Hostname: page-component-848d4c4894-nr4z6 Total loading time: 0 Render date: 2024-06-09T10:11:37.393Z Has data issue: false hasContentIssue false
  • English
  • Français

Analysis of Long Noncoding RNAs-Related Regulatory Mechanisms in Duchenne Muscular Dystrophy Using a Disease-Related lncRNA-mRNA Pathway Network

Published online by Cambridge University Press:  01 January 2024

Bing Xu ,
Chunlei Zheng Open the ORCID record for Chunlei Zheng [Opens in a new window]  and
Kanhaiya Singh
Bing Xu
Affiliation:
Department of Pediatrics, Tongxiang First People’s Hospital, Tongxiang, Zhejiang 314500, China
Chunlei Zheng*
Affiliation:
Department of Orthopedics, Tongxiang First People’s Hospital, Tongxiang, Zhejiang 314500, China
*
Correspondence should be addressed to Chunlei Zheng; zhengchunlei20@163.com

Abstract

Objective. This study aimed to investigate the molecular regulatory mechanisms underpinning Duchenne muscular dystrophy (DMD). Methods. Using microarray data, differentially expressed long noncoding RNAs (DELs) and DMD-related differentially expressed mRNAs (DEMs) were screened based on the comparative toxicogenomics database, using a cutoff of |log2 fold change| > 1 and false discovery rate (FDR) < 0.05. Then, protein-protein interaction (PPI), coexpression network of lncRNA-mRNA, and DMD-related lncRNA-mRNA pathway networks were constructed, and functional analyses of the genes in the network were performed. Finally, the proportions of immune cells infiltrating the muscle tissues in DMD were analyzed, and the correlation between the immune cells and expression of the DELs/DEMs was studied. Results. A total of 46 DELs and 313 DMD-related DEMs were identified. The PPI network revealed STAT1, VEGFA, and CCL2 to be the top three hub genes. The DMD-related lncRNA-mRNA pathway network comprising two pathways, nine DELs, and nine DMD-related DEMs showed that PYCARD, RIPK2, and CASP1 were significantly enriched in the NOD-like receptor signaling pathway, whereas MAP2K2, LUM, RPS6, PDCD4, TWIST1, and HIF1A were significantly enriched with proteoglycans in cancers. The nine DELs in this network were DBET, MBNL1-AS1, MIR29B2CHG, CCDC18-AS1, FAM111A-DT, GAS5, LINC01290, ATP2B1-AS1, and PSMB8-AS1. Conclusion. The nine DMD-related DEMs and DELs identified in this study may play important roles in the occurrence and progression of DMD through the two pathways of the NOD-like receptor signaling pathway and proteoglycans in cancers.

Type
Research Article
Information
Genetics Research , Volume 2022 , 2022 , e6
Creative Commons
Creative Common License - CCCreative Common License - BY
This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Copyright
Copyright © 2022 Bing Xu and Chunlei Zheng.

1. Introduction

Duchenne muscular dystrophy (DMD), a severe and progressive X-linked disease that is characterized by muscular dystrophy, is caused by mutations in the dystrophin gene and has a global prevalence of approximately 0.005% of male births [Reference Nascimento Osorio, Medina Cantillo, Camacho Salas, Madruga Garrido and Vilchez Padilla1]. Deficiency of the dystrophin protein in muscles results in the progressive loss of muscle tissue and impaired muscle function, rendering the muscles more vulnerable to mechanical damage [Reference Suthar and Sankhyan2]. Complications, including skeletal muscle wasting, cardiomyopathy, and respiratory insufficiency, are often observed, with cardiopulmonary failure being the most common cause of DMD-related mortality [Reference Meyers and Townsend3]. Multidisciplinary symptomatic treatment is used in the clinical management of DMD [Reference Messina and Vita4]. Gene-based therapies and myogenic cell transplantation are the current therapeutic approaches for DMD [Reference Sun, Shen, Zhang and Xie5].

Long noncoding RNAs (lncRNAs) are noncoding transcripts longer than 200 nucleotides that are not translated into proteins [Reference Jarroux, Morillon and Pinskaya6]. LncRNAs are important modulators of messenger RNA (mRNA) and modulate mRNA expression at the posttranscriptional level by functioning as competing endogenous RNAs (ceRNAs) that competitively bind to microRNAs (miRNAs) [Reference Panni, Lovering, Porras and Orchard7]. LncRNAs regulate dystrophin expression and thus represent a potential therapeutic strategy to alleviate dystrophin protein deficiency [Reference Brusa, Magri, Bresolin, Comi and Corti8]. LncRNA H19 reportedly ameliorates muscular dystrophy by suppressing the degradation of the dystrophin protein [Reference Zhang, Li and Hu9]. In a recent study, bioinformatics analyses were used to comprehensively examine DMD-related differentially expressed lncRNAs (DELs) by constructing a ceRNA network of lncRNA-miRNA-mRNA using gene expression omnibus (GEO) datasets [Reference Xu, Hao, Xiong and He10]. However, DMD-related lncRNAs and their regulatory molecular mechanisms involving lncRNAs, mRNAs, and pathways have not been fully investigated.

In the present study, gene expression profiles of DMD samples were analyzed using a series of bioinformatics tools to identify differentially expressed mRNAs (DEMs) and DELs in DMD. Next, protein-protein interaction (PPI) networks of the DMD-related DEMs and DMD-related lncRNA-mRNA pathway were constructed. Additionally, alterations in different types of immune cells and the correlation between immune cells and crucial factors in DMD were also analyzed. Finally, the expression of the important factors (lncRNAs and mRNAs) identified in the study was verified in the training and validation sets. Thus, the study identified several potential biomarkers and therapeutic targets for DMD and has significant implications for the development of novel treatment strategies based on them.

2. Materials and Methods

2.1. Data Source and Preprocessing

The gene expression profiles from the GSE38417 and GSE6011 datasets were downloaded from the NCBI Gene Expression Omnibus (GEO, https://www.ncbi.nlm.nih.gov/geo/). The GSE38417 dataset, which was used as a training set, included muscle tissue samples from 16 patients with DMD and six normal control individuals (Affymetrix Human Genome U133 Plus 2.0 Array platform). However, the GSE6011 dataset, which served as a validation set, contained 36 muscle tissue samples (22 patients with DMD and 14 normal control individuals) and was analyzed using the GPL96 detection platform. Then, the mRNA and lncRNA expression was reannotated according to the annotation files from the Ensembl genome browser database (https://asia.ensembl.org/index.html).

2.2. Screening DEMs and DELs in DMD

Using the limma package of the R software (version 3.6.1, https://bioconductor.org/packages/release/bioc/html/limma.html), the raw expression datasets were subjected to log2 transformation and then normalized using the “between array normalization” function. Subsequently, DEMs and DELs were screened with a strict cutoff of false discovery rate (FDR) < 0.05 and |log2 fold change (FC)| > 1. Then, bidirectional hierarchical clustering analysis based on the centered Pearson correlation was performed on the identified DEMs and DELs using the pheatmap package (https://cran.r-project.org/web/packages/pheatmap/index.html) in R 3.6.1.

Using the keyword “Duchenne muscular dystrophy,” DMD-related genes were searched in the comparative toxicogenomics database (CTD) [Reference Davis, Grondin and Johnson11] (https://ctdbase.org/), with the inference score threshold set at >3. Following comparison with the identified DEMs, the overlapping DMD-related DEMs were selected for gene ontology (GO) terms [12] and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway [Reference Kanehisa, Furumichi, Tanabe, Sato and Morishima13] enrichment analyses using the online tool Database for Annotation, Visualization, and Integrated Discovery (DAVID, version 6.8, https://david.ncifcrf.gov/). An FDR of <0.05 was set as the threshold for a significant difference.

2.3. Constructing a PPI Network

The overlapping DMD-related DEMs identified using the screening were subjected to searches for interactions of the proteins they encoded using the search tool for the retrieval of interacting genes [Reference Chen, Yu, Zhao, Qiu, Xie and Tao14] (STRING, version 11.0, https://string-db.org/), and the PPI pairs with a combined score >0.6 were retained to construct a PPI network [Reference Zhu, Zhang, Gao, Hu, Wang and Hao15]. Cytoscape (version 3.9.0) was used to visualize the PPI network. Subsequently, the topological features of the network, including the average shortest path length, degree centrality (DC), closeness centrality (CC), and betweenness centrality (BC) [Reference Zhu, Zhang, Gao, Hu, Wang and Hao15], were calculated to determine the hub genes in the PPI network using the CentiScaPe plugin (version 2.2) of Cytoscape software (https://apps.cytoscape.org/apps/centiscape). Degree refers to the number of interactions (edges) at a node (protein), and node genes with the highest degrees were defined as hub genes.

The molecular complex detection (MCODE) version 1.4.2; https://apps.cytoscape.org/apps/mcode) plugin in the Cytoscape software was used to mine the functionally related and highly interconnected modules using the following parameters: degree cutoff = 2, node score cutoff = 0.2, and K-core = 2. Then, the functional analysis of the genes in these modules was performed using the BINGO plugin (version 2.44; https://apps.cytoscape.org/apps/bingo) in the Cytoscape software with FDR <0.05.

2.4. Constructing a Coexpression Network of lncRNA-mRNA

Correlations between the DELs and overlapping DMD-related DEMs were analyzed by calculating Pearson’s correlation coefficient (PCC) using the cor.test (https://stat.ethz.ch/R-manual/R-devel/library/stats/html/cor.test.html) using the R software. Then, the lncRNA-mRNA pairs with P < 0.05 and PCC > 0.9 were chosen to construct a lncRNA-mRNA coexpression network, and the network was visualized using Cytoscape software. Next, the genes in the coexpression network were subjected to GO term and KEGG pathway enrichment analyses using the DAVID online tool (version 6.8) with FDR < 0.05.

2.5. Constructing a DMD-Related lncRNA-mRNA Pathway Network

The DMD-related KEGG pathways in the CTD database were searched using the keyword “duchenne muscular dystrophy.” The DMD-related KEGG pathways thus identified were compared with the KEGG pathways that were significantly enriched by the genes in the coexpression network, and the overlapping KEGG pathways and overlapping KEGG pathway-related DEMs and DELs were retained to construct a DMD-related lncRNA mRNA pathway network.

2.6. Correlation between the Crucial Factors and DMD-Related Immune Cell Types

The infiltration of immune cells in muscle tissues is an important feature of DMD [Reference Tripodi, Villa, Molinaro, Torrente and Farini16]. The proportion of different types of immune cells infiltrating the muscle tissue of patients with DMD and normal control individuals was calculated using CIBERSORT [Reference Chen, Khodadoust, Liu, Newman and Alizadeh17] (https://cibersort.stanford.edu/index.php), and the differences between the two sets of samples were compared using t-tests. Finally, the correlation between the significantly different immune cell types and the expression of the factors (mRNAs and lncRNAs) in the DMD-related lncRNA mRNA pathway network was analyzed by calculating the PCC.

3. Results

3.1. Identification and Functional Analysis of DELs and DEMs in DMD

Following analysis, 1708 lncRNAs and 17326 mRNAs were annotated in the GSE38417 dataset. Based on the cutoffs of FDR < 0.05 and |log2 FC| > 1, 46 DELs and 2125 DEMs were screened between DMD and normal control samples (Figure 1(a)). The 46 DELs and 2125 DEMs were bidirectionally hierarchically clustered, and the results showed that these DELs and DEMs effectively differentiate DMD samples from normal control samples (Figure 1(b)). Based on the CTD database, we obtained 2992 DMD-related genes with an inference score >3. These genes were compared with the 2125 DEMs, and 313 overlapping DMD-related DEMs were identified (Figure 1(c)).

FIGURE 1: Identification of differentially expressed lncRNA (DELs) and differentially expressed mRNAs (DEMs) in Duchenne muscular dystrophy (DMD). (a) Volcano plots of DELs and DEMs in DMD based on the thresholds of false discovery rate (FDR) < 0.05 and |log2 fold change| > 1. Blue and red spots represent downregulated and upregulated DELs/DEMs, respectively. (b) Bidirectional hierarchical clustering heatmap of DELs (upper) and DEMs (lower). (c) A Venn diagram of DEMs and DMD-related genes identified in the comparative toxicogenomics database (CTD) with the inference score >3 set as the threshold.

These overlapping DMD-related DEMs were then subjected to functional analysis and found to be significantly enriched in 24 GO terms of biological processes and 14 KEGG signaling pathways (FDR < 0.05), including “positive regulation of ERK1 and ERK2,” “inflammatory response,” “response to hypoxia,” “TNF signaling pathway,” “NOD-like receptor signaling pathway,” “PI3K-Akt signaling pathway,” “p53 signaling pathway,” “FoxO signaling pathway,” “Ras signaling pathway,” and “MAPK signaling pathway” (Figures 2(a) and 2(b)).

FIGURE 2: Functional analyses of the overlapping DMD-related DEMs. (a) Significantly enriched gene ontology (GO) terms of biological processes. (b) Significantly enriched Kyoto Encyclopedia of Genes and Genomes (KEGG) signaling pathways of the DMD-related DEMs. The horizontal axis indicates the number of the DMD-related DEMs, and the vertical axis indicates the name of items. The larger the dots, the greater the number of DEMs. The darker the dot, the more significant the enrichment of process.

3.2. Analysis of the PPI Network Based on the DMD-Related DEMs

The 313 DMD-related DEMs were subjected to a search for interactions between the proteins encoded by these DMD-related DEMs using the STRING database, and 695 interaction pairs with a combined score >0.6 were identified and used to build a PPI network, which contained 216 nodes and 695 edges (Figure 3(a)). After calculating the centrality parameters of each node in the network, the top 20 genes were identified (listed in Table 1), and the top three hub genes were STAT1 (degree = 41), VEGFA (degree = 32), and CCL2 (degree = 28).

FIGURE 3: Construction of a protein-protein interaction (PPI) network based on the overlapping DMD-related DEMs. (a) A PPI network based on the DMD-related DEMs. A change in the color of the spots from blue to red indicates a change in the degree of significant difference from downregulation to upregulation. (b) Four network clusters extracted from the PPI network. Blue and red spots represent downregulated and upregulated proteins, respectively.

TABLE 1: The centrality parameters of the top20 genes in the protein-protein interaction network.

Furthermore, we extracted four network clusters using MCODE (Figure 3(b)), including clusters 1 (18 nodes and 133 edges), 2 (10 nodes and 33 edges), 3 (9 nodes and 29 edges), and 4 (5 nodes and 10 edges). Following enrichment and annotation of the biological processes, the significantly related biological functions of each cluster were obtained (Supplementary Table 1). Clusters 1, 2, 3, and 4 were significantly related to immune response, skeletal system development, defense response, and response to organic substances, respectively.

3.3. Analysis of the Coexpression Network of lncRNA mRNA

To investigate the relationship between the identified DELs and overlapping DMD-related DEMs in the pathogenesis of DMD, we calculated the PPCs of the DELs and overlapping DMD-related DEMs. Based on the thresholds of P < 0.05 and PPC > 0.9, 308 lncRNA-mRNA pairs were obtained, and a coexpression network of lncRNA-mRNA was constructed (Figure 4(a)). Next, the genes in this coexpression network were subjected to functional analysis. With the FDR set at <0.05, 21 GO terms of biological process and nine KEGG signaling pathways were found to be significantly enriched and included “positive regulation of ERK1 and ERK2,” “T cell receptor signaling pathway,” “inflammatory response,” “response to hypoxia,” “NOD-like receptor signaling pathway,” “PI3K-Akt signaling pathway,” “HIF-1 signaling pathway,” and “NF-kappa B signaling pathway” (Figure 4(b)).

FIGURE 4: Establishment of a coexpression network of lncRNA-mRNA and functional analyses of the factors in this coexpression network. (a) A coexpression network of lncRNA-mRNA was constructed with the lncRNA-mRNA pairs (P < 0.05 and PCC > 0.9). Diamonds and circles represent lncRNAs and mRNAs, respectively. A change in spot color from blue to red indicates a change in the degree of significant difference from downregulation to upregulation. (b) Significantly enriched GO terms of biological process (left) and KEGG signaling pathways (right) of the genes in this coexpression network.

3.4. Analysis of the DMD-Related lncRNA-mRNA Pathway Network Composed of Nine mRNAs, Nine lncRNAs, and Two Pathways

We searched the CTD database using the keyword “Duchenne muscular dystrophy” and identified 45 KEGG signaling pathways to be closely associated with DMD. Following comparison with KEGG pathways significantly enriched by the genes in the coexpression network, two overlapping KEGG pathways were obtained, including the NOD-like receptor signaling pathway and proteoglycans in cancers. Subsequently, the interaction between DEMs in the overlapping KEGG signaling pathways and the coexpression relationship with lncRNAs were integrated, and a DMD-related lncRNA-mRNA pathway network was proposed with two KEGG pathways, nine DMD-related DEMs (LUM, HIF1A, PYCARD, RIPK2, CASP1, PDCD4, RPS6, MAP2K2, and TWIST1), and nine DELs (ATP2B1-AS1, GAS5, MBNL1-AS1, MIR29B2CHG, PSMB8-AS1, FAM111A-DT, CCDC18-AS1, DBET, and LINC01290) (Figure 5). From the lncRNA-mRNA pathway network, PYCARD, RIPK2, and CASP1 were found to be significantly enriched in the NOD-like receptor signaling pathway, whereas MAP2K2, LUM, RPS6, PDCD4, TWIST1, and HIF1A were significantly enriched in proteoglycans in cancers.

FIGURE 5: Construction of a DMD-related lncRNA-mRNA pathway network composed of two KEGG pathways, nine DMD-related DEMs, and nine DELs. Diamonds, circles, and yellow squares represent DELs, DMD-related DEMs, and KEGG pathways, respectively.

3.5. Analysis of the Proportion of DMD-Related Immune Cell Types and Correlation between Immune Cell Types and Factors in the lncRNA-mRNA Pathway Network

Based on the expression of all the genes detected in the samples, the CIBERSORT algorithm was used to calculate the proportion of various immune cell types in each sample, and 19 immune cell types were identified. The proportions of regulatory T cells (Tregs, P = 0.0139), activated natural killer (NK) cells (P = 0.0461), and neutrophils (P = 0.0251) were significantly decreased, whereas the proportions of M2 macrophages (P = 0.02) and resting myeloid dendritic cells (P = 0.0007) were significantly increased (Figure 6(a)) in the DMD samples compared to those in the normal control samples. Furthermore, analysis of the correlation between the five significantly different immune cell types, nine DMD-related DEMs, and nine DELs in the DMD-related lncRNA-mRNA pathway network revealed that activated NK cells, neutrophils, and Tregs were positively correlated with the expression of four DELs (CCDC18-AS1, DBET, MBNL1-AS1, and MIR29B2CHG) and one DMD-related DEM (MAP2K2), whereas the cells were negatively correlated with the expression of the other five DELs and eight DMD-related DEMs. However, resting myeloid dendritic cells and M2 macrophages showed an opposite trend (Figure 6(b)).

FIGURE 6: Correlation between immune cell types and the factors in the lncRNA-mRNA pathway network. (a) Proportion of the various immune cell types in the DMD samples and normal control subjects. (b) Correlation analysis between the five significantly different immune cell types (in patients with DMD and control individuals) and the factors (DELs and DMD-related DEMs) in the lncRNA-mRNA pathway network.

3.6. Validation of the Identified Crucial Factors in the GSE38417 and GSE6011 Datasets

The expression of the crucial factors in the lncRNA-mRNA pathway network was analyzed in the training (GSE38417) and validation (GSE6011) sets. In GSE38417, the expression of lncRNA ATP2B1-AS1, FAM111A-DT, GAS5, LINC01290, and PSMB8-AS1 significantly increased (P < 0.05), whereas that of CCDC18-AS1, DBET, MBNL1-AS1, and MIR29B2CHG was significantly decreased (P < 0.05, Figure 7(a)) in the DMD samples compared to that in the normal control samples. Additionally, in GSE38417, the levels of mRNA CASP1, HIF1A, LUM, PDCD4, PYCARD, RIPK2, RPS6, and TWIST1 were significantly upregulated in DMD samples compared to those in normal controls (P < 0.05), whereas mRNA MAP2K2 was markedly downregulated in DMD samples (P < 0.05, Figure 7(b)). Because GSE6011 is based on the GPL96 annotation platform, few lncRNAs could be annotated on this platform. Therefore, we only verified the expression of nine DMD-related DEMs in the validation set (GSE6011). We found that the trend of expression of these DMD-related DEMs (except for CASP1, RIPK2, and RPS6) in GSE6011 was consistent with that in GSE38417 (Figure 7(c)).

FIGURE 7: Validation of the crucial factors (lncRNAs and mRNAs) in the training (GSE38417) and validation (GSE6011) sets. (a) Expression of the nine DELs in the DMD-related lncRNA-mRNA pathway network between DMD and normal control samples in GSE38417. (b) Expression of the nine DMD-related DEMs between DMD and normal control samples in GSE38417. (c) Expression of the nine DMD-related DEMs between DMD and normal control samples in GSE6011. *P < 0.05; **P < 0.01; ***P < 0.001.

4. Discussion

DMD is a fatal X-linked genetic disorder characterized by progressive muscular wasting resulting from dystrophin protein deficiency [Reference Landfeldt, Thompson, Sejersen, McMillan, Kirschner and Lochmuller18]. The involvement of lncRNAs in the pathophysiology of DMD is being increasingly investigated because of their pivotal roles in the regulation of dystrophin protein expression. In the present study, we identified 46 DELs and 313 DMD-related DEMs in DMD. A PPI network was constructed based on these DMD-related DEMs, and STAT1, VEGFA, and CCL2 were identified as the top three hub genes. Based on the DMD-related DEMs and DELs, a coexpression network of lncRNAs and mRNAs was generated, and the factors in this coexpression network were found to be significantly enriched in 21 GO terms of biological processes and nine KEGG signaling pathways. According to the CTD database, NOD-like receptor signaling and proteoglycans in cancer pathways were found to be closely related to DMD. The DMD-related lncRNA-mRNA pathway network was found to consist of two pathways, nine DMD-related DEMs and nine DELs. Furthermore, five immune cell types were found to be significantly different between the DMD and normal control samples. Thus, this study provides important insights into the underlying mechanisms of lncRNAs in the pathogenesis of DMD and provides potential targets that could aid in the rational design of lncRNA-based therapies for DMD.

Lack of dystrophin results in chronic inflammation, which is closely associated with muscle degeneration in the pathogenesis of DMD [Reference Hoxha19, Reference Tulangekar and Sztal20]. In our proposed DMD-related lncRNA-mRNA pathway networks, PYCARD, RIPK2, and CASP1 were significantly enriched in the NOD-like receptor signaling pathway, whereas MAP2K2, LUM, RPS6, PDCD4, TWIST1, and HIF1A were significantly enriched in proteoglycans in cancers. The NOD-like receptor (NLR) assembles a protein complex called the NLR family pyrin domain-containing 3 (NLRP3) inflammasome in response to certain infectious and sterile stimuli [Reference Wang, Yang and Xiao21]. NLPR3 inflammasomes are upregulated as a result of dystrophin deficiency and play a key pathogenic role in DMD [Reference Boursereau, Abou-Samra, Lecompte, Noel and Brichard22]. The NLR signaling pathway has a critical role in inflammation [Reference Liu, Lu and Liu23]. PYCARD encodes the inflammasome adaptor apoptosis-associated speck-like protein containing a C-terminal caspase recruitment domain (ASC), a component of the NLRP3 inflammasomes, which is involved in downstream signaling of the inflammasome pathway [Reference Ritchey, Hai, Han, Barnard and Smith24]. Receptor-interacting serine/threonine protein kinase 2 (RIPK2) regulates NLRP3 and inflammation [Reference Gao, Chen and Fan25], and the NOD/RIPK2 inflammatory signaling pathway has been found to confer susceptibility to osteoarthritis [Reference Jurynec, Gavile and Honeggar26]. CASP1, encoding caspase 1, is a key component of NLRP3 inflammasomes [Reference Liu, Zhao, Feng, Zeng and Lin27] and is positively related to active ulcerative colitis [Reference Qiu, Liu and Fang28]. Our results suggest that PYCARD, RIPK2, and CASP1 may be involved in regulating DMD-related inflammation via the NLR signaling pathway.

The dystrophin protein is a component of the dystrophin-associated glycoprotein complex that connects actin filaments, intermediate filaments, and microtubules to transmembrane protein complexes to stabilize muscle cells [Reference Carmen, Maria, Morales-Medina, Vallelunga, Palmieri and Iannitti29]. Proteoglycans, a class of highly glycosylated proteins that are expressed in almost all tissues, are involved in tissue homeostasis and remodeling of the stromal microenvironment during physiological and pathological processes, such as tissue regeneration, angiogenesis, and cancer [Reference Dituri, Gigante, Scialpi, Mancarella, Fabregat and Giannelli30]. MAP2K2 encodes mitogen-activated protein kinase 2, an important component of the MAPK pathway. The MAPK pathway is an important regulator of myofiber death [Reference Wissing, Boyer and Kwong31]. MAPK/ERK signaling plays a critical role in various biological activities by phosphorylating various substrates in the cytoplasm and nucleus and has been reported to play a central role in organ regeneration [Reference Wen, Jiao and Tan32]. Ribosomal protein S6 (RPS6) controls mRNA translation, and phospho-RPS6 has been identified as a surrogate marker of the activated PI3K/AKT/mTORC1 pathway found in many cancer types [Reference Yi, You, Park, Lee and Seong33]. Programmed cell death 4 (PDCD4) mediates inflammation and regulates the MAPK pathway [Reference Lv, Qi, Wang, Feng, Deng and Ci34]. Wang et al. [Reference Wang, Ma, Zhang, Song, He and Wang35] demonstrated that the PDCD4/HO-1 pathway is involved in oxidative stress and inflammation in atherosclerosis. Lumican (LUM), a member of the small leucine-rich proteoglycan family, is present in muscle tissues and plays dual roles as an oncogene and tumor suppressor [Reference Appunni, Anand, Khandelwal, Gupta, Rubens and Sharma36]. TWIST1 has been reported to regulate inflammatory genes in skeletal muscle [Reference Mudry, Massart, Szekeres and Krook37] and mediate epithelial-to-mesenchymal transition and cell migration [Reference Fan, Fan and Wang38]. The activation of HIF1α (encoded by HIF1A) is an important cause of vascular dysfunction and impaired angiogenesis in DMD [Reference Nguyen, Conotte, Belayew, Decleves, Legrand and Tassin39]. Additionally, interaction between HIF1A and TWIST was observed in the DMD-related lncRNA-mRNA pathway network. HIF1α silencing reportedly leads to downregulation of TWIST1 [Reference Wang, Xie and Zhang40]. Taken together, these results suggest that MAP2K2, LUM, RPS6, PDCD4, TWIST1, and HIF1A may play important roles in DMD pathophysiology by regulating the MAPK pathway and proteoglycans in cancer.

In addition, nine DELs, including DBET, MBNL1-AS1, MIR29B2CHG, CCDC18-AS1, FAM111A-DT, GAS5, LINC01290, ATP2B1-AS1, and PSMB8-AS1, were identified in the DMD-related lncRNA-mRNA pathway network. The lncRNA DBET is associated with the epigenetic etiology of facioscapulohumeral muscular dystrophy [Reference Vizoso and Esteller41]. Li et al. showed that the downregulation of lncRNA MBNL1-AS1 suppressed skeletal muscle cell apoptosis [Reference Li, Wang, Li, Zhao and Yu42]. GAS5 reportedly promotes wound healing by activating the HIF1A/VEGF pathway [Reference Peng, He and Liu43]. Dysregulation of the lncRNAs MBNL1-AS1, CCDC18-AS1, LINC01290, MIR29B2CHG, and PSMB8-AS1 has been observed in many cancers [Reference Su, Chen and Ruan44–Reference Yao, Zhao, Lv, Yang and Wang48]. NF-κB signaling is activated during DMD pathogenesis and leads to inflammation and muscle degeneration [Reference Donovan, Zimmer, Offman, Grant and Jirousek49]. A substantial body of evidence suggests that lncRNA ATP2B1-AS1 protects against inflammation by targeting the NF-κB signaling pathway [Reference Song, Zhang, Li and Ji50, Reference Wang, Tan and Zhu51], which implies that lncRNA ATP2B1-AS1 may ameliorate inflammation and muscle degeneration in the pathogenesis of DMD via NF-κB signaling. However, the role of FAM111A-DT has not been identified. Therefore, these nine DELs may play important roles in the occurrence and development of DMD through the NLR signaling pathway and proteoglycans in cancers. However, the specific mechanisms of these nine lncRNAs in DMD pathogenesis need to be investigated further.

In DMD, the immune system is activated and several types of immune cells, such as CD4+ and CD8+ T cells, Tregs, and NK cells, invade the skeletal muscles [Reference Tripodi, Villa, Molinaro, Torrente and Farini16]. A recent study showed that lncRNA CCDC18-AS1 is positively correlated with activated myeloid dendritic cells and neutrophils [Reference Fan, Huang and Chen52], in concordance with our results that showed that the expression of lncRNA CCDC18-AS1 was positively correlated with the proportion of activated NK cells, neutrophils, and Tregs but negatively correlated with resting myeloid dendritic cells and M2 macrophages. Our study also showed that the proportions of Tregs, activated NK cells, neutrophils, M2 macrophages, and resting myeloid dendritic cells were significantly different between patients with DMD and normal control individuals.

However, this study has some limitations. First, the number of samples in the GSE38417 dataset was small, and additional validation cohorts need to be analyzed in future studies. Second, as this study involved secondary analysis of microarray data, further experimental studies are needed to verify the results of our study. Additionally, the specific mechanisms of the identified DMD-related DEMs and DELs need to be further explored in vitro and in vivo.

5. Conclusions

By performing a series of bioinformatics analyses on microarray data, a DMD-related lncRNA-mRNA pathway network was constructed. Nine DELs (DBET, MBNL1-AS1, MIR29B2CHG, CCDC18-AS1, FAM111A-DT, GAS5, LINC01290, ATP2B1-AS1, and PSMB8-AS1), nine DMD-related DEMs (PYCARD, RIPK2, CASP1, MAP2K2, LUM, RPS6, PDCD4, TWIST1, and HIF1A), and two KEGG pathways (NLR signaling pathway and proteoglycans in cancer) were identified to play a crucial role in the pathogenesis of DMD. Our findings provide novel insights into the regulatory relationships between lncRNAs, mRNAs, and pathways in the pathogenesis of DMD. The DMD-related DEMs and DELs identified in this study need to be further investigated for their clinical use as potential biomarkers and therapeutic targets for DMD.

Data Availability

The datasets used and analyzed during the current study are available from the corresponding author upon request.

Conflicts of Interest

The authors declare that they have no conflicts of interest.

Supplementary Materials

Supplementary table 1: Detailed information of the four PPI network clusters.

References

Nascimento Osorio, A., Medina Cantillo, J., Camacho Salas, A., Madruga Garrido, M., and Vilchez Padilla, J., “Consensus on the diagnosis, treatment and follow-up of patients with Duchenne muscular dystrophy,” Neurologia, vol. 34, no. 7, pp. 469–481, 2019.CrossRefGoogle ScholarPubMed
Suthar, R. and Sankhyan, N., “Duchenne muscular dystrophy: a practice update,” Indian J Pediatr, vol. 85, no. 4, pp. 276–281, 2018.CrossRefGoogle ScholarPubMed
Meyers, T. A. and Townsend, D., “Cardiac pathophysiology and the future of cardiac therapies in Duchenne muscular dystrophy,” International Journal of Molecular Sciences, vol. 20, no. 17, pp. 4098–4128, 2019.CrossRefGoogle ScholarPubMed
Messina, S. and Vita, G. L., “Clinical management of Duchenne muscular dystrophy: the state of the art,” Neurological Sciences, vol. 39, no. 11, pp. 1837–1845, 2018.CrossRefGoogle ScholarPubMed
Sun, C., Shen, L., Zhang, Z., and Xie, X., “Therapeutic strategies for Duchenne muscular dystrophy: an update,” Genes, vol. 11, no. 8, pp. 837–925, 2020.CrossRefGoogle ScholarPubMed
Jarroux, J., Morillon, A., and Pinskaya, M., “History, discovery, and classification of lncRNAs,” Advances in Experimental Medicine and Biology, vol. 1008, pp. 1–46, 2017.CrossRefGoogle ScholarPubMed
Panni, S., Lovering, R. C., Porras, P., and Orchard, S., “Non-coding RNA regulatory networks,” Biochimica et Biophysica Acta (BBA) - Gene Regulatory Mechanisms, vol. 1863, no. 6, Article ID 194417, 2020.CrossRefGoogle ScholarPubMed
Brusa, R., Magri, F., Bresolin, N., Comi, G. P., and Corti, S., “Noncoding RNAs in Duchenne and Becker muscular dystrophies: role in pathogenesis and future prognostic and therapeutic perspectives,” Cellular and Molecular Life Sciences, vol. 77, no. 21, pp. 4299–4313, 2020.CrossRefGoogle Scholar
Zhang, Y., Li, Y., Hu, Q. et al., “The lncRNA H19 alleviates muscular dystrophy by stabilizing dystrophin,” Nature Cell Biology, vol. 22, no. 11, pp. 1332–1345, 2020.CrossRefGoogle ScholarPubMed
Xu, X., Hao, Y., Xiong, S., and He, Z., “Comprehensive analysis of long non-coding RNA-associated competing endogenous RNA network in Duchenne muscular dystrophy,” Interdisciplinary Sciences: Computational Life Sciences, vol. 12, no. 4, pp. 447–460, 2020.Google ScholarPubMed
Davis, A. P., Grondin, C. J., Johnson, R. J. et al., “Comparative toxicogenomics database (CTD): update 2021,” Nucleic Acids Research, vol. 49, no. D1, pp. D1138–D1143, 2021.CrossRefGoogle ScholarPubMed
Gene Ontology Consortium, “The gene ontology resource: 20 years and still going strong,” Nucleic Acids Research, vol. 47, no. D1, pp. D330–D338, 2019.CrossRefGoogle Scholar
Kanehisa, M., Furumichi, M., Tanabe, M., Sato, Y., and Morishima, K., “KEGG: new perspectives on genomes, pathways, diseases and drugs,” Nucleic Acids Research, vol. 45, no. D1, pp. D353–D361, 2017.CrossRefGoogle ScholarPubMed
Chen, Q., Yu, D., Zhao, Y., Qiu, J., Xie, Y., and Tao, M., “Screening and identification of hub genes in pancreatic cancer by integrated bioinformatics analysis,” Journal of Cellular Biochemistry, vol. 120, no. 12, Article ID 19496, 2019.CrossRefGoogle ScholarPubMed
Zhu, J., Zhang, X., Gao, W., Hu, H., Wang, X., and Hao, D., “lncRNA/circRNA-miRNA-mRNA ceRNA network in lumbar intervertebral disc degeneration,” Molecular Medicine Reports, vol. 20, no. 4, pp. 3160–3174, 2019.Google ScholarPubMed
Tripodi, L., Villa, C., Molinaro, D., Torrente, Y., and Farini, A., “The immune system in Duchenne muscular dystrophy pathogenesis,” Biomedicines, vol. 9, no. 10, pp. 1447–1517, 2021.CrossRefGoogle ScholarPubMed
Chen, B., Khodadoust, M. S., Liu, C. L., Newman, A. M., and Alizadeh, A. A., “Profiling tumor infiltrating immune cells with CIBERSORT,” Methods in Molecular Biology, vol. 1711, pp. 243–259, 2018.CrossRefGoogle ScholarPubMed
Landfeldt, E., Thompson, R., Sejersen, T., McMillan, H. J., Kirschner, J., and Lochmuller, H., “Life expectancy at birth in Duchenne muscular dystrophy: a systematic review and meta-analysis,” European Journal of Epidemiology, vol. 35, no. 7, pp. 643–653, 2020.CrossRefGoogle ScholarPubMed
Hoxha, M., “Duchenne muscular dystrophy: focus on arachidonic acid metabolites,” Biomedicine & Pharmacotherapy, vol. 110, pp. 796–802, 2019.CrossRefGoogle ScholarPubMed
Tulangekar, A. and Sztal, T. E., “Inflammation in Duchenne muscular dystrophy-exploring the role of neutrophils in muscle damage and regeneration,” Biomedicines, vol. 9, no. 10, pp. 1366–1411, 2021.CrossRefGoogle ScholarPubMed
Wang, C., Yang, T., Xiao, J. et al., “NLRP3 inflammasome activation triggers gasdermin D-independent inflammation,” Sci Immunol, vol. 6, no. 64, Article ID eabj3859, 2021.CrossRefGoogle ScholarPubMed
Boursereau, R., Abou-Samra, M., Lecompte, S., Noel, L., and Brichard, S. M., “Downregulation of the NLRP3 inflammasome by adiponectin rescues Duchenne muscular dystrophy,” BMC Biology, vol. 16, no. 1, pp. 33–17, 2018.CrossRefGoogle ScholarPubMed
Liu, P., Lu, Z., Liu, L. et al., “NOD-like receptor signaling in inflammation-associated cancers: from functions to targeted therapies,” Phytomedicine, vol. 64, Article ID 152925, 2019.CrossRefGoogle ScholarPubMed
Ritchey, B., Hai, Q., Han, J., Barnard, J., and Smith, J. D., “Genetic variant in 3′untranslated region of the mouse pycard gene regulates inflammasome activity,” Elife, vol. 10, Article ID e68203, 2021.CrossRefGoogle ScholarPubMed
Gao, C., Chen, J., Fan, F. et al., “RIPK2-Mediated autophagy and negatively regulated ROS-NLRP3 inflammasome signaling in GMCs stimulated with high glucose,” Mediators of Inflammation, vol. 2019, Article ID 6207563, 13 pages, 2019.CrossRefGoogle ScholarPubMed
Jurynec, M. J., Gavile, C. M., Honeggar, M. et al., “NOD/RIPK2 signalling pathway contributes to osteoarthritis susceptibility,” Annals of the Rheumatic Diseases, vol. 81, no. 10, pp. 1465–1473, 2022.CrossRefGoogle ScholarPubMed
Liu, J., Zhao, M., Feng, X., Zeng, Y., and Lin, D., “Expression and prognosis analyses of CASP1 in acute myeloid leukemia,” Aging (Albany NY), vol. 13, no. 10, Article ID 14088, 2021.Google ScholarPubMed
Qiu, P., Liu, L., Fang, J. et al., “Identification of pharmacological autophagy regulators of active ulcerative colitis,” Frontiers in Pharmacology, vol. 12, Article ID 769718, 2021.CrossRefGoogle ScholarPubMed
Carmen, L., Maria, V., Morales-Medina, J. C., Vallelunga, A., Palmieri, B., and Iannitti, T., “Role of proteoglycans and glycosaminoglycans in Duchenne muscular dystrophy,” Glycobiology, vol. 29, no. 2, pp. 110–123, 2019.CrossRefGoogle ScholarPubMed
Dituri, F., Gigante, G., Scialpi, R., Mancarella, S., Fabregat, I., and Giannelli, G., “Proteoglycans in cancer: friends or enemies? A special focus on hepatocellular carcinoma,” Cancers, vol. 14, no. 8, 1902 pages, 2022.CrossRefGoogle ScholarPubMed
Wissing, E. R., Boyer, J. G., Kwong, J. Q. et al., “P38α MAPK underlies muscular dystrophy and myofiber death through a Bax-dependent mechanism,” Human Molecular Genetics, vol. 23, no. 20, pp. 5452–5463, 2014.CrossRefGoogle ScholarPubMed
Wen, X., Jiao, L., and Tan, H., “MAPK/ERK pathway as a central regulator in vertebrate organ regeneration,” International Journal of Molecular Sciences, vol. 23, no. 3, 1464 pages, 2022.CrossRefGoogle ScholarPubMed
Yi, Y. W., You, K. S., Park, J. S., Lee, S. G., and Seong, Y. S., “Ribosomal protein S6: a potential therapeutic target against cancer?,” International Journal of Molecular Sciences, vol. 23, no. 1, 48 pages, 2021.CrossRefGoogle ScholarPubMed
Lv, H., Qi, Z., Wang, S., Feng, H., Deng, X., and Ci, X., “Asiatic acid exhibits anti-inflammatory and antioxidant activities against lipopolysaccharide and d-galactosamine-induced fulminant hepatic failure,” Frontiers in Immunology, vol. 8, pp. 785–812, 2017.CrossRefGoogle ScholarPubMed
Wang, X., Ma, L., Zhang, S., Song, Q., He, X., and Wang, J., “WWP2 ameliorates oxidative stress and inflammation in atherosclerotic mice through regulation of PDCD4/HO-1 pathway,” Acta Biochimica et Biophysica Sinica, vol. 54, no. 8, pp. 1057–1067, 2022.CrossRefGoogle ScholarPubMed
Appunni, S., Anand, V., Khandelwal, M., Gupta, N., Rubens, M., and Sharma, A., “Small Leucine Rich Proteoglycans (decorin, biglycan and lumican) in cancer,” Clinica Chimica Acta, vol. 491, pp. 1–7, 2019.CrossRefGoogle ScholarPubMed
Mudry, J. M., Massart, J., Szekeres, F. L. M., and Krook, A., “TWIST1 and TWIST2 regulate glycogen storage and inflammatory genes in skeletal muscle,” Journal of Endocrinology, vol. 224, no. 3, pp. 303–313, 2015.CrossRefGoogle ScholarPubMed
Fan, J., Fan, Y., Wang, X. et al., “PLCε regulates prostate cancer mitochondrial oxidative metabolism and migration via upregulation of Twist1,” Journal of Experimental & Clinical Cancer Research, vol. 38, no. 337, pp. 337–416, 2019.CrossRefGoogle ScholarPubMed
Nguyen, T. H., Conotte, S., Belayew, A., Decleves, A. E., Legrand, A., and Tassin, A., “Hypoxia and hypoxia-inducible factor signaling in muscular dystrophies: cause and consequences,” International Journal of Molecular Sciences, vol. 22, no. 13, pp. 7220–7233, 2021.CrossRefGoogle ScholarPubMed
Wang, Y. C., Xie, H., Zhang, Y. C. et al., “Exosomal miR-107 antagonizes profibrotic phenotypes of pericytes by targeting a pathway involving HIF-1α/Notch1/PDGFRβ/YAP1/Twist1 axis in vitro,” American Journal of Physiology - Heart and Circulatory Physiology, vol. 320, no. 2, pp. H520–H534, 2021.CrossRefGoogle ScholarPubMed
Vizoso, M. and Esteller, M., “The activatory long non-coding RNA DBE-T reveals the epigenetic etiology of facioscapulohumeral muscular dystrophy,” Cell Research, vol. 22, no. 10, pp. 1413–1415, 2012.CrossRefGoogle ScholarPubMed
Li, X. F., Wang, Z. Q., Li, L. Y., Zhao, G. Q., and Yu, S. N., “Retracted article: downregulation of the long noncoding RNA MBNL1-AS1 protects sevoflurane-pretreated mice against ischemia-reperfusion injury by targeting KCNMA1,” Experimental and Molecular Medicine, vol. 50, no. 9, pp. 1–16, 2018.Google Scholar
Peng, W. X., He, P. X., Liu, L. J. et al., “LncRNA GAS5 activates the HIF1A/VEGF pathway by binding to TAF15 to promote wound healing in diabetic foot ulcers,” Laboratory Investigation, vol. 101, no. 8, pp. 1071–1083, 2021.CrossRefGoogle ScholarPubMed
Su, J., Chen, D., Ruan, Y. et al., “LncRNA MBNL1-AS1 represses gastric cancer progression via the TGF-β pathway by modulating miR-424-5p/Smad7 axis,” Bioengineered, vol. 13, no. 3, pp. 6978–6995, 2022.CrossRefGoogle ScholarPubMed
Mo, X., Hu, D., Li, Y. et al., “A novel pyroptosis-related prognostic lncRNAs signature, tumor immune microenvironment and the associated regulation axes in bladder cancer,” Frontiers in Genetics, 13, Article ID 936305, 2022.CrossRefGoogle ScholarPubMed
Chen, Z., Feng, R., Kahlert, U. D. et al., “Construction of ceRNA networks associated with CD8 T cells in breast cancer,” Frontiers Oncology, vol. 12, Article ID 883197, 2022.Google ScholarPubMed
Rishehri, M., Etemadi, T., Pisheh, L., Koufigar, G., and Azadeh, M., “Quantitative expression of SFN, lncRNA CCDC18-AS1, and lncRNA LINC01343 in human breast cancer as the regulator biomarkers in a novel ceRNA network: based on bioinformatics and experimental analyses,” Genetics Research, vol. 2022, Article ID 6787791, 13 pages, 2022.CrossRefGoogle Scholar
Yao, J. M., Zhao, J. Y., Lv, F. F., Yang, X. B., and Wang, H. J., “A potential nine-lncRNAs signature identification and nomogram diagnostic model establishment for papillary thyroid cancer,” Pathology and Oncology Research, vol. 28, Article ID 1610012, 2022.CrossRefGoogle ScholarPubMed
Donovan, J. M., Zimmer, M., Offman, E., Grant, T., and Jirousek, M., “A novel NF-κB inhibitor, edasalonexent (CAT-1004), in development as a disease-modifying treatment for patients with Duchenne muscular dystrophy: phase 1 safety, pharmacokinetics, and pharmacodynamics in adult subjects,” The Journal of Clinical Pharmacology, vol. 57, no. 5, pp. 627–639, 2017.CrossRefGoogle ScholarPubMed
Song, K. Y., Zhang, X. Z., Li, F., and Ji, Q., “Silencing of ATP2B1-AS1 contributes to protection against myocardial infarction in mouse via blocking NFKBIA-mediated NF-κB signalling pathway,” Journal of Cellular and Molecular Medicine, vol. 24, no. 8, pp. 4466–4479, 2020.CrossRefGoogle ScholarPubMed
Wang, L., Tan, Y., Zhu, Z. et al., “ATP2B1-AS1 promotes cerebral ischemia/reperfusion injury through regulating the miR-330-5p/TLR4-MyD88-NF-κB signaling pathway,” Frontiers in Cell and Developmental Biology, vol. 9, Article ID 720468, 2021.Google ScholarPubMed
Fan, F., Huang, Z., and Chen, Y., “Integrated analysis of immune-related long noncoding RNAs as diagnostic biomarkers in psoriasis,” PeerJ, vol. 9, Article ID e11018, 21 pages, 2021.CrossRefGoogle ScholarPubMed
Figure 0

FIGURE 1: Identification of differentially expressed lncRNA (DELs) and differentially expressed mRNAs (DEMs) in Duchenne muscular dystrophy (DMD). (a) Volcano plots of DELs and DEMs in DMD based on the thresholds of false discovery rate (FDR) < 0.05 and |log2 fold change| > 1. Blue and red spots represent downregulated and upregulated DELs/DEMs, respectively. (b) Bidirectional hierarchical clustering heatmap of DELs (upper) and DEMs (lower). (c) A Venn diagram of DEMs and DMD-related genes identified in the comparative toxicogenomics database (CTD) with the inference score >3 set as the threshold.

Figure 1

FIGURE 2: Functional analyses of the overlapping DMD-related DEMs. (a) Significantly enriched gene ontology (GO) terms of biological processes. (b) Significantly enriched Kyoto Encyclopedia of Genes and Genomes (KEGG) signaling pathways of the DMD-related DEMs. The horizontal axis indicates the number of the DMD-related DEMs, and the vertical axis indicates the name of items. The larger the dots, the greater the number of DEMs. The darker the dot, the more significant the enrichment of process.

Figure 2

FIGURE 3: Construction of a protein-protein interaction (PPI) network based on the overlapping DMD-related DEMs. (a) A PPI network based on the DMD-related DEMs. A change in the color of the spots from blue to red indicates a change in the degree of significant difference from downregulation to upregulation. (b) Four network clusters extracted from the PPI network. Blue and red spots represent downregulated and upregulated proteins, respectively.

Figure 3

TABLE 1: The centrality parameters of the top20 genes in the protein-protein interaction network.

Figure 4

FIGURE 4: Establishment of a coexpression network of lncRNA-mRNA and functional analyses of the factors in this coexpression network. (a) A coexpression network of lncRNA-mRNA was constructed with the lncRNA-mRNA pairs (P < 0.05 and PCC > 0.9). Diamonds and circles represent lncRNAs and mRNAs, respectively. A change in spot color from blue to red indicates a change in the degree of significant difference from downregulation to upregulation. (b) Significantly enriched GO terms of biological process (left) and KEGG signaling pathways (right) of the genes in this coexpression network.

Figure 5

FIGURE 5: Construction of a DMD-related lncRNA-mRNA pathway network composed of two KEGG pathways, nine DMD-related DEMs, and nine DELs. Diamonds, circles, and yellow squares represent DELs, DMD-related DEMs, and KEGG pathways, respectively.

Figure 6

FIGURE 6: Correlation between immune cell types and the factors in the lncRNA-mRNA pathway network. (a) Proportion of the various immune cell types in the DMD samples and normal control subjects. (b) Correlation analysis between the five significantly different immune cell types (in patients with DMD and control individuals) and the factors (DELs and DMD-related DEMs) in the lncRNA-mRNA pathway network.

Figure 7

FIGURE 7: Validation of the crucial factors (lncRNAs and mRNAs) in the training (GSE38417) and validation (GSE6011) sets. (a) Expression of the nine DELs in the DMD-related lncRNA-mRNA pathway network between DMD and normal control samples in GSE38417. (b) Expression of the nine DMD-related DEMs between DMD and normal control samples in GSE38417. (c) Expression of the nine DMD-related DEMs between DMD and normal control samples in GSE6011. *P < 0.05; **P < 0.01; ***P < 0.001.

Supplementary material: File

Xu and Zheng supplementary material
Download undefined(File)
File 21.7 KB