Understanding the mechanisms underlying obesity in remodeling the breast tumor immune microenvironment: from the perspective of inflammation

Obesity is a well-known modifiable risk factor for breast cancer and is considered a poor prognostic factor in pre- and post-menopausal women. While the systemic effects of obesity have been extensively studied, less is known about the mechanisms underlying obesity-associated cancer risk and the local consequences of obesity. Thus, obesity-induced inflammation has become the focus of research interest. Biologically, the development of cancer involves a complex interaction with numerous components. As the tumor immune microenvironment changes due to obesity-triggered inflammation, an increase in infiltration occurs for proinflammatory cytokines and adipokines, as well as adipocytes, immune cells, and tumor cells in the expanded adipose tissue. Complicated cellular-molecular crosstalk networks change critical pathways, mediate metabolic and immune function reprogramming, and have a significant role in tumor metastasis, proliferation, resistance, angiogenesis, and tumorigenesis. This review summarizes recent research findings on how inflammatory mediators in the in situ tumor microenvironment regulate the occurrence and development of breast cancer in the context of obesity. We analyzed the heterogeneity and potential mechanisms of the breast cancer immune microenvironment from the perspective of inflammation to provide a reference for the clinical transformation of precision targeted cancer therapy.


Introduction
Cancer accounts for 21% of all deaths, making it a major public health problem worldwide. Breast cancer (BC) is the most common malignancy, the incidence of which is increasing. Specifically, BC comprised approximately one-third of newly-diagnosed tumors among women in 2022 1,2 . BC, which is characterized by excessive epithelial cell proliferation, is a highly heterogeneous and multifactorial disease with a genetic predisposition, associated metabolic disorders, and known environmental factors 3,4 . It is commonly acknowledged that the tumor microenvironment (TME) is critically involved in cancer initiation, development, and response to therapies 5,6 .
The TME is a heterogeneous and dynamic cellular milieu composed of resident and infiltrating cells and molecules 7,8 . Collective evidence has revealed the importance of cellular and molecular interaction networks within the TME for cancer cell survival 9 , and recent studies have focused on elucidating the mechanisms by which stromal inflammatory mediators secreted by TME components influence tumor progression.
As shown in Figure 1, the biological properties of tumor cells are inextricably regulated by dynamic shifts in systemic metabolic status due to intimate interactions between tumor cells and their surroundings. Obesity is a chronic metabolic disease that is defined by the accumulation and reprogramming of adipose tissue. Extensive studies have reported a correlation between obesity and breast tumor-related events, such as poor survival, therapy resistance, recurrence, and metastasis [10][11][12][13][14][15][16][17][18][19] ; however, the underlying mechanisms have not been established. Notably, obese patients usually have impaired inflammatory responses and immune regulation. Moreover, a correlation exists between body mass index (BMI) and the proinflammatory state of breast tissue (P = 0.004) 20 , which has a significant role in worsening the prognosis [21][22][23] . Inflammation is one of the hallmarks of carcinogenesis, and researchers have revealed that an inflammatory milieu could favor BC initiation and invasiveness 24,25 . Obesity stimulates adipose tissue and BC cells (BCCs) to release inflammatory mediators, predisposing the BCCs to a proinflammatory state, driving oxidative stress, and creating paracrine and autocrine feedback loops 26 . In summary, the obese environment creates  a distinctive inflammatory milieu in adipose tissue that favors BC initiation, primary growth, invasion, and metastatic progression in conjunction with systemic endocrine alterations.
Tumor progression is typically orchestrated by crosstalk between the tumor and stroma 27 . The TME in patients with BC is dominated by stromal cells, which actively participate in tumorigenic malignant transformation 5,12 . Breast ductal epithelial cells are immersed in an adipose environment containing plentiful, specialized adipocytes 29,30 . The intimate bidirectional relationship between BCCs and para-cancerous adipocytes has been verified 31,32 . In addition to being an energy resource, adipose tissue is a bioactive organ associated with endocrine, metabolic, and immune systems that regulate systemic energy and metabolic homeostasis, as well as contribute to the composition of the extracellular matrix (ECM) through a complex signaling network 33,34 . Adipocytes secrete numerous soluble inflammatory mediators that bind to cancer cell receptors and affect BCCs via paracrine signaling pathways 35,36 . Preclinical models have linked these secreted factors in primary BC to reduced differentiation and poor clinical outcomes 37,38 . Moreover, adipose tissue is composed of heterogeneous cells, such as adipose-derived stem cells (ADSCs), cancer-associated adipocytes (CAAs), and tumor-associated macrophages (TAMs), that are reprogrammed to exhibit different molecular and cellular characteristics during BC initiation and progression, and increase cancer cell proliferation and invasion capacities by producing inflammatory mediators 39,40 . Defining these unique differences in adipose components might contribute to a better understanding of how the immunosuppressive BC TME arises.
The mechanisms underlying obesity and inflammation influencing BC are largely unknown. Adipose tissue is an essential modulator of BC biology. Understanding how the obesity-associated in situ breast inflammatory microenvironment controls cancer cell behavior and identifying the regulatory elements involved could aid in developing prognostic or therapeutic targets.

Tumor stemness and carcinogenesis in the obese inflammatory microenvironment
Obesity increases the BC incidence, particularly for postmenopausal women 41,42 (Figure 2). The risk of breast tumor malignant progression may be accelerated by alterations in the mammary tissue prior to tumor formation 40,43 . BC emerges due to complicated interactions between the environment and genetics that modify the immune and inflammatory systems to promote carcinogenesis.

Tumorigenesis
Inflammation of white adipose tissue (WAT) is correlated with higher estrogen and aromatase gene expression, both of which are significant risk factors for BC in obese postmenopausal women. Martínez-Chacón et al. 38 reported that IL-10 inhibits aromatase gene expression in mesenchymal stem cells (MSCs) and ADSCs by suppressing TNF-αstimulated ERK1/2 activation. Increased adiposity caused by ovarian hormone deficiency in ovariectomized mice results in enhanced expression of aromatase in mammary adipose tissue mediated by decreased local IL-10. Similarly, Alhallak et al. 44 reported that compared to normal weight women, IL-10 and FOXP3+ T regulatory levels in breast adipose tissue of obese premenopausal women were upregulated along with the expression of tumor suppressive markers in breast tissue. In vitro application of IL-10 was shown to promote apoptosis and inhibit proliferation in the mammary epithelial cell lines, MCF10A and HMEC. In summary, IL-10 may act as a potential protective mediator against BC; however, the current evidence is insufficient and additional research is warranted. Furthermore, the above-cited studies revealed a difference in obesity-associated oncogenesis between pre-menopausal and post-menopausal women, which may reflect hormone levels.
Another study showed that CCL2 augments estrogen production and cancer risk by activating the aromatase promoter, pI.4, via the ERK1/2 pathway. The mammary fat pads of animals with weight gain have significantly higher amounts of IL-6, CCL2, and leptin, which are positively linked with mammary aromatase reporter activity in experiments with human primary breast adipose stem cells (ASCs) and mouse mammary adipose explants 45 . Additionally, local estrogen biosynthesis via aromatase and inflammation are linked to COX-2-dependent PGE2 production in obese breast tissue. Gonçalves et al. 46 reported that an obesogenic diet augments proinflammatory mediators (CCL2, IL-6, COX-2, and PGE2) in breast tissue, along with the formation of crownlike structures of the breast (CLS-B) and upregulation of local aromatase and estrogen, which accelerate COX-2-dependent breast tumor formation.
Other inflammatory mediators associated with obesity that may contribute to an increased risk of BC are also of concern. Human triple-negative BC (TNBC)-derived conditioned media increases the expression of the CAArelated cytokines (CCL2, CCL5, IL-1β, and IL-6), as well as the immune inflammatory regulators, COX-2, HIF-1α, VEGF-α, and PD-L1 in ADSCs during adiposity. The epithelial-mesenchymal transition (EMT) biomarker, Snail, and activation of Smad2 and NF-κB are also correlated with the CAA phenotype, which in turn has a role in  tumorigenesis 47 . Tumor-bearing breast adipose tissue has a distinct physiologic state compared with tumor-free breasts, and the capacity to secrete more IL-8, differentially express genes associated with inflammation and the integrin and Wnt signaling pathways, thereby contributing to tumorigenesis 27 . Roubert et al. 48 showed that BMI is positively correlated with the release of IL-1, IL-6, and TNF-α, which stimulates Wnt signaling in mammary tissue and promotes BC 48 . Furthermore, reduced breast epithelial tumor suppressor TGF-β1 activity and elevated TGF-β1 within the ECM of obese mammary tissue in vivo enhance BC risk, possibly due to interactions with decorin expressed on macrophages surrounding the ducts and lobules in obese breast tissue 49 .

Tumor cell stemness
Some stem or progenitor cell types may increase the probability of developing BC 50 . One study demonstrated that visfatin induces monocyte M2 differentiation via ERK/CXCL1 and enhances BCC tumorsphere formation and stemness 51 .
In addition, when co-culturing human-derived adipocytes with established and primary BCCs, the adipocytes secrete more IL-6 to promote Src activation, thus leading to upregulation of Sox2, c-Myc, and Nanog, which results in the emergence of tumor stem cell traits 52 . As demonstrated by Tiwari et al. 53 , obesity induces a pro-inflammatory metabolic activation phenotype in mammary adipose tissue macrophages in vivo, leading to the secretion of IL-6 in an NADPH oxidase 2-dependent manner and promotes stem-like features to alter the niche and support tumor formation via glycoprotein 130 signaling in TNBC cells. ADSCs exhibit tropisms similar to MCF-7 cells, in which secreted macrophage inflammatory protein (MIP)-1δ and MIP-3α induce tumorsphere formation in vitro and promote tumorigenicity in vivo 54 . In addition, adipokines may have an important role in forming stem characteristics. Avtanski et al. 55 showed that resistin induces MCF-7 and MDA-MB-231 cells to acquire cancer stem cell (CSC)-like properties. The adipokine, adipsin, promotes proliferation and CSC-like characteristics in xenograft models, suggesting that adipsin contributes to the TME and CSC niche in breast malignancies 56 . A-FABP released from adipose tissue promotes tumor stemness by activating the IL-6/STAT3/ALDH1 pathway 57 . Mishra et al. 58 also showed that leptin promotes BCC stemness, probably through the induction of TGF-β1 expression and secretion.

Inflammatory mediators from expanded adipose tissue affect BC growth and survival
In addition to the roles in immune regulation and inflammatory responses, obesity-derived inflammatory mediators also facilitate BCC growth and survival. In fact, adipokines may be involved in mediating this effect. Leptin stimulates the proliferation and viability of both ER-positive and TNBC cells, while also promoting anti-proliferative, cytotoxic, and pro-apoptotic effects in MCF-7 cells, similar to TNF-α and IFN-γ 59 . Accordingly, inflammatory mediators may affect diverse molecular subtypes of BC via different mechanisms.
There is also evidence that leptin treatment of MCF10A cells leads to increased survival and reduces apoptosis associated with FAK/AKT phosphorylation 60 . Visfatin is an adipokine, the serum levels of which are elevated in various cancers 61 . NAD generated by extracellular visfatin increases SIRT1 activity and p53 deacetylation, which induces BCC proliferation 62 .
Chemokines maintain an ability to induce targeted chemotaxis in effector cells, and the role of chemokines in promoting tumor proliferation has gained increased attention. In a xenograft model, CCL2-induced inflammation dramatically increases tumor development and facilitates the establishment of a desmoplastic stroma in a xenograft model by early recruitment of macrophages and CAFs into the TME. Fibrocytes may be a potential target inside the TME to minimize tumor fibrosis and improve the treatment response in obese BC patients 65 . Kim et al. 66 reported significant IL-6 staining in fibroblast-like cells within the peritumoral region, in addition to a positive correlation with CAF markers, thus implying that IL6-positive fibroblast-like cells may originate from pre-adipocytes and mature adipocytes.
Chemokines mediate tumor resistance by participating in the formation of in situ regional fibrosis, which in turn leads to tumor growth. Xenograft-bearing mice fed a high-fat diet (HFD) exhibit increased expression of IL-6 and TNF-α to sustain the local chronic inflammatory microenvironment and favor M1 macrophages, which together with nicotine increase macrophage infiltration and anti-inflammatory cytokines, such as IL-10, IL-13, and IL-4, promote BC growth and invasive behaviors 67 . IL-6 and CXCL1/2/3 are increased and MCP-2 is decreased in human BC-conditioned media, and this effect correlates with tumor stage and histologic grade 68 . The inflammation-related genes, IL6, Ptx3, IL33, and Timp1, are upregulated after 3T3-L1 adipocytes are co-cultured with BCCs, thus promoting proliferation in ER-positive BCCs 69 . Obesity-triggered chemokines are involved in tumor growth; however, there is a lack of in-depth studies with respect to the specific underlying mechanisms.
ADSCs are an important member of the adipose microenvironment. Crosstalk between the secretome and TME may have a role in tumor growth. ADSCs expressing CXCR4 are recruited by SDF-1α (CXCL12) secreted from chemoresidual TNBC cells. These ADSCs then initiate signaling to drive tumor cell proliferation by secreting FGF2 and activating ERK 70 . ADSC-secreted CXCL5 is a key factor in promoting the proliferation of MCF-7 and MDA-MB-231 cells 71 . Another study showed that ADSCs facilitate the growth of MCF-7 and ZR-75-30 cells in vivo by secreting the pro-angiogenic cytokines, CXCL1 and IL-8, but have a limited effect on MDA-MB-231 cells 72 . Furthermore, co-injection of c-Kit + ADSCs with 4T1 or endothelial progenitor cells in vivo elevate IL-3 and SDF-1 levels and increase tumor volume 73 . These results may provide a foundation for inhibiting ADSCs or the secretome in the clinical setting.

Obesity-associated inflammation promotes tumor angiogenesis
Angiogenesis is of great significance for tumor development because blood vessels in the TME supply tumor cells with nutrients and oxygen 74 . Tumor cells have been shown to attract bone marrow-derived vascular endothelial progenitor cells (BM-EPCs) and promote angiogenesis via the VEGF/ HIF-1α pathway 6 . Pro-inflammatory mediators also maintain angiogenesis and tumor progression, regardless of VEGF blockade [75][76][77] . Many of these variables are also increased in obese patients, thus creating a complex microenvironment where cytokines, chemokines, and adipokines are released to cause inflammation and recruit additional pro-inflammatory and pro-angiogenesis mediators [78][79][80] . Numerous studies have reported correlations between obesity and enhanced tumor angiogenesis, highlighting the importance of angiogenesis for the progression of obesity-driven BC 81,82 . To suppress tumor angiogenesis, several anti-angiogenic agents have been developed, which unfortunately have not yielded the desired results in clinical trials 83 .
One of the most pressing issues in the effort to improve therapies is how to overcome resistance to anti-angiogenic drugs in cancer patients. Using a zebrafish model of metastasis, Rodriguez et al. 84 showed that IL-8 generated by breast adipocytes might synergize with LFA-1, integrin, CCL5, or CCL2 to transform the BC microenvironment into a proinflammatory and pro-angiogenic state, while inducing resistance against anti-VEGF treatment in ER-positive BCCs. Incio et al. 85 reported that in vivo IL-6 inhibition and FGF-2 adjustment eliminate obesity-induced resistance to anti-VEGF therapy by impacting tumor cell growth, restoring tumor vasculature, alleviating hypoxia, and diminishing immunosuppression. Hsieh et al. 86 also proposed that IL-8 improves the pro-angiogenic effects of breast adipocytes.
VEGF-A promotes neovascularization, increases vascular permeability, and is closely related to angiogenesis. The obesity-associated inflammatory cytokines, TNF-α, INF-γ, and leptin, increase VEGF-A levels and affect angiogenesis in MDA-MB-231 cells 59 . Elevated IL-6 secreted by adipocytes co-cultured with cancer cells promotes angiogenesis via upregulation of STAT3 and VEGF 52 . Moreover, c-Kit + ASCs also improve VEGF levels and vascular formation 73 .
In addition to directly promoting VEGF-A expression, several potential mechanisms may also mediate tumor angiogenesis. Kolb et al. 87 reported that obesity increases the number of tumor-infiltrating macrophages with activated NLRC4 inflammasomes and boosts IL-1β production, thus promoting angiogenesis and cancer progression by upregulating ANGPTL4 expression via NF-κB and MAPK activation. Indeed, targeting ANGPTL4 separately or in conjunction with anti-VEGF treatment may be a more effective strategy for treating obese patients with BC. Another study showed that a pro-inflammatory setting with elevated IL-6 and IL-12 expression and upregulated stress-induced pp38 MAPK and pERK1/2 encourages BC angiogenesis in HFD xenografts 67 . Moreover, WAT-derived GM-CSF and MMP9 promote immunosuppression and intratumor vascularization 88 .
Inflammatory mediators also promote tumor angiogenesis by affecting the activity and function of vascular endothelial cells. Wang et al. 72 co-injected human ADSCs and BCCs in vivo and demonstrated that ADSCs secrete CXCL1 and IL-8 to increase the migration and tube formation of human umbilical vein endothelial cells (HUVECs) by signaling via the receptors, CXCR1 and CXCR2, which in turn enhance angiogenesis and tumor growth. In addition, paracrine activation of integrin β1-ERK1/2-HIF1α-VEGFA signaling by MSCs injected into adipose tissue promote HUVEC proliferation, migration, and angiogenesis 89 . Obesity results in the release of increased inflammatory mediators acting on vascular endothelial cells to promote viability, thereby favoring angiogenesis.

Obesity-induced inflammation transforms the mammary gland into a microenvironment conducive to tumor aggressiveness and metastasis
Metastasis remains the predominant cause of the high mortality rate among BC patients, although the 5-year survival rate has greatly increased over the past few decades 1 . Tumor dissemination is a complicated process in which tumor cells escape from the primary site, persist in the peripheral circulation, extravasate to distant sites, and ultimately multiply within the target organs. Several factors, including tumor cell characteristics and the local microenvironment in the target area, may affect metastasis. At present, insights into the underlying mechanisms by which obesity-induced metastasis occurs are largely lacking, yet comprehending the connection between obesity and metastasis is imperative to develop targeted therapies for obese individuals.

Chemokines
Chemokines and chemokine receptors have key roles in TME formation, leukocyte recruitment, angiogenesis, and metastasis, and also mediate interactions between tumor cells and tumor surroundings 90,91 . Targeting CC-chemokine ligand 5 (CCL5) in vitro reverses the enhanced motility of TNBC cells upon co-culturing with human adipocytes. CCL5 is also detectable in TNBC peri-tumoral adipose tissue and correlates with lymph node positivity, distant metastases, and reduced overall survival among patients with TNBC 92 . Adipocytes within breast tissue release increased CCL2 when lipid metabolism is altered. Patients with high expression of atypical chemokine receptor 2 (ACKR2) may bind and internalize CCL2 for intracellular degradation to limit tumor metastasis, thus leading to low recruitment of CCR2+ monocytes along with high recruitment of CCR2+ NK cells 93 . This result may provide a biomarker for the clinical management of patients. The interactions between adipokines and chemokines also has an essential role. Wang et al. 51 demonstrated that visfatininduced M2 differentiation in monocytes via CXCL1/ERK phosphorylation enhance BCC viability, migration, and EMT in vitro and in orthotopic mouse models. In vitro experiments showed that leptin leads to BC bone metastases by stimulating the SDF-1/CXCR4 axis, which has been validated using clinical samples 94 . Moreover, 4T1 cells treated with HFD mouse adipose-conditioned medium exhibit enhanced migration ability through CXCL12 and CCL25, with an overall decrease in immune cell infiltration and activation in tumor sentinel lymph nodes, suggesting a high risk of metastasis and immune escape 95 . In summary, obesity may be a risk factor for chemokines to enhance tumor metastasis and immunity exemptions.
Numerous substances, including chemokines, metabolites, and exosomes, are secreted by CAAs and may contribute to BC malignancy 96 . Human CAA-derived leukemia inhibitory factor (LIF) promotes BCC migration and invasion, elevates the expression of BC-derived CXCL1/2/3/8, and activates ERK1/2 signaling to further drive LIF expression in CAAs via NF-κB/STAT3 transcription. This effect forms a positive feedback loop between adipocytes and BCCs. Targeting LIF in combination with CXCR2 greatly decreases in vivo BC lung metastasis 97 . Another study showed that ectopic IL-8 expression in tumor-adjacent adipocytes converted adipocytes into CAAs, with high levels of NF-κB and NF-κB targets (leptin/ IL-8/IL-1 and an active pre-oncogenic STAT3-dependent EMT phenotype), on BCCs and in orthotopic tumor xenografts in mice. Additionally, inactivating CAAs and blocking the IL-8dependent pro-carcinogenic effects was shown to be achieved by targeting CXCR1/2 in mammary adipocytes in vitro 98 . CAA-derived molecules may promote tumor progression through communication with chemokines secreted by tumor cells.

NF-κB signaling pathway
The relationship between the NF-κB signaling pathway and pro-inflammatory mediators has been studied extensively and has been shown to regulate inflammation and cellular immune responses. Microarray data have revealed that co-culture with adipocytes induces migration and pro-inflammatory genes involved in NF-κB signaling in the MDA-MBA-231 cell line 99 . The expression of obesity-related mediators (SphK1 and S1PR1) is enhanced in metastatic lesions of syngeneic and spontaneous breast tumor obese mice, along with elevated levels of TNF-α and IL-6. Targeting the SphK1/S1P/ S1PR1 axis mediated by lipid metabolism attenuates key proinflammatory cytokines and macrophage infiltration as well as obesity-induced tumor progression 100 . Treatment with 17β-estradiol, leptin, IL-6, and TNF-α (ELIT) for obesityrelated inflammation also decrease mitochondrial function and increase oxidative stress, aggressiveness, and motility in cell lines with low estrogen receptor beta (ERβ) expression 101 . Consistent with this finding, human and mouse adipocytederived conditioned media increase BCC migration and proliferation by activating the PI3K-AKT-mTOR pathway and upregulating the expression of target genes, such as TNF-α, IL-1, and IL-6, suggesting that tumor-derived molecules regulate the TME in addition to the tumor, thereby aggravating diseases 102 . Another study indicated that upregulation of IL-6 and TNF-α is disrupted by obesity 95 , indicating that obesity may affect the NF-κB signaling pathway through different mechanisms, which warrant further study.
In addition to directly affecting tumor cells, the NF-κB pathway is also involved in adipocyte-mediated tumor progression and metastasis via paracrine signaling. The growth and migration of 4T1 cells are stimulated by 3T3-L1 adipocyte-conditioned medium, which could be rescued by inhibiting macrophage chemoattractant protein (MCP-1/CCL2), IL-6, IL-1, and plasminogen activators 86 . Pre-adipocytes upregulates IL-6 secretion and promotes proliferation, migration, and invasion of MCF10DCIS.com cells, which can be attenuated by blocking IL6-mediated cross-linking in vitro and in a xenograft tumor model 66 . These results indicated that human adipocyte-derived IL-6 and leptin boost BC proliferation and metastasis, and induce an EMT phenotype in vitro by activating PI3K-AKT and IL-6/JAK-STAT3 signaling 103,104 . Additionally, using a zebrafish model of metastasis, Rodriguez et al. 84 demonstrated that the release of IL-8 by breast adipocytes might cause neutrophils to adopt a pro-tumorigenic phenotype, which would then promote the dissemination of ER-positive BCCs. In conclusion, the NF-κB signaling pathway may promote tumor metastasis by affecting tumor cells and adipocytes.

ADSCs
Multipotent ADSCs have attracted extensive attention because of the regenerative properties in plastic surgery; however, these features are also connected to metastasis and tumor initiation, raising concerns about safety in clinical applications 105 . Human ADSC-secreted IL-6 significantly stimulates proliferation, EMT, transmigration, and 3D invasion of human primary normal and tumor epithelial cells 106 . Blocking ADSC/ Th2-derived IL-4 depleted metastatic ADSCs and elevated the expression of DUSP4 by inhibiting NF-κB and impairing the RAS/P38-MAPK pathway, thus delaying the proliferative and invasive phenotype of BCCs and driving the transformation of metastatic cells into non-metastatic cells 107 . By crossreacting with the transforming growth factor (TGF)-β/Smad and PI3K/AKT pathways, the paracrine impact of ADSCs promotes EMT of MCF7 cells 108 . Moreover, Chang et al. 109 discovered that adipose tissue-derived mesenchymal stem cells (AT-MSCs) obtained from type 2 diabetes mellitus (T2DM) donors differentiate into cancer-associated fibroblasts (CAFs) when co-cultured with BCCs under hypoxic conditions, in turn promoting BCC proliferation and in vivo metastasis, as well as the expression of fibroblastic markers, which was associated with TGFβ-Smad2/3 signal activation. Understanding the inflammatory characteristics of ADSCs may help develop applications in the clinical setting.

Immunomodulatory factors
Immune cells interact with tumor cells via immunomodulatory factors in the TME, which affects tumor behavior and treatment efficiency. A diet-induced obese mouse model showed elevated collagen production and proliferative rates of lung stromal cells, and the upregulated CSF2-and TGF-β1-induced recruitment and invasion of myeloid cells, as well as an immunosuppressive macrophage phenotype, respectively, which are correlated with pre-metastatic niche formation and lung metastasis of primary mammary tumors 110 . As a growth factor for hematopoietic and immune cells, granulocyte-macrophage colony stimulating factor (GM-CSF) mobilizes stem cells and causes macrophage/ granulocyte differentiation in addition to regulating inflammation and autoimmune disease 111 . Liu et al. 112 identified G-CSF and the G-CSF target gene matrix metallopeptidase (MMP) 2 and 9 in CAAs through transcriptomic sequencing. Treatment of TNBC cells with G-CSF enhance EMT, migration, and invasion through activation of STAT3 in cooperation with IL-6 and GM-CSF. TGF-β activates ERK and phosphorylates SMAD4 in obese patients, which allows USP9x to block the binding and mono-ubiquitination of SMAD4, thereby maintaining SMAD4 while promoting TGF-β/SMAD3-mediated transcription of EMT markers. Free fatty acids (FFAs) further facilitate the process and promote TGFβ-dependent cancer invasion and metastasis 113 . In addition, Wolfson et al. 114 reported that increased TGF-β1 in the mammary adipose tissue of obese mice activate SMAD3 signaling and inhibit miR-140 transcription, thus preventing miR-140 from targeting SMAD3 for degradation, resulting in the formation of a fibrotic microenvironment that impacts stemness, invasion, and proliferation of ductal epithelial cells. According to Quail et al. 115 , obesity-produced IL-5 boosts CSF2 expression by IL5Rα+ monocyte and neutrophil transport to the lungs of mice. In addition, obesity-related pro-metastatic effects can be reversed by blockade of GM-CSF or weight loss. After co-culturing with BC, human-derived WAT-derived ADSCs dramatically upregulate the expression of GM-CSF and MMP9, neutralization of which significantly reduced local and metastatic BC progression in vivo. This process is related to reducing macrophages and myeloid-derived suppressor cells (MDSCs) and can be inhibited by metformin, suggesting that proteins may be novel targets for this widely used drug 88 . Among all immune cells present in the BC TME, monocytes or TAMs account for the majority and are correlated with obesity and a poor prognosis in BC patients. Targeting or reversing the differentiation into monocyte-macrophage immunosuppressive phenotypes and the immunoregulatory factor, TGF-β, may have clinical application.

Adipokines
Adipokines mediate tumor development through tumor-stromal interactions in the TME. As a cytokine-like hormone, leptin serves as a bridge between BC and obesity. Knockdown of ObR in BCCs induce a less proliferative and invasive tumor phenotype 116 . Treatment with leptin induce EMT, migration, and invasion of BCCs and MCF10A cells, which correlate with increased FAK/AKT phosphorylation 60 . Gelsomino et al. 117 showed that aromatase inhibitor (AI) anastrozole-resistant MCF-7 BCCs (AnaR) express higher levels of leptin along with AnaR receptors and that targeting leptin signaling reduce growth and viability. Another study showed that leptin reduced cell aggregation and increased cell proliferation, migration, invasion, and EMT of cancer and epithelial cell lines, possibly by inducing TGF-β1 expression and secretion 58 . MDSCs are known promoters of cancer progression. Obesity-induced inflammation upregulates leptin in vivo, which induces the accumulation of excessive MDSCs and further facilitates spontaneous growth and metastasis by suppressing the activation and function of tumor-reactive T lymphocytes 118 .
Other types of adipokines also have an important role in tumor invasion and metastasis. Resistin stimulates motility in MCF-7 and MDA-MB-231 cells through cytoskeletal remodeling and EMT reprogramming 55 . A-FABP released from adipose tissue directly targets BCCs and enhances tumor aggressiveness in vivo by activating the IL-6/STAT3/ALDH1 pathway 57 . In addition, A-FABP expression in CD11b+F4/80+MHCII-Ly6C-phenotype TAMs facilitates protumor IL-6/STAT3 signaling by regulating the NFkB/miR-29b pathway, and A-FABP deficiency significantly decreases the growth and spread of breast tumors in transgenic and syngeneic tumor models 119 . Consequently, adipokines may be a potential target for the precise treatment of BC, but specific mechanisms need to be investigated further.
The impact of obese stroma on tumor cells and exposure to heightened inflammatory mediators present in the TME lead to alterations in the mammary epithelium that possibly predispose women to the developing of more aggressive BC.

Inflammation triggered by obesity makes the breast TME resistant to therapeutic approaches
A significant barrier to treating cancer is the resistance to chemotherapy and molecular-targeted therapies. Despite high response rates to initial therapy, many cancers eventually lose sensitivity to original treatment methods, resulting in metastasis and mortality. The TME is crucial in intrinsic drug resistance pathways 120 , and patients who are obese are more likely to develop treatment resistance 98,121 . Prior research has demonstrated that peri-tumor adipocytes contribute to the development of a radio-resistant phenotype in BC 122 . A recent study also found adipocytes to actively convert the chemotherapeutic agent daunorubicin into a less effective metabolite 123 .

Resistance to chemotherapy and endocrine therapy
Tumors adopt various pathways to inhibit chemotherapy efficacy, in which obesity-associated inflammatory pathways may be involved. The anti-proliferative effects of tamoxifen on MCF-7 cells are counteracted by co-culture with mature adipocytes from obese women, which significantly increase expression of TNF-α and IL-6 124 . Further evidence supports the role of obesity-induced inflammatory responses (IL-6, TNF-α, and leptin) in tamoxifen-acquired BC resistance 85 . Human ADSC-secreted CXCL1 in vitro downregulates miR-106a and upregulates ABCG2, a transporter of doxorubicin efflux in TNBC, thus conferring doxorubicin resistance 125 . The inflammatory adipocytokines (leptin, CCL2, IL-1β, and resistin) secreted by adipocytes and FFAs released by MDA-MB-231 cells confer acquired doxorubicin resistance in BC by generating inflammation and lipid metabolic reprogramming in the TME 61 . Mentoor et al. 126 suggested that obesity-induced resistin inhibits de novo fatty acid synthesis and lipolysis in mammary adipose tissue of mice, further aggravating local inflammatory reactions via the NF-κB pathway and promoting TNBC survival together with reduced doxorubicin efficacy in a paracrine manner. In addition, the adipokine, visfatin, enhances cell viability and prevents the reduction of survivin, a well-known candidate for chemoresistance of cancer cells, in TNFα-induced apoptosis and PARP cleavage 63 .

Resistance to radiotherapy
Radiotherapy is a mainstream BC treatment, and the TME is known to influence radiosensitivity. Inflammation is a crucial aspect of radiotherapy because inflammation can be triggered by released debris and the production of proteins induced by radiotherapy to enhance immune system elimination of cancer cells 127 . Autotaxin (ATX) is mainly secreted by mammary adipocytes in BC; however, as a component of the wound healing response, the ATX-lysophosphatidate (LPA)-inflammatory cycle shields cells against radiation-induced death and facilitates fibrosis 128 . Exposure of rat-and human-derived adipose tissue to γ-radiation produces a significant inflammatory response, and inflammatory cytokines secreted by breast tumors, including IL-1β, IL-6, IL-10, and TNF-α, stimulate ATX secretion by activating the NF-κB pathway. This response, in turn, generates LPA and stimulates further activation of NF-κB, which results in a positive feedback loop that promotes BC growth and metastasis 129 . Meng et al. 130 confirmed that repeated fractions of radiotherapy activate the above-outlined cycle, further increasing concentrations of VEGF, CXCL10, CCL11, and G-CSF in irradiated fat pads, thereby reducing the efficacy of additional fractions. Moreover, targeting ATX in a syngeneic orthotopic mouse model of BC decreased CCL11, IL-9, IL-12p40, M-CSF, and IFN-γ, which reversed pro-survival signals and radio-resistance in cancer cells. Furthermore, ATX inhibition is synergistic with irradiation and doxorubicin to decrease tumor growth and the number of Ki67-positive cells 131 . Inhibition of the adipose-related ATX pathway restores the pro-inflammatory properties of radiotherapy, and is therefore effective when combined with other therapeutic modalities.

Potential roles of inflammatory mediators in modulating immune responses
In recent years the development of immunotherapies based on immune checkpoint blockade (ICB) has dramatically reshaped the landscape of cancer therapy; however, despite the improved prognosis in a subset of cancer patients using currently-available immunotherapeutics, the response rate among TNBC patients is < 20% 132,133 , largely owing to the immunosuppressive TME. Therefore, a major focus of research involves discovering and addressing variables that contribute to the immunosuppressive TME. Obesity may be one such variable, with evidence suggesting that obesity perturbs the immune system and allows macrophages and T cells to differentiate into phenotypes that favor tumor growth 22 . Nevertheless, one study also showed that obese cancer patients have better overall survival, progression-free survival, and response rates after ICB treatment 134 . Accordingly, further study is needed to determine how obesity affects the outcomes of cancer patients receiving immunotherapy.
T and B cells are the main types of lymphocytes and have various biological functions, such as the direct killing of target cells, the production of cytokines and antibodies, and immune regulation. Gibson et al. 135 reported that obesity increases the concentration of CXCL1 in the mammary TME, drives CXCR2-mediated chemotaxis, and accumulates granulocytic MDSCs (G-MDSCs) to stimulate CD8+ T-cell apoptosis via Fas/FasL, leading to immunotherapeutic resistance in obese mice. By increasing intracellular IFN-γ and reducing PD-1 expression, blocking the IL-4 pathway promotes CD8+ T-cell cytotoxicity, thus sensitizing BCCs to anti-cancer therapy and enhancing the immune responses 107 . Another study indicated that the expression of leptin, CXCR4, and CCR9 in obese TNBC patients is negatively correlated with CD8+ T-cell infiltration 95 . Furthermore, ADSCs suppress B-cell proliferation, reduce the TNF-α+:IL-10+ B-cell ratio in a contactdependent manner, and alter the cytokine profile of B cells to an anti-inflammatory profile 136 . IL-6 inhibition reverses the recruitment of immunosuppressive regulatory T cells induced by anti-VEGF treatment in obese mice, suggesting that IL-6 suppression may augment immunotherapy in the context of obesity 85 . Additionally, pro-tumoral IL-17+ γδT cells multiply in obese mouse tumors, exhibit high lipid absorption, and intracellular lipid storage, indicating that metabolic programming regulates the immune cell lineage and function 137 . Moreover, using 4T1 tumor models in the presence or absence of adipocytes, Liu et al. 91 noted that TAA-secreted CCL2 recruits monocytes and macrophages to become immunosuppressive MDSCs and M2 macrophages. Targeting CCL2 in vivo increases T-cell infiltration, ameliorates the immunosuppressive TME, and facilitates immunotherapy.
Monocytes and macrophages are involved in innate immunity and have an important role in antigen presentation and immune regulation; however, an increasing number of studies have shown that myeloid cells have a large number of immunosuppressive phenotypes that have an important role in mediating tumor immune escape. Obesity-related M1 macrophages secrete IL-6 via a JAK/STAT-dependent pathway to promote PD-L1 expression in TNBC, and telmisartan reverses this process by activating peroxisome proliferator-activated receptor (PPAR-γ) and inhibiting NF-κB P65, thus highlighting the potential application in adjuvant TNBC immunotherapy 138 . Moreover, AnaRsecreted leptin enhances macrophage motility and induces an M2-like phenotype via CXCR4 signaling, highlighting the clinical advantage of targeting the cytokine network in obesity-associated hormone-resistant breast tumors 117 . Consistent with this finding, loss of ObR reduces macrophage recruitment, resulting in decreased CCL2 in xenograft tumors and co-culture experiments. In addition, the absence of Ob/ObR signaling regulates the immunosuppressive TME, as shown by reduced expression of PD-L1/PD-1/ arginase 1 and increased phagocytosis capability in macrophages 116 . Accordingly, obesity-associated inflammatory signals promote the formation of an immunosuppressive microenvironment and tumor immune escape by inducing macrophage differentiation into a pro-tumor phenotype and increasing immune checkpoint expression.

Discussion and future perspectives
Despite the well-documented association between obesity and cancer, the underlying mechanisms are unclear. Recognition of obesity as a chronic inflammatory condition, however, implies potential roles for inflammatory cytokines in obesity-associated BC (Figure 3). Nevertheless, there remain many unanswered questions.
To determine the disease extent and develop successful treatments, it is essential to recognize cancer as a complex and interconnected process. The inflammatory and obese microenvironment of BC is not confined to the effects of the adipocyte secretome. Stromal cells in the microenvironment release inflammatory mediators to affect cancer behavior and survival. Similarly, cancer cells also induce changes in surrounding cells via the secretome 139 . Interactions among immune cells, cancer cells, adipocytes, the immunome, and the metabolome generate intriguing dynamics and lead to the secretion of important molecules that are likely to be potential subjects for inflammation-targeted therapies in the future ( Table 1). Although many studies have identified cells and molecules that may have roles in mediating BC, there has been a lack of continuous observations and cell tracking to distinguish different cell types, determine the specific source of inflammatory molecules, or determine whether other microenvironmental factors are involved. Due to the many molecular BC subtypes, diverse interactions and heterogeneity are displayed 72 , yet this finding seldom receives any notice from researchers. Moreover, different adipocyte subpopulations between pre-and post-menopausal breast tissues likely utilize distinct mechanisms to make neighboring epithelial cells sense activated inflammatory signaling and exhibit diverse responses 44 . In brief, variables that may influence the heterogeneity of the TME warrant consideration.
Tumors grow in a complex microenvironment containing diverse cell types and inflammatory mediators. Beyond the local TME, a systemic inflammatory environment can also influence the disease course by interfering with homeostasis in a variety of tissues across the body. Many studies, however, do not consider in vivo inflammation, instead identifying interactions between tumor cells and adipocytes using co-cultivation methods. This approach may overlook the systemic effects of metabolism and immunity and lead to different consequences. Therefore, the complex interplay of inflammatory mediators in the context of an immunocompetent host remains to be explored. Additionally, it is unknown whether any of the current findings can be applied to the microenvironment of human BC; further exploration is needed in more reliable models, such as xenograft mice and organoids.
As a systemic disease, complex interactions occur between obesity and the endocrine and metabolic systems of the body, and the comprehensive effect on the tumor immune microenvironment (TIME) needs to be further explored. Obesity causes altered glucose metabolism and increases circulating FFAs, which results in hyperinsulinemia and insulin resistance, elevates the secretion of pro-inflammatory cytokines, causes excessive activation of insulin receptors, and promotes tumor proliferation. Insulin resistance often occurs along with obesity, and like obesity, insulin resistance influences the TIME through inflammatory pathways, such as IKKβ or NLRP3 140,141 . In addition, Wang et al. 142 Figure 3 Summarized potential inflammatory pathways that may be involved in obesity-associated TIME. Continuous cell proliferation and differentiation in a microenvironment containing inflammatory cells, growth factors, activated matrices, and DNA-damaged materials are crucial for inflammation to encourage cancer formation. The TIME is the essential condition for differentiation, survival, and metastasis, and inflammatory cells are closely involved in regulating the TIME. The remodeled breast TIME secretes inflammatory mediators that crosstalk with cells by acting on surface receptors, activating inflammation-related signaling pathways, and modifying the behavior of tumor, immune, and stromal cells. In an inflammatory microenvironment, there are three primary stages for tumor formation and development (initiation, promotion, and metastasis). The inflammatory microenvironment involves immune cells, inflammatory cells, and various produced bioactive chemicals, such as TNF-α, IFN-γ, interleukin, and chemokines, which participate in cancer cell tumorigenesis, growth, angiogenesis, metastasis, and resistance to therapeutic approaches.
proteins and neutrophil infiltration, promoting tumor progression. The study suggested that insulin resistance in the TIME is equally heterogeneous; however, additional studies are needed to verify the findings and the roles obesity and insulin resistance have in the inflammatory TIME.
BC exhibits high genotypic and phenotypic diversity, and heterogeneous cell types within the TME exhibit dynamic and tumor-promoting behaviors during cancer progression. Therefore, identifying interactions between individual cells and cytokines, as well as microenvironmental regulators of tumor progression, is essential to develop predictive molecular signatures and more effective treatments. Sophisticated multiomics technologies, such as genomics, proteomics, metabolomics, single-cell omics, and spatial transcriptomics, comprehensively analyze and reflect tumor heterogeneity at the single-cell level. High-throughput data can be used to analyze the spatial structure of the TIME at the level of single-cell clustering, with far-reaching significance for the realization of precise cancer therapy.
Accordingly, for an integrated understanding of cancer, research into how the systemic environment affects tumor biology is essential. Accurate and efficient analysis of the complex TME and evaluating the meaning of biological targets or biomarkers for clinical applications has been challenging for researchers. Predictive models are data-driven algorithms that combine multiple important outcome-related predictors to assess the prognosis, curative efficacy, and risks of developing a disease based on a variety of mathematical modeling approaches. For example, an integrated analysis of multiple databases from Zhao et al. 143 identified 10 hub genes associated with the extent of immune cell infiltration into the TIME and patient prognosis. The genes may serve as a risk and survival predictor of BC TIME. By applying mathematical methods to clinical medicine, increasingly accurate predictive With the continuous advances in technology, nextgeneration sequencing, including whole exon and RNA sequencing, has been gradually utilized in clinical practice, thus providing massive data that help us explore biological factors related to the treatment response and prognosis. Moreover, pathologic images of human tumor tissue contain a large amount of information that people cannot fully distinguish and understand. Artificial intelligence can extract complex information from visual data, digitally analyze histologic tumor slices, and infer molecular and genetic changes in tumor tissues to provide a wide range of clinically-relevant applications. For example, researchers recently constructed a simplified mathematical model to predict systemic immune cell connectivity and the dynamics of immune interactions using a high-throughput surface receptor screening method, thus providing a comprehensive view of the human immune system at the single-cell level that may provide opportunities for clinical therapeutic interventions 144 . High-throughput mathematical and digital models perform well in elucidating biological processes and identifying key factors of clinical significance, which is a subject that needs future investigation and has profound clinical translational value.

Conclusions
During cancer development, different types of stromal cells establish essential information exchanges with luminal and basal epithelial cells present in the mammary ducts through paracrine signaling or cell-cell communication within the mammary adipose microenvironment. A thorough understanding of this epithelial-stromal interaction might uncover new pathways and mechanisms involved in tumor progression, with an increasing number of studies shifting the targets from the internal communication between cancer cells to the interaction between malignancies and the surroundings. Given the complicated function of the aforementioned obesity-related inflammatory mediators in the TME, addressing the paracrine metabolic reprogramming and immunomodulatory pathways might be an efficient strategy to combat BC. On the premise of clarifying the heterogeneity of the TME, combining cell therapies inhibiting targeted inflammatory molecules with other conventional chemotherapy, immunotherapy, and radiotherapy strategies is a promising direction for treating BC.