Diverse cell stresses induce unique patterns of tRNA up- and down-regulation: tRNA-seq for quantifying changes in tRNA copy number

Abstract

Emerging evidence points to roles for tRNA modifications and tRNA abundance in cellular stress responses. While isolated instances of stress-induced tRNA degradation have been reported, we sought to assess the effects of stress on tRNA levels at a systems level. To this end, we developed a next-generation sequencing method that exploits the paucity of ribonucleoside modifications at the 3'-end of tRNAs to quantify changes in all cellular tRNA molecules. Application of this tRNA-seq method to Saccharomyces cerevisiae identified all 76 expressed unique tRNA species out of 295 coded in the yeast genome, including all isoacceptor variants, with highly precise relative (fold-change) quantification of tRNAs. In studies of stress-induced changes in tRNA levels, we found that oxidation (H2O2) and alkylation (methylmethane sulfonate, MMS) stresses induced nearly identical patterns of up- and down-regulation for 58 tRNAs. However, 18 tRNAs showed opposing changes for the stresses, which parallels our observation of signature reprogramming of tRNA modifications caused by H2O2 and MMS. Further, stress-induced degradation was limited to only a small proportion of a few tRNA species. With tRNA-seq applicable to any organism, these results suggest that translational control of stress response involves a contribution from tRNA abundance.

Figure 1.

Two-step ligation method for the conversion of tRNA to cDNA. Bulk tRNAs extracted from S. cerevisiae are first 3′-end ligated with a DNA adaptor. After reverse transcription, a second DNA adaptor is ligated to the 3′-end of the resulting cDNA, and the resulting species are PCR amplified followed by standard Illumina sample preparation and analysis.

Figure 2.

Distribution of read lengths of untreated total small RNA. Normalized read counts (as proportion of total reads) from MiSeq analysis of total small RNA from unexposed cells were binned according to their read lengths before alignment of uniquely assignable sequences (dark gray bars) and after alignment (light gray bars).

Figure 3.

Hierarchical cluster analysis of either H₂O₂ or MMS induced changes in tRNA expression levels in cells. Expression levels of tRNAs extracted from either H₂O₂- or MMS-exposed cells were calculated from tRNA-Seq and compared to control tRNAs from untreated conditions. Hierarchical cluster analysis was performed on fold-change data. The bottom left color bar indicates the range of fold-change values. Asterisks denote significant changes at P < 0.1 in at least one exposure condition as determined by Student's t-test.

Figure 4.

Mismatch rates at m²₂G sites in S. cerevisiae tRNA. The mismatch error rate in tRNAs known to possess the modified base m²₂G at position 26 was quantified from tRNA-seq data for control (black bars), H₂O₂-exposed (white bars) and MMS-exposed (gray bars) cells. Data represent mean ± SD for six biological replicates, with P values determined by Student's t-test; SD: standard deviation.