Coronary atherosclerotic heart disease, or coronary artery disease (CAD), is the most common form of cardiovascular disease. CAD is the most common cause of human death, both domestically and abroad (1). To date, the gold standard for the diagnosis of CAD is the performance of coronary angiography, which has the highest accuracy. Still, this means of confirming the diagnosis is invasive and can only be done by cardiologists who have specialized studies in the cardiac intervention (2). Moreover, pharmacological treatment is essential throughout, in addition to medical intervention and surgical bypass surgery (3). In addition, cancer has a very high morbidity and mortality rate, and recent studies have found that cancer patients often have a combination of heart disease, especially CAD (4). However, little has been reported about the relationship between CAD and cancer. Therefore, the exploration of potential biomarkers associated with CAD and the exploration of possible mechanisms and common therapeutic targets between CAD and cancer deserve further investigation.

Atherosclerosis is the most basic pathological mechanism of CAD (5, 6). The immune response plays an integral role in the formation as well as in the development of atherosclerosis (7, 8). Among them, T lymphocytes are the primary immune cell type in atherosclerotic plaques, with a predominance of cytotoxic CD8+ T cells (9, 10). Animal experiments have shown that cytotoxic CD8+ T cells promote the formation of necrotic cores and exacerbate the progression of unfavorable plaques (11). It has been found that antibody-mediated depletion of CD8+ T cells ameliorates atherosclerosis (11). Also, CD8+ T cells can induce apoptosis and inflammatory responses by secreting cytokines such as tumor necrosis factor-α and IFN-γ (12, 13). In addition, CD8+ T cell function is tightly controlled by immune checkpoint proteins. Regulators of immune checkpoint proteins, such as the E3-ligase Casitas B-cell lymphoma B (CBL-B), control the progression of human atheromatous plaques mainly by limiting the activation CD8+ T cells (14). In contrast, co-suppressors that inhibit immune checkpoints, such as cytotoxic T-lymphocyte-associated protein 4 (CTLA-4), programmed cell death protein 1 (PD-1), and programmed cell death one ligand 1 (PD-L1), can increase cytotoxic CD8+ T cells and thus aggravate atherosclerotic plaques (15). Tumor immunotherapy most notably enhances the activity of CD8+ T cells to obtain a sustained and effective anticancer response (16). In the tumor microenvironment (TME), CTLA-4 inhibits the initiation of initial T cells, while PD-1 and PD-L1 accelerate effector T cell depletion (17). Blocking CTLA-4 or PD-1 enhances CD8+ T cell-mediated immune responses (18). In addition, immunomodulatory antibodies or cancer vaccines can restart CD8+ T-cell activity (19). In conclusion, immunotherapy has been widely used in the clinical treatment of cancer but is limited to laboratory studies on CAD.

Bioinformatics analysis methods have been utilized to conduct relevant explorations to gain insight into the underlying pathophysiological mechanisms of diseases. Among them, weighted gene co-expression network analysis (WGCNA) is a widely used bioinformatics analysis tool that can mine the gene expression modules most relevant to the immune response to disease, from which hub genes can be further screened (20). On the other hand, CIBERSORT is another bioinformatics analysis method that can identify the type of immune cell composition (21). In recent years, the combined application of WGCNA and CIBERSORT to explore potential biomarkers has been commonly used in oncology. However, few studies have used it to investigate potential biomarkers associated with CAD immune responses, especially to explore biomarkers related to CD8+ T cells.

We used the Gene Expression Omnibus (GEO) database to download the gene expression dataset GSE12288 from patients with CAD and first analyzed the enrichment of all sample genes in this dataset using GSVA. This dataset was then analyzed using WGCNA to identify the modules most associated with CD8+ T cell infiltration levels and candidate hub genes. Then, GO, and KEGG analyses were performed on this module to confirm the association of CAD with CD8+ T cells. In addition, we further explored the relevance of these candidate hub genes to CD8+ T cells using other datasets, GSE34198 and GSE66360, thus identifying three hub genes in CAD with CD8+ T cells. Finally, we also explored their role in different cancer types. This is the first time we have used WGCNA combined with immune infiltration analysis to identify potential molecular markers associated with CD8+ T cells in CAD while further exploring their possible immune mechanisms in CAD and different cancers. This discovery may provide a valid theoretical basis for the future development of novel immunotherapies for CAD and cancer and provide new insights into the possible link between CAD and cancer.

Despite the current high level of medical care available, CAD and cancer remain important causes of human mortality worldwide. A growing number of studies show that immune responses are essential in developing both CAD and cancer. It is well known that immunotherapy is an integral part of anti-cancer treatment. Several common types of immunotherapies, such as immune checkpoint blockade (ICB), chimeric antigen receptor (CAR)-T cell therapy, cytokine therapy, immunomodulatory antibodies, monoclonal antibodies, and cancer vaccines, have achieved better anti-cancer responses (32, 33). In contrast, no immunotherapies have been identified that can be applied to the clinical treatment of CAD. In recent years, findings exploring immunotherapeutic approaches for CAD, such as the Canakinumab Anti-Inflammatory Thrombosis Outcome Study (CANTOS) and the Cardiovascular Inflammation Reduction Trial (CIRT), have shown that only blocking specific signaling pathways associated with atherogenesis can be genuinely therapeutic (34, 35). This finding has further stimulated interest in exploring potential target biomarkers for immune correlates of CAD. In addition, patients with cancer combined with CAD may have accelerated CAD progression due to the pro-inflammatory response and hypercoagulability brought about by cancer itself (36). Moreover, anticancer therapy in cancer patients may also increase the risk of developing CAD due to its cardiotoxic effects (37). However, there is little evidence regarding the exact mechanism of interaction between CAD and cancer to date.

This study used WGCNA in the GSE12288 dataset to find the most relevant modules to CD8+ T cells for the following analysis. FBXO7, RAD23A, and MKRN1 were verified and identified as hub genes by using GSE34198 and GSE66360 datasets. Their expression levels were positively correlated with CD8+ T-cell infiltration in CAD and multiple cancers. Among them, FBXO7, as a member of the F-box protein family, can be involved in the pathological process of Parkinson’s disease (PD) development through various mechanisms such as indirect ubiquitination and disruption of autophagic function (38, 39). FBXO7 is a component of the SCF (SKP1/cullin-1/F-box protein)-E3 type ubiquitin ligase complex, which leads to the production of the corresponding substrate ubiquitination (40). Previous correlations have reported that FBXO7 is expressed at high levels in cancer and is a potential oncogene (41). It is associated with immune escape and immunotherapy resistance in cancer patients (42). RAD23A, also known as RAD23 and HHR23A, acts as a junction protein during ubiquitin-mediated proteasomal degradation, increasing the degradation of target proteins and enhancing the response of CD8+ T cells (43). Human tumor virus E6-associated protein (E6AP) acts as an E3 ligase in ubiquitination. HHR23A participates in the ubiquitination process by binding to E6AP (44). rAD23A increases cancer chemoresistance and the chance of recurrence by upregulating autophagy (45). Makorin Ringinger Protein 1 (MKRN1) is an E3-type ubiquitin ligase. It has been shown that MKRN1 promotes p21 protein ubiquitination and proteasome pathway degradation, thereby preventing intermittent hypoxia (IH)-induced myocardial apoptosis (46). MKRN1 can disrupt p53 and thus apoptosis through ubiquitination (47). P53 is known to regulate autophagy, according to previous studies (48). Of course, further experiments are needed to verify whether there is a link between MKRN1, p53, and autophagy. MKRN1 positively regulates biological processes mediated by Wnt/β-linked proteins and plays an essential role in the proliferation, migration, and invasion of cancer cells (49). In addition, it has been reported that MKRN1 may be involved in regulating the development of gastric carcinogenesis (50). depletion of MKRN1 may inhibit tumor growth (51). It has been previously found that ubiquitination is a physiological process by which ubiquitin binds to target proteins to cause a series of reactions in substrates. It plays a crucial role in regulating immune responses (52). Moreover, ubiquitination plays an essential role in the pathogenesis of CAD, and its expression level increases with the severity of CAD (52, 53). On the other hand, autophagy is a conserved evolutionary process that provides energy and biological components by digesting its dysfunctional organelles (54, 55). Early correlations reported that autophagy could be stimulated in surviving cardiomyocytes to produce autophagy for self-preservation in the context of long-term chronic myocardial ischemia (56, 57). In addition, ubiquitination and autophagy, two major pathways of intracellular protein degradation, can be presented to the surface of MCH class I molecules, thereby activating CD8+ T cells and thus inducing an immune response in the organism (58). The ubiquitination of three hub genes, all associated with E3 ligases. Targeting the E3 ligase inhibits tumor progression (59). In the early stages of cancer, autophagy inhibits carcinogenesis development, while it promotes tumor survival and metastasis in later stages (60). Thus, our study provides a potential direction for the involvement of FBXO7, RAD23A, and MKRN1 in the immune response to CAD through ubiquitination or autophagy. Meanwhile, they may play an oncogenic role in cancer and are closely related to CD8+ T cells.

In addition, we found that FBXO7, RAD23A, and MKRN1 were significantly positively correlated in cardiac tissues and most cancers. No correlation studies have been found on the role of these three core genes in the association between CAD and cancer. Our study provides a new direction for the joint research between CAD and cancer while possibly providing possible options for optimizing the treatment of potential CAD in cancer patients. Of course, these need further experimental validation and clinical studies. It has now been confirmed that the efficacy of immunotherapy such as ICB and CAR T-cell therapy for tumors is mainly influenced by epigenetic regulation. And epigenetic modifications associated with CD8+ T cells can enhance immunotherapy (61). As we all know, epigenetics primarily focuses on methylation modifications (62). Therefore, we performed methylation analysis of the three core genes and showed that except for MKRN1, which was not statistically significant, FBXO7 showed hypomethylation, while RAD23A was the exact opposite. This provides a new theory to study the methylation process of FBXO7 and RAD23A and improve potential CAD immune efficacy in cancer patients.

To further understand the linkage of hub genes, we identified 19 common transcription factors associated with FBXO7, RAD23A, and MKRN1 by constructing a TF-gene network. FOXO3 has been reported to be a lifespan-associated gene that increases mortality in elderly patients with CAD (63). At first, FOXO3 was mainly considered a tumor suppressor. However, recent studies have found that it is also involved in maintaining starter cells in leukemia and colon cancer metastasis (64, 65). It was found that blocking STAT3 activity may be an effective strategy for treating CAD (66). STAT3 is often over-activated in various cancers and is associated with tumor immunosuppression and poor prognosis (67). Targeting STAT3 is emerging as a potentially promising therapeutic modality for many cancers (68). These two transcription factors may serve as targets for CAD-cancer interactions. Another 17 transcription factors are not studied for their association with CAD. In addition, there is growing evidence that competitive endogenous RNAs (ceRNAs) play a key role in CAD and cancer (69). However, there is still a lack of ceRNA networks that can predict the diagnosis and treatment of CAD. In the present study, we identified three common microRNAs (hsa-miR-335-5p, hsa-miR-340-5p, and hsa-miR-514a-3p) associated with FBXO7, RAD23A, and MKRN1. It has been suggested that miR-335-5p may regulate the process of cardiomyocyte differentiation by activating the WNT and TGFβ signaling pathways (70). It can be used as a potential biomarker for various cancers, for example, osteosarcoma, breast cancer, and non-Hodgkin’s lymphoma (71–73). In lung cancer, reduced expression of has-miR-340-5p may mediate cisplatin resistance (74). miR-514a-3p promotes melanoma growth (75). All three microRNAs have been studied in cancer in a limited number of ways. However, none of them is associated with CAD so far. At the same time, we also reverse inferred four common LncRNAs (AC021078.1, NEAT1, LINC00943, TUG1) of the common microRNAs. Among them, it was found that NEAT1 may inhibit CAD cell apoptosis by activating the miR-140-3p/MAPK1 pathway (76). neat1 promotes the progression and metastasis of several cancers (77, 78). linc00943 is associated with gastric cancer and clear cell renal cell carcinoma (79, 80). tug1 regulates endothelial cell proliferation and migration through the Wnt pathway to stimulate diabetic atherosclerosis (81). tUG1 is involved in tumorigenesis and tumor progression (82). The ceRNA regulatory network constructed in this study may play a role in CAD and cancer development and progression. The TF-gene and ceRNA network suggests that three hub genes may act as a bridge between CAD and cancer, which of course requires further experimental studies.

Meanwhile, to predict potentially effective therapeutic agents for CAD, we applied the CMAP database to identify 11 therapeutic agents that may reverse abnormally high expression of hub genes associated with CAD. Among them, doxorubicin, although improving prognosis and mortality in cancer patients, increased cardiotoxic effects (83). Gabapentin alleviates the painful reaction caused after coronary artery bypass grafting (CABG) in patients with CAD (84). Delsoline plays an essential role in cardiac remodeling and heart failure progression after myocardial infarction (85). For the 11 drugs identified in this study, no relevant studies have been found so far in cardiovascular disease, except for the above three drugs for which there are few relevant reports. Our findings may provide an essential basis for future exploration of effective drugs for CAD treatment.

In addition to the above studies, to further understand the potential biological pathways between CAD and cancer, we performed GSVA analysis on all samples in the dataset GSE12288. Genes in this dataset were significantly enriched in biological pathways such as il6 jak stat3 signaling, interferon-gamma response (IFN-γ), and il2 stat5 signaling. Single gene GSEA analysis of the screened three hub genes was also performed. The results showed that they were significantly enriched in multiple pathways such as Inflammatory response, interferon-alpha response, interferon-gamma response, TNF signaling via NF-kB, allograft rejection, and pi3k Akt mTOR signaling, but also in various cardiomyopathies such as dilated cardiomyopathy and hypertrophic cardiomyopathy. A literature review revealed that flavanones could inhibit the IL6/JAK/STAT3/SOCS3 signaling pathway and improve atherosclerosis (86). The IL6/JAK/STAT3 signaling pathway plays an essential role in the proliferation, survival, and invasion of tumor cells (87, 88). Meanwhile, targeting the IL-6/JAK/STAT3 signaling axis downregulates the expression levels of the CD8+ T cell surface receptors PD-1 or PD-L1 (89). Thus, IL-6/JAK/STAT3 may directly inhibit tumor cell growth or indirectly enhance the antitumor immune response effects of immune checkpoint inhibitors. CD8+ T cells may promote atherosclerosis by increasing the expression level of IFN-γ (11). CD8+ T cells can also induce cytotoxic effects in cancer cells by releasing IFN-γ (90). In the presence of IFN-γ, cancer cells can release several immunosuppressive mediators, including PD-L1 and transcriptional activator 3 (STAT3). These mediators negatively feedback to inhibit the release of IFN-γ, thereby suppressing the activity of CD8+ T cells (91, 92). IL-2 activates STAT5 in CD8+ T cells (93). H9T, an agonist of IL-2, leads to increased expression of CD8+ T cells. In animal models of melanoma and acute lymphoblastic leukemia, CAR-modified CD8+ T cells expanded with H9T have strong antitumor activity (94). The pro-inflammatory cytokine IFN-α can lead to atherosclerosis flare-up and progression (95). the antitumor effect of IFN-α is mediated mainly by CD8+ T cells (96). the inflammatory response has been considered a cancer marker and is closely associated with cancer progression (97). In addition, it is an essential pathogenic mechanism of atherosclerosis (5). MicroRNA-499 was found to cause endothelial cell injury by activating tnfa signaling via NF-kB (98). And NF-kB is also involved in tumorigenesis and development (99). Triclosan (TCS) can damage endothelial cells by inhibiting the PI3K/Akt/mTOR signaling pathway (100). The PI3K/Akt/mTOR pathway is usually activated in cancer, and inhibition may have an anti-cancer effect (101). These findings reaffirm that the development of CAD is associated with multiple immune processes mediated by CD8+ T cells. Moreover, these immune processes play a crucial role in cancer.

In conclusion, we identified FBXO7, RAD23A, and MKRN1 as biomarkers in CAD associated with immune responses mediated by CD8+ T cells. This finding helps to understand the part of the immune response in the development of CAD. At the same time, they play the role of potential therapeutic targets in various cancers. FBXO7, RAD23A and MKRN1 may act as hub genes linking CAD and cancer in terms of immunity. It may provide some theoretical basis for identifying new drug targets and developing new therapeutic approaches to reduce the incidence of CAD in cancer patients. Of course, there are some limitations to this study. On the one hand, FBXO7, RAD23A, and MKRN1 have not been reported in CAD. The mechanism of their roles in CAD needs to be studied in depth. On the other hand, the functions of these three hub genes in CAD and cancer are only based on our bioinformatics analysis, and future in vitro or in vivo experiments are needed to validate these findings further.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/supplementary materials.

JZ, YW, XX, YanW, and SZ substantially contributed to conception and design. SZ and YW conducted the literature search. SZ and YW contributed to the acquisition of data or analysis and interpretation of data. SZ wrote the article. YW performed data analysis and drafted. JZ revised the article. All authors gave the final approval of the version to be submitted.

This work was supported by scientific research funding from the 2018 Science and Technology Plan Project of the Yunnan Provincial Science and Technology Department. Grant Program [2018FE001 (–285)].

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.