Nonsense-mediated mRNA decay controls the changes in yeast ribosomal protein pre-mRNAs levels upon osmotic stress

Abstract

The expression of ribosomal protein (RP) genes requires a substantial part of cellular transcription, processing and translation resources. Thus, the RP expression must be tightly regulated in response to conditions that compromise cell survival. In Saccharomyces cerevisiae cells, regulation of the RP gene expression at the transcriptional, mature mRNA stability and translational levels during the response to osmotic stress has been reported. Reprogramming global protein synthesis upon osmotic shock includes the movement of ribosomes from RP transcripts to stress-induced mRNAs. Using tiling arrays, we show that osmotic stress yields a drop in the levels of RP pre-mRNAs in S. cerevisiae cells. An analysis of the tiling array data, together with transcription rates data, shows a poor correlation, indicating that the drop in the RP pre-mRNA levels is not merely a result of the lowered RP transcription rates. A kinetic study using quantitative RT-PCR confirmed the decrease in the levels of several RP-unspliced transcripts during the first 15 minutes of osmotic stress, which seems independent of MAP kinase Hog1. Moreover, we found that the mutations in the components of the nonsense-mediated mRNA decay (NMD), Upf1, Upf2, Upf3 or in exonuclease Xrn1, eliminate the osmotic stress-induced drop in RP pre-mRNAs. Altogether, our results indicate that the degradation of yeast RP unspliced transcripts by NMD increases during osmotic stress, and suggest that this might be another mechanism to control RP synthesis during the stress response.

Figure 1. Tiling array analysis of exon and intron signals after osmotic stress in relation to non-stress.

(A) Schema of the PM index (PMi) calculation using the intron intensity signal ratio after a 15-minute incubation in the presence of 0.4 M NaCl (stressed cells) in relation to non-stressed cells. The mRNAs without changes, with increases or with a decrease, in the intron signal in relation to the exon signal between the stress and non-stress conditions take values close to zero, or positive or negative PMi values, respectively. All these situations are illustrated with a real example (bottom panels). (B) Global distribution of ICGs (intron-containing genes) according to their PMi (see Material and Methods). The black boxes in the histogram represent the distribution of intron-containing ribosomal protein (RP) genes.

Figure 2. Quantitative RT-PCR analysis of PMi for ribosomal protein genes in wild-type and hog1 or cbc1 mutants.

(A) A decrease in ribosomal protein (RP) pre-mRNAs in response to osmotic stress. The behaviour of transcripts RPL30, RPL28, RPL33A, RPL33B, RPL34A and YDR367W across a time course lasting 5, 10 and 15 minutes after treatment with 0.4 M NaCl was examined by quantitative RT-PCR using specific primers to intron regions and exon regions. The PMi value of the comparison made between stressed (0.4 M NaCl, 15 min) and non-stressed cells per gene, obtained by quantitative RT-PCR, is found in parentheses. Data and error bars represent the average and standard deviation of 3 independent experiments. (B) Representation of PMi for the wild-type, hog1 mutant and cbc1 mutant cells treated with 0.4 M NaCl for 15 minutes in relation to untreated cells. The intron and exon levels of the transcripts RPL30, RPL28, RPL37A, RPS13, RPL33A, RPL34A and RPL33B were examined by quantitative RT-PCR using specific primers to those regions. Data and errors bars represent the average and standard deviation of 3 independent experiments.

Figure 3. Analysis of the relation between ribosomal protein genes transcription and pre-mRNA processing during osmotic stress.

(A) Clustering RP genes according to their transcription rate (TR) profiles. Wild-type cells were grown in YPD until the exponential phase and were then treated with 0.4 M NaCl. The samples taken at 0, 2, 4, 6, 8, 10 and 15 minutes of osmotic shock were processed to measure the TR of all the yeast genes. The data set series for intron-containing RP genes refers to their respective 0 time on a logarithm scale. A relative repression (stress/non-stress ratio, on the log₂ scale) is shown in green (saturated green indicates a decrease of at least 6-fold) and relative induction is depicted red (saturated red indicates an increase of at least 6-fold). RP genes were ordered and grouped into three distinct subclasses by K-means clustering using Euclidean distance. The PMi for all the RP genes are represented in the left panel using a blue-yellow heat map (saturated blue indicates a PMi of ≤−1, while saturated yellow indicates a PMi of ≥1). Grey squares indicate missing values. (B) Comparison of RP genes ΔTR and their PMi in the cells treated with 0.4 M NaCl for 15 minutes related to non-stressed cells. The regression line, equation and correlation coefficient are shown in the graph. The RP genes that have been analyzed by quantitative RT-PCR are indicated as red diamonds.

Figure 4. Upf1 and exonuclease Xrn1 contribute to the decrease in ribosomal protein pre-mRNA during osmotic stress.

(A) The total and pre-mRNA relative levels are shown for the non-stressed cells of a wild-type strain and strains with deletions in RRP6, XRN1 or UPF1. (B) Representation of PMi in the cells treated with 0.4 M NaCl for 15 minutes in relation to untreated cells for the wild-type strain and the same mutant strains as in (A). The intron and exon levels of the transcripts RPL30, RPL28, RPL37A, RPS13, RPL33A, RPL34A, RPL33B and YDR367W were examined by quantitative RT-PCR using specific primers to those regions. Data and errors bars represent the average and standard deviation of 3 independent experiments.

Figure 5. Kinetics of ribosomal protein pre-mRNAs during osmotic stress in cells containing mutations in NMD components.

The behaviour of the transcripts RPL30, RPL28, RPL33A, RPL33B, RPL34A and YDR367W after 0, 5, 10 and 15 minutes treatment with 0.4M NaCl were examined in the wild-type, upf1, upf2 and upf3 and xrn1 mutants by quantitative RT-PCR, as described in Figure 2A. Data and errors bars represent the average and standard deviation of 3 independent experiments.

Figure 6. Scheme of the changes in ribosomal protein (RP) gene expression upon osmotic stress.

The RP pre-mRNA steady-state depends on both the RP gene transcription rate (TR) by RNA polymerase II and the RP pre-mRNA splicing and degradation rates. The RP mRNA produced by splicing can be used for the synthesis of ribosomal proteins through translation or can be sent to degradation. Osmotic stress provokes a decrease in the RP gene TR and an NMD-dependent increase in the RP pre-mRNA degradation rate. Additionally, osmotic stress increases the RP mRNA degradation rate (lower than the increase in the pre-mRNA degradation rate) and lowers the RP mRNA translation rate. The effect of osmotic stress on RP pre-mRNA splicing and RP decay has not been directly estimated. As a result of these changes, the RP pre-mRNA levels dropped considerably and, to a lesser extent, the RP mRNA level also lowered under osmotic stress. However, RP levels remained constant as a result of the growth inhibition provoked by osmotic stress. See the text for further discussion.