Introduction
Epithelial-derived tumors primarily thrive in a dynamic stroma composed of stromal cells, immune cells, and soluble cytokines. The tumor microenvironment (TME) supplies all the necessary incentives for tumor survival, growth, and aggressiveness [
1]. Macrophages are a heterogeneous group of cells in the immune system that play an important role in the body defense. Tumor-associated macrophages (TAMs) are macrophages that infiltrate into TME. In breast cancer (BC), TAMs account for more than 50% of the tumor volume and are the most abundant immune cell type in TME [
2]. The number of TAMs is associated with poor prognosis in BC [
3]. TAMs act in nearly the whole stage of tumor progression. Currently, two activation states of macrophages with different polarizations have been identified: classical activated (M1) and alternatively activated (M2) macrophages [
4]. In the early stages of tumor development, TAMs usually exhibit an M1 phenotype with high expression of iNOS and interleukin (IL)-12, and low expression of CD206, Arg-1 and IL-10. M1 macrophages enhance the immune response to recognize and destroy cancer cells through phagocytosis and cytotoxicity, presenting anti-tumor effects. In the advanced stage of tumor progression and metastasis, TAMs usually favor the M2 phenotype, which is characterized by low expression of iNOS and IL-12, and high expression of CD206, Arg-1 and IL-10, promoting tissue repair and angiogenesis in favor of tumor progression [
5,
6]. M2-type TAMs produce anti-inflammatory factors that encourage immune escape of cancer cells and further contribute to tumor progression [
7].
Adipocytes are the cell type with the largest proportion in the mesenchymal stroma of BC. Studies have shown that TAMs play an essential role in the interplay between adipocytes and BC cells [
8]. Adipocytes in the vicinity of BC tissue can be converted into cancer-associated adipocytes (CAAs), which promote cancer cell proliferation, enhance angiogenesis and change the extracellular matrix by secreting various cytokines such as IL-6, hepatocyte growth factor (HGF) and chemokine (C–C motif) ligand 2 (CCL2), playing an active role in the process of tumorigenesis and progression [
9]. Among them, CCL2 is a common macrophage chemokines and induces M2-type macrophage differentiation, thus promoting BC progression and metastasis [
10]. In addition, CAAs can secrete lactate and adenosine accumulated in the TME, and these metabolites have been shown to further induce macrophage to polarize towards M2-type [
11,
12].
IL-6 mediates chronic inflammation and provides a favorable microenvironment for tumor growth. Studies have shown that circulating levels of IL-6 are correlated with the aggressive characteristics of BC patients and could lead to a worse prognosis in BC patients [
13]. It was reported that IL-6 promoted the polarization of monocytes into M2-type macrophages, further enhancing the invasiveness of BC [
14]. IL-6 is a strong activator of STAT3. When IL-6 binds to IL-6R and the co-receptor gp130, it activates STAT3 and the activated p-STAT3 is rapidly transferred to the nucleus, thereby activating the inflammatory cascade and oncogenic pathways [
15]. STAT3 was proven to induce polarization of M2-type macrophages in ovarian cancer [
16]. In gastric cancer, IL-6/STAT3 signaling could promote M2 macrophage differentiation [
17]. IL-6-dependent activation of STAT3 is of importance in the progression of multiple tumors, including BC.
Recently, attention has focused on the function of CAAs in regulating immune cell recruitment. Studies on the role and mechanism of CAAs on monocytes/macrophages are not clear. Given the critical role of CAAs and TAMs in determining tumorigenesis and metastasis of cancer, we explored the interaction between CAAs, macrophages, and BC cells by utilizing a mouse BC cell line, a mouse subcutaneous tumor-bearing model and human BC tissue specimens. Our study aimed to determine the ability and the underlying mechanisms of CAAs in recruiting and polarizing macrophages to BC TME compared to normal mature adipocytes and BC cells, further affecting BC malignancy.
Materials and methods
Clinical BC samples
The 11 human BC samples from patients diagnosed with invasive BC undergoing mastectomy at Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology were collected for hematoxylin/eosin (H&E) and immunohistochemistry (IHC) staining. The Image J software (National Institute of Health, MD, USA) was used to measure the size of adipocytes in H&E-stained sections. The clinicopathological details of the 11 BC patients were displayed in Table S
1.
Cell culture and preparation of CAAs
4T1 cells and RAW 264.7 cells were collected from the Cell Bank of Chinese Academy of Sciences (Shanghai, China). 3T3-L1 preadipocytes, MDA-MB-231 and MCF-7 cells were purchased from Procell Life Science & Technology (Wuhan, China). Breast cancer cells and 3T3-L1 preadipocytes were cultivated in DMEM-H medium with 10% fetal bovine serum (FBS, Gibco, USA). RAW 264.7 cells were cultured in MEM-α medium with 10% fetal bovine serum. Through adipogenic differentiation of 3T3-L1 preadipocytes, mature adipocytes were obtained and identified by Oil Red O staining of the lipid droplets (Servicebio, Wuhan, China) [
18]. Then, CAAs were acquired by co-culture of the mature adipocytes (upper chamber) with 4T1 cells (lower chamber) in a transwell co-culture system (pore size 0.4 μm, Corning, USA) for 3 days.
Acquisition of conditioned medium
CAAs, mature adipocytes, and 4T1 cells were cultured in an FBS-free medium for 24 h to collect the CAA-conditioned medium (CM), AD-CM and 4T1-CM, respectively. Then, the collected CM was centrifuged at 3000 rpm for 10 min. To generate CAA-edu-RAW CM, AD-edu-RAW CM and 4T1-edu-RAW CM, RAW 264.7 cells were first cultured in CAA-CM, AD-CM and 4T1-CM for 72 h, respectively. Then, the CM was changed into FBS-free medium for 24 h. The FBS-free medium was collected as CAA-edu-RAW CM, AD-edu-RAW CM and 4T1-edu-RAW CM, respectively, with centrifugation at 3000 rpm for 10 min. 1% FBS was added to the CM in the following experiments.
Proliferation and wound-healing assay
The cell counting kit (CCK-8; Yeasen, Shanghai, China) was used to assess cell proliferation according to the manufacturer’s instructions. Breast cancer cells were seeded at 5 × 103/well in 96-well plates and cultivated in different CM. Then, each well was added with CCK-8 reagent at 0,12, 24 and 36 h. The enzyme-labeled instrument (Bio Tek, VT, USA) was utilized to test the optical density (OD) at 450 nm.
Breast cancer cells were inoculated at 1 × 106/well in a 24-well plate and cultured to confluence. The10 μL pipette was used to scratch a straight line in the cells of each well. Then, cells were cultured with or without CM for 24 h. The Image J software was employed for analyze the wound healing condition of 5 different fields obtained by a microscope (SOPTOP CX40, Ningbo, China).
Transwell migration and invasion assay
A total of 5 × 104 4T1 cells were cultivated in the upper chamber of the transwell system (pore size 8 μm, Corning, USA) in a 24-well plate for the transwell migration assay. The lower chamber of the transwell system was added with 500 μL CM. Then, the migrated 4T1 cells were dyed with crystal violet after incubation in the transwell system for 48 h. A microscope (SOPTOP CX40, Ningbo, China) was used to acquire the photographs of the migrated cells. The 33% glacial acetic acid was employed to wash the dyed cells. Finally, the OD values of each well were taken by the enzyme-labeled instrument at 590 nm.
As for the transwell invasion assay, the transwell chambers were coated with Matrigel (R&D system, MN, USA) of 3 mg/mL first. Then, a total of 2 × 105 4T1 cells were cultured in the upper chamber of the transwell system (pore size 8 μm, Corning, USA) in a 24-well plate. The protocols of the following steps were in accordance with the transwell migration assay.
Quantitative real-time PCR (qRT-PCR)
TRIzol (Takara, Shiga, Japan) was applied to lyse the cultured cells. Total RNAs were extracted based on the manufacturer’s protocol of the 1st Strand cDNA Synthesis Kit (Yeasen, Shanghai, China) for subsequent cDNA synthesis. QuantStudio1 (ABI Q1, CA, USA) and SYBR GreenTM Master Mix (Yeasen, Shanghai, China) were employed for qRT-PCR. The main primer sequences applied for qRT-PCR were displayed in Table S
2.
Western blotting assay
IP lysis buffer with PMSF (Servicebio, Wuhan, China) was used to lyse cultured cells for western blotting assay. Then, protein quantification of the lysates was performed with the BCA Protein Assay Kit (Servicebio, Wuhan, China). Total proteins were first isolated on SDS-PAGE and the SDS-PAGE was cut according to protein molecular weight. And then, proteins were transferred onto the PVDF membranes (0.22 μm; Millipore, MA, USA). The PVDF membranes were blocked in 5% BSA and incubated at 4 °C overnight in specific primary antibodies. Subsequently, the membranes were incubated in HRP-conjugated secondary antibodies for 1 h at room temperature. Finally, ECL chemiluminescent reagents (Yeasen, Shanghai, China) were utilized to detect the protein bands. The original western blotting images were included in the Additional file
5 of the supplementary materials. The main antibodies applied in the western blotting assay were listed in Table S
3.
Animal models
BALB/C female mice of 6-8w purchased from Beijing Vital River Laboratory Animal Technology (Beijing, China) were involved in this study. All mice were bred at 21–25 ℃ and at 12/12 h of alternating light and dark. The animal experiments were performed with the approval of the Institutional Animal Care and Use Committee, Huazhong University of Science and Technology.
For constructing the subcutaneous tumor-bearing model, ten mice were randomly divided into two groups: the CAAs group and the control group. A total of 5 × 105 4T1 cells and 2.5 × 106 adipocytes were injected into the axilla of the mouse subcutaneously in the CAAs group. A total of 5 × 105 4T1 cells were injected in the same location of the mouse in the control group. Mice were monitored and weighed every other day starting on day 1 after inoculation. The volume of the tumor was measured every 3 days when small and hard tumor nodules could be palpated. Four weeks after inoculation, mice were executed to obtain tumor tissues. Tumor tissues were weighed and were fixed in 4% paraformaldehyde and embedded in paraffin for following experiments.
For the tail vein pulmonary metastasis model, 4T1 cells were pre-cultured with CAA-CM or the control medium for 48 h and were injected into the tail vein of each mouse with 5 mice per group. After 14 days, all mice were sacrificed to isolate the lungs. Then, the lungs were embedded in paraffin for H&E staining to examine the pulmonary metastasis condition of each group.
H&E, IHC and immunofluorescence (IF) assay
For the H&E staining assay, sample tissues were first fixed in 4% paraformaldehyde and were embedded in paraffin. Then, paraffin sections were cut and stained with hematoxylin and eosin. Representative photographs of H&E-stained sections were captured by the microscope (SOPTOP CX40, Ningbo, China).
For IHC assay, sections were first dissociated in xylene, and rehydrated in ethanol solutions with different concentrations. Sections were then heated in citrate buffer to recover the antigen and were blocked in 5% BSA. they were incubated with primary antibodies overnight at 4 °C, followed by HRP-conjugated secondary antibodies at room temperature for 30 min. Then the immunoreactivity of sections was visualized using DAB and counterstained with hematoxylin. Photographs of the sections were taken with the microscope (SOPTOP CX40, Ningbo, China).
For IF staining, the procedures prior to incubation of the secondary antibodies were consistent with IHC assay. Subsequently, sections were incubated in the fluorescently labeled secondary antibodies at room temperature for 1 h. The nuclei were then re-stained by nuclear 4,6-diamidino-2-phenylindole (DAPI). Photographs of the sections were taken by the fluorescence microscope (Olympus, Japan). The main antibodies used in the IHC and IF assay were involved in Table S
3.
Enzyme-linked immunosorbent assay (ELISA)
According to the manufacturer's introduction, mouse IL-6 ELISA kits (Invitrogen, CA, USA) were used to measure the secretion levels of IL-6 in mouse serum of CAAs group and the control group, and CAA-CM and the control medium.
Statistical analyses
The statistical analyses of the experimental data were accomplished by Graphpad software (version 9.0, CA, United States). Data were obtained from at least three replicate assays and presented as mean ± standard deviation (means ± SD). Student’s-t test was used to compare the differences between two independent samples. One Way ANOVA was employed to analyze the differences among multiple groups. P < 0.05 was considered statistically significant.
Discussion
The continuous interaction between tumor cells and immune cells in the TME has an important role in tumor progression and metastasis. In addition, there are also interactions between the stromal cells and immune cells in the TME. In recent years, adipocytes at the invasive front of the tumor have been suggested to be associated with the recruitment and functional regulation of immune cells, thus influencing tumor cell behaviors [
10]. In the present study, it was found that in human BC samples, adipocytes located at the invasive front of the tumor were smaller in size, and adipocytes underwent de-differentiated phenotypic changes into CAAs when co-cultured with BC cells. In vitro cellular assays revealed that CAA-CM could enhance the malignant behaviors of 4T1 cells. In the mouse model, CAAs could accelerate tumor growth and pulmonary metastasis. We also demonstrated that macrophages cultured with CAAs could further promote the malignant behaviors of 4T1 cells. Furthermore, we uncovered a potential mechanism that CAA-derived IL-6 could induce M2 macrophage polarization by activating STAT3 signaling.
CAAs have been first put forward by Diart et al. upon co-culture of murine 3T3-F442A adipocytes with tumor cells [
22]. The research indicated that CAAs displayed a modified phenotype including a decrease in lipid content, a reduction in adipocyte markers (C/EBPα), and presented an activated condition with the characteristics such as overexpression of inflammatory cytokines (IL-6 and IL-1β) and proteases (matrix metalloproteinase-11) [
22]. In turn, CAAs altered the features of tumor cells into a more aggressive phenotype [
22]. Subsequently, Fujisaki et al. confirmed the findings of Dirat et al. in CAAs which were derived from BC patients, and further found that CAAs decreased in size and exhibited an immature and proliferative phenotype in the presence of cancer cells, and contributed to cancer cell migration via adipokines including IL- 6 and MCP-1 [
23]. The morphometric study of adipocytes on BC verified that peritumoral adipocytes were smaller when compared with adipocytes of the normal tissues through an image analysis with photonic microscopy [
24]. In our study, the phenotypic alteration of adipocytes with BC cells in vitro and in human BC samples were consistent with the characteristics of CAAs in the previous studies. Our study also indicated that CAAs could enhance the malignant behaviors BC cells both in vitro and in vivo. So far, studies on the interaction between CAAs and BC cells have mainly focused on the cellular level. In the present study, through co-injecting adipocytes and BC cells subcutaneously, we constructed a mouse tumor-bearing model and demonstrated that CAAs could promote breast tumor growth and malignant progression in vivo.
CAAs are important sources of adipokines, chemokines, cytokines, and exosomes that could promote tumor growth and progression [
9]. Prior studies have noted the ability of CAAs in secreting high abundance of IL-6, IL-1β, CCL2, chemokine (C–C motif) ligand 5 (CCL5), tumor necrosis factor-α (TNF-α), vascular endothelial growth factor (VEGF) and leptin, which could enhance tumor invasion and immune escape [
10,
22]. Increase secretion of IL-6 is one of the important phenotypic alterations of CAAs. Excessive secretion of IL-6 by tumor cells or mesenchymal cells in TME has been demonstrated to promote tumor growth, metastasis, and therapeutic resistance in multiple tumors, including BC [
25,
26]. In the immune system, IL-6 has a dual effect: a pro-inflammatory or an anti-inflammatory factor, mainly depending on the local immune microenvironment [
17,
27]. In some cases, IL-6 could induce M1 macrophage. It has been suggested that IL-6 induced macrophage expressed M1 macrophage marker, leading to an inflammatory TME thereby enhancing radiosensitivity in HPV
+ head and neck cancer [
28]. However, there is controversy regarding the role of IL-6 in inducing M1/M2 macrophage polarization. Braune et al. indicated that IL-6 in obesity adipose tissue acted as a Th2 cytokine by stimulating M2 macrophage polarization [
29]. In addition, Weng et al. found that in triple-negative breast cancer (TNBC) cells, IL-6 stimulated by MCT-1 promoted monocytic THP-1 polarizing into M2 macrophages to enhance TNBC cell invasiveness in IL-6/IL-6R signaling [
14]. The function of IL-6 in regulating macrophage polarization was complex and might be influenced by the local microenvironment.Here, we found that CAA-derived IL-6 could induce M2 macrophage polarization in vitro, which was in agreement with IL-6 involvement in M2 polarization in TNBC TME.
Macrophages are critical components of the host defense system [
30]. So far, two polarization states of macrophages have been identified, including the pro-inflammatory classical activated (M1) macrophages and the anti-inflammatory alternative activated (M2) macrophages [
31]. TAMs usually exhibited an M2 phenotype and promoted tumor progression by stimulating immunosuppression [
31]. The activated state of TAMs was transient, leading to the functional plasticity of TAMs [
1]. It is crucial to understand the role of macrophage heterogeneity and plasticity in the pro-tumor progression of CAAs. Adipocytes have been confirmed to affect macrophage polarization in adipose tissue and TME [
32]. It has been recently reported that the uptake and oxidation of FFAs, released by abnormal catabolism of CAAs, were correlated with anti-inflammatory, immunosuppressive, and pro-tumor polarization of M2 macrophages [
33]. CAA-secreted adenosine, accumulating in the tumor-associated adipose microenvironment, could reduce the classical polarization of macrophages or induce M2-type polarization and promote monocyte recruitment into tumors when it is binded to A2A or A2B receptors [
11,
34]. In addition, lactate, the aerobic glycolytic end product of CAAs, regulated the polarization state of M2 macrophages through ERK/STAT3 signaling [
12]. These results corroborated the findings of our study that CAAs were important participants in regulating M2 macrophage polarization. In this study, CAA-induced CD206 upregulation was found both in human samples and at the cellular level. However, the immunomodulatory function of CD206 has not been fully elucidated. The current study indicated that CD206 deficiency led to upregulation of the pro-inflammatory cytokine in serum mouse model, suggesting that CD206 might possess anti-inflammatory effects [
35]. Based on this, CAA-induced increased expression of CD206 in macrophages with anti-inflammatory effects, which might be an indirect immunosuppressive mechanism of CAAs. It was reported that CD206 was more related to M2a macrophages, mediating TH2-type immune response [
36]. Whereas, M2c macrophages, mainly expressed CD163, were more associated with the immune regulation and tissue remodeling [
35,
37]. Therefore, further studies using CD163 as a marker are recommended to determine whether CAAs are more likely to induce immunosuppressation-associated M2c macrophages.
In TNBC, TAMs promote cancer growth and progression through multiple mechanisms and can directly and indirectly regulate PD-1/PD-L1 expression in TME [
30]. The study of Zhang et al. indicated that in hepatocellular carcinoma, serum IL-6 could upregulate PD-L1 expression in macrophages, which in turn caused immunosuppression in an orthotopic tumor transplantation model [
38]. Similarly, in the present study, we found that CAAs could upregulate PD-L1 expression in CAA-induced macrophages, and the upregulated PD-L1 might further enhance immunosuppression in TME of BC. There is abundant room for further progress in determining the complex mechanisms among CAAs, macrophages and BC cells. In addition to the soluble factor-mediated cell interaction in the present study, further research should be undertaken to investigate whether CAAs could directly induce immunosuppressive TME and inhibit T lymphocyte proliferation. In addition, it is necessary to study the correlation between the number of M2 macrophages and CAAs in human BC samples to determine whether CAAs and M2 macrophages coexist and thus assess the potential prognosis of BC.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.