2'-O-ribose methylation of transfer RNA promotes recovery from oxidative stress in Saccharomyces cerevisiae

Abstract

Chemical modifications that regulate protein expression at the translational level are emerging as vital components of the cellular stress response. Transfer RNAs (tRNAs) are significant targets for methyl-based modifications, which are catalyzed by tRNA methyltransferases (Trms). Here, Saccharomyces cerevisiae served as a model eukaryote system to investigate the role of 2'-O-ribose tRNA methylation in the cell's response to oxidative stress. Using 2'-O-ribose deletion mutants for trms 3, 7, 13, and 44, in acute and chronic exposure settings, we demonstrate a broad cell sensitivity to oxidative stress-inducing toxicants (i.e., hydrogen peroxide, rotenone, and acetic acid). A global analysis of hydrogen peroxide-induced tRNA modifications shows a complex profile of decreased, or undetectable, 2'-O-ribose modification events in 2'-O-ribose trm mutant strains, providing a critical link between this type of modification event and Trm status post-exposure. Based on the pronounced oxidative stress sensitivity observed for trm7 mutants, we used a bioinformatic tool to identify transcripts as candidates for regulation by Trm7-catalyzed modifications (i.e., enriched in UUC codons decoded by tRNAPheGmAA). This screen identified transcripts linked to diverse biological processes that promote cellular recovery after oxidative stress exposure, including DNA repair, chromatin remodeling, and nutrient acquisition (i.e., CRT10, HIR3, HXT2, and GNP1); moreover, these mutants were also oxidative stress-sensitive. Together, these results solidify a role for TRM3, 7, 13, and 44, in the cellular response to oxidative stress, and implicate 2'-O-ribose tRNA modification as an epitranscriptomic strategy for oxidative stress recovery.

Fig 1. 2'-O-methylation in Saccharomyces cerevisiae.

2’-O-ribose methyltransferases and target nucleotides (black-filled) in a two-dimensional cloverleaf tRNA structure based on the conventional numbering system (Machnicka, Olchowik et al. 2014). Also shown are other commonly modified nucleotides in tRNAs (dot-filled).

Fig 2. Growth and survival of 2'-O-ribose trm mutants is inhibited in the presence of oxidative stress-inducing agents.

A) Wild type (wt) BY4741 and trm mutant strains were grown from a single colony overnight and then serially diluted in water as indicated. Five μl of each dilution was then plated on YPD agar plates containing acetic acid (20 mM), rotenone (20 μM), and hydrogen peroxide (2 mM), and the cells were grown for 3 days at 30°C. B) Cells were treated with hydrogen peroxide (40 mM), rotenone (400 μM), and acetic acid (80 mM) and non-viable cells were identified by trypan blue staining four and eight hours post-exposure. Untreated cells (i.e., zero hours post-exposure) served as the control. Cell death is shown as the mean number of trypan blue positive cells divided by total cells (i.e., trypan blue positive plus negative), expressed as a percent, with error bars representing the standard deviation of the mean (n = 3). Significant differences in cell death percentages were determined using the t-test: trm mutants versus wild type under the same treatment condition and at the same time post-exposure (n = 3, *P < 0.05, **P < 0.01, ***P < 0.001).

Fig 3. Colony survival is decreased in 2'-O-ribose trm mutants after exposure to oxidative stress-inducing agents.

A) Wild type (BY4741), trm3, trm7, trm13, and trm44 mutant strains were grown to stationary-phase and exposed to acetic acid (20 mM), rotenone (100 μM), and hydrogen peroxide (10 mM). One hour post-exposure, the cultures were serially by a factor of 10⁻⁵, plated (100 μl), and the number of colonies was recorded after growth at 30°C for three days. Percent colony survival is shown as the mean for the toxicant relative to each individual strain under untreated conditions; error bars are standard deviations (n = 6 for wild type and n = 3 for trm mutants). Significant differences in colony formation and survival were determined using the t-test: trm mutants versus wild type under the same treatment condition (*P < 0.05, **P < 0.01, ***P < 0.001). B) trm mutants strains (as indicated) were transformed with their corresponding galactose (Gal)-inducible open reading frames (+ TRM-ORF). Cells grown to stationary phase in selection media (-Ura) + raffinose (carbon source) were diluted (3 fold) in media (-Ura) containing either galactose (+ Gal) or dextrose (+ Dex), and allowed further growth to stationary phase. Then, cells were split into separate tubes for one-hour control (water) or hydrogen peroxide (20 mM) treatments at 30°C, and similarly diluted (i.e., 10⁻⁵), plated, with colonies recorded three days later. Percent colony survival is shown as the mean ± standard deviations (TRM3 and TRM13, n = 3; TRM7, n = 4). Significant differences in colony survival were determined using a Student t-test: cells exposed to hydrogen peroxide, +Gal (TRM "on") versus +Dex (TRM "off", *P < 0.005, **P <0.001).

Fig 4. Analysis of 2'-O-ribose methylation in S. cerevisiae exposed to hydrogen peroxide.

Wild type (wt) BY4741 and trm strains were cultured to log-phase growth (OD_600nm = 0.6) and either left untreated (open bars) or exposed to hydrogen peroxide (20 mM for 1 hour, grey-filled bars). Ribonucleoside modifications were identified by direct infusion electrospray ionization mass spectrometry (ESI-MS). Quantitative measurements of 2'-O-methyladenosine (Am, panel A), 2'-O-methyluridine (Um, panel B), 2'-O-methylcytidine (Cm, panel C), and 2'-O-methylguanosine (Gm, panel D) were made using the abundance versus proxy method (AvP), defined as the absolute signal intensity over the sum of the absolute intensities of the canonical bases. Fold changes represent increases or decreases in AvP values relative to wt untreated cells and were calculated using the AvP values from S1 and S2 Tables, with error bars indicating the standard error of the mean change (n = 3).

Fig 5. 2'-O-ribose trm mutants have elevated levels of intracellular reactive oxygen species (ROS).

The indicated trm mutant strains were cultured to log-phase growth (OD_600nm = 0.6) and stained with 2',7'-dichlorofluorescin (DCFDA) to measure intracellular ROS. The fluorescence intensity emitted by oxidized DCFDA at 517–527 nm was determined by flow cytometry (≥ 4000 cells per population), and is shown as the mean ± standard deviation (n = 4). Significant differences in mean fluorescence intensities were determined using the t-test: trm mutants versus wild type under similar treatment conditions (**P < 0.01, ***P < 0.001).

Fig 6. Colony survival is decreased in UUC-enriched deletion mutants after exposure to oxidative stress-inducing agents.

Wild type (BY4741), crt10, fre5, hxt2, hir3, and gnp1 mutant strains were grown to stationary-phase (i.e., OD₆₀₀ = 3.0) and exposed to hydrogen peroxide (10 mM) to induce oxidative stress. After 1 hour of exposure, the cultures were serially diluted to 9 x 10² cells per ml, plated (100 μl), and the number of colonies was recorded after growth at 30°C for three days. In the central figure, percent survival is shown relative to untreated wt. The inset shows the magnitude of the decrease in survival (i.e., percent survival untreated minus treated conditions for individual strains). Significant differences in colony formation and survival were determined using a Student t-test: UUC-enriched mutants versus wild type under similar treatment conditions (n = 5, *P < 0.002, **P < 0.001).

Fig 7. A hypothetical model for how 2'-O-ribose tRNA methylation promotes oxidative stress recovery.

In this model, 2'-O-ribose Trms are responsive to oxidative stress exposure by unknown signaling events or pathways as indicated by the question mark. Trms then target various tRNAs for 2'-O-ribose methylation at positions that have previously been implicated in increased tRNA stability, amino acid charging, and enhanced translation of transcripts enriched in certain codons. Possible biological processes that may be involved in oxidative stress recovery via this type of translational regulation were identified at a bioinformatic level. The precise role of Trm-dependent 2'-O-ribose methylation for enhanced decoding of transcripts associated with these biological processes has yet to be determined.