In-Vivo Imaging of the Drosophila Wing Imaginal Disc over Time: Novel Insights on Growth and Boundary Formation

In developmental biology, the sequence of gene induction and pattern formation is best studied over time as an organism develops. However, in the model system of Drosophila larvae this oftentimes proves difficult due to limitations in imaging capabilities. Using the larval wing imaginal disc, we show that both overall growth, as well as the creation of patterns such as the distinction between the anterior(A) and posterior(P) compartments and the dorsal(D) and ventral(V) compartments can be studied directly by imaging the wing disc as it develops inside a larva. Imaged larvae develop normally, as can be seen by the overall growth curve of the wing disc. Yet, the fact that we can follow the development of individual discs through time provides the opportunity to simultaneously assess individual variability. We for instance find that growth rates can vary greatly over time. In addition, we observe that mechanical forces act on the wing disc within the larva at times when there is an increase in growth rates. Moreover, we observe that A/P boundary formation follows the established sequence and a smooth boundary is present from the first larval instar on. The division of the wing disc into a dorsal and a ventral compartment, on the other hand, develops quite differently. Contrary to expectation, the specification of the dorsal compartment starts with only one or two cells in the second larval instar and a smooth boundary is not formed until the third larval instar.     

Recruitment of cells into the Drosophila wing primordium by a feed-forward circuit of vestigial autoregulation

The Drosophila wing primordium is defined by expression of the selector gene vestigial (vg) in a discrete subpopulation of cells within the wing imaginal disc. Following the early segregation of the disc into dorsal (D) and ventral (V) compartments, vg expression is governed by signals generated along the boundary between the two compartments. Short-range DSL (Delta/Serrate/LAG-2)-Notch signaling between D and V cells drives vg expression in ;border' cells that flank the boundary. It also induces these same cells to secrete the long-range morphogen Wingless (Wg), which drives vg expression in surrounding cells up to 25-30 cell diameters away. Here, we show that Wg signaling is not sufficient to activate vg expression away from the D-V boundary. Instead, Wg must act in combination with a short-range signal produced by cells that already express vg. We present evidence that this vg-dependent, vg-inducing signal feeds forward from one cell to the next to entrain surrounding cells to join the growing wing primordium in response to Wg. We propose that Wg promotes the expansion of the wing primordium following the D-V segregation by fueling this non-autonomous autoregulatory mechanism.     

A Model for Selection of Eyespots on Butterfly Wings

Unsolved Problem

The development of eyespots on the wing surface of butterflies of the family Nympalidae is one of the most studied examples of biological pattern formation.However, little is known about the mechanism that determines the number and precise locations of eyespots on the wing. Eyespots develop around signaling centers, called foci, that are located equidistant from wing veins along the midline of a wing cell (an area bounded by veins). A fundamental question that remains unsolved is, why a certain wing cell develops an eyespot, while other wing cells do not.

Key Idea and Model

We illustrate that the key to understanding focus point selection may be in the venation system of the wing disc. Our main hypothesis is that changes in morphogen concentration along the proximal boundary veins of wing cells govern focus point selection. Based on previous studies, we focus on a spatially two-dimensional reaction-diffusion system model posed in the interior of each wing cell that describes the formation of focus points. Using finite element based numerical simulations, we demonstrate that variation in the proximal boundary condition is sufficient to robustly select whether an eyespot focus point forms in otherwise identical wing cells. We also illustrate that this behavior is robust to small perturbations in the parameters and geometry and moderate levels of noise. Hence, we suggest that an anterior-posterior pattern of morphogen concentration along the proximal vein may be the main determinant of the distribution of focus points on the wing surface. In order to complete our model, we propose a two stage reaction-diffusion system model, in which an one-dimensional surface reaction-diffusion system, posed on the proximal vein, generates the morphogen concentrations that act as non-homogeneous Dirichlet (i.e., fixed) boundary conditions for the two-dimensional reaction-diffusion model posed in the wing cells. The two-stage model appears capable of generating focus point distributions observed in nature.

Result

We therefore conclude that changes in the proximal boundary conditions are sufficient to explain the empirically observed distribution of eyespot focus points on the entire wing surface. The model predicts, subject to experimental verification, that the source strength of the activator at the proximal boundary should be lower in wing cells in which focus points form than in those that lack focus points. The model suggests that the number and locations of eyespot foci on the wing disc could be largely controlled by two kinds of gradients along two different directions, that is, the first one is the gradient in spatially varying parameters such as the reaction rate along the anterior-posterior direction on the proximal boundary of the wing cells, and the second one is the gradient in source values of the activator along the veins in the proximal-distal direction of the wing cell.     

Differential division rates and size control in the wing disc

We have generated wing disc compartments that contain marked fast growing M+ clones surrounded by slow dividing M/+ cells. Under these conditions the interactions between fast and slow dividing cells at the clone borders frequently lead to cell competition. However, our assay suppressing apoptosis indicates that cell competition plays no major role in size control. We argue that cells within a compartment proliferate according to their genotype independently of each other and that their contribution to the final structure will depend solely on their proliferation rate. This model is supported by a computer simulation that predicts values similar to those found experimentally. Our results on the growth of M+ clones within compartments and on the expression of developmental genes like vestigial and wingless suggest the existence of a non-cell autonomous mechanism that functions at the level of the entire cell population. It measures the population size in each moment, determines the corresponding expression levels of developmental genes and establishes the time to arrest growth.     

In vitro culture of Drosophila imaginal disc cells: aggregation, sorting out, and differentiative abilities

This paper describes the aggregation in vitro of cells dissociated from imaginal discs and demonstrates the sorting out of undifferentiated cells from different imaginal discs and from differently determined regions of the same imaginal disc, as well as the abilities of such cells to undergo pattern reconstruction when injected into larvae. Dissociated cells begin to aggregate by 1.5 hr of rotation. By 5 hr of rotation, large aggregates of loosely associated cells appear. By 18 hr the aggregates have condensed and taken on a characteristic epithelial structure. To study sorting out in undifferentiated cells, we combined a histochemical stain for acid phosphatase with the use of the acid phosphatase null mutant acphn-11. We performed cell mixing experiments with 0-2 (prospective notum) and 2-8 (prospective wing) fragments, with the A and P (prospective anterior and posterior) fragments of the dorsal mesothoracic disc and with mixtures of cells from ventral prothoracic and dorsal mesothoracic discs. We found that prospective anterior and posterior dorsal mesothoracic cells do not sort out, but that prospective notum and wing and leg and wing cells do. The results from differentiated implants are consistent with those from undifferentiated mixes.     

Targeted expression of the signaling molecule decapentaplegic induces pattern duplications and growth alterations in Drosophila wings.

In the wing imaginal disc, the decapentaplegic (dpp) gene is expressed in a stripe of anterior cells near the anterior-posterior compartment boundary, and it is required solely in these cells for the entire disc to develop. In some viable segment polarity mutants, alterations in dpp expression have been demonstrated that correlate with changes in wing morphology. To test the hypothesis that the abnormal patterns of dpp expression are responsible directly for the mutant phenotypes, we have expressed dpp in ectopic places in wing imaginal discs, and we have found that dpp is able to cause overgrowth and pattern duplications in both anterior and posterior compartments of the wing disc. The alterations of the anterior compartment are strikingly similar to those observed in some viable segment polarity mutants. Thus, ectopic dpp alone can account for the phenotype of these mutants. We also show that ectopic expression of the segment polarity gene hedgehog (hh) gives similar morphological changes and activates dpp expression in the anterior compartment. This strongly suggests that the organizating activity of hh is mediated by dpp. We propose that the expression of dpp near the anterior-posterior compartment boundary is directed by the interaction between patched and hh, and that dpp itself could act as a general organizer of the patterning in the wing imaginal disc.     

Compartments and the control of growth in the Drosophila wing imaginal disc

The mechanisms that control organ growth are among the least known in development. This is particularly the case for the process in which growth is arrested once final size is reached. We have studied this problem in the wing disc of Drosophila, the developmental and growth parameters of which are well known. We have devised a method to generate entire fast-growing Minute(+) (M(+)) discs or compartments in slow developing Minute/+ (M/+) larvae. Under these conditions, a M(+) wing disc gains at least 20 hours of additional development time. Yet it grows to the same size of Minute/+ discs developing in M/+ larvae. We have also generated wing discs in which all the cells in either the anterior (A) or the posterior (P) compartment are transformed from M/+ to M(+). We find that the difference in the cell division rate of their cells is reflected in autonomous differences in the developmental progression of these compartments: each grows at its own rate and manifests autonomous regulation in the expression of the developmental genes wingless and vestigial. In spite of these differences, ;mosaic' discs comprising fast and slow compartments differentiate into adult wings of the correct size and shape. Our results demonstrate that imaginal discs possess an autonomous mechanism with which to arrest growth in anterior and posterior compartments, which behave as independent developmental units. We propose that this mechanism does not act by preventing cell divisions, but by lengthening the division cycle.     

Analysis of cell proliferation in Drosophila wing imaginal discs using mosaic clones

Experimental data on spatial and temporal distributions of mosaic clones in Drosophila wing imaginal disc were analyzed. Long-lived proliferation centers (PR1, PR2, and PR3) and areas with decreased proliferation activity were found in the notum region of the disc. Simulation of the growth kinetics of mosaic patches demonstrated that the cell cycle in proliferation centers PR2 and PR3 was shorter than the average cycle in the disc and in the center PR1. A nonrandom clustering of rapidly dividing cells was observed in the PR2, but not in the other cases. The reason why the cell-cycle duration and the clustering of dividing cells may not coincide is discussed in terms of the recruitment of nondividing cells into the cell cycle. The simulation of the time course of the first and second moments of the size distribution of mosaic clones allowed the variance of cell-cycle progression rates to be determined and demonstrated that a model with a continuous cell-cycle rates gave a better fit to the data than the transition probability model of Smith and Martin.     

A position-specific cell surface antigen in the drosophila wing imaginal disc

The antibody produced by the hybrid cell line DK.1A4 recognizes an antigen present initially on all the epithelial cells of the D. melanogaster wing imaginal disc. This antigen becomes progessively restricted to cells in the dorsal region of the disc during the final larval instar. The presence of the antigen does not correlate with the specific adult structures to which the cells will eventually contribute, but rather with the position of the cells in the disc. In late discs, the line bounding the region in which the antigen persists corresponds to the boundary between the dorsal and ventral compartments as revealed by a clonal analysis of the undifferentiated disc. Together, these data suggest that the antigen's disappearance may be specific to the cells of the ventral compartment of the wing disc.     

Dissociation and sorting out of Drosophila imaginal disc cells

Previous attempts to study sorting out of Drosophila imaginal disc cells have been hampered by an inability to thoroughly dissociate these cells and the need to use cuticular markers which require several days of in vivo culture. This study overcomes these limitations by using a new dissociation procedure and a genetic marker for undifferentiated cells, the succinate dehydrogenase8 (sdh8) mutation. Dissociated and reaggregated cells from wing and leg imaginal discs segregated or "sorted out" from one another after only 24 hr of in vivo culture. It was also found that leg cells from different body segments may sort out, but to a lesser degree than wing and leg cells. Mixtures of wing and haltere cells did not sort out, in contrast to previous reports. These results constitute the first unambiguous study of sorting out with Drosophila imaginal disc cells and indicate that dorsally situated imaginal cells share a recognition specificity which is different from that of ventral imaginal cells.     

On the constancy of cell division rate in the root meristem

This review examines under what circumstances the rate of cell division among cells of the root meristem is known to vary. First, methods are compared that have been used to quantify cell division rate. These can be grouped as being either cytological, in which the rate of accumulation of cells in a particular phase of the cell cycle is determined based on some form of cytological labeling, or kinematic, in which the rate of cell accumulation is determined from the net movement of cells. Then, evidence is reviewed as to whether cell division rates vary between different tissues or cell types, between different positions in the root, or finally between different environments. The evidence is consistent with cells dividing at a constant rate, and well documented examples where cell division rate changes substantially are rare. The constancy of cell division rate contrasts with the number of dividing cells, which varies extensively, and implies that a major point for cell cycle control is governing the exit from the proliferative state at the basal boundary of the meristem.     

Connecting Hh,Dpp and EGF signalling in patterning of the Drosophila wing; the pivotal role of collier/knot in the AP organiser

Hedgehog (Hh) signalling from posterior (P) to anterior (A) cells is the primary determinant of AP polarity in the limb field in insects and vertebrates. Hh acts in part by inducing expression of Decapentaplegic (Dpp), but how Hh and Dpp together pattern the central region of the Drosophila wing remains largely unknown. We have re-examined the role played by Collier (Col), a dose-dependent Hh target activated in cells along the AP boundary, the AP organiser in the imaginal wing disc. We found that col mutant wings are smaller than wild type and lack L4 vein, in addition to missing the L3-L4 intervein and mis-positioning of the anterior L3 vein. We link these phenotypes to col requirement for the local upregulation of both emc and N, two genes involved in the control of cell proliferation, the EGFR ligand Vein and the intervein determination gene blistered. We further show that attenuation of Dpp signalling in the AP organiser is also col dependent and, in conjunction with Vein upregulation, required for formation of L4 vein. A model recapitulating the molecular interplay between the Hh, Dpp and EGF signalling pathways in the wing AP organiser is presented.     

Genetic analysis of transdetermination in Drosophila. I. The effects of varying growth parameters using a temperature-sensitive mutation

The heat-sensitive mutation of Drosophila melanogaster l(3)c4(3)hs1, causes mutant larvae raised at a restrictive temperature to have abnormally large wing discs. The large size of these discs is a disc-autonomous property and results from an increase in the number rather than the size of wing disc cells. We have used wing discs from this mutant to further investigate properties of transdetermination which had previously been investigated with nonmutant discs. Transdetermination can occur in nonmutant discs when the proliferative phase of imaginal disc development is extended by wounding discs and culturing them in vivo. The results indicate that additional proliferation in the absence of wounding does not lead to transdetermination. There is a correlation between the extent of growth of a cultured disc and the probability that it will undergo transdetermination. The results suggest that this correlation does not depend on a differential rate of cell division. Finally, the results indicate that the cells which give rise to transdetermination are at an equivalent developmental stage no later than that characteristic of eye-antenna disc cells before the third larval instar.     

Roles for scalloped and vestigial in regulating cell affinity and interactions between the wing blade and the wing hinge

The scalloped and vestigial genes are both required for the formation of the Drosophila wing, and recent studies have indicated that they can function as a heterodimeric complex to regulate the expression of downstream target genes. We have analyzed the consequences of complete loss of scalloped function, ectopic expression of scalloped, and ectopic expression of vestigial on the development of the Drosophila wing imaginal disc. Clones of cells mutant for a strong allele of scalloped fail to proliferate within the wing pouch, but grow normally in the wing hinge and notum. Cells overexpressing scalloped fail to proliferate in both notal and wing-blade regions of the disc, and this overexpression induces apoptotic cell death. Clones of cells overexpressing vestigial grow smaller or larger than control clones, depending upon their distance from the dorsal-ventral compartment boundary. These studies highlight the importance of correct scalloped and vestigial expression levels to normal wing development. Our studies of vestigial-overexpressing clones also reveal two further aspects of wing development. First, in the hinge region vestigial exerts both a local inhibition and a long-range induction of wingless expression. These and other observations imply that vestigial-expressing cells in the wing blade organize the development of surrounding wing-hinge cells. Second, clones of cells overexpressing vestigial exhibit altered cell affinities. Our analysis of these clones, together with studies of scalloped mutant clones, implies that scalloped- and vestigial-dependent cell adhesion contributes to separation of the wing blade from the wing hinge and to a gradient of cell affinities along the dorsal-ventral axis of the wing.     

msh specifies dorsal cell fate in the Drosophila wing

Drosophila limbs develop from imaginal discs that are subdivided into compartments. Dorsal-ventral subdivision of the wing imaginal disc depends on apterous activity in dorsal cells. Apterous protein is expressed in dorsal cells and is responsible for (1) induction of a signaling center along the dorsal-ventral compartment boundary (2) establishment of a lineage restriction boundary between compartments and (3) specification of dorsal cell fate. Here, we report that the homeobox gene msh (muscle segment homeobox) acts downstream of apterous to confer dorsal identity in wing development.