Pervasive and dynamic protein binding sites of the mRNA transcriptome in Saccharomyces cerevisiae

Abstract

Background: Protein-RNA interactions are integral components of nearly every aspect of biology, including regulation of gene expression, assembly of cellular architectures, and pathogenesis of human diseases. However, studies in the past few decades have only uncovered a small fraction of the vast landscape of the protein-RNA interactome in any organism, and even less is known about the dynamics of protein-RNA interactions under changing developmental and environmental conditions.

Results: Here, we describe the gPAR-CLIP (global photoactivatable-ribonucleoside-enhanced crosslinking and immunopurification) approach for capturing regions of the untranslated, polyadenylated transcriptome bound by RNA-binding proteins (RBPs) in budding yeast. We report over 13,000 RBP crosslinking sites in untranslated regions (UTRs) covering 72% of protein-coding transcripts encoded in the genome, confirming 3' UTRs as major sites for RBP interaction. Comparative genomic analyses reveal that RBP crosslinking sites are highly conserved, and RNA folding predictions indicate that secondary structural elements are constrained by protein binding and may serve as generalizable modes of RNA recognition. Finally, 38% of 3' UTR crosslinking sites show changes in RBP occupancy upon glucose or nitrogen deprivation, with major impacts on metabolic pathways as well as mitochondrial and ribosomal gene expression.

Conclusions: Our study offers an unprecedented view of the pervasiveness and dynamics of protein-RNA interactions in vivo.

Figure 1

gPAR-CLIP identifies transcriptome-wide RBP crosslinking sites. (a) Schematic of the gPAR-CLIP protocol. (b) Reproducibility of crosslinking sites generated from replicate gPAR-CLIP libraries prepared from yeast grown in synthetic defined media (abbreviated as WT gPAR-CLIP hereafter). Pearson correlation coefficient is indicated. Inset: distribution of log₂ crosslinking site RPM ratios between replicates. Replicate error σ = 1.3-fold. (c) Length distribution of crosslinking sites in WT gPAR-CLIP libraries. Dotted line: average crosslinking site length of 23 nucleotides. (d) Pearson correlation coefficients of total mRNA-seq and gPAR-CLIP read coverage between 5' UTR, CDS, and 3' UTR regions as well as correlation coefficients of ribosome depleted (-ribosome) mRNA-seq and gPAR-CLIP read coverage between replicate WT libraries.

Figure 2

gPAR-CLIP captures known RBP crosslinking signatures. (a) Overlap of crosslinking sites identified in Puf3p PAR-CLIP and wild-type (WT) gPAR-CLIP. Puf3p PAR-CLIP crosslinking sites with >1% T-to-C conversion rate (Additional file 8) were considered captured by gPAR-CLIP if at least 50% of their nucleotides overlapped with a WT gPAR-CLIP crosslinking site with FDR <1%. (b) Identification of known Puf3p binding sites on COX17 mRNA in WT gPAR-CLIP and Puf3p PAR-CLIP. (c-e) Aggregate gPAR-CLIP crosslinking site coverage of the first 300 nucleotides of 2,626 annotated 5' UTRs (c), 51 annotated ribosomal gene introns centered at the branch point (BP) 3' end (d), and 4,241 3' UTRs centered on the poly(A) junction (e).

Figure 3

RBP crosslinking sites exhibit global sequence conservation. (a) Average ribosome depleted mRNA-seq and gPAR-CLIP read distributions across 5' UTR, CDS, and 3' UTR regions for all libraries. Error bar: 1 standard deviation. RPKM: reads per million mapped reads per kilobase. (b) Cumulative distribution of CLS values from wild-type (WT) libraries. (c) Proportion of Ts in crosslinking site binned by crosslinking site coverage (reads per million mapped reads (RPM)). The dashed red line indicates average T content of all crosslinking sites. (d) Number of conserved blocks in 3' and 5' UTRs overlapping 100% with WT gPAR-CLIP crosslinking sites (χ² P-values indicated). Control blocks were randomly generated within 3' and 5' UTRs to match the number and size of conserved blocks. (e) Two major gPAR-CLIP crosslinking sites in ATG8 3' UTR (top) and TOM40 3' UTR (bottom) overlapping conserved blocks. (f) Mean phastCons scores for Ts ranked and binned by CLS. Control lines represent mean phastCons scores of randomly ranked and binned Ts with no CLS, repeated ten times.

Figure 4

RBP crosslinking sites share global structural characteristics. (a) Mean unpaired probability scores for Ts ranked and binned by CLSs. Control lines represent mean unpaired probability of randomly ranked and binned Ts with no CLS. Pearson correlation coefficients: 5' UTR R² = 0.933, CDS R² = 0.976, 3' UTR R² = 0.986. (b) Crosslinking site pairedness visualized as a heatmap. Columns represent nucleotide positions within crosslinking sites. Rows represent average unpaired probability for 100 crosslinking sites in that bin. Select secondary structure predictions from low, middle, and high CLS regions are indicated with the crosslinking site colored. (c) Percentage of Ts ranked and binned by CLSs in conserved secondary structural elements as defined by RNAz [43,44]. Control lines represent percentage of randomly ranked and binned Ts with no CLS in conserved secondary structural elements.

Figure 5

Nutrient deprivation induces widespread but distinct RBP-crosslinking and mRNA changes. (a,b) Changes in 3' UTR crosslinking site coverage upon glucose (a) or nitrogen (b) starvation. Standard deviations of intra-replicate variation: wild-type (WT), 1.31-fold; glucose starvation, 1.24-fold; nitrogen starvation, 1.15-fold. (c) Overlap of 3' UTR crosslinking site changes affected by glucose or nitrogen starvation conditions. (d,e) Global changes in mRNA abundance upon glucose (d) or nitrogen (e) starvation. (f) Overlap of mRNAs with 3' UTR crosslinking site changes affected by glucose or nitrogen starvation conditions. (g,h) Enriched Gene Ontology terms for mRNAs with 3' UTR crosslinking sites with decreased (g) or increased (h) RBP occupancy upon glucose or nitrogen starvation or both. Grey lines indicate P-value of 0.05.

Figure 6

Glucose starvation induces RBP-crosslinking and mRNA changes associated with mitochondrial processes. (a) Changes in 3' UTR crosslinking site coverage versus changes in the corresponding mRNA upon glucose starvation. Crosslinking sites on genes annotated with the 'mitochondrial membrane' GO term are colored blue. Dotted lines indicate ≥4-fold changes in crosslinking site coverage (vertical) or ≥2-fold change in mRNA expression (horizontal). (b) ALD4 3' UTR contains four crosslinking sites that decrease two- to eight-fold in RBP occupancy upon glucose starvation and overlap with conserved blocks (red diamonds in (a)). ALD4 mRNA expression is up-regulated upon glucose starvation. (c) STM1 3' UTR contains one crosslinking site that increases in coverage upon glucose starvation (red diamond in (a)). STM1 mRNA expression is down-regulated upon glucose starvation. WT, wild type.

Figure 7

Nitrogen starvation induces specific RBP-crosslinking and mRNA changes associated with ribosomes and translation-related processes. (a) Global changes in 3' UTR crosslinking site coverage versus changes in the corresponding mRNA upon nitrogen starvation. Crosslinking sites on genes annotated with 'ribosome biogenesis' GO term are colored red. Dotted lines indicate ≥4-fold changes in crosslinking site coverage (vertical) or ≥2-fold change in mRNA expression (horizontal). (b) INO1 3' UTR contains one crosslinking site that increases in coverage upon nitrogen starvation and falls within a conserved block (blue diamond in (a)). INO1 mRNA expression is down-regulated upon nitrogen starvation. (c) AGP3 3' UTR contains three crosslinking sites that are lost and two crosslinking sites that appear upon nitrogen starvation (blue diamonds in (a)). AGP3 mRNA expression is up-regulated upon nitrogen starvation.