Trans-Golgi network and endosome dynamics connect ceramide homeostasis with regulation of the unfolded protein response and TOR signaling in yeast

Abstract

Synthetic genetic array analyses identify powerful genetic interactions between a thermosensitive allele (sec14-1(ts)) of the structural gene for the major yeast phosphatidylinositol transfer protein (SEC14) and a structural gene deletion allele (tlg2Delta) for the Tlg2 target membrane-soluble N-ethylmaleimide-sensitive factor attachment protein receptor. The data further demonstrate Sec14 is required for proper trans-Golgi network (TGN)/endosomal dynamics in yeast. Paradoxically, combinatorial depletion of Sec14 and Tlg2 activities elicits trafficking defects from the endoplasmic reticulum, and these defects are accompanied by compromise of the unfolded protein response (UPR). UPR failure occurs downstream of Hac1 mRNA splicing, and it is further accompanied by defects in TOR signaling. The data link TGN/endosomal dynamics with ceramide homeostasis, UPR activity, and TOR signaling in yeast, and they identify the Sit4 protein phosphatase as a primary conduit through which ceramides link to the UPR. We suggest combinatorial Sec14/Tlg2 dysfunction evokes inappropriate turnover of complex sphingolipids in endosomes. One result of this turnover is potentiation of ceramide-activated phosphatase-mediated down-regulation of the UPR. These results provide new insight into Sec14 function, and they emphasize the TGN/endosomal system as a central hub for homeostatic regulation in eukaryotes.

Figure 1.

Synthetic genetic interactions. (A) The sec14-1^ts allele exhibits strong synthetic interactions with tlg2Δ, gsg1Δ, and snc2Δ (genotypes indicated at left). Indicated double mutants were generated by meiotic segregation, and they were derived from tetrads with four viable spores and where all markers segregated in a Mendelian manner. Cells were spotted in a 10-fold dilution series on YPD and incubated at either permissive (30°C), semipermissive (33.5°C), or restrictive (35 and 37°C) temperatures for sec14-1^ts strains. Growth results were recorded after 48 h. (B) Loss of Tlg2 SNARE activity exacerbates sec14-1^ts associated growth defects. Growth properties of sec14-1^ts tlg2Δ cells carrying YCp(URA3), YCp(TLG2), YCp(tlg2^Q278R), and YCp (tlg2^R13Q) were determined. Cells were spotted in a 10-fold dilution series on uracil-free minimal medium and incubated at 33.5°C for 72 h. (C) tlg2Δ compromises kes1Δ-mediated bypass Sec14. Yeast strains of indicated genotype were spotted in 10-fold dilution series on YPD and incubated at 30, 33.5, 35, or 37°C as indicated. Growth results were recorded after 48 h.

Figure 2.

Endocytic trafficking is defective in sec14-1^ts mutants. (A) GFP-Snc1 accumulates in intracellular compartments in sec14-1^ts mutants. Wild-type and isogenic sec14-1^ts yeast harboring YCp(GFP-SNC1) were grown to mid-logarithmic growth phase in uracil-free minimal medium at 30°C. Cells were incubated with FM4-64 (10 μM) for 10 min after which cells were intoxicated with NaN₃/NaF (10 mM final, each). Cells were washed twice with NaN₃/NaF (10 mM final, each), and FM4-64 and GFP-Snc1 profiles were imaged using a Nikon E600 microscope. GFP-Snc1, FM4-64 and merge profiles are shown, as indicated. (B) GFP-Snc1 and FYVE-dsRed localization profiles are shown. Wild-type and isogenic sec14-1^ts yeast harboring YCp(GFP-SNC1) were grown to mid-logarithmic growth phase in uracil-free minimal medium at 30°C. GFP-Snc1 and FYVE-dsRed profiles were imaged using a Nikon E600 microscope. GFP-Snc1, FYVE-dsRed, and merge profiles are shown, as indicated. Arrows highlight examples of where GFP-Snc1 and FYVE-dsRed compartments are juxtaposed. (C) GFP-Snc1 localization in bypass Sec14 mutant yeast. Yeast strains with the indicated genotypes harboring YCp(GFP-SNC1) were grown to mid-logarithmic growth phase in uracil-free minimal medium at 30°C and then either maintained at 30°C or shifted to 37°C for 60 min. The GFP-Snc1 profile was imaged using a Nikon E600 microscope.

Figure 3.

Early secretory block in sec14-1^ts tlg2Δ yeast. (A) CPY trafficking is blocked at the ER in sec14-1^ts tlg2Δ double mutants, and exit from the ER is restored in the isogenic sec14-1^ts tlg2Δ kes1Δ strain. Yeast cultures were grown in minimal media at 25°C, shifted to 37°C for 2 h, and radiolabeled with [³⁵S]-amino acids for 30 min. Chase (10 min) was initiated by introducing excess unlabeled methionine and cysteine, and it was terminated with ice-cold trichloroacetic acid (final concentration, 5%). The core glycosylated p1 CPY, the fully modified p2 CPY, and the mCPY are indicated at right. Genotypes are at top. (B) Early secretory block in sec14-1^ts tlg2Δ mutants is poorly reversible. Yeast strains with the indicated genotypes were grown in minimal media at 25°C, shifted to 37°C for 2 h, and radiolabeled with [³⁵S]-amino acids for 30 min. Chase was initiated by addition of excess unlabeled methionine and cysteine, and cultures were concomitantly shifted to 25°C. Chase at 25°C was for 1, 3, and 5 h, as indicated. (C) Thin-section electron microscopy. Yeast strains with the indicated genotypes were cultured to early logarithmic growth phase in YPD medium at 30°C. The cultures were then shifted to the 37°C for an additional 2 h. Cells were fixed, embedded in Spurr's resin, stained with uranyl acetate, and imaged using a transmission electron microscope (20 kV). Representative images are shown (bar, 1 μm). Distended ER is highlighted by arrows.

Figure 4.

UPR activity in sec14-1^ts tlg2Δ double mutants. (A) Down-regulation of the UPR in sec14-1^ts tlg2Δ relative to parental sec14-1^tscells. β-Galactosidase expression was followed as reporter of UPR activity in yeast strains of the indicated genotype that harbor the pJT30 YCp(UPRE::LACZ) plasmid where LACZ transcription is under control of the KAR2 enhancer (Wilkinson et al., 2000). Cells were grown at the permissive temperature (30°C) or shifted to the restrictive temperature (37°C), for sec14-1^ts, 2 h before analysis. As control for UPR activity, wild-type yeast were treated with DTT (final concentration, 5 mM) for 60 min before assay. (B) The UPR cannot be activated in sec14-1^ts tlg2Δ yeast by challenge with ER stress agents. Yeast strains of indicated genotype were cultured, and UPR activity was assayed, as described in A, with the modification that cultures were either mock treated, challenged with DTT (5 mM), or challenged with 10 μg/ml tunicamycin 1 h before analysis. (C) UPR silencing is downstream of Ire1p function. cDNA was synthesized from total RNA isolated from either wild-type or sec14-1^ts tlg2Δ cells grown in the presence or absence of 5 mM DTT. In all cases, the yeast strains were grown at 30°C overnight and shifted to 37°C for 2 h before DTT challenge. Endogenous HAC1 and ACT1 mRNA levels were monitored by PCR as reporters of UPR activity and as normalizing factor, respectively. (D) HAC1^I expression rescues sec14-1^ts tlg2Δ-associated growth defects. The sec14-1^ts tlg2Δ double mutant was transformed with either vector control (pRS313) or YCp HAC1^I (pRC43), as indicated, transformants were spotted in 10-fold dilution series on YPD, and incubated at either permissive (30°C) or restrictive (33.5°C) temperatures. Growth was scored after 72 h. (E) Hac1^I transcriptional activation potency is reduced in sec14-1^ts tlg2Δ cells. UPR-dependent β-galactosidase expression (from the pJT30 reporter plasmid) for wild-type and sec14-1^ts tlg2Δ yeast in the absence or presence of HAC1^I expression (sustained by the YCpHAC1^I plasmid pRC43). Cells were grown at the permissive temperature (30°C) and challenged with the indicated temperatures for 2 h before analysis. As standard for UPR activity, wild-type yeast were challenged with DTT (final concentration, 5 mM) for 1 h before assay.

Figure 5.

Transcriptional profiling of sec14-1^ts tlg2Δ yeast. Gene expression profiles were compared for sec14-1^ts tlg2Δ (Cy5) double mutants relative to wild-type yeast (Cy3) by using Gene Spring version 7.3.1 software (Agilent Technologies). (A and B) Aberrant regulation of the UPR regulon in sec14-1^ts tlg2Δ mutants. The expression profile of a selection of genes whose expression is either up-regulated (A) or down-regulated (B) by the UPR (as reported by Travers et al., 2000) is shown. Light gray bars present the expression of genes in sec14-1^ts tlg2Δ mutants relative to wild-type, whereas dark gray bars represent gene expression in wild-type cells treated with DTT (5 mM) for 60 min relative to mock-treated wild-type cells (Travers et al., 2000). (C) TOR signaling is defective in sec14-1^ts tlg2Δ mutants. The expression profile is shown for genes encoding ribosomal structural proteins (RP). Expression maps show the log₁₀ expression profile of genes, after Lowess normalization, of sec14-1^ts tlg2Δ (Cy5) double mutants relative to wild-type yeast (Cy3). Analysis was performed using Gene Spring version 7.3.1 software (Agilent Technologies). Gray lines represent ± twofold expression difference. (D) sec14-1^ts tlg2Δ mutants are hypersensitive to rapamycin. Yeast strains of the indicated genotype were spotted in a 10-fold dilution series on either YPD alone or YPD supplemented with 2.5 nM rapamycin. Cells were grown at 30°C and growth was scored after 72 h. (E) Partial elevation of phosphor-eIF2α levels in sec14-1^ts tlg2Δ cells. Yeast total protein lysates were prepared from yeast strains of the indicated genotype, resolved by SDS-PAGE and blotted to nitrocellulose. Blots were probed with either anti-phospho eIF2α or anti-Sec61p antibody. (G) Down-regulation of the general amino acid control (GAAC) response in sec14-1^ts tlg2Δ mutants. The expression profile is shown for genes encoding amino acid biosynthetic enzymes. Expression maps show the log₁₀ expression profile of genes, after Lowess normalization, of sec14-1^ts tlg2Δ (Cy5) double mutants relative to wild-type yeast (Cy3). Analysis was performed using Gene Spring version 7.3.1 software (Agilent Technologies). Gray lines represent ± 2-fold expression difference. (F) Gcn4 is not derepressed in sec14-1^ts tlg2Δ mutants. β-Galactosidase expression was followed as reporter of GCN4 derepression in yeast strains of the indicated genotype that harbor the p180 YCp(GCN4::LACZ) plasmid where LACZ transcription and translation is under GCN4 regulated control (Hinnebusch, 1993, 1997). Cells were grown at the permissive temperature (30°C) and then shifted to the restrictive temperature (37°C), for sec14-1^ts, 2 h before analysis.

Figure 6.

Derangements in ceramide homeostasis in sec14-1^ts tlg2Δ yeast correlate with UPR silencing. Dihydroceramide (A) and phytoceramide (B) levels are elevated in sec14-1^ts tlg2Δ yeast. Quantitative lipidemic approaches were used to measure endogenous ceramides in yeast of indicated genotype. Lipids were extracted from three pooled cultures (30 OD_{600 nm}) of each yeast strain grown overnight at 30°C and either maintained at 30°C, or shifted to 37°C for 2 h, before lipid extraction. Bulk dihydroceramides and phytoceramides are represented as femtomoles of ceramide per nanomole of Pi. (C and D) Derangements in dihydro- and phytoceramides are corrected in sec14-1^ts tlg2Δ yeast by genetic ablation of KES1 or by increased dosage of the Ypc1 ceramidase. Endogenous ceramide profiles are shown for sec14-1^ts tlg2Δ kes1Δ and sec14-1^ts tlg2Δ YEp(YPC1) yeast relative to wild-type and sec14-1^ts tlg2Δ strains. Culture conditions and parameters for lipid extraction are as described in A and B. (E) Functional ablation of Kes1 restores UPR competence to sec14-1^ts tlg2Δ yeast. β-Galactosidase assays were performed on wild-type, sec14-1^ts tlg2Δ, sec14-1^ts tlg2Δ kes1Δ, and sec14-1^ts tlg2Δ cki1Δ yeast cells harboring the YCp(UPRE::LACZ) UPR-reporter plasmid. Cells were grown at the permissive temperature (30°C) overnight, and then they were shifted to the restrictive temperature (37°C), for 2 h before assay. As positive control for UPR activity, wild-type yeast were challenged with 5 mM DTT for 1 h before assay. (F) Increased Ypc1 ceramidase activity restores UPR activity to sec14-1^ts tlg2Δ yeast. cDNA was synthesized from total RNA isolated from mock treated wild-type yeast, and from wild-type, sec14-1^ts tlg2Δ, and sec14-1^ts tlg2Δ YEp(YPC1) strains challenged with 5 mM DTT. In all cases, the indicated strains were grown at 30°C overnight and shifted to 37°C for 2 h before DTT challenge. KAR2 and ACT1 mRNA levels were monitored by PCR as reporters of UPR activity and as normalizing factor, respectively.

Figure 7.

Ceramide synthesis and regulation of the UPR in sec14-1^ts tlg2Δ yeast. (A) β-Galactosidase expression was followed as reporter of UPR activity in wild-type, sec14-1^ts tlg2Δ, and sec14-1^ts tlg2Δ lag1Δ yeast that harbor the pJT30 YCp(UPRE::LACZ) plasmid where LACZ transcription is under control of the KAR2 enhancer. Cells were grown at the permissive temperature (30°C) or shifted to the restrictive temperature (37°C) for sec14-1^ts, 2 h before analysis. As control for UPR activity, wild-type yeast were treated with DTT (final concentration, 5 mM) for 60 min before assay. (B) β-Galactosidase UPR reporter assays were performed on wild-type and sec14-1^ts tlg2Δ yeast strains carrying YCp(UPRE::LACZ). The yeast strains were grown at 30°C to mid-logarithmic growth phase, harvested and subcultured into fresh synthetic complete growth medium either supplemented with dimethyl sulfoxide or with fumonisin B1 (100 μM) and cultured for an additional 6 h. Cultures were subsequently shifted to 37°C 2 h before assay. Where indicated, yeast were challenged with 5 mM DTT (final concentration) for 1 h before analysis.

Figure 8.

Ceramide homeostasis and enhanced catabolism of inositol SLs. (A) Inositol-SLs are labile in sec14-1^ts tlg2Δ double mutants. Wild-type, sec14-1^ts, sec14-1^ts tlg2Δ, and sec14-1^ts tlg2Δ kes1Δ yeast were radiolabeled to steady state with ³[H]inositol (20 μCi/ml) at 30°C. Subsequently, cells were washed and resuspended in glucose-SD media and incubated for 3 h at 37°C in the presence of unlabeled inositol (12 μM). Lipids were extracted from cells and subject to alkaline methanolysis. Inositol-SLs were quantified by liquid scintillation counting and relative inositol-SL stability for each strain is expressed as the ratio of [³H] after chase:[³H] before chase. Values represent averages (and standard deviations) from three independent experiments. (B) Inositol SL profiles are altered in tlg2Δ mutants. Wild-type, sec14-1^ts, tlg2Δ, sec14-1^ts tlg2Δ, and sec14-1^ts tlg2Δ kes1Δ yeast were labeled to steady state with [³H]inositol (20 μCi/ml). After a 2-h shift to the restrictive temperature for the sec14-1^ts allele (37°C), lipids were extracted from 5 OD_{600 nm} of cells and subject to alkaline methanolysis. TLC profiles for the inositol-SLs for yeast strains of indicated genotype are shown and each lipid assignment (indicated at left) was determined by R_f. (C) Quantitative inositol SLomics. Complex SLs were extracted from 25 OD₆₀₀ of yeast and analyzed by mass spectrometry. Lipid mass was calculated by comparison with internal standards. Each SL species is represented as the log₁₀ of the mutant:wild-type ion intensity ratio. Species shown in red report decreased mass in mutant relative to wild-type, whereas green indicates increase. (D) Incorporation of a ppn1Δ allele into the sec14-1^ts tlg2Δ genetic background restores UPR activity. UPR activity was monitored via the UPRE::LACZ reporter in wild-type, sec14-1^ts tlg2Δ, sec14-1^ts tlg2Δ isc1Δ, and sec14-1^ts tlg2Δ ppn1Δ yeast strains as indicated. Yeast were cultured at the permissive temperature (30°C) to mid-logarithmic growth phase and then shifted to the restrictive temperature (37°C) for 2 h before analysis. As control for UPR activity, wild-type yeast were challenged with DTT (final concentration, 5 mM) for 60 min before assay.

Figure 9.

Ceramide effects on the UPR and the Sit4 phosphatase. (A) Incorporation of sit4Δ into the sec14-1^ts tlg2Δ genetic background restores the UPR. β-Galactosidase expression was followed as reporter of UPR activity in wild-type, sec14-1^ts tlg2Δ and sec14-1^ts tlg2Δ sit4Δ yeast that harbor YCp(UPRE::LACZ). Cells were grown at the permissive temperature (30°C) to mid-logarithmic growth phase and then shifted to 37°C for 2 h before analysis. As control for UPR activity, wild-type yeast were treated with DTT (final concentration, 5 mM) for 1 h before assay. (B) Rapamycin sensitivity in sec14-1^ts tlg2Δ sit4Δ and sec14-1^ts tlg2Δ kes1Δ mutants. Yeast (genotypes indicated) were spotted in 10-fold dilution series on either YPD or YPD supplemented with 2.5 or 5 nM rapamycin. Cells were incubated at 30°C and growth was scored after 72 h. (C) eIF2α phosphorylation is reduced in sec14-1^ts tlg2Δ kes1Δ cells. Lysates were prepared from yeast strains of the indicated genotype, fractionated by SDS-PAGE, and blotted to nitrocellulose. Blots were probed with anti-phospho-eIF2α or anti-Sec61p antibody.

Figure 10.

Sec14 function in the TGN/endosomal dynamics. Plausible execution points for Sec14 in trafficking from the TGN to the plasma membrane, from endosomes to the plasma membrane, and within the TGN/endosomal system are indicated by hatched X. Details are described in the text. Catastrophic defects in TGN/endosomal function result in inositol SL (SL) turnover in TGN/endosomal compartments and enhanced generation of ceramide (Cer) therein. Expanded TGN/endosomal Cer pools act via Sit4 to depress the transcriptional (Tx) components of the UPR. Endosome/TGN trafficking defects additionally result in defective TOR signaling.