rRNA adenine methylation requires T07A9.8 gene as rram-1 in Caenorhabditis elegans

Abstract

RNAs are post-transcriptionally modified in all kingdoms of life. Of these modifications, base methylations are highly conserved in eukaryote ribosomal RNA (rRNA). Recently, rRNA processing protein 8 (Rrp8) and nucleomethylin (NML) were identified as factors of N¹-methyladenosine (m¹A) modification in yeast 25 S and mammalian 28 S rRNA, respectively. However, m¹A modification of rRNA is still poorly understood in Caenorhabditis elegans (C. elegans). Here, using the liquid chromatography/tandem mass spectrometry analysis and RNA immunoprecipitation assay, we have identified that the m¹A modification is located around position 674 (A674) of 26 S rRNA in C. elegans. Furthermore, quantitative PCR-based analysis revealed that T07A9.8, a C. elegans homolog of yeast Rrp8 and human NML, is responsible for m¹A modification at A674 of 26 S rRNA. This m¹A modification site in C. elegans corresponds to those in yeast 25 S rRNA and human 28 S rRNA. Intriguingly, T07A9.8 is not associated with pre-rRNA transcription under normal nutrient conditions. Since the m¹A modification of 26 S rRNA requires T07A9.8 in C. elegans, we designated the gene as rRNA adenine methyltransferase-1 (rram-1).

C. elegans, methylation, N¹-methyladenosine, pre-rRNA transcription, rRNA

RNA is a fundamental molecule that conveys the genetic information from DNA into protein. Messenger RNA (mRNA) is transcribed from DNA and translated by ribosomes, which are composed of ribosomal proteins and ribosomal RNAs (rRNAs). During translation, transfer RNAs (tRNAs) bring specific amino acids to elongating polypeptides through codon-specific pairing with mRNA on the ribosomes. Therefore, RNAs represent central factors in gene expression.

RNAs are post-transcriptionally modified in all kingdoms of life. To date, over 100 different types of RNA modifications have been identified (1–3). These modifications influence the structural and functional roles of RNA and help to maintain cellular homeostasis, to mediate stress responses and to participate in other biological functions as well (4–9). Furthermore, recent findings have revealed that some RNA modifications are dynamically and reversibly regulated (10–13). Although a wide variety of RNA modifications exist, RNA methylation accounts for approximately two thirds of the known chemical modifications occurring with RNA.

RNA methylations arise at carbon and nitrogen of the nucleobases (base methylations) and on the oxygen of the ribose 2′-OH (2′-O-ribose methylation) (14). Because methods for detecting and quantifying RNA methylation have been developed, new enzymes and modification sites are being identified (15). Among these modifications, the N¹-methyladenosine (m¹A) modification was previously discovered in mRNA, tRNA and rRNA from several species (16–18), and has been shown to modulate the structural stability and function of these RNA species (19, 20). Within a few years, the methylation at N¹ position of adenosine is reversibly regulated through demethylases (16, 21–23), emphasizing the biological significance of this modification.

Nucleomethylin (NML) is a nucleolar protein that controls rRNA transcription through epigenetic regulation and protects cells from energy deprivation-induced apoptosis (24). Recently, we have shown that NML is also required for the m¹A modification of 28 S rRNA, thereby modulating ribosomal subunit formation and cell proliferation in mammalian cells (25). Rrp8, a yeast orthologue of mammalian NML, is responsible for the m¹A modification in yeast 25 S rRNA and the null mutant of this gene shows growth defects in yeast (26). Database searching for Caenorhabditis elegans (C. elegans) protein has shown that T07A9.8 is a putative homolog of yeast Rrp8 and mammalian NML protein (27). However, the contribution of T07A9.8 to the m¹A modification remains to be elucidated. Here, using an RNA immunoprecipitation (RIP) assay and a site-specific reverse transcription-quantitative PCR (RT-qPCR)-based method, we showed that T07A9.8 is required for the m¹A modification of 26 S rRNA but is not contributed to the alteration of pre-rRNA transcription under normal nutrient conditions. On the basis of its function, we named the T07A9.8 gene as rRNA adenine methyltransferase-1 (rram-1).

Materials and Methods

Worms

Worm strains were maintained as described previously (28, 29). The Bristol N2 strain was obtained from the Caenorhabditis Genetics Center and used as a wild-type strain in this study. The bacterial strain Escheria coli OP50 was used as a food source for C. elegans.

RNA interference (RNAi) experiments

T07A9.8 #1 RNAi clone was isolated from Ahringer’s RNAi library and verified by sequencing the insert. For T07A9.8 #2 clone, a 500-bp nucleotide fragment, nucleotides 21-520 of the T07A9.8 gene was amplified. The T07A9.8 DNA fragment was cloned into the L4440 vector.

HT115 (DE3) bacteria were transformed with the RNAi vector expressing dsRNA of the genes of interest and cultured on LB plates with 50 μg/ml ampicillin and 10 μg/ml tetracycline. Single colonies were inoculated in 3 ml of LB containing ampicillin and grown for over-night. These cultures were seeded on NGM plates containing 25 μg/ml carbenicillin and 1 mM isopropylthiogalactoside. To induce dsRNA expression, these plates were incubated at room temperature overnight. Synchronized L1 larvae were grown on the plates for 48 hr for PCR experiments, and for 72 hr for Western blotting.

Anti-bodies

The following anti-bodies were used in this study: anti-β-actin (code number M177-3; MBL, Japan); anti-1-methyladenosine (code number D345-3; MBL) and normal mouse IgG (catalog number sc-2025; Santa Cruz Biotechnology, USA). The rabbit anti-C. elegans T07A9.8 anti-body was raised against a synthetic peptide corresponding to 30–45 amino acids of the T07A9.8 protein. Horseradish peroxidase-linked (HPL) sheep anti-mouse IgG (catalog number NA931; GE Healthcare, UK) and HPL donkey anti-rabbit IgG (catalog number NA934; GE Healthcare) were used as secondary anti-bodies for Western blotting.

Quantification of RNA modifications by liquid chromatography/tandem mass spectrometry (LC-MS/MS)

Large RNAs were extracted using ISOGEN II (Nippon Gene, Japan) according to the manufacturer’s protocols. To quantify the methylated ribonucleotides in the rRNA, 26 S rRNA and 18 S rRNA were isolated from large RNAs using agarose gel electrophoresis and gel extraction with the NucleoSpin® gel and PCR clean-up system (MACHEREY-NAGEL, Germany) according to the instruction manual. One unit of nuclease P1 (Wako Pure Chemicals, Japan) was added to the heat-denatured RNA sample (1 to 2 μg) in 10 mM ammonium acetate buffer (pH 5.3) and then incubated for 2 h at 45°C. Subsequently, the sample was dephosphorylated by additions of venom phosphodiesterase I (0.0002 unit, Sigma, USA) and bacterial alkaline phosphatase (0.3 unit, TOYOBO, Japan) in 0.1 M ammonium bicarbonate buffer (pH 7.9) for 2 h at 37°C, as previously described (30). As an internal standard, 10 pmol of 5-bromouridine (br⁵U, Tokyo Chemical Industry Co., Ltd., Japan) were added into the digestion mixture and enzyme proteins were subsequently removed by acetone precipitation. The supernatant was evaporated to dryness and the resulting nucleoside residues were dissolved with 15 μl of HPLC-grade water (Wako Pure Chemicals).

LC-MS/MS analyses were performed on a Shimadzu Nexera™ UHPLC system coupled to LCMS-8050™ triple quadrupole mass spectrometer (Shimadzu, Japan). LC separations were carried out on a Inertsil ODS-HL™ column (3 μm, 2.1 × 150 mm, GL Science, Japan) with a Inertsil ODS-HL™ Cartridge Guard Column (3 μm, 3.0 × 10 mm, GL Science) at 30°C, at a flow rate of 0.2 ml/min. The mobile phase consisted of solvent A (0.1% formic acid) and solvent B (80% acetonitrile in 0.1% formic acid). Above water-dissolved samples (5 μl) were injected and eluted starting with 100% solvent A/0% solvent B for 6 min, followed by a 20 min linear gradient of 0 to 10% solvent B, a 1 min linear gradient of 10% to 100% solvent B, 5.5 min with 100% solvent B and 6 min re-equilibration with the initial mobile phase conditions. For determining of elution positions of ribonucleosides on the chromatogram, standard chemicals of adenosine (A, Sigma), m¹A (Santa Cruz Biotechnology), N⁶, N⁶-dimethyladenosine (⁠${\text{m}}^6{}_2\text{A}$⁠, Compton, UK), 2′-O-methyladenosine (Am, Carbosynth, UK), cytidine (C, Sigma), 5-methylcytidine (m⁵C, Sigma), and br⁵U were used, and their retention times were revealed as 19.45 min, 9.47 min, 30.34 min, 23.04 min, 3.9 min, 9.12 min and 24.0 min, respectively.

The mass spectrometer was operated using an ion-spray source at 300°C in the positive mode with unit resolution for Q1 and Q3, and other optimized parameters: interface voltage, 4.0 kV; interface current, 0.1 μA; flow rate of nebulizer gas, 3 l/min; flow rate of heating gas, 10 l/min; flow rate of drying gas, 10 l/min; collision gas (Ar), 270 KPa; DL temperature, 250°C; heat block temperature, 400°C; conversion dynode potential, 10 kV; detector potential, 2.44 kV. The multiple reaction monitoring transitions and parameters (cone voltage and collision energy of precursor and product ions, respectively) for each nucleoside is listed in Methylated nucleosides database (MNSDB, http://www.agbi.tsukuba.ac.jp/∼akiftara/MNSDB/).

Western blotting

Worms were lysed in RIPA buffer (20 mM Tris-HCl [pH 7.5], 150 mM NaCl, 2 mM EDTA, 0.8% NP-40, 0.1% SDS, 0.5% sodium deoxycholate) containing protease inhibitor cocktail (Nacalai Tesque, Japan). The extracts were subsequently separated by 15% SDS-PAGE. The proteins were transferred to Immobilon-P polyvinylidene difluoride transfer membranes (Millipore, USA). After blocking with 0.3% skim milk in TBS-T buffer (20 mM Tris-HCl [pH 7.5], 150 mM NaCl, 0.1% Tween-20) for 1 h at room temperature, the membranes were incubated overnight at 4°C with the indicated primary anti-bodies. After washing with TBS-T buffer, the membranes were incubated with HPL secondary anti-bodies for 1 h at room temperature and then washed again with TBS-T buffer. Protein signals were detected with Luminata Forte Western HRP substrate (Millipore).

RIP experiments

Total RNA was extracted using ISOGEN II (Nippon Gene) and treated with DNase I (Nippon Gene) according to the manufacturer’s instructions. Extracted RNAs were fragmented using the NEBNext Magnesium RNA Fragmentation Module (New England BioLabs, USA). After ethanol precipitation with 20 µg of glycogen, fragmented RNAs were resuspended in IPP buffer (10 mM Tris-HCl [pH 7.5], 150 mM NaCl, 0.1% NP-40). Solutions containing 1 µg anti-m¹A anti-body, 50 µl of 50% slurry protein G Sepharose 4 Fast Flow (GE Healthcare) and 950 µl IPP buffer were added to the pre-cleaned RNA solutions. The mixture was rotated overnight at 4°C. The beads were washed 3 times with IPP buffer. Immunoprecipitated RNAs were purified by phenol/chloroform extraction. After ethanol precipitation, RNAs were analysed by RT-qPCR. Total RNAs were used as input for normalization of RT-qPCR data.

A site-specific RT-qPCR-based method

Methylation of 26 S rRNA was analysed by a site-specific RT-qPCR-based method (25). Reverse transcription (RT) was performed using 500 ng total RNA in the presence of 50 units of ReverTra Ace (TOYOBO), 1.25 µM of the methylated site unanchored or anchored primer and either low (0.25 µM) or high (250 µM) concentrations of deoxynucleotide triphosphates (dNTPs). RT reactions were performed at 42°C for 60 min, and then stopped by incubation at 99°C for 5 min. RT-qPCR was performed with SYBR Premix Ex Taq II (TaKaRa, Japan) and the Thermal Cycler Dice Real Time System (TaKaRa). The methylation levels were calculated as 2^{(CTlow-CThigh)}, where the threshold cycle value obtained with the RT-qPCR reaction at the low dNTP concentration was normalized to that obtained at the high dNTP concentration. The sequences of primers used in this assay are listed in Supplementary Table S2.

RT-qPCR experiments

Extracted total RNA (1 μg) was reverse transcribed with random primers (TaKaRa) and ReverTra Ace according to manufacturer’s instructions. RT-qPCR was performed by using primers for the indicated target region (Supplementary Table S3). Each gene-expression level was normalized to the act-1 expression level.

Statistical analysis

Data are presented as the mean ± SD. The statistical significances were determined by one-way analysis of variance (ANOVA), followed by Dunnett’s test. A P-value < 0.05 was considered statistically significant.

Results

m¹A modification is present in C. elegans 26 S rRNA

We firstly isolated the 26 S rRNA via agarose gel electrophoresis following extraction (Supplementary Fig. S1A). To determine the existence of N¹-methylated adenosine in 26 S rRNA in C. elegans (Fig. 1A), we performed LC-MS/MS analysis. A single peak corresponding to m¹A modification was clearly detected in 26 S rRNA (Fig. 1B). We estimated the ratio of m¹A to unmodified adenosine and found that the m¹A/adenosine (m¹A/A) ratio was 0.588% ± 0.164 (Fig. 1C). Additionally, we also assessed the amount of 2′-O-methyladenosine (Am) and 5-methylcytidine (m⁵C), which are already identified in C. elegans 26 S rRNA (31, 32). The ratio of m¹A/A was comparable to those of Am/A and m⁵C/C (1.379% ± 0.084 and 0.954% ± 0.217, respectively) (Fig. 1B and C). Furthermore, N⁶, N⁶-dimethyladenosine (⁠${\text{m}}^6{}_2\text{A}$⁠), which is a unique modification of 18 S rRNA in yeast and human, was also observed in C. elegans 18 S rRNA, but not in 26 S rRNA (Supplementary Fig. S1B and C). These results support the validity of present experimental conditions. Consequently, we demonstrated that the 26 S rRNA of C. elegans is methylated at N¹ position of adenosine.

Fig. 1

m¹A modification is present in C. elegans 26 S rRNA. (A) Chemical structure of m¹A. Addition of a methyl group occurs on the N¹ atom in the purine base of adenosine. (B) Representative LC-MS/MS chromatograms for the m¹A, Am and m⁵C in 26S rRNA isolated from N2 worms and for each standard (0.1 pmol). (C) LC-MS/MS quantification of m¹A, Am and m⁵C in 26S rRNA isolated from N2 worms. The ratio of each modified nucleoside is indicated as a percentage of the unmodified form. Mean values ± SD are shown (n = 3).

The methyltransferase-like domain of T07A9.8 is highly conserved

Recently, NML and Rrp8 were identified as enzymes responsible for the m¹A modification of rRNA in humans and yeast, respectively (25, 26). However, an enzyme that catalyzes adenosine N¹-metylation in nematodes has not been identified. We performed a database search of C. elegans proteins, based on amino-acid sequence similarity to the mammalian NML protein and identified a single putative homolog, T07A9.8, which encodes a 343-amino acid protein, consistent with a previous report (27). Alignment of the T07A9.8 amino acid sequence with homologs revealed that T07A9.8 shares 34% and 29% amino acid identity with the human NML and yeast Rrp8 proteins, respectively, and showed high conservation within the putative RNA methyltransferase domain composed of four methyltransferase consensus motifs (motif I, post-I, II and III) (Fig. 2).

Fig. 2

Sequence alignment of C. elegans T07A9.8 and its putative homologs. Sequence alignment of T07A9.8 with its putative homologs, S. cerevisiae Rrp8 and H. sapiens NML. The dark shading highlights amino acids that are identical in at least two sequences.

T07A9.8 does not affect pre-rRNA transcription

Human NML functions as not only an epigenetic suppressor of rRNA gene transcription but also a rRNA methyltransferase (24, 25). To assess the functional role of T07A9.8, we examined whether this function is conserved in the T07A9.8 protein. Initially, we performed the knockdown of T07A9.8 with two different RNAi clones and revealed that its expression was effectively reduced at the protein and mRNA levels (Fig. 3A and Supplementary Fig. S2). Next, we evaluated pre-rRNA transcription in nematodes. Because pre-rRNA cleavage occurs within the 3’-region at early steps (33), we designed qPCR primers flanking the region and performed to detect the pre-rRNA levels (Fig. 3B, upper panel). T07A9.8 RNAi did not induce obvious changes in the pre-rRNA levels (Fig. 3B, lower panel), suggesting that T07A9.8 does not regulate pre-rRNA transcription at least under normal feeding conditions. Moreover, the levels of three major rRNA species (26 S, 18 S and 5.8 S rRNA) did not significantly change in T07A9.8 RNAi worms (Fig. 3C and Supplementary Fig. S3). Therefore, these observations suggest that, unlike human NML, T07A9.8 does not contribute to pre-rRNA transcription under the feeding conditions examined.

Fig. 3

Knockdown of T07A9.8 does not change pre-rRNA transcription. (A) Protein levels of T07A9.8 in C. elegans fed E. coli expressing either the L4440 empty-vector (control) or T07A9.8 (#1 or #2) dsRNA were analysed by Western blotting. (B) Schematic representation of the C. elegans pre-rRNA gene is shown above. Light-gray boxes indicate the 18S, 5.8S and 26S rRNA regions. The qPCR primers for quantification of pre-rRNA level are exhibited as the black arrows. The levels of pre-rRNA in C. elegans fed E. coli expressing either the L4440 empty-vector (control) or T07A9.8 (#1 or #2) dsRNA were analysed by RT-qPCR. Mean values ± SD are shown (n ≥ 3, biological replicates). ns; the value is not significant (one-way ANOVA followed by Dunnett’s test). (C) The levels of 26S, 18S and 5.8S rRNAs in C. elegans fed E. coli expressing either the L4440 empty-vector (control) or T07A9.8 (#1 or #2) dsRNA were analysed by RT-qPCR. Mean values ± SD are shown (n ≥ 3, biological replicates). ns; the value is not significant (one-way ANOVA followed by Dunnett’s test).

m¹A modification(s) exists within the sequences around A674 of 26 S rRNA

Previous reports from our group and other groups have identified nucleotides with the m¹A modification in rRNA (A645 and A2142 of 25 S rRNA in yeast and A1309 of 28 S rRNA in humans) (25, 26, 34). As shown in Fig. 4A, m¹A modification sites in 25 S rRNA of Saccharomyces cerevisiae and 28 S rRNA of Homo sapiens are well conserved in C. elegans 26 S rRNA.

Fig. 4

T07A9.8 is involved in m¹A modification around the position 674 in 26 S rRNA. (A) Sequence alignment of partial C. elegans 26S rRNA with S. cerevisiae 25S rRNA and H. sapiens 28S rRNA. The dark shading highlights nucleotides that are identical in at least two sequences. The putative m¹A modification sites are indicated by asterisks. (B, C) The m¹A levels around A674 (B) or A2246 (C) in 26S rRNA of C. elegans fed E. coli expressing either the L4440 empty-vector (control) or T07A9.8 (#1 or #2) dsRNA were investigated by RIP assay using an anti-m¹A anti-body (αm¹A). Mean values ± SD are shown (n = 3). **P < 0.01; ***P < 0.005, ns; the value is not significant (one-way ANOVA followed by Dunnett’s test).

To investigate the participation of T07A9.8 in the m¹A modification of 26 S rRNA, we performed RIP assays to pull down m¹A-modified rRNA. RNA was purified from worms, with or without T07A9.8 knockdown and precipitated using an anti-body against m¹A. RIP assays indicated that the region adjacent to adenosine at 674 of 26 S rRNA was enriched in m¹A-precipitated RNA more than that in control IgG (Fig. 4B). We next examined the involvement of T07A9.8 in the m¹A modification. Compared to control RNAi, both T07A9.8 RNAi clones significantly diminished anti-m¹A (αm¹A) anti-body precipitated-RNA levels (Fig. 4B). In contrast, primers flanking the adenosine at position 2246 of 26 S rRNA did not amplify the corresponding DNA fragment from RNA co-immunoprecipitated with αm¹A anti-body (Fig. 4C). A database search of the C. elegans genome, based on amino acid sequence similarity to an yeast BMT2 protein, a responsible factor for m¹A2142 modification of yeast 25 S rRNA (34), revealed that it is not conserved in C. elegans. Together with the results of Fig. 4C, these findings suggest that the m¹A modification is unlikely to occur around A2246 in C. elegans 26S rRNA. Therefore, we concluded that m¹A modifications occurred around A674 of the 26 S rRNA and that T07A9.8 is involved in the methylation.

26 S rRNA is methylated at N¹ atom in the adenosine at position 674

To further investigate the m¹A modification of 26 S rRNA in nematodes, we used a site-specific RT-qPCR-based method. This method is based on the principle that modified residues (including m¹A or 2′-O-m) induce a pause in the reverse transcriptase reaction at low but not at high dNTP concentrations (Supplementary Fig. S4). RT reactions at low or high dNTP concentrations were performed using a specific reverse primer, as shown in Fig. 5A, and methylation levels were calculated using the qPCR cycle threshold values obtained with cDNA at a low dNTP concentration normalized by that obtained at a high dNTP concentration. The primers used for the RT reaction were also utilized for qPCR. Consistent with RIP assays, the methylation levels neighboring position A674 were detected (15.16 ± 0.47) (Fig. 5B and Supplementary Table S1). RNAi against T07A9.8 markedly reduced the methylation levels neighboring position A674 (8.07 ± 0.79 and 6.68 ±1.53 for RNAi constructs #1 and #2, respectively) (Fig. 5B and Supplementary Table S1). In addition to position A674, we examined the methylation levels around A2246. We found that they were very low and not altered by T07A9.8 RNAi (Fig. 5C and Supplementary Table S1). These data suggested that a nucleotide(s) around A674 is modified in a T07A9.8-dependent manner, whereas a nucleotide(s) around A2246 is not highly modified. Taken together with the RIP experiments, we concluded that adenosine around position 674 of 26 S rRNA is methylated at the N¹ position in a T07A9.8-dependent process.

Fig. 5

T07A9.8 contributes to the methylation around position 674 in 26 S rRNA. (A) Partial C. elegans 26 S rRNA sequences and primers used in a site-specific RT-qPCR-based method (a, b). The predicted methylation sites are indicated by rectangles. (B, C) The methylation levels around position 674 (B) or 2246 (C) of 26 S rRNA from C. elegans fed E. coli expressing either the L4440 empty-vector (control) or T07A9.8 (#1 or #2) dsRNA were evaluated by RT-qPCR using indicated primers. Mean values ± SD are shown (n ≥ 3, biological replicates). ***P < 0.005, ns; the value is not significant (one-way ANOVA followed by Dunnett’s test).

To determine the position of the modified residue around A674, we used several specific reverse primers for the RT reaction and qPCR that had shifted alignment sites to the target RNA (Fig. 6A). We performed an RT reaction using the Rv676 primer, which covers -2 nucleotides downstream of predicted m¹A modification site (A674), at the low and high concentrations of dNTPs and calculated the methylation levels from the qPCR data. We found that the methylation level was high (16.25 ± 1.86) and that T07A9.8 RNAi constructs #1 and #2 reduced the methylation level to 6.51 ± 2.91 and 5.24 ± 2.06, respectively (Fig. 6B and Supplementary Table S1). Similarly, the methylation levels obtained after an RT reaction using the Rv675 primer, followed by qPCR, were also high and significantly decreased by T07A9.8 RNAi (Fig. 6B and Supplementary Table S1). Moreover, we performed RT reaction using the Rv674 primer, which covers adenosine at position 674. The methylation of this residue does not influence the RT reaction when using this primer. The methylation levels were strikingly lower in control-knockdown, compared to those in control-knockdown using the Rv675 and Rv676 primers (Fig. 6B and Supplementary Table S1). Further, T07A9.8 RNAi did not alter the methylation levels. Therefore, we named the T07A9.8 gene as rRNA adenine methyltransferase-1 (rram-1).

Fig. 6

T07A9.8 is required for the methylation at position 674 in 26 S rRNA. (A) Partial C. elegans 26S rRNA sequences and primers used in a site-specific RT-qPCR-based method. The identified and predicted methylation sites are indicated by rectangles. (B) The methylation levels at a region of 26S rRNA from C. elegans fed E. coli expressing either the L4440 empty-vector (control) or T07A9.8 (#1 or #2) dsRNA were evaluated by RT-qPCR using indicated primers. Mean values ± SD are shown (n ≥ 3, biological replicates). ***P < 0.005, ns; the value is not significant (one-way ANOVA followed by Dunnett’s test).

It has been reported that guanosine 668 (G668) of 26 S rRNA is methylated at the 2′-O-position (32). To validate the method with one nucleotide resolution, we tested it with the Rv668 primer that covers G668, which resulted in a further reduction in the methylation levels versus those detected using the Rv674 primer (Fig. 6B and Supplementary Table S1). Additionally, rram-1 RNAi did not alter the methylation levels, suggesting that RRAM-1 is specifically involved in the m¹A modification at 674 position. Collectively, these data indicate that A674 of 26 S rRNA is N¹-methylated in a manner dependent on RRAM-1.

Discussion

We showed that m¹A modification is present in 26 S rRNA in C. elegans. In addition, our findings indicate that A674 of 26 S rRNA is N¹-methylated in a RRAM-1-dependent manner, whereas the m¹A modification of A2246 could not be detected, unlike in yeast. Additionally, the RNAi down-regulation of rram-1 did not alter the pre-rRNA transcription levels. These results provide the first evidence for the m¹A modification of rRNA in C. elegans.

Several studies have used an RT-qPCR-based method to examine the presence of modifications in the sequences flanked by the qPCR primers (25, 35, 36). This method was based on the principle that some kinds of methylated nucleotides prevent polymerase advance during the RT step at low dNTP concentrations. However, these studies did not identify the position of the modified nucleotide because of the use of a single primer for RT. To resolve this issue, we modified ‘Reverse Transcription at Low dNTP concentrations followed by PCR (RTL-P) assay’ invented by Dong et al. (37). Although they used electrophoresis to detect the DNA product of RT-PCR, we applied qPCR method, for which methylation sites could be simply identified. For the RT reaction, we used primers that were shifted downstream of a putative modification residue, one base at a time. Intriguingly, when the 3′-end of the primer covers just on the modified residue such as m¹A674 and Gm668 of 26 S rRNA, the methylation levels detected were markedly reduced (Fig. 6B and Supplementary Table S1). These data suggest that our method can be used to identify the position of methylated residues at single-nucleotide resolution. This tool will help map RNA modifications, including m¹A and 2′-O-methylation.

Yeast Rrp8 and mammalian NML have been identified as a methyltransferase responsible for the m¹A modification of 25 S rRNA (26) and 28 S rRNA (25), respectively. We provide evidence that RRAM-1 is required for the m¹A modification at A674 of 26 S rRNA in C. elegans. Indeed, the protein sequence alignment shows a high degree of conservation in the C-terminal region, which corresponds to the methyltransferase-like domain. Meanwhile, RRAM-1 is much less contributed to pre-rRNA transcription than methylation under normal nutrient conditions. Although NML is recruited to histone H3 (when dimethylated at Lys9) through its N-terminal region and acts as an epigenetic inhibitor of rRNA transcription under low glucose conditions in cooperation with SIRT1 and SUV39H1 (24), a low conservation of the N-terminal region might lead to the loss of function regarding transcriptional regulation.

In yeast, it has been reported that the adenosine ribonucleoside substitution to uridine at position 645 of 25 S rRNA, corresponding to A674 of 26 S rRNA in C. elegans, induces the reduction of 60 S ribosomal subunit biogenesis (26). Furthermore, we have previously revealed that mammalian NML contributes to the 60 S ribosomal subunit formation (25). Considering previous findings, it is conceivable that the RRAM-1 mediated m¹A modification at position 674 in C. elegans 26 S rRNA may be involved in the proper assembly of the 60 S ribosomal subunit.

To date, several genome-wide RNAi screening analysis including T07A9.8 (rram-1) gene have been performed. Intriguingly, rram-1 has been identified as one of the genes related to the lifespan regulation (38, 39). However, the mechanisms have not been fully understood. In the present study, we revealed that RRAM-1 plays a critical role in N¹-adenine methylation of 26 S rRNA but not in the pre-rRNA transcription in normal nutrient conditions. These results raise the possibility that a reduction of N¹-adenine methylation of 26 S rRNA induced by rram-1 RNAi might be involved in the extended lifespan. Further study will be needed to clarify the contribution of RRAM-1 to the lifespan.

Author Contributions

W.Y., K.H. and A.F. designed the study. A.F. supervised the study and obtained funding. K.H., H.W., N.S., M.M. and S.A. performed knockdown experiments. W.Y. performed biochemistry experiments. N.N. and K.K. established the LC-MS/MS methods. K.H., W.Y. and A.F. wrote the manuscript. All authors reviewed the results and approved the final version of the manuscript.

Supplementary Data

Supplementary Data are available at JB Online.

Acknowledgements

The authors thank the Caenorhabditis Genetics Center for providing the strains used in this study. The authors thank the members of the Fukamizu Laboratory for helpful discussions.

Funding

This work was supported by Grants-in-Aid for Scientific Research on Inovative Areas (23116001 and 23116004 to A.F.) from the Ministry of Education, Culture, Sports, Science and Technology, Grants-in-Aid for Scientific Research (A) (17H01519 to A.F.), Grants-in-Aid for Scientific Research (C) (26450119 and 17K07746 to K.H., and 26350957 and 17K01942 to K.K.) and Grant-in-Aid for JSPS Research Fellow (17J00231 to W.Y.) from the Japan Society for the Promotion of Science.

Conflict of Interest

None declared.

Abbreviations

Abbreviations
 
• Am

  2’-O-methyladenosine

 
• br⁵U

  5-bromouridine

 
• Gm

  2’-O-methylguanosine

 
• LC-MS/MS

  liquid chromatography/tandem mass spectrometry

 
• m¹A, N¹

  methyladenosine

 
• ${\text{m}}^6{}_2\text{A},N^6$

  N⁶-dimethyladenosine

 
• m⁵C

  5-methylcytidine

 
• RIP

  RNA immunoprecipitation

Author notes

Wataru Yokoyama and Keiko Hirota authors equally contributed to this work.

Accession numbers: Amino acid and nucleotide sequence data indicated are available in the NCBI Reference Sequence/GenBank databases under the accession numbers as follows,

Amino acid sequence: C. elegans T07A9.8; NP_499920.1, S. cerevisiae RRP8; NP_010368.5, H. sapiens NML; NP_056139.1

Nucleotide sequence: C. elegans 26S rRNA; NR_000055.1, S. cerevisiae 25S rRNA; NR_132209.1, H. sapiens 28S rRNA; U13369.1