The cotranscriptional assembly of snoRNPs controls the biosynthesis of H/ACA snoRNAs in Saccharomyces cerevisiae

Abstract

The carboxy-terminal domain (CTD) of RNA polymerase II large subunit acts as a platform to assemble the RNA processing machinery in a controlled way throughout the transcription cycle. In yeast, recent findings revealed a physical connection between phospho-CTD, generated by the Ctk1p kinase, and protein factors having a function in small nucleolar RNA (snoRNA) biogenesis. The snoRNAs represent a large family of polymerase II noncoding transcripts that are associated with highly conserved polypeptides to form stable ribonucleoprotein particles (snoRNPs). In this work, we have studied the biogenesis of the snoRNPs belonging to the box H/ACA class. We report that the assembly factor Naf1p and the core components Cbf5p and Nhp2p are recruited on H/ACA snoRNA genes very early during transcription. We also show that the cotranscriptional recruitment of Naf1p and Cbf5p is Ctk1p dependent and that Ctk1p and Cbf5p are required for preventing the readthrough into the snoRNA downstream genes. All these data suggest that proper cotranscriptional snoRNP assembly controls 3'-end formation of snoRNAs and, consequently, the release of a functional particle.

FIG. 1.

Naf1p is recruited to actively transcribed H/ACA snoRNA genes. (A) ChIP analysis was performed on the YAF10 yeast strain expressing a TAP-tagged version of NAF1 (panel RNase−). The cross-linked chromatin was amplified before (Input) and after (Naf1p) immunoprecipitation. An RNase treatment step was added before immunoprecipitation (panel RNase+). A nontranscribed intergenic region from chromosome V (band *) was used as an internal control by coamplification with each of the gene-specific primers. The diagram on the side shows the oligonucleotides specific for the 5′ end (“a”) and the coding region (“b”) of SNR10, SNR30, and SNR13 genes. (B) Histogram displaying the degree of Naf1p cross-linking on the snoRNA genes. The data are presented as the average (standard deviation, <20%) of signals of five PCRs performed on five preparations of immunoprecipitated DNA, normalized against the controls.

FIG. 2.

Cbf5p and Nhp2p are associated with the sites of H/ACA snoRNA transcription. Strains containing TAP-tagged versions of either Cbf5p (A), Nhp2p (B), or Gar1p (C) were analyzed by ChIP of snoRNA genes. Cross-linked chromatin was used for PCR amplification before (panel Input) and after (panel Ip) immunoprecipitation with primers specific for SNR10, SNR30, and SNR13 genes (see diagram in Fig. 1). A nontranscribed intergenic region from chromosome V (band *) was used as an internal control by coamplification with each of the gene-specific primers. The data are presented as the average (standard deviation, <20%) of signals of the PCRs performed on five preparations of immunoprecipitated DNA, normalized against the controls.

FIG. 3.

Ctk1 is required for cotranscriptional recruitment of H/ACA snoRNP in vivo. ChIP analyses were carried out with snoRNP TAP-tagged proteins in both wild-type (black bars) and Ctk1p deletion (gray bars) strains. PCR analysis of immunoprecipitated chromatin was performed on the SN30 (“a” and “b” as diagrammed in Fig. 1) and SNR13 genes. (A) ChIP analysis of Naf1p in CTK1 and ctk1Δ strains. (B) ChIP analysis of Cbf5p in CTK1 and ctk1Δ strains. The upper panels show the amplification products of the snoRNA regions specified above each lane. A nontranscribed intergenic region from chromosome V (band *) was used as an internal control by coamplification with each of the gene-specific primers. The data are presented as the average (standard deviation, <20%) of signals of the PCRs performed on three preparations of immunoprecipitated DNA, normalized against the controls.

FIG. 4.

Analysis of readthrough products in a Cbf5p-depleted strain. (A) Schematic representation of the Tet::Cbf5 allele (top of the panel) and growth of the Tet::Cbf5 (gray squares) and isogenic (black squares) strains following transfer to doxycycline (Dox)-containing medium (bottom of the panel). Cell density (optical density at 600 nm [OD₆₀₀]) was measured at the indicated time, and the cultures were periodically diluted to be continuously kept in exponential growth. (B) Northern hybridization for the detection of CBF5 mRNA and 35S and 18S rRNAs from the Tet::Cbf5 strain (lanes 0 to 24 h); U6 was used as a normalization control. (C) Northern analysis of snoRNAs (box H/ACA, snR30, snR10, snR5, and snR189; box C/D, snR13) and of the control U6 snRNA. (D) RT-PCR analysis was carried out with set of primers described below the panel on total RNA extracted from following strain: Tet::Cbf5, grown in a complete medium (lanes 0) and then shifted in doxycycline-containing medium for 24 h (lanes 24); ssu72-2, grown at the permissive (lanes 25°C) and restrictive temperatures for 2 h (lanes 37°C); and CTK1 and ctk1Δ (lanes CTK1 and ctk1Δ). No products were observed when reverse transcriptase was omitted during cDNA synthesis (data not shown).

FIG. 5.

The snoRNA factory. The snoRNP proteins are required for a quality control mechanism in which the functional assembly of box C/D and H/ACA snoRNAs is monitored during transcription. The snoRNA primary transcript attracts the 3′-end formation complex through snoRNP components. The Nop1p/Cbf5p-3′-end machinery interaction may contribute to negatively regulate poly(A) polymerase (Pap1p) at snoRNA cleavage sites. In the case of box C/D snoRNAs, the specific partner has been identified in the Ref2p factor while for H/ACA it is still unidentified.