Precise registration of gene expression boundaries by a repressive morphogen in Drosophila

Abstract

Background: Morphogen gradients are thought to create concentration thresholds that differentially position the expression boundaries of multiple target genes. Despite intensive study, it is still unclear how the concentration profiles within gradients are spatially related to the critical patterning thresholds they generate.

Results: Here we use a combination of quantitative measurements and ectopic-misexpression experiments to examine the transcriptional-repression activities of the Hunchback (Hb) protein gradient in Drosophila embryos. Our results define five expression boundaries that are set primarily by differences in Hb concentration and two boundaries that are set by combinatorial mechanisms involving Hb and at least one other repressor.

Conclusions: Hb functions as a repressive morphogen, but only within a specific range of concentrations ( approximately 40% to approximately 4.4% of maximum Hb concentration), within which there are at least four distinct concentration thresholds. The lower limit of the range reflects a position where the slope of the gradient becomes too shallow for resolution by specific target genes. Concentrations above the upper limit do not contribute directly to differential-repression mechanisms, but they provide a robust source that permits proper functioning of the gradient in heterozygous embryos that contain only one functional hb gene.

Figure 1

Spatial relationships between the Hb gradient and its repression targets. Wild type embryos (left panels) were stained to simultaneously detect the Hb protein gradient (red) and the mRNA expression patterns of target genes (green). Quantitative analyses of the ROI (yellow box) for each embryo are shown on the right. Peaks representing the eve stripes repressed by the anterior Hb gradient (eve 3 and eve 4) are marked by blue stars (L).

Figure 2

Temporal dynamics between refinement of the Hb gradient and target gene positioning. Wild type Hb gradients (red) and the mRNA expression patterns (green) of the gap genes kni (A-C, n=11) and Kr (D-F, n=19) are shown at two time points, early in nuclear cycle 14 (A, D) or midway through the same cycle (B, E). Quantitative profiles of these expression patterns are shown on the right (C, F). Positions where target gene mRNA patterns are repressed to a level of 50% of maximum are marked with vertical dashed black lines. Thin lines represent data from early cycle 14 embryos, while thick lines represent mid-cycle 14 embryos.

Figure 3

Target genes responses to ventral mis-expression of Hb. A-F. Hb protein staining patterns are shown for embryos containing one (A), two (C), or four (E) copies of the sna-hb transgene. Quantitative analyses of the ROIs (yellow boxes for the endogenous Hb profile, and green boxes for ventral expression domains) are shown to the right of each embryo (B, D, F). G-O. RNA expression patterns of target genes eve (green) and kni (red) (G, n=4), eve (green) and nub (red) (J, n=5) and Kr (red) (M, n=9) were examined in embryos containing 1 (G and M) or 2 copies (J) of the sna-hb transgene. Quantitative data for the endogenous expression patterns were collected from the longitudinal ROIs (yellow boxes), and are represented by the plots in H, K and N. ROIs represented by the purple boxes in G, J and M were used to quantify relative sensitivities to Hb-mediated responses, and are represented in the graphs in I, L and O. The grey triangles above I, L, and O show the orientation of the ventral gradient(s) with respect to the graphs below.

Figure 4

Threshold-dependent positioning of eve 3 and eve 4. A. Hb concentration profiles are shown for a wild type embryo (red curve) a hb/+ embryo (blue curve), along with a predicted hb/+ profile generated by dividing the expression levels at every position of the wild type profile by two (green curve). The AP positions where levels of Hb correspond to 50% repression of eve 3 and eve 4 in wild type embryos are marked by red dashed lines. The predicted positions of these boundaries in hb/+ embryos are marked with blue dashed lines. Red and blue arrows mark the distance between the anterior boundaries of eve 3 and eve 4 in wild type and hb/+ embryos respectively. B, C. eve RNA patterns in wild type and hb/+ embryos. The predicted and tested positions of eve 3 and eve 4 in hb/+ embryos are compared in D (n=9 each). Error bars represent standard deviations. E. The distance between the anterior boundaries of eve 3 and eve 4 in wild type (red) embryos is significantly shorter than in hb/+ embryos (blue) (p<0.0001). F. Data from the ROIs (yellow boxes in B and D) are plotted for comparisons of the positioning of the stripes. Red and blue plots represent data from B and C respectively. The arrows represent the distance between the measured anterior boundaries of eve 3 and eve 4 in B and C respectively.

Figure 5

Hb functions with other gap proteins to repress the anterior boundaries of gt and Kr. Wild type (yw) and homozygous mutant embryos containing sna-hb, sna-Kr, sna-gt, or combinations of two transgenes (indicated on the upper right hand corner of each panel) were stained by in situ hybridization to detect various RNA expression patterns (indicated on the lower left hand corner of each panel). The sna expression boundary provides a landmark for identifying position along the dorsal-ventral axis, and was used to position and orient the ROIs (yellow boxes). Data collected from the ROIs is shown schematically in the graphs to the right of each embryo photograph. Embryos in Q and R were stained with Kr probe only in order to better visualize the posterior expansion of the Kr domain along the ventral surface caused by mis-expression of Hb. The grey triangles above B and L show the orientation of the ventral gradient(s) with respect to all graphs below.

Figure 6

A. A morphogenetic range for the Hb repression gradient. The Hb gradient profile shown (red) is derived from a wild type embryo at mid-cycle 14. The positions of four boundaries (blue vertical lines) lie within the morphogenetic range (light blue shading), while two other boundaries (green vertical lines) lie outside this region.