Fig 1.
Loss of function of CG4825 affects lipid storage and cell growth.
(A) Knockdown of CG4825 in salivary glands reduces cell size and causes lipid droplet accumulation in 3rd instar larval salivary glands. (B) Quantification of the salivary gland size (n = 5 for each group). (C) The knockdown efficiency of ppl>CG4825 RNAi determined by qRT-PCR (each group has three biological repeats with 25 flies in each repeat). (D) Staining of lipid droplets in 3rd instar larval salivary glands with knockdown of CG4825 driven by AB1-GAL4. (E) The gene structure of CG4825. The UTR regions are shown as grey boxed regions while the black and red boxes represent the protein coding region. The red boxed regions indicate the predicted PSS domain. The transposon insertion sites of CG4825KG06018 and CG4825MI01234 and the RNAi target regions are marked. (F) The mRNA level of CG4825 in wild-type and CG4825KG06018 mutant 3rd instar larval salivary glands (each group has three biological repeats, and each repeat contains 25 flies). (G) Accumulation of lipid droplets in CG4825KG06018 mutant 3rd instar larval salivary gland cells. In A, D and G, BODIPY (green) stains lipid droplets and DAPI (blue) stains the nuclei. In B, C and F, data are shown as mean ± SEM, and were compared with the unpaired Welch Two Sample t-test. * p < 0.05, ** p < 0.01, *** p < 0.001. Scale bar represents 100 μm.
Fig 2.
CG4825 is the sole Drosophila phosphatidylserine synthase (PSS).
(A) The evolutionary phylogenetic tree of PSS and its homologs in different species. The scale bar indicates amino acid divergence between sequences. Bootstrap number is 1,000 for resampling. (B) Schematic of the phospholipid and glycerolipid synthesis pathways. The black arrows indicate the common pathways in different species, the green arrow shows the pathway in yeast, the red arrows indicate the pathways in mammals, and the blue arrows indicate the pathways that occur in mitochondria. (C) S2 cells were cultured with medium containing 10 μM NBD-PA, NBD-PC, or NBD-PE for pulse-chase analysis. The S2 cells were harvested before the pulse-chase treatment (P) and 0, 1, 2, and 4 hours after pulse-chase. The extracted total lipids from S2 cells were separated on TLC plates for 1~2 hour. NBD-PE is converted to NBD-PS while NBD-PC is not. M, marker. Noted that a contaminating signal, which is likely from the plate of Typhoon 9500 imager, is present in all experiments. (D-E) Lipid profiling of control and pss RNAi 3rd instar larval salivary glands. The levels of the lipid species were normalized by calculating the mole fraction of the total polar lipids. In pss RNAi, the levels of PS, PG and CL are reduced, the levels of PE, PA, PI, DAG and TAG are increased, and the level of PC is unchanged. Data are shown as mean ± SEM. Data were compared with the unpaired Welch Two Sample t-test. * p < 0.05, ** p < 0.01, *** p < 0.001.
Fig 3.
The cell growth defect of pss RNAi is caused by impaired insulin pathway activity.
(A) Images of the tGPH reporter in control and pss RNAi 3rd instar larval salivary glands. The same laser power and exposure time were used during the imaging. Scale bar represents 100 μm. (B) Quantification of the PM:cytoplasm PIP3 ratio in control and pss RNAi salivary glands (n = 5). Data are shown as mean ± SEM and were compared with Welch Two Sample t-test. *** p < 0.001. (C) Western blot of Akt and pAkt in ppl>control RNAi, and pss RNAi 3rd instar larval salivary glands. Three replicates were performed and a representative result from one replicate is shown here. A total of 10 μg protein was loaded. (D) Immunofluorescent staining of Akt in 3rd instar larval salivary gland cells. Scale bar represents 25 μm. (E) Overexpression of AktCA rescues the reduced salivary gland size, but not the ectopic lipid accumulation, of pss RNAi. Scale bar represents 100 μm. (F) Quantification of the size of salivary gland cells with RNAi of pss and over-expression of Akt (n = 5 for each group). Data are shown as mean ± SEM. Data were compared with One-way ANOVA. Multiple comparisons of means were conducted with Tukey Contrasts. * p < 0.05, *** p < 0.001, ns: not statistically significant. (G) Overexpression (OE) of Pisd (CG5991) in pss RNAi suppresses the lipid accumulation and enhances the cell size reduction. In E and F, BODIPY (green) stains lipid droplets and DAPI (blue) stains the nuclei. Scale bar represents 100 μm. (H) Quantification of the size of salivary gland cells in different genetic backgrounds (n = 5 for each group). Data are shown as mean ± SEM. Data were compared with One-way ANOVA. Multiple comparisons of means were conducted with Tukey Contrasts. * p < 0.05, *** p < 0.001.
Fig 4.
pss RNAi affects mitochondrial integrity.
(A) Images of mitoEYFP in 3rd instar larval salivary glands of ppl-Gal4/+, ppl>control RNAi and pss RNAi. Green (mitoEYFP), mitochondria; blue (DAPI), nuclei. Scale bar represents 20 μm. (B) Immunofluorescent staining of mitochondria with anti-ATP5A antibody in 3rd instar larval salivary glands. The nuclei are labeled with DAPI. Scale bar represents 25 μm. (C) Fluorescence signal of mitoEYFP reporter in 3rd instar larval salivary glands of ppl-GAL4/+, ppl>control RNAi or Tom40 RNAi. The mitoEYFP signal is absent in Tom40 RNAi salivary gland cells. Scale bar represents 20 μm. (D-E) EM images of mitochondria in 3rd instar larval salivary glands of ppl-GAL4/+ and pss RNAi. The red arrows mark cristae. Scale bar represents 1 μm. M, mitochondria; LD, lipid droplet. (F) The knockdown efficiency of human PTDSS1 and PTDSS2 in HeLa cells (n = 3). Data are shown as mean ± SEM. Data were compared with the unpaired Welch Two Sample t-test. * p < 0.05, ** p < 0.01. (G) Images of mitoEYFP in HeLa cells with RNAi of human PTDSS1 and PTDSS2. The tubular structures are fragmented in cells with double RNAi of human PTDSS1 and PTDSS2. Scale bar represents 10 μm.
Fig 5.
The metabolic shift from phospholipid synthesis to TAG synthesis contributes to ectopic lipid storage in salivary glands with pss RNAi.
(A) Staining of lipid droplets with BODIPY in 3rd instar larval salivary gland with RNAi of Lipin. DAPI stains nuclei. Scale bar represents 100 μm. (B) Staining of lipid droplets with Nile Red shows that overexpression (OE) of CdsA suppresses the lipid accumulation in 3rd instar larval salivary glands of pss RNAi. DAPI stains nuclei. Scale bar represents 100 μm. (C) Lipid profiling of 3rd instar larval salivary glands of control, pss RNAi, PisdOE, and pss RNAi with PisdOE. Data are shown as mean ± SEM. Data were compared with one-way ANOVA. Multiple comparisons of means were conducted with Tukey Contrasts. * p < 0.05, ** p < 0.01, *** p < 0.001.
Fig 6.
The mitochondrial defect of pss RNAi is not rescued by Pisd overexpression.
(A) Fluorescence images of mitoEYFP in 3rd instar larval salivary glands of different genetic backgrounds. The absence of mitoEYFP signal in pss RNAi salivary glands is not rescued by Pisd (CG5991) overexpression. mitoEYFP (green) marks mitochondria and DAPI (blue) stains the nuclei. Scale bar represents 20 μm. (B) Immunostaining of mitochondria with anti-ATP5A antibody in 3rd instar larval salivary glands of different genetic backgrounds. Anti-ATP5A antibody (green) marks mitochondria and DAPI (blue) stains the nuclei. Scale bar represents 20 μm. (C) The EM structure of mitochondria in 3rd instar larval salivary gland cells of control, pss RNAi, and pss RNAi with Pisd overexpression (PisdOE). Scale bar represents 1 μm. Red arrows in ppl-GAL4/+ and ppl>pss RNAi salivary gland cells mark the crista junction sites, and the red arrow in the salivary gland cell with pss RNAi and PisdOE marks the filamentous meshes or crista fragments. (D) Images of mitoEYFP in 3rd instar larval salivary glands of different genetic backgrounds. Overexpression of CdsA does not restore the mitoEYFP signal in salivary glands of pss RNAi. Scale bar represents 20 μm. (E) Images of mitoEYFP in 3rd instar larval salivary glands of different genetic backgrounds. The loss of mitoEYFP signal in pss RNAi is not rescued by the overexpression of AktCA. Scale bar represents 20 μm. (F) Illustration of the cellular PS transport pathways. (G) Images of mitoEYFP in 3rd instar larval salivary glands of different genetic backgrounds. The loss of mitoEYFP signal in pss RNAi is partially rescued by PI4KIIIα RNAi.
Fig 7.
Proposed models of the mechanisms underlying the different pss RNAi phenotypes.
(A-B) Loss of pss function leads to reduced Akt, increased DAG and TAG, and impaired mitochondrial structure and function. (C) The overexpression of Pisd in pss RNAi may lead to increased engagement of phospholipids in the PS-PE-PS cycle locally, probably at the ER or mitochondria. Therefore, Pisd overexpression reduces the level of PS at the PM and the amount of DAG available for TAG synthesis, leading to smaller cells and decreased lipid storage compared to pss RNAi alone (B). (D) RNAi of PI4KIIIα rescues the defective mitochondrial import of mitoEYFP in pss RNAi by increasing the transport of PS to the mitochondria.