In Silico Identification of MicroRNAs Targeting Epidermal Growth Factor Receptor in the Initiation and Progression of Hepatocellular Carcinoma

INTRODUCTION

Hepatocellular carcinoma (HCC) is the most prevalent form of primary liver cancer and represents a major global health concern. It is the second leading cause of cancer-related deaths.^[1] The incidence of HCC has been steadily rising, mainly due to chronic liver conditions such as hepatitis B and C infections, alcohol abuse, and non-alcoholic fatty liver disease. These risk factors contribute to chronic inflammation, liver cirrhosis, and, ultimately, the malignant transformation of hepatocytes.^[2] Despite advances in diagnosis and treatment, the prognosis for HCC remains poor, primarily due to late-stage detection, high recurrence rates, and limited therapeutic options.^[3] The pathogenesis of HCC is a complex, multistep process involving the interplay of genetic predispositions, viral infections, and environmental factors. The common and most important process among these is the dysregulation of crucial signalling pathways that drive cellular proliferation, survival, and resistance to apoptosis.^[4,5] Among these, the epidermal growth factor receptor (EGFR) signalling pathway has been identified as a critical player in HCC progression.^[6] Overexpression of EGFR has an important role in HCC progression from the earliest phases of the illness to the final phases of hepatocarcinogenesis progression. The EGFR pathway influences both the early stage of inflammation and HCC progression.^[7] EGFR overexpression is frequently observed in HCC and is associated with increased tumour aggressiveness, poor prognosis, and resistance to standard therapies.

Recent studies have focused on understanding the molecular underpinnings of EGFR-mediated oncogenesis in HCC, including its impact on the tumour microenvironment and immune evasion. An increasingly significant area of interest is the role of microRNAs (miRNAs) in cancer biology.^[8] MiRNAs are small noncoding RNAs that post-transcriptionally modulate gene expression by binding to the 3ʹ-untranslated region (3ʹ- UTR) of target mRNAs, leading to their degradation or inhibition of translation.^[9] Dysregulation of miRNAs has been implicated in various aspects of cancer biology, including cell proliferation, apoptosis, metastasis, and immune modulation.^[10] In HCC, specific miRNAs have been shown to directly target EGFR and its downstream effectors, influencing tumour growth and response to therapy.^[11]

Given the critical role of miRNAs in modulating EGFR signalling and their potential as therapeutic targets, there is a pressing need for comprehensive studies to identify miRNAs that specifically target EGFR in HCC. In silico approaches, which involve computational tools and databases to predict miRNA–mRNA interactions, offer a powerful platform for such investigations. These studies can provide insights into the regulatory networks governing HCC pathogenesis and identify novel biomarkers for diagnosis and therapy. In this regard, the present study aims to identify miRNAs targeting EGFR in HCC using in silico tools. By integrating bioinformatics tools with existing genomic and transcriptomic data, this study seeks to identify the role of miRNAs that may serve as potential therapeutic targets or prognostic markers in HCC. Understanding these molecular interactions will advance our knowledge of HCC and pave the way for developing miRNA-based therapeutic strategies to improve patient outcomes.

MATERIAL AND METHODS

RESULTS

Differential expression of EGFR may not be significantly correlated to overall patient survival

The UALCAN analysis revealed the differential EGFR expression in LIHC by comparing 307 primary tumours with 50 normal tissue samples. EGFR expression is significantly downregulated in primary tumours of LIHC [Figure 1a], while overall survival is not significantly associated with EGFR expression in LIHC patients (p ¼ 0.54); [Figure 1b]. These results indicate that EGFR expression may not serve as a reliable indicator to determine the survival outcomes in liver carcinoma patients.

Close

Figure 1:

(a) Differential expression and (b) overall survival of epidermal growth factor receptor (EGFR) in liver hepatocellular carcinoma.

Export to PPT

Gene interaction and TF regulatory network of EGFR

The STRING database analysis identified a network of 21 genes that show experimental evidence of interaction with the EGFR, forming a complex web of biological relationships [Supplementary Table S1]. This initial network consisted of 21 nodes and 93 edges. To gain deeper insight into the topological features and biological significance of these interactions, the network was imported into Cytoscape for further analysis. The network was refined within Cytoscape, resulting in a structure of 21 nodes and 73 edges, and then examined for its modules [Figure 2a]. The analysis revealed two distinct modules within the network. The first module included nine highly interconnected genes (EGFR, STAT3, ERBB2, PTPN11, CBL, CBLB, GRB2, PTK2, and SRC), indicating functional clusters or pathways that play significant roles in cellular processes. In addition to gene interactions, regulatory mechanisms involving TFs were explored using Network Analyst. This analysis identified 31 TFs that regulate EGFR through different mechanisms [Table 1]. The regulatory interactions between these TFs and EGFR were also mapped using Cytoscape, creating a network of 31 nodes and 30 edges [Figure 2b].

Supplementary Table S1

Download

Close

Figure 2:

Protein–protein interaction network of (a) epidermal growth factor receptor (EGFR) associated genes and (b) regulating transcription factors.

Export to PPT

Table 1: List of epidermal growth factor receptor (EGFR) regulating transcription factors (TFs)

Transcription factors	Mode of interaction
AR	Activation
BCL3	Activation
BRCA1	Repression
CREBBP	Unknown
EGR1	Activation
ESR1	Unknown
HDAC1	Repression
HDAC3	Unknown
HOXB7	Activation
JUN	Activation
JUN	Activation
KLF10	Repression
LRRFIP1	Repression
MTA1	Activation
NFKB1	Activation
NR3C2	Activation
PGR	Unknown
PML	Repression
PPARG	Repression
RELA	Activation
SP1	Repression
STAT1	Unknown
STAT3	Activation
TFAP2A	Unknown
TP53	Activation
TSG101	Repression
VDR	Repression
WT1	Repression
YBX1	Activation
YY1	Activation

miRNAs targeting EGFR expression in HCC

The GSE147889 dataset, generated using the GPL21263 (3D- Gene Human miRNA V21_1.0.0) platform, contains noncoding RNA profiling from 97 HCC tumours and surrounding non-tumorous (control) tissues obtained through surgical resection. The GEO2R analysis of this dataset revealed 18 DE-miRNAs (both up- and downregulated) potentially playing key roles in the tumorigenic processes of HCC [Figure 3; Supplementary Table S2]. To explore the relationship between these DE-miRNAs and the EGFR, miRWalk was employed to predict miRNAs that could target EGFR. A total of 731 miRNAs with potential EGFR-binding sites were identified, and further validated using multiple databases, such as TargetScan, MiRDB, and miRTarBase, that confirmed 11, 44, and 6 miRNAs, respectively. The validated miRNAs, such as hsa- miR-216b-5p, hsa-miR-214–3p, hsa-miR-325, hsa-miR-199a- 3p, hsa-miR-200a-3p, and hsa-miR-199b-3p, were identified as “hub miRNAs” as they are differentially expressed in HCC and are also predicted to target EGFR.

Supplementary Table S2

Download

Close

Figure 3:

Volcano plot showing the differentially expressed microRNAs in GSE147889. padj, adjusted p value.

Export to PPT

miRNAs regulate EGFR through direct and indirect interactions

All 18 identified DE-miRNAs were analysed for their potential to target EGFR-associated genes and TFs involved in EGFR regulation using the miRWalk database. Based on this analysis, a regulatory network was constructed to map DE-miRNAs’ influence on EGFR signalling in HCC. This network consists of 52 nodes, including 15 DE-miRNAs, 14 genes, and 23 TFs, interconnected by 141 edges, providing a clear and detailed representation of how these miRNAs may affect EGFR [Figure 4].

Close

Figure 4:

Regulatory network of microRNA, transcription factors (TFs), and epidermal growth factor receptor (EGFR) associated genes. The yellow- coloured V represents miRNAs, purple-coloured triangles represent EGFR-associated genes, and the green-coloured diamond represents TFs.

Export to PPT

DISCUSSION

HCC is the most common form of liver cancer and one of the leading causes of cancer-related mortality globally. The increasing incidence of HCC is attributed mainly to chronic hepatitis infections and nonalcoholic fatty liver disease. Understanding the molecular mechanisms driving HCC is essential for improving patient outcomes and addressing this significant public health challenge.^[17] One critical player in HCC pathogenesis is EGFR, which regulates key cellular processes such as proliferation, survival, and migration. The differential expression of EGFR in HCC suggests its involvement in both tumour initiation and progression, influenced by factors such as genetic mutations, epigenetic alterations, and changes in the tumour microenvironment. Elucidating these mechanisms is vital for developing targeted therapies to inhibit EGFR signalling, potentially improving treatment outcomes in HCC.^[18,19]

Epigenetic alterations, particularly the role of miRNAs in regulating several genes in various carcinomas, have garnered significant attention in recent years. miRNAs are critical regulators in various carcinomas, acting as tumour suppressors or oncomiRs by modulating the expression of genes and their signalling pathways.^[20] The inhibition of oncomiRs, their regulators, or their target genes can offer alternative therapeutic strategies for HCC. This study identified hsa-miR 199a/b-3p, hsa-miR-200a-3p, hsa-miR 214–3p, hsa-miR 216b-5p, and hsa-miR 325 as “hub miRNAs” through an in-silico analysis. Notably, all these miRNAs, except hsa-miR-216b-5p, were found to be downregulated along with EGFR in HCC.

The role of miR-199 in inhibiting HCC proliferation and metastasis is well established, and its downregulation is correlated with tumour growth and poor prognosis. miR 199a/b-3p has been shown to suppress HCC growth by targeting p21-activated kinase 4, disrupting the Raf/MEK/ ERK pathway.^[21] Restoration of miR-199a/b-3p expression has inhibited tumour growth in nude mice models without significant toxicity, highlighting its therapeutic potential.^[22]

Additionally, miR-199 targets key oncogenes involved in HCC progression, such as mTOR and c-Met. mTOR, a critical cell growth and metabolism regulator, can block the G1–S cell cycle transition.^[23] MiR-199a-3p also targets the c-Met oncogene, which is overexpressed in HCC and is crucial for cell cycle progression and invasiveness.^[24]

Downregulation of miR-200a in HCC has also been extensively reported, indicating its significant role in tumour progression. Konno et al.^[24] found that methylated miRNAs, particularly miR-200c-3p, are elevated in gastrointestinal cancers, with the methylated form of miR-200c-3p more effectively suppressing its target genes. Moreover, previous studies have demonstrated that the miR-200 family regulates the Wnt/β-catenin signalling pathway both directly by targeting core components and indirectly by modulating ZEB1/2 contributes to the regulation of oncogenic processes in HCC.^[25] While a significant upregulation of miR–214–3p in the plasma of HCC patients has been reported, suggesting its potential as a biomarker, other studies have observed the downregulation of miR–214–3p, consistent with these in silico findings.^[26]

The concurrent downregulation of key miRNAs and their target genes in cancer suggests a complex regulatory mechanism beyond the typical miRNA-mediated repression of gene expression. One possible explanation is the shared transcriptional regulation, where the same TFs or epigenetic modifications control both miRNAs and their target genes. In cancer, these regulatory elements may become dysfunctional or silenced, leading to the simultaneous reduction in the expression of both miRNAs and their target genes.^[27] Another potential mechanism involves the interplay between feedback loops and the tumour microenvironment. In some cases, miRNAs are part of feedback loops where they regulate TFs or signalling pathways that, in turn, control their expression.^[28] Disruption of this loop could lead to the downregulation of both miRNAs and their targets. This dual downregulation may indicate that both the miRNAs and their targets are part of a broader pathway being suppressed in cancer, potentially contributing to oncogenesis through mechanisms beyond direct miRNA–target interactions.^[29–31]

Further, miRNAs often interact with TFs to regulate gene expression networks, vital for controlling gene expression across various biological processes.^[32] TFs typically bind to specific DNA sequences in promoter regions to either activate or repress the transcription of genes. On the other hand, miRNAs exert their regulatory effects post-transcriptionally by inhibiting translation and promoting mRNA degradation, acting as negative regulators of gene expression.^[33] The interaction between miRNAs and TFs creates complex regulatory networks that maintain a balance in gene expression and are integral to various biological processes and diseases. Our results revealed that miRNAs, such as miR-199a-3p and miR–214–3p, can modulate the activity of TFs that control EGFR expression and other oncogenic pathways in HCC. This interplay is crucial for maintaining cellular homeostasis, and its disruption can lead to tumorigenesis.

The in-silico network biology study provides a comprehensive framework for understanding the complex interactions between miRNAs, TFs, and their target genes. In-depth studies of such networks help identify the key regulatory nodes and potential therapeutic targets. While these in silico analyses offer valuable insights into the potential roles of miRNAs and their target genes in cancer, these findings require further experimental validation. Computational predictions are based on existing data and algorithms, which may not fully capture the complexity of biological systems. Therefore, integrating in silico predictions into experimental data is crucial for advancing and validating the understanding of potential miRNA functions in HCC and for developing effective therapies.

Although the study has not confirmed the role of identified miRNAs in regulating EGFR in the laboratory, the predicted interactive network analysis indicates the potential for developing a specific drug therapy for the treatment of HCC. Further, the findings of this study hold a significant clinical implication for developing miRNA-targeted therapies that modulate gene expression involved in tumorigenesis, metastasis, and drug resistance in HCC. miRNA replacement strategies and miRNA antagonists targeting oncomiRs present promising approaches for tumour suppression and improving drug sensitivity. These therapies can effectively regulate key signalling pathways involved in cancer progression by inhibiting oncogenic miRNAs or restoring tumour suppressor miRNAs. Additionally, miRNAs that influence chemoresistance offer the potential to enhance the efficacy of current treatments by overcoming drug resistance. As predicted in this study, the potential of miRNAs to target multiple genes and TFs offers an advantage over conventional therapies. Further studies to understand the in-depth role of these miRNAs in cancer and their integration into personalised treatment regimens could revolutionise HCC therapy, enhancing efficacy while minimising side effects.

CONCLUSION

This in silico study explored the regulation of EGFR in HCC and identified the hub miRNAs that may influence HCC progression by modulating oncogenic pathways and the tumour microenvironment. However, these miRNAs’ role in targeting EGFR has to be confirmed by in vitro and in vivo experiments for developing them as promising therapeutic targets. Overall, this study provides a valuable framework for understanding the complex interactions of miRNAs, TFs, and their target genes, and scope for the developing of miRNA-based therapies for HCC after experimental validation.