Cell –cell communication is not only a common strategy for cell fate specification in vertebrates, but plays important roles in invertebrate development as well. We report here on experiments testing the compatibility of mechanisms specifying cell fate among six different Drosophila species. Following interspecific transplantation, the development of single ectodermal cells was traced in order to test their abilities to proliferate and differentiate in a heterologous environment. Despite considerable differences in cell size and length of cell cycle among some of the species, the transplants gave rise to fully differentiated clones that were integrated into the host tissue. Clones comprised cells of epidermal and/or neural histotypes, indicating that mechanisms mediating the epidermal/neural dichotomy in the ectoderm are conserved between the species. Cells of the neural lineages differentiated into neurones, glia, or both. Moreover, heterologous neurones sent out axons that followed major pathways along nerves and within the neuropile, demonstrating their ability to recognize positional cues in the heterologous CNS of the host.

The relative importance of cell lineage and cell communication as strategies for cell determination differs among organisms and cell types. In insect development, cellular interactions have been frequently shown to play an essential role in the specialization of cell fate (Hafen et al. 1987; Campos-Ortega, 1988; Hooper and Scott, 1989).

The first step towards neurogenesis in insects is the delamination of a distinct population of neuroblasts from the neurogenic ectoderm (Wheeler, 1893; Bate, 1976; Hartenstein and Campos-Ortega, 1984; Doe and Goodman, 1985a). For neuroectodermal cells, cell-cell interactions mediate the choice between the neurogenic and epidermogenic developmental pathways (Doe and Goodman 1985b; Technau and Campos-Ortega, 1986; Campos-Ortega, 1988). However, these cells are capable of switching their fates for an extended period of time during embryogenesis (Technau et al. 1988). In the second major phase of neurogenesis, most of the delaminated neuroblasts divide asymmetrically to generate a characteristic set of ganglion mother cells, each of which divides once more to give rise to a pair of sibling cells (Poulson, 1950; Goodman and Spitzer, 1979; Hartenstein et al. 1987). Cell ablations in the grasshopper have shown that ganglion mother cells are determined by their lineage, whereas the determination of daughter neurones involves interactions with their sibling (Doe and Goodman, 1985b; Kuwada and Goodman, 1985). Neural progeny attain their final form by growing axonal and dendritic projections, which depends at least in part on recognition by the growth cone of positional cues either on cell surfaces or in the extracellular matrix. Cell interactions may lead to the differential expression of these recognition molecules on neural surfaces (Bastiani et al. 1985, 1987). Cell communication therefore plays an important role for many of the major developmental events that guide an undifferentiated neuroectodermal progenitor to generate functional neural progeny.

Interspecific chimeras have been extensively used to study the functional compatibility of mechanisms that involve interactions among certain components in the developing organisms (e.g. Gardner, 1968; Le Douarin et al. 1975; Santamaria, 1975; Tanaka and Landmesser, 1986). These studies present information about how far these mechanisms are conserved among species. We tested in vivo the compatibility of mechanisms underlying determinative events during neurogenesis using single cell transplantations among six different Drosophila species. The results presented here show that the mechanisms leading to (i) neural/epidermal dichotomy in the ectoderm, (ii) neuronal/glial determination of neural progeny and (iii) pathfinding of outgrowing neurites are compatible among these species.

Stocks and culture conditions

The following Drosophila species were selected for the experiments:

Except for D. melanogaster and D. hydei, the species came originally from the (meanwhile discontinued) Drosophila Species Center in the University of Texas at Austin; since 1982 they were maintained in the laboratory of K. Sander (University of Freiburg, FRG), from whom we obtained the strains.

All strains were maintained on standard Drosophila culture medium supplemented with yeast. D. busckii, D. pseudoobscura and D. latifasciaeformis require a moist atmosphere; therefore, their vials were kept in a humid chamber.

Eggs were collected on agar plates supplemented with yeast. For D. hydei, D. busckii and D. latifasciaeformis stimulation of egg deposition was best when agar plates were covered with moistened filter paper. D. latifasciaeformis preferred 1 –2 drops of a dry white wine on the filter paper. Agar plates were changed at one hour intervals.

Measurement of developmental periods

To determine the time until complete cellularization of embryos (stage 7, all embryonic stages refer to those of Campos-Ortega and Hartenstein, 1985) about 60 eggs from a 30-minute collection were mechanically dechorionated, covered with Voltalef oil and inspected with a microscope at 25 °C. The periods for individuals to reach stage 7 were averaged. After completion of embryogenesis, hatching larvae were transferred to culture vials. Embryonic, larval and pupal developmental periods were then averaged for each species.

Cell measurements

Cells of stage 7 embryos were dissociated in a drop of Schneider’s Drosophila medium. Diameters of 50 cells were obtained by viewing with micrometer oculars and were averaged for each species.

Cell labelling and transplantations

Cell labelling and transplantations were carried out according to Technau (1986). Embryos were mechanically dechorionated and mounted on a coverslip in appropriate orientation. After sufficient desiccation, the embryos were covered with 10S Voltalef oil. Donors at the syncytial blastoderm stage were injected with a 3:1 mixture of 10% HRP and 10% FITC – dextran. Transplantations were carried out immediately after cell closure when donors and hosts had reached stage 7. Tips of transplantation capillaries were ground to a 30 – 40° bevel. The inner diameter was adjusted to the cell size of the donor. Single cells were removed and implanted within segment anlagen T3/A1, which map at about 50% egg length (%EL) of the D. melanogaster fate map (Technau and Campos-Ortega, 1985; Hartenstein et al. 1985). Transplantations were either homotopic, from/to the ventral neurogenic region, between 0% (ventral midline) and 30% of the ventrodorsal dimension (%VD), or heterotopic from the dorsal epidermal anlage (80 –100% VD) into the ventral neurogenic ectoderm (0 –30% VD) of unlabelled recipients (Fig. 1). Transplantation of single cells was visualized with a fluorescence microscope. Hosts were allowed to develop to embryonic stage 17 or 2nd/3rd instar larval stages.

Fig. 1.

Fate map of the early gastrula stage of Drosophila melanogaster (according to Technau and Campos-Ortega, 1985; Hartenstein et al. 1985). All transplantations were performed when donors and hosts were at this stage immediately after completion of cell formation. Stippled area designates the neurogenic ectoderm. Transplantations of cells were attempted within segment anlagen T3/A1 which map around 50% egg length (%EL). Transplantations were either homotopic (1) from/to the ventral neurogenic region (vNR) between 0% (ventral midline) and 30% of the ventrodorsal dimension (%VD) or heterotopic (2) from the dorsal epidermal anlage (dEpi; 80 – 100 %VD) into the vNR (0 – 30 %VD). Cl – 3, gnathal segments; Tl – 3, thoracic segments; Al-10, abdominal segments.

Fig. 1.

Fate map of the early gastrula stage of Drosophila melanogaster (according to Technau and Campos-Ortega, 1985; Hartenstein et al. 1985). All transplantations were performed when donors and hosts were at this stage immediately after completion of cell formation. Stippled area designates the neurogenic ectoderm. Transplantations of cells were attempted within segment anlagen T3/A1 which map around 50% egg length (%EL). Transplantations were either homotopic (1) from/to the ventral neurogenic region (vNR) between 0% (ventral midline) and 30% of the ventrodorsal dimension (%VD) or heterotopic (2) from the dorsal epidermal anlage (dEpi; 80 – 100 %VD) into the vNR (0 – 30 %VD). Cl – 3, gnathal segments; Tl – 3, thoracic segments; Al-10, abdominal segments.

Preparation and staining of donors

Stage 17 embryos were prefixed according to Zalokar and Erk (1977). The vitelline membrane was mechanically removed and embryos were fixed in 5 % glutaraldehyde in phosphate buffer (PB). Larvae were dissected and central nervous systems fixed in 5 % glutaraldehyde in PB. Following staining for HRP, the preparations were embedded in Epon, either on slides or in glass capillaries of 0.2 mm inner diameter (hemispheres were removed from larval CNSs). The embedding in capillaries allows rotation of the preparation under the compound microscope (E. R. Macagno, personal communication).

For electron microscopy, larval CNSs were postfixed in 2 % OsO4 in PB. Regions with HRP-labelled cells were cut into light golden sections and stained with lead citrate (Reynolds, 1963).

Labelled clones in whole-mount preparations were photographed using Nomarski optics, or were drawn with the aid of a camera lucida (Zeiss). Transmission electron micrographs were obtained with a Siemens Elmiskop 101.

Comparisons of the general course of development among the selected species

In all experiments, cell transplantations were performed when donors and hosts were at the same developmental stage, shortly after cell closure at the beginning of gastrulation. Although developmental periods to this stage are almost invariant within a given species, they vary considerably between species (Fig. 2). This is also true for the length of the total developmental period. These periods are not correlated with the body sizes, however. For example, development is longest for the small species D. busckii and D. latifasciaeformis. The relative contributions of the embryonic, larval and pupal phases to total development also vary between species (Fig. 2); in D. melanogaster embryogenesis requires 9.5 % of the total time for development, whereas in D. busckii it only requires 5.6%. It is likely that the differences in developmental periods correspond to the cell cycles of the respective species. This is in agreement with the fact that cell sizes (Table 1) and body sizes vary proportionally among the species.

Table 1.

Diameters of dissociated cells from the ventral ectoderm of Stage 7 embryos

Diameters of dissociated cells from the ventral ectoderm of Stage 7 embryos
Diameters of dissociated cells from the ventral ectoderm of Stage 7 embryos
Fig. 2.

Developmental times of the six Drosophila species selected for the experiments. Bars on the left side represent the times until completion of cell formation (in minutes) at the beginning of gastrulation (Stage 7) when transplantations were performed. Bars on the right side represent the total developmental times (in hours) including embryonic and postembryonic (larval and pupal) stages.

Fig. 2.

Developmental times of the six Drosophila species selected for the experiments. Bars on the left side represent the times until completion of cell formation (in minutes) at the beginning of gastrulation (Stage 7) when transplantations were performed. Bars on the right side represent the total developmental times (in hours) including embryonic and postembryonic (larval and pupal) stages.

Spatial distribution of clones

In all experiments, we transplanted cells to a region within a transverse ring at 50 % EL that corresponds to segment anlagen T3/A1 of the D. melanogaster fate map (Hartenstein et al. 1985; Fig. 1). There was a close correlation between the location of the implants and the final position of their differentiated clones. 74% of all resulting cell clones were located either in T3 or in Al, 23 % were in T2 or A2 and 3 % outside these regions. This suggests that, with respect to segment boundaries, the cells stay within the region of implantation. In addition, we found that the D. melanogaster fate map seems to apply to the other Drosophila species as well.

Clones derived from cells of the ventral neurogenic ectoderm following homotopic transplantation

The ventral epidermis and the ventral nervous system derive from the neurogenic ectoderm (Poulson, 1950). In accordance with this, in D. melanogaster homotopically transplanted single cells from the neurogenic ectoderm give rise to epidermal or neural lineages. The existence of mixed clones consisting of epidermal and neural subclones indicates that divergence into either lineage may occur even after the transplanted cell has undergone an additional mitosis (Technau and Campos-Ortega, 1986). We also performed homotopic intraspecific transplantations of neurogenic ectodermal cells of D. virilis. The cells behave like their homologues in D. melanogaster in that they give rise to clones with the same histotypes (Table 2).

Table 2.

Frequency of clonal histotypes (number of cases) derived from cells of the ventral neurogenic ectoderm following homotopic transplantation

Frequency of clonal histotypes (number of cases) derived from cells of the ventral neurogenic ectoderm following homotopic transplantation
Frequency of clonal histotypes (number of cases) derived from cells of the ventral neurogenic ectoderm following homotopic transplantation

In order to test the developmental potentials of individual precursor cells in a heterologous environment, we performed a series of transplantations of neuroectodermal cells from D. melanogaster into five other Drosophila species (D. virilis, D. busckii, D. hydei, D. latifasciaeformis, D. pseudoobscura) as well as from these species into D. melanogaster. The species chosen for study are members of four different subgenera (see Materials and methods). In all experimental combinations, transplants gave rise to neural as well as epidermal clones, and mixed clones developed in most of the experiments (Table 2). Therefore, in all chimeric combinations tested, single precursor cells of the neurogenic ectoderm were able to follow the neurogenic and epidermogenic developmental pathways despite being surrounded by cells of other Drosophila species. However, the general success rate of transplanted cells in producing differentiated clones varies considerably among the experiments. Whereas in the intraspecific transplantations (mel →mel, vir →vir; see Table 2), the ratio of numbers of clones obtained to the total number of transplantations was more than 25%, it varied in interspecific transplantations between only 3 % (bus → mel) and 17% (vir →mel).

Clones derived from cells of the dorsal epidermal anlage following heterotopic transplantation

The homotopic transplants demonstrated that a precursor cell from the neurogenic ectoderm retains the ability to develop in a heterologous environment. However, the decision to acquire one of the two alternative fates could be imposed upon the cell prior to removal from the donor. Alternatively, it could result from interactions with the heterologous host cells.

To address this issue, we removed cells from the dorsal ectoderm of D. melanogaster that, upon homotopic transplantation, only give rise to epidermal clones (Technau and Campos-Ortega, 1986), and implanted them into the ventral neurogenic ectoderm of each of the other five species. We also carried out the reverse experiments, transplantations from each of the five species into D. melanogaster. Under all 10 chimeric conditions, as well as in the two intraspecific control experiments (mel →mel, vir →vir; Table 3) the hetero-topically transplanted donor cells gave rise to epidermal and neural clones. Mixed epidermal/neural clones developed in most of the experimental series. This suggests that the transplanted cells communicate with their heterologous neighbours and that the mechanisms leading to the epidermal/neural dichotomy of ectodermal cells must be similar among different Drosophila species.

Table 3.

Frequency of clonal histotypes (number of cases) derived from cells of the dorsal epidermal anlage following heterotopic transplantation into the ventral neurogenic region. (For abbreviations see Table 2)

Frequency of clonal histotypes (number of cases) derived from cells of the dorsal epidermal anlage following heterotopic transplantation into the ventral neurogenic region. (For abbreviations see Table 2)
Frequency of clonal histotypes (number of cases) derived from cells of the dorsal epidermal anlage following heterotopic transplantation into the ventral neurogenic region. (For abbreviations see Table 2)

Table 3 shows the distribution of clonal histotypes. The ratios between neural and epidermal clones varied considerably from experiment to experiment. This variability roughly parallels that between the homotopic transplantation experiments (Table 2), despite the fact that, in general, the relative abundance of neural clones was lower following heterotopic transplantations. For example, in both experimental series, the proportion of neural clones was high following transplantations vir →vir, vir →mel and lat →mel, as compared with transplantation mel →mel in which it was relatively low. Again, as with the series of homotopic transplantations, the success rate of transplanted ceils to produce differentiated clones was much higher following intraspecific transplantations (vir →vir 40% and mel →mel 23%) than for interspecific transplantations (varying between 4 % for mel →lat and 19 % in experiment mel →vir). It therefore seems that while the cell – cell interactions that determine cell fate are similar for different species, they are not identical. Alternatively, more general factors required for cell viability and proliferation might impede successful integration of heterologous cells.

Distribution of clone sizes

Clone sizes (referring to the number of progeny) reflect the proliferative activities of the transplanted progenitor cell. We asked whether the proliferation of a cell is affected by transplantation to a region containing cells of different size (see Table 1) and cell cycles (as judged from the respective developmental periods; Fig. 2). Tables 4 and 5 present the size distributions of neural and epidermal clones following homotopic (a) and heterotopic (b) transplantations.

Table 4.

Size distribution of neuronal clones following homotopic transplantation of cells from the ventral neurogenic region (a) and heterotopic transplantation of cells from the dorsal epidermal anlage into the ventral neurogenic region (b). For abbreviations of donor and host species see Table 2. Above each column donor is first, host second

Size distribution of neuronal clones following homotopic transplantation of cells from the ventral neurogenic region (a) and heterotopic transplantation of cells from the dorsal epidermal anlage into the ventral neurogenic region (b). For abbreviations of donor and host species see Table 2. Above each column donor is first, host second
Size distribution of neuronal clones following homotopic transplantation of cells from the ventral neurogenic region (a) and heterotopic transplantation of cells from the dorsal epidermal anlage into the ventral neurogenic region (b). For abbreviations of donor and host species see Table 2. Above each column donor is first, host second
Table 5.

Size distribution of epidermal clones following homotopic transplantation of cells from the ventral neurogenic region (a) and heterotopic transplantation of cells from the dorsal epidermal anlage into the ventral neurogenic region (b). For abbreviations of donor and host species see Table 2. Above each column donor is first, host second.

Size distribution of epidermal clones following homotopic transplantation of cells from the ventral neurogenic region (a) and heterotopic transplantation of cells from the dorsal epidermal anlage into the ventral neurogenic region (b). For abbreviations of donor and host species see Table 2. Above each column donor is first, host second.
Size distribution of epidermal clones following homotopic transplantation of cells from the ventral neurogenic region (a) and heterotopic transplantation of cells from the dorsal epidermal anlage into the ventral neurogenic region (b). For abbreviations of donor and host species see Table 2. Above each column donor is first, host second.

In general, the sizes of neural clones varied between 1 and 20 cells, the majority of clones containing less than 10 cells. The distribution of clone sizes was comparable for intra- and interspecific transplantations, thus, proliferative capabilities of neural progenitors did not seem to be affected by the heterologous environment. We have not yet tested whether the temporal pattern of mitosis of the donor cells is autonomous or whether it becomes synchronized with that of the surrounding host cells.

Following intraspecific transplantations, the sizes of epidermal clones varied between 2 and 8 cells (mel → mel, vir →vir; Table 5a,b) corresponding to a maximum of 3 symmetric divisions of epidermoblasts. This was true for homotopic (ventral →ventral; Table 5a) as well as for heterotopic (dorsal →ventral) transplantations (Table 5b; for D. melanogaster see also Technau and Campos-Ortega, 1986). Following interspecific transplantations, however, clones comprising more than 8 cells (up to 15 cells; Fig. 3A) were also found, most of which originated from heterotopically transplanted epidermoblasts (10 out of 11 cases). Interestingly, clones with more than 8 cells were only found when cells were transplanted into D. melanogaster (vir →mel, hyd → mel, bus →mel and lat →mel) and not in the reciprocal experiments (mel →vir, mel →hyd, mel →bus and mel-→lat). In several cases of interspecific (primarily heterotopic) transplantations, the epidermoblasts failed to divide but nonetheless became integrated and differentiated as single cells in the epidermis of the host. These data suggest that, under normal conditions, cell interactions restrict the sizes of epidermal clones to a maximum of 8 cells and that the functions controlling these interactions are not fully compatible between D. melanogaster and the species D. virilis, D. hydei, D. busckii and D. latifasciaeformis.

Fig. 3.

(A) Epidermal clone consisting of 15 cells derived from a dorsal ectodermal cell of D. hydei, which was heterotopically transplanted into the ventral neurogenic region of a D. melanogaster host. The clone does not respect the compartment border. Arrows point to denticle belts. (B) Ventrolateral clone of tracheal cells obtained following heterotopic dorsal-to-ventral transplantation of a progenitor cell from D. hydei into D. melanogaster. Bar, (A,B) 30 μm.

Fig. 3.

(A) Epidermal clone consisting of 15 cells derived from a dorsal ectodermal cell of D. hydei, which was heterotopically transplanted into the ventral neurogenic region of a D. melanogaster host. The clone does not respect the compartment border. Arrows point to denticle belts. (B) Ventrolateral clone of tracheal cells obtained following heterotopic dorsal-to-ventral transplantation of a progenitor cell from D. hydei into D. melanogaster. Bar, (A,B) 30 μm.

Differentiation of neural cell types

(A) Neuronal versus glial cell types

Neurons and glia, due to their distinct cell morphologies and different spatial distributions in the CNS, are easily distinguished in our preparations (Figs 4,5). Irrespective of whether cells were transplanted intra- or interspecifically, we found principally three types of neural clones in the ventral nervous system (Tables 2,3): clones exclusively containing neuronal cells (N), mixed clones containing neuronal as well as glial cells (NG) and clones exclusively composed of glial cells (G). Therefore, among the cells delaminating from the ventral neurogenic ectoderm, there are neuroblasts, glioblasts and common precursors for both cell types. In contrast to neuroblasts, which can produce up to 20 daughter cells (Table 4), glioblasts produced a maximum of four progeny (data not shown). The distribution of the three types of clones resulting from homotopic transplantations is summarized in Table 2. Glial cells were also found among clones derived from cells of the dorsal ectoderm upon heterotopic transplantation into the ventral neurogenic region (Table 3).

Fig. 4.

Examples of neuronal clones at late embryonic stages demonstrating that growing fibers follow well-defined pathways. (A) Camera-lucida drawing of a clone comprising motoneurones (ventrolateral view). The four cell bodies are close to the ventral midline (dashed-dotted line). Their fibers project dorsally into the neuropile where they send branches anteriorly along the connective on each side. Axons are leaving the CNS (broken line) through segmental nerves of T2 and T3 on each side to project to somatic muscles. Solid line outlines the body wall. The progenitor cell was taken from the ventral neurogenic ectoderm of D. melanogaster and homotopically transplanted into D. busckii. (B) Clone of two cells located midventrally in the third thoracic neuromere. The clone was obtained following intraspecific, homotopic transplantation in D. virilis. (C,D) Lateral (C) and ventral (D) view of a clone consisting of the same type of cells as in B. In this case the clone is derived from a dorsal ectodermal precursor cell of D. latifasciaeformis that was heterotopically transplanted into the ventral neurogenic ectoderm of D. melanogaster. In both cases (B and C,D) the two cells send a fiber dorsally into the ipsilateral neuropile where it bifurcates sending one branch anteriorly and the other posteriorly along the connective. Bar, (A) 30 μm; (C –D) 10 μm.

Fig. 4.

Examples of neuronal clones at late embryonic stages demonstrating that growing fibers follow well-defined pathways. (A) Camera-lucida drawing of a clone comprising motoneurones (ventrolateral view). The four cell bodies are close to the ventral midline (dashed-dotted line). Their fibers project dorsally into the neuropile where they send branches anteriorly along the connective on each side. Axons are leaving the CNS (broken line) through segmental nerves of T2 and T3 on each side to project to somatic muscles. Solid line outlines the body wall. The progenitor cell was taken from the ventral neurogenic ectoderm of D. melanogaster and homotopically transplanted into D. busckii. (B) Clone of two cells located midventrally in the third thoracic neuromere. The clone was obtained following intraspecific, homotopic transplantation in D. virilis. (C,D) Lateral (C) and ventral (D) view of a clone consisting of the same type of cells as in B. In this case the clone is derived from a dorsal ectodermal precursor cell of D. latifasciaeformis that was heterotopically transplanted into the ventral neurogenic ectoderm of D. melanogaster. In both cases (B and C,D) the two cells send a fiber dorsally into the ipsilateral neuropile where it bifurcates sending one branch anteriorly and the other posteriorly along the connective. Bar, (A) 30 μm; (C –D) 10 μm.

Fig. 5.

Neuronal (A,C) and glial elements (B,D) derived from neurogenic ectodermal cells of D. virilis following homotopic transplantation into D. melanogaster. (A) Whole-mount preparation of a late embryo with a heterologous neuronal fiber following a longitudinal tract in the ventral nerve cord. The fiber carries specializations (arrowheads). Whole mount in B shows a heterologous glial cell located dorsomedially in the thoracic ventral nerve cord of a late embryo. (C,D) Electron micrographs demonstrating integration of heterologous neural elements in the ventral nerve cord of a late third instar larva. Neuronal fiber in C shows a longitudinal section through a specialization or a small branch of a fiber fasciculating with fibers in the host neuropile. (D) Donor-derived glial elements (closed arrows) intermingle with corresponding host elements (open arrows) at the dorsal neuropile border. Bar, (A) 10 μm; (B) 30 μm; (C,D) 2 μm.

Fig. 5.

Neuronal (A,C) and glial elements (B,D) derived from neurogenic ectodermal cells of D. virilis following homotopic transplantation into D. melanogaster. (A) Whole-mount preparation of a late embryo with a heterologous neuronal fiber following a longitudinal tract in the ventral nerve cord. The fiber carries specializations (arrowheads). Whole mount in B shows a heterologous glial cell located dorsomedially in the thoracic ventral nerve cord of a late embryo. (C,D) Electron micrographs demonstrating integration of heterologous neural elements in the ventral nerve cord of a late third instar larva. Neuronal fiber in C shows a longitudinal section through a specialization or a small branch of a fiber fasciculating with fibers in the host neuropile. (D) Donor-derived glial elements (closed arrows) intermingle with corresponding host elements (open arrows) at the dorsal neuropile border. Bar, (A) 10 μm; (B) 30 μm; (C,D) 2 μm.

(B) Axonal pathfindinq

Further differentiation of neuronal progeny cells results in hundreds of cell types which can be individually characterized by their specific axonal projections and dendritic arborizations. Development of these typical projection patterns involves, among other things, selective fasciculation of outgrowing fibers along preexisting pathways (e.g. Bastiani et al. 1985). We asked wether the underlying mechanisms of cell recognition might function between heterologous neural cells. Locations and projection patterns of obtained clones were compared to those of known cell types from the intraspecific transplantations.

Although neural clones developed with different success rates, in all 10 chimeric situations tested, neuronal progeny of donor cells developed processes that were able to fasciculate with heterologous fiber tracts within connectives, commissures and segmental nerves. In several cases, cells developed into motoneurones with characteristic dendritic arborizations within the neuropile and their axons leaving the CNS through segmental nerves and projecting to somatic muscles (Fig. 4A). Fig. 4B and 4C,D show two clones probably representing the same cell types. Each clone consisted of two midventrally located neurones. Their fibers projected dorsally into the neuropile (Fig. 4C), where they bifurcated, sending one branch anteriorly and the other posteriorly along the connective (Fig. 4B,D). The clone in Fig. 4B is derived from a cell of the ventral neurogenic ectoderm of D. virilis following intraspecific and homotopic transplantation. Clone in Fig. 4C,D is derived from a progenitor cell transplanted from the dorsal epidermal anlage of D. latifasciaeformis into the ventral neurogenic ectoderm of D. melanogaster. Thus, mechanisms of cell recognition in the process of selective fasciculation appear to be compatible among the species tested. This is true irrespective of whether the progenitor cell stems from the dorsal or ventral ectoderm of the donor.

(C) Histotypical and functional integration

During larval growth, the fly nervous system increases considerably in volume. Larval neuronal clones stemming from a single transplanted embryonic progenitor extend their dendritic arborizations during this period (Prokop and Technau, unpublished results). We found that this was also true for the differentiating interspecifically transplanted neural precursors, by transplanting ectodermal cells from D. virilis into D. melanogaster and allowing them to develop up to the third instar larval stage. The CNSs were then removed and stained for HRP.

At the light-microscopic level, neuronal fibers showed spine-like specializations within the neuropile (Fig. 5A). At the ultrastructural level, cell bodies of D. virlis-derived neurones were seen to be well integrated into the cortex of the D. melanogaster host. Labelled donor axons were in close appositions to their heterologous neighbours (Fig. 5C). The spine-like structures often contained mitochondria. Glial cells intermingled with glia of the host and acquired typical positions in the CNS, e.g. medially (Fig. 5B) or at the periphery of the neuropile (Fig. 5D). Unfortunately, unsatisfactory tissue preservation, possibly due to the HRP-staining reaction, did not allow us to identify synapses without ambiguity. However, we are fairly confident that these indirect cues suggest that these cells are able to establish functional contacts.

Other structures formed by interspecifically transplanted cells

Several somatic muscles and, in some cases, visceral muscles were found following homotopic intra- and interspecific transplantation of ventral cells (Table 2). Since the neurogenic ectoderm is close to the invaginated mesoderm, we believe that muscle clones are derived from mesodermal cells which were inadvertantly removed from the donors (as discussed in Technau et al. 1988). This is in accordance with the fact that these clones never occurred upon heterotopic transplantations of dorsal ectodermal cells into the ventral ectoderm (Table 3). Sizes and location of these chimeric muscle clones seemed to be normal and, in all cases, became attached to the epidermal apodemes of the host.

In some chimeras, the transplanted cell gave rise to sensory organs. In addition, two clones contributing to the tracheal system developed following heterotopic dorsal to ventral transplantation (Fig. 3B).

We have tested the compatibility of functions controlling successive determinative events during larval CNS development in six different Drosophila species belonging to four different subgenera. We transplanted single ectodermal progenitors, to exclude inductive influences from homologous cells. We found that transplanted progenitors gave rise to clones of differentiated cells that became integrated into the host tissues in all chimeric combinations tested, albeit with lower frequencies than for intraspecific transplantations.

The analysis of interspecific chimeras provides information about the compatibility between species of mechanisms controlling certain aspects of the development of cells or tissues. Such studies have already been successful in chick/quail chimeras (Le Douarin et al. 1975; Alvarado-Mallart and Sotelo, 1984; Tanaka and Landmesser, 1986; Balaban et al. 1988), in the mouse (Gardner, 1968; Goldowitz, 1986), in salamander (Harris, 1980), and in Xenopus (O’Gorman and Hunt, 1986). In Drosophila, compatibility between nuclei and cytoplasm has been demonstrated by tranplantation of nuclei between embryos of different Drosophila species (Santamaria, 1975). Transplantations of cytoplasm from oocytes and nurse cells from various Drosophila species into eggs of mutant Drosophila melanogaster have been performed in order to test the functional compatibility of factors specifying anterior/posterior polarity (K. Sander, personal communication; Schrbder, 1989).

The first step in neurogenesis requires a choice by cells of the neurogenic ectoderm of two alternative pathways of development, namely to become a neuroblast or an epidermoblast. This decision depends on interactions with neighbouring cells (Doe and Goodman, 1985b; Technau and Campos-Ortega, 1986,1987). In Drosophila, this communication process is under the control of a group of neurogenic genes (Wright, 1970; Lehmann etal. 1981) that seem to mediate an epidermalizing signal (Campos-Ortega, 1988) causing about 75 % of the cells of the neurogenic ectoderm to switch from a primary neurogenic to a secondary epidermogenic fate (Hartenstein and Campos-Ortega, 1984). Upon transplantation into a heterologous background (interspecific transplantation), cells of the ventral neurogenic ectoderm are still able to give rise to epidermal and/or neural clones. This may not necessarily result from interactions with cells of the heterologous host, since cells could have been determined as epidermoblasts or neuroblasts while still in the donor. But, as previously shown for D. melanogaster, epidermoblasts and neuroblasts do not become irreversibly committed to their fates (Technau et al. 1988). Accordingly, cells of the dorsal epidermal anlage, which normally only give rise to epidermal clones, may also produce neural or mixed epidermal/neural clones when heterotopically transplanted into the ventral neurogenic ectoderm (Technau and Campos-Ortega, 1986). When we performed the same heterotopic transplantation experiment among different species, cells behaved like those in the corre-ponding intraspecific transplantation experiment. We therefore believe that their epidermal or neural fate is actually determined by interactions with the heterologous host cells, and that the functional elements involved are conserved between D. melanogaster and the species tested. On the other hand, we cannot exclude that dorsal ectodermal cells may acquire a neural fate because they are released from an inhibiting factor when implanted into the ventral neurogenic region.

Following delamination from the neurogenic ectoderm, neuroblasts begin to proliferate (Doe and Goodman, 1985a). Distribution of neural clone sizes was in the same range for inter- and intraspecific transplantations. Thus, proliferative capabilities of neuroblasts did not seem to be influenced by surrounding heterologous cells although cell sizes and mitotic cycles differed considerably between some of the species. In grasshoppers, it has been shown that neuroblasts give rise to characteristic invariant lineages. Experimental delay of delamination from the neurogenic ectoderm does not prevent a neuroblast from producing its complete characteristic lineage (Doe and Goodman, 1985b).

Size distribution of epidermal clones, on the other hand, was clearly affected in a heterologous background. In a homologous environment, sizes of epidermal clones varied between 2 and 8 cells (corresponding to 3 symmetric divisions; see also Technau and Campos-Ortega, 1986). Following interspecific transplantations, the distribution of clone sizes differed in two ways: many of the transplanted cells did not divide at all but, on the other hand, there were several clones with more than 8 cells. Almost all of the large clones were derived from dorsal epidermoblasts, heterotopically transplanted into the ventral neurogenic ectoderm. Therefore, dorsal epidermoblasts are potentially capable of producing more than 8 progeny. This suggests that, in the case of epidermoblasts, control of proliferative activity is not cell-autonomous. Instead, it seems that under normal conditions, interactions among cells of the epidermal primordium are involved in controlling their mitotic activity. When epidermoblasts from D. virilis, D. hydei, D. busckii or D. latifasciaeformis are brought into a D. melanogaster environment, the assumed interactions do not appear to function properly. This suggests that the underlying mechanisms are not entirely compatible between these species. This incompatibility seems to be unidirectional, in that normal restrictions in clone sizes are imposed on epidermoblasts from D. melanogaster by cells from D. virilis, D. hydei, D. busckii and D. latifasciaeformis, the reverse is not the case.

A further basic step towards defining the fate of neural clones is the choice of cells to develop as neurones or glia. We found three categories of clones, namely, neuronal, glial and mixed neuronal/glial clones, indicating that both separate and common progenitor cells exist for neurones and glia (see also Prokop and Technau, 1990). Since glial clones generally contain only few cells, determination of a neural precursor as a glioblast would have to involve restrictions of its proliferative capabilities. Our heterotopic transplantations reveal the potential of dorsal ectodermal cells to switch from their normal epidermal fate to a neural one. Those selecting the neural pathway may produce neurones or glia. It is likely that cell –cell interactions determine whether a neural precursor acts as a neuroblast or glioblast. We found the same general behaviour for intra- and interspecifically transplanted cells, indicating that the mechanisms leading to the neuronal/glial dichotomy of neural daughter cells also function in heterologous Drosophila hosts.

During differentiation, neuronal and glial progeny cells acquire their typical shape, in that neurones develop axonal and dendritic fibers whereas glial cells form sheath-like protrusions. However, the ability of the cells to form these structures following interspecific transplantation does not in itself indicate that this process results from communication with the cells of the respective host. Since neuronal differentiation is also observed in vitro (Seecof et al. 1973; Furst and Mahowald, 1985), it is possible that interspecifically transplanted cells simply grow under the ideal ‘culture conditions’ of the respective hosts. However, in our experiments, the donor cells do differentiate into identifiable cell types with characteristic morphologies and locations within the host CNS. To attain their characteristic shapes, neurons have to project their axons along well-defined pathways. This process of axonal pathfinding involves cell recognition and selective fasciculation (e.g. Bastiani et al. 1985). In vertebrate systems, the formation of normal retinotectal projections has been demonstrated for chick/quail (Alvarado-Mallart and Sotelo, 1984) and salamander chimeras (Harris, 1980). Further, quail motoneurones innervate leg muscles of the chick with high precision (Tanaka and Landmesser, 1986). Thus, the functions controlling cell recognition during neuronal pathfinding are compatible between the respective species. Consistent with this idea is our finding that in all chimeric combinations tested, neuronal cells of the donor were able to fasciculate along specific pathways within the CNS of the host.

Our ultrastructural data revealed that the differentiated donor cells became fully integrated into the host tissue. They were closely apposed to the membranes of the surrounding host cells, thus facilitating the putative cell – cell interactions. Although we were unable to test for the presence of synapses, the fact that the cells formed axonal and dendritic spines may be taken as an indirect indication that synaptic contacts are actually formed between the heterologous cells. Furthermore, donor cells are still alive at late postembryonic stages. Cells that succeed in forming connections with the target cells are less likely to die than those that fail (Hollyday et al. 1977; Oppenheim, 1981).

As a functional test at the level of a single cell, our experiments demonstrate the ability of cells of five different Drosophila species to interact appropriately with cells of D. melanogaster. These interactions result in fully differentiated clones with identifiable cell types, although with different efficiencies. This suggests that mechanisms of cell –cell interactions mediating particular steps during neurogenesis, are highly conserved among different Drosophila species.

We thank Klaus Sander, Rachel Kraut, Eduardo Macagno and Laura Wolszon for carefully reviewing the text, Jose A. Campos-Ortega for working facilities and comments on the manuscript, Eva Varus for technical assistance, Klaus Sander for D. virilis, D. pseudoobscura, D. busckii and D. latifasciaeformis and Oswald Hess for D. hydei. This work was supported by a grant from the Deutsche Forschungsgemein-schaft to G.M.T. (DFG, Te 130/1 –2).

Alvarado-Mallart
,
R. M.
and
Sotelo
,
C.
(
1984
).
Homotopic and heterotopic transplantation of quail tectic primordia in chick embryos: Organization of the retinotectal protections in the chimeric embryos
.
Devi Biol
.
103
,
378
398
.
Balaban
,
E.
,
Teillet
,
M. A.
and
Le Douarin
,
N. M.
(
1988
).
Application of the chick-quail chimera system to the study of brain development and behaviour
.
Science
241
,
1339
1342
.
Bastiani
,
M. J.
,
Doe
,
C. Q.
,
Helfand
,
S. L.
and
Goodman
,
C. S.
(
1985
).
Neuronal specificity and growth cone guidance in grasshopper and Drosophila embryos
.
Trends Neurosci
.
8
,
257
266
.
Bastiani
,
M. J.
,
Harrelson
,
A. L.
,
Snow
,
P. M.
and
Goodman
,
C. S.
(
1987
).
Expression of fasciclin 1 and II glycoproteins on subset of axon pathways during neuronal development in the grasshopper
.
Cell
48
,
745
755
.
Bate
,
C. M.
(
1976
).
Embryogenesis of an insect nervous system. I. A map of the thoracic and abdominal neuroblasts in Locusta migratoria
.
J. Embryol. exp. Morph
.
35
,
107
123
.
Campos-Ortega
,
J. A.
(
1988
).
Cellular interactions during early neurogenesis of Drosophila melanogaster
.
Trends Neurosci
.
11
,
400
405
.
Campos-Ortega
,
J. A.
and
Hartenstein
,
V.
(
1985
).
The Embryonic Development of Drosophila melanogaster. Springer, Berlin Heidelberg New York
Doe
,
C. Q.
and
Goodman
,
C. S.
(
1985a
).
Early events in insect neurogenesis. I. Development and segmental differences in the pattern of neuronal precursor cells
.
Devi Biol
.
111
,
193
205
.
Doe
,
C. Q.
and
Goodman
,
C. S.
(
1985b
).
Early events in insect neurogenesis. II. The role of cell interactions and cell lineage in the determination of neuronal precursor cells
.
Devi Biol
.
111
,
206
219
.
Furst
,
A.
and
Mahowald
,
A. P.
(
1985
).
Differentiation of primary embryonic neuroblasts in purified neural cell cultures from Drosophila.
Devl Biol
.
109
,
184
192
.
Gardner
,
R. L.
(
1968
).
Mouse chimeras obtained by injection of cells into the blastocyst
.
Nature
220
,
596
597
.
Goldowitz
,
D.
(
1986
).
Inter-specific mouse chimeras to study mutant gene action and cell lineage in the CNS
.
Soc. Neurosci Abstr
,
12
,
1583
.
Goodman
,
C. S.
and
Spitzer
,
N. C.
(
1979
).
Embryonic development of identified neurones; differentiation from neuroblast to neurone
.
Nature
280
,
208
214
.
Hafen
,
E.
,
Basler
,
K.
,
Edstroem
,
J.-E.
and
Rubin
,
G. M.
(
1987
).
Sevenless, a cell-specific homeotic gene of Drosophila, encodes a putative transmembrane receptor with a tyrosine kinase domain
.
Science
236
,
55
63
.
Harris
,
W. A.
(
1980
).
The effects of eliminating impulse activity on the development of the retinotectal projection in salamanders
.
J. comp. Neurol
194
,
303
317
.
Hartenstein
,
V.
and
Campos-Ortega
,
J. A.
(
1984
).
Early neurogenesis in wild-type Drosophila melanogaster
.
Willhelm Roux’s Arch, devl Biol
.
193
,
308
325
.
Hartenstein
,
V.
,
Rudloff
,
E.
and
Campos-Ortega
,
J. A.
(
1987
).
The pattern of proliferation of the neuroblasts in the wild-type embryo of Drosophila melanogaster
.
Roux’s Arch, devl Biol
.
196
,
473
485
.
Hartenstein
,
V.
,
Technau
,
G. M.
and
Campos-Ortega
,
J. A.
(
1985
).
Fate-mapping in wild-type Drosophila melanogaster. III. A fate map of the blastoderm
.
Roux’s Arch, devl Biol
.
194
,
213
216
.
Holliday
,
M.
,
Hamburger
,
V.
and
Farris
,
J.
(
1977
).
Localization of motor neuron pools supplying identified muscles in normal and supernumerary legs of chick embryo
.
Proc. natn. Acad. Set. U.S.A
.
74
,
3582
3586
.
Hooper
,
J. E.
and
Scott
,
M. P.
(
1989
).
The Drosophila patched gene encodes a putative membrane protein required for segmental patterning
.
Cell
59
,
751
765
.
Kuwada
,
J. Y.
and
Goodman
,
C. S.
(
1985
).
Neuronal determination during embryonic development of the grasshopper nervous system
.
Devi Biol
110
,
114
126
.
Le Douarin
,
N. M.
,
Renaud
,
D.
,
Teilet
,
M. A.
and
Le Douarin
,
G. H.
(
1975
).
Cholinergic differentiation of presumptive adrenergic neuroblasts in interspecific chimeras after heterotopic transplantations
.
Proc. natn. Acad. Sci. U.S.A
.
72
,
728
732
.
Lehmann
,
R.
,
Dietrich
,
U.
,
Jiminez
,
F.
and
Campos-Ortega
,
J. A.
(
1981
).
Mutations of early neurogenesis in Drosophila
.
Willhelm Roux’s Arch, devl Biol
.
190
,
226
229
.
O’gorman
,
O.
and
Hunt
,
R. K.
(
1986
).
Clones and polyclones in retinas of Xenopus interspecific cell-injection chimerae
.
Soc. Neurosci. Abstr
.
12
,
1583
.
Oppenheim
,
R. W.
(
1981
).
Neuronal cell death and some related regressive phenomena during neurogenesis: A selective historical review and progress report
. In
Studies in Developmental Biology
(ed.
W. M.
Cowan
), pp.
74
133
.
New York
:
Oxford University Press
Poulson
,
D. F.
(
1950
).
Histogenesis, organogenesis and differentiation in the embryo of Drosophila melanogaster
. In
Biology of Drosophila
(ed.
M.
Demerec
), pp.
168
274
.
New York
:
Wiley
.
Prokop
,
A.
and
Technau
,
G. M.
(
1990
).
The origin of imaginal neuroblasts in the ventral nerve cord of Drosophila melanogaster. In preparation
.
Reynolds
,
E. S.
(
1963
).
The use of lead citrate at high pH as an electron opaque stain in electron microscopy
.
J. Cell Biol
.
17
,
208
212
.
Santamaria
,
P.
(
1975
).
Transplantation of nuclei between eggs of different species of Drosophila. Wilhelm Roux’s Archiv
.
devl Biol
.
178
,
89
98
.
SchrÖder
,
R.
(
1989
).
Partielle Polaritatsumkehr in bicoid- Embryonen von Drosophila melanogaster nach heterospezifischer Nfihrzellplasma Transplantation oder Zentrifugierung. Diploma thesis, Univ. Freiburg
.
Seecof
,
R. L.
,
Donady
,
J. J.
and
Teputz
,
R. L.
(
1973
).
Differentiation of Drosophila neuroblasts to form ganglion-like clusters of neurons in vitro
.
Cell Diff
.
2
,
143
149
.
Tanaka
,
H.
and
Landmesser
,
L. T.
(
1986
).
Interspecies selective montoneuron projection patterns in chick-quail chimeras
.
J. Neurosci
.
6
,
2880
2888
.
Technau
,
G. M.
(
1986
).
Lineage analysis of transplanted individual cells in embryos of Drosophila melanogaster. I. The method
.
Roux’s Archiv. devl Biol
.
195
,
389
398
.
Technau
,
G. M.
,
Becker
,
T.
and
Campos-Ortega
,
J. A.
(
1988
).
Reversible commmitment of neural and epidermal progenitor cells during embryogenesis of Drosophila melanogaster
.
Roux’s Archiv. devl Biol
.
197
,
413
418
.
Technau
,
G. M.
and
Campos-Ortega
,
J. A.
(
1985
).
Fate-mapping in wild-type Drosophila melanogaster. II. Injections of horseradish peroxidase in cells of the early gastrula stage
.
Wilhelm Roux’s Archiv. devl Biol
.
194
,
196
212
.
Technau
,
G. M.
and
Campos-Ortega
,
J. A.
(
1986
).
Lineage analysis of transplanted individual cells in embryos of Drosophila melanogaster. II. Commitment and proliferative capabilities of neural and epidermal cell progenitors
.
Roux’s Arch devl Biol
.
195
,
445
454
.
Technau
,
G. M.
and
Campos-Ortega
,
J. A.
(
1987
).
Cell autonomy of expression of neurogenic genes of Drosophila melanogaster
.
Proc. natn. Acad Sci. U.S.A
.
84
,
4500
4504
.
Wheeler
,
W. M.
(
1893
).
A contribution to insect embryology
.
J. Morphol
8
,
1
160
.
Wright
,
T. R. F.
(
1970
).
The genetics of embryogenesis in Drosophila
.
Adv. Genetics
15
,
262
395
.
Zalokar
,
M.
and
Erk
,
I.
(
1977
).
Phase-partition fixation and staining of Drosophila eggs
.
Stain Technol
.
52
,
89
95
.