Abstract
Background/Aim: Cervical cancer (CC) is a high-risk disease in women, and advanced CC can be difficult to treat even with surgery, radiotherapy, and chemotherapy. Hence, developing more effective treatment methods is imperative. Cancer cells undergo a renewal process to escape immune surveillance and then attack the immune system. However, the underlying mechanisms remain unclear. Currently, only one immunotherapy drug has been approved by the Food and Drug Administration for CC, thus indicating the need for and importance of identifying key targets related to immunotherapy. Materials and Methods: Data on CC and normal cervical tissue samples were downloaded from the National Center for Biotechnology Information database. Transcriptome Analysis Console software was used to analyze differentially expressed genes (DEGs) in two sample groups. These DEGs were uploaded to the DAVID online analysis platform to analyze biological processes for which they were enriched. Finally, Cytoscape was used to map protein interaction and hub gene analyses. Results: A total of 165 up-regulated and 362 down-regulated genes were identified. Among them, 13 hub genes were analyzed in a protein-protein interaction network using the Cytoscape software. The genes were screened out based on the betweenness centrality value and average degree of all nodes. The hub genes were as follows: ANXA1, APOE, AR, C1QC, CALML5, CD47, CTSZ, HSP90AA1, HSP90B1, NOD2, THY1, TLR4, and VIM. We identified the following 12 microRNAs (miRNAs) that target the hub genes: hsa-miR-2110, hsa-miR-92a-2-5p, hsa-miR-520d-5p, hsa-miR-4514, hsa-miR-4692, hsa-miR-499b-5p, hsa-miR-5011-5p, hsa-miR-6847-5p, hsa-miR-8054, hsa-miR-642a-5p, hsa-miR-940, and hsa-miR-6893-5p. Conclusion: Using bioinformatics, we identified potential miRNAs that regulated the cancer-related genes and long noncoding RNAs (lncRNAs) that regulated these miRNAs. We further elucidated the mutual regulation of mRNAs, miRNAs, and lncRNAs involved in CC occurrence and development. These findings may have major applications in the treatment of CC by immunotherapy and the development of drugs against CC.
Cervical cancer (CC) occurs mainly because of human papillomavirus (HPV), a relatively small, non-enveloped virus with a diameter of 55 nm (1). Surgery and concurrent chemoradiotherapy have curative rates of 80% for early-stage (stage I-II) CC and 60% for stage III CC (2). However, surgery, radiotherapy, and chemotherapy have major side effects, and their curative rates for advanced CC are low. To date, only one immunotherapeutic drug has been approved by Food and Drug Administration for treating CC. Therefore, identifying key targets related to immunotherapy for treating CC is imperative (3).
There are two types of immunity, innate immunity and acquired immunity. Innate immunity is present since birth and acts as the body’s first line of defense by activating nonspecific immune responses against foreign particles by releasing cytokines. In contrast, acquired immunity develops over time after coming in contact with foreign particles. It is specific and time-dependent and can respond to different external stimuli (4). Nonetheless, the process of immune escape occurs, which plays a major role in cancer initiation and progression. Tumors escape the immune system by constantly renewing themselves and attacking the host’s immune system. The main factors that drive tumor immune escape are as follows: heat shock proteins inhibit apoptosis, antigen-presenting cells support tumor growth and development, changes in major histocompatibility complex (MHC) expression affect the immune response and induce T-cell dysfunction, and the tumor regulation of the B-cell response promotes immune escape and affects tumor targets of innate immunity, such as natural killer cells and tumor-associated macrophages (5).
Based on the aforementioned information, the prevention and treatment of cancer by blocking immune evasion may be an effective treatment approach. Owing to processes such as immune escape as well as a lack of effective anticancer drugs, developing new therapies that can target immunity-associated factors to prevent and treat cancer is crucial. Hence, we aimed to identify such factors and elucidate the underlying mechanisms in order to gain insights into CC prevention and treatment.
Materials and Methods
Microarray data. We downloaded sample gene expression profiles of normal cervical and CC tissues from the GEO datasets GSE131027 (GSM3760006, GSM3760015, GSM3760053, and GSM3760071) and GSE42764 (GSM1049569 and GSM1049570), which included two normal cervical and four CC tissues. These datasets were available at https://www.ncbi.nlm.nih.gov/gds/, a free online analysis platform for a microarray atlas database (6, 7).
Screening and cluster analysis of DEGs. We used the Transcriptome Analysis Console (TAC) software to screen the datasets of the normal cervical and CC tissues obtained, and the critical value criteria were set as follows: fold change (FC) |log2| >4, p<0.05, and false discovery rate <0.05 (8, 9). TAC was further used to generate a heat map, and the free online platform Bioinformatics (http://www.bioinformatics.com.cn/) was used to generate volcano maps. The Bioinformatics platform includes diverse common databases, analysis methods, and analyses of various types of sequencing data.
GO enrichment analysis of DEGs. We used the DAVID 6.8 online platform [DAVID: Functional Annotation Tools (ncifcrf.gov)] to assess DEG functions. GO analysis was performed, which was divided into three domains, biological process (BP), cellular composition, and molecular function. We mainly analyzed BP terms (10, 11).
DAVID is an online platform for analyzing the functions of high-throughput genes and the GO analysis of DEGs (12). Statistical significance was set at p=0.001. Furthermore, Sangerbox 3.0 and GraphPad Prism V. 8.0.2 were used to draw images. Sangerbox 3.0 is a free, secure, stable, and highly interactive online platform, which can be used to effectively analyze biomedical data. And we used Sangerbox 3.0 to draw bubble map. GraphPad Prism V. 8.0.2 is a scientific graphing tool that helps design scientific graphs, integrate curves, and organize data. And GraphPad Prism V.8.0.2 was used to draw bar graphs.
Construction of a protein–protein interaction (PPI) network and screening of hub genes. The corresponding DEGs obtained from the GO analysis were uploaded to STRING version 11.5 for PPI network and module analyses (13). STRING (version 11.5) contains 67.6 million proteins from 14,094 organisms and more than 20 billion interactions. STRING is an online platform that can identify interactions between known and predicted proteins and can export TSV files (14). It can be opened in Excel and Cytoscape, can be used for subsequent analysis. Our data files were then uploaded to Cytoscape v. 3.8.2 for visualization. Cytoscape is a powerful visualization tool that shows a network of interactions between all genes. The network is the core of Cytoscape; nodes represent genes, and connections between nodes represent interactions (14). The number of shortest paths passing through a node is represented by the betweenness centrality (BC) value, and the number of gene interactions is denoted by the degree (9). We decided to set a BC value >0.05 and a degree higher than the average as standards for a follow-up analysis (9).
Expression and prognosis of hub genes in CC. UALCAN (ualcan.path.uab.edu/analysis) is a comprehensive network database that can analyze the expression of hub genes in cancer and their effects on cancer survival (15). We uploaded the identified hub genes to UALCAN and analyzed differences in hub gene expression between the normal cervical and CC tissues, as well as their effect on the survival rate.
Identification of miRNAs and construction of mRNA–miRNA networks. We used the online prediction platform MIRwalk3.0 [Home - MIRwalk (Uni-Heidelberg. De] to predict miRNAs that target the identified hub genes (16). MIRwalk3.0 can accurately predict interactions between miRNA target genes. It includes three databases, MIRwalk3.0, TargetScan, and miRDB. Herein, we used MIRwalk3.0 and miRDB to simultaneously predict miRNAs, which were uploaded to Cytoscape v. 3.8.2. An mRNA–miRNA network was constructed. miRNAs that simultaneously targeted more than two genes were screened.
Identification of lncRNAs and construction of miRNA–lncRNA network. We used StarBase V2.0, a comprehensive miRNA–lncRNA prediction platform, to predict lncRNAs targeting important miRNAs (17, 18). We identified lncRNAs that simultaneously targeted two or more miRNAs and used Cytoscape v. 3.8.2. to build an miRNA–lncRNA network.
Results
Hierarchical clustering and volcano maps of DEGs. We identified 527 DEGs, including 362 down-regulated genes and 165 up-regulated genes, in the normal cervical tissues, heat and volcano maps of these DEGs were generated (Figure 1). Each dot in the volcano map represents a gene; down-regulated genes are shown in blue, whereas up-regulated genes are shown in red. The farther up the volcano, the greater the difference (Figure 1A). In the heat map, two samples on the left represent the normal cervical tissue samples, whereas four samples on the right represent the CC tissue samples (Figure 1B). A heat map can more intuitively show changes in the global expression of several samples of multiple genes and the clustering relationship of multiple samples or gene expression.
Hierarchical clustering and volcano plot of differentially expressed genes (DEGs). (A) Compared to normal cervical tissues, the red dots represent up-regulated genes, and the blue dots represent down-regulated genes. (B) GSM1049569 and GSM1049570 are normal cervical tissue samples, and GSM3760006, GSM3760015, GSM3760053, and GSM3760071 are CC tissue samples. The vertical axis is the DEG cluster, the horizontal axis is the sample cluster, the orange color represents up-regulated genes, and the blue color represents down-regulated genes.
Enrichment analysis of the GO function. The occurrence and progression of CC are closely related to immune escape; therefore, we performed GO analysis to screen genes related to immune escape in the normal cervical and CC tissues (Table I) and explored the mechanisms underlying CC occurrence and development (Figure 2A and B). The screened genes related to immune escape were uploaded to the STRING online database to construct a PPI network. The downloaded TSV file was uploaded to Cytoscape for network visualization and hub gene analysis. The root BC value (>0.05) was used as a screening threshold. However, the degree value of the hub gene was greater than the average degree value of all genes (degree=3.915). Therefore, we screened the following 13 hub genes: annexin A1 (ANXA1), apolipoprotein E (APOE), androgen receptor (AR), complement C1q C chain (C1QC), calmodulin like 5 (CALML5), CD47 molecule (CD47), cathepsin Z (CTSZ), heat shock protein 90 alpha family class A member 1 (HSP90AA1), heat shock protein 90 beta family member 1 (HSP90B1), nucleotide binding oligomerization domain containing 2 (NOD2), Thy-1 cell surface antigen (THY1), toll like receptor 4 (TLR4), and vimentin (VIM) (Figure 3 and Table II).
Immune escape associated with GO.
Gene Ontology (GO) enrichment analysis. (A) The x-axis denotes the biological process (BP), and Y axis denotes the number of genes involved in the BP. (B) The horizontal axis represents enrichment multiples, and the vertical axis represents BPs. The greater the number of different genes in a BP, the greater the diamond shape. The false discovery rate value decreases from red to blue.
Network interactions of hub genes. Immune escape from the protein-protein interaction (PPI) network. Cytoscape 3.8.2 was used to obtain the results. Darker nodes from yellow to green indicate higher betweenness centrality (BC). Nodes represent the degrees of the nodes from small to large. Larger nodes and darker colors indicate more linked genes.
Screening of hub genes.
Expression of hub genes in CC. To examine the expression profiles of the identified hub genes in CC, we uploaded them to the UALCAN online database for further analysis. The expression of VIM, HSP90AA1, APOE, and HSP90B1 was consistent with our initial prediction of differential gene expression. VIM was down-regulated (Figure 4A), whereas HSP90AA1 (Figure 4B), APOE (Figure 4C), and HSP90B1 (Figure 4D) were up-regulated in CC.
Expression of hub genes in cervical cancer (CC) tissues. Analysis of hub gene expression in CC using UALCAN (online ATCGA database for analysis and mining, based on PERL-CGI, JavaScript, and CSS built). The blue color represents normal cervical tissue, and the red color represents CC tissue. The expression of (A) VIM was down-regulated, whereas those of (B) HSP90AA1, (C) APOE, and (D) HSP90B1 were up-regulated in CC tissues.
Effect of hub genes on the prognosis and survival rate of patients with CC. UALCAN was used to analyze the effect of hub genes with the same expression as that of the initial DEGs on CC prognosis and patient survival. The results showed that the prognostic survival rates associated with VIM, HSP90AA1, and APOE were not significantly different, suggesting that they may not affect CC prognosis or survival (Figure 5A-C). When HSP90B1 was highly expressed (Figure 5D), the prognosis and survival rates of patients with CC decreased significantly.
Effect of hub genes on the prognosis and survival rate of patients with cervical cancer (CC). UALCAN was used to determine the effect of hub genes on the prognosis and survival rate of patients with CC. (A) VIM, (B) HSP90AA1, and (C) APOE did not affect these factors. However, the higher the expression of (D) HSP90B1, the lower the prognosis and survival rate of patients with CC.
Prediction of miRNAs that targeted hub genes. miRNAs can reportedly regulate mRNA expression via complementary pairing (19). Therefore, we used the MIRwalk3.0 online prediction platform to predict miRNAs that could target the 13 hub genes (Table III). The MIRwalk3.0 and miRBD databases showed the 3′ UTR as the target gene-binding region. The identified miRNAs were uploaded to Cytoscape to construct an interaction network. miRNAs that simultaneously targeted more than two hub genes (hsa-miR-520d-5p, hsa-miR-499b-5p, hsa-miR-642a-5p) were used for subsequent analyses (Figure 6).
Analysis of miRNAs targeting hub genes
Interaction between predicted miRNAs and their targeted mRNA networks. Cytoscape V. 3.8.2 software was used to visualize the relationship between the identified miRNAs and their target mRNAs. Red nodes indicate mRNAs, blue nodes indicate miRNAs targeting only one mRNA, pink nodes indicate miRNAs targeting two mRNAs, and orange nodes indicate miRNAs targeting all three mRNAs.
Prediction of lncRNAs that targeted key miRNAs. Previous studies have suggested that lncRNAs can interact with miRNAs and thereby regulate target gene expression; therefore, we used StarBase 2.0 to predict which lncRNAs targeted our previously analyzed miRNAs to regulate gene expression (hsa-miR-520d-5p, hsa-miR-499b-5p, hsa-miR-642a-5p). The results were then uploaded to Cytoscape, which revealed that one lncRNA (NEAT1) simultaneously targeted all three miRNAs (hsa-miR-520d-5p, hsa-miR-499b-5p, hsa-miR-642a-5p), whereas two lncRNAs (XIST and NORAD) simultaneously targeted two miRNAs (XIST targeted hsa-miR-520d-5p, hsa-miR-499b-5p and NORAD targeted hsa-miR-520d-5p, hsa-miR-642a-5p) (Figure 7 and Table IV).
Interaction between predicted lncRNAs and their targeted miRNA networks. Cytoscape V. 3.8.2 software was used to visualize the relationship between the identified lncRNAs and their target miRNAs. Red nodes indicate miRNAs, blue nodes indicate lncRNAs targeting only one miRNA, yellow nodes indicate lncRNAs targeting two miRNAs, and purple nodes indicate lncRNAs targeting all three miRNAs.
Analysis of lncRNAs targeting corresponding miRNAs.
Discussion
Tumors escape from the immune system and create an environment conducive to their survival and metastasis by modifying the cytotoxic resistance of an immune effector or suppressing the immune mechanism by regulating a host immunogen so that the host cannot defend against it. Furthermore, tumor immune escape can occur in many ways, such as via changes in tumor antigenicity, tumor cell death pathways, the tumor acquisition of stem cell-like phenotypes, and signal transduction pathways that promote cancer immune escape (5, 20). High-stake HPVs, such as HPV16 and HPV18, may integrate into the host genome, interfere with innate immune responses by reducing interferon production, impede STING and other key pathways, and inhibit the expression of HPV antigens via class I MHC molecules, which are crucial for eliciting the immune response (21). Treatment options such as adoptive T-cell therapy and immune checkpoint suppression are efficient against CC, and they have exhibited improved patient survival (22). Therefore, immunotherapy is necessary for CC treatment.
All GO terms selected in this study were related to immunity in order to explore whether CC can escape immunity via these BPs, thereby leading to CC occurrence and progression. Studies have shown that neutrophils are key effector cells of innate immunity, which can eliminate pathogens via degranulation inside and outside the cell. Neutrophils can promote the occurrence and development of cancer via DNA damage, angiogenesis promotion, and immunosuppression (23, 24). Some studies have suggested that the NF-
B family is involved in innate and adaptive immune responses and is a central mediator of inflammatory processes. Additionally, miRNAs affect tumorigenesis by regulating cell proliferation, metastasis, angiogenesis, and apoptosis (25, 26). A literature review shows that the original T cells remain in a quiescent state; however, when stimulated by antigens, T cells exit the quiescent state, become activated, and play a key role in adaptive immune responses. When T cells are exposed to cancer cells for a long time, they become dysfunctional and render ineffective in removing cancer cells (27-29). Some studies show that interferon-γ plays an important role in mammalian resistance to pathogens and can regulate several aspects of the immune response, such as stimulating the bactericidal activity of phagocytes, as well as presenting antigens via MHC molecules and affecting cell apoptosis. Interferon-γ can increase PD-L1 expression and facilitate IDO expression, an immunosuppressive metabolite, thus allowing cancer cells to establish a drug-resistant state (30, 31). Research has shown that toll-like receptors are a protective immune sentinel that recognize unmethylated double-stranded DNA, single-stranded RNA, and lipoproteins of pathogens, thus inducing the secretion of inflammatory cytokines and eliminating invaders. Among these, activated TLR5 exerted antitumor effects in mouse xenografted human breast cancer (32, 33).
In the innate immune system, the body’s first line of defense against viral invasion, pattern recognition receptors detect viral RNA or DNA and secrete proinflammatory factors in infected cells. Components of the innate immune response in tumors can be manipulated during tumor development, allowing cancer cell growth and metastasis, thus limiting the immune response (34, 35). After pathogen invasion, the body exhibits a protective immune response to prevent their invasion, with inflammation being one of the earliest defensive strategies adopted by the body. The inflammatory response promotes cancer clearance and is closely related to cancer cell proliferation, immune escape, and metastasis (36, 37). Interleukin-1 is a major mediator of the innate immune response, which plays a crucial role in human inflammatory diseases. In cancer, interleukin-1 promotes cancer development and metastasis (38, 39). Furthermore, interleukin-6 maintains homeostasis and elicits an immune response to defend the body against infection. It also leads to chemoresistance in gastric cancer cells via the activation of Jak1-STAT3 (40, 41). The findings confirmed that all GO terms selected in this study were related to immunity. Further, we discussed how hub genes enriched in these BPs regulated CC initiation and progression and which miRNAs and lncRNAs regulated hub gene expression.
We performed bioinformatics to investigate the hub genes related to immune evasion, including ANXA1, APOE, AR, C1QC, CALML5, CD47, CTSZ, HSP90AA1, HSP90B1, NOD2, THY1, TLR4, and VIM. We found that changes in their expression indicated the occurrence of CC and promoted immune escape. A previous literature review revealed that ANXA1 is highly expressed in triple-negative breast cancers and promotes tumor growth (42). ANXA1 is strongly expressed in the normal cervical squamous epithelium and significantly decreased with cervical tumor progression, and ANXA1 may be an effective candidate for the detection of CIN lesions and evaluation of cervical squamous cell carcinoma cell differentiation (43). ANXA1 plays a crucial role in cellular communication between the host defense and neuroendocrine systems. In both systems, after the protein has been transported from the cytoplasm to the cell surface of a neighboring cell, its action is performed extracellularly via membrane-binding receptors at neighboring sites (44). Furthermore, CTSZ is involved in diseases, such as cancer and rheumatoid arthritis, and is an important indicator of inflammation (45). Previous studies have shown that the increased level of CTSZ mRNA may be closely associated with osteoporosis. The higher the expression of CTSZ mRNA, the higher the degree of osteopenia or osteoporosis. However, the effect of the changes in CTSZ expression on osteoporosis was not reported, which may require further study (46). Its increased expression is associated with carcinogenesis, and its serum levels are higher in patients with lung cancer than in normal subjects, with a low general survival rate (47, 48). HSP90AA1, a heat shock protein, is another important cause of cancer initiation and progression. HSP90AA1 expression in surrounding tissues and levels of changes in precancerous colorectal lesions determine whether it can cause malignancy (49). HSP90AA1 plays a key role in drug resistance of human osteosarcoma. Suppression of HSP90AA1 mRNA inhibits protein expression, thereby decreasing the drug resistance of osteosarcoma cells. HSP90AA1 may have the same mechanism in CC but it is not clear yet (50). In liver metastases of colorectal cancer, VIM expression is decreased compared with that in normal liver tissue, which improves the survival of patients with high VIM expression after treatment (51). Previous studies have shown that VIM can increase its protein levels by enhancing its mRNA expression in ovarian cancer and may contribute to the development of ovarian cancer. VIM likely has the same function in CC, which is not clear and requires exploration (52). APOE can promote the growth and metastasis of lung and ovarian cancers through immunomodulation, and disordered APOE expression was observed in gastric and thyroid cancers (53, 54). Some researchers have proposed that APOE plays a major role in the development of thyroid cancer. By inhibiting the mRNA expression of APOE, its protein levels can be decreased, which can inhibit glycolysis and thereby inhibit the proliferation of thyroid cancer cells. Furthermore, APOE is responsible for the development of several cancers; therefore, APOE is highly likely to become a new target for CC treatment (55). androgen receptor (AR) can be amplified, mutated, and spliced to acquire drug resistance and stimulate prostate cancer cells to develop tumors (56, 57). In prostate cancer, Hsp27 and AR can interact, and the protein levels of Hsp27 can positively regulate AR. However, the mRNA expression of AR does not change significantly after the addition of Hsp27 inhibitors (58). HSP90B1 is involved in breast cancer cell proliferation and brain tumor metastasis, and its upregulation is strongly associated with the poor prognosis of patients with non-small cell lung cancer (59, 60). HSP90B1 plays a major role in human osteosarcoma. A previous study showed that Increased protein levels of HSP90B1 plays a pro-cancerous role in human osteosarcoma (61). Changes in C1QC expression are directly associated with the prognostic and pathological features of osteosarcoma (62). Previous studies have shown that the up-regulated mRNA expression of C1QC in melanoma leads to better prognosis in patients, and it may play a regulatory role in inhibiting tumor growth and promoting tumor apoptosis. However, the effects of increased protein levels of C1QC in patients with melanoma is not known (63). TLR4 is closely associated with colon carcinogenesis and promotes the proliferation of ovarian cancer cells. The upregulation of TLR4 can activate the TLR4/NF-
B signaling pathway to inhibit epithelial ovarian cancer (64, 65). Some researchers speculate that TLR4 can increase immunity and exhibit antitumor effects in MyD88 mice. The protein levels of TLR4 are increased by increasing its mRNA expression (66). Increase NOD2 expression in patients with metastatic cervical squamous cell carcinoma is closely associated with a low survival rate. The upregulation of NOD2 promotes proliferation and migration and activates related signaling pathways, leading to the development of cancer (67). NOD2 mRNA is highly expressed in human squamous CC. NOD2 can promote the development and metastasis of CC by activating the ERK-P65 signaling pathway. However, the effect of the protein levels of NOD2 on CC is not known, which requires further studies (67). The upregulation of CALML5 inhibits the migration of CC and is related to the occurrence of breast cancer in women (68, 69). In a previous study, immunohistochemistry was performed to detect the levels of CALML5 in patients with thymic carcinomas. It is not known whether CALML5 can increase the drug sensitivity of cisplatin and increase the killing effect of cisplatin on thymic carcinoma (70). KRT1 is associated with immune cell infiltration; the higher its expression, the lower the survival rate of patients with melanoma. KRT1 is found in the upper layer of the epidermis and on the surface of endothelial cells and plays a major role in the structural integrity of the skin. The expression of KRT1 is higher in primary melanoma than in metastatic melanoma. Most CCs originate from squamous epithelial cells; therefore, KRT1 is likely to play a significant role in CC (71). A literature review reported that CD47 is highly expressed in various cancers, with low expression in myelodysplastic syndrome and high expression in acute myeloid leukemia, indicating its relevance in the transformation of myelodysplastic syndrome to acute myeloid leukemia (72). CD47 has been studied as an anticancer agent for a long time. In the mouse CC xenotransplantation experiment, the downregulation of CD47 could lead to the inhibition tumor activity, The mRNA expression of CD47 is inhibited and its protein level is decreased at the same time. CD47 is highly expressed in almost all cancers; therefore, CD47 is also promising as a new target for CC immunotherapy (73). Increased THY1 expression is correlated with the poor prognosis of patients with gastric cancer. It is a marker of ovarian cancer stem cells and is associated with the proliferation, self-renewal, and poor prognosis of epithelial ovarian cancer (74, 75). THY1 regulates the proliferation, migration, and apoptosis of gastric cancer cells. In gastric cancer, THY1 expression increases in response to THY1 mRNA expression while promoting cell proliferation and migration. Inhibiting THY1 with inhibitors can promote the apoptosis of cancer cells and inhibit cell migration and proliferation (76).
As a type of small noncoding RNA, miRNAs can affect the expression of target genes involved in many BPs through target gene 3′ UTR binding. After the viral invasion, the expression of host miRNAs changes, leading to immune escape and creating an environment that is conducive to virus survival (77). In this study, we identified 12 miRNAs that simultaneously targeted more than two hub genes and lncRNAs that targeted the corresponding miRNAs. The key miRNAs that simultaneously targeted more than two hub genes included hsa-miR-520d-5p, hsa-miR-499b-5p, and hsa-miR-642a-5p. The miR-520 family plays a major role in the occurrence and development of many cancers. The expression of miR-520d-5p was significantly decreased in colorectal cancer compared with that in normal tissues but was activated when miR-520d-5P targeted CTHRC1 and SP1, thereby suppressing proliferation, migration, and invasion in colorectal cancer. The inhibition of miR-520d-5p showed the opposite results, suggesting that miR-520d-5p is a novel antitumor molecular target for the treatment of colorectal cancer (78). Whether miR-520d-5p affects the development and progression of CC by changing the expression of AR, CD47, and NOD2 should be studied further. In this study, we found that miR-499b-5p inhibited RWDD4, which is a key gene in bladder cancer and glioma and also inhibited the cell proliferation, migration, and invasion of glioma, which may provide new insights into the treatment of glioma (79). Our study also provides further scope for investigating whether miR-499b-5p targets AR, HSP90AA1, and TLR4 to regulate the initiation and progression of CC (80). Whether miR-642a-5p targets the regulation of CD47, HSP90B1, and TLR4 also holds potential for further developing the treatment methods of CC.
The human genome is mostly transcribed, with only a small part coding for proteins. Many transcripts contain RNA families, including long non-coding RNAs (lncRNAs), which are over 200 bases long. lncRNAs play a role in various molecular processes, including those related to cancer (81). Previous studies have shown that lncRNAs regulate miRNAs and genes (82). Therefore, lncRNAs targeting the corresponding miRNAs were identified, including NEAT1, which targeted the three miRNAs simultaneously, and XIST and NORAD, which targeted two miRNAs simultaneously. The identified lncRNAs may provide new therapeutic targets for the immunotherapy of patients with CC.
Conclusion
Our results indicate that the mutual regulation of mRNAs, miRNAs, and lncRNAs may play a role in the occurrence and development of CC. Bioinformatics analysis was performed to identify and analyze 13 genes that may regulate the development of CC, namely ANXA1, APOE, AR, C1QC, CALML5, CD47, CTSZ, HSP90AA1, HSP90B1, NOD2, THY1, TLR4, and VIM. These genes are known to play crucial roles in the development of other cancers; therefore, they are likely to play a similar role in CC. Thus, our results reveal new therapeutic targets for the immunotherapy of CC.
Acknowledgements
This work was supported by the Project of the Natural Science Foundation of Heilongjiang Province of China (LH2021C061). This work was supported by the KRIBB Research Initiative Program (KGM5242322). This research was a part of the project titled [Development and advancement of mass production process for Ishige okamurae a functional material for improving sensitive skin condition], funded by the Ministry of Oceans and Fisheries, Korea.
Footnotes
↵# These Authors contributed equally to this work.
Conflicts of Interest
The Authors declare no conflicts of interest.
Authors’ Contributions
YHH, DYM, and TK contributed to the conception of the study, writing the manuscript, and performing the literature search. SJL, YYM, DYM, SYS, MHJ, and HNS performed the data analysis. YHH, DYM, and TK performed the analysis and quality assessment of the study. All Authors read and approved the final manuscript.
Funding
This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (2020R1I1A2052417), KRIBB-RBM0112314.
- Received March 29, 2023.
- Revision received April 27, 2023.
- Accepted May 16, 2023.
- Copyright © 2023 The Author(s). Published by the International Institute of Anticancer Research.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).
