The effects of NCBP3 on METTL3‐mediated m6A RNA methylation to enhance translation process in hypoxic cardiomyocytes

Abstract Hypoxia as a crucial pathogenesis factor usually results in huge harmful effects on cardiac injury and dysfunction. Our previous study has uncovered the global transcriptome and translatome profiles of cardiomyocytes in vitro and in vivo to response to hypoxia by RNA sequencing and ribosome profiling sequencing. We observe a series of differential expressed genes between transcription and translation, which may be attributed to the hypoxia‐specific binding affinity of nuclear cap‐binding subunit 3 (NCBP3) at 5' untranslation region of target genes. Although we observe that NCBP3 can facilitate translational process in myocardium under hypoxia stress, the underlying molecular mechanism of NCBP3 for gene translation modulation remains unclear. In this study, we performed NCBP3 immunoprecipitation for mass spectrum and found that METTL3 and eIF4A2 particularly interacted with NCBP3 in hypoxic rat H9C2 cardiomyocytes. Furthermore, we observed that METTL3‐mediated N6‐methyladenosine (m6A) methylation was elevated in hypoxia, but compromised by NCBP3 or METTL3 knockdown. Finally, we also demonstrated that NCBP3/METTL3/eIF4A2 regulatory axis plays a specific role in cardiomyocytes undergoing hypoxic stress. Taken together, we unmasked NCBP3, a novel hypoxia‐specific response protein functions as a scaffold to coordinate METTL3 and eIF4A2 for enhancing gene translation by m6A RNA methylation in cardiomyocytes upon hypoxic stress.

models. [2][3] The debate over whether hypoxia benefits or harms cardiomyocyte proliferation has never been resolved, and the underlying mechanism still remains obscure.
Cells experience a series of physiological changes to adapt hypoxic stress via shifting cellular processes from general housekeeping functions to specialized hypoxia-response pathways activation.
Transcriptomic changes responding upon hypoxia are relatively well illustrated, whereas gene translation as one important regulatory means still remains a mystery. There are several proposed hypoxic translation machineries: open reading frame-mediated mRNA regulation 4 endoplasmic reticulum-mediated mRNA selection 5 IRESdependent translation initiation 6 , ribosomal hypoxia-response elements (rHRE) in the mRNA 7 and RNA binding proteins recruitment at UTRs for mRNA stabilization. 8 To monitor the localized protein synthesis and explore cotranslational folding and targeting, ribosome profiling technique (polyriboseq) based on sucrose-gradient separation of polysome-associated RNAs is allowed to assess the coding potential of mRNAs. 9 Our previous study 10 has compared the global transcriptome and translatome in hypoxic myocardium in vitro and in vivo via RNA-seq and polyribo-seq. A large number of differential expressed genes between transcriptome and translatome indicate the complicated posttranscriptional and translational reactions responding to hypoxia.
We have found that nuclear cap-binding subunit 3 (NCBP3) specifically occupies at 5' UTR of genes which are up-regulated in translatome compared with transcriptome. NCBP1, 2 and 3 constitute the cap-binding complex that binds all capped RNA and are necessary for RNA processing and intracellular localization. 11 NCBP3 has been demonstrated to engage in interacting with components of the exon junction complex and transcription and export complex including eIF4A3, MAGOH, RBM8A ALYREF and DDX39B to enhance mRNA biogenesis. 12 However, the current understandings of NCBP3 for gene translation in hypoxic condition are unknown. In this study, we continue with the investigation of myocardial ischaemia and focus on the function of NCBP3 on gene translational process, as well as the connection between NCBP3 and hypoxia. The data may help us to deeply understand the regulatory network of gene translation responding to hypoxia.

| Immunoprecipitation (IP)
The whole-cell lysates or nuclear extracts were mixed with 1 μg NCBP3

| Gel filtration assay
Gel filtration assay was performed as previously described. 13 In brief, hypotonic buffer with NP-40 was used to separate the cytoplasm and nuclei of 1x10 8 GT1-7 cells,the nuclear proteins were Each fractions with 1 ml were collected and detected NCBP3, METTL3 and eIF4A2 by Western blot assay.

| Western blot (WB) assay
The protein lysate was subjected to SDS/PAGE and transferred to PVDF membranes (Bio-Rad Laboratories). The membrane was blocked with 5% fat-free milk in PBST for 30 min, followed by incubation overnight at 4°C with final dilutions of primary antibodies against NCBP3, METTL3, eIF4A2 or GAPDH (#60004-1, Proteintech Group). After that, the membrane was washed three times and then incubated with HRP-conjugated secondary antibodies (Proteintech Group). The blotting bands were developed with ECL plus immunoblotting detection reagents (Thermo Fisher Scientific) and captured using ImageJ.

| Liquid chromatography-tandem mass spectrometry (LC-MS/MS)
Enzymatic hydrolysation was initially conducted by 12 ng/µl trypsin at 37°C for 16 h. Whole peptides were dissolved within 0.1% formic acid and passed through RP-C18 column for desalination and washed by 0.1% formic acid-acetonitrile. Peptide library was processed through trap column, analysis column and RP-C18 column by EASY-nLC1200 system. Enzymatic peptides were analysed using Q-Exactive TM . 300-1600 m/z precursor scan was performed after standard solution adjustment. Full scan model of data-dependent acquisition was conducted followed by MS2 scan; for MS1 phase,

| Dot plot
Dot plot was performed as previously described. 14 In brief, total RNA or poly (A) + mRNA was isolated as described above. RNA After being blocked with 5% non-fat milk, the membrane was incubated with specific m6A antibody (1:1000, Millipore) overnight at 4°C. Dot blots were hatched with HRP-conjugated anti-mouse immunoglobulin G (IgG) for 1 h before visualized by an imaging system (Bio-Rad Laboratories).

| RNA immunoprecipitation (RIP) assay
RNA immunoprecipitation or MeRIP experiments were conducted as previously described. 15 In brief, 1 × 10 7 H9C2 cells were har-  16 and checked the quality of clean reads using Fastqc. 17 Next, clean reads were aligned to the latest mouse genome assembly mm10 using Hisat2 v2.0.5 (non-default parameters: --rna-strandness RF --dta). 18 The transcripts were assembled and the expression levels were estimated with FPKM values using the StringTie algorithm (non-default parameters: --rf). 19 Differential mRNA and lncRNA expression among the groups was evaluated using an R package Ballgown, 20 and the significance of differences by the Benjamini & Hochberg (BH) pvalue adjustment method was computed. Gene annotation was described by Ensembl genome browser database (http://www.ensem bl.org/index.html). The R package ClusterProfiler was used to annotate the differential genes with gene ontology (GO) terms and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways. 21 The peaks of m6A at certain gene were browsed by Integrative Genomics Viewer (IGV).

| Quantitative polymerase chain reaction assay
Quantitative polymerase chain reaction (qPCR) was performed using CFX Fast real-time PCR system (Bio-Rad Laboratories). The following cycle parameters were used for all experiments: 30 s at 94°C for pre-denaturation, 20 s at 94°C, 30 s at 60°C, and 30 s at 72°C for total 45 cycles. The relative levels of mRNA for each specific gene were normalized to GAPDH. Table S1 shows the sequences for all primer sets used in these experiments.

| Statistical analysis
The results were presented as the mean ± SD. The significance of difference among the groups was assessed by Student's t test. All analysis was processed by SPSS 20 software. The p-value less than 0.05 was considered as statistically significant.

| Characterization of RNA-binding proteins interacted with NCBP3 in hypoxic H9C2 cells
Our previous study has unmasked that NCBP3 can specially recognize 'GAAGCUGCC' at 5' UTR of mRNAs, and further affects a large number of genes' translation in hypoxic H9C2 cells. 10 To further understand the role of NCBP3 in translational process, IP was conducted to pull down NCBP3 ( Figure S1A) followed by mass spectrum ( Figure S1B). The differential binding protein profiling of NCBP3 between normoxic and hypoxic rat H9C2 cardiomyocytes was identified (log2FC>1 or <−1, p < 0.05) ( Figure 1A). We found that the binding abilities with NCBP3 strengthened in 67 proteins while weakened in 35 proteins in hypoxic condition compared with normal control (Table S2). The correlations among top 20 differential expressed genes (DEGs) indicated that these binding proteins were likely to function as several complexes to coordinate with NCBP3 in post-transcriptional regulation responding to hypoxia ( Figure 1B).
GO analysis suggested that the DEGs were mainly related to RNA metabolism, stabilization, transcription factor activity, RNA polymerase regulation upon hypoxia, translation initiation process, 5' NAD-cap decapping as well as RNA N6-methyladenosine (m6A) methyltransferase complex, which was in accord with the probable biological events in our system ( Figure 1C). It was noteworthy that m6A RNA methylation and cap-dependent translational process were shown in the central position within the directed acyclic graph of functional regulation ( Figure 1D). Consistently, protein-protein interaction analysis displayed that METTL3 and eIF4A2 located in the central position of protein network and connected with multiple proteins ( Figure 1E). Taken together, we preliminary characterized a unique binding protein profiling of NCBP3 in hypoxic H9C2 cells.

| Coordination between METTL3 and eIF4A2 by NCBP3 in hypoxic condition
In order to verify the reliability of mass spectrum, we pulled down NCBP3 and detected METTL3 and eIF4A2 by WB. First, we could determine that the interaction of NCBP3 with METTL3 and eIF4A2 was both substantially strengthened in hypoxic condition compared with normal control (Figure 2A). Next, we observed that NCBP3 deficiency indeed affected the interplay between METTL3 and eIF4A2 in hypoxic but failed in normoxic H9C2 cells ( Figure 2B). Gel filtration

| The m6A profiling regulated by METTL3mediated RNA methylation in hypoxic H9C2 cells
Although METTL3 has been widely acknowledged as an essential enzyme for m6A RNA methylation, how METTL3 works with eIF4A2 in the current system is obscure. To this end, we conducted dot plot and MeRIP-seq to investigate the potential connection between transcriptome-wide m6A RNA methylation and eIF4A2 in hypoxic condition. We observed that m6A RNA methylation levels could be induced under hypoxia compared to normal control, but compromised by additional NCBP3 or METTL3 knockdown ( Figure 3A).
However, in normoxic condition, only METTL3 but not NCBP3 knockdown indeed reduced the global m6A RNA methylation ( Figure 3A).
Moreover, MeRIP-seq data indicated the four profiles of m6A differentially enriched genes (log2FC >1 or <−1, p < 0.05) that were higher or lower in hypoxia compared with control, and combined with lower or higher compared with hypoxia plus NCBP3 knockdown ( Figure 3B). Surprisingly, we notice that the profile C containing 85 genes with highly enriched m6A methylation in hypoxia and reduced in NCBP3 knockdown was 87.6% identical with DEGs of translatome data between normoxic and hypoxic H9C2 cells in our previous study 10 ( Figure 3C). In profile C, genes with differential m6A RNA methylation at 5' UTR were all included in RNA-seq data ( Figure 3D Figure 3G).
Taken together, our data determined that METTL3 contributed to m6A RNA methylation at 5' UTR to facilitate eIF4A2 location on target genes in hypoxic cardiomyocytes.

| NCBP3-METTL3-eIF4A2 as a myocytespecific regulatory axis to regulate gene expression responding to hypoxia
Although the importance of NCBP3 for coordination with METTL3 and eIF4A2 in hypoxic H9C2 cells has been determined, whether this regulatory axis is conserved and can be extended to other We observed that NCBP3 was specially increased in monocytes, smooth muscle tissues and cardiomyocytes with hypoxic condition ( Figure 4A), indicating that NCBP3 was likely to be a mesoderm-  Figure 4C). Besides that, RIP-qPCR was conducted to determine that the presence of robust occupancies of NCBP3, eIF4A2 and METTL3 at 5' UTR in human genes of Mnat1, Fgf22, Vegfa and Pdgfb, which were homologous with rat, also existed in human cardiomyocytes with hypoxic stress ( Figure 4D).
Collectively, these results indicated that NCBP3 exerted as a special scaffold to drive the coordination between METTL3 and eIF4A2 in cardiomyocytes under the hypoxia stress.

| DISCUSS ION
There are multiple steps in the process of gene expression from eukaryotic nucleus to cytoplasm. The canonical cap-binding complex comprised of NCBP1 and its binding partner NCBP2 is acquired cotranscriptionally by the precursors of all mRNAs. As an alternative accessory protein for NCBP2, NCBP3 is also capable of binding m 7 Gcap and recruiting NCBP1 via an RNase-insensitive manner in vitro. 22 Moreover, NCBP3 can also contribute to RNA splicing and exonexon junctions coordinated with exon-junction complexes. 23 In our mass spectrum data, the components of NCBP2, eIF4A3, MAGOH, RBM8A ALYREF and DDX39B are all identified both in normoxia and hypoxia H9C2 cells. But we further find the substantial interaction of eIF4A2 and METTL3 with NCBP3 especially in hypoxic stress, indicating a hypoxia-specific role in cardiomyocytes although the potential connection of these novel binding proteins with NCBP3 is never studied. IP results confirm that the interaction among NCBP3, eIF4A2 and METTL3 is likely to only appear in a very particular system (hypoxic stress). In present study, we document that NCBP3 only exerts as a special scaffold to link with eIF4A2 and METTL3,  represents the statistical significance N6-methyladenosine (m6A) RNA methylation is prevalently and abundantly enriched in RNA in eukaryotic cells. 26 A number of scholars pointed out that m6A RNA methylation has been found to be involved in multiple cellular processes, such as RNA maturation, alternative splicing and protein synthesis. [27][28] The growing evidence suggests that abnormally regulation of m6A may profoundly contribute to the carcinogenesis. 29 A series of enzymes termed as m6A 'writers', 'erasers' and 'readers' based on the different functions have been documented to install m6A RNA methylation modification. The catalytic process of METTL3 has been considered the most common m6A pathway, especially in mRNA. The dysregulation of METTL3 has been determined to be implicated in many aspects of human diseases which have prompted many researchers to explore its possible molecular mechanism. METTL3 associated with ribosomes can promote translation via translation initiation machinery at 5' UTR 30 or mRNA circularization at stop codon. 31 In current case, we determine that METTL3 recruited by NCBP3 can improve m6A RNA methylation levels and then facilitate eIF4A2 for translation process. The other enzymes for m6A RNA methylation are not involved in posttranscription regulatory complex modulated at 5' UTR by NCBP3. This is because the genes with differential translation compared with their transcription have a NCBP3 binding motif at 5' UTR, but it cannot exclude the putative functions of other enzymes such as YTHDF3 at exon-exon junctions or 3' UTR 32 for regulating translation. We believe that the differential translation of a proportion of genes induced by hypoxia is supposed to respond to hypoxic stress by the synergic and complicate effects of multiple RNA-binding proteins.
Overall, we unmasked NCBP3, a novel hypoxia-specific response protein functions as a scaffold to coordinate METTL3 and eIF4A2 for enhancing gene translation by m6A RNA methylation in cardiomyocytes upon hypoxic stress. NCBP3 can be considered as a new therapeutic target for acute myocardial infarction prevention.

ACK N OWLED G EM ENTS
This work is supported by National Natural Science Foundation of China (grant no. 81870303).

CO N FLI C T O F I NTE R E S T S
All the authors have declared no conflict of interest in this study.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.