Genetic dissection of the phosphoinositide cycle in Drosophila photoreceptors

Abstract

Phototransduction in Drosophila is mediated by phospholipase C-dependent hydrolysis of PIP_2-, and is an important model for phosphoinositide signalling. Although generally assumed to operate by generic machinery conserved from yeast to mammals, some key elements of the phosphoinositide cycle have yet to be identified in Drosophila photoreceptors. Here, we used transgenic flies expressing fluorescently tagged probes (P4M and Tb^R332H), which allow in vivo quantitative measurements of PI4P and PIP₂ dynamics in photoreceptors of intact living flies. Using mutants and RNA interference for candidate genes potentially involved in phosphoinositide turnover, we identified Drosophila PI4KIIIα (CG10260) as the PI4-kinase responsible for PI4P synthesis in the photoreceptor membrane. Our results also indicate that PI4KIIIα activity requires rbo (the Drosophila orthologue of Efr3) and CG8325 (orthologue of YPP1), both of which are implicated as scaffolding proteins necessary for PI4KIIIα activity in yeast and mammals. However, our evidence indicates that the recently reported central role of dPIP5K59B (CG3682) in PIP₂ synthesis in the rhabdomeres should be re-evaluated; although PIP₂ resynthesis was suppressed by RNAi directed against dPIP5K59B, little or no defect was detected in a reportedly null mutant (dPIP5K¹⁸ ).

Keywords: PI4-kinase; PIP2; PIP5-kinase; Phototransduction; TTC7.

Fig. 1.

PIP₂ and PI4P recovery time courses in control flies. (A) Tb^R332H YFP fluorescence in response to a 30 s blue excitation light measured from rhabdomere patterns in the DPP of otherwise wild-type fly. In a fully dark-adapted fly (red trace) fluorescence decays over ∼20 s as PIP₂ is depleted and the Tb^R332H probe translocates out of the rhabdomere. After M to R reconversion by red (R) light and variable periods in the dark: (Δt=10-200 s), the blue excitation was repeated and the return of the probe to the rhabdomeres (reflecting PIP₂ resynthesis) monitored from the instantaneous fluorescence (arrows). (B) Similar protocol from a fly expressing one copy of GMR-Gal4 and UAS-wRNAi (GMRw control for RNAi experiments). (C,D) Similar traces from flies expressing the PI4P-specific P4M-GFP probe on wild-type (C) and GMRw (D) backgrounds. (E,F) Normalised time course of recovery from traces as in A-D after varying times in the dark. (E) PIP₂ monitored with Tb^R332H; mean±s.e.m.; wt, n=17; GMRw, n=27. (F) PI4P monitored with P4M (n=9). (G,H) Time for 50% recovery (t_½) for Tb^R332H (G) and P4M (H) probes (from time course data as in E and F). For both PIP₂ (Tb^R332H) and PI4P (P4M). There were distinct effects on kinetics of depletion and recovery attributable to GMR-Gal4 expression, emphasising the need for GMR-Gal4 controls for UAS-RNAi experiments. (I) The canonical phosphoinositide cycle with identified and candidate genes indicated.

Fig. 2.

PIP₂ and PI4P recovery time courses in RNAi flies. (A,C) Time courses of Tb^R332H (A) and P4M (C) recovery in flies expressing RNAi constructs for various candidate genes under control of GMR-Gal4 (mean±s.e.m.). (B,D) Time to 50% recovery (t_½) for resynthesis of both PI4P (P4M) and PIP₂ (Tb^R332H) were substantially and significantly slowed in flies expressing UAS-RNAi constructs directed against PI4KIIIα, rbo, YPP1, dPIP5K, dPIS and cds (P<0.0001, one-way ANOVA, Dunnett's multiple comparison test). P4M data for cds^KK not shown as there was no detectable recovery. RNAi directed at other candidate PI4-kinases (fwd and PI4KIIα) or either fly homologue of TMEM150A (CG4025 and CG7790 data pooled) had little or no effect. Flies were progeny of crosses between Tb^R332H;GMRw or P4M;GMRw and the respective VDRC RNAi lines or the progenitor control (KK).

Fig. 3.

ERG recordings. (A) ERG responses to 1 s flashes of increasing intensity in GMRw×UAS-PI4KIIIα-RNAi flies (n=5) and GMRw controls (n=8). (B) Resulting response intensity functions (mean±s.e.m.). Maximum intensity (10⁰) was equivalent to ∼10⁷ effectively absorbed photons/s. (C) ERG from PI4KIIIα-RNAi fly exposed to PIP₂-depleting stimulus (30 s saturating blue excitation followed by 5 s red light to photoreisomerise M to R). Repeated brief (250 ms) dim red test flashes monitored loss and recovery of sensitivity. Inset shows similar protocol in a GMRw/+ control fly. (D) Normalised time course of recovery following PIP₂-depleting stimuli from PI4KIIIα-RNAi flies and also from YPP1-RNAi, cds^KK-RNAi, rbo-RNAi, rbo^ts at 37°C (n=4-10 flies as indicated) as well as rbo^ts at 22°C, GMRw and RNAi parent controls (n=6 each).

Fig. 4.

PIP₂ and PI4P resynthesis in rbo^ts mutants is blocked at 37°C. (A-C) Representative fluorescence traces from wild type (A) and rbo^ts mutants expressing Tb^R332H to monitor PIP₂ (B) and P4M to monitor PI4P (C). Red traces are from initial dark-adapted state and the remaining traces after different times in the dark following depletion (5-200 s as indicated). Top series of traces at room temperature (22°C), bottom traces after warming to 37°C for 3 min. In wild type (A), both depletion and recovery were markedly accelerated at 37°C whereas in rbo^ts (B), Tb^R332H depletion was similarly accelerated at 37°C (red traces), but the apparent partial recovery now showed an increase during blue excitation rather than decay. (C) Most P4M fluorescence was lost during the 3 min warming period in the dark, and thereafter no recovery could be detected. The lower traces (A-C measured at 37°C) have been corrected (i.e. increased) for the 20% reduction in GFP fluorescence at 37°C. (D,E) Averaged recovery time courses for Tb^R332H and P4M from data as in A-C normalised to peak fluorescence. Graphs show the mean±s.e.m. from n=8-14 flies per plot. (F) Time to 50% recovery (t_½) for Tb^R332H (PIP₂) and P4M (PI4P) from plots for each fly. Note acceleration of recovery of both in wild-type at 37°C and slower recovery for P4 M in rbo^ts compared with wild type at the permissive temperature (22°C). t_½ data for rbo^ts at 37°C not shown because no flies recovered sufficient fluorescence.

Fig. 5.

In vivo PIP₂ dynamics are barely affected in dPIP5K mutant or overexpressing eyes. (A) Normalised PIP₂ resynthesis time courses measured with Tb^R332H probe from mosaic dPIP5K¹⁸ mutant eyes (mean±s.e.m.; n=23 flies), compared with heterozygote siblings (n=23) and wild-type controls recorded on same days (n=13). (B) Recovery time course from flies overexpressing dPIP5K (oe) (driven by Rh1Gal4 n=10) compared with sibling controls (non-Rh1Gal4 F1 from same cross) or Rh1Gal4;Tb^R332H parent controls pooled (n=7). (C) Summary of time to 50% recovery (t_½) of PIP₂ (i.e. Tb^R332H-YFP fluorescence) in dPIP5K¹⁸ mosaics and overexpressing flies. On average, PIP₂ resynthesis time courses in dPIP5K¹⁸ mosaic eyes were slightly slower than in controls, but data showed considerable overlap, reaching statistical significance only with respect to wild type, but not sibling heterozygote controls. ns, not significant.

Fig. 6.

ERG recordings from dPIP5K¹⁸ mosaics. (A,B) Representative electroretinogram (ERG) responses to 1 s flashes (indicated by bars) of increasing intensity from dPIP5K¹⁸ mosaic eyes and dPIP5K¹⁸/+ sibling controls from the same cross. (C) V/log I function (ERG amplitudes at end of 1 s flash; mean±s.e.m.; n=12); amplitudes were slightly reduced and sensitivity (intensity required to elicit 50% V_max response) ∼3-fold reduced in mosaics, but the most conspicuous phenotype was the lack of ‘on’ and ‘off’ transients, indicating that synaptic transmission was blocked. (D) ERG V/log I functions from hdc mutants, which also lack synaptic transmission (data replotted from Dau et al., 2016), showed a similar reduction in amplitude and sensitivity compared with wild-type controls; however, this difference can be attributed to the lack of synaptic feedback to the photoreceptors. Maximum intensity (10⁰) was equivalent to ∼10⁷ effective photons/s.

Fig. 7.

Whole-cell recordings from photoreceptors from dPIP5K¹⁸ mosaic eyes. Whole-cell recordings from dissociated ommatidia from dPIP5K¹⁸ mosaic eyes (blue) and controls (dPIP5K¹⁸/+ and wild type, recorded over the same time period pooled). (A) Responses to 1 ms flashes (arrow) containing ∼30 effective photons (means of responses from 10 flies) were virtually identical. (B) Peak amplitudes (P=0.73, two-tailed unpaired t-test) and time-to-peak (P=0.28) of responses were statistically indistinguishable. Red symbols are data from rare homozygote ‘escapers’. (C) Averaged quantum bumps (each is the average of 200-250 bumps from 4-5 cells, aligned by rising phase) in dPIP5K¹⁸ and control were again nearly identical. (D) Quantum efficiency and bump amplitudes (each point from a different cell) were statistically indistinguishable (P=0.47 and P=0.54, respectively). Red symbols are data from homozygote escapers. (E) Responses to 1 s flashes of light of increasing intensity. Mean±s.e.m. plotted in F for peak (above) and plateau (below, last 200 ms of response) were indistinguishable. dPIP5K¹⁸, n=5; control, n=7 cells.