A PI4KIIIα protein complex is required for cell viability during Drosophila wing development

Abstract

Phosphatidylinositol 4 phosphate (PI4P) and phosphatidylinositol 4,5 bisphosphate [PI(4,5)P₂] are enriched on the inner leaflet of the plasma membrane and proposed to be key determinants of its function. PI4P is also the biochemical precursor for the synthesis of PI(4,5)P₂ but can itself also bind to and regulate protein function. However, the independent function of PI4P at the plasma membrane in supporting cell function in metazoans during development in vivo remains unclear. We find that conserved components of a multi-protein complex composed of phosphatidylinositol 4-kinase IIIα (PI4KIIIα), TTC7 and Efr3 is required for normal vein patterning and wing development. Depletion of each of these three components of the PI4KIIIα complex in developing wing cells results in altered wing morphology. These effects are associated with an increase in apoptosis and can be rescued by expression of an inhibitor of Drosophila caspase. We find that in contrast to previous reports, PI4KIIIα depletion does not alter key outputs of hedgehog signalling in developing wing discs. Depletion of PI4KIIIα results in reduced PI4P levels at the plasma membrane of developing wing disc cells while levels of PI(4,5)P₂, the downstream metabolite of PI4P, are not altered. Thus, PI4P itself generated by the activity of the PI4KIIIα complex plays an essential role in supporting cell viability in the developing Drosophila wing disc.

Keywords: Cell death; Drosophila; Imaginal discs; Phosphoinositides; Plasma membrane; Signalling; Wing development.

Fig 1. Knockdown of dPI4KIIIα affects shape and venation patterning of adult wings

Ai, Bi, Ci) Control wings of the indicated Gal4 drivers for anterior compartment (ci-Gal4), posterior compartment (hh-Gal4) and anterior cells close to A/P compartment boundary of (ptc-Gal4), respectively. L1-L6 are the longitudinal veins; ACV and PCV are anterior and posterior cross vein respectively. Aii, Bii and Cii) ci> dPI4KIIIα, hh> dPI4KIIIαⁱ and ptc> dPI4KIIIαⁱ wings. In case of Aii and Bii the respective compartment is shortened. In Cii (ptc>dPI4KIIIα ⁱ) the inter-vein between L₃ and L₄ is shortened and the area around the ACV has disorganised extra cuticle deposition leading to a ‘smudged’ ACV. Aiii, Biii, Ciii) Rescue with the human PI4KIIIα transgene. Aiv, Biv, Civ) Ectopic expression of human PI4KIIIα in the respective domains of otherwise wild type discs. D) Quantification of the L3-L4 inter-vein distance of the wings represented in Cii, n≥20 for each genotype for quantification. All experiments were repeated thrice and one set with corresponding data numbers (n) is reported here. For the current set, Aii n=20, Bii n=8, Cii n=70. The region of genetic manipulation is demarcated by blue dotted lines. In case of C, the L3-L4 inter-vein distance shortening is pointed out with black arrow heads and the ‘smudging’ of ACV with black arrows.

Fig 2. Changes in plasma membrane PI4P levels observed upon dPI4KIIIα complex knockdown Ai and iii)

Confocal images of S2R+ cells transiently transfected with pUAST-P4M::GFP attB plasmid and reagent control respectively. Aii and iv) Corresponding DIC images of the same cells.20-25 cells were checked for expression of the probe. 48 hr post transfection live cells were imaged for GFP, n=20 cells. B) Western blot from the same set of transfected cells showing the expression of the correct sized GFP tagged P4M protein. Anti GFP antibody was used for detection of P4M::GFP. The experiment was repeated twice and one such set is reported. Ci and ii) ptc >P4M::GFP probe alone and ptc >dPI4KIIIα ⁱ, P4M::GFP probe in the third instar larval wing disc. The cartoon in the inset shows third instar wing disc; ptc-gal4 domain is marked by white line and the rectangle marks the region used for analysis. D) Quantification of fluorescence intensity in wild type and dPI4KIIIα knockdown backgrounds. 5-6 discs per genotype were used and 10-20 cells per wing disc were used for quantification. The experiment was repeated twice, and one set is reported here. Ei, ii and iii) Representative wing images of the control, dPI4KIIIα ⁱ alone and the same RNAi along with the PI4P probe P4M::GFP respectively. Note that the phenotype is not modulated by presence of the PI4P probe (compare ii and iii) n=5 for each genotype. Scale bar is 10 μm.

Fig 3. Venation defect seen upon dPI4KIIIα complex knockdown is distinct from changes in PI(4,5)P₂ levels

A) q-PCR based detection of transcripts of different PIP5Ks, sktl, and dpip5k with respect to RP49 transcripts in the wild type wing disc. The experiment was repeated twice and result of one such experiment is reported here. Bi and Di) Control wing images. Bii and Dii) Overexpression of kinase dead sktl (sktl^K/D) and sktl knockdown (sktlⁱ) respectively with the consequent venation defect marked by black arrows. C and E) Quantification of the L3-L4 inter-vein distance in case of sktl^K/D::RFP over expression and sktlⁱ mediated knockdown. For all adult wings experiment n≥10. F) Comparison of severity of shortening of L3-L4 inter-vein distance in case of different genetic manipulations of sktl and dPI4KIIIα. Gi, ii, iii) Representative images of wing discs expressing the PI(4, 5)P₂ probe PH-PLCd::mCherry in wild type, dPI4KIIIα knockdown and sktl knockdown using the ptc driver at 25°C. Inset carton is a third instar wing disc; white line is the ptc domain and rectangle marks the area from which cells were used for analysis. The scale bar represents 10 μm. H) Quantification of PH-PLCd::mCherry fluorescence intensity.

Fig 4. dTTC7 and StmA work in conjunction with dPI4KIIIα to affect venation patterning

Ai and Bi) ptc-Gal4 controls. Aii and Bii) Wings of ptc >Efr3/stmA ⁱ and ptc >dTTC7 ⁱ RNAi respectively. The inter-vein area as well as the ACV region show defects as seen in ptc>dPI4KIIIα^I wings in the same domain. Aiii) Partial rescue of the stmA ⁱ phenotype with mouse ortholog of Efr3 (mEFR3B). Biii) Rescue of the dTTC7 ⁱ phenotype with the human ortholog hTTC7B. C and D) Quantification of the shortening of the inter-vein and the partial rescue for stmA and dTTC7 respectively. Ei, ii and iii) Third instar wing discs expressing P4M::GFP in wild type, stmA ⁱ, and dTTC7 ⁱ driven by ptc. The inset cartoon shows the third instar wing disc with the white line marking the ptc domain and rectangle the area from which cells were chosen for analysis. Since there was no enrichment of the probe in plasma membrane of the ptc>dTTC7ⁱ and ptc>stmAⁱ, it was not possible to quantify it. Scale bar is 10μm.

Fig 5. Developmental timing of the venation defect

Ai-iv and Bi-iv) represent ptc>mCD8::GFP to denote the region of RNAi manipulation. Ai’-iv’) Discs stained with Blistered antibody marking all the inter-vein cells and making the pro-veins visible by negative staining (n=10 for each genotype and the experiment was repeated thrice with the figure representing data from one such trial). Note there is no detectable difference between control (Ai’) and the RNAi manipulated (Aii’-iv’) discs. The region of interest i.e., the presumptive L3-L4 inter-vein is marked by white dotted lines. Bi’-iv’) Blistered staining in indicated genotype at 6 hrs after pupae formation (APF). Developmental time point is mentioned below the figure panel and genotype of each wing is mentioned within each panel. Bii’-iv’ wings have smaller L3-L4 distance (marked by white arrowhead) as compared to Bi’ (control). Ai”-iv” and Bi”-iv”) Merged images. C, D, E) Quantifications of the L3-L4 inter-vein distance of dPI4KIIIα, dTTC7 and stmA RNAi at 6 hrs APF (The experiment was repeated thrice and n=10 for each genotype). Scale bar is 50μm.

Fig 6. Depletion of dPI4KIIIα during development does not affect cell division

A, B) Anti phosphohistone 3 (PH3) staining for third instar and 6 hr APF wing discs respectively. Aii and Bii) ptc>dPI4KIIIα ⁱ Ai and Bi) Corresponding controls. The experiment was repeated twice n=8 for control and n=10 for RNAi genotype. C i, ii and Di, ii) Anti PH3 stained third instar larval wing discs driving dPI4KIIIα depletion in the hh and ci domains respectively. Ci and Di are the controls and Cii and Dii are ci>dPI4KIIIα ⁱ and hh>dPI4KIIIα ⁱ depleted discs, respectively. The yellow line demarcates the anterior and posterior compartments. The boundaries of wing discs are denoted by white dotted line. Posterior is to the left. Scale bar is 50μm. E Mitotic Index of ci and ci>dPI4KIIIα ^I 3^rd instar wing discs is represented. Y-axis shows the mitotic index calculated as described in methods. Genotypes represented on X-axis. Each point represents the index as calculated from a single disc, error bars indicate standard error of mean (SEM). n = 5 discs per genotype.

Fig. 7. Depletion of the components of dPI4KIIIα complex causes cell death by apoptosis

All the wing discs are stained with apoptosis marker Dcp-1 (Drosophila Caspase 1). Ai-Fi) Regions showing dPI4KIIIα, stmA, and dTTC7 RNAi manipulation by corresponding Gal4 drivers. A-D) Pupal wing discs (5.5-6 hrs APF). Ai’-Fi’) Corresponding controls. Aii’) ptc>dPI4KIIIα ⁱ, Aiii’) ptc>dPI4KIIIα ⁱ; hPI4KIIIα pupal wing discs (5.5-6 hrs APF). Aii’ wing discs are Dcp-1 positive while in the rescue genotype of Aiii there is no Dcp-1 staining. Bi) Control, Bii’ and iii’) stmA and dTTC7 RNAi in the ptc domain. Bii’ and iii’ stain positive for Dcp-1. Ci’) Control, Cii’) ci>dPI4KIIIα ⁱ. Di’) Control, Dii’) hh>dPI4KIIIα ⁱ.E-F) Third instar larval wing discs stained with antibody against Dcp-1. Ei’) Control, Eii’ and iii’) ap>dPI4KIIIα ⁱ and ap>dTTC7ⁱ. Crosses in experiment E were carried out at 29°C for maximal knockdown o the genes concerned. Fi’) Control, Fii’) ap>stmAⁱ. Crosses in experiment F were carried out at 25°C as 29°C did not yield viable larvae. Scale bar is 50 μm.

Fig 8. Depletion of dPI4KIIIα complex shows no effect on the readouts of hedgehog signalling in larval stages

Ai-ii and Ci-ii) Third instar wing disc images showing expression of ptc>mCD8::GFP. Ai’-ii’ and Ci’-ii’) Same discs stained for Ci and Ptc in control and ptc>dPI4KIIIα ⁱ, respectively. All discs are positioned with anterior to the right and posterior to the left. Ei and ii) Third instar wing disc of control and ptc>dPI4KIIIα ⁱ, respectively stained for lacZ antibody to show the expression pattern of Dpp-lacZ. Three repeats for each staining was done and n=7-10 for each genotype. B, D, F) Quantification of spread of indicated antigens, with6-7 discs used per experiment. The quantification was repeated thrice with three different biological sets. G and H) Data with ap Gal4 driver. Gi-ii and Hi-ii) Third instar wing disc images showing expression of ap>mCD8::GFP to demarcate region of RNAi manipulation. Gi’-ii’ and iii’) Same discs stained for the Ptc in control, ap>dPI4KIIIα ⁱ and apt>dTTC7ⁱ respectively. Hi’-ii’) Same discs stained for the Ptc in control and apt>stmA ⁱ respectively. All experiments were repeated thrice. For the graphs B, D and F the regions of the wing disc chosen and the line profiles for intensity measurements are depicted by cartoons in corresponding panels. White dotted lines mark the wing disc boundaries. Scale bar is 50 μm.