aPKC mutant embryos derived from germ-line clones show loss of epithelial integrity in the epidermis. (A) Cuticle preparation of a wild-type embryo showing a contiguous cuticle with ventral denticle bands (arrowheads) and the head skeleton (arrow). (B-F) Cuticles of aPKC mutant embryos derived from germ-line clones show only crumbs of cuticle (B,C,F) or rudimentary patches of contiguous cuticle (D,E; between white lines), indicative of a breakdown of epithelial tissue structure. The phenotypes of the aPKC^psu69 (B) and aPKC^psu141 (C) class I alleles are indistinguishable from the null allele aPKC^k06403 (F), whereas the phenotypes of the aPKC^psu265 (D) and aPKC^psu417 (E) class II alleles are much milder. Arrowheads in D point to rudimentary denticle bands. The cuticle phenotype of par-6^Δ226 null mutant embryos derived from germ-line clones (G) resembles the aPKC null mutant phenotype. Anterior is to the left in all panels. Scale bar:100 μm.

Apical-basal polarity of the ectodermal epithelium is disrupted in aPKC and par-6 mutant embryos derived from germ-line clones. The ectodermal epithelium of wild-type and mutant embryos at stage 7 was stained for aPKC (A-E), PAR-6 (F-J), Baz (K-O), Nrt, Crb and Dlg (P-T′). Scale bars: 10 μm (A-O); 10 μm (P-T). (U) Ultrastructure of the wild-type embryonic epidermis at stage 11. There is a contiguous electron dense ZA (U′, red arrows) in the apical portion of the lateral membrane. By contrast, aPKC^psu141 mutant embryos at the same stage do not possess a contiguous ZA (V,V′). Instead, spot adherens junctions are found in ectopic positions along the lateral membrane (V″, arrow). Scale bars: 0.5 μm (U-V″). In all images apical is to the top.

Neuroblast polarity in aPKC and par-6 mutant embryos derived from germ-line clones. Wild-type and mutant NBs at metaphase were stained for aPKC (red; A-G), PAR-6 (red; H-N), Baz (red) and Mira (blue; O-U″). DNA was stained with YOYO-1 (green). Note that spindle orientation deduced from the orientation of the metaphase plate and the position of the Baz crescent was occasionally abnormal (S,T). Scale bar: 5 μm. In all panels, apical is up.

aPKC mutant phenotypes in oogenesis. (A) In wild-type egg chambers, the follicular epithelium has a regular, monolayered structure and PAR-6 is localized to the apical cortex of the follicle cells. (B) Egg chambers of homozygous aPKC^psu69 mutant females show both multilayering and large gaps (arrowheads) in the follicle epithelium. Note that PAR-6 is still localized apically in follicle cells that are in direct contact with the germ-line cells. Scale bar: 10 μm (A,B). (C-E′) The phenotype of the hypomorphic aPKC alleles in homozygous mutant follicle cell clones is much milder than that of the aPKC null allele. Follicle cell clones of aPKC^psu417 (C,C′) and of aPKC^psu69 (D,D′) show a normal monolayered tissue structure and correct localization of the mutant aPKC (red) and PAR-6 (blue) to the apical cortex. Occasionally, gaps are observed in the follicle epithelium of egg chambers carrying clones of aPKC mutant cells (D,D′,D″, arrowheads). In contrast to the hypomorphic aPKC alleles, follicle cell clones of the null allele aPKC^k06403 (E,E′,E″) show multilayering of the epithelium and a complete loss of the apical localization of Bazooka (blue). Clones of homozygous mutant follicle cells are marked by the absence of GFP (green). Scale bar: 10 μm (C-E″). (F-H) Oocyte determination occurs normally in aPKC^psu417 (F) and aPKC^psu265 (G) germ-line clones, but not in the the null allele aPKC^k06403 germ-line clone (H), which shows loss of the oocyte in most egg chambers. Oocytes are marked by staining for Orb (red) and homozygous mutant cells are marked by absence of GFP (green, arrows). Scale bar: 10 μm (F-H). (I-K) Anterior-posterior polarity of the oocyte is normal in aPKC^psu69 and aPKC^psu417 germ-line clones. In wild-type oocytes (I), Gurken (red) is localized to the anterior dorsal cortex and Staufen (blue) to the posterior cortex of the oocyte. The localization of Gurken and Staufen in germ-line clones of aPKC^psu69 (J) and aPKC^psu417 (K) is indistinguishable from that in wild-type. Germ-line clones are marked by absence of GFP staining (insets) in nurse cells and oocytes. Scale bar: 50 μm (I-K). Anterior is to the left in all panels.

Biochemical properties of mutant aPKC proteins. (A) Structure of the aPKC protein. The position of the mutations in the four new aPKC alleles is indicated by arrowheads. (B) The GFP-aPKC^psu69 protein does not bind to PAR-6. S2 cells were transfected with wild-type GFP-aPKC and with the four mutant versions of GFP-aPKC. Untransfected S2 cells were used as negative control. The cell lysates were subjected to immunoprecipitation (IP) followed by western blotting (Blot) with the indicated antibodies. Preimmune serum (pre; for PAR-6 IPs) or anti-β-galactosidase antibodies (β-Gal; for GFP IPs) were used as negative controls. (C) The mutant aPKC^psu141 protein is not recognized by the phospho-specific antibody directed against the phosphorylated threonine residue T422. Lysates from S2 cells transfected with GFP-aPKC or with GFPaPKC^psu141 were subjected to western blotting with either an antibody that recognizes aPKC irrespective of its phosphorylation state (aPKC) or with an antibody that specifically recognizes aPKC phosphorylated at T422 (aPKCpT422). A T422 phospho-specific band corresponding to GFPaPKC, but not to GFPaPKC^psu141 was detected. Untransfected S2 cells were used as negative controls. The endogenous aPKC band served as loading control. This observation was confirmed in western blots of adult head extracts of wild-type and aPKC^psu141 homozygous mutant flies (right). (D) Three of the four mutant aPKC proteins show strongly reduced kinase activity. In vitro kinase assays using [γ-³²P]ATP were performed on anti-GFP immunoprecipitates of lysates of S2 cells transfected with either wild-type GFP-aPKC, with any of the four mutant GFP-aPKC versions and with GFP-tagged kinase-dead aPKC (GFP-aPKC^K293A). Untransfected cells and cells transfected with wild-type GFP-aPKC but immunoprecipitated with anti-HA antibody were used as negative controls. Both autophosphorylation of GFP-aPKC and phosphorylation of a GST-Baz fusion protein containing the aPKC target site S980 were measured by autoradiography. One representative autoradiogram is shown. The quantitative analysis of four independent in vitro kinase assays is shown in the bar diagrams below the autoradiographs. Values are percentage phosphorylation relative to that of the wild-type GFP-aPKC protein. The input for the kinase assay (top) is shown at the bottom. (E) Three of the four mutant aPKC proteins are kinase-dead with respect to phosphorylation of S980 of Baz. In vitro kinase assays were performed with non-radioactive ATP as phosphate donor and phosphorylation of the GST-Baz fusion protein was detected in western blots using a phospho-specific antibody against S980 of Baz. The immunoprecipitated GFP-aPKC from the same experiment was also detected by western blotting. Quantification of the blots with the phospho-specific Baz antibody from four independent experiments and the GST-Baz input corresponding to the kinase assay (top) are shown at the bottom.

Structure of human PKCι and thermodynamic cycles. (A) Overall structure of human PKCι with modeled structure elements (blue) and mutated residues (red). The enlarged view (right) compares the ATP binding pocket of the wild-type structure (green) with that of the A272V mutant (dark blue). (B) Thermodynamic cycles used in the free energy calculations described in the text. WT and Mut. refer to the wild-type and each of the considered mutant structures, respectively; the indices (f) and (u) refer to the folded and unfolded states, respectively. The color codings of the free energy double differences (ΔΔG_bind and ΔΔG_fold) correspond to the respective mutated amino acids in A.