Article

ISSN 0006-2979, Biochemistry (Moscow), 2025, Vol. 90, No. 11, pp. 1741-1756 © Pleiades Publishing, Ltd., 2025.

Russian Text © The Author(s), 2025, published in Biokhimiya, 2025, Vol. 90, No. 11, pp. 1862-1878.

1741

Activity of METTL4 Methyltransferase

Is Crucial for Maintaining Optimal Splicing

Efficiency in HeLa S3 Cells

Anastasiia K. Bolikhova

1,2,3,a

, Andrey I. Buyan

3,4#

, Maria A. Khokhlova

1,3

Sofia S. Mariasina

2,3,5

, Anton R. Izzi

1,6

, Alexander Y. Rudenko

Marina V. Serebryakova

, Alexander M. Mazur

Olga A. Dontsova

1,2,3,8

, and Petr V. Sergiev

1,2,3

Skolkovo Institute of Science and Technology, Center for Life Sciences, 121205 Skolkovo, Russia

Belozersky Institute of Physico-Chemical Biology, Lomonosov Moscow State University,

119991 Moscow, Russia

Department of Chemistry, Lomonosov Moscow State University, 119991 Moscow, Russia

Institute of Protein Research, Russian Academy of Sciences,

142290 Pushchino, Moscow Region, Russia

Research and Educational Resource Center “Pharmacy”, RUDN University, 117198 Moscow, Russia

Faculty of Bioengineering and Bioinformatics, Lomonosov Moscow State University,

119991 Moscow, Russia

Institute of Bioengineering, Research Center of Biotechnology of the Russian Academy of Sciences,

119071 Moscow, Russia

Department of Functioning of Living Systems, Shemyakin-Ovchinnikov Institute of Bioorganic Chemistry,

Russian Academy of Sciences, 117997 Moscow, Russia

e-mail: anastasia_b7@mail.ru

Received August 8, 2025

Revised November 11, 2025

Accepted November 13, 2025

Abstract— Methyltransferases that modify spliceosomal small nuclear RNAs (snRNAs) play a crucial role in

the cell by ensuring proper maturation of snRNAs, which is essential for optimal function of spliceosome.

In this study, we investigated the enzyme METTL4, which catalyzes N6-methylation of 2′-O-methyladenosine

at position 30 of U2 snRNA. Function of both the protein and the modification in splicing remains unclear.

We demonstrated that inactivation of the METTL4 gene in HeLa S3 cells leads to significant changes in al-

ternative splicing, general slowdown in spliceosome activity, and intron accumulation. In the cells lacking

METTL4, expression of the set of genes associated with ribosomal RNA maturation is reduced, and the number

of coilin-positive structures, most likely Cajal bodies, is decreased in the nuclei of these cells.

DOI: 10.1134/S0006297925602382

Keywords: splicing, small nuclear RNAs (snRNAs), METTL4, U2 snRNA, methylation

* To whom correspondence should be addressed.

# These authors contributed equally to this study.

INTRODUCTION

N6-methyladenosine (m

A) is one of the most

common modified nucleotides in eukaryotic RNA [1].

A is found in various types of RNA, including mes-

senger RNA (mRNA), transfer RNA (tRNA), ribosomal

RNA (rRNA), long non-coding RNAs, and small nuclear

RNAs (snRNAs) [2-6].

Addition of a methyl group to the N6 atom of

adenosine could perform diverse regulatory functions

depending on position of the nucleotide and type of

the modified RNA. In the nucleus, the pattern of m

distribution in pre-mRNA correlates with different

BOLIKHOVA et al.1742

BIOCHEMISTRY (Moscow) Vol. 90 No. 11 2025

splicing rates of introns [7]; in the cytoplasm, recog-

nition of m

A could either enhance translation effi-

ciency or lead to mRNA degradation [8, 9].

In the case of non-coding RNAs, modifications,

including m

A, could play regulatory and function-

al roles. For example, cessation of N6-methylation of

A1832 in 18S rRNA or A4220 in 28S rRNA disrupts

ribosome assembly, which, in turn, affects translation

efficiency [10, 11]. The modified m

A43 residue in U6

snRNA directly participates in splicing by stabilizing

interaction between the U6 snRNA and the 5′ splice

site [12, 13].

The range of enzymes that perform N6-methyla-

tion of adenosine is extremely diverse [14]. The meth-

yltransferase complex METTL3–METTL14, which has

several RNA targets and is involved in regulation of

cell differentiation, is well studied [15]. Another such

methyltransferase is METTL4, whose function has not

yet been fully deciphered. The role of this enzyme,

according to the current understanding, may vary

depending on the type and state of the cell. Under

hypoxia, METTL4 causes DNA methylation, regulating

the cluster of genes associated with hypoxia response

[16, 17]. Formation of m

A in the mitochondrial DNA

by METTL4 plays an important role in the develop-

ment of inflammation [18]. METTL4 can methylate

mRNA and microRNA under certain conditions [19,

20]. However, the constitutive and evolutionarily

conserved function of METTL4 is formation of m

from 2′-O-methyladenosine (Am) at position 30 of U2

snRNA [21,  22]. In some organisms, such as worms

and fruit flies, the enzyme that 2′-O-methylates the

30th adenine in U2 snRNA is absent, and METTL4

orthologs modify A30 instead of Am30 [23,  24]. The

30th adenine of U2 snRNA is located immediately up-

stream the branch point recognition site [25], making

modification of this residue a potential mechanism

for regulating efficiency and accuracy of splicing.

At the later stages of splicing, m

Am30 is located

next to the duplex formed by the complementary

parts of U2 and U6 [26], potentially affecting stability

and conformation of the U2–U6 complex.

In this work, we aimed to study how inactiva-

tion of the gene encoding the METTL4 protein af-

fects HeLa S3 cells. We confirmed earlier observa-

tions that disappearance of METTL4 in human cells

leads to cessation of N6-methylation of Am30 in U2

snRNA [27]. We demonstrated significant changes in

alternative splicing in the HeLa S3 ΔMETTL4 cell line,

showed that inactivation of METTL4 is accompanied

by the decrease in the number of Cajal bodies (CBs),

and general decrease in the accuracy and efficiency

of splicing. Thus, METTL4 appears to play an im-

portant role in maturation of the U2 snRNA, and its

activity is important for optimal functioning of the

spliceosome.

MATERIALS AND METHODS

Oligonucleotide synthesis. All oligonucleotides

(primers, probes) were synthesized by Lumiprobe

LLC (Russia and EU); all sequences are provided in

Table S1 in the Online Resource 1.

Cell culturing. HeLa S3 cells were cultured in

a DMEM/F12 medium (Gibco, USA). Before use, the

medium was supplemented with fetal bovine serum

(50  mL per 500  mL of medium; Gibco), a mixture

of penicillin and streptomycin antibiotics (5  mL per

500  mL of medium; Gibco), and GlutaMAX (5  mL per

500  mL of medium; Gibco). Cells were cultured at

37°C, 5%  CO

. Passaging was performed by scraping

upon reaching 70-80% confluency.

Inactivation of METTL4 gene. The METTL4 gene

was inactivated in the HeLa S3 cell line using the

CRISPR-Cas9 method. A guide RNA (gRNA) sequence

was generated using CRISPR guide RNA design (https://

benchling.com). This gRNA (5′-GGAGCCGTTCGTAATGC-

GAG-3′) targets exon  4. Sense and antisense oligonu-

cleotides (MET4_hum_CrF and MET4_hum_CrR) were

annealed and cloned into a pX458 vector (Addgene:

48138) [28].

HeLa S3 cells were transfected with the plasmid

using a Lipofectamine™ 3000 reagent (Invitrogen,

USA). After 48  h, GFP-positive cells were sorted using

a BD FACSAria™ III (BD Biosciences, USA) and plat-

ed as single clones. Individual clones were analyzed

by PCR (MET4_hum_ChF and MET4_hum_ChR) and

Sanger sequencing of the target region.

Western blotting. The following antibodies were

used to detect proteins: anti-METTL4 (Invitrogen),

anti- GAPDH (Abcam, UK).

Isolation and analysis of U2 snRNA. Total RNA

was isolated from cells using an ExtractRNA reagent

(Evrogen, Russia). U2 snRNA was isolated using a

5′-biotinylated oligonucleotide (U2_BIO), as described

in Laptev et al. [29] and in the Supplementary Meth-

ods section in the Online Resource 1.

Fluorescence in situ hybridization (FISH). DNA

probes complementary to U2 snRNA targets and la-

beled with Cy5 fluorescent dye were designed. HeLa

S3 cells were seeded onto poly-lysine slides, fixed

with a fresh 4% paraformaldehyde (PFA) solution.

Cells were permeabilized in PBS (Sigma-Aldrich, USA)

containing 0.01% Triton X-100. Background blocking

was performed with a 1% glycine solution in PBS.

Pre-hybridization was carried out in a ULTRAhyb™-

Oligo buffer (Thermo Fisher Scientific, USA); hybrid-

ization was performed in the same buffer with ad-

dition of 100  ng of probe. After washing, cells were

stained with a mixture of DAPI (1  mg/mL) and Mowiol

(1  :  1000; Sigma-Aldrich). Samples were analyzed us-

ing a Celena X fluorescence microscope (Logos Biosys-

tems, South Korea). The staining protocol is detailed

IMPORTANCE OF METTL4 FOR MAINTAINING EFFICIENT SPLICING 1743

BIOCHEMISTRY (Moscow) Vol. 90 No. 11 2025

in the Supplementary Methods section in the Online

Resource 1.

Immunofluorescence of Cajal Bodies. HeLa S3

cells were seeded onto poly-lysine slides. Cells were

fixed with a 4% PFA solution. Cells were permeabi-

lized in PBST (PBS containing 0.05% Tween-20) and

blocked with a 1% bovine serum albumin solution

in PBST. Cells were stained with primary antibod-

ies against coilin (Abcam) and secondary antibodies

against rabbit antibodies labeled with Cy5 fluorescent

dye (Thermo Fisher Scientific). After staining with a

mixture of Mowiol and DAPI, samples were analyzed

using a Nikon C2 confocal microscope (Nikon, Japan)

in 9 planes, after which the stack was projected onto

a single plane. DAPI-stained nuclei were used to de-

tect regions of interest (ROI), and number of CBs

within each ROI was automatically counted. Count-

ing was performed using ImageJ software; the script

is available at: https://github.com/bio-mother/Cajal-

bodies-counting-jython-. The staining protocol is de-

tailed in the Supplementary Methods section in the

Online Resource 1.

RT-PCR and qPCR. For total RNA extraction,

HeLa S3 cells were grown to a concentration of 1 mil-

lion/mL (~40% confluency); cells were collected and

resuspended in 300  µL of QIAzol reagent (QIAGEN,

Germany) per 1 million cells. RNA samples were ex-

tracted according to the manufacturer’s instructions.

After removal of genomic DNA using DNase  I

(Thermo Fisher Scientific), cDNA synthesis was per-

formed using a Maxima First Strand cDNA Synthe-

sis Kit for RT-qPCR (Thermo Fisher Scientific). PCR

was performed using a DreamTaq PCR Master Mix

(2X) (Thermo Fisher Scientific) according to the rec-

ommended protocol; annealing temperature for all

primers was 60°C. PCR products were separated on

a 1% agarose gel at 200  V. For real-time PCR, a Maxi-

ma SYBR Green/ROX qPCR Master Mix (Thermo Fisher

Scientific) was used.

Isolation of a newly synthesized RNA. Cells

were seeded before labeling at a concentration of

1  million/mL. The culture medium was supplement-

ed with 5-ethynyluridine (5-EU) at concentration of

0.5  mM (200  mM 5-EU stock solution in DMSO was

stored at −20°C). Cells were incubated with 5-EU for

10 and 20 min. After incubation, cells were collect-

ed, separated from the medium, and resuspended in

300  µL of a QIAzol lysis reagent (QIAGEN, Germany)

per 1 million cells. RNA samples were extracted ac-

cording to the manufacturer’s instructions. Newly

synthesized RNA labeled with 5-EU was purified us-

ing a Click-iT™ Nascent RNA Capture Kit (Invitrogen,

USA) according to the protocol described previously

[30]; a brief description of the protocol is available

in the Supplementary Methods section in the Online

Resource 1.

RNA sequencing library preparation. A NEBNext

Ultra II RNA Library Prep Kit for Illumina (NEB, USA)

was used for library preparation. Beads were mixed

with a first-strand buffer, random primers, and wa-

ter according to the manufacturer’s protocol and

incubated at 94°C for 12  min to obtain longer RNA

fragments. Subsequent steps were performed accord-

ing to the manufacturer protocol; the resulting cDNA

library was sequenced using an Illumina Hiseq1500

Sequencing System (Illumina, USA).

High-throughput sequencing data analysis. De-

multiplexed fastq files from Illumina were processed

using TrimGalore (version 0.6.6) to remove adapters

and low-quality bases from the ends of reads and

filter them by a minimum length of 20 nucleotides

[31]. After quality control using FastQC (version 0.11.9),

human genome GRCh38 indices were generated using

STAR (version 2.7.10) -runMode genomeGenerate and

gencode annotation (version 42) with parameters two-

passMode Basic and others specified in the ENCODE

standard long RNA-seq pipeline (see STAR manual) [32,

33]. To identify chimeric reads, default STAR-Fusion

parameters were added, including

--chimSegmentMin 12,

--chimJunctionOverhangMin 8,

--chimOutJunctionFormat 1,

--alignSJstitchMismatchNmax 5-1 5 5,

--chimMultimapScoreRange 3,

--chimMultimapNmax 20,

--chimScoreJunctionNonGTAG-4,

--chimNonchimScoreDropMin 10,

--alignInsertionFlush Right.

To estimate the proportion of reads mapping

to exons, they were counted on genes using featu-

reCounts from the Rsubread package (v.2.12.0) [34].

Only uniquely aligned reads were retained using sam-

tools (version 1.3.1) -q255, and those corresponding to

small and repetitive RNA genes (including ribosomal,

transfer, small nuclear and nucleolar, and microRNAs)

were filtered out according to gencode comprehen-

sive (v.42), RNAatlas, UCSC RepeatMasker, sno/miRNA,

and tRNA annotations [35-37]. For this, bedtools (ver-

sion 2.30.0) intersect was used with parameters -split,

-v, and -f 0.9 (minimum fraction of the read over-

lapping with small RNA) [38]. Next, gene and exon

coverage were calculated using featureCounts with

allowMultiOverlap  =  T and GTF.featureType  =  “gene”

or GTF.featureType  =  “exon,” respectively, using the

same gencode comprehensive annotation as in the

alignment step. Finally, genes with a non-zero total

read count in each sample and at least 3cpm (counts

per million) in at least one sample were retained.

All plots presented in the article were generated

using the R package ggplot2 (version 3.4.0). Principal

component analysis was performed using fviz_pca_ind

from the R package factoextra (version 1.0.7).

BOLIKHOVA et al.1744

BIOCHEMISTRY (Moscow) Vol. 90 No. 11 2025

Splicing rate assessment. To assess kinetics of

splicing at the level of individual donor and acceptor

sites, proportion of unspliced reads among all over-

lapping regions was calculated using the R package

FRASER (v.1.10.0) [39]. For this, bam files contain-

ing uniquely aligned reads were imported into R

using the FraserDataSet function, after which read

counting was performed using countRNAData with

parameters keepNonStandardChromosomes  =  FALSE,

filter  =  FALSE, and the genome from the BSgenome.

Hsapiens.UCSC.hg38 package (version 1.4.4). Next, for

each splicing site, Ψ values were calculated using the

calculatePSIValues function by default. These values

reflect proportion of unspliced reads overlapping the

RNA region of interest among all reads crossing it.

Finally, the splicing rate at 10 and 20 min was cal-

culated as 1  −  (Ψ_{10  min/20  min}  −  Ψ_{total RNA})

and adjusted by multiplying by (1  −  Ψ_{total RNA})

to account for intron retention events.

Splicing accuracy assessment. Splicing accuracy

was assessed by calculating the Shannon entropy of

Ψ values for 5′ and 3′ splice sites using FRASER (ver-

sion 1.10.0) [39]; a detailed description of the meth-

od is available in the Supplementary Methods section

in the Online Resource 1.

Differential expression analysis. Differential ex-

pression analysis was performed using edgeR (version

3.42.4). Read counts on gene exons were normalized

using calcNormFactors, followed by model parameter

estimation and testing using estimateDisp, glmQLFit,

and glmQLFTest.

Gene set enrichment analysis (GSEA) was per-

formed using the fgsea package (version 1.26.0) for

the genes ranked by log

FC (for differential expres-

sion) or negative log

Pv with the sign of deltaPSI

(for alternative splicing), using gene sets from MSigDB

(2022.1) [40].

We also calculated number of the reads on in-

trons and exons using gencode v.42 annotation with

addition of intronic regions, using the featureCounts

function with parameters allowMultiOverlap  =  T and

GTF.featureType  =  “exon” or “intron.” Previously, we

filtered out reads mapped to repetitive genes and

small RNA genes. Calculation of the number of reads

on introns and exons was used to assess changes in

the intron and exon expression levels in the ΔMETTL4

cells compared to the wild-type (WT) cells relative

to the overall change in gene expression (i.e., to as-

sess changes in intron usage considering differential

gene expression). This was done using edgeR (version

3.42.4) in the same way as for differential expression

analysis, but using the diffSpliceDGE function.

Alternative splicing. Alternative splicing

events were identified using rMATS (version 4.1.2)

with parameters: -t single, -variable-read-length,

-ReadLength 120, and gencode v.42 annotation [41].

Complementarity of the splice sites to snRNA was

determined as proposed in Bolikhova et al. [30]; a

brief description is available in the Supplementary

Methods section in the Online Resource 1.

RESULTS

METTL4 methyltransferase is responsible for

the formation of m

Am30 in U2 snRNA. Previous

studies have shown that METTL4 methyltransferase

in human cells is responsible for formation of m

from Am at position 30 of the spliceosomal U2 snRNA

[21, 27]. To confirm this function, we created the HeLa

S3 cell line with inactivated METTL4 gene (HeLa S3

ΔMETTL4). CRISPR/Cas9 technology and guide RNA

targeting exon 4 of the gene were used to create the

cell line (the cleavage site was chosen so that a frame-

shift mutation would eliminate the main structural

part of the protein, including the SAM binding site).

A monoclonal cell line was selected in which a single

nucleotide insertion occurred in both alleles encod-

ing METTL4 (Fig. 1a, “i1”). The high-throughput RNA

sequencing data from the HeLa S3 WT and HeLa S3

ΔMETTL4 cells confirmed presence of the insertion in

the mRNA encoding METTL4. Western blot analysis

confirmed absence of the METTL4 protein in the HeLa

S3 ΔMETTL4 cells compared to the wild-type cells

(HeLa S3 WT; Fig. 1b). Disappearance of METTL4 did

not affect growth rate of the HeLa S3 cells (Fig.S1 in

the Online Resource 1). In agreement with the previ-

ous studies, inactivation of METTL4 led to disappear-

ance of one methyl group in the corresponding region

of U2 snRNA, as shown by mass spectrometry (Fig.1c).

Absence of METTL4 methyltransferase does

not affect quantity or localization of the spliceo-

somal snRNAs. Modifications could significantly af-

fect processing, localization, and stability of certain

RNA groups [42,  43]. Using RT-qPCR, we showed that

expression of the U2 snRNA in the HeLa S3 ΔMETTL4

cells does not change compared to the wild type

(Fig. 2a). Similarly, expression of other spliceoso-

mal snRNAs does not change with disappearance of

METTL4 (Fig.2a). Localization of various snRNAs also

does not depend on the presence of METTL4 in the

cell, as shown by RT-qPCR; total RNA for the reaction

was isolated separately from the cytoplasm and nu-

clei of the HeLa S3 WT and HeLa S3 ΔMETTL4 cells

(Fig. 2b). Same distribution of the U2 snRNA in both

cell lines was additionally confirmed by staining with

FISH probes specifically binding to U2 (Fig. 2c).

Part of the U2 snRNA maturation process oc-

curs within Cajal bodies (CBs) – nuclear structures

where snRNA modification and final stages of the

small nuclear ribonucleoprotein (snRNP) maturation

take place [44-46]. It is also known that disruption

IMPORTANCE OF METTL4 FOR MAINTAINING EFFICIENT SPLICING 1745

BIOCHEMISTRY (Moscow) Vol. 90 No. 11 2025

Fig.  1. Confirmation of the HeLa S3 ΔMETTL4 phenotype. a) Sanger sequencing of the target region in the METTL4 gene

(“i1” in the cell line with gene inactivation indicates insertion of one nucleotide). gRNA indicates the sequence corre-

sponding to the guide RNA; Protospacer adjacent motif (PAM) is the site recognized by the CRISPR/Cas9 system. b)Amount

of METTL4 protein in total lysates of the HeLa S3 WT and HeLa S3 ΔMETTL4 cell lines, shown by Western blotting.

GAPDH was used as a loading control. c)Mass spectra of U2 snRNA after RNase T1 treatment. The fragment CUAAGmAUCA

AmGp (calculated average mass-to-charge ratio (m/z) is 3619) corresponds to the modified form of U2 snRNA; the frag-

ment CUAAGmAUCAAmGp (calculated average m/z is 3619) corresponds to the form without 2′-O-methylation at position 30.

Peaks corresponding to these fragments are indicated by arrows on the graph. The spectra reflect disappearance of the

modification in the cell line with METTL4 inactivation.

of snRNP maturation could affect the number of

CBs in the cell [47]. The HeLa S3 WT and HeLa S3

ΔMETTL4 cells were stained with antibodies against

coilin (a protein that forms the structure of CBs [48])

and DAPI, which stains DNA. By analyzing the num-

ber of coilin-positive signals in the 1070 WT and

780 ΔMETTL4 nuclei, we found a significant reduc-

tion in the number of such signals in the cells lack-

ing methyltransferase (Fig. 2d and e; images of cells

stained for coilin, with nuclei indicated, are available

in Fig. S1 in the Online Resource 1).

Disappearance of METTL4 leads to noticeable

changes in gene expression. Although METTL4 does

not directly affect the quantity of snRNAs, changes

in the expression levels of various genes are one

of the main mechanisms allowing the cell to re-

spond to changing conditions [49]. By performing

high-throughput RNA sequencing of the HeLa S3 WT

BOLIKHOVA et al.1746

BIOCHEMISTRY (Moscow) Vol. 90 No. 11 2025

Fig. 2. Overall effect of METTL4 inactivation on quantity and maturation of spliceosomal snRNAs. a)  Analysis of snRNA

expression in the HeLa S3 WT and HeLa S3 ΔMETTL4 cell lines performed using RT-qPCR. Six biological replicates were

used for each cell line. Expression level of the target was normalized to the expression level of the housekeeping gene

GAPDH. The graph shows mean expression value and spread of values. Gray bars denote WT, purple bars denote ΔMETTL4.

Statistical significance was determined using the two-sided Mann–Whitney U test; ns, p  >  0.05. b)  Analysis of distribution

of spliceosomal snRNAs in the HeLa S3 WT and HeLa S3 ΔMETTL4 cells performed using RT-qPCR. Three biological rep-

licates were used for each cell line. Y-axis shows the ratio of the amount of specific RNA in the nuclear and cytoplasmic

fractions. Bars reflect the mean and spread of values. Gray bars denote WT, purple bars denote ΔMETTL4; ns, p  >  0.05.

c)  Micrographs of the HeLa S3 WT and HeLa S3 ΔMETTL4 cells: nuclei stained with DAPI (left panel, blue), U2 snRNA

stained with Cy5-labeled FISH probes (middle panel, purple); right panel is a merged image. d)Distribution of the number

of coilin-positive signals in the nuclei of the HeLa S3 WT (gray) and HeLa S3 ΔMETTL4 (purple) cells. Sample sizes: 1070

and 780 nuclei for WT and ΔMETTL4, respectively. Statistical significance was determined using the unpaired Student’s

t-test; ***p  ≤  0.001. e)  Micrographs of the HeLa S3 WT and HeLa S3 ΔMETTL4 cells taken in 9 planes and projected onto a

single plane: nuclei stained with DAPI (left panel, blue), cells stained with antibodies against coilin (middle panel, purple);

the right panel is a merged image.

IMPORTANCE OF METTL4 FOR MAINTAINING EFFICIENT SPLICING 1747

BIOCHEMISTRY (Moscow) Vol. 90 No. 11 2025

Fig. 3. Analysis of expression changes occurring upon METTL4 inactivation. a) Results of differential expression analysis

of 13,813 genes (with cpm ≥ 3 in at least three samples) in the HeLa S3 ΔMETTL4 cells compared to the wild-type cells.

Vertical dashed lines correspond to |log

FC| = 0.5; horizontal line denotes FDR = 0.05. In total, 1,217 genes are more ac-

tively expressed in the ΔMETTL4 cells compared to the wild type (highlighted in blue), and 946 are less actively expressed

(highlighted in red). Genes with qPCR-confirmed differential expression are marked with a dark border and additionally

labeled. b) Confirmation of high-throughput sequencing results. RT-qPCR was performed in six replicates for each cell

line. Expression level of the target gene was normalized to the expression level of the housekeeping gene GAPDH. Gray

bars denote WT, purple bars denote ΔMETTL4. Statistical significance was determined using the one-sided Mann-Whitney

U test; * p ≤  0.05, ** p ≤  0.01. c)  Results of gene set enrichment analysis (GSEA) for 15 pathways with increased (blue) and

decreased (red) regulation and FDR < 0.05. NES – normalized enrichment score.

and HeLa S3 ΔMETTL4 cells (Fig.S2a, TableS2 in the

Online Resource1) and comparing the obtained tran-

scriptomes, we found a number of genes that are more

or less actively expressed in the cells without METTL4

compared to the wild type cells (Fig. 3a); some find-

ings were confirmed using RT-qPCR (Fig. 3b). When

conducting gene set enrichment analysis (GSEA) of

the differentially expressed genes with gene ontol-

ogy terms, we found activation of the pathways as-

sociated with the development of immune response

(“Positive regulation of Natural Killer cell activation”

and “Interferon alpha response”; Fig.  3c) in the HeLa

S3 ΔMETTL4. Among the pathways inhibited in the

HeLa S3 ΔMETTL4 cells are those associated with RNA

processing (e.g., “rRNA metabolic process” and “RNA

processing”; Fig. 3c).

BOLIKHOVA et al.1748

BIOCHEMISTRY (Moscow) Vol. 90 No. 11 2025

Fig. 4. Analysis of alternative splicing in the HeLa S3 ΔMETTL4 cells compared to the wild-type cells. a)  Overall changes

identified by examining 68,005 alternative splicing events (6,020 alternative 3′ splice sites, 4,046 alternative 5′ splice sites,

45,727 exon skipping events, 4,732 intron retention events, 7,480 mutually exclusive exons), where at least one isoform

had coverage of ≥10 reads in all replicates. Red dots indicate events with FDR <  0.05 and |ΔPSI|  >  5%. Number of alter-

native splicing events exceeding the threshold values is indicated in red on the graph. Events successfully confirmed by

RT-qPCR are additionally highlighted with a dark border and labeled. b) Confirmation of the high-throughput sequenc-

ing results using RT-PCR followed by gel electrophoresis. Each experiment was performed for three biological replicates.

For each event, the electrophoresis of RT-PCR products in agarose gel is shown below, and quantitative assessment of

the product ratio on the gel is shown above. Gray bars denote WT, purple bars denote ΔMETTL4. Statistical significance

was determined using the one-sided Mann–Whitney U test; *  p  ≤  0.05.

IMPORTANCE OF METTL4 FOR MAINTAINING EFFICIENT SPLICING 1749

BIOCHEMISTRY (Moscow) Vol. 90 No. 11 2025

Alternative splicing shifts in the HeLa S3

ΔMETTL4 cells. Presence of modified bases in spliceo-

somal snRNAs plays an important role in both regula-

tion of the splice site and branch point recognition, as

well as in maintaining interactions that ensure struc-

ture of the spliceosome [6]. Earlier studies suggested

that modification of A30 in the U2 snRNA specifically

affects splicing, as this residue is located immediate-

ly before the branch point recognition site [21, 27].

We compared the ratio of alternative splicing prod-

ucts in the HeLa S3 WT and the HeLa S3 ΔMETTL4

cells. The analysis revealed a significant number of

differentially spliced genes, some of which were con-

firmed using RT-PCR (Fig.  4, a and b; Table  S3 in the

Online Resource 1).

It can be noted that the largest number of events

changing upon the METTL4 inactivation belongs to

“Exon Skipping,” with both increased inclusion and

exclusion of the exon in the HeLa S3 ΔMETTL4 com-

pared to the wild type depending on the specific splic-

ing event.

Splicing changes in the HeLa S3 ΔMETTL4 cells

are general in nature. Modifications of snRNAs can

have both general and specific effects on splicing [6,

13]. We examined which groups of alternative exons

are most susceptible to change upon the METTL4 in-

activation. For the alternative exons that are includ-

ed to a greater, equal, or lesser extent in the HeLa

S3 ΔMETTL4 cells in comparison with the wild-type

cells, we compared lengths of the exons themselves,

as well as of the surrounding introns (Fig. 5a). As a

result, it was found that the long alternative exons

are on average more represented in the ΔMETTL4

(ΔPSI  <  −5%), while the short ones are more likely to

be excluded in the cells without the methyltransferase

(ΔPSI  >  5%). At the same time, length of the following

intron is slightly greater for the events characterized

by exon skipping in the ΔMETTL4. In the next step,

for similar groups of the alternatively spliced exons,

we compared complementarity of the 5′ splice site of

the upstream and downstream introns to the sequenc-

es of U1, U5, and U6 snRNAs, as well as the region

including the branch point – U2 snRNA (Fig.S3a and

S3b in the Online Resource 1). No significant correla-

tion between complementarity and splicing changes

in the ΔMETTL4 was found. A similar analysis was

performed for the second most changing group of

splicing events – intron retention (Fig. 5b; Fig. S3c in

the Online Resource 1). The main and only pattern

is that the longer introns are more often retained in

the cell line without the METTL4 methyltransferase.

To demonstrate potential functional role of splic-

ing changes, we identified biological processes en-

riched in the alternatively spliced genes (GSEA); the

analysis results are shown in Fig. 5c. As in the case

of differential expression, pathways related to RNA

maturation and immune response were found. A pos-

sible explanation may be that some of the splicing

changes are direct consequences of the changes in

gene expression upon METTL4 inactivation, and vice

versa, that the alternative splicing of certain genes

could directly affect expression profile of the cell.

We also determined whether the METTL4 meth-

yltransferase has a general effect on splicing by com-

paring efficiency and accuracy of splicing of all in-

trons in the cell. As a measure of splicing efficiency,

we used the average proportion of exon representa-

tion relative to the total gene coverage (Fig.5d). As a

measure of splicing accuracy, the Shannon entropy of

Ψ values of 5′ and 3′ splice sites was calculated (see

“Materials and Methods” section; Fig.5e). We showed

that the average splicing efficiency decreases upon

METTL4 inactivation (the median proportion of exon

reads in total RNA of HeLa S3 WT is 0.9097649, in

total RNA of the HeLa S3 ΔMETTL4 it is 0.9001537;

p < 10

−15

), but the average splicing accuracy remains

unchanged.

Splicing rate in the HeLa S3 ΔMETTL4 cells is

significantly lower than in the wild-type cells. Splic-

ing is an extremely dynamic process, and changes in

it may not be noticeable when considering only total

RNA. To determine how METTL4 affects the splicing

rate, we used the analysis method described in Bolik-

hova et al. [30]. Briefly, the newly synthesized RNA

after 10 and 20 min of labeling cells with 5-EU was

isolated and subjected to high-throughput sequenc-

ing (Fig.  S2a and S2b in the Online Resource  1). By

comparing the degree of splicing of each intron in

the newly synthesized RNA separately after 10 and

20 min of labeling and in the poly-A fraction of total

RNA (used as an indicator of isoform ratio in mature

RNA), we calculated the splicing rate for each donor

and acceptor site. The rate values obtained for the

RNA after 10 and 20 minutes of labeling were similar

(Fig.S2c and S2d in the Online Resource  1), confirming

applicability of this method for assessing splicing rate

of each site. Analysis of the values for RNA isolated

after 10min of labeling is presented in the main text,

and after 20 min in Fig. S4 in the Online Resource 1.

The average splicing rate in the HeLa S3 ΔMETTL4

cell line was lower than in the wild type (Fig.  6a;

Fig.  S4a in the Online Resource 1). By analogy with

the analysis of the total poly-A fraction, we calculated

the Shannon entropy of Ψ values of 5′ and 3′ splice

sites in the newly synthesized RNA (Fig. 6b; Fig. S4b

in the Online Resource  1). As seen in the graphs, in

the newly synthesized RNA isolated after 10  min, the

splicing accuracy in the ΔMETTL4 is significantly

lower than in the wild type (Fig. 6b), but this effect

disappears after 20 min of labeling. Thus, only in the

“youngest” RNAs the presence of METTL4 does affect

splicing accuracy, and this effect is quickly neutralized

BOLIKHOVA et al.1750

BIOCHEMISTRY (Moscow) Vol. 90 No. 11 2025

Fig. 5. Splicing changes upon METTL4 inactivation. a)  Distribution of lengths of alternatively included exons (central graph),

as well as the upstream and downstream introns (left and right graphs, respectively) depending on the degree of inclusion/

exclusion of the alternative exon in the HeLa S3 ΔMETTL4 cells compared to the wild type. b)  Distribution of lengths of

alternative introns (central graph), as well as the upstream and downstream exons (left and right graphs, respectively)

depending on the degree of retention of the alternative intron in the HeLa S3 ΔMETTL4 cells compared to the wild type.

c)  Results of gene set enrichment analysis (GSEA) of alternatively spliced genes with FDR < 0.05. NES, normalized enrich-

ment score. d) Analysis of global splicing efficiency by assessing average proportion of exon representation for 15,035

genes with cpm ≥ 3 in at least three samples. e) Analysis of splicing accuracy through assessment of Shannon entropy

of 5′ and 3′ splice site Ψ values. In total, 56,062 splice sites were considered, for each of which the cpm of the splicing

product was ≥1 in at least one sample. Statistical significance was determined using the one-sided Wilcoxon rank test;

ns, p > 0.05; * p ≤ 0.05; *** p ≤ 0.001; **** p≤10

−15

. Gray color denotes WT, purple denotes ΔMETTL4.

IMPORTANCE OF METTL4 FOR MAINTAINING EFFICIENT SPLICING 1751

BIOCHEMISTRY (Moscow) Vol. 90 No. 11 2025

Fig. 6. Analysis of splicing rate and accuracy (RNA isolated after 10min of 5-EU labeling) in the HeLa S3 WT and the HeLa

S3 ΔMETTL4 cells. a)  Comparison of average splicing rates. Values for 68,288 splice sites are shown, for each of which

the cpm of the splicing product is ≥10 in at least three samples. b)  Comparison of splicing accuracy through assessment

of Shannon entropy of 5′ and 3′ splice sites Ψ values. In total, 10,481 splice sites were considered, for each of which the

cpm of the product is ≥1 in at least one sample. Statistical significance was determined using the one-sided Wilcoxon

rank test; **** p≤10

−15

. Gray color denotes WT, purple denotes ΔMETTL4. c)  Relationship between the difference in splic-

ing rates (ΔMETTL4 minus WT) and the average splicing rate (divided into deciles). ART ANOVA p < 10

−15

. d)  Relation-

ship between the difference in splicing rates (ΔMETTL4 minus WT) and the expression level of the corresponding genes

(log

(TPM), values divided into deciles). ART ANOVA p < 10

–15

. Dots correspond to the median values; the lower and upper

error bars show 25% and 75% quantiles, respectively. Dashed line reflects the overall trend of the median. Significances

of the Tukey test for the difference between the first and last groups are shown on the graphs (***  p  ≤  0.001).

over time, for example, due to degradation of the in-

correctly spliced RNA.

To better understand splicing features in the

HeLa S3 ΔMETTL4 line, we performed correlation

analysis of the degree of change in the splicing rate

with various characteristics of the corresponding in-

trons and exons (only significant results are present-

ed). It was shown that the slow-down of splicing in

the absence of METTL4 is more pronounced for the

initially “slower” introns (Fig.  6c; Fig.  S4c in the On-

line Resource 1). On the other hand, splicing of the

introns that are part of the highly expressed genes,

which is initially quite fast [30], slows down slight-

ly more in the absence of METTL4 (Fig. 6d; Fig. S4d

in the Online Resource 1).

DISCUSSION

Modification of snRNAs plays an important role

in the regulation of splicing, ensuring correct recog-

nition of introns and maintaining optimal structure

of the spliceosome [6,  13, 50]. In this work, we con-

firmed that the METTL4 methyltransferase catalyzes

the N6-methylation of Am at position 30 of the U2

snRNA in human cells [21,  22]. To understand the

role of METTL4 methyltransferase in the life of a eu-

karyotic cell, we analyzed the changes occurring in

the HeLa S3 cells after inactivation of the METTL4

gene. Growth rate of the cells without METTL4 did

not differ from growth rate of the wild-type cells,

which is consistent with the results previously shown

in the HEK293T cell line [27].

Modification of certain RNAs is often a mech-

anism of post-transcriptional regulation of the ex-

pression of the corresponding genes, while changes

in the expression level of snRNAs themselves could

also be an important factor [51,  52]. We did not find

fundamental differences in the expression level of

spliceosomal snRNAs between the wild-type cells

and the cells without METTL4 (Fig.  2a). Maturation

of snRNAs following synthesis is, in turn, a complex,

BOLIKHOVA et al.1752

BIOCHEMISTRY (Moscow) Vol. 90 No. 11 2025

multi-stage process, part of which occurs in the nu-

cleus and part in the cytoplasm [53-55]. The ratio of

the amount of snRNAs in the nucleus and cytoplasm

turned out to be independent of METTL4 (Fig.  2,

b and c). At the same time, we managed to detect

a statistically significant decrease in the number of

coilin-positive signals in the HeLa S3 ΔMETTL4 cells

compared to the HeLa S3 WT (Fig.  2d). Coilin is a

known marker of CBs, but it could also be part of oth-

er intranuclear structures – histone locus bodies [56].

Histone locus bodies are non-membrane intranuclear

structures that play a role in maturation of the his-

tone mRNAs; CBs are similar structures that play an

important role in maturation of the snRNPs [45, 57].

It has been shown that disruption of the final stag-

es of snRNP maturation by reducing the number of

proteins involved in this process could lead to the in-

crease in the number of CBs, thereby helping to main-

tain optimal concentration of the spliceosome compo-

nents [47]. Decrease in the number of coilin-positive

signals hints at the presence of a mechanism by

which inactivation of METTL4 affects maturation of

the U2 snRNA, associated with the presence of U2 in

CBs. However, one cannot rule out a more complex,

possibly not snRNA-mediated, connection between

METTL4 and functioning of the histone locus bodies.

Thus, this issue requires further study.

The most obvious process that could be affected

by snRNA modification is splicing [6]. m

Am30 in the

U2 snRNA is located immediately upstream the branch

point recognition site [30], while at the later stages

of splicing, this residue is located next to the duplex

formed by the complementary parts of U2 and U6 [26].

We found that in the cells without METTL4, alternative

splicing is significantly altered, with the largest num-

ber of events changing upon disappearance of the MET-

TL4 belonging to the exon skipping, and distribution

between the more included and excluded exons in the

ΔMETTL4 is equally probable (Fig.4, a and b). It was

noted that splicing in the absence of METTL4 proceeds

more slowly (Fig. 6a), leading, apparently, to a small

but general accumulation of introns for most genes

(Fig.5d). It is important to note the implicit pattern of

the splicing slowdown. Initially, the “slower” introns

are most susceptible to the disappearance of METTL4

(Fig.  6b; Fig.  S3b in the Online Resource  1), which

could be explained by less optimized splicing in these

regions. However, the RNA molecules of more highly

expressed genes, which are normally characterized by

increased splicing rates [30], are more susceptible to

the slow-down upon METTL4 inactivation. A possible

mechanism partially explaining the observed phenom-

ena may be decrease in the concentration of the ma-

ture snRNPs associated with reduction in the number

of CBs. Such change in the availability snRNPs would

more strongly affect both less efficient splicing sites,

which initially attract snRNPs weaker, and the sites of

highly represented RNAs, which, due to their greater

quantity, require higher concentration of snRNPs to

maintain optimal splicing rates.

Modified nucleotides of snRNAs under certain

conditions could directly or indirectly regulate tran-

scription [58, 59]. In the HeLa S3 ΔMETTL4 cells, we

observe a decrease in the expression of a number of

genes associated with rRNA maturation compared to

the wild type. It should be noted that small nucleolar

RNAs (snoRNAs), such as U3, also pass through CBs

during their maturation [48,  60]. At the same time,

snoRNAs play an important role in the maturation of

ribosome subunits [61-63], which could explain con-

nection between the disruption of snRNA maturation

and inhibition of the expression of genes associated

with rRNA maturation. At the same time, in the cells

with METTL4 inactivation, more active expression of

a group of genes involved in the development of in-

flammatory response is observed. A similar effect in

response to the inhibition of splicing, leading to in-

crease in the amount of intracellular double-stranded

RNA due to hybridization of the repetitive elements

within introns, has been described in the literature

before [64].

In summary, inactivation of the METTL4 gene

in the HeLa S3 cells leads to disruption of the U2

snRNA maturation and, through a previously unstud-

ied mechanism, causes decrease in the number of

coilin-positive signals in the cells, which may affect

expression of the genes associated with rRNA matu-

ration. At the same time, the ΔMETTL4 cells exhibit

decrease in the splicing rate, theoretically explained

both by the decrease in concentration of the mature

snRNPs, including due to the reduction in the CB num-

ber, and by the “imbalance” of the spliceosome due

to the disappearance of the N6-methyl group of Am

at position 30 of the U2 snRNA.

CONCLUSION

N6-methylation of adenosine, leading to forma-

tion of m

A, is an extremely common modification

found in various types of eukaryotic RNA. m

A can

both perform a regulatory function, affecting stability

and transport of mRNA, and play a structural role,

for example, for nucleotides that are part of ribo-

somes. In this work, we studied the role of the MET-

TL4 methyltransferase, one of whose most conserved

functions is N6-methylation of Am at position 30 of

the U2 snRNA.

We showed that disappearance of METTL4 in

the HeLa S3 cells leads to a slow-down in splicing,

thereby causing intron accumulation. A possible

reason for the slow-down in splicing is disruption

IMPORTANCE OF METTL4 FOR MAINTAINING EFFICIENT SPLICING 1753

BIOCHEMISTRY (Moscow) Vol. 90 No. 11 2025

of snRNP maturation associated with the decrease in

the number of CBs. Another possible consequence

of the decrease in the number of CBs is reduc-

tion in concentration of the RNA genes associated

with rRNA maturation, also observed in the cells

without METTL4.

Although importance of METTL4 for splicing effi-

ciency in the HeLa S3 cells is clear, the exact mecha-

nism linking methylation, splicing, and maturation of

various RNAs remains to be investigated.

Abbreviations

5-EU 5-ethynyluridine

Am 2′-O-methyladenosine

CBs Cajal bodies

gRNA Guide RNA

A N6-methyladenosine

Am N6,2′-O-dimethyladenosine

qPCR quantitative polymerase chain reaction

RT-qPCR reverse transcription quantitative PCR

snRNA small nuclear RNA

snRNP small nuclear ribonucleoprotein

Supplementary information

The online version contains supplementary material

available at https://doi.org/10.1134/S0006297925602382.

Acknowledgments

We are grateful to the Lomonosov Moscow State Uni-

versity Development Program for providing access

tothe Celena X multiparametric cell analysis system.

Contributions

A. K. Bolikhova and A. I. Buyan designed and conduct-

ed experiments, analyzed data, and wrote the man-

uscript. M. A. Khokhlova conducted experiments and

analyzed data. S. S. Mariasina designed and conduct-

ed experiments, analyzed data, and wrote the manu-

script. A. R. Izzi, A. Y. Rudenko, M. V. Serebryakova, and

A. M. Mazur conducted experiments. O. A. Dontsova su-

pervised the work. P. V. Sergiev supervised the work,

conceptualized the study, and wrote the manuscript.

Funding

This research was financially supported by the state

assignment of Lomonosov Moscow State University.

The experimental work was supported by the Russian

Science Foundation (grant no.21-64-00006-P).

Ethics approval and consent to participate

This work does not contain any studies involving hu-

man and animal subjects.

Conflict of interest

The authors of this work declare that they have

noconflicts of interest.

REFERENCES

1. Jiang, X., Liu, B., Nie, Z., Duan, L., Xiong, Q., Jin, Z.,

Yang,C., and Chen,Y. (2021) Therole of m6A modifi-

cation in the biological functions and diseases, Signal

Transduct. Target. Ther., 6, 74, https://doi.org/10.1038/

s41392-020-00450-x.

2. Boo, S.H., and Kim, Y.K. (2020) Theemerging role of

RNA modifications in the regulation of mRNA stabili-

ty, Exp. Mol. Med., 52, 400-408, https://doi.org/10.1038/

s12276-020-0407-z.

3. Suzuki, T. (2021) The expanding world of tRNA

modifications and their disease relevance, Nat. Rev.

Mol. Cell Biol., 22, 375-392, https://doi.org/10.1038/

s41580-021-00342-0.

4. Sergiev, P. V., Aleksashin, N. A., Chugunova, A. A.,

Polikanov, Y.S., and Dontsova, O.A. (2018) Structural

and evolutionary insights into ribosomal RNA

methylation, Nat. Chem. Biol., 14, 226-235, https://

doi.org/10.1038/nchembio.2569.

5. Kazimierczyk, M., and Wrzesinski, J. (2021) Long

non-coding RNA epigenetics, Int.J. Mol. Sci., 22, 6166,

https://doi.org/10.3390/IJMS22116166.

6. Morais,P., Adachi,H., and Yu, Y.T. (2021) Spliceosomal

snRNA epitranscriptomics, Front. Genet., 12, 652129,

https://doi.org/10.3389/FGENE.2021.652129/XML.

7. Louloupi, A., Ntini,E., Conrad,T., and Ørom, U.A. V.

(2018) Transient N-6-methyladenosine transcrip-

tome sequencing reveals a regulatory role of m6A in

splicing efficiency, Cell. Rep., 23, 3429-3437, https://

doi.org/10.1016/j.celrep.2018.05.077.

8. Wang,X., Zhao, B.S., Roundtree, I.A., Lu,Z., Han,D.,

Ma, H., Weng,X., Chen, K., Shi,H., and He,C. (2015)

N6-methyladenosine modulates messenger RNA

translation efficiency, Cell, 161, 1388-1399, https://

doi.org/10.1016/j.cell.2015.05.014.

9. Wang,X., Lu,Z., Gomez,A., Hon, G.C., Yue,Y., Han,D.,

Fu, Y., Parisien, M., Dai, Q., Jia, G., Ren, B., Pan, T.,

and He, C. (2013) N6-methyladenosine-dependent

regulation of messenger RNA stability, Nature, 505,

117-120, https://doi.org/10.1038/nature12730.

10. Peng,H., Chen,B., Wei,W., Guo,S., Han,H., Yang,C.,

Ma,J., Wang,L., Peng,S., Kuang,M., and Lin,S. (2022)

N6-methyladenosine (m6A) in 18S rRNA promotes

fatty acid metabolism and oncogenic transformation,

Nat. Metab., 4, 1041-1054, https://doi.org/10.1038/

s42255-022-00622-9.

11. Ma,H., Wang,X., Cai,J., Dai,Q., Natchiar, S.K., Lv,R.,

Chen, K., Lu, Z., Chen, H., Shi, Y. G., Lan, F., Fan, J.,

Klaholz, B. P., Pan, T., Shi, Y., and He, C. (2019) N

6-Methyladenosine methyltransferase ZCCHC4 medi-

ates ribosomal RNA methylation, Nat. Chem. Biol., 15,

88-94, https://doi.org/10.1038/S41589-018-0184-3.

12. Warda, A. S., Kretschmer, J., Hackert, P., Lenz, C.,

Urlaub, H., Höbartner, C., Sloan, K. E., and Bohnsack,

M.T. (2017) Human METTL16 is a N6-methyladenos-

BOLIKHOVA et al.1754

BIOCHEMISTRY (Moscow) Vol. 90 No. 11 2025

ine (m6A) methyltransferase that targets pre-mR-

NAs and various non-coding RNAs, EMBO Rep., 18,

2004-2014, https://doi.org/10.15252/EMBR.201744940.

13. Ishigami, Y., Ohira, T., Isokawa, Y., Suzuki, Y., and

Suzuki, T. (2021) A single m6A modification in U6

snRNA diversifies exon sequence at the 5′ splice

site, Nat. Commun., 12, 3244, https://doi.org/10.1038/

s41467-021-23457-6.

14. Yang, Y., Hsu, P. J., Chen, Y.S., and Yang, Y. G. (2018)

Dynamic transcriptomic m6A decoration: Writers,

erasers, readers and functions in RNA metabolism,

Cell. Res., 28, 616-624, https://doi.org/10.1038/S41422-

018-0040-8,.

15. Liu, J., Yue, Y., Han, D., Wang, X., Fu, Y., Zhang, L.,

Jia,G., Yu,M., Lu, Z., Deng,X., Dai,Q., Chen,W., and

He, C. (2014) A METTL3-METTL14 complex mediates

mammalian nuclear RNA N6-adenosine methylation,

Nat. Chem. Biol., 10, 93-95, https://doi.org/10.1038/

nchembio.1432.

16. Lai, J. C. Y., Hsu, K. W., and Wu, K. J. (2024) Inter-

rogation of the interplay between DNA N6-meth-

yladenosine (6mA) and hypoxia-induced chroma-

tin accessibility by a randomized empirical model

(EnrichShuf), Nucleic Acids Res., 52, 13605-13624,

https://doi.org/10.1093/NAR/GKAE1152.

17. Hsu, K. W., Lai, J.C., Chang, J. S., Peng, P. H., Huang,

C. H., Lee, D. Y., Tsai, Y. C., Chung, C. J., Chang, H.,

Chang, C.H., Chen, J.L., Pang, S.T., Hao,Z., Cui, X.L.,

He,C., and Wu, K.J. (2022) METTL4-mediated nuclear

N6-deoxyadenosine methylation promotes metastasis

through activating multiple metastasis-inducing tar-

gets, Genome. Biol., 23, 249, https://doi.org/10.1186/

s13059-022-02819-3.

18. Zheng, L., Chen, X., He, X., Wei, H., Li, X., Tan, Y.,

Min,J., Chen,M., Zhang,Y., Dong,M., Yin,Q., Xue,M.,

Zhang, L., Huo, D., Jiang, H., Li, T., Li, F., Wang, X.,

Li, X., and Chen, H. (2025) METTL4-Mediated mito-

chondrial DNA N6-methyldeoxyadenosine promot-

ing macrophage inflammation and atherosclerosis,

Circulation, 151, 946-965, https://doi.org/10.1161/

CIRCULATIONAHA.124.069574.

19. Sang, A., Zhang, J., Zhang, M., Xu, D., Xuan, R.,

Wang,S., Song,X., and Li,X. (2024) METTL4 mediat-

ed-N6-methyladenosine promotes acute lung injury by

activating ferroptosis in alveolar epithelial cells, Free

Radic. Biol. Med., 213, 90-101, https://doi.org/10.1016/

j.freeradbiomed.2024.01.013.

20. Van den Homberg, D.A.L., van der Kwast, R.V.C.T.,

Quax, P.H.A., and Nossent, A.Y. (2022) N-6-methylad-

enosine in vasoactive microRNAs during hypoxia;

a novel role for METTL4, Int.J. Mol. Sci., 23, 1057,

https://doi.org/10.3390/IJMS23031057.

21. Chen, H., Gu, L., Orellana, E. A., Wang, Y., Guo, J.,

Liu, Q., Wang, L., Shen, Z., Wu, H., Gregory, R. I.,

Xing, Y., and Shi, Y. (2020) METTL4 is an snRNA

m6Am methyltransferase that regulates RNA splicing,

Cell. Res., 30, 544-547, https://doi.org/10.1038/S41422-

019-0270-4,.

22. Luo, Q., Mo, J., Chen, H., Hu, Z., Wang, B., Wu, J.,

Liang,Z., Xie,W., Du,K., Peng,M., Li,Y., Li,T., Zhang,Y.,

Shi, X., Shen, W. H., Shi, Y., Dong, A., Wang, H., and

Ma,J. (2022) Structural insights into molecular mech-

anism for N6-adenosine methylation by MT-A70 fam-

ily methyltransferase METTL4, Nat. Commun., 13,

5636, https://doi.org/10.1038/s41467-022-33277-x.

23. Gu, L., Wang, L., Chen, H., Hong, J., Shen, Z.,

Dhall, A., Lao, T., Liu, C., Wang, Z., Xu, Y., Tang,

H. W., Chakraborty, D., Chen, J., Liu, Z., Rogulja, D.,

Perrimon, N., Wu, H., and Shi, Y. (2020) CG14906

(mettl4) mediates m6A methylation of U2 snRNA in

Drosophila, Cell Discov., 6, 44, https://doi.org/10.1038/

s41421-020-0178-7.

24. Sendinc, E., and Shi, Y. (2023) RNA m6A methyla-

tion across the transcriptome, Mol. Cell., 83, 428-441,

https://doi.org/10.1016/j.molcel.2023.01.006.

25. Van der Feltz, C., and Hoskins, A. A. (2019) Struc-

tural and functional modularity of the U2 snRNP in

pre-mRNA splicing, Crit. Rev. Biochem. Mol. Biol.,

54, 443-465, https://doi.org/10.1080/10409238.2019.

1691497.

26. Karunatilaka, K. S., and Rueda, D. (2014) Post-tran-

scriptional modifications modulate conformational

dynamics in human U2-U6 snRNA complex, RNA, 20,

16-23, https://doi.org/10.1261/RNA.041806.113.

27. Goh, Y.T., Koh, C.W.Q., Sim, D.Y., Roca,X., and Goh,

W. S. S. (2020) METTL4 catalyzes m6Am methylation

in U2 snRNA to regulate pre-mRNA splicing, Nucleic

Acids Res., 48, 9250-9261, https://doi.org/10.1093/NAR/

GKAA684.

28. Ran, F. A., Hsu, P. D., Wright, J., Agarwala, V., Scott,

D.A., and Zhang,F. (2013) Genome engineering using

the CRISPR-Cas9 system, Nat. Protoc., 8, 2281-2308,

https://doi.org/10.1038/nprot.2013.143.

29. Laptev,I., Shvetsova,E., Levitskii,S., Serebryakova,M.,

Rubtsova, M., Zgoda, V., Bogdanov, A., Kamenski, P.,

Sergiev, P., and Dontsova, O. (2020) METTL15 inter-

acts with the assembly intermediate of murine mito-

chondrial small ribosomal subunit to form m4C840

12S rRNA residue, Nucleic Acids Res., 48, 8022-8034,

https://doi.org/10.1093/NAR/GKAA522.

30. Bolikhova, A.K., Buyan, A.I., Mariasina, S.S., Rudenko,

A.Y., Chekh, D.S., Mazur, A.M., Prokhortchouk, E.B.,

Dontsova, O.A., and Sergiev, P.V. (2024) Study of the

RNA splicing kinetics via invivo 5-EU labeling, RNA,

30, 1356-1373, https://doi.org/10.1261/rna.079937.123.

31. Krueger, F., James, F., Ewels, P., Afyounian, E.,

Weinstein, M., Schuster-Boeckler, B., Hulselmans,

G., sclamons (2023) FelixKrueger/TrimGalore: ver-

sion0.6.8, https://doi.org/10.5281/ZENODO.7579519.

32. Dobin, A., Davis, C. A., Schlesinger, F., Drenkow, J.,

Zaleski,C., Jha,S., Batut,P., Chaisson,M., and Gingeras,

T.R. (2013) STAR: Ultrafast universal RNA-seq aligner,

IMPORTANCE OF METTL4 FOR MAINTAINING EFFICIENT SPLICING 1755

BIOCHEMISTRY (Moscow) Vol. 90 No. 11 2025

Bioinformatics, 29, 15-21, https://doi.org/10.1093/

BIOINFORMATICS/BTS635.

33. Loveland, J. E., Mudge, J. M., Sisu, C., Wright, J. C.,

Armstrong,J., Barnes,I., Berry,A., Bignell,A., Boix,C.,

Carbonell Sala, S., Cunningham, F., Di Domenico, T.,

Donaldson,S., Fiddes, I.T., García Girón,C., Gonzalez,

J. M., Grego, T., Hardy, M., Hourlier, T., Howe, K.,

Hunt, T., et al. (2021) GENCODE, Nucleic Acids Res.,

49, D916-D923, https://doi.org/10.1093/NAR/GKAA1087.

34. Liao,Y., Smyth, G.K., and Shi,W. (2019) TheRpackage

Rsubread is easier, faster, cheaper and better for align-

ment and quantification of RNA sequencing reads,

Nucleic Acids Res., 47, e47, https://doi.org/10.1093/

NAR/GKZ114.

35. Danecek, P., Bonfield, J. K., Liddle, J., Marshall, J.,

Ohan, V., Pollard, M. O., Whitwham, A., Keane, T.,

McCarthy, S. A., Davies, R. M., and Li, H. (2021)

Twelve years of SAMtools and BCFtools, Gigascience,

10, giab008, https://doi.org/10.1093/GIGASCIENCE/

GIAB008.

36. Lorenzi,L., Chiu, H.S., Avila Cobos,F., Gross,S., Volders,

P. J., Cannoodt, R., Nuytens, J., Vanderheyden, K.,

Anckaert, J., Lefever, S., Tay, A. P., de Bony, E. J.,

Trypsteen,W., Gysens,F., Vromman,M., Goovaerts,T.,

Hansen, T. B., Kuersten, S., Nijs, N., Taghon, T.,

Vermaelen, K., Bracke, K., Saeys, Y., Meyer, T.,

Deshpandeet,N., etal. (2021) TheRNA Atlas expands

the catalog of human non-coding RNAs, Nat. Biotech-

nol., 39, 1453-1465, https://doi.org/10.1038/S41587-

021-00936-1.

37. Karolchik, D., Hinrichs, A. S., Furey, T. S., Roskin,

K.M., Sugnet, C.W., Haussler,D., and Kent, W.J. (2004)

The UCSC table browser data retrieval tool, Nucleic

Acids Res., 32, D493-D496, https://doi.org/10.1093/

NAR/GKH103.

38. Quinlan, A.R., and Hall, I.M. (2010) BEDTools: Aflex-

ible suite of utilities for comparing genomic features,

Bioinformatics, 26, 841-842, https://doi.org/10.1093/

BIOINFORMATICS/BTQ033.

39. Mertes, C., Scheller, I. F., Yépez, V. A., Çelik, M. H.,

Liang, Y., Kremer, L. S., Gusic, M., Prokisch, H., and

Gagneur, J. (2022) Author Correction: Detection of

aberrant splicing events in RNA-seq data using

FRASER, Nat. Commun., 13, 3474, https://doi.org/

10.1038/S41467-022-31242-2.

40. Subramanian, A., Kuehn, H., Gould, J., Tamayo, P.,

and Mesirov, J. P. (2007) GSEA-P: A desktop applica-

tion for gene set enrichment analysis, Bioinformatics,

23, 3251-3253, https://doi.org/10.1093/bioinformatics/

btm369.

41. Wang, Y., Xie, Z., Kutschera, E., Adams, J. I., Kadash-

Edmondson, K. E., and Xing, Y. (2024) rMATS-turbo:

an efficient and flexible computational tool for alter-

native splicing analysis of large-scale RNA-seq data,

Nat. Protoc., 19, 1083-1104, https://doi.org/10.1038/

S41596-023-00944-2.

42. Lirussi,L., Demir,Ö., You, P., Sarno, A., Amaro, R. E.,

and Nilsen,H. (2021) RNA metabolism guided by RNA

modifications: the role of SMUG1 in rRNA quality

control, Biomolecules, 11, 76, https://doi.org/10.3390/

BIOM11010076.

43. Qiu, L., Jing, Q., Li, Y., and Han, J. (2023) RNA mod-

ification: mechanisms and therapeutic targets,

Mol. Biomed., 4, 25, https://doi.org/10.1186/S43556-

023-00139-X.

44. Darzacq, X., Jády, B. E., Verheggen, C., Kiss, A. M.,

Bertrand, E., and Kiss, T. (2002) Cajal body-specific

small nuclear RNAs: a novel class of 2′-O-methyla-

tion and pseudouridylation guide RNAs, EMBO J.,

21, 2746-2756, https://doi.org/10.1093/EMBOJ/

21.11.2746.

45. Nesic,D., Tanackovic,G., and Krämer,A. (2004) Arole

for Cajal bodies in the final steps of U2 snRNP bio-

genesis, J. Cell Sci., 117, 4423-4433, https://doi.org/

10.1242/JCS.01308.

46. Staněk, D. (2017) Cajal bodies and snRNPs – friends

with benefits, RNA Biol., 14, 671-679, https://doi.org/

10.1080/15476286.2016.1231359.

47. Novotný, I., Malinová, A., Stejskalová, E., Matějů, D.,

Klimešová, K., Roithová, A., Švéda, M., Knejzlík, Z.,

and Staněk,D. (2014) SART3-dependent accumulation

of incomplete spliceosomal snRNPs in Cajal bodies,

Cell Rep., 10, 429-440, https://doi.org/10.1016/j.celrep.

2014.12.030.

48. Staněk, D. (2023) Coilin and Cajal bodies, Nucleus,

14, 2256036, https://doi.org/10.1080/19491034.2023.

2256036.

49. Casamassimi,A., and Ciccodicola,A. (2019) Transcrip-

tional regulation: molecules, involved mechanisms,

and misregulation, Int.J. Mol. Sci., 20, 1281, https://

doi.org/10.3390/IJMS20061281.

50. Duchemin, A., O’Grady, T., Hanache, S., Mereau, A.,

Thiry, M., Wacheul, L., Michaux, C., Perpète, E.,

Hervouet, E., Peixoto, P., Ernst, F. G. M., Audic, Y.,

Dequiedt,F., Lafontaine, D.L.J., and Mottet,D. (2021)

DHX15-independent roles for TFIP11 in U6 snRNA

modification, U4/U6.U5 tri-snRNP assembly and

pre-mRNA splicing fidelity, Nat. Commun., 12, 6648,

https://doi.org/10.1038/s41467-021-26932-2.

51. Vazquez-Arango, P., and O’Reilly, D. (2018) Variant

snRNPs: new players within the spliceosome system,

RNA Biol., 15, 17-25, https://doi.org/10.1080/15476286.

2017.1373238.

52. Devany, E., Park, J. Y., Murphy, M. R., Zakusilo, G.,

Baquero, J., Zhang, X., Hoque, M., Tian, B., and

Kleiman, F. E. (2016) Intronic cleavage and polya-

denylation regulates gene expression during DNA

damage response through U1 snRNA, Cell Discov., 2,

16013, https://doi.org/10.1038/celldisc.2016.13.

53. Köhler, A., and Hurt, E. (2007) Exporting RNA from

the nucleus to the cytoplasm, Nat. Rev. Mol. Cell Biol.,

8, 761-773, https://doi.org/10.1038/nrm2255.

BOLIKHOVA et al.1756

BIOCHEMISTRY (Moscow) Vol. 90 No. 11 2025

54. Pánek, J., Roithová, A., Radivojević, N., Sýkora, M.,

Prusty, A. B., Huston, N., Wan, H., Pyle, A. M.,

Fischer, U., and Staněk, D. (2023) The SMN complex

drives structural changes in human snRNAs to en-

able snRNP assembly, Nat. Commun., 14, 6580, https://

doi.org/10.1038/s41467-023-42324-0.

55. Zieve, G. W. (1987) Cytoplasmic maturation of

the snRNAs, J. Cell Physiol., 131, 247-254, https://

doi.org/10.1002/JCP.1041310215.

56. Nizami, Z., Deryusheva, S., and Gall, J. G. (2010)

The Cajal body and histone locus body, Cold Spring

Harb. Perspect. Biol., 2, a000653, https://doi.org/

10.1101/CSHPERSPECT.A000653.

57. Geisler, M. S., Kemp, J. P., and Duronio, R. J. (2023)

Histone locus bodies: a paradigm for how nuclear

biomolecular condensates control cell cycle regu-

lated gene expression, Nucleus, 14, 2293604, https://

doi.org/10.1080/19491034.2023.2293604.

58. Mauer, J., Sindelar, M., Despic, V., Guez, T., Hawley,

B. R., Vasseur, J. J., Rentmeister, A., Gross, S. S.,

Pellizzoni, L., Debart, F., Goodarzi, H., and Jaffrey,

S.R. (2019) FTO controls reversible m6Am RNA meth-

ylation during snRNA biogenesis, Nat. Chem. Biol.,

15, 340-347, https://doi.org/10.1038/S41589-019-0231-8.

59. Karijolich, J., Kantartzis, A., and Yu, Y.T. (2010) RNA

modifications: a mechanism that modulates gene

expression, Methods Mol. Biol., 629, 1-19, https://

doi.org/10.1007/978-1-60761-657-3_1.

60. Verheggen, C., Lafontaine, D. L., Samarsky, D.,

Mouaikel, J., Blanchard, J. M., Bordonné, R., and

Bertrand,E. (2002) Mammalian and yeast U3 snoRNPs

are matured in specific and related nuclear com-

partments, EMBO J., 21, 2736-2745, https://doi.org/

10.1093/EMBOJ/21.11.2736.

61. Dragon, F., Gallagher, J. E., Compagnone-Post,

P. A., Mitchell, B. M., Porwancher, K. A., Wehner,

K. A., Wormsley, S., Settlage, R. E., Shabanowitz, J.,

Osheim, Y., Beyer, A. L., Hunt, D. F., and Baserga,

S. J. (2002) A large nucleolar U3 ribonucleoprotein

required for 18S ribosomal RNA biogenesis, Nature,

417, 967-970, https://doi.org/10.1038/nature00769.

62. Kass, S., Tyc, K., Steitz, J. A., and Sollner-Webb, B.

(1990) The U3 small nucleolar ribonucleoprotein

functions in the first step of preribosomal RNA pro-

cessing, Cell, 60, 897-908, https://doi.org/10.1016/

0092-8674(90)90338-F.

63. Lackmann,F., Belikov,S., Burlacu,E., Granneman,S.,

and Wieslander, L. (2018) Maturation of the 90S

pre-ribosome requires Mrd1 dependent U3 snoRNA

and 35S pre-rRNA structural rearrangements, Nucle-

ic Acids Res., 46, 3692-3706, https://doi.org/10.1093/

nar/GKY036.

64. Bowling, E. A., Wang, J. H., Gong, F., Wu, W., Neill,

N. J., Kim, I. S., Tyagi, S., Orellana, M., Kurley, S. J.,

Dominguez-Vidaña, R., Chung, H. C., Hsu, T. Y.,

Dubrulle, J., Saltzman, A. B., Li, H., Meena, J. K.,

Canlas, G. M., Chamakuri, S., Singh, S., Simon, L. M.,

Olson, C., Dobrolecki, L., Lewis, M., Zhang, B., Gold-

ing, I., Rosen, J. M., Young, D. W., Malovannaya, A.,

Stossi, F. Miles, G. et al. (2021) Spliceosome-targeted

therapies trigger an antiviral immune response in tri-

ple-negative breast cancer, Cell, 184, 384-403, https://

doi.org/10.1016/j.cell.2020.12.031.

Publisher’s Note. Pleiades Publishing remains

neutral with regard to jurisdictional claims in published

maps and institutional affiliations. AI tools may have

been used in the translation or editing of this article.