Sulfate assimilation regulates hydrogen sulfide production independent of lifespan and reactive oxygen species under methionine restriction condition in yeast

High-throughput genetic screening-based identification of genes required for hydrogen sulfide production under MR condition

MR induces lifespan extension and triggers an increase in endogenous hydrogen sulfide levels [13]. Hydrogen sulfide is particularly considered to be an essential mediator of MR benefits; however, its role in MR still remains unclear. To address this, we exploited an MR media that was previously shown to extend the lifespan of yeast [6]. This MR media was first used to examine the change in hydrogen sulfide production in the presence of different methionine concentrations using a lead acetate paper assay. As expected, hydrogen sulfide production was observed to be consistently higher in yeast cells grown using MR media when compared to those grown using regular media (Fig. S1).

Next, we conducted high-throughput hydrogen sulfide screening for the MR media with the yeast knockout collection (Fig. 1a). For this screening process, we used the methylene blue assay, in which the blue color changes to colorless upon hydrogen sulfide production. During this screening, two optical densities were measured simultaneously; one at OD₆₆₃ to determine hydrogen sulfide production and another at OD₆₀₀ to determine cell mass. These measurements were then used to calculate hydrogen sulfide production per cell. Compared to that in regular media, most of the tested strains, including wild-type, consistently showed low cell mass during the stationary phase (Fig. S2) and increased hydrogen sulfide production (Fig. 1a) in the MR media. Consequently, average hydrogen sulfide production per cell dramatically increased in the MR media. To identify the genes that regulated the increase in hydrogen sulfide production under the MR condition, two analytical strategies were applied to the screening data (Fig. 1a): 137 deletion mutant strains were selected to produce lesser hydrogen sulfide than the wild-type in MR media (Table S2; fold-change < -1.5; FDR-adjusted p-value < 0.05) and 81 deletion mutant stains were selected to exhibit no increase in hydrogen sulfide production under MR condition (Table S3; hydrogen sulfide production ratio between MR and regular media < 1; FDR-adjusted p-value < 0.05). Interestingly, 50 genes resided in the intersection between two independently performed comparisons (Fig. 1a). Next, we carried out Gene Ontology analysis for a total of 168 genes obtained from both analyses and found that genes involved in methionine biosynthesis (11 genes; p-value = 0.0000) and vacuolar acidification (7 genes; p-value = 0.0331) were significantly enriched (Fig. 1b). Therefore, we surmised that these two pathways (methionine biosynthesis and vacuolar acidification) possibly played a role in regulating hydrogen sulfide production under MR condition.

Deletion of MET1, MET3, MET5 or MET10 reduces hydrogen sulfide production under MR condition

Based on the high-throughput methylene blue assay, methionine biosynthesis and vacuolar acidification were predicted to be involved in the regulation of hydrogen sulfide production under MR condition. Next, an alternative hydrogen sulfide detection assay was further performed to confirm these results owing to the pH dependency of methylene blue color change. Thus, we measured hydrogen sulfide production in deletion mutant strains associated with methionine metabolism (Fig. 2a) and vacuolar acidification under various methionine concentrations (50, 25, 12.5, and 5 mg/L) using lead acetate assay (Fig. 2b, S3, and S4). Methionine concentration showed a strong negative correlation with hydrogen sulfide production in wild-type yeast cells (Fig. 2b), which indicated reduced methionine uptake to be very effective in increasing hydrogen sulfide production. This phenomenon was also observed in most deletion mutant strains whose vacuolar acidification pathway genes were knocked out (Fig. S4). In contrast, all deletion mutant strains associated with sulfate assimilation, including met1Δ, met3Δ, met5Δ, and met10Δ, failed to increase hydrogen sulfide production to wild-type levels under MR conditions (Fig. 2b). These differences in hydrogen sulfide production were analyzed in following 3 ways and are shown in the 3D plot (Fig. 2c). For each strain, we subtracted the hydrogen sulfide amounts under regular condition from those under MR condition ([H₂S]_MR – [H₂S]_Reg; x-axis). For regular condition, we subtracted the wild-type hydrogen sulfide amount from each deletion mutant hydrogen sulfide amount ([H₂S]_{mutant in regular} – [H₂S]_{WT in regular}; y-axis). The values for z-axis ([H₂S]_{mutant in MR} – [H₂S]_{WT in MR}) were calculated in the same way as those for the y-axis, except that they were obtained under MR condition. Interestingly, the four deletion mutant strains exhibiting defective sulfate assimilation were observed in the bottom left corner of the 3D plot, thus showing greatly reduced hydrogen sulfide production when compared to the wild-type under MR condition. Consequently, it also meant that the sulfate assimilation pathway was crucial for increasing hydrogen sulfide production under MR condition. Therefore, we surmised that the sulfate assimilation pathway may possibly have a regulatory role in upregulating hydrogen sulfide production when methionine uptake is reduced.

Increased hydrogen sulfide production is not directly associated with lifespan extension of yeast under the MR condition

MR mimics many aspects of CR-mediated benefits, including lifespan extension and increased hydrogen sulfide production [10,13]. Increased hydrogen sulfide levels are particularly considered to be a mediator of lifespan extension under CR condition. In this regard, we assessed the effect of increased hydrogen sulfide production on lifespan under MR condition. Previously, DeLuna and his colleagues identified yeast aging genes from the stationary phase survival data of yeast single gene knockout mutants using an automated chronological lifespan assay [14]. Upon correlating our hydrogen sulfide production data against this genome-wide chronological lifespan data, no significant correlation between hydrogen sulfide production and lifespan was observed in the deletion mutant strains (correlation value was -0.009 and -0.151 under regular and MR conditions; Fig. S5). To demonstrate this experimentally, deletion mutants that exhibited defective sulfur assimilation and failed to increase hydrogen sulfide production in MR media were examined. Although these deletion mutants showed shorter lifespans than the wild-type in regular media, their lifespans were observed to extend at par with that of the wild-type under MR condition (Fig. 3 and Table S3). As a result, MR-mediated lifespan extension was concluded to be maintained independent of hydrogen sulfide production in all tested deletion mutants (Fig. 3f).

To better understand the role of hydrogen sulfide in lifespan, we added a hydrogen sulfide donor, sodium hydrosulfide (NaHS), to wild-type cells grown in regular and MR media, i.e., the cells were treated thrice (6, 24, and 48 h after inoculation) with 5 µM NaHS. However, lifespan was not observed to extend by adding this small concentration of NaHS under both regular and MR conditions (Fig. S6). Next, we tested the effect of a much higher concentration of NaHS (50 μM) on wild-type, met3Δ, and met5Δ strains at the culture starting point (Fig. 4 and Fig. S7). Whereas NaHS markedly increased hydrogen sulfide production, lifespan remained unchanged after NaHS treatment in all tested strains under regular and MR conditions. Therefore, we concluded that increased extracellular hydrogen sulfide production was not directly related to lifespan extension under the MR condition.

Recent studies suggested that hydrogen sulfide protected cells from oxidative stress by reducing ROS or increasing antioxidant production [15]. Hence, we examined the relation between ROS and MR-mediated lifespan extension and hydrogen sulfide production. Under regular condition, deletion mutants produced about 2 to 6-fold higher ROS levels than the wild-type (Fig. 5). However, no noticeable difference in the ROS levels was observed between mutants and the wild-type upon methionine restriction (Fig. 5). Thus, all strains, including the wild-type and deletion mutants, showed decreased ROS levels via MR. Although this suggested a relation between MR-mediated decrease in ROS levels and lifespan extension under MR condition, it did not necessarily associate the process with changes in hydrogen sulfide production via MR.

Yeast strains and culture media

Saccharomyces cerevisiae BY4741 (MATa his3Δ1 leu2Δ0 met15Δ0 ura3Δ0) and deletion strain collections (BY4741 background) were used in this study (EUROSCARF, Germany). Unless otherwise noted, yeast cells were grown in synthetic medium (Table S4) containing 50 mg/L or 5 mg/L of methionine for regular and MR conditions, respectively [21].

High-throughput genetic screening for hydrogen sulfide production

Hydrogen sulfide screening was conducted using methylene blue assay as described by Winter G and Curtin C [22] with minor modifications. Briefly, cells were precultured to the stationary phase in 200 μL of YPD (1% yeast extracts, 2% peptone, and 2% glucose) prior to inoculation. Assays were performed in a microtiter plate at a total volume of 250 μL per well. Each well contained 247 μL of medium and 3% methylene blue reaction mix [23]. The experiments were carried out in triplicate.

Two optical densities were measured: OD₆₆₃ for the hydrogen sulfide indicator and OD₆₀₀ for the cell mass of the wild-type and 3,502 single gene deletion strains. The raw data for hydrogen sulfide production were calculated using the following formula:

Hydrogen sulfide production of strain A = OD₆₆₃ of strain A / OD₆₀₀ of strain A

Selection of genes involved in hydrogen sulfide production under MR condition

To prevent bias from growth defective strains, we eliminated mutant strains exhibiting OD₆₀₀ values less than one-third the wild-type OD₆₀₀ value (Fig. S8). To identify single gene deletion mutants showing lower hydrogen sulfide production than the wild-type under MR condition, we compared the fold-change between mutant and wild-type hydrogen sulfide production and determined the p-value using Student’s t-test followed by false discovery rate (FDR) adjustment. When the adjusted p-value was less than 0.05 and the fold-change between mutant and wild-type was less than -1.5, genes (137 gene deletion strains; Table S1) were categorized to be significantly involved in hydrogen sulfide production under MR condition. We also compared hydrogen sulfide production in each mutant strain under MR and control conditions. When the ratio of hydrogen sulfide production between the MR and control conditions was less than 1.0 and FDR-adjusted p-value was less than 0.05, genes (81 gene deletion strains; Table S2) were categorized to be significantly involved in hydrogen sulfide increase via MR.

Lead acetate assay

Commercially prepared lead acetate strip papers (Merck Millipore, Germany) or manually prepared lead acetate strip papers (Whatman filter paper soaked in 300 mM lead acetate solution and then dried) were used to detect the hydrogen sulfide emitted from the yeast batch culture. The paper was attached to the culture flask lids and incubated for 1 to 12 h at 30 °C until the paper darkened (due to lead sulfide formation). A color representation from the raw images (Fig. S3) was picked using Photoshop and is presented (Fig. 2b).

Chronological lifespan assay

Lifespan assay was conducted using propidium iodide staining as described previously [24,25]. Three yeast colonies of each yeast strain were seeded into 10 mL of 2% glucose-containing rich media (YPD) and incubated overnight. The seed culture was re-inoculated into 20 mL of YPD and the yeast cells were grown until the end of the lifespan assay. As the indicated time-point on the viability graph, cells were harvested, washed with phosphate-buffered saline (PBS). And stained with 5 µg/mL of propidium for 20 min at 30 °C. Fluorescence was detected using FACS Verse (BD, USA).

ROS measurement

Cells were harvested at day 6 and incubated in PBS solution containing 50 μM dihydroethidium (Sigma, USA) at 30°C for 20 minutes. Samples were washed thrice with PBS solution and resuspended in 1 mL of PBS. Fluorescence signal was detected in the FL3 channel using flow cytometry (FACSVerse).