Identification of a novel element required for processing of intron-encoded box C/D small nucleolar RNAs in Saccharomyces cerevisiae

Abstract

Processing of intron-encoded box C/D small nucleolar RNAs (snoRNAs) in metazoans through both the splicing-dependent and -independent pathways requires the conserved core motif formed by boxes C and D and the adjoining 5'-3'-terminal stem. By comparative analysis, we found that five out of six intron-encoded box C/D snoRNAs in yeast do not possess a canonical terminal stem. Instead, complementary regions within the flanking host intron sequences have been identified in all these cases. Here we show that these sequences are essential for processing of U18 and snR38 snoRNAs and that they compensate for the lack of a canonical terminal stem. We also show that the Rnt1p endonuclease, previously shown to be required for the processing of many snoRNAs encoded by monocistronic or polycistronic transcriptional units, is not required for U18 processing. Our results suggest a role of the complementary sequences in the early recognition of intronic snoRNA substrates and point out the importance of base pairing in favoring the communication between boxes C and D at the level of pre-snoRNA molecules for efficient assembly with snoRNP-specific factors.

FIG. 1

Structures of EFB1 episomal constructs expressing the tagged U18 snoRNA. (A) Diagrammatic representation of the pGALU18wt plasmid, containing the transcriptional unit of the EFB1 gene between the GAL1 promoter and the CYC1 terminator. The Cbs construct is a mutant derivative which carries an A-to-C (shown in bold) substitution of the branch nucleotide, which abolishes splicing. The dark box within the U18 snoRNA coding region represents the tag sequence (47). Complementary sequences A and B are indicated. The lengths of the different portions of the constructs are indicated in nucleotides. pA indicates the polyadenylation site within the CYC1 terminator (term). (B) Representation of the U18 host intron secondary structure. The putative interaction between A and B sequences is shown along with the U18 box C/D motif formed by the C and D boxes and the adjoining 5′-3′-terminal stem. Nucleotides belonging to the U18 coding sequence are shown in bold. Numbering starts from the first nucleotide of the intron, and the positions of the 5′ splice site (5′ss), branch site (bs), and 3′ splice site (3′ss) are indicated. (C) Representation of the putative secondary structures of the different mutants used in this study; mutations are boxed. ΔS disrupts A-B pairing by inverting sequence A; RS restores pairing by inverting sequence B to make it complementary to the inverted A sequence; 5′-3′S lengthens the terminal stem complementarity; bC carries a mutation (lowercase letters) within conserved box C.

FIG. 2

(A) Analysis of U18 processing in a wt strain. Total RNA was extracted from strain CH1462 carrying the indicated wt or Cbs constructs (schematically represented above the lanes) at the indicated times of galactose induction (expressed in hours above the lanes; hr-gal). Total RNA (5 μg) was resolved on a 6% acrylamide–7 M urea gel, electroblotted onto a nylon membrane, and hybridized with the anti-tag oligonucleotide. The different processing products are diagrammed on the side (exons are depicted as dark boxes, U18 is shown as an open box, and intronic sequences and the 5′ untranslated region are shown as lines). (B) Control hybridization to U5 snRNA. L and S indicate the long and short forms of U5 snRNA, respectively. The oligonucleotide used is indicated below each panel. Lane M: pBR322 plasmid DNA, MspI digested; molecular sizes are indicated on the left in nucleotides.

FIG. 3

Analysis of mRNA production from the different U18 host pre-mRNAs. Total RNA extracted from strain CH1462 transformed with the indicated wt (A) and Cbs (B) constructs before or 1 h after galactose induction (hr-gal) was analyzed by primer extension with oligonucleotide E3′ complementary to the downstream exon of the plasmid-borne EFB1 gene (see Materials and Methods). A schematic representation of the EFB1 pre-mRNA is shown below, together with the position of the oligonucleotide used for the primer extension. The different elongation products are indicated on the side. The asterisk marks the position of a nonspecific stop within the U18 coding sequence, which is observed upon elongation of full-length pre-mRNA. Lanes M: pBR322 plasmid DNA, MspI digested.

FIG. 4

(A) Analysis of U18 processing in the debranching-deficient strain dbr1-Δ. Total RNA was extracted from strain dbr1-Δ carrying the indicated wt constructs (schematically represented above the lanes) at the indicated times of galactose induction (expressed in hours above the lanes; hr-gal). Total RNA (5 μg) was resolved on a 6% acrylamide–7 M urea gel, electroblotted onto a nylon membrane, and hybridized with the anti-tag oligonucleotide. The different processing products are diagrammed on the side, as in Fig. 2. The lariat species is depicted as a circle lacking the extension from the branch point to the 3′ splice site, since this is the major form found in strain dbr1-Δ (13). (B) Control hybridization to U3 snRNA. The oligonucleotide used is indicated below each panel. Lanes M: pBR322 plasmid DNA, MspI digested.

FIG. 5

Analysis of U18 processing in the RNT1 and rnt1-Δ strains. Total RNA was extracted from strains RNT1 and rnt1-Δ carrying the indicated wt or Cbs constructs at the indicated times of galactose induction (expressed in hours above the lanes; hr-gal). Total RNA (5 μg) was resolved on a 6% acrylamide–7 M urea gel, electroblotted onto a nylon membrane, and hybridized with the anti-tag oligonucleotide. U18 snoRNA is indicated on the side. (B) Control hybridization to U5 snRNA: L and S indicate the long and short forms, respectively, of U5 snRNA (10).

FIG. 6

(A) Analysis of U18 processing in the 5′→3′ exonuclease-deficient strain rat1-1. Total RNA was extracted from strain rat1-1 carrying the indicated wt or Cbs constructs (schematically represented above the lanes) at the indicated times of galactose induction (expressed in hours above the lanes; hr-gal). Strain rat1-1 was shifted to the restrictive temperature of 37°C for 2 h before addition of galactose. Total RNA (5 μg) was resolved on a 6% acrylamide–7 M urea gel, electroblotted onto a nylon membrane, and hybridized with the anti-tag oligonucleotide. The different processing products are diagrammed on the side, as in Fig. 2. (B) Control hybridization to U3 snRNA. The oligonucleotide used is indicated below each panel. Lanes M: pBR322 plasmid DNA, MspI digested. (C) Levels of accumulation of U18 in the wt (CH1462; lanes 1, 4, and 7), rat1-1 (lanes 2, 5, and 8), and dbr1-Δ (lanes 3, 6, and 9) strains. Total RNAs extracted from the different strains transformed with the indicated constructs 1 h after galactose induction were analyzed by primer extension with the anti-tag oligonucleotide. Strain rat1-1 was shifted to the restrictive temperature of 37°C for 2 h before the addition of galactose.

FIG. 7

Analysis of snR38 accumulation in the wt CH1462 and the mutant dbr1-Δ strains. Total RNA was extracted from strains CH1462 (left) and dbr1-Δ (right) carrying the pGAL38wt-tag and pGAL38ΔS-tag constructs at the indicated times of galactose induction (expressed in hours above the lanes; hr-gal). Total RNA (5 μg) was resolved on a 6% acrylamide–7 M urea gel, electroblotted onto a nylon membrane, and hybridized with the 38 anti-tag oligonucleotide. The different processing products are diagrammed. Lane M: pBR322 plasmid DNA, MspI digested.

FIG. 8

Model for intron-encoded box C/D snoRNA processing in yeast. This model applies to the intronic snoRNAs lacking a canonical terminal stem. The host intron displays nonconserved complementary sequences (gray inverted arrows). The box C/D snoRNA is depicted as a thick line, and its conserved boxes C and D are depicted as small open boxes. The host pre-mRNA mainly undergoes splicing (symbolized by the association with U1 and U2 snRNPs), releasing mRNA and the lariat. This is quickly linearized by Dbr1p, producing free 5′ and 3′ ends readily attacked by exonucleases (represented as “Pacmen”; the nuclear 5′→3′ exonuclease is Rat1p). Concomitantly with or immediately following linearization, formation of the external stem directs proper folding of the pre-snoRNA molecule (base pairing interactions are depicted by horizontal bars), allowing association of snoRNP factors with boxes C and D. Nevertheless, it cannot be excluded that snoRNP assembly may begin at some previous steps during the splicing process. This assembly protects the pre-snoRNA from exonuclease digestion and specifies the snoRNA mature ends. In the concurrent splicing-independent pathway, whose existence is evidenced by the partial insensitivity to the dbr1-Δ mutation, folding of newly synthesized host pre-mRNA molecules induced by the external stem ensures their recognition as pre-snoRNA substrates. Association with snoRNP factors is proposed to direct the endonucleolytic cleavages of the snoRNA-flanking sequences (arrowheads; their position is based on earlier results with U18 snoRNA [47]).