mTORC1-Sch9 regulates hydrogen sulfide production through the transsulfuration pathway

Inhibiting mTORC1-Sch9 inhibits H₂S production

Sch9 is a direct substrate of yeast mTORC1 and depletion of SCH9 extends yeast lifespan through mechanisms shared with lifespan extension by calorie restriction (CR) [13, 16]. Since H₂S mediates the benefits of CR, we first compared H₂S production in Δsch9 mutant cells to WT cells. While WT cells released measurable amounts of H₂S, Δsch9 cells produce barely detectable amounts of H₂S (Figure 1A and 1B). H₂S production was recovered if a functional SCH9 gene was added back to the mutant cells (Figure 1A and 1B), thus, showing that Sch9 activity is required for H₂S production. Western blotting for Sch9 was used to verify that H₂S production correlated with the concentration of Sch9 protein present in cells (lower panels, Figure 1A).

Decreased H₂S production by Δsch9 cells was also observed by measuring the reduction of methylene blue in different yeast strain backgrounds (BY4741 and BY4742) (Figure 1C). Measurement of intracellular H₂S in these two yeast strains by using WSP-1 fluorescent showed the same trends (Figure 1D). Consistent with previous studies showing that CR enhanced H₂S production in yeast [7], we observed a significant increase of H₂S production in WT cell under CR (Figure 1E). However, similar to no restriction condition, H₂S production was still significantly impaired in Δsch9 cell even under CR (Figure 1E).

To investigate if phosphorylation of Sch9, which correlates with its kinase activity and is required for H₂S production, two inhibitors, Myriocin and Rapamycin, that indirectly lower Sch9 activity were used to inhibit the phosphorylation of Sch9 at the activation loop and the hydrophobic motif, respectively [19] (Figure 2A). Treatment with myriocin at 0.75 μM, but not 0.25 μM, inhibited H2S production (Figure 2B). Similarly, Rapamycin treatments at 10 or 50 nM also resulted in decreased H2S production (Figure 2C). These data indicate that the phosphorylation of Sch9 at the both activation loop and the hydrophobic motifs are required to the regulate H₂S production.

Unlike mammalian cells, yeast cells convert extracellular sulfate to sulfide through the sulfur assimilatory pathway with enzymes encoded by MET14, MET16, and MET5/10 in addition to conserved the TSP pathway [7]. To verify if the TSP pathway is involved in decreased H₂S production by mTORC1-Sch9 inhibition in yeast, we monitored the effect of rapamycin in H₂S production by MET14, MET16 and MET5 mutants. As shown in Figure 2D, all three mutants have decreased H₂S production upon mTORC1 inhibition by rapamycin, suggesting that interfering with the sulfur assimilatory pathway does not change the inhibitory role of rapamycin in H₂S production and the TSP pathway is likely involved.

The role of mTORC1-Sch9 on the production of hydrogen sulfide is not caused by the alteration of methionine metabolism

The mTORC1 pathway plays an important role in regulating cell growth in response to amino acid availability [19, 20]. Additionally, methionine metabolism contributes to H₂S production through TSP pathway and methionine restriction extends lifespan from yeasts to humans [3, 8]. Therefore, we asked if there is a lowered free methionine pool in Δsch9 cell which may contribute to the decreased H₂S production. The effects of methionine on H₂S production were investigated in BY4741, a strain with a defective MET15 gene which prevents the synthesis of methionine from sulfate in the medium, and BY4742 with functional MET15. BY4741 cells produced more H₂S in the present of 2 mg/L methionine than in the present of 20 mg/L methionine while BY4742 cell only produced barely detectable H₂S at both conditions (Figure 3A). This suggests that methionine restriction indeed contributes to H₂S production. It is worth to noting that when the methionine concentration in the medium was increased to 100 mg/L, H₂S production increased in BY4742 cells. This may be due to the extreme abundance of substrates for H₂S production. Similarly, in a different yeast background TB50a, decreasing the methionine concentration in the medium from 20 mg/L to 5 mg/L or 0 gave rise to significant H₂S production while Δsch9 cell remained defective in H₂S production under those conditions (Figure 3B). Also, increasing methionine to 80 mg/L partially recovered H₂S production by Δsch9 cell probably due to the extreme abundance of substrates.

To investigate if there is a high level of intracellular methionine to inhibit H₂S production by Δsch9 cell, a methionine probe plasmid pUG35-eGFP was constructed by putting eGFP expression under MET17 promoter which efficiency is inhibited by high concentration of intracellular methionine [21]. Indeed, the expression of eGFP protein by pUG35-eGFP was inhibited by increasing methionine concentration of medium (Figure 3C upper and Figure 3D left panels), while eGFP level was not altered by exogenous methionine in the absence of the MET17 promoter (Figure 3C lower and Figure 3D right panels) indicating that the altered eGFP levels are not due to the degradation of the protein. However, when pUG35-eGFP was transformed into Δsch9 cell the expression of eGFP was similar to that when the probe was present in WT cells (Figure 3D left panels), indicating that the intracellular level of methionine in Δsch9 cell is not higher than that in WT cells and does not contribute to the decreased H₂S production.

mTORC1-Sch9 regulates H₂S production by regulating cysteine metabolism

Cysteine is another sulfur-containing amino acid whose metabolism is closely related to H₂S production [3]. To verify that an altered intracellular level of cysteine is involved in the regulation of H₂S production by mTORC1-Sch9, we first investigated how H₂S production is affected by different levels of cysteine supplementation. Adding 100 mg/L cysteine to SDC medium lacking methionine significantly decreased H₂S production in WT TB50a cell (Figure 4A). And increasing cysteine concentration to 500 mg/L restored H₂S production likely due to substrate abundance (Figure 4A). Deletion of SCH9 caused significant inhibition of H₂S production under cysteine limited or over-supplied conditions (Figure 4A). These data suggest that the inhibition of mTORC1-Sch9 renders H₂S production less sensitive to exogenous cysteine, probably due to increased endogenous cysteine. Indeed, unlike the intracellular methionine level which was not changed in Δsch9 cell (Figure 3D), the intracellular cysteine level was ~50% higher in Δsch9 cell than WT cell (Figure 4B). And when a functional SCH9 gene was added back to mutant cells the intracellular cysteine level decreased (Figure 4B), showing that Sch9 regulates intracellular cysteine metabolism. When the level of exogenous cysteine was 100 mg/L, the intracellular cysteine level was still higher in Δsch9 cells than WT cells until the level of exogenous cysteine reached 500 mg/L (Figure 4C). Together, these data indicate that the decreased H₂S production by mTORC1-Sch9 inhibition is most likely due to an increase in the level of intracellular cysteine.

mTORC1-Sch9 regulates H₂S production via transsulfuration pathway

Increased intracellular cysteine usually inhibits the expression of enzymes in the transsulfuration pathway that required for H₂S production [7]. To investigate if the expression of the transsulfuration enzymes is altered in Δsch9 cell, the mRNA levels of CYS3 and CYS4 which encodes Cystathionine gamma-lyase and Cystathionine beta-synthase respectively in yeast were compared in Δsch9 and WT cells. Indeed, the mRNA levels of both CYS3 and CYS4 decreased to about 50% in Δsch9 cell compared to them in WT cell. And it can be reversed by adding SCH9 back to mutant cells (Figure 5A). Similarly, inhibiting mTORC1-Sch9 by rapamycin also decreased the expression of both CYS3 and CYS4 (Figure 5B). These data suggest that inhibiting mTORC1-Sch9 which increases intracellular cysteine level does down-regulate transsulfuration pathway.

To verify the role of Cys3 and Cys4 in H₂S production regulation by mTORC1-Sch9, CYS3 or CYS4 was overexpressed in Δsch9 or WT cells treated with rapamycin and H₂S production was monitored (Figure 5C to 5E). Overexpressing CYS3 significantly increased H₂S production by WT cell. While Δsch9 cell with empty vector produced little H₂S, overexpressing CYS3 restored H₂S production to a level similar to WT cells overexpressing CYS3 (Figure 5C and 5D). Similarly, overexpressing CYS3 also restored H₂S production in rapamycin treated WT cell (Figure 5E). However, although overexpressing CYS4 restored H₂S production in rapamycin treated WT cell, it did not restore H₂S production in Δsch9 cell (Figure 5C). These data support the hypothesis that the CYS3 and CYS4 genes in the transsulfuration pathway mediate H₂S production and are regulated by the mTORC1-Sch9 pathway.

Inhibiting mTORC1 reduces H₂S production and expression of transsulfuration pathway enzymes in human cells.

The anti-aging effects of H₂S production mediated by transsulfuration pathway and regulated by mTORC1 are evolutionary conserved from yeast to mammals [7, 11–13]. To assess if inhibiting mTORC1 also interferes with H₂S production in mammalian cells, we examined human 293T and HeLa cells. To measure H₂S production in cultured 293T and HeLa cells by using the lead acetate strip assay the growth medium was supplemented with the CGL/CBS substrates Cys and cofactor pyridoxal-5′-phosphate (PLP). The supplementation of Cys/PLP decreased the cell viability (Supplementary Figure 1), but significantly increased H₂S production. With these assay conditions rapamycin treated cells have decreased the cell death induced by Cys/PLP and higher cell density (Supplementary Figure 1). However, less H₂S is produced in both 293T and HeLa cells with rapamycin treatments (Figure 6A and 6B), similar to what we observed in yeast cells (Figure 2D). These data suggest that inhibiting mTORC1 in mammalian cells may also decrease H₂S production.

To determine if reduction of H₂S by inhibition of mTORC1 activity in mammalian cells is also accompanied by down-regulation of transsulfuration pathway enzymes, the expression of CGL and CBS was monitored by RT-qPCR in HeLa cell with or without rapamycin treatments for 1 to 4 days. Similar to what we observed in yeast cells (Figure 5B), rapamycin treatment reduced the mRNA level of both CGL and CBS significantly (Figure 6C and 6D). Consistent with mRNA levels, CGL and CBS protein levels were significantly decreased after 3 to 4 days of rapamycin treatments (Figure 6E), indicating that the expression of both transsulfuration pathway enzymes is reduced upon mTORC1 inhibition in HeLa cell.

Yeast strains, plasmids, and media

The S. cerevisiae strains and plasmids used in this study are listed in Table 1 and Table 2. Strains were grown at 30 °C in YPD medium (1% yeast extract, 2% peptone, 2% glucose) or synthetic dextrose complete medium (SDC) which contains no cysteine [26]. Cells transformed with plasmids carrying URA3 were grown in the SDC medium lacking uracil. For inducing the expression of DNA sequences inserted into pYES2, cells were grown in the galactose-inducing medium (2% glucose, carbon source of SDC medium, was replaced to 1% Galactose and 1% Sucrose).

Protein extraction and western blotting

For protein extraction, trichloroacetic acid yeast cell extracts were prepared according to a method described previously [17, 27]. HeLa cell with indicated treatments were washed with PBS and lysed in Laemmli buffer. Standard SDS-PAGE and Western blotting protocols were performed with primary antibodies including polyclonal rabbit anti-Sch9 antibodies (1:2000, a gift from Dr. Robert C. Dickson of University of Kentucky), monoclonal mouse anti-GFP antibodies (1:5000, Zen BioScience, China), monoclonal mouse anti-actin (1:5000, Zen BioScience, China), polyclonal rabbit anti-LC3B (1:2000, Abcam, USA), monoclonal rabbit anti-CGL (1:2000, Abcam, USA) or monoclonal rabbit anti-CBS (1:2000, Abcam, USA). Secondary antibodies include alkaline phosphatase-linked anti-rabbit or anti-mouse IgG (1:2000, Zen BioScience, China).

RNA extraction and RT-qPCR analysis

Total RNA was extracted from yeast cell or HeLa cell with RNAiso Plus (TaKaRa Bio, China) or Beyozol reagent (Beyotime) respectively. For yeast cell, cells (OD600nm of 5) were pretreated with 50 U lyticase at 30 °C for 30 min to increase extraction efficiency. Reverse transcription reactions were conducted using a PrimeScript RT reagent kit with gDNA eraser (Takara Bio, China). The primers for real-time quantitative PCR (RT-qPCR) are indicated in Table 3. RT-qPCR experiments were performed using SYBR Premix Ex Taq II (TaKaRa Bio, China) and Bio-Rad CFX manager RT-qPCR system. Data were collected and analyzed by Bio-Rad CFX manager software.

H₂S assay for yeast cell

Lead acetate strips for measuring H₂S were pasted on the top of inner wall of 125 ml culture flasks containing 25 ml media. 1 L culture flasks containing 250 ml media were used for TB50a cells which do not intensively produce H₂S. H₂S reacts with the lead acetate on the strip, creating a darkened band. The length of the darkened band is proportional to the amount of H₂S produced during fermentation [28, 29]. The stripes were replaced every day or remained for entire experimental periods as described.

The methylene blue assay described previously for H₂S detection was also performed in centrifuge tubes [30]. 36ml cells at OD600nm of 0.05 were divided in two parts. One of them was added with 2ml medium to monitor growth rate as measured by the absorbance at 600 nm and the another was the addition of 2 ml methylene blue reaction mix (1 mg/ml methylene blue, 100 mM citric acid buffer at pH 4.5), reacting with H₂S dissolved in medium. Methylene blue decolorization by H₂S were monitored at 663 nm and normalized to biomass.

Intracellular free H₂S levels were also determined using the H₂S fluorescent probe WSP-1 (Cayman, USA) [31]. 1 OD of cells were harvested and washed with PBS, then incubated with 10 mM WSP-1 for 1 h at 30 °C in dark. After washed three times with PBS (2.7 mM KCl, 1.75 mM KH₂PO₄, 10 mM Na₂HPO₄ and 136.75 mM NaCl, pH 7.4), the pellet was resuspended in 1 ml PBS. 10 μl of cells was added to a microscope slide, and WSP-1 fluorescence was monitored using CEWEI LWD200-37FT fluorescence microscope (CEWEI, China). Alternatively, the fluorescence was measured at 465/515 nm excitation/ emission using an f4500 fluorescence spectrometer (Hitachi, Japan). The fluorescence intensity was normalized according to the OD value.

H₂S assay for cultured human cells

293T or HeLa cell was grown in DMEM (Invitrogen) supplemented with 10% FBS, 50 U/mL penicillin, and 50 μg/mL streptomycin with or without 250 nM rapamycin. To measure H₂S production, growth media was supplemented with or without 10 mM Cys and 10 mM pyridoxal-5′-phosphate (PLP) and a lead acetate strip was placed above the media in a 25 ml cell culture flask incubated in a CO₂ incubator at 37 °C for indicated time.

Measurement of the Cys content

The extraction and estimation of cysteine content in yeast cells were done as described previously [32]. Cells grown in SDC-Ura or medium containing different concentrations of Cys were harvested and washed with PBS twice by centrifugation. Cell pellets were dried at 50 °C until constant weight was achieved. Dried cells were lysed in liquid nitrogen and then 1ml of 10% TCA was added to 200mg cell powder. The homogenates were centrifuged at 2800×g for 60min. Acid ninhydrin were added to the extract, and reaction mixture was kept in boiled water for 10 min. After fast cooling, the A560nm absorbance of the reaction mixture was measured. The amount of cysteine in each reaction was determined using a standard curve.

Statistical analysis

All data from at least three independent experiments. Error bars are presented as averages ± SD. Statistical analysis and comparisons were performed using two-tailed, unpaired Student t tests.