Breast cancer is the prevalent carcinoma in human and the immunoprofile in TME has been proven to play critical roles in tumor occurrence, development, prognosis and treatment [5, 6]. Infiltrating immune cells contain T cells (CD4 T resting, CD4 T activate, Tregs, T helper, CD8 and Tγ-δ), B cells (B naive and memory), macrophages (M0, M1 and M2), NK cells (NK resting and activate), DC cells (DC resting and activate), monocytes, neutrophils and eosinophils [22–25]. Studies indicated that macrophages had a wide range of subtypes in past decades with various cell markers, such as CD206, CD64, MHC Ⅱ, MerTK, Siglec-1, Ly6C, Lyve1 and CX3CR1 [26–28]. Due to different localized environments or precipitating factors, macrophages were divided into more sophisticated subtypes, such as M2a, M2b, M2c, M2d and TAM [29]. TAMs were under the intense discussion for decades due to high heterogeneity and various functions in different cancer models [30, 31]. Appearing cell markers for TAMs, such as CD206, PM-2K, CD40, COX2 and CCR2, led to deeper investigations of tumor associated macrophages [32–34].
In the current study, we used typical cell markers for different immune cell types, namely CD3 for T cells, CD19 for B cells, CD68 for macrophages, CD68+TLR4+ for M1, CD68+CD163+ for M2 and CD68+CD40+ for TAM, to investigate the detailed immunoprofiles of breast cancer with MI/IHC method, compared with non-carcinoma, para-cancer and different pathological and molecular types. However, the selection of macrophages subtypes, especially TAMs, was still ambiguous, because studies showed that TAMs resembled not only M2 subtype, but also contained M1-like subtype with diverse cell markers [35, 36]. Of note, in the benign samples of this study, we also clearly identified the CD68+CD40+ TAM subtype (SFigure 2A, green arrows). Thus, the novel and specific TAM cell markers are still urgent need to be found.
We collected 140 cases of breast tissues containing benign and malignant diseases, but we only had two DCIS and two male patients, which had limitations in investigating these clinic classifications. Also, we lost 18 samples during the staining process. However, our available samples were statistically sufficient with molecular equilibrium and pathological types. Similar with the bioinformatic analysis results (Figure 7), our sample generally indicated the immunoprofiles of breast cancer, and even showed more precise differences among various sample types.
Compared between benign and malignant samples, our data indicated that the immune cells transferred from gland/para-gland to stromal areas (Figure 2 and SFigure 2) and the percentages of macrophages and T cells increased significantly, while B cells decreased (Figure 2). Studies showed that regulations of interstitial components of tumor tissues were crucial to tumorigenesis and metastasis [37, 38], and our data further verified the re-localization of immune cells, especially M2 in altering the TME and promoted the tumor-promoting changes in the matrix formation and angiogenesis [39–41]. Due to the shortage of MI/IHC method, we did not further classify subtypes of T cells and B cells, which might show diverse infiltrating levels as results of bioinformatic analysis (Figure 7). In the current study, we concluded that M2 was the highest subtype in tumor environment than M1 and TAM (Figure 2C and 2G). However, some studies indicated that TAM was the dominant macrophage subtype in breast cancer [42]. First of all, as discussed above, different cell markers used to label TAMs contributed to the contrary results, even regarding CD206+ or CD163+ as TAM, which was normally regarded as M2 markers [34, 43]. Second, we found that even T cells or B cells showed overlapped staining with both CD3+ and CD19+ in MI/IHC. We analyzed the immune cell phenotypes using double-positive mode, during which we discarded the non-specific double-positive cells and only count single positive cells. Of note, we found that CD68+CD40+ TAM also had CD163 or TLR4 positive binding, indicating that TAMs had a high level of heterogeneity.
By comparing benign with malignant samples, different pathological types, molecular types, para-cancer with cancer samples, and different stages, our study indicated that macrophages, especially the M2 subtype, were the most obviously regulated immune cells that were stably and highly expressed in breast cancer with a malignant degree-dependent manner. Some studies also correlated the percentages of M2 in peripheral circulation with the poor prognosis in breast cancer patients [44]. Our MI/IHC data together with bioinformatic analysis results consistently further indicated that the infiltrating levels of M2 were correlated with tumorigenesis and malignant degree. To further investigated the macrophage regulation or differentiation mechanisms, we compared the DEGs between breast cancer and normal tissues, overlapped with immune-related gene database, and obtained 26 potential DEG-IRGs (Figure 8). Combining our data of macrophage characteristics, we further applied GO analysis targeting macrophage biological regulations, and obtained single macrophage related biological function, namely macrophage apoptosis process, IRF3, IRF7 and NOD2 were all upregulated in breast cancer. IRF3 and IRF7 were key transcription factors involving in diverse pathways, including immune regulation, interferon response, and viral response [45, 46]. NOD2 is a member of the Nod1/Apaf-1 family, and was mainly involved in the inflammatory response and NFκB activation [47]. Their dysregulation in breast cancer and correlation with macrophage functions indicated the potential targeting values in cancer therapy.