The exploration of the TME and immunological treatment in breast cancer have become increasingly important in recent years. In our study, although the TME did not correlate with the stages of breast cancer, we verified that it was highly associated with the subtypes of breast cancer and gene mutations (CDH1, TP53 and PTEN) and possessed immunological characteristics. The combined analysis of OS, time-dependent ROC, and the PPI network revealed that the genes of the top 3 GO terms of the upregulated DEGs from the high vs low immune score groups were associated with better prognosis in breast cancer, and 15 of them were related to good prognosis in breast cancer, especially CD226 and KLRC4-KLRK1. High CD226 and KLRC4-KLRK1 expression levels were identified, and their correlation with better OS in specific stages or subtypes of breast cancer was validated.
The TME is crucial in tumor initiation, progression and drug resistance in breast cancer, and the infiltration of nontumor cells in the TME was calculated by the ESTIMATE algorithm. The ESTIMATE algorithm developed based on gene expression data is valid and effective in various cancers, such as prostate cancer, breast cancer, and colon cancer[9, 10]. By utilizing the breast cancer cohorts in the TCGA database and ESTIMATE algorithm-derived scores, we first analyzed the relationship between stromal/immune scores and different stages or subtypes of breast cancer and found that stromal/immune scores were highly correlated with subtypes of breast cancer, especially in the basal-like subtype, and the luminal B and HER2-enriched subtypes also showed relevance with immune scores. This finding demonstrated that TME variation correlated with subtypes of breast cancer and possessed immunological characteristics, which is consistent with the findings of a previous study[11]. However, the high immune-related scores did not predict a better prognosis in breast cancer subtypes. This is because prognosis is more influenced by various factors, such as age, race, never being pregnant or having a first child after age 30, type of surgical procedure, initial tumor size, clinical lymph mode status, and neoadjuvant chemotherapy[12, 13].
Second, we analyzed the relationship between the TME and gene mutations. Although our data demonstrated that CDH1, TP53 and PTEN mutations were closely associated with the TME, BRCA1 and BRCA2 mutations did not show significant differences in TME variation. BRCA1 and BRCA2 mutant genes predispose individuals to an elevated risk of breast cancer, and those with a family history of cancer are recommended to undergo gene detection based on the National Comprehensive Cancer Network (NCCN) guidelines[14]. Thus, BRCA1 and BRCA2 mutations are not highly frequent in sporadic cases, and the design of upcoming trials could stratify patients by BRCA status to avoid potential bias. CDH1, TP53 and PTEN are tumor suppressor genes, and their alteration presents poor survival and worse prognosis in breast cancer[15–17]. Our study findings coincide with those of a previous exploration that after the exclusion of BRCA1 or BRCA2, the TME correlates with TP53, CDH1, and PTEN mutations, and their mutations revealed a high or moderately increased risk of breast cancer[18].
Third, we screened the DEGs generated from the comparison of high vs low stromal/immune scores. According to the GO term analysis, the upregulated genes from the high vs low immune score groups were mainly involved in lymphocyte/leukocyte activation or proliferation and cytokine-cytokine interaction. Similarly, the upregulated genes from the high vs low stromal score groups were mainly involved in ECM organization and cytokine activity. When further exploring the DEGs in the TME, many previous studies have ignored genes with low expression but exhibit high significance in antitumor activity and are related to better prognosis in patients. Thus, to eliminate the factor of tumor cells downregulating the expression of antitumor genes during tumor progression, we focused on the genes of the top GO terms of the upregulated DEGs based on functional enrichment.
Fourth, through step-by-step screening via OS analysis, time-dependent ROC analysis, and PPI network analysis, we revealed that the top GO term genes of the upregulated DEGs from the high vs low immune score groups exhibited better prognosis in breast cancer, which could be explained by the fact that the immune system plays an important role in cancer development and therefore potentially offers novel targeted therapies in antitumor treatment[19]. Ultimately, we found 2 important TME genes with good prognosis (CD226 and KLRC4-KLRK1).
CD226, also known as DNAM-1, is an activating receptor expressed on various immune cells, such as CD4 + and CD8 + T lymphocytes, regulatory T cells (Tregs), monocytes, macrophages, and NK cells[20, 21]. CD226 serves as a costimulator, enhances T cell and NK cell activation[20], and exhibits significance in innate/adaptive immune regulatory networks. When combined with its ligand CD115 or CD112 upregulated in tumor cells[22], CD226 facilitates the cytotoxicity of NK cells[23]. In addition, in Treg-mediated tumor immune escape, Tregs express relatively high levels of TIGIT and low levels of CD226 compared with effector T cells (Teffs), resulting in a high ratio of TIGIT/CD226 expression and accelerating tumor development. In contrast, augmenting CD226 expression and reversing the TIGIT/CD226 ratio would predict a good clinical outcome[24]. However, reduction in CD226 expression decreases the immune regulatory capacity, and CD226-deficient CTL or NK cells exhibit markedly less cytotoxic activity against DNAM-1 ligand-expressing tumors[25]. Accumulating evidence has shown that CD226 plays a pivotal role in tumor recognition and cancer immune surveillance[26], even promoting antitumor immune responses mediated by NK and T cells.
Killer cell lectin-like receptor subfamily C4 - killer cell lectin-like receptor subfamily K1 (KLRC4-KLRK1) belongs to killer cell lectin-like receptor family[27] and represents naturally occurring read-through transcription between neighboring KLRC4 and KLRK1. KLRC4 lacks a significant portion of the KLRC4 coding sequence but encodes the KLRK1 protein. Once tumorigenesis occurs, the amount of KLRK1 ligand increases immediately. KLRK1 (or NKG2D) is also an activating receptor expressed by NK cells and T cell subsets. In addition, KLRK1 could augment the cytotoxicity of NK cells/T cells or synergize with immune checkpoint inhibitors to eliminate tumor cells[28]. However, tumor cells evoke a range of mechanisms to evade KLRK1 surveillance system detection and impair the clinical benefits of immunotherapy in various cancers[29],[30]. The downregulation of KLRK1 hampered NK cell cytotoxicity[31]. Conversely, blocking the shedding of ligands by tumors or the release of KLRK1 ligand-bearing exosomes might restore the expression of KLRK1 receptors on NK cells and T cells and improve their activity[32].
It was reported that the KLRK1 axis is becoming an emerging target in cancer immunotherapy[33, 34], and the overexpression of CD226 or KLRK1 on NK cells resulted in efficient anti-sarcoma activity[35]. These studies were in accordance with our exploration to provide a molecular basis for the development of CD226- and KLRC4-KLRK1-targeted antitumor immune therapeutics. When considering that the results from the TCGA database were verified by the GEO database in part and that the samples from GEO were much fewer than those from TCGA database, further exploration should include more breast cancer patients.
Tumorigenesis is initiated by 3 steps: cancer cell elimination by various immune cells, such as NK cells and CD8 + T cells, then immune pressure leading to the selection of tumor cell variants and finally, immune escape by inhibiting effector cells or inducing tolerogenic cells[36]. Escaping antitumor immunity is a hallmark for the progression of breast cancer. In the TME, tumor cells interact with various types of immune cells by activating immune checkpoint pathways[37, 38]. Immune checkpoints (CTLA-4, PD-1/PDL1, LAG3, TIM3 and TIGIT) are orchestrated by a series of costimulatory and inhibitory signaling molecules and then modulate effector T lymphocyte (Teff) activity. Recent advancements in antibodies against immune checkpoints have highlighted the benefits of immune checkpoint inhibitors in both animal studies and clinical trials[39].
In this study, we also detected immune checkpoint genes, such as CTLA-4, PD-1, LAG3 and TIM3, and found that their high expression was not associated with good prognosis in breast cancer (data not shown). This finding is probably related to the patients’ status of checkpoint gene expression and immune state of the TME[8]. To achieve unbiased results, the identification of immune checkpoint gene expression status is critical. The investigation into the relationship between immune checkpoint gene expression and the TME immune state (such as tumor-infiltrating lymphocytes and the T cell receptor repertoire) would provide key insights into checkpoint blockade therapy.
At present, there are multiple antitumor approaches: inverting tumor immunosuppression (for example, employing immune checkpoint inhibitors and the direct induction of the Teff immune response), using immune-based therapies targeting specific immune cell types (for example, improving cytotoxic efficacy and immune surveillance through NK cells or Teffs), reducing the number of immunosuppressive myeloid cells, inhibiting Tregs, and altering the function of myeloid cells[7].
Despite combining multidisciplinary treatment strategies, breast cancer patients still have a comparably high mortality rate. An improved understanding of the immune-related genetic profile of the TME in breast cancer and the identification of new immunological targets are critical for improving clinical outcomes. CD226, KLRC4-KLRK1 and subsequent new targets seem to be promising avenues for promoting antitumor targeted therapy in breast cancer.