Altered rRNA processing disrupts nuclear RNA homeostasis via competition for the poly(A)-binding protein Nab2

Abstract

RNA-binding proteins (RBPs) are key mediators of RNA metabolism. Whereas some RBPs exhibit narrow transcript specificity, others function broadly across both coding and non-coding RNAs. Here, in Saccharomyces cerevisiae, we demonstrate that changes in RBP availability caused by disruptions to distinct cellular processes promote a common global breakdown in RNA metabolism and nuclear RNA homeostasis. Our data shows that stabilization of aberrant ribosomal RNA (rRNA) precursors in an enp1-1 mutant causes phenotypes similar to RNA exosome mutants due to nucleolar sequestration of the poly(A)-binding protein (PABP) Nab2. Decreased nuclear PABP availability is accompanied by genome-wide changes in RNA metabolism, including increased pervasive transcripts levels and snoRNA processing defects. These phenotypes are mitigated by overexpression of PABPs, inhibition of rDNA transcription, or alterations in TRAMP activity. Our results highlight the need for cells to maintain poly(A)-RNA levels in balance with PABPs and other RBPs with mutable substrate specificity across nucleoplasmic and nucleolar RNA processes.

Figure 1.

Transcriptomes in RNA decay and ncRNA processing mutants are similar. (A) FISH images showing nuclear accumulation of poly(A)-RNA in control, rrp6Δ, dis3-1, csl4-ph, enp1-1 and srm1-ts cells after 90 min of growth at 37°C. Nucleus shown in blue (DAPi) and poly(A)-RNA in red. Scale bar = 2 μm. (B) Hierarchical clustering of genes significantly differently expressed at P <0.01, log₂(FC), in ts mutant strains after 90 min at 37°C with respect to control in ribo-depleted RNA libraries. Bar on the left indicates transcript identity, mRNA (gray) and pervasive transcripts (blue) with each row representing an individual transcript and each column a mutant. Red shades indicate positive and blue negative value changes in gene expression. log₂(FC) values greater than the absolute value of 6 are encoded as maximum saturation. (C) Correlation matrix displaying calculated correlation coefficients of transcriptome-wide log₂(FC) between strains after 90 min at 37°C in ribo-depleted RNA libraries. For B and C, log₂(FC) values correspond to the average from at least two biological replicates using all reads quantified against each feature.

Figure 2.

Polyadenylation of ncRNAs is increased in Csl4 and Enp1 mutants. (A) Plot displaying calculated log₂(FC) values for snoRNAs in csl4-ph, enp1-1 and srm1-ts cells after 90 min at 37°C with respect to control in both oligo-dT purified and ribo-depleted RNA libraries from at least two biological replicates. Average change across all transcripts is indicated by the solid line, dark circles indicate significance at P < 0.05, and gray dots were not significance at P < 0.05. (B) Metagene analysis showing average read density across snoRNA genes in both oligo-dT purified (left) and ribo-depleted (right) RNA-seq libraries with respect to the transcription start site (TSS) and transcription end site (TES). Data is presented as RPKM, reads per kilo base per million mapped reads, to normalize for gene length and library size. (C) Heatmap displaying calculated log₂(FC) values for rRNA transcripts generated from RDN25 (25S), RDN18 (18S), RDN58 (5.8S), and RDN5 (5S), plus associated transcribed spacers (ETS1, ITS1, ITS2, ETS2), within the rDNA locus (see schematic) in csl4-ph, enp1-1 and srm1-ts cells after 90 min at 37°C with respect to control in oligo-dT purified RNA libraries. Transcripts that could not be calculated due to a lack of sufficient read counts are indicated as not determined (n.d.).

Figure 3.

rRNA and snoRNA species are polyadenylated in csl4-ph and enp1-1 mutants and polyadenylated pre-rRNAs associate with Nab2. (A) Total RNA was extracted from control, csl4-ph, rrp6Δ, dis3-1, enp1-1, srm1-ts, and mex67-5 mutants after 90 min at 37°C. Poly(A)-RNA species were enriched from 770 μg of total RNA using oligo-dT and analyzed by northern blotting. Labeled oligonucleotide probes were used to detect indicated pre-rRNAs, rRNA processing intermediates, mature rRNAs, ACT1 pre- and mature mRNAs, and snoRNAs (U14 and snR30). Probe numbers are listed beside each image and sequences are provided in Supplementary Table S6. (B) RNA species co-isolating with affinity-purified Nab2-PrA in control, csl4-ph, enp1-1 and srm1-ts cells after 90 min at 37°C detected by northern blot analysis using indicated probes as in (A). Input material, flow through (FT), and Nab2 bound (eluate) samples are included for comparison. In A and B, panels marked with a * in the upper right hand corner were obtained following an extended exposure time. (C) Localization of plasmid expressed Nab2-GFP (pBM458) and Gar1-GFP (pBM547) compared to poly(A)-RNA detected by FISH in control, csl4-ph, enp1-1, and srm1-ts cells after 90 min at 37°C. Nucleus is shown in blue (DAPi), poly(A)-RNA in red, and the GFP-tagged fusion protein in green. Scale bar = 2 μm.

Figure 4.

Poly(A)-RNA accumulation and Nab2 localization phenotypes are influenced by rRNA synthesis and TRAMP-dependent polyadenylation. (A) FISH image showing localization of poly(A)-RNA (red) and Nab2-GFP (green, expressed from the endogenous NAB2 locus) in control, csl4-ph, enp1-1 strains 90 min at 37°C without (top, DMSO control) and with addition of rapamycin (bottom, 100 ng/ml rapamycin) 15 min prior to temperatures shift compared to DAPi (blue). (B) FISH image showing localization poly(A)-RNA (red) and Nab2-GFP (green, expressed from the endogenous NAB2 locus) in enp1-1 with a trf4Δ or trf5Δ deletion or csl4-ph with a trf5Δ deletion after 90 min at 37°C as compared to DAPi (blue). Scale bars = 2 μm.

Figure 5.

Nab2 protein interactions are altered in csh4-ph and enp1-1 cells. Nab2-PrA associated proteins identified by AP-MS from control, csl4-ph, and enp1-1 cells after 90 min at 37°C. Proteins were included if identified in at least one of the tested strains as a high confidence interaction (i.e., at least five exclusive spectrum counts (ESC) in two biological replicates). Each protein is presented as a colored shape, with the color indicating an associated biological process (see legend). Shapes are defined as follows: • indicates a protein matching the defined high confidence criteria in both samples being compared; ✦ indicates a protein that was identified in both samples, but did not meet the high-confidence criteria in one of the samples being compared; ▴ indicates a protein identified in both samples, but at a level that is below the high-confidence criteria for both samples being compared; ▪ indicates a protein identified in only one of the samples with high confidence; ★ indicates a protein identified in only one of the samples at a level below the defined criteria. Due to the inability to calculate a log₂(FC) in the cases where no peptides were identified (▪ and ★), these data points are located to the extremes of the graph to indicate the fold change is >10 or ←10.

Figure 6.

Poly(A)-RNA accumulation is decreased, but Nab2-ncRNA associations persist in an enp1-1 strain overexpressing Pab1. (A) FISH image showing localization of poly(A)-RNA (red) in control, csl4-ph, enp1-1, and mex67-5 strains with or without overexpression of Nab2-GFP (green) with respect to DAPi (blue) after incubation at 37°C for 90 min. Nab2-GFP expression was induced by removal of methionine from the growth media for 2 h prior to temperature shift. Scale bar = 2 μm. (B–D) FISH image showing localization of poly(A)-RNA (red) in control, csl4-ph, enp1-1, and mex67-5 strains overexpressing Yra1-GFP (pBM763) or Pab1-GFP (pBM766) in green from a 2μ-plasmid versus the empty plasmid control (pBM005) with respect to DAPi (blue) after incubation at 37°C for 90 min. In panel D, the arrowhead denotes a cell with low Pab1-GFP expression that also shows a poly(A)-RNA accumulation phenotype. Scale bars = 2 μm. (E) Northern blots showing total RNA, poly(A)-RNA, and RNA species co-isolating with affinity-purified Nab2-PrA extracted from control, csl4-ph, enp1-1, and srm1-ts cells overexpressing Pab1 (pBM874) from a 2μ-plasmid versus an empty vector control (pBM007) after 90 min at 37°C. For the Nab2-PrA associated material, input material, flow through (FT), and Nab2 bound (eluate) samples are included for comparison. Probe numbers are listed beside each image and sequences are provided in Supplementary Table S6. Panels marked with a * in the upper right hand corner were obtained following an extended exposure time.

Figure 7.

Excess PABP activity suppresses secondary enp1-1 phenotypes. (A) Cumulative distribution plots of log₂(FC) values for the indicated classes of transcripts observed for control (black lines), csl4-ph (blue lines), and enp1-1 (green lines) strains overexpressing Pab1-GFP from a 2μ-plasmid (pBM766, dashed lines) versus an empty vector control (pBM005, solid line). (B) Metagene analysis showing average read density across snoRNA genes in control (black lines), csl4-ph (blue lines), and enp1-1 (green lines) strains overexpressing Pab1-GFP from a 2μ-plasmid (pBM766, dashed lines) versus an empty vector control (pBM005, solid line) with respect to the transcription start site (TSS) and transcription end site (TES). Data are presented as RPKM, reads per kilo base per million mapped reads, to normalize for gene length and library size. For both A and B, values were calculated using read data from three biological replicates as compared to the control strain with the empty plasmid as measured by 3′-TAG-RNA-seq. (C) Relative maximum growth rate of control, csl4-ph, enp1-1, and mex67-5 strains in liquid culture when overexpressing Pab1-GFP (pBM766) from a 2μ-plasmid at 32°C as compared to the same strains carrying an empty vector (pBM005). P-values for control versus csl4-ph (P = 0.0070, t = 5.091, and df = 4), control versus enp1-1 (P = 0.0009, t = 8.760, and df = 4), and control versus mex67-5 (P = 0.5593, t = 0.6360, and df = 4) were calculated from an unpaired t-test with a two-tailed distribution from three biological replicates.