Profiling condition-specific, genome-wide regulation of mRNA stability in yeast

Abstract

The steady-state abundance of an mRNA is determined by the balance between transcription and decay. Although regulation of transcription has been well studied both experimentally and computationally, regulation of transcript stability has received little attention. We developed an algorithm, MatrixREDUCE, that discovers the position-specific affinity matrices for unknown RNA-binding factors and infers their condition-specific activities, using only genomic sequence data and steady-state mRNA expression data as input. We identified and computationally characterized the binding sites for six mRNA stability regulators in Saccharomyces cerevisiae, which include two members of the Pumilio-homology domain (Puf) family of RNA-binding proteins, Puf3p and Puf4p. We provide computational and experimental evidence that regulation of mRNA stability by these factors is modulated in response to a variety of environmental stimuli.

Fig. 1.

MatrixREDUCE data flow. Using microarray data for 758 pairs of experimental conditions and the downstream 200 nt of every yeast ORF as input to MatrixREDUCE, we identified PSAMs corresponding to putative mRNA stability regulators and inferred TFAPs across all microarray conditions. For each discovered PSAM, “responder analysis” (see Methods) produced a list of putative target genes that are likely to contain functional binding sites for each transfactor modeled by a PSAM.

Fig. 2.

Discovered 3′ PSAMs logos. The PSAMs were discovered by MatrixREDUCE using a large set of steady-state mRNA expression data and the 200 nt downstream of every ORF.

Fig. 3.

Gene Ontology analysis. The enrichment of the PSAM targets in each category was quantified by using the cumulative hypergeometric distribution. The color scale corresponds to Bonferroni-corrected P values. The ontology to which a category belongs is indicated by C (subcellular component), P (biological process), or F (molecular function). Only a representative selection of all significant functional categories is shown. The significantly enriched functional categories for each PSAM were generally consistent with their inferred activity profiles (Fig. 4).

Fig. 4.

Transfactor activity profiles (TFAPs). (A–D) The direction and degree of the correlation between PSAM match score and mRNA expression data are represented by a color scale ranging from bright blue (strong negative correlation) to bright yellow (strong positive correlation). These correlations correspond to changes in activity of the mRNA stability-regulating transfactors whose binding specificities are modeled by the PSAMs. The triangles represent the progression of time within time course experiments. References for the microarray data sets used for the displayed analysis are as follows: diauxic shift (31); stationary phase, ethanol, fructose, glucose, sucrose, menadione, heat shock, diamide, nitrogen depletion, DTT, and amino acid and adenine starvation (32); peroxide (35); proteosome inhibition by PS-341 (36); H-2k(b) expression and tunicamycin (38); rapamycin (39); and puf4 deletion transcriptional arrest (5).

Fig. 5.

Regulation of Puf3p activity in response to a change in carbon source. Shown are Northern blot analyses of the decay of MFA2 mRNA or the hybrid MFA2/COX17 mRNA expressed from wild-type or puf3Δ yeast grown in media containing 2% glucose or 2% ethanol. Minutes after transcriptional repression are indicated above the set of blots, with the half-lives (t_1/2) as determined from multiple experiments.

Fig. 6.

Regulation of Puf3p activity in response to rapamycin. Data from Northern blot analyses of MFA2 (A) or the hybrid MFA2/COX17 (B) mRNA decay are plotted, with minutes following transcriptional repression on the x axis and the fraction of RNA remaining as compared to the steady-state RNA level at time 0 on the y axis. Decay was monitored with or without rapamycin treatment for 60 min before transcriptional repression as follows. (A) MFA2 without rapamycin (filled square) and MFA2 with rapamycin (open square). (B) MFA2/COX17 without rapamycin (filled circle) and MFA2/COX17 with rapamycin (open circle). Data points are averages of multiple experiments.