Antisense pairing and SNORD13 structure guide RNA cytidine acetylation

Abstract

N4-acetylcytidine (ac⁴C) is an RNA nucleobase found in all domains of life. The establishment of ac⁴C in helix 45 (h45) of human 18S ribosomal RNA (rRNA) requires the combined activity of the acetyltransferase NAT10 and the box C/D snoRNA SNORD13. However, the molecular mechanisms governing RNA-guided nucleobase acetylation in humans remain unexplored. After applying comparative sequence analysis and site-directed mutagenesis to provide evidence that SNORD13 folds into three main RNA helices, we report two assays that enable the study of SNORD13-dependent RNA acetylation in human cells. First, we demonstrate that ectopic expression of SNORD13 rescues h45 in a SNORD13 knockout cell line. Next, we show that mutant snoRNAs can be used in combination with nucleotide resolution ac⁴C sequencing to define structure and sequence elements critical for SNORD13 function. Finally, we develop a second method that reports on the substrate specificity of endogenous NAT10-SNORD13 via mutational analysis of an ectopically expressed pre-rRNA substrate. By combining mutational analysis of these reconstituted systems with nucleotide resolution ac⁴C sequencing, our studies reveal plasticity in the molecular determinants underlying RNA-guided cytidine acetylation that is distinct from deposition of other well-studied rRNA modifications (e.g., pseudouridine). Overall, our studies provide a new approach to reconstitute RNA-guided cytidine acetylation in human cells as well as nucleotide resolution insights into the mechanisms governing this process.

Keywords: N4-acetylcytidine; SNORD13; epitranscriptome; modification; ribosome.

FIGURE 1.

(A) NAT10 and SNORD13 are required for deposition of ac⁴C in helix 45 of 18S rRNA. (B) Schematic of complementarity between human SNORD13 (purple) and 18S rRNA (blue). The site of cytidine acetylation (C1842) is specified.

FIGURE 2.

(A) Sequence alignment of vertebrate pre-SNORD13 homologs. Antisense regions 18S-A, 18S-B, and 18S-C are highlighted in red, C and D-box motifs involved in putative kink-turn formation in yellow, and putative stem regions in gray. The small verticle black arrow indicates the 3′ end of SNORD13 as experimentally determined by RACE analysis. (B) Proposed SNORD13 base-pairing interactions with colors coded as above. Bases written in uppercase are part of stable, fully processed SNORD13 while those written in lowercase correspond to the 3′-extension of the transient pre-SNORD13 intermediate. (C) Proposed maturation of pre-SNORD13, containing a 3′-extension which masks its 5′-antisense region, to mature SNORD13 capable of interacting with 18S rRNA.

FIGURE 3.

(A) Schematic for analysis of SNORD13 stability by RNase A/T1 assay. (B) Sequence of human SNORD13s analyzed. Antisense regions 18S-A, 18S-B, and 18S-C are highlighted in red, C and D-box motifs involved in k-turn formation in yellow, and putative stem regions in gray. In addition to wild-type (“WT,” black) sequence, disruptive mutations (“mut,” red), and rescue mutations (“comp,” red) were explored for stems I–IV. C and D box mutations were also analyzed (sequences provided in Supplemental Information). (C) Results of RNase A/T1 stability assay. Top box indicates protected RNA fragment corresponding to full length, transfected human versions of SNORD13. Bottom box indicates protected RNA fragments corresponding to endogenously expressed mouse SNORD13 background.

FIGURE 4.

(A) Schematic for analysis of SNORD13 function using quantitative ac⁴C sequencing. (B) Sequence of human SNORD13s analyzed. In addition to wild-type (“WT”) sequence, disruptive mutations (18S-A, 18S-B*, 18S-C mutants), and a mutant with increased complementarity (“18S-A/18S-B full comp”) were explored for rescue of ac⁴C in SNORD13 KO cells. Sequences and verification of expression are provided in Supplemental Figure S4. Note the finding that the 18S-A mutant is expressed when its stem V is disrupted, implies stem V is not strictly required for SNORD13 synthesis. (C) Stem IV is required for accumulation of mutant SNORD13s. Mutation of SNORD13 in mutant 18S-B disrupts stem IV. Reintroduction of complementarity in construct 18-B* (bottom) allows accumulation and testing of function. Expression of 18S-B* is verified by RNase A/T1 mapping in Supplemental Figure S4c. (D) Rescue of SSU-ac⁴C1842 by SNORD13 antisense mutants. Mutants rescue values are normalized relative to the WT SNORD13, which was set to equal 100%. Background misincorporation rates in SNORD13 KO cells were 0%–4%. Values represent n = three biological replicates, analyzed by two-tailed Welch's t-test (ns = not significant, [*] P < 0.05, [**] P < 0.01, and [***] P < 0.001). Exemplary sequencing traces are provided in Supplemental Figure S4d. (E) Rescue of SSU-ac⁴C1842 by SNORD13 stem comp mutants. Structures of comp mutants are provided in Figure 3B. Values represent n = three biological replicates, analyzed by two-tailed Welch's t-test (ns = not significant, [*] P < 0.05, [**] P < 0.01, and [***] P < 0.001).

FIGURE 5.

(A) Schematic for analysis of endogenous SNORD13 substrate specificity using a Pol I-transcribed pre-rRNA h45-ITS1 minigene substrate and quantitative ac⁴C sequencing. (B) Structure of h45-ITS1 substrates with mutations lying proximal to natural ac⁴C site. Values in red represent ac⁴C-dependent misincorporation rates normalized relative to the WT h45-ITS1 sequence, which was set to equal 100%. (C) Summary of percent misincorporation observed upon mutation of 5′-CCG-3′ consensus sequence in h45-ITS1 substrates. (D) Bar graph of mutants specified in Figure 5B and E. Values represent n = three biological replicates, analyzed by two-tailed Welch's t-test in comparison to WT (ns = not significant, [*] P < 0.05, [**] P < 0.01, [***] P < 0.001, and [****] P < 0.0001) (E) Structure of “triple mutant” h45-ITS1 substrate engineered to have increased complementarity to SNORD13. (F,G). Structure of h45-ITS1 substrates with bases inserted 5′ and/or 3′ relative to the 5′-CCG-3′ consensus sequence. (H) Structure of h45-ITS1 substrates in which the 5′-CCG-3′ consensus sequence is shifted 1 bp (+1 bp) or 3 bp (+3 bp). (I) Bar graph of mutants specified in Figure 5F–H. Values represent n = three biological replicates, analyzed by two-tailed Welch's t-test in comparison to WT (ns = not significant, [*] P < 0.05, [**] P < 0.01, [***] P < 0.001, and [****] P < 0.0001). Exemplary sequencing traces provided in Supplemental Figure S6 and Supplemental Figure S8b.

FIGURE 6.

(A) Schematic for analysis of orthogonal SNORD13-substrate pairs. Exemplary sequencing traces are provided in Supplemental Figure S9. (B) Bar graph of ac⁴C levels at site corresponding to SSU-1842 in orthogonal SNORD13-substrate pairs. Values represent ac⁴C-dependent misincorporation rates normalized relative to the WT h45-ITS1 and WT SNORD13 pair, which was set to equal 100%. Values represent n = three biological replicates, analyzed by two-tailed Welch's t-test in comparison to WT (ns = not significant, [*] P < 0.05, [**] P < 0.01, [***] P < 0.001, and [****] P < 0.0001). Sequences of SNORD13 18S-A mutant and h45-ITS1 18S-A mutant are provided in the Supplemental Information. (C) Overview of SNORD13 structure-function relationships probed in this study.

Supuni Thalalla Gamage