Fig 1.
R8 selection and MF specificity of the CK2 mimic M8-S159D.
(A) Cell fate acquisition at different stages of the MF. Ato expression is upregulated in response to proneural enhancement (PE), which is followed by lateral inhibition (LI) through the E(spl) repressors. Color codes of cell fates is shown in inset, and expression domains of Gal4 drivers used in the studies are indicated relative to stages of the MF; the vertical dashed line denotes sequential recruitment of secondary photoreceptors. (B) The reduced eye phenotype of CK2 variants (random insertions) or M8*, the product of the E(Spl)D allele, upon expression at stage-1 (hH10Gal4) and at stage-2/3 (scaGal4 and 109-68Gal4). The reduced eye is denoted by number of ommatidia (facets) remaining; WT denotes facet counts between 750–800 and no perturbation of the hexagonal architecture of the adult eye. Note that the reduced eye of M8-S159D only manifests at stage-2/3, whereas that of M8* occurs at stage-1. Schematic to the right of panel B depicts the domains of E(spl)-M8 and CtD deletion in M8*, and sequence of the P-domain highlighting the CK2 site (S159DCD) that was altered to generate the CK2 mimic M8-S159D. The domains are indicated and include a C-terminal WRPW motif (Gro-binding). (C, D) Scanning EM of the adult eye at 200x. Overexpression of the CK2 mimic M8-S159D elicits a reduced eye at stage-2/3 of the MF (C), but not at stage-1 (D). (E) For each genotype, the number of images analyzed for eye size (ommatidial/facet counts) is indicated. Genotypes shown in panels C and D were compared to controls, and ** denotes P-value < 0.001.
Fig 2.
Conservation of Ser residues in the P-domain of E(spl)-M8.
(A) Alignment of the P-domain of E(spl)-M8 from the indicated Drosophila species, stalk-eyed flies Teleopsis Dalmanii, and mammalian HES6. The evolutionary tree depicts the divergence time of Drosophila species; MYR denotes 106-years. The blue and yellow shaded boxes denote the consensus for MAPK and CK2, respectively (shown at top of alignment). Conserved Ser residues are in bold red. Note that mammalian HES6 harbors six additional residues in the linker separating the MAPK and CK2 sites, and this region harbors the Asp-Asp (DD) motif in the CtD (grey box and black arrow). (B) Transgenic lines for CK2 and MAPK variants were generated using ΦC31 and inserted at site 68E. The nomenclature and predicted behavior of the M8 variants are indicated. Inset on right shows yeast two-hybrid interactions with Groucho; “+” denotes productive interaction.
Fig 3.
An Ala residue at the MAPK site neutralizes the reduced eye of M8-S159D, but permits bristle suppression.
(A) Scanning EM of the adult eye at 200x magnification. The CK2 mimic with an Ala at the MAPK site (M8-S151A+S159D) fails to elicit a reduced eye upon expression at stage-2/3 of the MF. The facet counts (inset) are equivalent to those in wild type flies (not shown). Note that M8-S151A+S159D elicits loss of many IOBs (dashed oval). (B, C) Nota of flies of the indicated genotypes exhibiting loss of MCs. Inset in panels B and C denote MCs remaining per heminotum. (D) The MAPK mimic (M8-S151D) and the dual-kinase mimic (M8-S151D+S159D) elicit embryonic lethality (EL) upon expression with scaGal4 at either 24°C or 18°C, reflecting expression in most PNCs.
Fig 4.
M8 variants with an Asp at the MAPK site elicit a reduced eye.
(A-D) Scanning EM of the adult eye at 200x magnification. M8 variants were expressed with 109-68Gal4 at stage-2/3 of the MF. Unlike scaGal4 (Fig 3), reduced expressivity (strength) of 109-68Gal4 in PNCs prevents embryonic lethality of the MAPK variants, and diminishes the reduced eye of the CK2 mimic M8-S159D (A). In contrast the MAPK-refractory protein M8-S151A+S159D elicits no reduced eye (D), whereas the MAPK-mimic M8-S159D (B) and the dual-kinase mimic M8-S151D+S159D (C) elicit a reduced eye. (E) For each genotype, the number of images analyzed for eye size (ommatidial/facet counts) is indicated. Genotypes shown in panels A-D were compared to control (109-68Gal4/+), and ** denotes P-value < 0.001. (A’-D’) Nota of flies expressing M8 variants indicated above panels A-D with 109-68Gal4. All M8 variants elicit MC loss with almost equal severity (F). For each genotype, MCs (per heminotum) were counted in the indicated number of flies. Genotypes shown in panels A’-D’ were compared to control (109-68Gal4/+), and ** denotes P-value < 0.001. Yellow inset indicates that the parental M8-lines had a wild-type eye, IOBs and MCs.
Fig 5.
M8 variants with Asp at the CK2 and MAPK sites perturb R8 birth.
Eye discs were stained with α-Sens and α-ELAV to label differentiated R8s and secondary photoreceptors, respectively. Arrows indicate direction of MF, dotted circles denote missing R8s and secondary photoreceptors, and arrowheads denote recruitment defective R8s. (A) Discs expressing the CK2-mimic M8-S159D show sporadic gaps in Sens+ELAV clusters, whereas discs expressing the MAPK-refractory variant M8-S151A+S159D (D) closely resemble 109-68Gal4/+ (E) or WT disc (data not shown). Both, the MAPK mimic M8-S151D (B) and dual kinase (CK2+MAPK) mimic M8-S151D+S159D (C) exhibit areas lacking Sens+ELAV clusters with the greatest severity. (A’-D’) magnified view of panels A-D, and (A”-D”) Sens channel in greyscale.
Fig 6.
An Asp at the MAPK site engenders M8 activity at stage-1 of the MF.
(A-C) Scanning EM of adult eye at 200x magnification. The MAPK mimic M8-S151D (A) and the dual-kinase mimic M8-S151D+S159D (B) elicit a reduced eye, whereas a WT eye is evidenced upon expression of M8-S151A+S159D, a variant refractory to MAPK (C). (A’-C’) 1000x magnification of the eyes in panels A-C. Note that M8 variants with an Asp at the MAPK site elicit the specification of extra IOBs (A’, B’), whereas such defects are infrequently seen upon expression of M8-S151A+S159D (C’). (D) For each genotype, the number of images analyzed for eye size (ommatidial/facet counts) is indicated. Genotypes shown in panels A-C were compared to control (hH10Gal4/+), and ** denotes P-value < 0.001.
Fig 7.
Halved egfr dosage mitigates the reduced eye of the CK2 mimic, but not the MAPK mimics.
(A-B) Scanning EM of adult eye at 200x magnification. The reduced eye of the CK2 mimic M8-S159D at stage-2/3 (A) is significantly attenuated (rescued) in the egfrf24/+ background (B). (C) Graph showing eye size; histograms labeled A and B correspond to the eyes in panels A and B. For each genotype, the number of images analyzed for eye size (ommatidial/facet counts) is indicated. Values for genotype shown in panel B were compared to that in panel A (the relevant control) and ** denote P-value < 0.001. Eye size in scaGal4/+; egfrf24/+ flies was not the control and included to illustrate a WT-eye in this genetic background. (A’-B’) Eye discs of genotypes in A and B were stained to label differentiated R8s (α-Sens) and secondary photoreceptors (α-ELAV). Arrows indicate direction of MF, dotted circles denote missing R8s and secondary photoreceptors, and arrowheads denote recruitment defective R8s. Panel (B”) shows a magnified image of (dashed) yellow box in B’ to highlight rescue of Sens+Elav clusters, although a few R8s fail to recruit Elav+ cells. In contrast, the reduced eye of the MAPK mimic M8-S151D (D) and the dual kinase (CK2+MAPK) mimic M8-S151D+S159D (E) are not rescued in the egfrf24/+ background. (F) For each genotype, the number of images analyzed for eye size (ommatidial/facet counts) is indicated. Genotypes shown in panels D and E were compared to their corresponding controls, i.e., values from Fig 4B and 4C, respectively, (‘n.s.’ denotes not significant). Eye size in 109-68Gal4/+; egfrf24/+ flies was not the control and included to illustrate a WT-eye in this genetic background.
Fig 8.
Deletion of the CK2 or MAPK sites in M8 elicits a reduced eye.
(A) Deletion variants of the P-domain of M8. M8-ΔCK2 lacks the SDCD motif and M8-ΔMAPK lacks the PLSP motif, but are otherwise full-length. Inset in panel A (shaded grey) shows effects of the two deletions upon expression with scaGal4 or 109-68Gal4; note embryonic lethality of M8-ΔMAPK upon expression with scaGal4, as also seen with M8-S151D, the Asp variant of the MAPK-site (Fig 3D). Inset (shaded yellow) shows that in the absence of expression M8-ΔCK2 and M8-ΔMAPK transgenic lines have WT eyes (>750 facets with normal hexagonal architecture). Stage-1 expression of M8-ΔCK2 fails to perturb the eye (B) whereas M8-ΔMAPK elicits a reduced eye phenotype (C). (D) Quantitative analysis of eyes in panels B and C, and the relevant control (hH10Gal4/+). For each genotype, the number of images analyzed for eye size (ommatidial/facet counts) is indicated. Genotypes shown in panels B and C were compared to controls (hH10Gal4/+) and ** denotes P-value < 0.001. Stage-2/3 expression of M8-ΔCK2 elicits a reduced eye with greater severity in egfr+ (E) compared to egfrf24/+ (F) backgrounds. (G) Quantitative analysis of eyes; values from genotype shown in panel E were compared to that in panel F, and ** denotes P-value < 0.001. In contrast, stage-2/3 expression of M8-ΔMAPK elicits a reduced eye with equal severity in egfr+ (H) or egfrf24/+ (I) backgrounds. Note that lethality of scaGal4/+; UAS- M8-ΔMAPK/+ flies (grey inset in panel A), necessitated expression with 109-68Gal4. (J) Quantitative analysis of eyes; values from genotype shown in panel H were compared to that in panel I (n.s. denotes not significant).Note that grey bars in panels G and J show that the Gal4 driver alone or in combination with egfrf24 (shown in yellow inset above graph) has a wild type eye (>750 facets with normal hexagonal architecture) and was not used as a control for statistical comparisons.
Fig 9.
MAPK site in M8 may be a target for the phosphatase PP2A.
(A) Co-expression of widerborst (wdb) rescues the reduced eye of the CK2 mimic M8-S159D, whereas that of LacZ elicits no rescue, as previously described [50]. Ommatidial (facet) counts were determined in 20 flies of the indicated genotypes. Values for genotypes upon co-expression of Wdb or LacZ were compared to the corresponding control (scaGal4/+; UAS-M8-S159D/+) flies (** denotes P-value < 0.001). In contrast, co-expression of Wdb does not rescue the reduced eye of the MAPK mimic M8-S151D (B) or the dual kinase (CK2+MAPK) mimic M8-S151D+S159D (C). (D) Quantitative analysis of eyes in panels B and C. For each genotype, the number of images analyzed for eye size (ommatidial/facet counts) is indicated. Genotypes shown in panels B and C were compared to their corresponding controls, i.e., values from Fig 4B and 4C, respectively (‘n.s.’ denotes not significant). Note that grey bar in panel D shows that the Gal4 driver alone or in combination with UAS-Wdb (shown in yellow inset above graph) has a wild type eye (>750 facets with normal hexagonal architecture) and was not used as a control for statistical comparisons.
Fig 10.
Regulation of M8 activity by reversible phosphorylation.
(A) In the ‘cis’-inhibited state, the CtD interacts with the HLH and/or Orange domains thereby blocking M8 repression of Ato. Phosphorylation by CK2 and MAPK serves as a conformational switch that converts M8 into an Ato repressor, whereas PP2A mediates inactivation. (B) Model for spatial and/or temporal regulation of M8 activity at stage-2/3 of the MF. EGFR/MAPK signaling ensures that M8 activation does not occur until stage-2/3 of the MF, thereby allowing ato-auto-activation to raise Ato levels to a threshold sufficient for the R8 fate. In this case, the phosphatase PP2A targets M8 to control either the ‘amplitude’ or ‘duration’ of active M8.