Global analysis of yeast RNA processing identifies new targets of RNase III and uncovers a link between tRNA 5' end processing and tRNA splicing

Abstract

We used a microarray containing probes that tile all known yeast noncoding RNAs (ncRNAs) to investigate RNA biogenesis on a global scale. The microarray verified a general loss of Box C/D snoRNAs in the TetO7-BCD1 mutant, which had previously been shown for only a handful of snoRNAs. We also monitored the accumulation of improperly processed flank sequences of pre-RNAs in strains depleted for known RNA nucleases, including RNase III, Dbr1p, Xrn1p, Rat1p and components of the exosome and RNase P complexes. Among the hundreds of aberrant RNA processing events detected, two novel substrates of Rnt1p (the RUF1 and RUF3 snoRNAs) were identified. We also identified a relationship between tRNA 5' end processing and tRNA splicing, processes that were previously thought to be independent. This analysis demonstrates the applicability of microarray technology to the study of global analysis of ncRNA synthesis and provides an extensive directory of processing events mediated by yeast ncRNA processing enzymes.

Figure 1

Longer oligonucleotide probes improve the sensitivity of detection. The intensity of fluorescence of wild-type RNA bound to probes on either short (top) or long (bottom) oligonucleotide microarrays, shaded according to the color bar on the right, is plotted for representative transcripts: (A) tRNAs, (B) 35S pre-rRNA and (C) snoRNAs. Schematic diagrams of the RNAs and flanking regions are shown below with boxes representing RNA sequence and thin lines indicating flanking sequence. (D) Box C/D RNA depletion in a TetO₇-BCD1 strain. The relative fluorescence of Box C/D and Box H/ACA snoRNA probes is shown, colored according to the scale on the right. Note that because the reciprocal of the mutant:wild-type ratio is plotted in this panel, increasing red color indicates depletion of snoRNAs.

Figure 2

Detecting known RNA processing events by microarray. Oligonucleotides corresponding to processed regions showing at least a 4-fold increase at least one of the mutant strains were subjected to hierarchical agglomerative clustering. Probes corresponding to accumulated sequences are colored red, according to the scale shown. The identities of the probes are shown in the black and white panel on the right. A fully labeled numerical version of this figure is available in the Supplementary Material.

Figure 3

The 5′ ends of RUF1 and RUF3 are processed by Rnt1p. (A) Relative fluorescence of probes from analysis of TetO₇-RNT1 and rnt1-1^ts strains is shown, with schematic diagrams of the RNAs below as described in Figure 1. (B) Northern blot analysis of RUF3 and RUF1 RNA in wild-type and Rnt1p-deficient strains. The blots were probed sequentially with two different probes (the positions of which are shown on the schematic diagrams). Two replicates of each sample were loaded side-by-side; U4 RNA was probed as a loading control. (C) Predicted structures of 5′ flank sequences of RUF1 and RUF3, based on Mfold structure prediction (25). The inset shows the full predicted structure of the 5′ leader. The AGNN tetraloop in each RNA is circled in red.

Figure 4

Microarray analysis reveals a partial dependence of tRNA splicing on tRNA 5′ end processing. (A) Relative fluorescence of probes complementary to select tRNAs are shown with schematic diagrams below. (B) Relative fluorescence of LeuCAA tRNA probes is shown for the nine microarray experiments shown in Figure 2A. (C) Northern blot analysis of LeuCAA tRNA species present in TetO₇-POP1 and POP4 strains is shown. The same blot was probed sequentially with three different probes, as indicated by the schematic diagrams at the top of each blot (with the red line indicating the position of the probe). The identity of each tRNA species is shown to the right.