miR155 deficiency reduces breast tumor burden in the MMTV-PyMT mouse model

Abstract

miRNA155 (miR155) has emerged as an important regulator of breast cancer (BrCa) development. Studies have consistently noted an increase in miR155 levels in serum and/or tissues in patients with BrCa. However, what is less clear is whether this increase in miR155 is a reflection of oncogenic or tumor suppressive properties. To study the effects of miR155 in a transgenic model of BrCA, we developed an MMTV-PyMT mouse deficient in miR155 (miR155^−/− PyMT). miR155^−/− mice (n = 11) exhibited reduced tumor number and volume palpations at ∼14–18 wk of age compared with miR155 sufficient littermates (n = 12). At 19 wk, mammary glands were excised from tumors for RT-PCR, and tumors were counted, measured, and weighed. miR155^−/− PyMT mice exhibited reduced tumor volume, number, and weight, which was confirmed by histopathological analysis. There was an increase in apoptosis with miR155 deficiency and a decrease in proliferation. As expected, miR155 deficiency resulted in upregulated gene expression of suppressor of cytokine signaling 1 (Socs1)—its direct target. There was a reduction in gene expression of macrophage markers (CD68, Adgre1, Itgax, Mrc1) with miR-155^−/− and this was confirmed with immunofluorescence staining for F4/80. miR155^−/− increased expression of M1 macrophage marker Nos2 and reduced expression of M2 macrophage markers IL-10, IL-4, Arg1, and MMP9. Overall, miR155 deficiency reduced BrCA and improved the tumor microenvironment through the reduction of genes associated with protumorigenic processes. However, given the inconsistencies in the literature, additional studies are needed before any attempts are made to harness miR155 as a potential oncogenic or tumor suppressive miRNA.

NEW & NOTEWORTHY To examine the effects of miR155 in a transgenic model of breast cancer, we developed an MMTV-PyMT mouse-deficient in miR155. We demonstrate that global loss of miR155 resulted in blunted tumor growth through modulating the tumor microenvironment. Specifically, miR155-deficient mice had smaller and less invasive tumors, an increase in apoptosis and a decrease in proliferation, a reduction in tumor-associated macrophages, and the expression of genes associated with protumoral processes.

INTRODUCTION

Breast cancer (BrCa) is the leading cause of cancer death in women and the most commonly diagnosed cancer worldwide with an estimated 2.3 million new cases in 2020 (1). Despite BrCa being a complex, heterogeneous, and multifactorial disease, evidence supports a critical role of the immune system in its progression and consequently provides a target for disease control (2–4). Indeed, breast tumorigenesis has been attributed, at least in part, to the ability of tumor cells to evade the immune system by releasing tumor promoting and immunosuppressive factors (3, 4). The complex tumor and host immune interactions within the tumor microenvironment (TME)—comprised of various immune cells, cytokines, and stroma—strongly influences tumor development, disease progression, survival, and even the response to therapy, all of which can deviate across subtypes and stages of BrCa (3–6). Interestingly, small noncoding microRNAs (miRNAs) have emerged as important regulators of these processes in BrCa and consequently may have promise to serve as novel anticancer strategies (7).

Dysregulated expression of small noncoding miRNAs have been demonstrated to function as oncogenes or tumor suppressors in various human cancers, including BrCa, as they regulate a myriad of genes involved in cellular processes including cell survival, proliferation, differentiation, adhesion, and motility (8–11). There has been a recent emphasis on the clinical potential of utilizing oncogenic and tumor suppressor miRNAs to identify disease state, cancer subtypes and stage, hormone receptor status, metastatic potential, and tumor invasiveness (9, 12–17). One miRNA in particular, miRNA155 (miR155), has emerged as an important regulator of BrCa development. In fact, it is being appraised as a potential noninvasive diagnostic biomarker for BrCa detection (18) and/or as an indicator of response to therapy (19). Studies have consistently found an increase in miR155 levels in serum and/or tissues in patients with BrCa (14, 18–21). However, what is less clear is whether this increase in miR155 is a reflection of oncogenic or tumor suppressive properties. Indeed, some studies have demonstrated that miR155 possesses oncogenic properties in BrCa (22), whereas others have reported tumor suppressive effects of miR155 in BrCa, including work by our group (23–27). The tumor promoting effects of miR155 have generally been attributed to its upregulation of proliferation and inhibition of apoptosis (22), whereas its tumor suppressive effects have been linked to its ability to regulate immune process (22–24, 26–28). Given the inconsistences in the literature, more work is required to truly understand the role of miR155 in BrCa and consequently its potential as a therapeutic target in BrCa—whether that be strategies to inhibit or overexpress this miRNA.

Although miR155 is upregulated in serum and tumor tissues of patients with BrCa and is closely associated with advanced BrCa or early BrCa relapse (29), its mechanistic role in mammary tumorigenesis continues to be unearthed using preclinical models. Thus, we investigated whether miR155 has any promise in reducing tumor burden in a transgenic BrCa model. Toward this, our laboratory developed a mouse mammary tumor virus polyoma virus middle T antigen (MMTV-PyMT) mouse with a global miR155 knockout. We found that miR155 deficiency significantly reduced breast tumor burden and improved the TME indicated by the expression levels of several immune related genes.

METHODS

Animals

Our laboratory developed MMTV-PyMT mice deficient in miR155 on a C57BL/6 background. Four-week-old female PyMT/miR155^−/− (n = 11) and their littermates PyMT/miR155^+/+ (PyMT; n = 12) were cared for in the Department of Laboratory Animal Resources (DLAR) at the University of South Carolina School of Medicine. Mice were housed 3–5/cage, maintained on a 12:12-h light-dark cycle in a low stress environment (22°C, 50% humidity, low noise), and given food and water ad libitum. All mice were fed an AIN-76A diet (Bioserv, Frenchtown, NJ, USA; Cat. No. F1515) (30, 31)—a purified, balanced diet that is phytoestrogen free. Dietary phytoestrogens have been shown to influence anxiety-related behaviors, fat deposition, blood insulin, leptin and thyroid levels as well as lipogenesis and lipolysis in adipocytes, all of which could nonspecifically impact tumorigenesis (32). Body weight, food, and water consumption were monitored weekly for the duration of the study (i.e., 15 wk). All experimental mice were euthanized at 19 wk of age; this time-point was selected 1) as all mice had developed tumors by this stage and 2) to prevent loss of experimental units to ulcers, infection, or moribundity that may skew our findings. The Institutional Animal Care and Usage Committee of the University of South Carolina approved all experiments, and all methods were performed in accordance with the American Association for Laboratory Animal Science.

Genotyping

Mice were genotyped for the PyMT transgene using the primer sequences as follows:

PYVT-1 GGAAGCAAGTACTTCACAAGGG, PYVT-2 GGAAAGTCACTAGGAGCAGGG and for miR155 as follows: WT Fwd 5′- GTGCTGCAAACCAGGAAGG-3′, WT Rev 5′- CTGGTTGAATCATTGAAGATGG-3′, miR155 5′- CGGCAAACGACTGTCCTGGCCG-3′. A snip of mouse tail was added to 150 µL of DirectPCR tail buffer (Viagen Biotech Inc, Los Angeles, CA) and 2 µL of Proteinase K (Viagen Biotech Inc, Los Angeles, CA) and digested at 55°C overnight. The next day samples were incubated at 95°C for an hour then added to a PCR cocktail for amplification. The PCR cocktail contained DNA template, upstream and downstream primers, ddH₂0, and GoTaq Green Master Mix (Promega Corp, Madison, WI). Samples were run on 2% agarose gel and compared with wild-type MMTV-PyMT negative control samples to determine genotype (520 base pair molecular weight for MMTV-PyMT-positive samples) and for miR155 compared with wild type (600 base pair for miR155 deficiency).

Tumor Palpations

Tumors were palpable and measured beginning at 12 wk of age until 18 wk of age, by the same investigator that was blinded to the experimental groups. Previous studies have reported that MMTV-PyMT mice develop palpable mammary tumors between 12 and 16 wk of age (33). On palpation of a tumor, calipers were used to measure the longest and shortest diameter of the tumor. The number of tumors within each mouse also was recorded and the tumor volume was estimated for each tumor using the formula: 0.52 × (largest diameter) × (smallest diameter)² as previously described (34).

Tissue Collection

At 19 wk of age, mice were euthanized by isoflurane overdose following a 4 h fast. Visible tumors were dissected from mammary glands, measured to determine tumor weight and tumor volume, and then fixed in 4% formaldehyde and paraffin embedded for histopathology. A portion of remaining thoracic mammary gland tissue also was removed. These tissues were either snap frozen in liquid nitrogen for gene expression analysis or fixed in 4% formaldehyde and paraffin embedded for histopathology, immunohistochemistry, and immunofluorescence.

Histopathology Analysis

The portion of the thoracic mammary gland that was excised from each mouse was fixed for 24 h in 4% formaldehyde, dehydrated with alcohol, and embedded in paraffin wax (FFPE). Sections of mammary glands and tumors were stained with hematoxylin and eosin (H&E; Fisher HealthCare, Irvine, CA; Cat. No.: 245–651 and 245–827) as previously described (35). All histological analyses were performed blindly by a certified pathologist (I.C.). Similar to our previous study, degree of dysplasia in the mammary gland was scored accordingly: no hyperplasia, adenoma/mammary intraepithelial neoplasia (MIN), and early and late invasive carcinoma (35).

Immunohistochemistry

To examine cell proliferation, immunohistochemical analysis of Ki67 (Abcam, Cat. No. ab16667) was performed according to manufacturer’s instructions with minor modifications. FFPE sections (5 µm) were deparaffinized and rehydrated before antigen retrieval for 30 min at 95°C with Rodent decloaker (Biocare Medical, Cat. No. RD913M). Sections were then incubated in 10% goat serum for 2 h at RT in a humid chamber to block nonspecific binding. Sections were labelled with primary antibody against Ki67 (1:100) in 1% BSA/TBS overnight at 4°C. Sections were then washed with TBS/0.025% Triton and incubated in 0.3% H₂O₂ to remove endogenous peroxidases, followed by being incubated in goat anti-rabbit IgG HRP-linked secondary antibody (1:100, Cell Signaling, Cat. No. 7074S) 1% BSA/TBS in for 2 h at RT. Color was developed with DAB (Vector Laboratories, Cat. No. SK-4100) as a chromogen for 6 min at RT followed by counterstaining using CAT hematoxylin and Tacha’s bluing for 30 s. Finally, sections were dehydrated and mounted with Permount medium (Fisher Chemical, Cat. No. SP15-100) and images were taken at 20× and 40× using Keyence All-in-One fluorescence Microscope BZ-X800. Percentage of Ki67 cells (DAB-positive nuclei) over total cells, relative to area, were quantified using Material Image Processing and Automated Reconstruction (MIPAR) software, more specifically, we used MIPAR’s immunohistochemistry recipe that quantifies positively stained DAB cells relative to area (36). Five images were assessed for each mouse (n = 3 mice/group).

TUNEL staining was carried out as per the manufacturer’s instructions using ApopTag Peroxidase In Situ Apoptosis Detection Kit (Millipore, Cat. No. S7100). Color was developed with DAB (Vector Laboratories, Cat. No. SK-4100) as a chromogen for 6 min at RT followed by counterstaining using 1% methyl green (R&D systems, Cat. No. 4800-30-18) for 10 min at RT and then washed with diH20 and 100% N-Butanol (Acros-Organic, Cat. No. 71-36-3). Sections were dehydrated and mounted with Permount medium (Fisher Chemical, Cat. No. SP15-100). Representative images were taken at ×20 and ×40 using Keyence All-in-One fluorescence Microscope BZ-X800. Percentage of apoptotic cells (TUNEL-positive nuclei) over total cells, relative to area, were quantified using Material Image Processing and Automated Reconstruction (MIPAR) software, more specifically, we used MIPAR’s immunohistochemistry recipe that quantifies positively stained DAB cells relative to area (36). Five images were assessed for each mouse (n = 3 or 4 mice/group).

Immunofluorescence

After deparaffinization and rehydration, paraffin sections were incubated with rodent decloaker (Biocare Medical, Cat. No. RD913M) for 30 min at 95°C for antigen retrieval and then blocked with 10% goat serum for 1 h at RT. Sections were labelled with primary antibody against F4/80 (1:100, Santa Cruz, Cat. No. sc-377009) overnight at 4°C and then detected with Alexa Fluor 488 goat anti-rat secondary antibody (1:200, Abcam, Cat. No. ab150157) for 30 min at RT in dark and counterstained with 4’,6-diamidino-2-phenylindole (DAPI). Fluorescent images were taken at 20× using Leica DM2500 fluorescent microscope.

Real-Time Quantitative PCR

Mammary glands were homogenized, and RNA was extracted using the RNeasy Lipid Tissue Mini Kit (Qiagen, Cat. No. 74804) given the mammary gland lipid content according to the manufacturer’s instructions. RNA sample quality and quantities were verified using a Nanodrop One Microvolume UV-Vis Spectrophotometer (Thermo Fisher Scientific, Waltham, MA, United States) and determined to be of good quality based on A260/A280 and 260/230 values (>1.8) before cDNA synthesis using High capacity Reverse Transcriptase kit (Applied Biosystems, Cat. No. 4368814). PCR analysis was carried out as per the manufacturer's instructions and all primers used were TaqMan Gene Expression Assays (Applied Biosystems, Foster City, CA). TaqMan reverse transcription reagents were used to reverse transcribe and to analyze mammary gland mRNA gene expression of the following markers for: apoptotic marker Bax; suppressor of cytokine signaling (Socs) 1—direct target of miR-155; total macrophages as indicated by cluster of differentiation-68 (CD68) and adhesion G-protein coupled receptor E1 (Adgre1) also known as F4/80 and EMR1, M1 macrophage marker as indicated by integrin subunit α × (Itgax), and M2 macrophage marker as indicated by mannose receptor C-type 1 (Mrc1); M1-associated markers as indicated by interleukin (IL) 1β, IL-6, interferon-γ (IFN-γ), tumor necrosis factor α (TNF-α), and nitric oxide synthase 2 (Nos2); M2-associated markers as indicated by interleukin (IL)-13, IL-10, IL-4, tumor growth factor-β1 (TGF-β1), arginase 1 (Arg1), vascular endothelial growth factor (VEGF), and matrix metalloproteinase 9 (MMP9); T cell and neutrophil markers as indicated by CD4, CD8a, forkhead box protein P3 (FOXP3), and lymphocyte antigen 6 complex locus G6D (Ly6g). RT-PCR was performed as previously described (37). Briefly, quantitative RT-PCR analysis was carried out as per the manufacturer’s instructions (Applied Biosystems, Foster City, CA) using Taq-Man Gene Expression Assays on a Qiagen Rotor-Gene Q. Data were normalized to PyMT controls and compared with five reference targets (Hmbs, B2M, TBP, H2afv, and 18 s), which were evaluated for expression stability using GeNorm.

Statistical Analyses

All data were analyzed using commercial software (GraphPad Software, Prism 7, La Jolla, CA). Body weights were analyzed using a two-way analysis of variance at each time point. Tumor palpations, end point tumor data, IHC, and all mRNA analysis of mammary gland fold change gene expressions were analyzed using a two-tailed t test. Statistical significance was set with an α value of P < 0.05. All data are presented as means ± standard error (SE).

RESULTS

Animal Characteristics and Tumor Progression

Bodyweight, number of palpable tumors, and volume of palpable tumors were measured weekly beginning at 4 wk of age until 18 wk of age for female PyMT (n = 12) and PyMT/miR155^−/− (n = 11) mice. No differences in body weight were detected between the groups at any time point (Fig. 1A). PyMT/miR155^−/− mice displayed significantly lower number of palpable tumors at 14–18 wk of age compared with PyMT mice (Fig. 1B, wk 14, P = 0.003; wk 15, P = 0.006, wk 16, P = 0.004; wk 17, P = 0.001; wk 18, P = 0.001). In addition, PyMT/miR155^−/− mice exhibited significantly reduced volume of palpable tumors at 14–18 wk of age (Fig. 1C, wk 14, P = 0.001; wk 15, P = 0.049, wk 16, P = 0.011; wk 17, P = 0.05; wk 18, P = 0.012).

Figure 1.

PyMT/miR155^−/− mice exhibit reduced tumor progression and tumor burden. Female PyMT (n = 12) and PyMT/miR155^−/− (n = 11) mice were euthanized at 19 wk of age. A: absolute body weights shown in grams (g). Tumors were palpated weekly from wk 12–18 to obtain tumor number (B) and tumor volume given in millimeters³ (mm³; C). Tumors were excised from mice for tumor number (D), tumor weight given in milligrams (mg; E), and tumor volume given in mm³ (F). Significance was set at P < 0.05. *Significantly different from PyMT using a Student’s t test.

miR155 Deficiency Reduced Tumor Burden

On euthanasia (19 wk of age), female PyMT (n = 12) and PyMT/miR155^−/− (n = 11) mice, mammary glands were excised, and tumors were counted, measured, and weighed. PyMT mice deficient in miR155 exhibited reduced tumor number (Fig. 1D, P = 0.020), weight (Fig. 1E, P = 0.011), and volume (Fig. 1F, P = 0.038) at end point. Specifically, the average tumor number at end point (19 wk of age) was 7.5 ± 0.45 tumors in PyMT mice compared with the significantly lower 5.9 ± 0.43 tumors in PyMT/miR155^−/− mice (Fig. 1D, P = 0.020). Furthermore, the average tumor weight at end point (19 wk of age) was 906.08 ± 20.27 mg in the PyMT mice compared with the significantly reduced weight of 390 ± 13.42 mg in PyMT/miR155^−/− mice (Fig. 1E, P = 0.011). Finally, PyMT/miR155^−/− mice also exhibited a significantly lower average of tumor volume at end point (19 wk of age), 538.20 ± 23.60 mm³, compared with 1069.86 ± 29.46 mm³ in PyMT mice (Fig. 1F, P = 0.038). Reduced tumor burden displayed in miR155-deficient mice was also confirmed by H&E staining (Fig. 2A). Histopathological mammary gland stage progression into the early and late invasive carcinoma grade was exhibited in 100% of PyMT mice, but only 40% of PyMT/miR155^−/− mice (Fig. 2B). The majority (60%) of miR155-deficient mice exhibited a lower degree of dysplasia characterized as adenoma/mammary intraepithelial neoplasia (MIN; Fig. 3B). Moreover, PyMT/miR155^−/− mammary carcinomas were well differentiated as opposed to carcinomas developed in the PyMT mice that were mostly poorly differentiated (Fig. 3C). Well differentiated tumors are associated with a better prognosis as they tend to grow and spread more slowly (38), which was exhibited in miR155-deficient mice via reduced tumor palpation volume. In comparison, poorly differentiated carcinomas in PyMT mice are associated with higher aggressiveness (increased mitotic index, lack of gland formation, high cellular pleomorphism) and poorer prognosis (38).

Figure 2.

miR155 deficiency reduced breast dysplasia. Mammary glands and tumors were excised from PyMT and PyMT/miR155^−/− (n = 11–12) mice and stained with hematoxylin and eosin (H&E). A: representative 40× images of mammary glands stained with H&E. Scale bar = 50 µm. B: degree of dysplasia was characterized as adenoma/mammary intraepithelial neoplasia (MIN) or early invasive carcinoma. C: representative 10× and 20× images of mammary tumors stained with H&E. Significance was set at P < 0.05. Scale bar = 100 µm.

Figure 3.

miR155 deficiency increased breast tumor apoptosis and decreased proliferation. Immunohistochemistry of breast tumor sections from PyMT and PyMT/miR155^−/− mice (n = 3–4/group) stained with TUNEL for detection of apoptotic cells and Ki67 to examine proliferation. A: representative 20× images with 40× insets of breast tumors stained with TUNEL (top) and Ki67 (bottom). Scale bar = 50 µm. B: percentage of apoptotic (positively stained TUNEL-DAB) cells relative to area. C: gene expression analysis of Bax, a proapoptotic marker, within the mammary gland associated tumor microenvironment using qPCR. D: percentage of Ki67-positive cells (positively stained Ki67-DAB) cells relative to area. Significance was set at P < 0.05. *Significantly different from PyMT using a Student’s t test.

miR155 Deficiency Increased Apoptosis and Decreased Proliferation

To further investigate the antitumor role of miR155 deficiency, breast tumor sections were stained with TUNEL for detection of apoptotic cells (Fig. 3A, top) and Ki67 to examine proliferation (Fig. 3A, bottom). miR155 deficiency significantly increased the percentage of apoptotic (positively stained) cells (Fig. 3B, P = 0.004; i.e., the number of TUNEL-positive cells over total cell number relative to area). In addition, gene expression of proapoptotic marker, Bax, also was significantly increased in PyMT/miR155^−/− breast tumors compared with PyMT (Fig. 3C, P = 0.044). PyMT/miR155^−/− tumors exhibited a significant decrease in the percentage of Ki67-positive cells (P = 0.002) compared with PyMT (Fig. 3D).

miR155 Deficiency Improved the TME

To investigate the antitumoral effects associated with miR155 deficiency, mammary glands were further analyzed. Given their role in tumorigenesis and documented link to miR155, we focused on examination of Socs1 (direct target of miR155), macrophage markers, and associated inflammatory mediators within the mammary TME (Fig. 4). We demonstrate that miR155 deficiency significantly upregulates Socs1, a direct target of miR155 (Fig. 4A, P = 0.019) and significantly reduces macrophage markers as indicated by lower expression of CD68 (0.007), Adgre1 (P = 0.042), Itgax (P < 0.001), and Mrc1 (P = 0.002; Fig. 4B). This reduction was confirmed with immunofluorescence staining for F4/80 (Fig. 4C). We next assessed the contribution of miR155 deficiency to M1 and M2 macrophage related molecules. Specifically, we report an enhancement of the M1 related marker iNOS (Nos2, P = 0.018; Fig. 4D); M1 macrophages are known to participate in the elimination of tumor cells via iNOS (Nos2) (39). There was a trend for a reduction in IL-1β (P = 0.052) expression in PyMT/miR155^−/− mice but no differences were detected for IL-6, IFN-γ, or TNF-α (Fig. 4C). On the contrary, mammary glands of mice deficient in miR155 exhibited a reduction in select M2 macrophage-related cytokines—important contributors to tumor growth and immunosuppression—as indicated by reduced IL-10 (P = 0.038), IL-4 (P = 0.017), Arg1 (P = 0.033), and MMP9 (P = 0.046; Fig. 4E) but not IL-13, TGFβ, or VEGF (Fig. 4E). Finally, mammary glands of PyMT/miR155^−/− mice exhibited significant reductions in CD8a (P = 0.005) and regulatory T-cell gene Foxp3 (P = 0.014; Fig. 4F) but no changes in markers associated with T-helper cells (CD4) and neutrophils (Ly6G; Fig. 4F).

Figure 4.

miR-155 deficiency improved the inflammatory tumor microenvironment. Mammary glands were excised from tumor and inflammatory genes were analyzed using qPCR for suppressor of cytokine signaling 1 (Socs1; A)—direct target of miR-155. B: macrophages as indicated by cluster of differentiation (CD) 68, adhesion G protein-coupled receptor E1 (Adgre1), M1 macrophages as indicated by integrin subunit α × (Itgax), and M2 macrophages as indicated by mannose receptor C-type 1 (Mrc1). C: representative 20× images of immunofluorescence staining of F4/80 in breast tumor tissues harvested from PyMT and PyMT/miR155^−/− mice. DAPI (blue) as an individual channel (top) and F4/80 (green) as an individual channel (middle) for visualization of nuclei, and merged (bottom). Scale bar = 100 µm. D–F: M1-associated markers as indicated by interleukin (IL) 1β, IL-6, interferon-γ (IFN-γ), tumor necrosis factor α (TNF-α), and nitric oxide synthase 2 (Nos2; D); M2-associated markers as indicated by IL-13, IL-10, IL-4, tumor growth factor-β1 (TGF-β1), Arginase 1 (Arg1), vascular endothelial growth factor (VEGF), and matrix metalloproteinase 9 (MMP9; E); and T-cell and neutrophil markers as indicated by CD4, CD8a, forkhead box protein P3 (FOXP3), and lymphocyte antigen 6 complex locus G6D (Ly6g; F). Data were normalized to PyMT and compared with five reference targets (B2M, TBP, HPRT, HMBS, and H2AFV), which were evaluated for expression stability using GeNorm. Significance was set at P < 0.05. *Significantly different from PyMT using a Student’s t test.

DISCUSSION

Identifying regulators of tumorigenesis in BrCa remains critical to identifying diagnostic markers and therapeutic targets. Currently, understanding the interaction between key immune modulators and cancer progression is at the forefront of novel cancer therapies. Moreover, understanding regulators of the TME, including miRNAs, provides key mechanistic insight into the pathogenesis of certain BrCa subtypes. One miRNA, miR155, has been documented to be elevated in serum and tumor samples of patients with BrCa (14, 18–21). However, whether this miRNA serves as a tumor promoter or tumor suppressor has not been established in the preclinical literature. Therefore, we sought to understand the role of miR155 in the progression of BrCa using a transgenic mouse model. We identified that global loss of miR155 resulted in blunted tumor growth through modulating the TME. Specifically, miR155-deficient mice had smaller and less invasive tumors, an increase in apoptosis and a decrease in proliferation, a reduction in tumor associated macrophages and the expression of genes associated with protumoral processes including Mrc1, IL-10, IL-4, Arg1, and MMP9.

Our findings of an oncogenic role for miR155 in a mouse model of BrCa are supported by previous reports. For example, Jiang et al. (40) reported that ectopic expression of miR155 promoted the proliferation of breast cancer cells and the development of tumors in nude mice. Similarly, Kong et al. (41) reported that upregulation of miR155 promotes tumor angiogenesis and is associated with poor prognosis and triple negative breast cancer. In tumor xenografts, it was reported that inhibition of miR155 significantly reduced proliferation of cancer cells further supporting an oncogenic role for miR155 in BrCa (42). Conversely, there also are several reports suggesting miR155 possesses antitumor or tumor suppressive properties. In fact, a previous report by our group indicated that miR155 deficiency enhanced tumor growth and metastasis in an orthotopic BrCa mouse model (EO771) (24). Similarly, a colorectal cancer mouse study published by our group documented that miR155 deficiency resulted in a greater number of polyps/adenomas, an increased symptom severity score, a higher grade of epithelial dysplasia, and a decrease in survival (43). Further, when miR155 was overexpressed, we reported suppression of breast tumor growth along with a reduction in lung metastasis (23). In addition, stable expression of miR155 in 4T1 murine breast tumor cells significantly reduced the aggressiveness of tumor cell dissemination by preventing tumor cell epithelial-to-mesenchymal transition (EMT) in vivo (44). Interestingly, however, when these tumor cells were administered into the bloodstream, miR155 promoted tumor formation in the lung (44). This highlights the complexity of the relationship between miR155 and mammary tumorigenesis.

In the current study, we have attributed the blunting of mammary tumorigenesis with miR155 deficiency to a reduction in macrophages—specifically a reduction in M2-like macrophages. M2-like macrophages promote breast tumorigenesis via secretion of anti-inflammatory cytokines that reduce inflammation and support immunosuppression, angiogenesis, tumor invasion, and metastasis (45, 46). miR155 is differentially expressed in monocytes—precursors to macrophages—and has been reported to regulate macrophage polarization (47). Indeed, accumulating evidence suggests that miR155 decreases protumor-M2 polarization in normal bone marrow derived macrophages (47); however, it remains unclear whether this holds true in the context of BrCa. In the current study, we show that miR155 deficiency reduced expression of CD68 and Adgre1, pan macrophage markers and this was confirmed with IF staining for F4/80. However, inconsistent with the finding that miR155 depletion promotes polarization to an M2 or protumor phenotype (47), we found that M2 macrophage marker Mrc1 (also known as CD206), Arg1 (the prototype marker for M2 macrophage expression), M2-associated cytokines (IL-4 and IL-10), and angiogenesis marker MMP9 were decreased in miR155-deficient mice. Supporting this finding, Kong et al. (41) reported that mammary fat pad xenotransplantation of ectopically expressed miR155 resulted in recruitment of tumor-associated macrophages.

Contrary to M2-like macrophages, M1-like macrophages largely promote antitumor immunity by stimulating T-helper 1 cell responses, recruiting cytotoxic CD8 T cells to tumors, and secreting proinflammatory cytokines and chemokines (IL-6, IL-1β, and TNF-α) (45). However, this relationship is context dependent as it is well-known that inflammation can promote tumorigenesis (48). miR155 is thought to regulate inflammatory processes via its actions on Socs1—its direct target (49). As expected we show that Socs1 is increased in miR155 mice. Theoretically, this should translate to a reduction in proinflammatory cytokines as Socs1 suppresses cytokine signaling. As such, we report a trend for a reduction in IL-1β in the current study but there was no change in IL-6, IFNγ, or TNFα. This may not be entirely surprising given the complex relationship between inflammation, macrophages, and tumor progression. Continuing with our discussion of M1 markers, we report an increase in Nos2 expression in miR155-deficient mice. Nos2 is expressed at substantial levels in proinflammatory macrophages. Just like macrophages, the role of Nos2 during tumor development is complex and even perplexing with both tumor suppressive and tumor promoting activities having been reported (50). Antitumor activities of Nos2 include apoptosis and cytotoxicity whereas protumor activities include metastasis and invasion (39, 50). It appears that the dual role of Nos2 is influenced by the cell situation and is environment dependent, with either induction or inhibition of Nos2 having antitumor potential depending on the TME (50). The increase in Nos2 expression in miR155-deficient mice in the current study was associated with a reduction in tumorigenesis. This is in contrast to a report that miR155 knockdown significantly accelerated tumor growth by impairing classic activation of tumor-associated macrophages (51). Thus, the role of miR155 on macrophage behavior is not yet firmly established, at least in the context of BrCa.

Previous studies, including those by our group, have found that miR155 plays a critical role on other immune cells that are known to impact mammary tumorigenesis. Indeed, we previously reported that miR155 is required for dendritic cells to exert effective functions in the antitumor response, including maturation, cytokine secretion, migration toward lymph nodes, and activation of T cells in an orthotopic EO771 BrCa mouse model (24). Unlike the current findings, miR155 deficiency was actually associated with enhanced tumor growth in the EO771 BrCa mouse model. A follow-up study confirmed this beneficial role for miR155 on dendritic cell function in BrCa using miR155 overexpression experiments (23). In the current study, we report a reduction in expression of CD8 and Foxp3 in miR155-deficient mice, which is in line with others reporting that miR155^−/− mice have immune cell defects (52). However, whether these responses impacted tumorigenesis in the PyMT/MMTV model could not be determined from the current study design. Based on the available literature it is undisputed that miR155 plays a critical role in immune cell responses. What is less clear is the role it plays on immune cell function in the context of BrCa initiation.

The inconsistencies on miR155 effects in BrCa may be due to variations in cancer subtypes and animal models used. A meta-analysis was performed to shed light on the inconsistences reported in the literature. A total of 13 studies that included 791 BrCa samples and 509 normal samples were included in this meta-analysis; as expected, there was a significant increase in miR155 in blood and tissue samples of patients with BrCa (21). Interestingly, in subtype analysis, the expression level of miR155 was significantly correlated with estrogen receptor α (ER), progesterone receptor (PR), HER2, lymph node metastasis, tumor size, and p53 status. In subtype analysis of different ER and PR status BrCa, miR155 was significantly less expressed in ER⁺ or PR⁺ BrCa (21). But it was highly expressed in HER2⁺ or lymph node metastasis positive BrCa compared with HER2⁻ or lymph node metastasis negative BrCa (21). Thus, there appears to be an interaction between miR155 levels and hormone status in BrCa. In the current study we used a PyMT/MMTV mouse model that presents with loss of ER⁺ and PR⁺ receptors and overexpression of HER2 as the cancer progresses to malignancy. Although the literature base on the interaction between miR155 and hormone status in BrCa is too small to draw firm conclusions, we can certainly speculate that this may be contributing to the divergent findings across animal studies.

There are several limitations to the current study. First, we used just one mouse model of spontaneous BrCa that is not representative of all BrCa models or patients. Given that there are disparities in findings across mouse models, additional studies using various mouse models that incorporate different BrCa subtypes are necessary to understand the impact of miR155 on BrCa. Second, our immune markers are limited to gene expression data and immunofluorescence; flow cytometry was not performed on the mammary tissue and thus more studies are needed to comprehensively elucidate miR155’s role in immune cell regulation in BrCa. Third, our miR155 knockout model was a global knockout model; it is possible that local and global miR155 have distinct effects on tumorigenesis. To tease out miR155 effects on specific immune cells in BrCa, studies utilizing cell-specific knockout and overexpression models are needed. Finally, our macrophage marker data and subsequent interpretations are limited to examination of one time-point; it is possible that the profile of these markers would be different if earlier time-points in the progression of BrCa were assessed. Despite these limitations, the current data using a novel mouse model of BrCa significantly contribute to the discussion on the role of miR155 on BrCa.

In summary, the clinical literature consistently supports an increase in miR155 levels in serum and/or tissues in patients with BrCa (14, 18–21). However, preclinical studies investigating whether this increase in miR155 is a reflection of oncogenic or tumor suppressive properties remain divisive. In the current study using a PyMT/MMTV mouse model that is deficient in miR155, we show that miR155-deficient mice had smaller and less invasive tumors, as well as a reduction in the expression of genes associated with protumoral processes including Mrc1, IL-4, IL-10, Arg1, and MMP9. Additional studies in this domain are needed to include examination of a possible interaction between miR155 and hormone status before any attempts are made to harness miR155 as a potential oncogenic or tumor suppressive miRNA.

DATA AVAILABILITY

Data will be made available upon reasonable request.

GRANTS

This work was supported in part by National Cancer Institute Grant 5F31CA183458 (to T.L.C.).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

T.L.C., K.T.V., R.T.E., D.F., and E.A.M. conceived and designed research; S.J.M., T.L.C., B.N.V., T.D.C., K.T.V., and R.T.E. performed experiments; S.J.M., T.L.C., T.D.C., and I.C. analyzed data; S.J.M., B.N.V., K.T.V., R.T.E., I.C., D.F., and E.A.M. interpreted results of experiments; S.J.M. and T.D.C. prepared figures; S.J.M., B.N.V., and E.A.M. drafted manuscript; S.J.M., T.L.C., B.N.V., T.D.C., K.T.V., R.T.E., I.C., D.F., and E.A.M. edited and revised manuscript; S.J.M., T.L.C., B.N.V., T.D.C., K.T.V., R.T.E., I.C., D.F., and E.A.M. approved final version of manuscript.