Fig 1.
Kre33 forms acetylation complexes with 18S rRNA and snoRNPs snR4 and snR45.
A) 3D surface model of the yeast 18S rRNA (3U5B pdb file together with UCSF-Chimera [43] were used to make the surface model). Both acetylated residues in the 18S rRNA, ac4C1280 and ac4C1773 are highlighted in red and blue spheres, respectively. B) Kre33 binds directly to 18S rRNA, snoRNAs snR4 and snR45 and specifically to leucine and serine tRNAs. Pie-chart representing the relative abundance of various RNA classes in the Kre33 data-set according to FPKMs (Fragments per kilo base transcript per million mapped reads). The RNA species listed on the right specifically cross-linked to Kre33 as found by further data-analysis. Relative abundance of some tRNAs (in brackets) was not above that of those in control CRAC-experiments. C) Kre33 cross linking sites on the 2D structure of 18S rRNA; Kre33 cross-links to the 5′ domain (pink), 3′ major domain (helix 34 (orange)) and 3′ minor domain (helix 44 and helix 45 (blue)). Cross-linked residues in these regions are highlighted in yellow and the acetylated cytosine residues are colored in purple. D-F) Line diagrams showing the total number of hits each time a nucleotide was mapped to the reference sequence (black, left y-axis) and the number of reads with a deletion of that nucleotide (red, right y-axis) plotted against the RNA sequence (x-axis). Apart from 18S rRNA (D), snoRNAs snR4 (E) and snR45 (F) were identified to cross-link to Kre33.
Fig 2.
U3, snR4, and snR45 copurify with Kre33.
A) RNA isolated from affinity-purified complexes of Kre33, Nop58 and Gar1 TAP-tagged strains were analyzed by northern blotting using probes hybridizing specifically to U3a, snR4, snR45 and snR10. T stands for total cell extract /input, S for supernatant/unbound fraction, and P for pellet /eluate. Mock-purification with the untagged parental strain (BY 4741) was used as a negative control. Nop56 is a core protein of all box C/D snoRNPs, whereas Gar1 is an integral constituent of box H/ACA snoRNPs. These two proteins were used as a positive (Nop58) and negative control (Gar1) for the Kre33-TAP pull-down. B) Quantification of immunoprecipitations with the percent enrichment calculated as the percent change in the signal of respective snoRNA bands in T vs P. Western blot with the PAP antibody exhibiting the specific pull-down of each TAP-tagged protein is shown below the Northern blot panels. C) Co-sedimentation of snR4, snR45, and Kre33 on a 5% to 25% sucrose gradients. Both snR4 and snR45 were detected by Northern blotting and Kre33 by Western blotting using an anti-TAP antibody.
Fig 3.
SnR4 and snR45 are distinctly involved in acetylation of 18S rRNA.
A) Predicted base-pairing of snR45 and U13 to helix 45 of yeast and human 18S rRNA, respectively. B) Overlaid RP-HPLC chromatograms of nucleosides derived from 18S rRNA of WT (black) and a strain carrying a SNR45 deletion (Δsnr45, red). C) Overlaid chromatograms of the nucleosides derived from fragments isolated using mung bean nuclease assay, containing ac4C1280 (oligo 34) and ac4C1773 (oligo 45) isolated from WT (black, green) and Δsnr45 (red, blue). D) Like snR45, snR4 can base-pair to 18S rRNA via extended complementarity to helix 34 proximal to residue ac4C1280. E) Overlaid chromatograms of the nucleosides derived from 18S rRNA of WT (black), Δsnr4 (cyan), Δsnr45 (red) and the double mutant Δsnr4Δsnr45 (blue).
Fig 4.
Secondary structures of snR4 and snR45.
Secondary structure modeling for snR4 and snR45 was based on phylogenetic analysis (see S2 and S3 Figs) as explained in Materials and Methods, and confirmed by in vivo chemical probing. DMS methylated residues were analyzed by primer extension as explained in Materials and Methods. A, C) Representative gels showing structure probing with DMS for A) snR4, and C) snR45. Bands corresponding to modified residues are marked and mapped on to the 2D structure of snR4 (B) and snR45 (D) (orange dots). Shown are the conserved regions with guide sequences GS1 and GS2 (red) and their interactions with 18S rRNA (blue), the C/D and C′/D′ motifs (black), the pseudo-knot (olive) and helices with strong phylogenetic support (teal) as well as modifications on the snoRNAs. Kre33 cross-linking sites identified by CRAC analysis are highlighted (yellow ovals) and the snoRNA-region protected by Kre33 is outlined (dark blue).
Fig 5.
Physical interaction of snR4 and snR45 with 18S rRNA is indispensable for acetylation of C1280 and C1773.
Mutations of snR4 (A) and snR45 (B) that were tested are boxed with the resulting acetylation efficiency indicated (%; red for negative; green for neutral; substituted nucleotides are red on white). Mutant snoRNAs were stably expressed as shown by Northern blot developed with a snoRNA-specific probe, C) for snR4 and D) for snR45. E) RP-HPLC chromatograms of 18S rRNA isolated from strain Δsnr4Δsnr45 expressing mutant snr4-a (red), -b (blue), -c (green) or -d (pink). Acetylation efficiencies are in brackets. F) RP-HPLC chromatograms of 18S rRNA isolated from strain Δsnr4Δsnr45 expressing mutant snr45-a (red), -b (green), -c (blue) or -d (cyan). Acetylation efficiencies are in brackets. 5S rRNA was used as loading control.
Fig 6.
Putative Helicase activity of Kre33 is vital for 18S rRNA acetylation.
A) Domain architecture of Kre33. B) In silico 3D structure of Kre33, predicted using Phyre [42] and prepared using UCSF Chimera [43]. C) The Walker A motif (P loop) in the helicase domain of Kre33 and TmcA is highly conserved. The substitution of lysine 289 to alanine (K289A) in Kre33 leads to a dramatic loss of 18S rRNA acetylation. D) Overlaid chromatograms of the nucleosides derived from 18S rRNA of WT and the helicase mutant (K289A).
Fig 7.
Loss of putative helicase activity affects the binding of snoRNAs in pre-ribosomal complexes.
A) Sucrose gradient distribution of snR4, snR45, snR40, snR55, snR49, snR51, snR57, snR41, snR10, snR30 and aberrant 23S rRNA in the isogenic WT and helicase mutant (K289A). All snoRNAs were detected by Northern blot using specific probes and WT-Kre33 and kre33-K289A on a Western blot with anti-His antibody (Qiagen). Aberrant 23S rRNA was detected by Northern Blot using a probe specific to 5′ETS (cf. Fig 8A). Loss of helicase activity of Kre33 affects the structure of snR4 and snR45. B) DMS structure probing of snR4 and snR45 in the WT and the K289A mutant (two independent experiments). Bands corresponding to nucleotides with altered (orange dots) and unaltered (blue dots) DMS reactivity are annotated and mapped on the 2D structures of snR4 (C) and snR45 (D).
Fig 8.
Loss of putative helicase activity of Kre33 affects early pre-rRNA processing and 18S rRNA modification by snR40 or snR55.
Early pre-rRNA processing requires the predicted helicase activity of Kre33 and does not depend on acetylation. A) Overview of the 35S primary transcript. 35S pre-rRNA contains 18S, 5.8S and 25S rRNA sequences separated by internal transcribed spacers (ITS1 and ITS2). B) Northern blot analysis of pre-RNA processing in WT and strains expressing mutant Kre33 in which its helicase (K289A) or acetylation (H545A, R637A) activity was abolished. The membrane was hybridized with radioactively labeled probes annealing to 5′ ETS, ITS1, or ITS2 sequences. The increased levels of 35S pre-rRNA and the accumulation of the aberrant 23S species in the case of kre33-K289A are indicative for reduced processing at sites A0, A1 and A2 (which is bypassed by cleavage at A3) resulting in very low steady state levels of 18S rRNA (EtBr panel). In line with defective 18S formation, absence of putative Kre33-helicase activity leads to a significant growth defect (C). D) Guide-sequence interaction of snR40 (green), snR55 (magenta) and snR4 (red) with 18S rRNA (black). The target sequences of these snoRNAs in h34 of 18S rRNA overlap with each other. Primer extension analysis of ribose methylation in helix 34 of 18S rRNA (E) and helix 33 of 25S rRNA (F). 32P-labeled primer complementary to nucleotides 1315 to 1336 of yeast 18S rRNA (E) and to nucleotides 947 to 967 of 25S rRNA (F) were used for methylation analysis of Um1269 (snR55) and Gm1271 (snR40) (E) in the 18S rRNA and Um898 (snR40) (F) in the 25S rRNA. Since the bands corresponding to Gm1271 were barely visible in comparison to Um1269 (E(i)), the levels for these bands were altered in the boxed section (E(ii)). Bands corresponding to Um1269 and Gm1271 were quantified using ImageJ software (http://imagej.nih.gov/ij/).