Longevity control by supersulfide-mediated mitochondrial respiration and regulation of protein quality

Abstract

Supersulfides, which are defined as sulfur species with catenated sulfur atoms, are increasingly being investigated in biology. We recently identified pyridoxal phosphate (PLP)-dependent biosynthesis of cysteine persulfide (CysSSH) and related supersulfides by cysteinyl-tRNA synthetase (CARS). Here, we investigated the physiological role of CysSSH in budding yeast (Saccharomyces cerevisiae) by generating a PLP-binding site mutation K109A in CRS1 (the yeast ortholog of CARS), which decreased the synthesis of CysSSH and related supersulfides and also led to reduced chronological aging, effects that were associated with an increased endoplasmic reticulum stress response and impaired mitochondrial bioenergetics. Reduced chronological aging in the K109A mutant could be rescued by using exogenous supersulfide donors. Our findings indicate important roles for CARS in the production and metabolism of supersulfides-to mediate mitochondrial function and to regulate longevity.

Keywords: Cysteinyl-tRNA synthetase; ER stress; Longevity; Mitochondrial energy metabolism; Supersulfides.

Graphical abstract

Fig. 1

Supersulfide production by yeast cysteinyl-tRNA synthetase (Crs1). (A)In vivo formation of various supersulfides in the yeast WT and Δcrs1 strain. Endogenous production of supersulfides was identified by means of HPE-IAM labeling LC-ESI-MS/MS analysis (n = 3). **p < 0.01, ***p < 0.001; Student's t-test. (B) CysSSH production by the recombinant WT and K109A mutant protein (n = 3). ***p < 0.001; Student's t-test. (C) Translation activity of the yeast K109A strain as determined by Western blot analysis. Protein samples were prepared for detection of Pgk1 (nuclear encoded gene) and Cox2 (mitochondrial genome-encoded gene). (D) Supersulfide metabolome analysis with the yeast WT and K109A strain. *p < 0.05, **p < 0.01, ***p < 0.001; Student's t-test. CysSH, cysteine; CysSSH, cysteine persulfide; GSH, reduced glutathione; GSSH, glutathione persulfide; HSH, hydrogen sulfide; HSSH, hydrogen disulfide; HS₂O₃⁻, thiosulfate.

Fig. 2

In vivo formation of various supersulfides in the yeast WT, K109 mutant, and K109A + Crs1 strain. (A)In vivo formation of various supersulfides in the yeast WT, K109A mutant, and K109A + Crs1 strain. K109A + Crs1 strain was generated by introduction of the WT Crs1 transgene into the mutant strain. Endogenous production of supersulfides was identified by means of HPE-IAM labeling LC-ESI-MS/MS analysis (n = 3). *p < 0.05, **p < 0.01, ***p < 0.001 ****p < 0.0001; Multiple t-test. n.s., not significant. (B) Gene expression of oxidative stress marker in the WT, K109 mutant, and K109A + Crs1 strain. The mRNA of Gpx2, the oxidative stress marker, was quantified using qPCR in WT, K109A, and K109A + Crs1 strain at late log phase. *p < 0.05, ***p < 0.001; Multiple t-test. n.s., not significant.

Fig. 3

Phenotypic analysis of the WT and the CRS1 K109A mutant strain. (A) Growth curves of the WT and the CRS1 K109A mutant. Cell growth was determined at the indicated time points by measuring the OD₆₀₀. ***p < 0.001; two-way ANOVA with Tukey's test. (B) A chronological lifespan assay of the WT and the K109A mutant. Chronological lifespan was determined by using colony formation units. The K109A mutant showed a significant variance in survival rates compared with the WT. The K109A mutant strain with plasmid expressed WT Crs1 showed the same survival rate as the WT. *p < 0.05 vs. WT, ^#p < 0.05 vs. K109A; two-way ANOVA with Tukey's test. (C) Mitochondrial or cytosolic Crs1 expression vectors were transformed into the K109A mutants, and the chronological lifespans of the mutants were measured. As a control, pRS416 was transformed into the K109A mutant. *p < 0.05 vs. WT, ^#p < 0.05 vs. K109A; two-way ANOVA with Tukey's test. (D) Chronological lifespans were measured in heterozygous mutants of CARS (Δcrs1/WT), HARS (Δhts1/WT), and EARS (Δgus1/WT). *p < 0.05 vs. WT; two-way ANOVA with Tukey's test.

Fig. 4

Crs1-dependent ER stress response in aging. (A) HAC1 splicing during aging. cDNAs were made from the cultured cells, followed by individual PCR with primers in upper panel (HAC1 Fw and HAC1 Rv). The PCR products were electrophoresed to determine the HAC1 splicing (middle and lower panels). Arrows indicate primer positions. M: DNA marker, Tm: tunicamycin-treated sample as the positive control of ER stress. (B)KAR2 induction during aging. Conventional real-time quantitative PCR analysis was used, and cDNAs from the WT and the K109A strain were measured by using the KAR2 primers (given in Supplementary Table S2). **p < 0.01 vs. WT; two-way ANOVA with Tukey's test. (C) Alteration of PDI enzyme activity by means of reduction. Recombinant PDI protein was reduced with TCEP and treatment with or without Na₂S₂ treatment, followed by measurement of PDI enzymatic activity by using RNase as a substrate. (D) LC-ESI-Q-TOF-MS chromatograms obtained via proteome analysis of peptide fragments from the recombinant PDI protein, which included cysteine residues. The disulfide bond or trisulfide bond of cysteine in the active center of the PDI protein was identified by monitoring at m/z 835.6369 and 843.8814, respectively. Peptide fragments containing cysteine residues are shown at the tops of the panels. Cysteine residues in the fragments are indicated by red type. CysS-SCys, disulfide bond. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 5

Crs1-dependent mitochondrial energy metabolism in aging. (A) ATP contents of the WT and the K109A mutant were measured during aging. *p < 0.05 vs. WT; two-way ANOVA with Tukey's test. (B) ATP contents of the WT, the K109A mutant, and K109A mutant + Crs1 were measured at stationary phase. **p < 0.01, ***p < 0.01; Multiple t-test. n.s., not significant.

Fig. 6

Effects of supersulfide donors on longevity. (A) Schematic showing the method used for regulation of longevity by exogenous supersulfide. Yeast cells were cultured at 30 °C for 3 days in SC medium. Thereafter, the persulfide donor Na₂S₂ (20 μM) was added to the culture every 6 days until day 30. A small amount of the culture was removed and diluted and plated on YPD plates to determine colony formation after 2 days of incubation at 30 °C. (B) Chronological survival curves of the WT and the K109A mutant with or without Na₂S₂ were constructed by counting the number of colony-forming units. *p < 0.05 vs. WT control, ^#p < 0.05 vs. K109A control; two-way ANOVA with Tukey's test.