Fig 1.
dZDHHC8 knockdown causes tissue overgrowth due to increased cell proliferation.
(A) Three independent RNAi lines targeting different regions of dZDHHC8 mRNA result in tissue overgrowth and downward bending of the wing when expressed in the dorsal compartment using apterous-Gal4 (apG4) at 25°C. Phenotype penetrance = 100% (26/26 for RNAi #1, 23/23 for RNAi #2, 32/32 for RNAi #3). (B) Expression of dZDHHC8 RNAi in the posterior compartment of the wing using engrailed-Gal4 (enG4>dZDHHC8 RNAi #1) increases posterior compartment size normalized to anterior when compared to control wings (enG4>+). Representative examples are provided on the right. Phenotype penetrance = 90% (1 of 10 RNAi wings had a P/A ratio smaller than the largest P/A ratio of control wings.) Error bars = stdev. n≥9. * ttest = 2x10-5. (C) The size of wing cells is not altered upon dZDHHC8 knockdown suggesting that tissue overgrowth is caused by enhanced cell proliferation. dZDHHC8 is knocked down in the posterior compartment with engrailed-Gal4 (enG4>dZDHHC8 RNAi #1). Cell size was determined by counting the number of cells (via trichomes) within a wing area of defined size. Error bars = stdev. n≥9. (D) Ubiquitous expression of dZDHHC8 RNAi using tubulin-Gal4 (tubG4>GFP, dZDHHC8 RNAi #2) often results in formation of extra vein material. Frequency of phenotypes in adult males of indicated genotypes are shown below representative images.
Fig 2.
(A) Schematic of the dZDHHC8 genomic locus, indicating transcript isoforms and knockout regions. Two independent dZDHHC8 knockout lines were generated: knockout 1 (KO1) which removes the majority of dZDHHC8 as well as CG34450, and knockout 2 (KO2) which removes a small region common to all dZDHHC8 isoforms including the catalytic domain. (B) dZDHHC8 knockout larvae do not have dZDHHC8 protein. dZDHHC8 protein levels were analysed by immunoblotting extracts from control (CT) or dZDHHC8 knockout (KO1) larvae, separated on a 6% gel with a dZDHHC8 specific antibody. dZDHHC8 appears as multiple high molecular weight bands. Arrows indicate dZDHHC8 protein absent in KO1. (C) dZDHHC8 mutants are larval lethal. Control (CT) and dZDHHC8 knockouts (KO1) were collected as L1 larvae and the number and developmental stage (judged by mouth hook morphology) of viable animals was determined after 1–11 days. KO1 animals start dying at the L2/L3 transition with very few animals reaching pupal stage. Some lethality of control animals is observed in this experiment, due to all animals being picked out of the food and sorted daily for scoring. Error bars = stdev. n = 3x60 L1/genotype. (D) dZDHHC8 phenotypes appear at the L2/L3 transition. Shown are control (CT) and dZDHHC8 mutant (KO1) larvae imaged at 60h after egg laying (AEL) (corresponds to wildtype larval stage 2) or 84h AEL (corresponds to wildtype larval stage 3). The developmental stage of the depicted larvae, judge by the mouth hook morphology is indicated. KO1 larvae are limited in their capability to molt from L2 to L3, either growing bigger for a few days as L2, or molting but never reaching the size of control L3 larvae. (E) dZDHHC8 mutants weigh less than controls. dZDHHC8 knockout larvae (KO1) are lighter than control (CT) larvae with the difference starting at the L2/L3 transition. Error bars = stdev. n = 3x20 larvae/genotype. (F) dZDHHC8 mutants are lean. Shown are total body triglyceride levels, normalized to total body protein, for control (CT) and dZDHHC8 knockout (KO1) L3 larvae. Error bars = stdev. n = 3x6 larvae/genotype. * ttest = 0.04.
Fig 3.
(A) Antibody against dZDHHC8 specifically detects dZDHHC8 in immunofluorescent stainings of Drosophila S2 cells treated with control (luciferase) dsRNA as there is no staining upon dZDHHC8 knockdown (dsRNA anti dZDHHC8). (B) Endogenous dZDHHC8 (red) was co-stained with markers for different cellular compartments (green). Overexpressed organelle markers are Golgi-tethered fringe-myc, HA-Rab5 marking endosomes, GFP-KDEL marking endoplasmic reticulum and dLamp1-GFP marking lysosomes. DAPI was used to stain nuclei. dZDHHC8 co-localises with the Golgi-tethered marker.
Fig 4.
Acyl-Biotin Exchange (ABE) assay identifies acylated proteins in Drosophila as well as potential dZDHHC8 targets.
(A) Schematic representation of an Acyl-Biotin Exchange (ABE) assay followed by mass spectrometry (ABE-MS) to identify acylated proteins (ie palmitoylated, stearoylated, etc). First, lysates are treated with N-ethylmaleimide to block free cysteine-residues on proteins. Acylation on proteins is then removed by hydroxylamine (H3NO) treatment yielding free cysteines that were previously acylated. Free cysteines subsequently react with thiol-reactive HPDP-biotin and biotinylated proteins are pulled down with strepavidin-beads. Eluted proteins are analysed by mass spectrometry or western blot (ABE-WB). Every ABE-assay includes a non-H3NO treated control to distinguish acylated proteins from unspecifically biotinylated proteins. (B) Csp and Ras64B are acylated. ABE assay was performed on lysates from L3 larvae expressing HA-tagged Csp or Ras64B under the control of heatshock-Gal4 24h after heatshock. Csp and Ras64B are acylated as they are specifically pulled down in the H3NO treated, but not in H3NO untreated control samples. (C) The stoichiometry of ras64B acylation is very low, in agreement with panel (B), as it cannot be detected by the acyl-PEG exchange (APE) method [20]. Detection of calnexin acylation in HeLa cells and of Csp acylation in larvae serve as positive controls for the method. (D) Scrib is palmitoylated, partly by dZDHHC8. Shown are endogenous scrib protein levels from control (CT) and dZDHHC8 knockout (KO1) larvae in lysates and after ABE assay. Scrib is detected after ABE in H3NO treated samples of CT and to a lesser extend of KO1 larvae, but not in control samples (-H3NO).
Fig 5.
dZDHHC8 affects Ras64B stability.
(A) Ras64B overexpression decreases viability of dZDHHC8 knockouts. Four days after L1 collection, the number of live Ras64B overexpressing larvae (hsG4; UAS-Ras64B), dZDHHC8 homozygous mutant larvae (KO1; hsG4) or larvae overexpressing Ras64B in dZDHHC8KO1 mutant background (KO1; hsG4; UAS-Ras64B) was determined. Ras64B overexpression in the dZDHHC8KO mutant background causes synergistic lethality. Error bars = stdev. n = 3x60 larvae/genotype. * ttest≤0.04. (B) ABE assay was performed on lysates from control (CT) and dZDHHC8 knockout (KO1) early L3 larvae expressing Ras64B under the control of heatshock-Gal4 24h after heatshock. KO1 larvae have less palmitoylated Ras64B-HA when compared to WT controls but also less total Ras64B-HA protein. (C) dZDHHC8 affects Ras64B stability. Protein levels of Ras64B-HA expressed with hsG4 in control (CT) and dZDHHC8 knockout (KO1) larvae were assayed by immunoblotting at different time points after heat-shock. ERK was used as a loading control. KO1 larvae have less Ras64B-HA protein than CT larvae at any time point. (D) Ras64B is palmitoylated on at least two cysteines. Larvae expressing HA-tagged wildtype Ras64B, or mutant Ras64B where each of the three cysteine residues present in Ras64B is mutated to alanine. All expression constructs are inserted in the same genomic locus using phiC31-directed integration, and their expression induced using hsG4 and collecting samples for ABE 24h after heat-shock. All Ras64B-HA C/A mutants are still palmitoylated, albeit less than WT Ras64B-HA. Note that Ras64B C46A protein levels are lower than wildtype Ras64B. (E-G) Ras64B is palmitoylated on all three cysteines C46, C120, C147. Assay performed as in (B). Each of the Ras64B-HA double cysteine mutants is still palmitoylated (E). Only mutating all three cysteines at the same time (F) or mutating the cysteine before and within the CAAX motif (G) abolishes palmitoylation. Note that protein levels of Ras64B versions carrying C46A are lower.