Breast cancer clinical outcomes and tumor immune microenvironment: cross-dialogue of multiple epigenetic modification profiles

Introduction

Breast cancer (BC) is the leading cause of tumor-associated death in females worldwide and one of the most common malignancies. It is a highly heterogeneous disease, exhibiting various characteristics across different individuals. Owing to difficulties in early diagnosis and the aggressive nature of tumor progression, a significant number of BC patients are often diagnosed at an advanced stage or with metastatic lesions [1]. As a result of early detection and treatment, 5-year relative survival rate was almost 100% for BC patients diagnosed primarily in stage I, compared to only 26% for BC patients initially diagnosed in stage IV [2]. According to PAM50, BC has five subtypes (i.e., luminal A, luminal B, Her2-enriched, normal-like, and basal-like) [3]. However, it has been observed that patients with the same molecular type and similar clinical characteristics can exhibit different prognoses and responses to chemotherapy or immunotherapy [4], suggesting other subtle factors affecting their prognosis and treatment outcome. Prognosis in BC patients can always be improved with an accurate diagnosis, detection, and treatment [5]. Hence, to enhance the prognosis and treatment of BC patients, it is essential to investigate new effective biomarkers and therapy modalities. The post-transcriptional modifications have attracted much attention in BC research because its important role in BC biology [2].

Messenger RNA (mRNA) and non-coding RNA are post-transcriptional modifications [6]. To date, numerous studies have identified more than 170 RNA modifications [7]. RNA modification disorders have been linked to a wide range of diseases, including BC [8]. Post-transcriptional alterations include N6-methyladenosine (m⁶A), 5-methylcytosine (m⁵C) and N1-methyladenosine (m¹A), which affect adenosine in various ways [6]. m⁶A is the most common variation in mRNA, initially found in the 1970s [9]. It has been described to be a changeable, normal adenosine alteration, particularly abundant in mammalian mRNA 3 ’UTR [10]. In addition, m⁶A has been found to be associated with various cellular processes such as RNA stabilization, metabolism, transcriptional control, and intracellular signaling. Its biological significance can be seen in several contexts, including embryonic self-renewal, the development of hematopoietic stem/progenitor cells, circadian rhythm regulation, heat shock response, neurofunction, and cancer. These examples highlight the diverse roles that m6A plays in essential biological processes [8]. m⁵C is a common human RNA mutation found in both mRNAs and ncRNAs [11]. So far, the human genome has been revealed to have overall m⁵C sites. This reversible post-transcriptional modification affects many molecular processes, including RNA-protein interactions, RNA stability, and translation efficiency [8]. It follows orders by “author” (methyltransferase), “reader” and “eraser” (demethylase). According to a recent study, NSUN6 mediates location-specific deposition of m⁵C in mRNA to regulate translation quality [12]. It is known that the m⁵C modification can be transformed into various kinds of alterations, such as hm⁵C, although the m⁵C demethylase is unclear [13]. In recent years, abnormal m⁵C has been discovered to be carcinogenic in several malignancies [8]. Preliminary results of m¹A were available more than 50 years ago. Adenosine becomes m¹A when a methyl group joins the N1 site. The Watson-Crick base pair interface is the location of the methyl group, disrupting the base pair, while the positive charge carried by m¹A affects local RNA structure or protein-RNA interactions [10]. M¹A is also found to be a highly enrichment effect for 5 ’UTR translation [14]. Based on earlier studies, M¹A was found in tRNA, mitochondrial transcripts, as well as rRNA and mRNA [15]; In eukaryotes, m¹A is abundant in tRNA and rRNA. Furthermore, recent studies have revealed that m¹A alteration may be impacts mRNA translation [16].

Cancer-related RNA methylation alterations have been recognized as a promising alternative for developing biomarkers that can be used in prediction, prognosis, and diagnosis. Discoveries of methylation associated with antitumor immunity are promising [17]. Methylation, like necrosis, is an important cellular response that regulates cancer progression, spread, and metastasis [6]. However, the mechanism that methylation modulators affect the prognosis, and the underlying molecular pathways of BC is not well understood. methylation landscapes help predict BC prognosis, and in recent years STOX1, UNKL, ZMAT3, and ZNF443 have all been identified as novel biomarkers of BC metastasis with strong methylation deregulation and association with metastasis [18]. Although methylation is critical for both oncogenic and anticancer pathways, few studies have addressed the significance of methylation in cancer, particularly in BC.

Cancer immunotherapy uses anti-tumor immune responses to identify and destroy tumor cells by stimulating the host immune system [19]. T cell-related immune responses only partially benefit a small percentage of cancer patients. Although T cells contribute to immune checkpoint inhibitors, they also stimulate anti-cancer responses [20, 21]. BC is known to be an immunologically dormant tumor, which slows obviously its response to immunotherapy. Encouraging preclinical trial results, along with years of clinical data, suggest that immunotherapies may hold the key to ushering breast cancer into an era of clinical intervention that has yet to be experienced [22]. Furthermore, there is growing evidence that immune infiltration in the tumor microenvironment (TME) plays an important role in determining BC progression. According to a large number of studies, the prognosis of patients with triple-negative BC was related to the quantity of T cells obviously in the tumor [23]. For this reason, the method we got from immune cell infiltration (ICI) profiles—clustering of BC samples relied on molecularly specific subgroups related to the ICI patterns—may be the most reliable and promising way to thoroughly assess tumor sensitivity to immunotherapy. It also facilitates personalized treatments to increase their effectiveness. But the investigation did not think over the incomplete context of BC’s TME, because it was mediated by ICI patterns.

With the rapid advancement of science, our research on cancer treatment, prognostic biomarkers, and the carcinogenic framework will continue to progress. We aim to investigate the prognostic significance, expression, and presentation of the regulatory axis of MRGs in BC. Our findings have the potential to offer further insights into the molecular mechanisms and prognostic biomarkers associated with BC.

Materials and Methods

Results

Discussion

Methylation, which involves significant modifications of proteins and nucleic acids, has been strongly associated with cancer [17]. Methylation, similar to necrosis, plays a crucial role in regulating cancer progression, metastasis, and spread as an important cellular response. RNA methylation can also have an impact on the tumor immunological microenvironment and the prognosis of patients [25–28]. For example, In BC cells, hypoxia-inducible factor-1α and 2α dependent ALKBH5 can be activated, which results in NANOG mRNA being demethylated and influences the formation and incidence of tumors [29]. when the m⁶A methyltransferase is silenced, the stability of gene expression is impacted by the inhibition of the TP53 signaling pathway (a critical tumor suppressor gene) [30]. Similarly, m¹A was reported to improve the translation efficiency and involved in determining the mRNA stability of BC [31]. DNMT3B and ALYREF (m5C regulators) were upregulated in BC and their high expression hint unfavorable prognosis [32]. The decreased levels of m6A (methylation) on ADAM19 in glioblastoma multiforme result in increased expression, which enhances the development and self-renewal of glioblastoma stem-like cells. This, in turn, ultimately leads to carcinogenesis in glioblastoma. Reduced m⁶A levels result in AKT activation, increased proliferation, and tumorigenicity in endometrial cancer cells in about 70% of endometrial cancers. rectal cancer patients with different consensus clusters of m⁶A RNA methylation regulators showed a significant difference in overall survival rate [27, 33]. Muraoka et al. demonstrated the viability of employing miR-34b/c methylation as a diagnosis and prognosis for malignant pleural mesothelioma in the context of lung cancer. Additionally, Lian et al. reported that RNA 5hmC methylation serves as a melanoma prognostic marker [34]. Five m⁶A RNA methylation regulatory factors, including HNRNPA2B1, HNRNPC, KIAA1429/VIRMA, RBM15, and METTL3, have been linked to lung cancer patients’ overall survival rates, according to a different study. Actually, as many genes affected by RNA methylation, including BCL2, FOXM1, HBXIP, and SOCS2, are crucial in regulating tumor development, RNA methylation is not only associated with carcinogenesis in diverse forms of cancer [35–37], Additionally, it is essential for tumor suppression since some loci’s lack of RNA methylation results in carcinogenesis [38]. According to a substantial number of studies, the overall effects and characteristics of different MRGs (methylated RNA genes) on the adjustment of tumor microenvironment (TME) infiltration have not been fully elucidated yet. However, the majority of research has concentrated on a single MRG or TME cell. According to recent research, MRG is highly aberrant at the genomic and transcriptional levels in BC [39, 40].

We distinguished two distinct molecular subtypes based on 57 MRGs. Those with MRG cluster A of OS have more severe OS than patients with MRG cluster B, and the two clusters of OS have quite different natures.

Changes in mRNA transcriptomes across different methylation subtypes are significantly related to biological pathways involving MRG and the immune system. We discovered three genetic clusters using the DEGs in the two MRG clusters. According to the data, MRG may be used to predict clinical outcomes and the effectiveness of immunotherapy in BC. We were able to demonstrate their prognostic abilities as a result and identify the trustworthy prognostic risk score. Higher and lower risk scores were seen in immune-activated and inhibition-driven MRG clusters, respectively. The TME, prognosis, mutation, and therapy susceptibility among various patient populations varied significantly. Then, utilizing a fusion of risk ratings and tumor stage, we produced a quantitative Normandy map that dramatically enhanced performance and made it simpler to apply risk scores. This prognostic model can be used to categorize BC patients, aid in a better understanding of the molecular causes of BC, and offer novel ideas for focused treatment. Our study, leveraging the CIBERSORT deconvolution method, has unveiled significant disparities in immune cell infiltration across distinct MRG groups in breast cancer, highlighting the complexity of immune landscapes within subtypes. The observed dominance of specific immune cells, such as cytotoxic T cells or regulatory T cells in certain groups, points to varying degrees of tumor aggressiveness or immunosuppression, critical for tailoring prognosis and immunotherapy approaches. Notably, a pronounced presence of cytotoxic T cells could indicate a favorable response to immune checkpoint inhibitors, steering therapeutic strategies. These patterns of differential immune infiltration might also mirror the spectrum of immune evasion mechanisms, a reflection of the tumor’s adaptive pressures on the immune system. Unraveling these mechanisms is pivotal for crafting bespoke immunotherapies aligned with the unique immune profiles of MRG-defined subgroups. The distinct immune infiltration landscapes underscore the integral role of immunogenomic analyses in oncology, enriching our understanding of tumor-immune dynamics and opening new avenues for personalized immunotherapeutic interventions, necessitating further exploratory studies to elucidate the functional implications of these immune cell infiltration differences.

After receiving standard therapy, BC patients have a poor prognosis because of elevated checkpoints, tumor-infiltrating cells, and neoantigens [41, 42]. The fact that patients with BC still have a heterogeneous prognosis despite recent therapeutic advancements emphasizes the crucial part TME plays in the development and progression of BC tumors [43]. Immune cells such as macrophages, lymphocytes, and granulocytes are among the key biological components of TME [44]. To aid in survival, these cells take part in a range of immunological reactions and actions, such as the coordination of inflammatory responses by tumors [45, 46]. Additional data point to a critical function for TME in cancer development, progression, and therapy resistance [47]. In the present investigation, immune activation-driven MRG (cluster B) was linked with lower risk scores than immunological inhibition-driven MRG (cluster A). In terms of the TME characteristics and relative richness of 22 immune cells, we discovered that these two molecular clusters and various risk scores differed significantly from one another.

These findings suggest that MRG plays a crucial role in the growth of BC. An increasing body of evidence supports the significance of potent T cells, memory T cells, and T cells in the immunological response against BC. CD4+ T cells play a significant role in tumor management, as they can either enhance or suppress anti-tumor responses. However, CD8+ T cells are generally considered to be the key drivers of anti-tumor immunity. They are believed to be primarily responsible for exerting anti-tumor effects and facilitating immune responses against cancer cells. For instance, by stimulating a variety of innate immune cell types, including CD8+ T cells, NK cells, and others, conventional CD4+ T cells can aid in the control of tumors [48]. Since tumor-infiltrating T cell concentrations in BC samples were higher than those in healthy tissues, this suggested a favorable prognosis. Increased infiltration of activated memory CD4+ and CD8+ T cells and T cells indicated that B cluster and low-risk groups had better prognoses and likely contributed to the advancement of BC. The importance of effector T cell, memory T cell, and T cell differentiation is becoming increasingly clear [25]. M1 macrophages and M2 macrophages are the two phenotypes of tumor-associated macrophages (as we all known, the former prevents the spread of cancer, and the latter promotes cancer progression). According to this study, patients with low-risk scores may benefit from immunotherapy because macrophages M1 was more abundant in the groups with low-risk scores. Immunosuppressive M2 macrophages support matrix remodeling, which encourages the growth of malignancies [49]. As in previous research, we discovered that M1 macrophage infiltration increased in MRG cluster B groups with low risk-score and good prognosis, but not in MRG cluster A groups with high risk-score and poor prognosis.

Although, it was shown that a large number of studies found that B cells support the immunological response. In Hodgkin lymphoma, for instance, studies have indicated that B cell enrichment is positively correlated with responsiveness to PD-1 inhibition, as described by Vari et al. This suggests that B cells may contribute to the effectiveness of immune checkpoint blockade therapies in certain cancer types [50]. According to Engelhard V et al., patients who reacted to immune checkpoint inhibition had significantly higher levels of the B cell-related genes than those who did not [51]. According to Hollern et al. immune checkpoint therapy induces T follicular helper cell activation of B cells to facilitate the anti-tumor response in these models [52]. Additionally, a favorable prognosis in BC was associated with tumor-infiltrating B cells. Numerous studies have demonstrated that T cells and B cells interact and go through cooperative selection, specialization, and clonal growth in tertiary lymphoid structures associated with tumors. Importantly, B cells can interact with T cells by presenting them with homologous tumor-derived antigens; the functional outcomes of such interactions are influenced by the B cell phenotype [53]. The findings of this study showed that B cells represent a unique immunotherapy target and may be an effective cancer-fighting tool, rather than merely incidental contributions to anti-cancer immunotherapy. This is consistent with our findings.

In this study, the expression levels of a part of immune cells were found to be different obviously in the risk model of MRGs. Our study revealed that the risk score was inversely correlated with resting dendritic cells, M1 macrophages, monocytes, plasma cells, active NK cells, follicular helper T cells, gamma delta T cells, regulatory T cells, and activated memory CD4 + T cells. This implies that BC immune cell infiltration is related to the risk model created using MRGs. Our study shows a significant correlation between most immune cells and 12 genes, The high-risk score group was related to a high stromal score, and the low-risk score group was closely associated with a high immune score. Yan-Fei Ma et al. demonstrated Early breast cancer was shown to have high levels of CD52 expression, and these levels were correlated with a good prognosis [54]. Higher infiltrations of M1 macrophages, monocytes, T follicular helper cells, and resting memory CD4 T cells were caused by overexpression of CD52. Downregulation of CD52 led to significant M2 macrophage infiltration [54]. Studies have shown that some TME and IGLL5 expression are linked. Additionally, three different types of TME are positively linked with IGLL5 expression, according to our immune infiltration findings. Potential for IGLL5 as a predictive biomarker of clear cell renal cell carcinoma [55]. According to Scannell Bryan et al., the NEK10 variant may play a role in the incidence of breast cancer [56]. Our study established a risk model for BC and found evidence of MRGs involvement; however, further clinical BC tissue samples and cell tests are required to corroborate this. A high-risk score is an independent risk factor for a poor prognosis in BC patients and is associated with patient outcomes.

The study has several limitations. First, all the analysis is based entirely on data from a common database. Therefore, an inherent bias in case selection may have influenced the results. Second, further prospective studies and additional in vitro and in vivo studies are needed to validate our findings. In addition, most datasets were unable to analyze data on important clinical variables (such as surgery, neoadjuvant chemotherapy, and chemoradiotherapy) that could affect the outcome of immune response and methylation status.

Conclusions

Through our comprehensive investigation of MRGs, we have discovered their influence on the tumor microenvironment (TME), clinicopathological characteristics, and various prognostic regulatory systems. Furthermore, we have identified the therapeutic roles of immunotherapy and commonly used anti-tumor medications. These findings emphasize the clinical significance of MRGs and provide novel insights to guide both immunotherapy and conventional anti-tumor strategies in BC patients.

Supplementary Materials

Supplementary Table 1

Supplementary Table 2

Supplementary Table 3

Supplementary Table 4

Supplementary Table 5

Supplementary File 1