surface. The distribution on the plastic surface is shifted to the lower rate range and there are no cells moving faster than 60μmh-1.

Fig. 4 shows trajectories of the cell nuclei on the ECM-coated, plastic or glass surfaces during a 5h culture period. It is apparent that cell trails on the ECM-coated surface are longer than those on the plastic or glass surfaces. In most cases, they make apparently random zigzag lines. At the periphery of outgrowth, however, the direction of movement has a tendency to move radially away from the centre of explant. Cell trails on the glass or plastic surfaces are relatively circular due to movement of the nuclei within the cytoplasm. Nuclear movement within the cytoplasm occurred even when the whole cell did not translocate on plastic or glass surfaces. Therefore, the real rate of cell movement on these substrata seems to be lower than the calculated values shown in Table 1.

Fig. 4.

Trajectories of the cell nuclei on the ECM-coated (A), plastic (B) or glass (C) surfaces for 5 h traced from the time-lapse video recording. Cell trails on A are much longer than those in B or C. Bars, 100 μm.

Fig. 4.

Trajectories of the cell nuclei on the ECM-coated (A), plastic (B) or glass (C) surfaces for 5 h traced from the time-lapse video recording. Cell trails on A are much longer than those in B or C. Bars, 100 μm.

Cell division

Mesodermal cells on the ECM-coated surface showed frequent cell division accompanied by vigorous blebbing and rounding up of the cell body. Table 2 shows the frequency of cell division observed during 5 h of the video recording. The frequency is much lower on the plastic or glass surfaces than on the ECM-coated surface.

Table 2.

Frequency (per 100 cells) of cell division by the mesodermal cells during 5 h culture on ECM, plastic or glass surfaces

Frequency (per 100 cells) of cell division by the mesodermal cells during 5 h culture on ECM, plastic or glass surfaces
Frequency (per 100 cells) of cell division by the mesodermal cells during 5 h culture on ECM, plastic or glass surfaces

20 examples of cell division on the ECM-coated surface were analysed to obtain the time course. Vigorous blebbing started around the cell body as soon as the cell body became rounded from more angular shapes and the cell processes started to detach from the substratum. Cytokinesis generally started 20 to 30 min after rounding up of the cell body. Blebbing resumed immediately before the completion of cytokinesis. The two daughter cells moved rapidly away from each other and resumed normal cell shape, generally with only one active lamellipodium, by 30 to 60min from the first sign of rounding up and blebbing.

T mutant cells

No apparent differences were observed in cell morphology between explants taken from T/T and normal embryos (Fig. 5). The mesodermal cells showed extensive migration with a spindle or stellar shape on the surface of ECM-coated dishes. In contrast, on plastic or glass surfaces, the cells took more flattened shapes without extensive outgrowth from the explants.

Fig. 5.

Phase-contrast micrographs of the mesodermal cells on the ECM-coated dish surface, after they completed the outgrowth from small pieces of the mesodermal cell layer isolated from the primitive streak region. (A) Control cells; (B) T/T cells. Bars, 50μm.

Fig. 5.

Phase-contrast micrographs of the mesodermal cells on the ECM-coated dish surface, after they completed the outgrowth from small pieces of the mesodermal cell layer isolated from the primitive streak region. (A) Control cells; (B) T/T cells. Bars, 50μm.

Table 3 shows the mean and standard deviation of the rate of cell movement analysed from time-lapse video recording. There was no significant difference between T/T cells and control cells on plastic or glass substrata, and the mean rate was uniformly low (approx. 20μmh-1). On the ECM-coated surface, however, mesodermal cells from normal embryos moved faster (approx. 50μm h-1 on day 8, and approx. 40μm h-1 on day 9) than the cells from T/T embryos (approx. 40μmh-1 on day 8, and approx. 25 μm h-1 on day 9). Statistical analyses by Student’s t-test showed these differences to be significant at the level of P = 0·005.

Table 3.

Motility (mean rate and S.D.; μmh -1) of the mesodermal cells isolated from T/T and normal embryos on ECM, plastic or glass substrata

Motility (mean rate and S.D.; μmh -1) of the mesodermal cells isolated from T/T and normal embryos on ECM, plastic or glass substrata
Motility (mean rate and S.D.; μmh -1) of the mesodermal cells isolated from T/T and normal embryos on ECM, plastic or glass substrata

Fig. 6 shows histograms illustrating distribution of the rate of cell movement by the T/T and control mesodermal cells of day-9 embryos. The distribution is apparently shifted to the lower cell speed range in the T/T cell group. One remarkable feature is that approximately 10% of cells moved rapidly (faster than 80μmh-1) in the control group, while there were no such fast-moving cells among the T/T mutant cell population. Such fast-moving cells were always present among control cells, but never found in the T/T cells when distribution of five groups each of the T/T and control cells were examined.

Fig. 6.

Histograms showing distribution of the rate of cell movement analysed from time-lapse video recording. 50 mesodermal cells were analysed in cultures of explants from a normal (A) and T/T (B) embryos on day 9. Mean values are shown with arrows.

Fig. 6.

Histograms showing distribution of the rate of cell movement analysed from time-lapse video recording. 50 mesodermal cells were analysed in cultures of explants from a normal (A) and T/T (B) embryos on day 9. Mean values are shown with arrows.

Behaviour of mesodermal cells in vivo and in vitro

Motility of mesodermal cells inside embryos was examined by using whole embryo culture and DIC optics (Nakatsuji et al. 1986). It showed that the mesodermal cells migrated away from the primitive streak at an average rate of 46μmh-1 within threedimensional spaces in the early primitive-streak-stage embryos. They moved slowly in a crowded area, but more rapidly in areas of lower population density of mesodermal cells and with larger cell-free spaces (Nakatsuji et al. 1986). They took spindle, but not flattened, cell shapes (Spiegelman & Bennett, 1974; Spiegelman, 1976; Batten & Haar, 1979; Poelmann, 1981; Franke et al. 1982, 1983; Nakatsuji et al. 1986). In the present study, we calculated the mean rate of cell movement in vitro by mesodermal cells from the primitive-streak-stage embryos to be 52μmh-1 on the ECM-coated surface, whereas it was much lower on plastic or glass surfaces. Cell shape was bipolar or stellar on the ECM-coated surface, but it was more flattened on plastic or glass surfaces. These results reveal a resemblance between the mesodermal cells on the ECM-coated surface and those within an embryo during gastrulation. Moreover, the mesodermal cells exhibited contact inhibition of movement in the present study. It may contribute to the directional migration of the mesodermal cells away from the primitive streak and their slowdown in a crowded area observed in vivo (Nakatsuji et al. 1986).

Cell divisions accompanied by extensive blebbing were frequently observed while mesodermal cells were migrating away from the primitive streak within an embryo (Nakatsuji et al. 1986). We also observed cell division by the mesodermal cells accompanied by vigorous blebbing during outgrowth from an explant on the ECM-coated surface. Such cell division occurred very frequently among the mesodermal cells cultured on the ECM-coated surface, but much less frequently on plastic or glass surfaces. Snow (1977, 1978a,b) has shown that the mesodermal cells in the primitive-streak-stage embryos are rapidly proliferating (estimated cell cycle time, 14–22h). The time course of cell division observed here in vitro is also similar to that in vivo (Nakatsuji et al. 1986). These results show that the behaviour of mesodermal cells on the ECM-coated surface is much more similar to that inside embryos than on plastic or glass surfaces.

ECM as the substratum for cell migration

In the primitive-streak-stage mouse embryo, mesodermal cells migrate away from the primitive streak through an extracellular space between the embryonic ectoderm and primitive endoderm layers (Nakatsuji et al. 1986). Their cell processes attach to the outer (basal) surface of the ectoderm layer, the inner (basal) surface of the endoderm layer or adjacent mesodermal cells (Spiegelman & Bennett, 1974; Spiegelman, 1976; Batten & Haar, 1979; Poelmann, 1981;. Tam & Meier, 1982; Franke et al. 1982, 1983; Nakatsuji et al. 1986).

Transmission electron microscopic studies have revealed that the embryonic ectoderm layer has a basal lamina which is almost continuous except for the primitive-streak region (Batten & Haar, 1979; Franke et al. 1982, 1983). The primitive endoderm layer has no continuous basal lamina or lamina-like structure. However, fragments of fuzzy amorphous material (Franke et al. 1983; K. Hashimoto & N. Nakatsuji, unpublished observation), which can be stained with tannic acid (Herken & Barrach, 1985), were found near the basal surface of the primitive endoderm layer. Immunohistochemical studies of mouse embryos (Adamson & Ayers, 1979; Wartiovaara, Leivo & Vaheri, 1979; Leivo, Vaheri, Timpl & Wartiovaara, 1980; Leivo, 1983; Dziadek & Timpl, 1985; Herken & Barrach, 1985) have shown the presence of laminin, fibronectin, type IV collagen and heparan sulphate proteoglycan at the basal lamina of the embryonic ectoderm layer or near the basal surface of the primitive endoderm layer. Therefore, it is very likely that the migrating mesodermal cells use the ECM containing these molecules as the substratum for cell attachment and movement, in a manner similar to the mesodermal cells of amphibian gastrulae (Nakatsuji, Gould & Johnson, 1982; Boucaut & Darribère, 1983; Boucaut, Darribère, Boulekbache & Thiery, 1984; Nakatsuji, 1984; Nakatsuji, Smolira & Wylie, 1985; Darribère, 1983; Boucaut, Darribère, Boulekbache, Shi & Boucaut, 1985), chick gastrulae (Critchley, England, Wakely & Hynes, 1979) and sea urchin gastrulae (Katow, 1986).

ECM of the ECM-coated dish used in this study was produced by bovine corneal endothelial cells (personal communication from the distributor, Funakoshi Pharmaceutical Co. Ltd, Tokyo). Immunohistochemical analyses reveal that main components of such basement membrane ECM are type III and IV collagen, fibronectin and laminin (Gospodarowicz et al. 1979; Gospodarowicz & Tauber, 1980; Vlodavsky, Lui & Gospodarowicz, 1980). The present study showed that the behaviour of mesodermal cells on the ECM-coated surface is very similar to that in vivo. Therefore, the ECM seems to contain adequate components necessary for the attachment and movement by mesodermal cells, although it is produced by cells unrelated to those of the mouse embryo.

An in vitro system for analysing abnormal morphogenesis

Flint & Ede (1982) analysed the behaviour of mesodermal cells from the mouse amputated mutant using an in vitro system. Morris (1973, 1975) observed the effects of vitamin A on the behaviour in vitro of mesodermal cells isolated from the primitive-streakstage rat embryos. In these studies, however, they used glass or plastic surfaces as the culture substrata. Use of an ECM-coated substratum might give further insight into effects of the mutations or teratogens on the cell behaviour. In the present study, we showed that the mesodermal cells from the T/T mutant embryos had a reduced rate of motility on the ECM-coated surface, compared to the cells from normal embryos, but such a difference was not detectable in the culture conditions using plastic or glass substrata.

Studies on viability of the T/T embryos have shown that the metabolic activity and cell division continue beyond day 10 (Yanagisawa & Fujimoto, 1977a). There is no statistical difference in the average generation time of cells for normal and T/T embryos on day 8 (Yanagisawa, Fujimoto & Urushibara, 1981). Although the mitotic activity at the posterior end of the mutant embryo on day 9 is lower than that in the anterior region (Yanagisawa & Fujimoto, 1977a), those tissues have potentiality to develop into teratomas composed of fully differentiated tissues when grafted into ectopic sites (Fujimoto & Yanagisawa, 1979). Thus, the reduced rate of movement in the T/T cells could not be attributed to general degeneracy.

The retardation of cell movement can be caused by many factors. One of them is a change in the cytoskeletal system, which might be affected by the mutation as suggested in the case of t9 mutation (Spiegelman & Bennett, 1974). Embryos homozygous for the t9 mutation show defective primitive streak differentiation similar to that of the T/T embryo, but at the earlier stages. No reduction in the microfilaments or microtubules, however, has been detected in the T/T embryonic cells (Spiegelman, 1976).

Another likely factor is a change in the cell surface molecules related to cell adhesion. Shur (1982) showed that cell surface glycosyltransferase activities, which have been inferred to play a role in cell migration (Shur, 1977a,b), were different in normal and T/T mesenchyme cells. Yanagisawa & Fujimoto (1977b) reported that reaggregation kinetics of the dissociated embryonic cells was different in the T/T cells and normal cells, suggesting some changes in the cell surface properties. Similar differences are found between normal and talpid3 mesenchyme cells in the wing bud of the fowl (Ede & Agerbak, 1968; Ede & Flint, 1975a). The talpid3 mutation also affects cell motility in vitro (Ede & Flint, 1975b).

One interesting finding concerning the role of ECM is that in the T/T embryos it is greatly decreased in all areas when they are observed with scanning electron microscopy (Jacobs-Cohen, Spiegelman & Bennett, 1983). Altered matrix molecules might result in faulty organization of ECM and abnormal cell migration. However, the results presented here suggest that anomalous cell migration is caused by the intrinsic character of the T/T mutant cells.

We thank Izumi Fuketa and Masako Nagatomo for excellent technical assistance, and Misae Hirayama for looking after the mouse mutant stocks. Karyotyping was carried out by Hirofumi Suemori.

Adamson
,
E. D.
&
Ayers
,
S. E.
(
1979
).
The localization and synthesis of some collagen types in developing mouse embryos
.
Cell
16
,
953
965
.
Armstrong
,
P. B.
(
1985
).
The control of cell motility during embryogenesis
.
Cancer and Metastasis Rev
.
4
,
59
80
.
Batten
,
B. E.
&
Haar
,
J. L.
(
1979
).
Fine structural differentiation of germ layers in the mouse at the time of mesoderm formation
.
Anat. Rec
.
194
,
125
142
.
Beddington
,
R.
(
1983
).
The origin of the foetal tissues during gastrulation in the rodent
. In
Development in Mammals
, vol.
5
(ed.
M. H.
Johnson
), pp.
1
32
.
Amsterdam
:
Elsevier
.
Bellairs
,
R.
(
1986
).
The primitive streak
.
Anat. Embryol
.
174
,
1
14
.
Bellairs
,
R.
,
Curtis
,
A.
&
Dunn
,
G.
(
1982
).
Cell Behaviour
.
Cambridge
:
Cambridge University Press
.
Bennett
,
D.
(
1975
).
T-locus of the mouse
.
Cell
6
,
441
454
.
Boucaut
,
J.-C.
&
Darribère
,
T.
(
1983
).
Fibronectin in early amphibian embryos. Migrating mesenchymal cells contact fibronectin established prior to gastrulation
.
Cell Tissue Res
.
234
,
135
145
.
Boucaut
,
J.-C.
,
Darribère
,
T.
,
Boulekbache
,
H.
&
Thiery
,
J.-P.
(
1984
).
Prevention of gastrulation but not neurulation by antibodies to fibronectin in amphibian embryos
.
Nature, Lond
.
307
,
364
367
.
Chesley
,
P.
(
1935
).
Development of the short-tailed mutant in the house mouse
.
J. exp. Zool
.
70
,
429
459
.
Critchley
,
D. R.
,
England
,
M. S.
,
Wakely
,
J.
&
Hynes
,
R. O.
(
1979
).
Distribution of fibronectin in the ectoderm of gastrulating chick embryos
.
Nature, Lond
.
208
,
4989
5000
.
Dan
,
K.
&
Okazaki
,
K.
(
1956
).
Cyto-embryological studies of sea urchins. III. Role of the secondary mesenchyme cells in the formation of the primitive gut in sea urchin larvae
.
Biol. Bull. mar. biol. Lab., Woods Hole
110
,
29
42
.
Darribère
,
T.
,
Boulekbache
,
H.
,
Shi
,
D. L.
&
Boucaut
,
J.-C.
(
1985
).
Immuno-electron-microscopic study of fibronectin in gastrulating amphibian embryos
.
Cell Tissue Res
.
239
,
75
80
.
Deppe
,
W.
,
Schierenberg
,
E.
,
Cole
,
T.
,
Krieg
,
C.
,
Schmitt
,
D.
,
Yoder
,
B.
&
Von Ehrenstein
,
G.
(
1978
).
Cell lineages of the embryo of the nematode Caenorhabditis elegans. Proc. natn. Acad. Sci. U.S.A
.
75
,
376
380
.
Dziadek
,
M.
&
Timpl
,
R.
(
1985
).
Expression of nidogen and laminin in basement membranes during mouse embryogenesis and in teratocarcinoma cells
.
Devi Biol
.
111
,
372
382
.
Ede
,
D. A.
&
Agerbak
,
G. S.
(
1968
).
Cell adhesion and movement in relation to the developing limb pattern in normal and talpid3 mutant chick embryos
.
J. Embryol. exp. Morph
.
20
,
81
100
.
Ede
,
D. A.
&
Flint
,
O. P.
(
1975a
).
Intercellular adhesion and formation of aggregates in normal and talpid3 mutant chick limb mesenchyme
.
J. Cell Sci
.
18
,
97
111
.
Ede
,
D. A.
&
Flint
,
O. P.
(
1975b
).
Cell movement and adhesion in the developing chick wing bud: Studies on cultured mesenchyme cells from normal and talpid3 mutant embryos
.
J. Cell Sci
.
18
,
301
313
.
Eisen
,
J. S.
,
Myers
,
P. Z.
&
Westerfield
,
M.
(
1986
).
Pathway selection by growth cones of identified motoneurones in live zebra fish embryos
.
Nature, Land
.
320
,
269
271
.
England
,
M. A.
(
1983
).
The migration of primordial germ cells in avian embryos
. In
Current Problems in Germ Cell Differentiation
(ed.
A.
Mclaren
&
C. C.
Wylie
), pp.
91
114
.
Cambridge
:
Cambridge University Press
.
Flint
,
O. P.
&
Ede
,
D. A.
(
1982
).
Cell interactions in the developing somite: In vitro comparisons between amputated (am/am) and normal mouse embryos
.
J. Embryol. exp. Morph
.
67
,
113
125
.
Franke
,
W. W.
,
Grund
,
C.
,
Kuhn
,
C.
,
Jackson
,
B. W.
&
Illmensee
,
K.
(
1982
).
Formation of cytoskeletal elements during mouse embryogenesis. III. Primary mesenchymal cells and the first appearance of vimentin filaments
.
Differentiation
23
,
43
59
.
Franke
,
W. W.
,
Grund
,
C.
,
Jackson
,
B. W.
&
Illmensee
,
K.
(
1983
).
Formation of cytoskeletal elements during mouse embryogenesis. IV. Ultrastructure of primary mesenchymal cells and their cell-cell interactions
.
Differentiation
25
,
121
141
.
Fujimoto
,
H.
&
Yanagisawa
,
K. O.
(
1979
).
Effect of the T-mutation on histogenesis of the mouse embryo under the testis capsule
.
J. Embryol. exp. Morph
.
50
,
21
30
.
Fujimoto
,
H.
&
Yanagisawa
,
K. O.
(
1983
).
Defects in the archenteron of mouse embryos homozygous for the T-mutation
.
Differentiation
25
,
44
47
.
Gluecksohn-Schoenheimer
,
S.
(
1944
).
The development of normal and homozygous Brachyury (T/T) mouse embryos in the extraembryonic coelom of the chick
.
Proc. natn. Acad. Sci. U.S.A
.
30
,
134
140
.
Gluecksohn-Waelsch
,
S.
&
Erickson
,
R. P.
(
1970
).
The T-locus of the mouse: Implications for mechanisms of development
.
Curr. Top. devl Biol
.
5
,
281
316
.
Gospodarowicz
,
D.
,
Greenberg
,
G.
,
Vlodavsky
,
I.
,
Alvarado
,
J.
&
Johnson
,
L. K.
(
1979
).
The identification and localization of fibronectin in cultured corneal endothelial cells: Cell surface polarity and physiological implications
.
Expl Eye Res
.
29
,
485
509
.
Gospodarowicz
,
D.
&
Tauber
,
J.-P.
(
1980
).
Growth factors and the extracellular matrix
.
Endocrine Rev
.
1
,
201
227
.
Grüneberg
,
H.
(
1958
).
Genetical studies on the skeleton of the mouse. XXIII. The development of Brachyury and Anury
.
J. Embryol. exp. Morph
.
6
,
424
443
.
Gustafson
,
T.
&
Kinnander
,
H.
(
1956
).
Microaquaria for time-lapse cinematographic studies of morphogenesis in swimming larvae and observations on sea urchin gastrulation
.
Expl Cell Res
.
11
,
36
51
.
Heasman
,
J.
,
Hynes
,
R. O.
,
Swan
,
A. P.
,
Thomas
,
V. A.
&
Wylie
,
C. C.
(
1981
).
Primordial germ cells of Xenopus embryos: The role of fibronectin in their adhesion during migration
.
Cell
27
,
437
447
.
Heasman
,
J.
&
Wylie
,
C. C.
(
1983
).
Amphibian primordial germ cells - what can they tell us about directed cell migration?
In
Current Problems in Germ Cell Differentiation
(ed.
A.
Mclaren
&
C. C.
Wylie
), pp.
73
90
.
Cambridge
:
Cambridge University Press
.
Herken
,
R.
&
Barrach
,
H.-J.
(
1985
).
Ultrastructural localization of type IV collagen and laminin in the seven-day-old mouse embryo
.
Anat. Embryol
.
171
,
365
371
.
Jacobs-Cohen
,
R. J.
,
Spiegelman
,
M.
&
Bennett
,
D.
(
1983
).
Abnormalities of cell cells and extracellular matrix of T/T embryos
.
Differentiation
25
,
48
55
.
Katow
,
H.
(
1986
).
Behavior of sea urchin primary mesenchyme cells in artificial extracellular matrices
.
Expl Cell Res
.
162
,
401
410
.
Le Douarin
,
N. M.
(
1984
).
Cell migration in embryos
.
Cell
38
,
353
360
.
Leivo
,
I.
(
1983
).
Structure and composition of early basement membranes: Studies with early embryos and teratocarcinoma cells
.
Med. Biol
.
61
,
1
30
.
Leivo
,
L
,
Vaheri
,
A.
,
Timpl
,
R.
&
Wartiovaara
,
J.
(
1980
).
Appearance and distribution of collagens and laminin in the early mouse embryo
.
Devi Biol
.
76
,
100
114
.
Morris
,
G. M.
(
1973
).
The ultrastructural effects of excess maternal vitamin A on the primitive streak stage rat embryo
.
J. Embryol. exp. Morph
.
30
,
219
242
.
Morris
,
G. M.
(
1975
).
Abnormal cell migration as a possible factor in the genesis of vitamin A-induced craniofacial anomalies
. In
New Approaches to the Evaluation of Abnormal Mammalian Embryonic Development
(ed.
D.
Neubert
&
H. J.
Merker
), pp.
678
687
.
Stuttgart
:
Georg Thieme
.
Nakatsuji
,
N.
(
1984
).
Cell locomotion and contact guidance in amphibian gastrulation
.
Amer. Zool
.
24
,
615
627
.
Nakatsuii
,
N.
,
Gould
,
A. C.
&
Johnson
,
K. E.
(
1982
).
Movement and guidance of migrating mesodermal cells in Ambystoma maculatum gastrulae
.
J. Cell Sci
.
56
,
207
222
.
Nakatsuji
,
N.
Johnson
,
K. E.
(
1982
).
Cell locomotion in vitro by Xenopus laevis gastrula mesodermal cells
.
Cell Motil
.
2
,
149
161
.
Nakatsuji
,
N.
,
Smolira
,
M. A.
&
Wylie
,
C. C.
(
1985
).
Fibronectin visualized by scanning electron microscopy immunocytochemistry on the substratum for cell migration in Xenopus laevis gastrulae
.
Devi Biol
.
107
,
264
268
.
Nakatsuji
,
N.
,
Snow
,
M. H. L.
&
Wylie
,
C. C.
(
1986
).
Cinemicrographic study of the cell movement in the primitive-streak mouse embryo
.
J. Embryol. exp. Morph
.
96
,
99
109
.
Poelmann
,
R. E.
(
1981
).
The formation of the embryonic mesoderm in the early post-implantation mouse embryo
.
Anat. Embryol
.
162
,
29
40
.
Porter
,
R.
&
Fitzsimons
,
D. W.
(
1973
).
Locomotion of Tissue Cells. Ciba Found. Symp
.
14
, (New Series).
Amsterdam
:
Elsevier
.
Sherman
,
M. I.
&
Wudl
,
L. R.
(
1977
).
T-complex mutations and their effects
. In
Concepts in Mammalian Embryogenesis
(ed.
M. I.
Sherman
), pp.
136
234
.
Cambridge, Mass
.:
MIT Press
.
Shur
,
B. D.
(
1977a
).
Cell-surface glycosyltransferases in gastrulating chick embryos. I. Temporally spatially specific patterns of four endogenous glycosyltransferase activities
.
Devi Biol
.
58
,
23
39
.
Shur
,
B. D.
(
19776
).
Cell-surface glycosyltransferases in gastrulating chick embryos. II. Biochemical evidence for a surface localization of endogenous glycosyltransferase activities
.
Devi Biol
.
58
,
40
55
.
Shur
,
B. D.
(
1982
).
Cell surface glycosyltransferace activities during normal and mutant (T/T) mesenchyme migration
.
Devi Biol
.
91
,
149
162
.
Snow
,
M. H. L.
(
1977
).
Gastrulation in the mouse: Growth and regionalization of the epiblast
.
J. Embryol. exp. Morph
.
42
,
293
303
.
Snow
,
M. H. L.
(
1978a
).
Techniques for separating early embryonic tissues
. In
Methods in Mammalian Reproduction
(ed.
J. C.
Daniel
, Jr
), pp.
167
178
.
New York
:
Academic Press
.
Snow
,
M. H. L.
(
1978b
).
Proliferative centres in embryonic development
. In
Development in Mammals
, vol.
3
(ed.
M. H.
Johnson
), pp.
337
362
.
Amsterdam
:
Elsevier
.
Spiegelman
,
M.
(
1976
).
Electron microscopy of cell associations in T-locus mutants
. In
Embryogenesis in Mammals
(ed.
K.
Elliott
&
M.
O’connor
), Ciba Found. Symp. 40 (New Series), pp.
199
220
.
Amsterdam
:
Elsevier
.
Spiegelman
,
M.
&
Bennett
,
D.
(
1974
).
Fine structural study of cell migration in the early mesoderm of normal and mutant mouse embryos. (T-locus: t9/t9)
.
J. Embryol. exp. Morph
.
32
,
723
728
.
Tam
,
P. P. L.
&
Meier
,
S.
(
1982
).
The establishment of a somitomeric pattern in the mesoderm of the gastrulating mouse embryo
.
Am. J. Anat
.
164
,
209
225
.
Thiery
,
J.-P.
,
Duband
,
J. L.
&
Tucker
,
G. C.
(
1985a
).
Cell migration in the vertebrate embryo: Role of cell adhesion and tissue environment in pattern formation
.
A. Rev. Cell Biol
.
1
,
91
113
.
Thiery
,
J.-P.
,
Boucaut
,
J.-C.
&
Yamada
,
K. M.
(
19856
).
Cell migration in the vertebrate embryo
. In
Molecular Determinants of Animal Form
(ed.
G. M.
Edelman
), pp.
167
193
.
New York
:
Alan R. Liss, Inc
.
Trinkaus
,
J. P.
(
1973
).
Surface activity and locomotion of Fundulus deep cells during blastula and gastrula stages
.
Devi Biol
.
30
,
68
103
.
Trinkaus
,
J. P.
(
1976
).
On the mechanism of metazoan cell movements
. In
The Cell Surface in Animal Embryogenesis and Development
(ed.
G.
Poste
&
G. L.
Nicolson
), pp.
225
329
.
Amsterdam
:
Elsevier
.
Trinkaus
,
J. P.
(
1984
).
Cells into Organs. The Forces that Shape the Embryo. 2nd edn. Englewood Cliffs: Prentice-Hall
.
Trinkaus
,
J. P.
(
1985
).
Protrusive activity of the cell surface and the initiation of cell movement during morphogenesis
.
Expl Biol. Med
.
10
,
130
173
.
Vlodavsky
,
I.
,
Lui
,
G. M.
&
Gospodarowicz
,
D.
(
1980
).
Morphological appearance, growth behavior and migratory activity of human tumor cells maintained on extracellular matrix versus plastic
.
Cell
19
,
607
616
.
Wartiovaara
,
I.
,
Leivo
,
I.
&
Vaheri
,
A.
(
1979
).
Expression of the cell surface-associated glycoprotein, fibronectin, in the early mouse embryo
.
Devi Biol
.
69
,
247
257
.
Whittingham
,
D. G.
&
Wales
,
R. G.
(
1969
).
Storage of two-cell mouse embryos in vitro
.
Austr. J. biol. Sci
.
22
,
1065
1068
.
Yanagisawa
,
K. O.
&
Fujimoto
,
H.
(
1977a
).
Viability and metabolic activity of homozygous Brachyury (T) embryos
.
J. Embryol. exp. Morph
.
40
,
271
276
.
Yanagisawa
,
K. O.
&
Fujimoto
,
H.
(
1977b
).
Differences in rotation-mediated aggregation between wild-type and homozygous Brachyury (7) cells
.
J. Embryol. exp. Morph
.
40
,
277
283
.
Yanagisawa
,
K. O.
,
Fujimoto
,
H.
&
Urushibara
,
H.
(
1981
).
Effects of the Brachyury (7) mutation on morphogenetic movement in the mouse embryo
.
Devi Biol
.
87
,
242
248
.