The landscape of N 1 -methyladenosine (m 1 A) modification in mRNA of the decidua in severe preeclampsia

Recent discoveries in mRNA modification have highlighted N 1 -methyladenosine (m 1 A), but its role in preeclampsia (PE) pathogenesis remains unclear. In this study, we utilized methylated RNA immunoprecipitation sequencing (MeRIP-seq) and RNA sequencing (RNA-seq) to identify m 1 A peaks and the expression profile of mRNA in the decidua of humans with early-onset PE (EPE), late-onset PE (LPE), and normal pregnancy (NP). We assessed the m 1 A modification patterns in preeclamptic decidua using 10 m 1 A modulators. Our bioinformatic analysis focused on differentially methylated mRNAs (DMGs) and differentially expressed mRNAs (DEGs) in pairwise comparisons of EPE vs NP, LPE vs NP, and EPE vs LPE, as well as m 1 A-related DEGs. The comparisons of EPE vs NP, LPE vs NP, and EPE vs LPE identified 3110, 2801, and 2818 DMGs, respectively. We discerned three different m 1 A modification patterns from this data. Further analysis revealed that key PE-related DMGs and m 1 A-related DEGs predominantly influence signaling pathways critical for decidualization, including cAMP, MAPK, PI3K-Akt, Notch, and TGF-β pathways. Additionally, these modifications impact pathways related to vascular smooth muscle contraction, estrogen signaling, and relaxin signaling, contributing to vascular dysfunction. Our findings demonstrate that preeclamptic decidua exhibits unique mRNA m 1 A modification patterns and gene expression profiles that significantly alter signaling pathways essential for both decidualization and vascular dysfunction. These differences in m 1 A modification patterns provide valuable insights into the molecular mechanisms influencing the decidualization process and vascular function in the pathogenesis of PE. These m 1 A modification regulators could potentially serve as potent biomarkers or therapeutic targets for PE, warranting further investigation.


Introduction
Preeclampsia (PE) is a multisystem syndrome that uniquely occurs during human pregnancy.It is characterized by the new onset of hypertension, proteinuria, and other signs of maternal vascular damage after 20 weeks of gestation.PE affects approximately 8% of first-time pregnancies worldwide each year, impacting eight million mother-infant pairs.It is recognized as one of the leading causes of maternal and perinatal morbidity and mortality [1,2].Severe PE (sPE) is diagnosed based on a further elevation of systolic pressure ≥ 160 mm Hg or diastolic pressure ≥ 110 mm Hg, or any of the following: thrombocytopenia, impaired liver function, progressive renal insufficiency, pulmonary edema, and the new onset of cerebral or visual disturbances [3].
Currently, the etiology of PE remains unclear, but dysregulated decidualization has been a subject of interest in studying the genesis of PE.Previous studies have confirmed and augmented the results through global transcriptional profiling of decidua in sPE [4].It was surprising to find that the decidual gene transcription profile was altered in sPE.Additionally, decidual cells from patients with sPE, when dedifferentiated in vitro, failed to redecidualize in culture [5].Therefore, the decidua is considered a prime candidate for the genesis of PE [6].In humans, decidualization, the formation of the decidua, is independent of the presence of a conceptus.It is a progressive process that involves hormonally regulated differentiation of human endometrial stromal cells (ESCs), morphologically transforming into enlarged round-shaped cells from a fibroblast-like population driven by genetic reprogramming [7].
N 1 -methyladenosine (m 1 A) is a recently identified mRNA modification that is found enriched around the first codon upstream of the first splice site and alternative translation initiation sites.This modification is highly conserved in human cells and is responsive to physiological conditions.It correlates positively with protein production, indicating a functional role in promoting translation of methylated mRNA [8].m 1 A modification is modified by the m 1 A methyltransferases, or "writers", such as TRMT10C, TRMT61B, TRMT6, and TRMT61A, and removed by the demethylases, or "erasers" including ALKBH1 and ALKBH3.It is recognized by m 1 A-binding proteins YTHDF1, YTHDF2, YTHDF3, and YTHDC1, also known as "readers" [9,10].These ten regulatory genes play an essential role in the process of modifying m 1 A. However, little is known about the precise location and biogenesis of m 1 A methylation in mRNA of decidual tissue in sPE.
To reveal mechanisms in decidualization and explore potential epigenetic biomarkers in sPE, this study aimed to investigate the m 1 A methylation atlas in mRNA in decidual tissue.Two strategies were employed: (i) an m 1 A-specific and in-depth bioinformatics analysis of m 1 A in mRNA obtained from individuals with sPE and normal pregnancies (NPs) using methylated RNA immunoprecipitation sequencing (MeRIP-seq) and (ii) an analysis of m 1 A modification patterns by RNA sequencing (RNA-seq) in decidual tissue of sPE and NPs.

Participants and samples
Nine decidua samples were collected from patients with an NP (n = 3), EPE (n = 3), and LPE (n = 3) at the Department of Obstetrics and Gynecology, Ren Ji Hospital, Shanghai Jiao Tong University School of Medicine, between January 2018 and July 2019.EPE was defined as the onset of symptoms before 34 weeks of gestation, while LPE was defined as the onset of symptoms at or after 34 weeks of gestation [3].Patients with multiple pregnancies and other complications, such as maternal diabetes, thyroid dysfunction, and abnormal placental structure were excluded.Human decidua basalis was collected from the placental bed during caesarean section, separated into centrifuge tubes, and snap-frozen in liquid nitrogen for further analysis.
MeRIP-seq and RNA-seq MeRIP-Seq was performed as described in the published procedure with slight modifications [11].Total RNA was isolated using Trizol Reagent (Life Technologies) and then quantified using the NanoDrop ND-1000 instrument (Thermo Fisher Scientific, MA, USA).The RNA purity was assessed by ensuring that its OD260/OD280 value fell between 1.8 and 2.1.Next, the total RNA was depleted of rRNA using Ribo-zero (Illumina) and fragmented using an RNA fragmentation reagent (Thermo Fisher Scientific).All of the RNA fragments were subjected to immunoprecipitation using the GenSeqTM m 1 A-MeRIP Kit (GenSeq Inc., China), following the manufacturer's instructions.The input samples, which underwent no immunoprecipitation, were used to generate RNA-seq libraries.Both the input and immunoprecipitation samples were used to construct libraries using the NEBNext® Ultra II Directional RNA Library Prep Kit (New England Biolabs, Inc., USA).The libraries were then sequenced on a HiSeq platform (Illumina).

Sequencing data analysis
Paired-end reads were obtained from the Illumina HiSeq 4000 sequencer and underwent quality control using Q30.
Clean reads were then generated after trimming the 3' adaptors and removing low-quality reads using the Cutadapt software (v1.9.3).These clean reads from all libraries were aligned to the UCSC HG19 reference genome using the Hisat2 software.For the m 1 A IP-seq data, the MACS software was utilized to identify methylated sites on RNAs (peaks) [12].Differential methylated sites were determined using the DiffReps software [13].Custom scripts were developed to select the peaks that overlapped with mRNA exons [14].The mRNAs that overlapped with these differentially methylated peaks and had a fold change (FC) of at least 2.0 (Log (FC) ≥ 1.0) and a P value less than or equal to 0.0001 were classified as differentially methylated mRNAs (DMGs).For the input-seq data, raw counts were obtained using the HTSeq software (v0.9.1) and normalized using EdgeR [15].The differentially expressed mRNAs (DEGs) were defined as those with an FC of at least 2.0 (Log (FC) ≥ 1.0) and a P value less than or equal to 0.01.

Bioinformatics analysis
Consensus clustering analysis was performed based on the expression of 10 m 1 A regulators using the Consensus Cluster Plus R package.We employed the consensus cumulative distribution function (CDF) and the delta area plot to evaluate cluster stability and robustness.Based on these criteria, we found that a value of K = 3 yielded the most stable and interpretable clustering solution.Consequently, the PE group was divided into three distinct clusters.Gene set variation analysis (GSVA) was then performed in the PE and NP groups using the GSVA package in R package.The gene sets for GSVA were obtained from the MSigDB database, specifically the "C2.cp.kegg.V 7.5.1.symbols" gene sets (accessed on 8 May 2022).The m 1 A-related DEGs were defined as the union of genes selected in m 1 A modification pattern pairwise comparisons, meeting the criteria of an FC of ≥2.0 (Log(FC) ≥1.0) with adjusted P values of <0.05.UpSetR was used for data exploration and generating set visualizations [16].The DMGs, DEGs, and m 1 A-related DEGs were then subjected to functional and pathway enrichment analysis using the Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) databases.GO, the world's largest source of information on the functions of genes (www.geneontology.org), and KEGG (www.genome.jp/kegg)provide insights into the high-level functions and utilities of biological systems, respectively.

External validation of the mRNA expression of key m 1 A-related DEGs and m 1 A modulators
We selected a validation set consisting of gene expression data from 60 preeclamptic patients and 65 normal controls sourced from the Gene Expression Omnibus (GEO) dataset, specifically GSE60438.To normalize the gene expression data, we employed the "Scale" function, which facilitated subsequent comparative analyses.

Ethical statement
The study was approved by the Ren Ji Hospital Research and Ethics Committee.All participants provided informed consent before the caesarean section.

Statistical analysis
Statistical analyses and visualizations were conducted using R software (version 3.5.3,https://www.r-project.org/) and GraphPad Prism v6.0 for Mac (GraphPad; San Diego, CA, USA).The Student's t-test was employed to assess the statistical differences between the two groups.For all analyses, a two-sided P value was utilized, with a significance level set at P < 0.05.

Overview of the m 1 A methylation map in pairwise comparisons
The pairwise comparison of m 1 A peaks and m 1 A methylated mRNAs in the three groups, EPE vs NP, LPE vs NP, and EPE vs LPE, revealed unique m 1 A peaks and unique m 1 A methylated mRNAs in each comparison.In the EPE group, a total of 6767 unique m 1 A peaks and 1659 unique m 1 A methylated mRNAs were identified.Similarly, in the NP group, there were 7429 unique m 1 A peaks and 1718 unique m 1 A methylated mRNAs in the EPE vs NP comparison (Figure 1A).In the LPE group, there were 5966 unique m 1 A peaks and 1510 unique m 1 A methylated mRNAs; and 8648 unique m 1 A peaks and 2159 unique m 1 A methylated mRNAs in the NP group in the LPE vs NP comparison (Figure 1B).Further, a total of 8184 unique m 1 A peaks and 2157 unique m 1 A methylated mRNAs were obtained in the EPE group; and 6162 unique m 1 A peaks and 1567 unique m 1 A methylated mRNAs in the LPE group for the EPE vs LPE comparison (Figure 1C).Subsequently, all the differentiated methylated m 1 A peaks in each comparison were mapped to human chromosomes.They were distributed across all chromosomes, with higher abundance observed in chr1, chr17, and chr19, while no m 1 A peaks were found in chrY (Figure 1D-1F).

GO and KEGG pathway analysis of DMGs
A comparison of EPE vs NP, LPE vs NP, and EPE vs LPE resulted in 3110, 2801, and 2818 DMGs, respectively.Table 1 presents the top 20 DMGs identified in the pairwise comparisons.To determine the biological significance of m 1 A modifications in preeclamptic decidua, GO and KEGG pathway analyses were conducted for the DMGs (Figures S1-S4).The GO analysis was divided into three functional groups: biological process (BP), cellular component (CC), and molecular function (MF).
To account for the bias caused by non-matched gestation age in the PE cases and NP controls, a Venn diagram was utilized to visualize the overlaps between the pairwise comparisons [17].A total of 526 overlapping DMGs were identified as the key PE-related DMGs (Figure 2A).Additionally, KEGG pathway and GO analyses were performed to further elucidate the influence of these key PE-related DMGs. Figure 2B displays the top ten significantly enriched KEGG pathways (P < 0.01), including calcium signaling pathway, phosphatidylinositol signaling system, axon guidance, protein digestion and absorption, and others.
The enriched GO terms, arranged according to BP, CC, and MF, are shown in Figure 2C1-2C3 (P < 0.001).The enriched BPs mainly include cell morphogenesis involved in neuron differentiation, regulation of small GTPase-mediated signal transduction, and regulation of Ras protein signal transduction, among others.In terms of CC annotation classification, the key PE-related DMGs are mainly associated with adherens junctions, extracellular matrix, and actin cytoskeleton, among others.Regarding MF, the key PE-related DMGs are mainly localized in actin binding, GTPase activator activity, and signaling adaptor activity, among others.

Consensus clustering and GSVA of m 1 A regulators
Based on the expression levels of m 1 A regulators, consensus clustering analysis was performed.K = 3 was determined to be the optimal selection for dividing PE patients into three clusters: cluster A, cluster B, and cluster C (Figure 3A). Figure 3B shows the expression of 10 m 1 A methylation regulators between EPE and LPE decidua.Compared to EPE, the expression of ALKBH3 was significantly lower in LPE decidua, while the expression of YTHDF1 was significantly higher.
Subsequently, GSVA enrichment analysis of the three m 1 A modification patterns was conducted to investigate the associated pathways and biological significance.Figure 3C shows that m 1 A cluster A is significantly enriched in amino sugar  and nucleotide sugar metabolism, glycosaminoglycan degradation, and galactose metabolism, among others.m 1 A cluster B is notably enriched in sulfur metabolism, ascorbate and aldarate metabolism, and tryptophan metabolism, among others (Figure 3D).Meanwhile, m 1 A cluster C is significantly enriched in proximal tubule bicarbonate reclamation, alpha linolenic acid metabolism, limonene and pinene degradation, butanoate metabolism, fatty acid metabolism, and beta alanine metabolism, among others (Figure 3E).

Functional annotation for m 1 A-related DEGs
A total of 940 m 1 A-related DEGs were identified among three m 1 A modification patterns (Figure 4A).According to the GO analysis, the m 1 A-related DEGs are notably enriched in glycosaminoglycan binding, heparan binding, extracellular matrix binding, and others (MF); extracellular region, extracellular space, plasma membrane, and others (CC); and extracellular matrix organization, epidermis development, angiogenesis, and others (BP) (Figure 4B).Additionally, based on the KEGG analysis, the m 1 A-related DEGs are significantly enriched in the PI3K-Akt signaling pathway, MAPK signaling pathway, Rap1 signaling pathway, cAMP signaling pathway, Ras signaling pathway, and others (Figure 4C).

Conjoint analysis of DMGs and DEGs
The conjoint analysis of DMGs and DEGs in pairwise comparisons was divided into four categories based on their methylation and expression levels: hypermethylatedupregulated, hypomethylated-downregulated, hypermethylateddownregulated, and hypomethylated-upregulated.The GO analysis of these four categorized DMGs and DEGs is displayed in Table 2.
The comparison between EPE and NP revealed that 356 genes expressed in EPE decidua showed synchronization between m 1 A methylation and transcription levels, while 45 genes showed non-synchronization.The differentially hypermethylated mRNAs were significantly involved in downregulated expressed pathways such as base excision repair and Notch signaling pathway.The upregulated expressed pathways included ECM-receptor interaction, PI3K-Akt signaling pathway, focal adhesion, protein digestion and absorption, and amoebiasis.On the other hand, the differentially hypomethylated mRNAs were significantly involved in downregulated expressed pathways, such as hypertrophic cardiomyopathy, dilated cardiomyopathy, focal adhesion, vascular smooth muscle contraction, and ECM-receptor interaction.The upregulated expressed pathways included oxytocin signaling pathway, lipid and atherosclerosis, nucleocytoplasmic transport, relaxin signaling pathway, and estrogen signaling pathway (Table 3).
The comparison between LPE and NP revealed that 144 genes expressed in LPE decidua showed synchronization between m 1 A methylation and transcription levels, while 62 genes showed non-synchronization.The differentially hypermethylated mRNAs were found to be significantly involved in non-homologous end-joining (downregulated expression), autophagy-animal, Fc gamma R-mediated phagocytosis, longevity regulating pathway-multiple species, synaptic vesicle cycle, and EGFR tyrosine kinase inhibitor resistance (upregulated expression).On the other hand, the differentially hypomethylated mRNAs were significantly involved in ABC transporters, synaptic vesicle cycle, TGF-β signaling pathway, and protein digestion and absorption (downregulated expression).Additionally, focal adhesion, ECM-receptor interaction, MAPK signaling pathway, AGE-RAGE signaling pathway in diabetic complications, and amoebiasis were identified as being upregulated (Table 3).
The comparison between EPE and LPE revealed that 218 genes expressed in EPE decidua were synchronized between m 1 A methylation and transcription levels, while 50 genes were unsynchronized.The differentially hypermethylated mRNAs were significantly involved in downregulated expression of aldosterone-regulated sodium reabsorption, endocrine and other factor-regulated calcium reabsorption, and steroid hormone biosynthesis.For the upregulated expression, they were involved in ECM-receptor interaction, protein digestion and absorption, human papillomavirus infection, focal adhesion, and small cell lung cancer.Regarding the differentially hypomethylated mRNAs, they were significantly involved in downregulated expression of vascular smooth muscle contraction, oxytocin signaling pathway, axon guidance, transcriptional misregulation in cancer, and hypertrophic cardiomyopathy.For the upregulated expression, they were involved in non-homologous end-joining, arginine biosynthesis, terpenoid backbone biosynthesis, and thyroid cancer (Table 3).

Validation of external datasets for the differential expression of key m 1 A-related DEGs and m 1 A modulators
For external validation of the key m 1 A-related DEGs, we consulted the GEO dataset, GSE60438, which comprised gene expression data from 60 PE patients and 65 normal samples.We observed significant differences in the mRNA expression of COL1A2, IGF2, LAMC2, NOS3, SCIN, HSD11B2, POLM, DES, FLNB, and COL17A1 between PE decidua and NP decidua (Figure S5).These findings further underscore the reliability of our study.Additionally, we also observed significant differences in the mRNA expression of ALKBH1, ALKBH3, YTHDC1, and YTHDF1 among 10 m 1 A modulators between preeclamptic decidua and normal pregnant decidua (Figure S5).

Discussion
Accumulated evidence reveals that disorders related to m 1 A are widely associated with the pathogenesis of many diseases.
To our knowledge, this study is the first to investigate the m 1 A landscape in preeclamptic decidua using MeRIP-seq.The results of this study demonstrate distinct m 1 A modification profiles in the decidua of PE compared to NP.A total of 3110, 2801, and 2818 DMGs were identified when comparing EPE and NP, LPE and NP, and EPE and LPE, respectively.Further, a combined analysis of MeRIP-seq and RNA-seq data was employed to identify genes associated with the differential expression of methylated m 1 A peaks and differentially expressed mRNA levels.In addition, we categorized three m 1 A modification patterns based on 10 m 1 A regulators.Functional annotation analysis of key PE-related DMGs and m 1 A-related DEGs revealed their involvement in decidualization and pathways related to vascular dysfunction.These findings highlight the strong relationship between m 1 A methylation and the decidualization process in preeclamptic decidua.Until now, there has been no consensus on the etiology of PE.Recent research has focused primarily on the early stage of decidualization, which allows for implantation.Human epithelial stroma cells (ESCs) derived from nonpregnant women with a history of sPE showed a lack of differentiation, as indicated by the absence of structural changes and secretory markers.Further, conditioned medium from sPE decidual cells was unable to induce cytotrophoblast invasion [5].Decidualization is a transient phase of the endometrium that makes it receptive to embryonic signaling and promotes implantation.This process occurs before the embryo reaches the uterine cavity and is driven by ovarian hormones such as estrogen and progesterone, as well as an increase in local cAMP production.The cAMP signaling pathway, activated by progesterone, plays a crucial role in sensitizing HESCs and stimulating the transcriptional activity of the progesterone receptor [18].Decidualization in HESCs is initiated through the rapid non-genomic ERK/MAPK pathway [19].In this study, we identified DMGs and DEGs that are closely associated with essential biological pathways involved in the decidualization process.
The Notch signaling mechanism is a highly conserved developmental pathway that is essential for implantation and placentation.In primates, HESCs respond to chorionic gonadotropin and progesterone by activating the Notch receptor 1 (NOTCH1) pathway.NOTCH1 induces the expression of α-smooth muscle actin and positively regulates cytoskeletal remodeling and the initial changes typical of the decidualization process [20].Dysfunction in both the loss and gain of Notch signaling leads to impairment of endometrial function [21].The PI3K/serine/threonine protein kinase (Akt) pathway is known to affect the synthesis of decidual markers [22], and inhibition of this pathway could be involved in the decrease in motility of HESCs during decidualization [23].TGF-β is composed of three isoforms, including TGF-β1, TGF-β2, and TGF-β3.Emerging data demonstrate an important interplay between the TGFβ signaling pathway and decidualization [24].TGF-β1 plays an essential role in regulating fetal-maternal immune tolerance during pregnancy [25].Elevated levels of decidual TGF-β1 suppress the activation of specific subsets of decidual natural killer cells, which in turn contributes to the uteroplacental pathology associated with the onset of PE [26].This study revealed that the differentially hypomethylated mRNAs in LPE decidua were significantly involved in the TGF-β signaling pathway, showing downregulated expression compared to NP decidua (Table 3).However, the role of the TGF-β signaling pathway in the pathological development of PE is still highly contradictory and requires further elucidation [27].
Further, maternal vascular dysfunction is another feature of PE that plays a role in its pathogenesis and can persist into the postpartum period [28].Potential abnormalities include impaired placentation, incomplete spiral artery remodeling, and endothelial damage, which consequently lead to placental ischemia [29].When placental ischemia occurs, bioactive factors are released into the maternal circulation and cause an imbalance between antiangiogenic soluble fms-like tyrosine kinase-1 and soluble endoglin, and proangiogenic vascular endothelial growth factor (VEGF), placental growth factor (PGF), and TGF-β.These bioactive factors target vascular smooth muscle and impact vascular contraction mechanisms, leading to increased vasoconstriction in PE [30].This study also revealed that differentially hypomethylated mRNAs in the EPE decidua were significantly involved in downregulated expression of vascular smooth muscle contraction pathways compared to both LPE and NP groups (Table 3).Additionally, the uterine vascular adaptation to pregnancy is modulated by ovarian hormones, including estrogen and relaxin, which affect both the endothelium and the vascular smooth muscle [31,32].This study discovered the differentially hypomethylated mRNAs in EPE decidua were significantly involved in the upregulated expression of pathways, including relaxin signaling pathway and estrogen signaling pathway (Table 3).Relaxin, a peptide hormone secreted by the corpus luteum and circulating during pregnancy, exhibits local expression in decidua cells along with its major receptor, relaxin/insulin-like family peptide receptor 1 (RXFP1) [33].Several clinical studies have documented a significant change in the levels of estradiol and relaxin during PE [34,35].Plenty of evidence demonstrates that estrogen-induced uterine vasodilation in pregnancy is through the estrogen-VEGF-NO pathway [32,36].Also, VEGF and PGF are emerging as essential mediators in relaxin-induced vasodilation [37,38].Therefore, the disturbance of estrogen signaling pathway will decrease uterine vascular resistance and induce uterine hemodynamic changes which are associated with PE.
Additionally, we have noted that the number of unique m 1 A methylated mRNAs in NPs is consistently higher than in both EPE and LPE (Figure 1).This suggests a broader range of m 1 A modifications in NPs compared to those with PE, potentially due to altered regulatory mechanisms or stress responses in preeclamptic conditions.Some researchers have utilized m 1 A modulators for cancer detection, prognostication, clinical staging, and as targets for drug-based disease treatment [39].Moreover, in our study, we have observed significant differences in the mRNA expression of ALKBH1, ALKBH3, YTHDC1, and YTHDF1 among 10 m 1 A modulators between preeclamptic decidua and normal pregnant decidua through external dataset validation (Figure S5).Therefore, it is imperative to undertake further research and development efforts to refine m 1 A modulation patterns in decidualization, enabling their versatile application.
There are clear limitations in this study, including the relatively small sample size for a heterogeneous disease like PE.Additionally, the patients in our study are primarily of Asian ethnicity, which may limit the generalizability of our findings to other populations, such as Caucasian or Sub-Saharan African groups.Furthermore, gestational age differences could contribute to the observed modifications.Although we included pairwise comparisons between normal pregnancies and late-onset preeclamptic cases, which had more comparable gestational ages than early-onset preeclamptic cases, the differences in gestational age between normal and preeclamptic decidua could influence the results.This makes it challenging to distinguish whether the observed modifications are due to the disease state or differences in gestational age and highlights the need for further studies to control for this variable more rigorously.

Conclusion
In summary, the present study has revealed that PE decidua exhibits significantly different mRNA m 1 A modification patterns and gene expression profiles.These differences impact various signaling pathways related to decidualization and vascular dysfunction.The distinct m 1 A modification patterns observed provide valuable insight into how the process of decidualization and vascular function contribute to the development of PE.These m 1 A modification regulators could potentially be used as biomarkers or therapeutic targets for PE, necessitating further investigation.

Figure 1 .
Figure 1.The overview of the m 1 A methylation map in pairwise comparisons.(A-C) The total number of unique and common methylated m 1 A peaks in EPE vs NP (A, upper section), LPE vs NP (B, upper section), and EPE vs LPE (C, upper section) comparisons; while the lower sections show the total number of unique, common, and non-m 1 A methylated mRNAs in EPE vs NP (A), LPE vs NP (B), and EPE vs LPE (C) comparisons.(D-F) The distribution of differentially methylated m 1 A sites with significance across chromosomes for three comparisons.From up to down are EPE vs NP (D), LPE vs NP (E), and EPE vs LPE (F).EPE: Early-onset preeclampsia; LPE: Late-onset preeclampsia; NP: Normal pregnancy.

Figure 2 .
Figure 2. GO and pathway analysis of key preeclampsia-related DMGs.(A) Venn diagram depicts the overlapping regions; (B) The top ten significant pathways; (C) A bar plot was generated for the top ten GO terms with the most significant P values for biological processes (C1), cellular component (C2), and molecular function (C3).The data shown are the negative log 10 (P value) within each category.EPE: Early-onset preeclampsia; LPE: Late-onset preeclampsia; NP: Normal pregnancy; GO: Gene ontology.

Figure 3 .
Figure 3.The consensus clustering and GSVA of m 1 A regulators in preeclamptic decidua.(A) Heatmap of consensus clustering of 10 m 1 A regulators in PE decidua; (B) The expression levels of 10 m 1 A methylation regulators between EPE and LPE (* P < 0.05); (C) Heatmap of GSVA for cluster A in PE and NP decidua; (D) Heatmap of GSVA for cluster B in PE and NP decidua; (E) Heatmap of GSVA for cluster C in PE and NP decidua.GSVA: Gene set variation analysis; PE: Preeclampsia; EPE: Early-onset preeclampsia; LPE: Late-onset preeclampsia; NP: Normal pregnancy.

Figure 4 .
Figure 4.The functional annotation of m 1 A-related DEGs.(A) Identification of m 1 A-related DEGs.The functional annotation of m 1 A-related DEGs using GO enrichment analysis (B) and KEGG enrichment analysis (C).DEG: Differentially expressed mRNAs; GO: Gene ontology; KEGG: Kyoto Encyclopedia of Genes and Genomes.

:
Authors declare no conflicts of interest.Funding: This study was supported by the National Key R&D Program of China (2019YFA0802600) and the National Natural Science Foundation of China (32170863 and 31871512) to C. Zhang.Support was also received from grants from the Shanghai Commission of Science and Technology (20DZ2270900) and Open Project of Shandong Provincial Key Laboratory of Reproductive Medicine (SDKL2017018).Data availability: The data that support the findings of this study are available from the corresponding author upon reasonable request.Submitted: 27 March 2024 Accepted: 23 June 2024 Published online: 03 July 2024

Table 2 .
The top five significant GO terms of DMGs and DEGs in conjoint analysis m1A modification in mRNA of decidua in preeclampsia 10 www.biomolbiomed.com

Table 2
m1A modification in mRNA of decidua in preeclampsia 11 www.biomolbiomed.com