The nuclear RNA polymerase II surveillance system targets polymerase III transcripts

Abstract

A key question in nuclear RNA surveillance is how target RNAs are recognized. To address this, we identified in vivo binding sites for nuclear RNA surveillance factors, Nrd1, Nab3 and the Trf4/5–Air1/2–Mtr4 polyadenylation (TRAMP) complex poly(A) polymerase Trf4, by UV crosslinking. Hit clusters were reproducibly found over known binding sites on small nucleolar RNAs (snoRNAs), pre-mRNAs and cryptic, unstable non-protein-coding RNAs (ncRNAs) ('CUTs'), along with ~642 predicted long anti-sense ncRNAs (asRNAs), ~178 intergenic ncRNAs and, surprisingly, ~1384 mRNAs. Five putative asRNAs tested were confirmed to exist and were stabilized by loss of Nrd1, Nab3 or Trf4. Mapping of micro-deletions and substitutions allowed clear definition of preferred, in vivo Nab3 and Nrd1 binding sites. Nrd1 and Nab3 were believed to be Pol II specific but, unexpectedly, bound many oligoadenylated Pol III transcripts, predominately pre-tRNAs. Depletion of Nrd1 or Nab3 stabilized tested Pol III transcripts and their oligoadenylation was dependent on Nrd1–Nab3 and TRAMP. Surveillance targets were enriched for non-encoded A-rich tails. These were generally very short (1–5 nt), potentially explaining why adenylation destabilizes these RNAs while stabilizing mRNAs with long poly(A) tails.

Figure 1

Nuclear RNA surveillance factors are crosslinked to many RNA targets. (A–D) High-throughput sequencing of cDNA libraries generated from crosslinked RNAs associated with purified Nrd1–HTP (based on 3.1 M mapped reads), Nab3–HTP (5.8 M mapped reads), Trf4–HTP (6.5 M mapped reads) and a no-tag control (23 K mapped reads). Sequencing data were mapped to the yeast genome using NOVOALIGN. Pie charts illustrate proportion of mapped reads corresponding to classes of RNAs in the indicated IP. The results from two independent, high-throughput analyses are combined. (E) Single track distribution of CRAC hits of the indicated IP along chromosome IV. (F) Bar diagram representing percentage of all mapped sequences carrying at least three non-encoded terminal A residues in the indicated purification from the combined data sets. (G) Length distribution of non-encoded tails on RNAs associated with the indicated proteins using the combined data sets and allowing up to 20% non-A residues. See also Supplementary Figure S1F and G.

Figure 2

Nrd1 and Nab3 bind consensus motifs. (A) Statistical overrepresentation scores of 4 mer sequences in Nab3 and Nrd1. (B–D) Density of clusters (B), deletions (C) and substitutions (D) around Nab3-binding motifs UCUU/CUUG or Nrd1-binding motifs GUAG/GUAA. See also Supplementary Figure S1H.

Figure 3

Crosslinking experiments recover known targets for the nuclear RNA surveillance machinery. Densities in hits per million of high-throughput sequencing reads of the indicated IP mapped to SNR13 (A), SNR3 (B), NRD1 (C), CTH2 (D), IGS1-R CUT (E) and SRG1-SER3 (F). Open reading frames are represented by arrows, ncRNAs by boxes. UTRs and terminator elements are indicated where appropriate. (E) 5S and 25S transcripts are represented by bold arrows and cryptic ncRNAs by slim arrows. See also Supplementary Figure S3.

Figure 4

Long non-coding asRNAs are targets for the nuclear RNA surveillance machinery. Densities of reads mapped to HPF1as RNA (A) and CAF17as RNA (B). Arrows represent major asRNA species, read-through products as detected in nrd1 and nab3 mutants (C, D) are illustrated as dashed arrows. (C, D) Northern analyses of total RNA from BY4741, GAL∷nrd1, GAL∷nab3, GAL∷nrd1 GAL∷nab3, trf4Δ, trf5Δ and rrp6Δ strains. Riboprobes were directed against indicated RNA species; sizes of detected RNAs are given. Ethidium bromide stain of 18S rRNA serves as loading control. (^*) Marks an unknown contaminant RNA. See also Supplementary Figure S4.

Figure 5

CUTs are preferentially targeted by Nrd1 and Nab3. Density of high-throughput sequencing reads mapped to CUTs, SUTs, mRNAs and intergenic regions, according to Xu et al (2009).

Figure 6

Pre-Rrp1 is polyadenylated and targeted by the surveillance machinery. (A) Densities of high-throughput sequencing reads mapped to pre-RPR1. (B) Two-dimensional structure of RPR1 according to Srisawat et al (2002). High-throughput sequencing reads mapped to RPR1 are highlighted. Trf4 crosslinks overlapped with Nab3 hits and have been omitted for clarity. (C) Alignment of high-throughput sequencing reads of RNAs in the indicated IP to pre-RPR1. Grey boxes mark the mature RPR1 sequence and numbering indicates the nucleotide position with respect to nucleotide +1 in the RPR1 gene. Mismatches and deletions in sequencing reads are displayed in red. Numbers in brackets indicate the frequency with which each specific sequence was recovered in reads per million mapped sequences. (D) Northern analyses of total and poly(A)⁺ RNA from GAL∷nrd1, GAL∷nab3 and BY4741 strains. Quantification of expression levels of polyadenylated pre-RPR1 relative to TSA1 mRNA is displayed. Ratio of expression after 12 h compared with 0 h is set to 1 for WT and given as average of three biological replicates with s.d.

Figure 7

Pre-tRNAs are targets for the nuclear RNA surveillance machinery. (A) Average densities of reads mapped to intron-containing tRNAs. tRNA exons and introns have various lengths; all exons and introns were divided into three bins and density of reads in each bin is displayed. (B, C) Average densities of reads of all tRNAs associated with the indicated proteins are plotted with respect to start (B) and end of the tRNA (C). (D, E) Alignments of high-throughput sequencing reads of RNAs associated with the indicated proteins to 3′ extended pre-tRNAs (D) and 5′ extended pre-tRNAs (E). Grey boxes mark mature tRNA sequences and numbering indicates nucleotide positions with respect to nucleotide +1 of the tRNA. Mismatches and deletions in sequencing reads are displayed in red. Nrd1 and Nab3 consensus-binding motifs are underlined. Numbers in brackets indicate the frequency with which each specific sequence was recovered in reads per million mapped sequences. (F, G) Northern analyses of total RNA from BY4741, GAL∷nrd1CIDΔ, GAL∷nrd1 and GAL∷nab3 strains. Oligonucleotide probes are given with the probe number in brackets. A schematic representation of the identified species is shown. (H) Quantification of pre-tRNA relative to mature tRNA is shown; expression at 12 versus 0 h is set to 1 for the WT. Average of three biological replicates is shown with s.d. See also Supplementary Figure S6.

Figure 8

Model for transcription termination and surveillance mediated by the Nrd1–Nab3, TRAMP and exosome complexes. (A) Nrd1–Nab3 interact with the CBC and CTD of Pol II to initiate transcription termination through recognition of consensus-binding motifs of sn(o)RNAs, CUTs and long ncRNAs. They subsequently recruit TRAMP and exosome complexes for oligoadenylation and degradation/processing. (B) Nrd1–Nab3 interact post-transcriptionally with aberrant Pol III transcripts, recognizing consensus-binding motifs or structural abnormalities in the RNA. They subsequently recruit TRAMP and exosome complexes for oligoadenylation and degradation.