Rrp5 binding at multiple sites coordinates pre-rRNA processing and assembly

Abstract

In vivo UV crosslinking identified numerous preribosomal RNA (pre-rRNA) binding sites for the large, highly conserved ribosome synthesis factor Rrp5. Intramolecular complementation has shown that the C-terminal domain (CTD) of Rrp5 is required for pre-rRNA cleavage at sites A0-A2 on the pathway of 18S rRNA synthesis, whereas the N-terminal domain (NTD) is required for A3 cleavage on the pathway of 5.8S/25S rRNA synthesis. The CTD was crosslinked to sequences flanking A2 and to the snoRNAs U3, U14, snR30, and snR10, which are required for cleavage at A0-A2. The NTD was crosslinked to sequences flanking A3 and to the RNA component of ribonuclease MRP, which cleaves site A3. Rrp5 could also be directly crosslinked to several large structural proteins and nucleoside triphosphatases. A key role in coordinating preribosomal assembly and processing was confirmed by chromatin spreads. Following depletion of Rrp5, cotranscriptional cleavage was lost and preribosome compaction greatly reduced.

Graphical abstract

Figure 1

Rrp5 Binding Sites on the Pre-rRNA (A) A translational fusion between Rrp5 and a His6-TEV-ProteinA (HTP) cassette was expressed from the endogenous RRP5 locus. Rrp5 comprises 12 S1 RNA-binding domains associated with 7 tetratricopeptide repeat (TPR) protein-binding domains. (B) Rrp5-HTP cells were exposed to UV crosslinking while actively growing in culture medium (in vivo) or following cell lysis and clearance of cell debris (in vitro). Crosslinked RNAs were trimmed and ligated to linkers followed by RT-PCR amplification and Illumina sequencing. The sequences obtained were aligned with the yeast genome, and identified target RNAs were sorted into functional categories. The percentage of total mapped reads in the sample is shown for each class. A total of 35 M mapped reads were recovered for the in vitro sample and 9 M for the in vitro sample. (C and D) Sequences obtained from in vitro (C) and in vivo (D) experiments were aligned with the rDNA (RDN37-1) encoding 35S pre-rRNA (nucleotides 1–6,858), and the frequency of recovery (hits per million mapped reads) is plotted for each individual nucleotide (shown in purple). Below the graph, the locations of mutations and deletions are shown in black. These generally represent precise binding sites. In the cartoons, the position of the mature 18S, 5.8S, and 25S rRNAs are indicated by thick lines. Peaks (a–h) are labeled. Peaks that are frequently recovered and present in a control experiment are indicated by a star (^∗). The two major peaks (c and d) are located in the ITS1.

Figure 2

Binding Sites for the Rrp5 NTD and CTD Regions (A) Schematic of the domain structure of Rrp5 and functional targets of the NTD and CTD. Red arrows indicate the requirements for these domains in the pre-rRNA cleavages. Note that for clarity, the pre-rRNA is drawn in a 3′-5′ orientation. (B) Schematic of the tagged versions of the N- and C-terminal domains of Rrp5. (C) NTD: P_GAL::RRP5 with plasmids expressing PTH-Rrp5NTD (Rrp5 aa 1–1,130) and untagged Rrp5CTD. CTD: P_GAL::RRP5 with plasmids expressing PTH-Rrp5CTD (Rrp5 aa 1,131—1,729) and untagged Rrp5NTD. Strains were grown in glucose medium and UV crosslinked in vivo, and RNA sequences were obtained and analyzed as in Figure 1. The percentage of total mapped reads in the sample is shown for each class of RNA. A total of 5.6 M mapped reads were recovered for the NTD and 0.5 M for the CTD. (D and E) rDNA sequences were treated as in Figure 1. Peaks that are frequently recovered and present in a control experiment are indicated by an asterisk (^∗). Peaks for the N- and C-terminal domains with the same coordinates as found for the full-length Rrp5 are indicated by the same letter in black. A peak of sequences recovered specifically with Rrp5NTD-HTP is labeled i. Peaks of sequences recovered specifically with the Rrp5CTD-HTP are labeled j, k, and l. Mutations are indicated as for Figures 1C and 1D.

Figure 3

Sites of Rrp5 Association with ITS1 (A) Mapped reads for Rrp5-HTP full-length in vitro (purple) and in vivo (gray). (B) Mapped reads for Rrp5NT-HTP (red) and Rrp5CT-HT (blue) in vivo. The graphs show hit density for each nucleotide over the ITS1 region, per 10,000 total reads mapped within ITS. Processing sites A₂ and A₃ are indicated on the graphs, as well as the major peaks annotated as c and d. (C) Predicted secondary structure of the yeast ITS1 region. The binding sites identified in vitro for full-length Rrp5-HTP (purple) and in vivo for Rrp5NTD-HTP (red) or Rrp5CTD-HTP (blue) are indicated. Mutated nucleotides are designated by dots alongside the sequence, using the same colors.

Figure 4

Sites of Rrp5 Association with the Pre-rRNA Predicted secondary structure of part of the 5′ ETS, 18S, 5.8S, and 25S regions. The binding sites of full-length Rrp5-HTP (purple), Rrp5NT-HTP (red), and Rrp5CT-HTP (blue) are indicated on the sequences. Mutated nucleotides found in the sequences of full-length Rrp5-HTP (purple), Rrp5NT-HTP (red), and Rrp5CT-HTP (blue) are indicated by dots alongside the sequence. Binding sites for snoRNAs U3 (yellow), U14 (green), and snR30 (brown) are also indicated.

Figure 5

Association of Rrp5 with snoRNAs (A) The snoRNAs identified in each of the data sets were clustered by relative recovery, using Cluster 3.0 and Java TreeView. Recovery as a fraction of total mapped reads is represented in the heatmap. Only snoRNAs that were significantly enriched in more than one sample and in at least one full-length Rrp5-HTP set are included. For full analysis, see Figure S2. Colored names indicate snoRNAs recovered with both Rrp5NTD-HTP and Rrp5CTD-HTP (purple), with only Rrp5NT-HTP (red), and with only Rrp5CT-HTP (blue). Classes of snoRNA (C/D or H/ACA) and functions, where known, are indicated. (B–E) Mapped reads for Rrp5-HTP (purple), Rrp5NTD-HTP (red), and Rrp5CTD-HTP (blue) were aligned with different snoRNAs, and the number of hits for each individual nucleotide is plotted per million mapped reads. Black bars below indicate known regions of interaction with the pre-rRNA. The green bar below SNR10 indicates a 7 nt region required for A0, A1, and A2 cleavage (Liang et al., 2010). In the case of U3A, the region shown is the genomic sequence, which includes a 5′-proximal intron (thin gray line), which splits the 5′ rRNA binding region (thick line).

Figure 6

Effects of Rrp5 on Pre-rRNA Compaction and Processing (A) Effects of Rrp5 depletion on pre-rRNA cotranscriptional events were analyzed by EM on chromatin spreads. A P_GAL::RRP5 strain was used, and Miller spreads were prepared from cells growing in galactose medium (n = 228) or following transfer to glucose medium for 4 hr (n = 301). Representative examples of the different phenotypes are presented. Arrows indicate the different cotranscriptional events observed: normal SSU formation (black); normal cotranscriptional cleavage (gray); aberrant, less dense SSU formation (blue); aberrant, small RNP particles (green); aberrant, larger, looser particles (pink). (B) All genes analyzed were assigned to one of three phenotypic categories. Distribution between each category is represented in columns, with normal rDNA morphology indicated in gray. (C) Model for Rrp5 action in pre-rRNA compaction and processing on the major ribosome synthesis pathway. (1 and 2) Rrp5 is a primary preribosome binding factor, and the crosslinking data strongly suggest that the A2 and A3 flanking sequences are the initial binding sites. (3) Rrp5 binds to additional pre-rRNA sites and structural proteins to assemble the large terminal balls that correspond to SSU processome complexes. (4) Following cotranscriptional cleavage at sites A0–A2, Rrp5 remains associated with the nascent transcript until cleavage at site B0. The released pre-rRNA undergoes rapid cleavage at site A3, accompanied or immediately followed by release of Rrp5. For simplicity, the alternative processing pathway of posttranscriptional cleavage at site A2 in the full-length 35S pre-rRNA is omitted.