Identification of Immune Cell Components in Breast Tissues by a Multiparametric Flow Cytometry Approach

Simple Summary The tumor microenvironment in breast cancer plays important roles in tumor development and treatment response, giving important information critical for disease management. Today, an analysis of the tumor microenvironment is included in routine histopathologic reporting for practical clinical application. This manuscript aimed to deepen the study of the tumor microenvironment, analyzing the immune cells in breast tumoral and benign pathologies. Indeed, using a deep immunophenotyping approach by flow cytometry, we have studied the immune cells at the level of breast tissues, identifying different immunophenotyping that could be useful in the diagnosis and follow up of breast pathologies. As possible targets are continually being discovered in the tumor microenvironment, a future approach to breast cancer diagnosis and therapy could likely combine cancer cell elimination and tumor microenvironment modulation. Abstract Immune cell components are able to infiltrate tumor tissues, and different reports described the presence of infiltrating immune cells (TILs) in several types of solid tumors, including breast cancer. The primary immune cell component cells are reported as a lymphocyte population mainly comprising the cytotoxic (CD8+) T cells, with varying proportions of helper (CD4+) T cells and CD19+ B cells, and rarely NK cells. In clinical practice, an expert pathologist commonly detects TILs areas in hematoxylin and eosin (H&E)-stained histological slides via light microscopy. Moreover, other more in-depth approaches could be used to better define the immunological component associated with tumor tissues. Using a multiparametric flow cytometry approach, we have studied the immune cells obtained from breast tumor tissues compared to benign breast pathologies. A detailed evaluation of immune cell components was performed on 15 and 14 biopsies obtained from breast cancer and fibroadenoma subjects, respectively. The percentage of tumor-infiltrating T lymphocytes was significantly higher in breast cancer patients compared to patients with fibroadenoma. Infiltrating helper T lymphocytes were increased in the case of malignant breast lesions, while cytotoxic T lymphocytes disclosed an opposite trend. In addition, our data suggest that the synergistic effect of the presence/activation of NK cells and NKT cells, in line with the data in the literature, determines the dampening of the immune response. Moreover, the lymphocyte-to-monocyte ratio was calculated and was completely altered in patients with breast cancer. Our approach could be a potent prognostic factor to be used in diagnostic/therapeutic purposes for the improvement of breast cancer patients’ management.


Introduction
Breast cancer remains the principal cause of death among women, with 2,261,419 newly diagnosed cases in 2020, of which 684,996 were mortality cases [1]. Based on gene expression and molecular features, breast cancer is classified into four subtypes cific IHC staining, can be more affordable to define the TILs immune components [19,20]. However, an additional way to describe the immune cell components could be a multiparametric flow cytometry (FCM) approach. Indeed, despite FCM having been extensively used for the characterization of normal and pathological lymphocytes in liquid specimens such as peripheral blood and bone marrow samples [21][22][23], fewer reports evaluated its usefulness for defining immune cell subpopulations in solid cancer tissues [24]. These protocols provide for an optimization phase of the breast tissue digestion protocol based on the experimental objective. In general, for solid tumors the experimental protocols foresee a mechanical digestion phase, through the use of scalpels, and an enzymatic digestion phase that favors the obtaining of a sample for single-cell analysis [25]. In the present manuscript, we have analyzed the immune cells components (T-lymphocytes, NK-lymphocytes and monocytes) obtained from breast tumor tissue compared with those obtained from benign breast pathologies using an FCM approach to provide in detail the profile of immune cells residing in malignant and benign breast lesions.

Patient Samples
The present study comprised 29 female patients who underwent breast surgery at the Clinica Villa Fiorita S.p.A, between June 2020 and May 2022 (Aversa, Italy). The study was approved by the Ethics Committee of IRCCS Pascale (Naples, Italy) with reference number 3/19 approved on 29 May 2019. All methods were performed in compliance with standard operating procedures and in accordance with the Declaration of Helsinki and each patient participated in the study by signing written informed consent. Table 1 summarizes the characteristics of the patients included, which was 15 women with malignant breast cancer and 14 women with benign fibroadenoma. Table 1. Clinico-pathological characteristics of the studied patients.

Hematoxylin and Eosin (H&E) Staining
Sections (4 µm) were deparaffinized with 2 changes of xylene for 10 min each. The breast cancer sections were hydrated by passing through a decreasing alcohol series (100, 95, and 70%). Slides were stained in hematoxylin for 10 min at room temperature and then washed under running tap water for 5 min. Acidified alcohol (1%) was used for differentiation (1% HCl in 70% alcohol) for 2 min. Sections were washed under running tap water until the sections were blue again by dipping in an alkaline solution followed by another tap water wash. Then the sections were stained in 1% eosin Y for 5 min at room temperature. Sections were washed in tap water for 3 min and dehydrated in increasing concentration of alcohol and cleared in xylene.

Breast Tissue Digestion and Flow Cytometry Analysis
Flow cytometry was used to evaluate the immune cells at the level of the tumor microenvironment. For the isolation of the immune cell components, a protocol established in our laboratory, based on previously used methods, was applied [26,27]. We used freshly resected tissue that was manually minced with a scalpel blade and then incubated overnight at RT in a shaker table bed (100 rpm/min) with 1.0 mg/mL collagenase A (Roche Diagnostic GmbH, Basel, Switzerland) and 10 µg/mL hyaluronidase from bovine testes, Type I-S (Sigma Aldrich, Saint Louis, MO, USA) diluted in DMEM Medium (Gibco; Thermo Fisher Scientific, Milan, Italy). Subsequently, a single-cell suspension was prepared by filtering through 70-µm nylon strainers (BD Biosciences, Franklin Lakes, NJ, USA). For surface staining, cells were then washed twice with PBS containing 2% FBS (Gibco; Thermo Fisher Scientific, Inc.) and then labeled together with the manufacturer's suggested dilutions (incubation, 30 min at R.T.) of the following antibodies: CD45 KO, CD8-FITC/CD4-PE/CD3-ECD Antibody cocktail, HLA-DR-PE-Cy5, CD19 PC7, CD 326 (EpCAM) APC, CD56 APC Alexa Fluor 700, CD 14 APC Alexa Fluor 750, and CD16-PB. All the antibodies were obtained from Beckman Coulter, Milan, Italy. For the evaluation of T-reg, we used Duraclone IM T-reg tube (B53346, Beckman Coulter). According to the Beckman-Coulter instruction for the FMO controls, we used VersaComp antibody capture beads in association with the single Compensation Kit provided in the DuraClone IM Treg kit. These beads allow the positive and negative signals associated with the background to be identified separately. The CytoFLEX instrument was used to perform the experiment. CytExpert software 2.4 (Beckman Coulter, Brea, CA, USA) was used to obtain the data. Kaluza Analysis software was used to analyze and characterize the immune cells obtained from the cytometry data.

Statistical Analysis
All statistical analyses were conducted using GraphPad 6 software. A value of p < 0.05 was considered to indicate a statistically significant difference. All experiments were repeated three times.

Breast Tissue Morphological Evaluation
Morphological analysis of H&E-stained breast tissue slides allowed us to define the degree of TILs, as presented in Figure 1 and Supplementary Figure S1. An experienced pathologist reviewed the slides and evaluated the lymphocyte infiltration into absent, low, and moderate. However, H&E analysis cannot define specific lymphoid subpopulations unless specific immunostaining protocols are applied [28]. Therefore, to better define the immune cell components residing in the breast tissue, we decided to exploit a 10-color flow cytometry approach described hereafter. degree of TILs, as presented in Figure 1 and Supplementary Figure S1. An experienced pathologist reviewed the slides and evaluated the lymphocyte infiltration into absent low, and moderate. However, H&E analysis cannot define specific lymphoid subpopulations unless specific immunostaining protocols are applied [28]. Therefore, t better define the immune cell components residing in the breast tissue, we decided t exploit a 10-color flow cytometry approach described hereafter.

Analysis of Breast-Infiltrating Immune Cells by Multicolor Flow Cytometry
The FCM analysis protocol ( Figure 2 and Supplementary Figure S2) was conceived to discriminate between immune cells (CD45 pos EpCAM neg ) and epithelial cell (CD45 neg EpCAM pos ) on 15 and 14 biopsies from malignant and benign breast tissues respectively. Live cells were selected according to the physical parameters FSC and SSC and then single cells were identified in the FSC-H vs. FSC-A dot plot (Figure 2A,B, lef plot). Immune cells and epithelial cells are then identified in the CD45 vs. EpCAM do plot (Figure 2A,B, right plot). We observed that the percentage of immune cells wa significantly higher in tissues obtained from patients' breast fibroadenomas (* p < 0.05 compared with breast cancer patients ( Figure 2C, left plot). Conversely, no significan

Analysis of Breast-Infiltrating Immune Cells by Multicolor Flow Cytometry
The FCM analysis protocol ( Figure 2 and Supplementary Figure S2) was conceived to discriminate between immune cells (CD45 pos EpCAM neg ) and epithelial cells (CD45 neg EpCAM pos ) on 15 and 14 biopsies from malignant and benign breast tissues, respectively. Live cells were selected according to the physical parameters FSC and SSC, and then single cells were identified in the FSC-H vs. FSC-A dot plot (Figure 2A,B, left plot). Immune cells and epithelial cells are then identified in the CD45 vs. EpCAM dot plot (Figure 2A,B, right plot). We observed that the percentage of immune cells was significantly higher in tissues obtained from patients' breast fibroadenomas (* p < 0.05) compared with breast cancer patients ( Figure 2C, left plot). Conversely, no significant differences were observed regarding epithelial cells ( Figure 2C, right plot). We went on to perform a deep-immunophenotyping of the CD45+ cells identified through the evaluation of the T-, B-and NK lymphocytes, as well as the monocyte subpopulations.

T Lymphocytes
T lymphocytes were identified as CD45 pos CD3 pos ( Figure 3A-D) events and were significantly increased (**** p < 0.0001) in tumor lesions when compared to fibroadenomas ( Figure 4A and Table 2). Furthermore, we defined the percentages of helper and cytotoxic T cells according to the CD4 and CD8 expression (Figures 3E and 4B,C and Table 2). We found that CD45 pos CD3 pos CD4 pos helper T cells were significantly higher in malignant breast lesions ( Figure 4B and Table 2). In contrast, CD45 pos CD3 pos CD8 pos cytotoxic T cells disclosed an opposite trend: the CD4/CD8 ratio is significantly unbalanced in the direction of breast cancer patients ( Figure 4D and Table 2). Moreover, the percentage of activated CD4 T lymphocytes, which was determined by evaluating those expressing the MHC class 2 membrane receptor (the Human Leukocyte Antigen DR isotype-HLA-DR,), is observed to be significantly less active in the tumor microenvironment than those present in fibroadenomas (Figures 3F,G and 4E and Table 2). This finding suggests that the tumor microenvironment may affect the ability of CD4+ T lymphocytes to activate and attract cytotoxic components to the tumor site, leading to the inability of the anti-cancer immune response to function properly.
Cancers 2022, 14, 3869 6 of 18 differences were observed regarding epithelial cells ( Figure 2C, right plot). We went on to perform a deep-immunophenotyping of the CD45+ cells identified through the evaluation of the T-, B-and NK lymphocytes, as well as the monocyte subpopulations. of breast cancer patients ( Figure 4D and Table 2). Moreover, the percentage of activated CD4 T lymphocytes, which was determined by evaluating those expressing the MHC class 2 membrane receptor (the Human Leukocyte Antigen DR isotype-HLA-DR,), is observed to be significantly less active in the tumor microenvironment than those present in fibroadenomas (Figures 3F,G and 4E and Table 2). This finding suggests that the tumor microenvironment may affect the ability of CD4+ T lymphocytes to activate and attract cytotoxic components to the tumor site, leading to the inability of the anti-cancer immune response to function properly.

T-Reg
T-reg cells have been associated with the clinical outcome of breast cancer patients [29,30]. For these reasons, we analyzed the Treg populations in breast cancer tissues and in the healthy counterpart of the same subject. As shown in Figure 5, CD45 pos /CD3 pos /CD4 pos / FoxP3 pos T-reg cells are significantly higher in tumor tissues with respect to the normal tissues ( Figure 5A,B,E). Moreover, the CD45 pos /CD3 pos /CD4 pos /FoxP3 pos /Helios pos T-reg subpopulation was evaluated. Figure 5C,D,F clearly show an increase in abundance of this T-reg in breast cancer tissues in comparison with the healthy breast parenchyma.

NK Lymphocytes
NK lymphocytes were selected according to the gating strategy shown in Figure 6. Both the NK cells (CD56 brigth CD16 pos ) and cytotoxic NK cells (CD56 dim CD16 pos ) were not significantly different between the two study groups ( Figure 6A,B) ( Figure 6A,B and Table 3). Conversely, when considering the NKT cells (CD56 pos /CD3 pos , Figure 6C), we found a significantly higher presence in the breast cancer patients than in the fibroadenomas (Figure 7A-C and Table 3). Similarly, type 1 NKT cells (CD56 pos /CD3 pos /CD4 pos , Figure 6D) are also significantly more present in subjects with tumors than in those with fibroadenoma ( Figure 7D and Table 3). The hyper-activation of NK cells and NK-T cells was also significantly higher in patients with tumors than in those with fibroadenomas ( Figures 6E,F and 7E,F and Table 3). In contrast, no variations in the percentage of B lymphocytes were observed between subjects with breast cancer and those with fibroadenoma (Supplementary Figure S3).

Monocytes
Finally, the tumor-associated monocyte component was analyzed in detail compared with fibroadenoma. First of all, the lymphocyte-to-monocyte ratio was calculated by selecting all lymphocytes as CD45 pos cells and monocytes as CD14 pos cells ( Figure 8A). As can be seen from Figure 9A and Table 4, the ratio is significantly higher in subjects with breast cancer, a fact also confirmed by the percentage of monocytes identified, which is significantly higher in fibroadenomas than in tumors ( Figure 9B and Table 4). No significant differences were found in the percentage of atypical CD14 pos /CD16 pos (Figures 8A and 9C and Table 4) monocytes, or in the degree of monocyte activation (Figures 8B and 9D  and Table 4).  significantly higher presence in the breast cancer patients than in the fibroadenomas (Figure 7A-C and Table 3). Similarly, type 1 NKT cells (CD56 pos /CD3 pos /CD4 pos , Figure 6D) are also significantly more present in subjects with tumors than in those with fibroadenoma ( Figure 7D and Table 3). The hyper-activation of NK cells and NK-T cells was also significantly higher in patients with tumors than in those with fibroadenomas ( Figures 6E,F and 7E,F and Table 3). In contrast, no variations in the percentage of B lymphocytes were observed between subjects with breast cancer and those with fibroadenoma (Supplementary Figure S3).

Monocytes
Finally, the tumor-associated monocyte component was analyzed in detail compared with fibroadenoma. First of all, the lymphocyte-to-monocyte ratio was calculated by selecting all lymphocytes as CD45 pos cells and monocytes as CD14 pos cells ( Figure 8A). As can be seen from Figure 9A and Table 4, the ratio is significantly higher in subjects with breast cancer, a fact also confirmed by the percentage of monocytes identified, which is significantly higher in fibroadenomas than in tumors ( Figure 9B and Table 4). No significant differences were found in the percentage of atypical CD14 pos /CD16 pos (Figures 8A and  9C and Table 4) monocytes, or in the degree of monocyte activation ( Figures 8B and 9D and Table 4).

Discussion
In physiological conditions, immune cells maintain breast tissue homeostasis by continuous immunosurveillance and play a central role in initiating inflammatory reactions [31]. In the case of BC, immune cells invade the cancer tissue, and infiltrating lymphocytes play a central in cancer growth, so that the percentage of TILs is currently accepted as a useful prognostic factor for managing breast cancer patients. Histological methods (hematoxylin-eosin or immunohistochemistry staining) are used to evaluate the TILs percentage in paraffin-embedded BC tissue sections [28,32]. Although the assessment is specific to the tumor region, it does not consider the different subpopulations infiltrating the tumor, unless time-consuming and rather complex analyses are made (i.e., IHC with specific markers for each subpopulation). Identifying which subpopulations are altered in the tumor microenvironment would help design targeted intervention strategies based on immunotherapy, which has been gaining importance in the treatment of neoplasms [33]. In this scenario, our manuscript aims at studying the tumor-associated lymphocyte subpopulations in comparison to those associated with benign formations (fibroadenomas) by a multiparametric FCM approach. We performed a deep-immunophenotyping of CD45 pos cells through the evaluation of immune cells, with particular reference to the lymphocyte subpopulations and monocytes. According to recent observations of our research group

Discussion
In physiological conditions, immune cells maintain breast tissue homeostasis by continuous immunosurveillance and play a central role in initiating inflammatory reactions [31]. In the case of BC, immune cells invade the cancer tissue, and infiltrating lymphocytes play a central in cancer growth, so that the percentage of TILs is currently accepted as a useful prognostic factor for managing breast cancer patients. Histological methods (hematoxylineosin or immunohistochemistry staining) are used to evaluate the TILs percentage in paraffin-embedded BC tissue sections [28,32]. Although the assessment is specific to the tumor region, it does not consider the different subpopulations infiltrating the tumor, unless time-consuming and rather complex analyses are made (i.e., IHC with specific markers for each subpopulation). Identifying which subpopulations are altered in the tumor microenvironment would help design targeted intervention strategies based on immunotherapy, which has been gaining importance in the treatment of neoplasms [33]. In this scenario, our manuscript aims at studying the tumor-associated lymphocyte subpopulations in comparison to those associated with benign formations (fibroadenomas) by a multiparametric FCM approach. We performed a deep-immunophenotyping of CD45 pos cells through the evaluation of immune cells, with particular reference to the lymphocyte subpopulations and monocytes. According to recent observations of our research group in peripheral blood [34], we found that in the case of BC onset the tumor microenvironment featured by a higher percentage of CD4 pos T cells in comparison to fibroadenoma, as previously reported [35][36][37]. However, the percentage of activated CD4+ T cells was lower in case of BC malignancies. In addition, considering that CD4 pos T cells primarily mediate antitumor immunity by stimulating CD8 pos [38,39], it is not surprising that our data highlighted that the percentage of cytotoxic CD8 pos T cells was significantly diminished in malignant tissues. In order to assess the presence of T-reg cells in the tumor microenvironment, we performed a complex comparative analysis between the microenvironment of the tumoral lesion and that of disease-free breast tissue from the same subject. T-reg cells are rare lymphocyte subpopulations. A large number of events must be used for their characterization at the FCM. For this reason, we could determine the T-reg subpopulations only in a limited number of subjects (three tumor tissues vs. three healthy counterparts). Although preliminary, given the number of subjects analyzed, our data show that the T-reg CD45 pos /CD3 pos /CD4 pos /FoxP3 pos and CD45 pos /CD3 pos /CD4 pos /FoxP3 pos /Helios pos subpopulations are a more stable immunosuppression phenotype at the level of the tumor microenvironment and more abundant in the tumor microenvironment than in the non-diseased microenvironment, in accordance with data in the literature [30]. This evaluation shows that our approach can also be used to determine the rare subpopulations in tumor tissues.
The ability of malignant transformations to promote the onset of a cancer immunotolerance was confirmed when analyzing the NK sub-settings. These cells contribute to various immune functions during cancer initiation and progression that can change in quality and magnitude depending on the disease stage [40][41][42][43]. Our data indicate that there are no differences in the percentage of NK cells or cytotoxic NK cells (CD56 pos /CD16 pos ) in breast cancer patients compared with fibroadenomas. On the contrary, a significant increase in NK-T cells and type I NK cells (CD56 pos /CD4 pos ) occur in the tumor tissues. Surprisingly, NK cells and NK-T cells turn out to be more activated by the tumor microenvironment. The synergistic effect of the presence/activation of NK cells, and in particular of NK-T cells, could be in line with data in the literature, in which these effects determine the amplification or dampening of the immune response. Moreover, the hyperstimulation of NK cells during tumor progression could be associated with anergy and/or skewing of the anti-cancer immunity, which could facilitate the tumor progression and immune escape [44]. Finally, we observed a significant reduction in the percentage of monocytes in tumor tissues, resulting in an imbalance in the lymphocyte/monocyte ratio, significantly higher in subjects with tumors than in those with fibroadenomas. The tumor microenvironment reprograms monocytes by reducing their ability to respond to stimuli and assuming immunosuppressive activity [45]. Furthermore, assessing the ratio of lymphocytes to monocytes is a prognostic factor that is becoming increasingly important in the clinical management of breast cancer patients, since it is associated with a poor prognosis [46]. Furthermore, there are still unclear indications on some immune components. Conflicting current scientific data suggests that CD 19+ B cells in breast cancer tissues may have positive [47], negative [48], or no significant performance [49] in humoral and cellular immunity. Furthermore, there are potential interconnections to be explored between CD19+ B cells and other lymphocytes that make the tissue microenvironment favorable or hostile to tumor progression [50].
Our approach aims to evaluate the specific subpopulations of the immune system involved in the anti-cancer response using flow cytometry. In association with normal clinical practice performed by pathological anatomy, this analysis could shed light on which cell types of the immune system are altered by the tumor microenvironment. Our approach is intended to support normal clinical practice because it could augment the information provided by routine investigations. In fact, the evaluation of TILs carried out by the pathologists provides a precise focus on the tumor portion, allowing identification of the quantity and/or quality of lymphocytes infiltrating the tumor [18,51], but does not provide a general view of the entire tissue fragment. On the contrary, our method, although it loses the ability to focus only on the region invaded by the tumor cells, provides an overall view of the entire tissue, allowing, with the same analysis, to evaluate many more parameters than a pathological anatomy investigation. In this way, truly targeted and personalized immunotherapy approaches could be developed to improve the management of breast cancer patients. For human breast tissue, isolation of single cells typically involves long mechanical and enzymatic dissociation, where the majority of the well-established protocols rely on a 16-24-h enzymatic dissociation, with varying concentrations of collagenase and/or hyaluronidase [25,27,52]. It is thus important to underline a limitation of our study. During the digestion processes, exposure to enzymatic action can lead to the disruption of cell surface markers, induction of apoptosis, and, in the worst case, a complete loss of certain immune cell subtypes [53,54]. Furthermore, comparing different enzymatic dissociation protocols, the focal technical variables, such as agitation speed, duration, and temperature, can affect the composition of the isolated cells, until inducing transcriptional changes that represents an important aspect to evaluate in gene expression studies sensitive to excessive stress caused by digestion conditions. Despite the limitations mentioned for the study of breast tumoral tissues, our approach could also be extended to other malignancies for which an understanding of the involvement of the anti-cancer immune response in the onset of the disease is still complicated. Finally, in the future, in order to better study the tumor microenvironment, other lymphocyte subpopulations (PD1, PD-L1, CTLA4, etc.) also need to be taken into account, to identify personalized approaches that are not only diagnostic but also therapeutic.

Conclusions
The study of the breast cancer microenvironment is important in the clinical management of breast cancer patients. Routine clinical practice is based on a qualitative assessment of TILs through the involvement of pathologists. Despite the innovative approaches available, such as immuno-specific IHC to study lymphocyte subpopulations, or machine learning approaches to analyze the entire histological slide, it is necessary to ensure an increasingly personalized approach to patient care. For this reason, a deepimmunophenotyping analysis performed by flow cytometry, directly on tissue recovered after biopsy, would open the way to new diagnostic and/or therapeutic approaches that could support the normal clinical routine in order to improve the management of breast cancer patients.
Author Contributions: L.C. conceived the study and contributed to the experimental and data analysis and drafting of the manuscript; G.S. contributed to the experimental activity, data analysis and drafting of the manuscript; M.D.: breast surgeon, provided the tissue samples, clinical data and reviewed the manuscript; G.D.: breast surgeon, provided the clinical data and reviewed the manuscript; G.M.: pathologist, contributed to the supply of tissue samples, patient reports, to the supply of hematoxylin-eosin slides and to the evaluation of lymphocyte infiltrates; M.R.: breast surgeon, provided the clinical data and reviewed the manuscript; S.V. contributed to the processing of biological samples and participated in the experimental activity; E.N. contributed to the revision of the manuscript; M.S. contributed to the direction general organization of the research activities, reviewed the manuscript, and gave the approval to the submission; P.M. conceived the study with L.C. and contributed to the data analysis and drafting of the manuscript. All authors have read and agreed to the published version of the manuscript.
Funding: This research was funded by "Progetti di Ricerca Corrente", funded by the Italian Ministry of Health.

Institutional Review Board Statement:
The study was conducted in accordance with the Declaration of Helsinki and approved by the Ethics Committee of IRCCS Pascale (Naples, Italy) (reference number 3/19 on 29 May 2019).

Informed Consent Statement:
Informed consent was obtained from all subjects involved in the study.