Fig 1.
Cell shape change analysis during Drosophila axis extension.
(A) Scanning electron microscopy micrographs (from Flybase, [36]) showing lateral views of gastrulating Drosophila embryos at stage six, seven, and eight. Anterior is to the left. The part of the germband undergoing convergent extension is labelled in purple and the main direction of extension indicated with a yellow arrow. The following landmarks or morphogenetic events are indicated: PC, pole cells; CF, cephalic furrow; ATF, anterior transverse furrow; PTF, posterior transverse furrow; PMI, posterior midgut invagination (also called posterior endoderm invagination). The box represents the ventral surface that we optically sectioned by confocal microscopy. (A’) Scanning electron microscopy micrograph showing a ventral view of a gastrulating embryo at stage eight, with the approximate position of the anterior and posterior field of views that we analyzed. Both views are bissected by the ventral furrow (VF) through which the mesoderm invaginates. (B) Schematics representing the small neighborhood of cells considered by the tracking algorithms. Cell shape changes (strain rates) are calculated by comparing two timepoints before and after a given time. A strain rate is the ratio of the change in length to the original length, divided by the time interval, with units of proportion per minute (pp/min). The cell shape change is represented here by two orthogonal vectors, showing elongation in one direction (blue, positive) and shrinkage in the perpendicular direction (red, negative). (C, C’) Example frames of movies of wild-type embryos at t = 10 min after GBE onset labeled with ubi-DE-cad-GFP, showing an anterior (C, wtLB009) and a posterior view (C’ wtCL051010). (C) For anterior views, the cephalic furrow (arrow) is used as an anterior landmark, and the scale shows distance from this landmark. The purple shading shows the region removed from the analysis, where cells stretch behind the cephalic furrow. (C’) For posterior views, the posterior-most edge of the embryo seen in the confocal stack is used as posterior landmark (arrow), with distance from it indicated in the scale. (D, D’) Outcome of tracking for anterior (D) and posterior views (D’), showing the polygons describing the cell outlines and the cell centroids from which are drawn the tracks giving the cell positions for the previous 2.5 min (5 timepoints). The tracks shown are track retained for the analysis, after removing tracks from mesodermal cells, mesectoderm cells, and for anterior views, from cells deformed by the cephalic furrow (corresponding to purple region in C). (E, E’) Spatiotemporal heat maps summarizing AP cell length change contributing to GBE, over the first 30 min of GBE (y-axis) and as a function of cell position in the AP axis (x-axis), for anterior (E) and posterior views (E’), averaged for five and four embryos, respectively. (F, F’) Graphs summarizing AP cell length change as a function of time for the first 30 min of GBE, for anterior (F) and posterior views (F’), averaged for five and four embryos, respectively. The ribbon’s width shows the standard error (see Materials and Methods). Data associated with this figure can be found in S1 Data.
Fig 2.
AP cell elongation patterns form an AP gradient.
(A, B) AP cell length change shown for each analyzed cell, for timepoints 7.5, 10, and 12.5 min after GBE onset in anterior (A, wtLB009) and posterior views (B, wtCL051010). The color of the dot at the center of each cell corresponds to the scale bar shown. (A’, B’) Spatial maps summarizing AP cell length change over the 7.5–12.5 min time interval, as a function of the position of cells in the AP (x-axis) and DV (y-axis) embryonic axes, for anterior (A’) and posterior views (B’) (average for five and four embryos per views, respectively). (C, D) Graphs summarizing AP cell length change over the 7.5–12.5 min time interval, as a function of cell position in the AP axis, for anterior (C) and posterior views (D). (E) Graphs summarizing AP cell length change (y-axis), at 2.5, 5, and 7.5 min after the onset of GBE, as a function of cell position in the AP axis (x-axis) for posterior views (average for four wild-type embryos). The ribbon’s width shows the standard error (see Materials and Methods). Data associated with this figure can be found in S2 Data.
Fig 3.
The AP cell elongation gradient is present in twi mutant embryos.
(A) Spatiotemporal map summarizing AP cell length change contributing to GBE, over the first 30 mins of GBE (y-axis), and as a function of cell position in the AP axis (x-axis), for twi mutants in anterior views (average for five embryos). (A’) Graph comparing AP cell length change as a function of time for the first 30 min of GBE, in wild-type (blue) and twi mutants (red) for anterior views (average for five embryos each). In these graphs and thereafter, the ribbon’s width indicates the standard error, and the grey-shaded boxes show where a difference is statistically significant (p < 0.05, see Materials and Methods). (B, B’) Equivalent plots as A, A’ for posterior views. (C, D) AP cell length change shown for each analyzed cell, for timepoints 7.5, 10, and 12.5 min after GBE onset in twi mutant embryos, for movie frames of an anterior (C, twiLB012) and a posterior view (D, twiCL140411). The color of the dot at the center of each cell corresponds to the scale bar shown. (C’, D’) Spatial maps summarizing AP cell length change over the 7.5–12.5 mins time interval, as a function of the position of cells in the AP (x-axis) and DV (y-axis) embryonic axes, for anterior and posterior views in twi mutants (average of five and three embryos per view, respectively). (E, F) Graphs comparing AP cell length change over the 7.5–12.5 min time interval, as a function of cell position in the AP axis, for wild-type (blue) and twi mutant (red) embryos, for anterior and posterior views. (G–G”) Graphs summarizing AP length change (y-axis), at 2.5, 5, and 7.5 min after the onset of GBE, as a function of cell position in the AP axis (x-axis) for wild-type (blue) and twi mutant (red) embryos, for posterior views (average for four and three embryos per genotype, respectively). Data associated with this figure can be found in S3 Data.
Fig 4.
Temporal relationships between morphogenetic movements during Drosophila gastrulation.
(A) Diagram showing the sites of the different morphogenetic movements on lateral views of stage six and eight embryos (see also Fig 1A and 1A’). The invaginating mesoderm and endoderm layers are shown in yellow and red, respectively. At stage six, these layers are in the process of invagination from the surface to the interior of the embryo, and by stage eight, both of these layers are fully internalized. PC, pole cells; CF, cephalic furrow; ATF, anterior transverse furrow; PTF, posterior transverse furrow; GBE, germband extension. (B) Graph summarizing the temporal mapping of the head mitoses and four morphogenetic events relative to the onset of GBE, for three wild-type and three twi mutant movies. (C–H’) All views are taken from a wild-type embryo labeled with the whole-membrane markers resille-GFP and spider-GFP and imaged by light sheet microscopy (SPIM). For each morphogenetic movement, two movie frames are shown, before and after the temporally mapped event. The position of the pole cells (PC) at the posterior of the embryos are indicated by a bracket. All embryos shown are at either late stage five or early stage six, except the embryo in D, which is at stage eight. At these early stages, embryonic cells are arranged in a single columnar layer surrounding a large yolk cytoplasm, visible as a bright signal in lateral views. (C, C’) Lateral views showing the mapping of individual cells (yellow dots) 1 min before (C) and 1.5 min after (C’) the onset of GBE. The arrowheads indicate the position of the cells before the start of GBE. (D–D”) Lateral views with a box highlighting the mitotic domain 2 in the head [40], with close-up in (D’) showing a cell about to divide and (D”) the two resulting daughter cells 3.5 min later. (E, E’) Lateral views showing the mapping of cells (red dots) 2 min before (E) and 2 min after (E’) the onset of apical constriction in the endoderm primordium. The close-ups show that the cells between the pair of red dots constricted their apices, causing the dots to move closer together. (F, F’) Ventral view showing the ventral furrow 2.5 min before (F) and 1.5 min after (F’) mesoderm sealing. (G, G’) Lateral view showing the mapping of the apical surface of a cell in comparison to its original position (green dots) on the dorsal surface 2.5 min before (G) and 4 min after (G’) the initiation of the posterior transverse fold (one of the two dorsal folds). The close-up shows the basalwards displacement of the apical surface. (H, H’) Lateral view showing the movement of individual cells (magenta dots) 1 min before (H) and 3 min after (H’) dorsal contraction. The arrowheads show the cells’ positions before contraction. Data associated with this figure can be found in S4 Data.
Fig 5.
The AP cell elongation gradient is present in Kr but not Kr; tsl double mutants.
(A) Graph comparing AP cell length change contributing to GBE as a function of time for the first 30 min of GBE, in wild-type (blue) and Kr mutants (red), in posterior views (average for four wild-type and three Kr embryos). In this graph and thereafter, the ribbon’s width indicates the standard error and the grey-shaded boxes show where a difference is statistically significant (p < 0.05, see Materials and Methods). (B, C) Spatiotemporal maps summarizing AP cell length change over the first 30 min of GBE (y-axis) and as a function of cell position in the AP axis (x-axis), for Kr and Kr; tsl mutants (average for three embryos of each genotype; individual movie plots in S5 Fig). (D) Graph summarizing AP cell length change as a function of time for the first 30 min of GBE for Kr (blue) and Kr; tsl mutants (red). (E, F) AP cell length change shown for each analyzed cell for timepoints 7.5, 10, and 12.5 min after GBE onset in a Kr (KrCL051112) and Kr; tsl mutant embryo (KrtslCL040713). The color of the dot at the center of each cell corresponds to the scale bar shown. (E’, F’) Spatial maps summarizing AP cell length change over the 7.5–12.5 min time interval, as a function of the position of cells in the AP (x-axis) and DV (y-axis) embryonic axes (average for three embryos per genotype). (G) Graph comparing AP cell length change over the 7.5–12.5 min time interval, as a function of cell position in the AP axis for Kr (blue) and Kr; tsl mutant (red) embryos (average for three embryos per genotype). (H, I) Spatiotemporal maps summarizing cell area changes with same axes as B, C for Kr and Kr; tsl mutant embryos. Data associated with this figure can be found in S5 Data.
Fig 6.
Apical constriction of the posterior endoderm primordium generates a tensile stress in acellular embryos.
(A, B) Examples of movies of the posterior lateral surface of acellular embryos, with the actomyosin cytoskeleton labelled with sqh-GFP (see also S8 Movie, from which example A is taken). The Myosin II signal forms a disorganized meshwork at the apical surface of the embryo, which concentrates in the region close to the PC (see arrows in B). Occasionally, the meshwork becomes more cable-like, orienting towards the presumptive posterior endoderm (identified by the position of the PC). (A’, B’) Particle Imaging Velocimetry (PIV) tracking of the Myosin II signal reveals flows towards the presumptive posterior endoderm. Note that the ventralward flows also seen here move towards the ventral presumptive mesoderm (see S8 Movie). Arrows represent the displacement from the previous timepoint, scaled by a factor of four. Magnitude is shown using a heat scale, with fastest flows in red. Times shown are from the start of the movies. Scale bars are 20 microns. (C) Cross section of the posterior of an acellular embryo stained for Myosin II (see S6 Fig), showing the concentration of Myosin II where the apical surface has contracted and the beginning of an invagination. PC are indicated. (D) Schematics showing the position of the laser cuts performed on the lateral surface of the presumptive germband in acellular embryos (approximately to scale). The cuts are along a line 20 microns long, positioned either at the anterior (ant.) or at the posterior (post.) of the embryo, either orthogonal to the posterior flows (magenta, called DV thereafter for simplicity) or parallel to them (blue, called AP thereafter). (E) Dot plot with box plot overlaid showing the normalized relaxation velocities (corrected for displacement) for each category of cuts (n = 13 for ant. AP; n = 11, ant. DV; n = 16, post. AP; n = 12, post. DV). A two sample t test was used for the statistics (comparison ant. AP and ant. DV, ns: p = 0.070; comparison ant. AP and post. AP, ns: p = 0.1225; *: p = 0.0173; **: p = 0.0011). For the box plots, the central red line is the median, the edges of the box are the 25th and 75th percentile, the whiskers extend to the most extreme points not considered outliers, and outliers are plotted individually (Data points considered outliers are those more than 2.7 standard deviations from the mean). Data associated with this Figure can be found in S6 Data.
Fig 7.
Our findings indicate that apical constriction and invagination of the endoderm primordium (red region) causes a tensile stress that is propagated to the germband and elongates the cells in AP (pink). Passive AP cell elongation and genetically-programmed polarized cell intercalation (blue) contribute together to Drosophila GBE (green arrow). endo, endoderm.