Unveiling transcription factor regulation and differential co-expression genes in Duchenne muscular dystrophy

Duchenne muscular dystrophy (DMD) is a severe and progressive inherited neuromuscular disease characterized by muscle fiber degeneration and central nervous system disorders that affects 1 in 3600–6000 live male births [1]. This disease is caused by mutations or dysregulation of the X-linked dystrophin gene. Absence of or defects of dystrophin protein result in disruption of the dystrophin-associated protein complex (DAPC), resulting in chronic inflammation and progressive muscle degeneration [2].

Although mutations of the DMD protein have been identified to be primarily responsible for the pathology, comprehensive understanding of the downstream mechanisms due to dystrophin absence is still lacking. Secondary changes in DMD involve calcium homeostasis [3], nitric oxide synthase [4], inflammation [5] and mast cell degranulation [6], suggesting that the pathological process of DMD is highly complicated. Gene expression profile analysis is a powerful strategy to investigate the pathophysiological mechanisms of DMD. Several gene expression profiling studies [7]-[11] have been performed earlier, providing insights into the pathology of DMD. However, most of them focused on expression level changes of individual genes, without considering the differences in gene interconnections. Since the majority of proteins function with other proteins, differential co-expression analysis (DCEA) would provide a better understanding of the mechanism underlying DMD progression. The rationale of DCEA is that changes in gene co-expression patterns between two disease statuses (case and control) provide hints regarding the disrupted regulatory relationships in patients. Therefore, integrate the information of Transcription factors (TFs) into DCEA may detect the regulatory causes of observed co-expression changes since TFs can regulate gene expression through binding the cis-elements in the target genes’ promoter regions. TFs are reported to be important in both normal and disease states [12]. Investigation of TF activities in mdx mice has proposed several TFs involved in DMD pathogenesis [13]. Thus, integrative analysis of TFs and target differential co-expression genes may provide better understanding on molecular mechanism of the secondary changes of DMD.

In this study, using microarray gene expression data collected from the Gene Expression Omnibus (GEO) database, we carried out an integrative bioinformatics analysis. Firstly, differentially expressed genes (DEGs) were detected and pathways enriched with DEGs were acquired. Secondly, differentially regulated link networks were constructed to integrate the TF-to-target information and the differential co-expression genes. Our results may provide new understanding of DMD and new targets for future therapeutic tests.

With the two downloaded datasets (GSE6011 and GSE3307), expression profile of 27 DMD patients (22 from GSE6011 and 5 from GSE3307) and 14 healthy controls were included in this study. Dystrophin protein was absent in all patients [9],[10]. A total of 454 genes were detected to be DEGs, including 284 up regulated genes and 170 down regulated ones. KEGG pathway analysis revealed that aberrantly regulated genes were enriched in 10 pathways (Table 1). Most of these pathways (7/10) are involved in the inflammatory/immune response process, such as the immune system, immune disease and infectious disease. In addition, a DMD related pathway, the viral myocarditis pathway, was included. The other two pathways are Cell adhesion molecules (CAMs) and Focal adhesion.

The IPA results of canonical pathways showed that the DEGs were enriched for epithelial adherens junction signaling, calcium signaling, nNOS signaling in skeletal muscle cells and some additional pathways (Figure 1). Consistent with the KEGG pathway analysis, aberrantly regulated genes were enriched in immune system pathways, such as calcium-induced T lymphocyte apoptosis, iCOS-iCOSL signaling in T helper cells, complement system, and antigen presentation pathway (Figure 1). Diseases and Disorders analysis (Table 2) revealed that “Neurological Disease” (P = 3.98E-15–1.20E-02) was the top disease affected by these DEGs followed by “Skeletal and Muscular Disorders” (P = 2.66E-13–1.07E-02). “Cellular Growth and Proliferation” (P = 1.38E-14–1.12E-02) was the top biological function mediated by these DEGs (Table 2).

DCEA results generated 610 pairs of DEGs regulated by at least one common TF. Among these pairs, 78 pairs DEGs were identified as differentially co-expressed. The expression values of a total of 25 pairs of DEGs were positively correlated with each other. A network was generated to illustrate the relationship between the differentially co-expressed genes and their regulatory TFs (Figure 2). Considering the primary causative effect of the aberrant expression of DMD, we construct a subnetwork with DMD and its differentially co-expressed genes (Figure 3) to illustrate genes and TFs that may play important roles in the secondary changes of DMD patients. As shown in Figure 3, among the DEGs which shared TFs with DMD, six genes were co-expressed with DMD, including ATP1A2, C1QB, MYOF, SAT1, TRIP10, and IFI6.

The pathophysiology of DMD is highly complex, involving the dysregulation of many downstream cascades. Gene expression profiling is powerful for investigating the secondary changes of DMD patients. However, previous studies mainly focused on individual gene expression changes, without considering the gene co-expression pattern or the TF information. In this study, with combined two DMD microarray datasets, we implemented DCEA and constructed a TF regulated network with the hope to provide new understanding of the pathogenesis.

Both KEGG pathway and IPA canonical pathway analysis of the DEGs revealed that pathways enriched with aberrantly regulated genes are mostly involved in the immune response processes. This may be due to the infiltration of immune cells into the muscles and the elevated levels of various inflammatory cytokines [18]-[20]. In addition, the calcium signaling and nNOS signaling in skeletal muscle cells pathway were found to be enriched with deregulated genes, which is consistent with previous findings [3],[4].

DCEA results generated 610 pairs of DEGs regulated by at least one common TF, including 78 pairs of co-expressed DEGs. As shown in Figure 2, many TFs are involved in the regulation of these DEGs. Since DMD is the causative gene, a subnetwork was constructed to illustrate important genes and TFs that may play important roles in the secondary changes of DMD patients (Figure 3). Among the DEGs which shared TFs with DMD, six genes were co-expressed with DMD, including ATP1A2, C1QB, MYOF, SAT1, TRIP10, and IFI6. Among them, ATP1A2, MYOF, and SAT1 have previously been reported to be involved in muscular dystrophy [21]-[23]. The relationship between the three left genes and DMD is still known. However, it is worth further investigation. Take IFI6 for example, this gene was highly expressed in muscle and bone and it is reported to inhibit cytochrome c release from mitochondria by regulating the Ca²⁺ channel which consequently attenuate apoptosis [24]. Alterations in intracellular Ca²⁺ homeostasis are involved in the alteration of apoptosis since a sustained increase in cytosolic Ca²⁺concentration accompanies with apoptosis in cells [25]. Previous study [26] have demonstrated the high influx of extracellular calcium through a dystrophin-deficient membrane, which may lead to subsequent muscle necrosis or apoptosis along with the inflammation response. Whether this gene contributes to the pathogenesis through its regulation of the Ca²⁺ channel needs further investigation. Among the TFs detected in the array, several TFs of DMD are dysregulated, including GATA2 and STAT5B. Further studies are needed to investigate their involvement in the disease and the mechanisms of other unmeasured TFs (Figure 3).

The ultimate cure for DMD will lie in the stable introduction of a functional dystrophin gene into the muscles of DMD patients, however, when gene therapy or transplantation of stem cells/ muscle precursor cells will be clinically available is unpredictable. In the interim, reducing secondary features of the pathologic progression of dystrophin deficiency could improve the quality and length of life for DMD patients. Dysregulation of the genes which were co-expressed with DMD may be caused by the malfunction of DMD. Therapeutic strategies aimed to compensate for the dysregulation of these genes may help to reduce the secondary features of DMD. Therefore, genes co-expressed with DMD we identified here may be considered as therapeutic targets in further investigations to treat the secondary effects. In addition, TFs that regulated these genes and DMD may also serve as potential therapeutic markers in future studies.

In summary, with combined microarray data from the GEO database, we conducted DECA and acquire the TF information of co-expressed DEGs to capture pathogenic characteristics of DMD. Our results may provide a new understanding of DMD and further contribute potential targets for future therapeutic tests.