Master Regulators of Causal Networks in Intestinal- and Diffuse-Type Gastric Cancer and the Relation to the RNA Virus Infection Pathway

Causal networks are important for understanding disease signaling alterations. To reveal the network pathways affected in the epithelial–mesenchymal transition (EMT) and cancer stem cells (CSCs), which are related to the poor prognosis of cancer, the molecular networks and gene expression in diffuse- and intestinal-type gastric cancer (GC) were analyzed. The network pathways in GC were analyzed using Ingenuity Pathway Analysis (IPA). The analysis of the probe sets in which the gene expression had significant differences between diffuse- and intestinal-type GC in RNA sequencing of the publicly available data identified 1099 causal networks in diffuse- and intestinal-type GC. Master regulators of the causal networks included lenvatinib, pyrotinib, histone deacetylase 1 (HDAC1), mir-196, and erb-b2 receptor tyrosine kinase 2 (ERBB2). The analysis of the HDAC1-interacting network identified the involvement of EMT regulation via the growth factors pathway, the coronavirus pathogenesis pathway, and vorinostat. The network had RNA–RNA interactions with microRNAs such as mir-10, mir-15, mir-17, mir-19, mir-21, mir-223, mir-25, mir-27, mir-29, and mir-34. The molecular networks revealed in the study may lead to identifying drug targets for GC.


Introduction
Various diseases have different molecular network dynamics.A number of signaling pathways and molecular networks have several activation statuses in diffuse-and intestinaltype gastric cancer (GC) [1][2][3].For instance, the epithelial-mesenchymal transition (EMT) regulation pathway alters in diffuse-and intestinal-type GC [1,2].The main features of EMT include the acquisition of anti-cancer drug resistance, recurrence, and metastasis of cancer.EMT and cancer stem cells (CSCs) share some properties in terms of anti-cancer drug resistance.The precise mechanism of how CSCs arise and the condition of the EMT's appearance are unknown.Some of the cells with drug resistance exhibit EMT features.The main population causing the drug resistance comprises CSCs that have a high expression of transporters such as ATP binding cassette (ABC) transporters [4][5][6].EMT, itself, is related to the development process induced by TGF-beta [7] and shares some properties with CSCs in terms of cancer malignancy, such as metastasis, recurrence, and drug resistance [8].
EMT and resistance toward therapeutics characterize metastatic GC [9].The molecular pathways related to EMT in metastatic and resistant GC have been identified.However, the causal networks in GC with poor prognosis are poorly understood [10][11][12].The diffusetype GC has a much poorer prognosis compared to the intestinal-type GC.A previous study revealed that a significant difference exists in the spread patterns of intestinal-and diffuse-type GC [13].The identification of the regulators of molecular networks in diffuseand intestinal-type GC is crucial for understanding drug resistance and finding therapeutic targets for diffuse-type GC [14].
In this study, we investigated the master regulators of the causal networks in diffuseand intestinal-type GC and found some interactions of molecular networks in diffuseand intestinal-type GC.The study elucidated an interesting relationship between diffuseand intestinal-type GC molecular networks.These findings may be useful for targeting the new therapeutics in metastatic GC characterized by EMT.Considering that EMT and drug resistance share some characteristics, we performed a close investigation of molecular networks in diffuse-and intestinal-type GC.

Causal Networks of Lenvatinib in Diffuse-and Intestinal-Type GC
The causal networks of lenvatinib in diffuse-type and intestinal-type GC are shown in Figure 6.The network included KIT proto-oncogene, receptor tyrosine kinase (KIT), STAT3, HRas proto-oncogene, GTPase (HRAS), and integrin subunit beta 1 (ITGB1) (Figure 6).The lenvatinib was predicted as being activated and inactivated in diffuse-type GC and intestinal-type GC, respectively (Figure 6).

Causal Networks of Lenvatinib in Diffuse-and Intestinal-Type GC
The causal networks of lenvatinib in diffuse-type and intestinal-type GC are shown in Figure 6.The network included KIT proto-oncogene, receptor tyrosine kinase (KIT), STAT3, HRas proto-oncogene, GTPase (HRAS), and integrin subunit beta 1 (ITGB1) (Figure 6).The lenvatinib was predicted as being activated and inactivated in diffuse-type GC and intestinal-type GC, respectively (Figure 6).

Discussion
Two distinct molecular subtypes, the mesenchymal and epithelial phenotypes, in GC have been identified [15].The mesenchymal-like type includes diffuse type with poor prognosis [16].The TP53-active and TP53-inactive types include patients with an intermediate prognosis, which raises an important issue that the subtypes with molecular hallmarks are risk factors of prognosis [16].
Lenvatinib, pyrotinib, HDAC1, mir-196, and ERBB2 have been identified as master regulators of causal networks in diffuse-and intestinal-type GC in the analysis.Since diffuse-type GC has relatively stable genomic features, the development of targeting therapies has been challenging [17].The combination of focal adhesion kinase inhibitor and mitogen-activated protein kinase (MAPK) kinase inhibitor was effective in inhibiting the tumor growth of human diffuse-type GC xenograft [17].Lenvatinib, a multi-kinase inhibitor to inhibit vascular endothelial growth factor receptor (VEGFR), fibroblast growth

Discussion
Two distinct molecular subtypes, the mesenchymal and epithelial phenotypes, in GC have been identified [15].The mesenchymal-like type includes diffuse type with poor prognosis [16].The TP53-active and TP53-inactive types include patients with an intermediate prognosis, which raises an important issue that the subtypes with molecular hallmarks are risk factors of prognosis [16].
Lenvatinib, pyrotinib, HDAC1, mir-196, and ERBB2 have been identified as master regulators of causal networks in diffuse-and intestinal-type GC in the analysis.Since diffuse-type GC has relatively stable genomic features, the development of targeting therapies has been challenging [17].The combination of focal adhesion kinase inhibitor and mitogen-activated protein kinase (MAPK) kinase inhibitor was effective in inhibiting the tumor growth of human diffuse-type GC xenograft [17].Lenvatinib, a multi-kinase inhibitor to inhibit vascular endothelial growth factor receptor (VEGFR), fibroblast growth factor receptor (FGFR), platelet-derived growth factor receptor (PDGFR) alpha, KIT, and rearranged during transfection (RET), showed an effect in the progression-free survival and response rate in patients with radioiodine-refractory thyroid cancer [18].Pyrotinib is an inhibitor of EGFR (ERBB1) and HER2/4 (ERBB2/4), which is approved for the treatment of breast cancer in China [19].Since both lenvatinib-and pyrotinib-oriented causal networks in diffuse-type GC were activated, it may also be targeted in the treatment of diffuse-type GC.The effects of the combination therapy with sorafenib and lenvatinib were limited in hepatocellular carcinoma.Careful examination in terms of drug resistance and effectiveness is required [20].
Histone modification is involved in drug resistance in lung cancer [21].Long noncoding RNA HRCEG, which HDAC1 regulates, inhibited cell proliferation and EMT in GC [22].Silencing of HDAC1 inhibited the proliferation of GC, which suggested the role of HDAC1 as a target for the treatment of GC [23].HDAC expression determines the sub-type of GC and is involved in tumor microenvironment characteristics and immunotherapy efficacy in GC [24].Class I HDAC inhibitor induced lipid peroxidation and ferroptosis, which inhibit tumor cell growth [25].HDAC inhibitors demonstrate an anti-cancer effect via the production of reactive oxygen species, of which dampening renders the resistance to HDAC inhibitors in cancer cells, which requires future investigation to reveal the mechanism of resistance acquisition of cancer [26,27].
The causal network of mir-196, a master regulator of the network, was activated in intestinal-type GC and inactivated in diffuse-type GC.Inhibitor of growth family member 5 (ING5), a class II tumor suppressor, is translationally targeted by miR-196, miR-196a, miR-196b-5p, miR-193a-3p, and miR-27-3p [28].ING5 promotes the autoacetylation of p53 and histone H3 and H4 to induce the transcription of Bax, growth arrest and DNA damage-inducible 45 (GADD45), p21, and p27 [28].miRNAs are involved in linking obesity and cancer [29].The investigation of miRNA-miRNA and miRNA-long non-coding RNA interaction revealed the link between PPARgamma cell signaling regulated by miR-130, miR-4663, miR-375, miR-494-3p, and miR-128-3p and MAPK cell signaling regulated by miR-143, miR-375, miR-196, and miR-128-3p [29].miR-196 is overexpressed in the intestinal epithelium of Crohn's disease patients, for which the relationship between cancer and Crohn's disease is unknown [30].miR-196 is upregulated in pancreatic cancer cells and activates the AKT signaling pathway, which is involved in the development of type 2 diabetes [31].The expression level of miR-196b was higher in pancreatic cancer cells than in cancer stroma, and the high expression of miR-196b decreased the overall survival rate, which suggested the role of miR-196b as a prognosis biomarker for pancreatic cancer [32].The relationship between miRNAs and causal networks may need to be further elucidated to reveal the malignancy.
Regulator effect network analysis in diffuse-type GC revealed the relationship between the network of infection and let-7, mir-15, mir-17, mir-34, mir-8, mir-497, miR-136-3p (miRNAs w/seed AUCAUCG), miR-3529-3p (miRNAs w/seed ACAACAA), miR-3680-3p (miRNAs w/seed UUUGCAU), and miR-7215-5p (miRNAs w/seed CUCUUUA).Let-7 plays a crucial role in the development of virus and cancer-associated virus infection [33].Since let-7 serves as a regulator of several cellular processes [33], it may be challenging to target let-7 in general to treat diseases; some specific targeting for disease-associated cells would be valuable.Let-7 regulates the self-renewal and tumorigenicity of breast cancer cells [34].Let-7 was decreased in breast tumor-initiating cells and increased with differentiation [34].Let-7 may be involved in reducing cancer cell resistance to chemotherapy by silencing the target molecules to inhibit the self-renewal [34].Let-7 has been identified as a starting point of the RNA revolution and a potential target for cancer and immune therapy [35].These various roles of let-7 are crucial for considering cancer therapeutics.
The limitation of our approach includes that HDAC1, found as one of the master regulators, is just one member of the class I HDACs in the HDAC classification, with HDAC 2 and 3 (and 8) being other members.Since vorinostat inhibits HDACs 1, 2, and 3, the roles of HDACs 2 and 3 might be important in the metastatic GC.Further advancement of studies in other members of class I HDACs is needed to highlight the role of HDAC inhibitors in drug resistance of GC [36].Panobinostat, an HDAC inhibitor, suppresses the cell proliferation, metastasis, and cell cycle progression of GC cells [37].Not only vorinostat but other possible HDAC inhibitors may also be helpful in treating GC.
In conclusion, several causal networks in diffuse-and intestinal-type GC have been identified in the study.EMT characterizes metastatic GC, and drug resistance is closely related to EMT.The master regulators of the causal network included lenvatinib, pyrotinib, HDAC1, mir-196, and ERBB2.Our study highlighted the new therapeutics targeting the master regulators of causal networks for metastatic GC characterized by EMT.In the future, further procedures, i.e., further steps including in vitro, ex vivo, and in vivo experiments and clinical studies, are needed to validate the findings of this study in the application of novel therapeutic targets.Furthermore, the approach based on molecular networks may identify possible side effects critical to new drugs.

Network Pathway Analysis
Data on intestinal-and diffuse-type GC from the TCGA cBioPortal for Cancer Genomics were uploaded and analyzed using Ingenuity Pathway Analysis (IPA) (https://digitalinsights.qiagen.com,accessed on 6 August 2024) (QIAGEN Digital Insights, Aarhus C, Denmark) [43].The 1099 causal networks in diffuse-and intestinal-type GC were generated by filtering with a cut-off z-score absolute value of 2 (As of July 2023).

Statistical Analysis
The RNA sequencing data on diffuse-and intestinal-type GC were analyzed using Student's t-test.The z-scores of intestinal-and diffuse-type GC samples were compared, and the difference was considered significant at p < 0.00001, following previous reports [1][2][3].

Figure 2 .
Figure 2. The causal network regulated by HDAC1 with depth two.(a) Regulators of the HDAC1regulated causal network in diffuse-type GC are shown with activation prediction; (b) regulators of the HDAC1-regulated causal network in intestinal-type GC are shown with activation prediction.Pale red or green color indicates upregulated or downregulated gene expression, respectively.Orange or blue color indicates predicted activation or inhibition, respectively.The intensity of colors indicates the degree of up-or downregulation.A solid or dashed line indicates direct or indirect interaction, respectively.Triangle, oval, rectangle, and diamond shapes indicate kinase, transcription regulator, ligand-dependent nuclear receptor, and enzyme, respectively.

Figure 3 .Figure 3 .
Figure 3.The analysis of the HDAC1-interacting network identified the involvement of the regulation of EMT by the growth factors pathway, coronavirus pathogenesis pathway, and vorinostat.(a) The HDAC1-interacting network in diffuse-type GC; (b) the HDAC1-interacting network in intestinal-type GC.Pale red or green color indicates upregulated or downregulated

Figure 4 .
Figure 4.The HDAC1-regulated causal network with depth three had RNA-RNA interactions with microRNAs such as mir-10, mir-15, mir-17, mir-19, mir-21, mir-223, mir-25, mir-27, mir-29, and mir-34.(a) The regulators of the HDAC1-regulated causal network with depth three in diffuse-type GC are shown; (b) the regulators of the HDAC1-regulated causal network with depth three in intestinaltype GC are shown.Pale red or green color indicates upregulated or downregulated gene expression, respectively.Orange or blue color indicates predicted activation or inhibition, respectively.The intensity of colors indicates the degree of up-or downregulation.A solid or dashed line indicates direct or indirect interaction, respectively.(Original images are available in the Supplementary Materials as Figures S3 and S4).

Figure 4 .
Figure 4.The HDAC1-regulated causal network with depth three had RNA-RNA interactions with microRNAs such as mir-10, mir-15, mir-17, mir-19, mir-21, mir-223, mir-25, mir-27, mir-29, and mir-34.(a) The regulators of the HDAC1-regulated causal network with depth three in diffuse-type GC are shown; (b) the regulators of the HDAC1-regulated causal network with depth three in intestinal-type GC are shown.Pale red or green color indicates upregulated or downregulated gene expression, respectively.Orange or blue color indicates predicted activation or inhibition, respectively.The intensity of colors indicates the degree of up-or downregulation.A solid or dashed line indicates direct or indirect interaction, respectively.(Original images are available in the Supplementary Materials as Figures S3 and S4).

Figure 5 .Figure 5 .
Figure 5. Regulator effect network of diffuse-type GC (Infection of cells, ID25).Pale red or green color indicates upregulated or downregulated gene expression, respectively.Orange or blue color indicates predicted activation or inhibition, respectively.The intensity of colors indicates the degree of up-or downregulation.A solid or dashed line indicates direct or indirect interaction, respectively.

Figure 6 .
Figure 6.Causal networks of lenvatinib in diffuse-and intestinal-type GC.(a) The causal network (depth three) of lenvatinib in diffuse-type GC; (b) the causal network (depth three) of lenvatinib in intestinal-type GC.Pale red or green color indicates upregulated or downregulated gene expression, respectively.Orange or blue color indicates predicted activation or inhibition, respectively.The intensity of colors indicates the degree of up-or downregulation.A solid or dashed line indicates direct or indirect interaction, respectively.(Original images are available in the Supplementary Materials as Figures S5 and S6).

Table 1 .
Causal networks of HDAC molecules as master regulators in diffuse-and intestinal-type gastric cancer (GC).

Table 2 .
MicroRNAs that have a direct relationship with the regulator effect network of diffuse-type gastric cancer.

Table 2 .
MicroRNAs that have a direct relationship with the regulator effect network of diffuse-type gastric cancer.