Steroid receptor coactivator-3 inhibition generates breast cancer antitumor immune microenvironment

C57BL/6J, C57BL/6J-SCID, B6 albino, BALB/cJ, and SCID mice were purchased from Jackson Laboratory. All animal studies were conducted with the approval of the Institutional Animal Care and Use Committee at Baylor College of Medicine. All mice were euthanized in a CO₂ chamber. Additionally, any animals that showed a loss of mobility, 10% weight loss, or other symptoms related to tumor growth were euthanized. Tumors were not allowed to grow to > 20% of body weight or ulcerate. All animals were euthanized, and tumors were harvested at a volume of no more than 4,000 mm³.

In the luciferase activity analysis of E0771 tumor growth in Fig. 1A, the mean luciferase activity in mice treated with the vehicle was 23,923,333, and the standard deviation was 6,012,964. The mean luciferase activity in mice treated with SI-2 (2.5 mg/kg) was 4,672,000, and the standard deviation was 2,324,967 on the 35^th day after drug treatment. Therefore, the power calculation based on these data indicated that the animal size (n = 2/group) was sufficient to see a significant effect on E0771 breast cancer growth (p < 0.05).

Regarding the tumor size determination (Fig. 1D), the tumor volume in mice treated with the vehicle was 1,250 mm³ ± 155 and the tumor volume in mice treated with SI-2 (2.5 mg/kg) was 590 mm³ ± 57 on the 35th day after drug treatment. Therefore, the power calculation indicated that the optimum number of animals needed to attain a significance level of p < 0.05 with a statistical power of 95% was one.

E0771 (CRL-3461, ATCC) and 4T1 cells were grown in RPMI 1640 containing 10% (vol/vol) fetal bovine serum (FBS). MDA-MB-468 cells were cultured in DMEM (high glucose) containing 10% (vol/vol) FBS. The E0771, 4T1, and MDA-MB-468 cell lines were maintained in a humidified incubator with 5% (vol/vol) CO₂ at 37 °C.

The firefly luciferase gene was cloned into the pSMPUW-Hygro Lentiviral Vector (Cell Biolabs, catalog number: VPK-214). Lentiviruses containing a luciferase expression cassette were produced in 293 TN cells (System Bioscience, catalog number: LV900A-1) by transient transfection of the pSMPUW-Hygro Lentiviral Vector carrying the luciferase gene and the ViraSafe™ Lentiviral Packaging System (Cell Biolabs, catalog number: VPK-206) with Lipofectamine 2000 (Thermo Fisher Scientific, catalog number: 11668030). The recombinant lentivirus titer was measured using Lenti-X™ GoStix™ Plus (ClonTech, catalog number: 631280). E0771 and 4T1 cells were transduced with lentiviral vectors carrying the luciferase expression cassette with TransDux MAX™ (System Bioscience, catalog number: LV860A-1). Luciferase-labeled E0771 and 4T1 cells were then selected in the presence of 300 µg/ml hygromycin. Luciferase gene expression in E0771 and 4T1 cells was validated using a luciferase activity assay kit (Promega, catalog number: E4550). Luciferase-labeled E0771 and 4T1 cells were maintained in RPMI 1640 medium supplemented with 10% FBS, penicillin/streptomycin, and 300 µg/ml hygromycin.

We purchased a pGIPZ vector containing an shRNA against the mouse SRC-3 gene (Thrmo Fisher, catalog number: V2LMN_24526). The pGIPZ vector containing a scrambled nontargeting (NT) shRNA (Thermo Fisher, catalog number: RHS4346) was used as the control. Lentivirus vectors expressing shSRC-3 and the NT shRNA were produced in 293 TN cells by transiently transfecting pGIPZ carrying shSRC-3 (or NT shRNA) and the ViraSafe™ lentiviral packaging system with Lipofectamine 2000. The recombinant lentivirus titer was measured using Lenti-X™ GoStix™ Plus. Luciferase-labeled E0771 and 4T1 cells were transduced with lentiviral vectors expressing shSRC-3 (or nontarget shRNA) with TransDux MAX™ and then selected in the presence of 1 µg/ml puromycin. The efficiency of SRC-3 KD in luciferase-labeled E0771 and 4T1 cells was determined by performing Western blotting with an SRC-3 antibody (rabbit, Cell Signaling Technology, catalog number: 2126; 1:1,000 dilution), and the results were compared to parental luciferase-labeled E0771 cells.

A total of 1 × 10⁵ luciferase-labeled E0771 cells, SRC-3 KD luciferase-labeled E0771 cells, and luciferase-labeled E0771 cells expressing NT shRNA were resuspended in 50 µl of phosphate-buffered saline (PBS). They were then injected into one of the 4^th mammary fat pads in female C57BL/6J mice (8 weeks old), female B6 albino (8 weeks old), or female C57BL/6J-SCID mice (8 weeks old), as described in previous studies [17]. For 4T1 cells, 1 × 10⁵ luciferase-labeled 4T1 cells, SRC-3 KD luciferase-labeled 4T1 cells, and luciferase-labeled 4T1 cells expressing NT shRNA were injected into one of the 4th mammary fat pads in female BALB/cJ female mice (8 weeks old). Then, tumor growth was monitored by evaluating luciferase activity using an in vivo imaging system (IVIS). When the luciferase activity was 1000 luciferase photons/second/cm² (~ 7 days after injection), the mice were randomized into four groups (n = 6 mice/group). Mice in the treatment group received an intraperitoneal injection of SI-2 (2.5, 5, or 10 mg/kg, twice a day for 35 days) or vehicle (phosphate-buffered saline containing 10% DMSO, 30% PEG300, and 1% Tween 20). Luciferase activity in luciferase-labeled E0771 and 4T1 breast tumors was evaluated twice a week for 5 weeks using IVIS.

Mice were anesthetized with a 1.5% isoflurane/air mixture using an inhalation anesthesia system (VetEquip). D-Luciferin (Xenogen, catalog number: 122799) was injected intraperitoneally at a dosage of 40 mg/kg mouse body weight. Ten minutes after the D-luciferin injection, mice were imaged using the IVIS (Xenogen) under continuous exposure to 1% to 2% isoflurane. Imaging parameters were maintained for comparative analysis. Grayscale and pseudocolor images showing bioluminescence were superimposed and analyzed using Living Image software (Version 4.4, Xenogen). A region of interest (ROI) was manually selected over the relevant areas showing a luciferase signal. The ROI area was kept constant across the experiments, and the luciferase activity was recorded as total luciferase photon counts/second/cm² within the ROI.

E0771 cells and 4T1 cells (1 × 10⁶ cells) were injected into the mammary fat pads of female mice (8 weeks old) to induce breast tumor formation. The tumor-injected female mice were randomly divided into two groups after detecting tumor growth using luciferase activity imaging (1000 luciferase photons/second/m2). Then, the mice in one group were treated with SI-2 (2.5 mg/kg, twice a day for 35 days), while mice in the control group were treated with vehicle alone.

Mice were considered deceased when the tumor had 20 × 10⁶ luciferase photons/second/cm² because the volume of a tumor with this luciferase activity is approximately 20 mm in diameter, which is the maximum allowable tumor size according to our institution’s animal care guidelines. Kaplan–Meier survival analysis results were plotted using GraphPad Prizm (Version 8.0).

Fresh mouse tumor tissues were fixed with 10% buffered formalin and then embedded in paraffin for subsequent sectioning. Tissue slices were deparaffinized by sequential soaking in xylene, ethanol, and water. Antigen retrieval was performed with antigen unmasking solution (Vector Laboratory, catalog number: H-3300). Endogenous peroxidase activity was blocked with a 3% (vol/vol) hydrogen peroxide solution. Sections were blocked with 2.5% (wt/vol) normal goat serum to reduce nonspecific antibody binding. Slices were incubated with primary antibodies, including anti-cleaved caspase 3 (rabbit, Cell Signaling Technology, catalog number: 9664, 1:300 dilution), anti-Ki-67 (rabbit, Cell Signaling Technology, catalog number: 9027; 1:1,000 dilution), anti-CD4 (rabbit, Abcam, catalog number: ab183685, 1:300 dilution), anti-CD8 (rabbit, Abcam, catalog number: ab217344, 1:300 dilution), anti-CD56 (rabbit, Novus Biologicals, catalog number: NBP2-38452, 1:300 dilution), anti-IFNG (rabbit, Novus Biologicals, catalog number: NBP2-66900, 1:300 dilution), anti-Cxcl9 (rabbit, Thermo Fisher Scientific, catalog number: PA5-81371, 1:100 dilution), and anti-Foxp3 (rabbit, Invitrogen, catalog number: PA1-16876, 1:300 dilution) antibodies, at 4 °C overnight. Slices were stained with HRP polymer-conjugated secondary antibodies (Vector Laboratories, catalog number: 7401 for rabbit antibodies). Slices were reacted with a freshly prepared DAB solution (Dako, catalog number: K3468). After staining, nuclear counterstaining was performed with Mayer's hematoxylin.

Tissue slices were deparaffinized by sequential soaking in xylene, ethanol, and water. Antigen retrieval was performed with antigen unmasking solution (Vector Laboratory, catalog number: H-3300). Slides were blocked with 2.5% (wt/vol) normal goat serum to reduce nonspecific antibody binding. Slides were incubated with anti-CD4 (rabbit, Abcam, catalog number: ab183685, 1:100 dilution) and anti-FOXP3 (rat, LSBio, catalog number: LS-C344878, 1:100 dilution) antibodies at 4 °C overnight. Afterward, the slides were washed with TBST [20 mM Tris (pH 7.5), 150 mM NaCl and 0.1% (w/v) Tween 20]. They were then incubated with goat anti-rabbit Alexa Fluor 488 (Thermo Fisher, catalog number: A11008, 1:500 dilution) and goat anti-rat Alexa Fluor 594 (Thermo Fisher, catalog number: A11007, 1:500 dilution) for 1 h at room temperature. Afterward, slides were washed with TBST and mounted with Antifade Mounting Medium containing DAPI (Vector, catalog number: H-2000).

Primary antibodies against the following proteins were used: ERα (Santa Cruz Biotechnologies, catalog number: SC-71064, 1:1000 dilution), ERβ (Santa Cruz Biotechnologies, catalog number: SC-373853, 1:1000 dilution), PR (Santa Cruz Biotechnologies, catalog number: SC-7208, 1:1000 dilution), HER-2 (Cell Signaling, catalog number: 2165, 1:1000 dilution), SRC-3 (Cell Signaling, catalog number: 2126, 1:1000 dilution), and β-Actin (Santa Cruz Biotechnologies, catalog number: SC-47778, 1:1000 dilution). Membranes containing proteins were incubated with a goat anti-rabbit IgG-HRP antibody (Sigma, catalog number: A0545, 1:5000 dilution), and the signals were visualized using an ECL plus kit (Thermo Scientific, catalog number: 11517271).

E0771 breast tumors were isolated from mice on the 21st day after SI-2 or vehicle treatment. Afterward, cell lysates were generated from breast tumors treated with SI-2 and vehicle using lysis buffer [20 mM Tris HCl (pH 7.5), 150 mM NaCl, 10% glycerol, 1% Nonidet P-40 (NP-40), and 2 mM EDTA]. Additionally, SRC-3 KD E0771 breast tumors and their control E0771 breast tumors were harvested on the 14th day after tumor cell injection. Cell lysates were generated from SRC-3 KD E0771 breast tumors and their control tumors with lysis buffer. According to the manufacturer's protocols, cytokine levels in cancers were determined with a Proteome Profiler Mouse Cytokine Antibody Array Kit (R&D Systems, catalog number: ARY006).

E0771 cells were inoculated into the wells of 96-well plates (1 × 10⁴ cells/well). The next day, each cell line was treated with serially diluted IL-1RA (0–200 ng/ml) and TIMP-1 (0–200 ng/ml). After 3 days, 10 μL of MTS reagent in Premix WST-1 Cell Proliferation Assay system (Takara, catalog number: MK400) was added to each well. MTS-treated plates were incubated for 2 h. Then, the optical density of the color in each well was measured at 450 nm using a microtiter plate reader.

E0771, 4T1, and MDA-MB-486 cells were plated into 96-well plates (1 × 10⁴ cells/well). The next day, each cell line was treated with serial dilutions of SI-2. On the 3rd day after SI-2 treatment, the proliferation of breast cancer cells was determined using the MTS assay.

SRC-3 KD and KD control E0771 cells were grown in RPMI 1640 containing 10% (vol/vol) FBS and 1 μg/ml puromycin. Total RNA was isolated from cells using a RNeasy Kit (Qiagen, catalog number: 74004). cDNA was generated from 1 μg of total RNA from cells using a high-capacity reverse transcription kit (Applied Biosystems, catalog number: 4368814). The expression levels of target genes were determined by quantitative PCR with SsoAdvanced Universal SYBR Green Supermix (BIO-RAD, catalog number: 1725271) and target gene primers as follows: Cxcl9 (5’-ACTCCAACACAGTGACTCAATAG -3' and 5’-CGTTCTTCAGTGTAGCAATGATTT-3'), Il-1ra (5’-CAAGCCTTCAGAATCTGGGATA-3', and 5’-CTCAGAGCGGATGAAGGTAAAG-3'), and Il-16 (5’-GCGTCAGTCATCTCCAACATAG-3' and 5’-CCAACACCTGCCTCTTTCTT-3').

To determine the role of SRC-3 in the modulation of TIME, we employed an E0771 murine mammary cancer cell line that was initially isolated from a spontaneous tumor in a C57BL/6 mouse [18]. Luciferase-labeled E0771 cells were generated by a lentivirus carrying a luciferase expression cassette to noninvasively evaluate tumor growth in mice. Since luciferase expression does not alter metabolism in cancer cells [19], the luciferase activity measured from luciferase-expressing E0771 breast tumors facilitated the practical quantification of tumor development in recipient mice. As shown in our previous study, continuous treatment of mice with SI-2 (2 mg/kg, twice a day, for 5 weeks) suppressed MDA-MB-468 breast cancer progression in severe combined immunodeficiency (SCID) mice without inducing toxicities in internal organs [11]. When cancer growth was evaluated by detecting luciferase activity (~ 1000 luciferase photons/second/cm²), E0771 breast cancer-bearing C57BL/6J female mice were treated with SI-2 (2.5 mg/kg, twice a day for 35 days) and drug vehicle as the control based on our previously established xenograft mouse model[11]. Comparative analyses of the luciferase activity images revealed that the E0771 cells robustly formed breast tumors in syngeneic C57BL/6J mice-administered vehicle treatment (Fig. 1A, B). However, SI-2 treatment significantly reduced imaging-based luciferase activity in E0771 breast tumors compared to vehicle-treated breast tumors (Fig. 1A, B). Therefore, SI-2 treatment effectively suppressed the growth of E0771 breast tumors in immune-intact syngeneic C57BL/6J mice. All E0771 breast cancer-bearing mice treated with vehicle were euthanized on 35 days after cancer cell injection because tumors were reached the maximum size of tumors (> 20 mm in diameter) (Fig. 1C, D). However, compared to vehicle-treated mice, SI-2 (2.5 mg/kg)-treated mice had a small tumor size but did not have the maximum tumor size (< 10 mm in diameter) on the 35th day after the injection of cancer cells (Fig. 1C, D). Collectively, this result substantiates that compared to vehicle treatment, SI-2 treatment effectively suppressed the growth of E0771 breast tumors in immune-intact C57BL/6J mice.

In addition to evaluating luciferase activity in images, we harvested tumors from animals treated with SI-2 and vehicle on the 35th day after tumor cell injection and determined tumor volumes. We then conducted a histological analysis of the harvested tissue. Consistent with the luciferase activity in the tumors, the analysis of the tumor volume revealed that compared to vehicle treatment, SI-2 treatment significantly reduced the tumor volume (2.1-fold, p = 0.023) (Fig. 1D). Hematoxylin and eosin (H&E) staining revealed SI-2 treatment-induced hypocellularity in cancers compared with vehicle treatment because an organized structure was not detected in the tumor masses treated with SI-2 (Fig. 1E). SI-2 effectively suppressed breast cancer progression in immune-intact mice.

E0771 cells from primary breast tumors metastasize to the lungs of C57BL/6J mice [20]. Thus, we determined whether SI-2 inhibited lung metastasis of primary E0771 breast tumors that developed in C57BL/6J mice. An analysis of luciferase activity images from our cancer-bearing mice revealed that luciferase activity from E0771 cells was detected in the lungs on the 35th day after vehicle treatment (Fig. 1F). However, unlike the vehicle treatment, luciferase activity in E0771 cells was not detected in the lungs of E0771 tumor-bearing mice on the 35th day after SI-2 treatment (Fig. 1F). We isolated lungs from cancer-bearing mice treated with SI-2 or vehicle and performed H&E staining of the tissue to validate the difference in lung metastasis identified by imaging luciferase activity. H&E staining analysis with the serial section of the lung revealed that tumor masses were detected in the lungs of cancer-bearing mice treated with vehicle, but cancer masses were not detected in the lungs of mice treated with SI-2 (Fig. 1G). In addition to suppressing the growth of primary breast tumors, SI-2 also effectively inhibited lung metastasis of E0771 breast tumors in C57BL/6J mice.

In contrast to the low-dose-treated animals (2.5 mg/kg of SI-2), mice were euthanized 10 days after administration of the high dose of SI-2 (10 mg/kg) because mice were notably sick (Fig. 1C). Compared to administration with 10 mg/kg, however, administration of 5 mg/kg SI-2 did not lead to sickness in these tumor-bearing mice. Moreover, the cancer-suppressive activity of 5 mg/kg SI-2 was less than that of 2.5 mg/kg SI-2 (Additional file 1: Fig. S1A). To understand this observation, the cytotoxicity of SI-2 was determined in C57BL/6J female mice (8 weeks old) treated with vehicle or 2.5, 5, or 10 mg/kg SI-2 twice a day for one week. Comparative analysis of the cytokine profile of blood in these animals revealed that 10 mg/kg SI-2 treatment significantly increased the levels of various cytokines in blood compared to the vehicle-treated animals (Fig. 1H and Additional file 2: Fig. S2). Therefore, the cytotoxicity of 10 mg/kg SI-2 leads to sickness in mice. SI-2 (5 mg/kg) also elevated the levels of cytokines in the blood compared to the vehicle, even though the fold changes in cytokines induced by 5 mg/kg SI-2 were less than those induced by 10 mg/kg SI-2 (Fig. 1H and Additional file 2: Fig. S2). However, 2.5 mg/kg SI-2 increased the levels of only three cytokines in the blood compared to the vehicle (Fig. 1H and Additional file 2: Fig. S2). 5 mg/kg of SI-2 slightly increased CD4 + T cells, but did not change the level of CD8 + T cells in breast tumors compared to 2.5 mg/kg of SI-2 (Additional file 1:Fig. S1B, C). Interestingly, 5 mg/kg of SI-2 significantly reduced CD56 + NK cell levels in E0771 breast tumors in C57BL/6J mice compared to 2.5 mg/kg of SI-2 (Additional file 1:Fig. S1D). Therefore, the high levels of blood cytokines likely reduce antitumor activity upon administration with 5 mg/kg of SI-2 compared to administration of 2.5 mg/kg of SI-2 by altering the tumor immune microenvironment.

According to a previous study, SI-2 effectively suppressed MDA-MB-468 breast cancer progression in SCID mice by inhibiting tumor cell proliferation [11]. In immune-intact C57BL/6J mice, SI-2 treatment also significantly reduced the percentage of proliferating Ki67-expressing cells compared to their vehicle-treated counterparts (Fig. 2A). Furthermore, compared with vehicle treatment, SI-2 treatment increased apoptosis in MDA-MB-468 cells in immunodeficient mice to inhibit tumor growth [11]. Therefore, we wanted to investigate whether SI-2 also increased apoptosis in aggressive E0771 breast tumors in C57BL/6J mice. Immunohistochemistry (IHC) with an anti-cleaved caspase 3 antibody revealed that SI-2 treatment significantly increased the percentage of cleaved caspase 3-expressing cells compared to vehicle treatment (Fig. 2B). Thus, SI-2 treatment promoted apoptosis in E0771 breast tumors to suppress breast tumor growth in immune-intact mice.

The differential TIME suppresses or stimulates breast cancer progression [21]. Thus, we wanted to determine whether SI-2 treatment altered the infiltrating immune cell repertoire in E0771 breast tumors to suppress their growth. Immunohistochemical staining revealed that SI-2 treatment increased the numbers of infiltrating CD4+, CD8+, and CD56+ cytotoxic immune cells in E0771 tumors compared to vehicle treatment (Fig. 3A–C). Cytotoxic CD4⁺ T cells directly or indirectly eliminate MHC class II-positive tumor cells [22]. CD8 + cytotoxic T cells also suppress MHC class I-positive tumor cells [23]. Therefore, SI-2 treatment suppresses E0771 breast tumor progression in immune-intact mice by recruiting CD4+ and CD8+ T cells into tumors. Hematopoietic expression of CD56 appears to be restricted to activated immune cells, such as natural killer (NK) cells, which exhibit cytotoxic activity that suppresses cancer progression [24, 25]. Cytotoxic immune cells produce interferon-gamma (Ifng) that enhances their cytotoxicity [26]. To define whether the recruited CD4+ and CD8+ T cells in breast tumors function as cytotoxic immune cells, Ifng levels were determined. IHC revealed that SI-2 treatment elevated the levels of Ifng in tumors compared to vehicle treatment (Fig. 3D). Therefore, SI-2 promoted the establishment of an antitumor immune environment at least in part by recruiting cytotoxic CD4+, CD8+ T cells, and CD56+ cells into the tumor tissue.

In contrast to its effects on cytotoxic immune cells, SI-2 treatment significantly reduced the number of Foxp3+ cells in E0771 breast tumors in C57BL/6J mice compared to vehicle-treated mice (Fig. 3E). Foxp3 is the crucial transcription factor that establishes regulatory T (Treg) cell identity and drives their immunosuppressive activity [27]. In addition to Treg cells, previous studies revealed that Foxp3 is also expressed in various cancers and promotes tumor progression by activating cancer-specific cellular pathways, such as the Wnt/β-catenin signaling pathway and EMT [28, 29]. Thus, we need to validate whether the reduction in Foxp3 expression correlated with a lower number of Treg cells. We conducted dual immunofluorescence staining to address this question. Both CD4 + /Foxp3 + and CD4-/Foxp3 + cells were detected in control E0771 breast tumors (Fig. 3F). CD4 + /Foxp3 + cells are Tregs, and CD4-/Foxp3 + cells might be E0771 breast cancer cells expressing Foxp3. Compared to the control, SI-2 treatment significantly reduced the numbers of both CD4 + /Foxp3 + and CD4-/Foxp3 + cells in E0771 breast tumors. Therefore, SI-2 treatment increased the ratio of cytotoxic lymphocytes to Tregs in E0771 tumors by reducing the Treg cell population, likely leading to enhanced antitumor immunity capable of limiting tumor progression.

We employed the 4T1 tumor-bearing syngeneic BALB/cJ mouse model to determine whether SI-2 could suppress the progression of another breast tumor line in a syngeneic mouse model of breast cancer [30]. First, we determined the expression levels of nuclear receptors, HER2, and SRC-3 in E0771 cells, 4T1 cells, and MDA-MB-468 cells. E0771 and 4T1 cells expressed estrogen receptor (ER)α, ERβ, and progesterone receptor (PR) at higher levels than MDA-MB-468 cells (Fig. 4A). Additionally, the SRC-3 expression levels in E0771 cells were similar to those in 4T1 cells, but HER-2 expression in E0771 cells was lower than that in 4T1 cells (Fig. 4A). Accordingly, the molecular profile of mouse E0771 and 4T1 cells was distinct from human MDA-MB-468 cell lines. Afterward, we determined the half-maximal inhibitory concentrations (IC₅₀) of SI-2 for E0771, 4T1, and MDA-MB-468 cells. SI-2 suppressed the growth of all breast cancer cells (Fig. 4B). The growth of E0771 cells was suppressed by a low concentration of SI-2 (IC₅₀: 4.5 nM) compared to 4T1 (IC₅₀: 51 nM) and MDA-MB-468 cells (IC₅₀: 21 nM) (Fig. 4B).

Luciferase-labeled 4T1 cancer cells were orthotopically injected into the mammary fat pad of immune-intact BALB/cJ female mice to determine whether SI-2 treatment also suppressed 4T1 breast tumor progression in syngeneic immune-intact BALB/cJ female mice in vivo. Animals were then intraperitoneally injected with 2.5 mg/kg SI-2 (twice a day) and the vehicle. SI-2 treatment significantly reduced the luciferase activity of 4T1 breast tumors in BALB/cJ mice compared to vehicle treatment (Fig. 4C, D). After the final luciferase imaging analysis, we harvested 4T1 breast tumors and determined their volumes. Consistent with the luciferase activity imaging analysis, SI-2 treatment significantly reduced 4T1 breast tumor volume compared to the vehicle (Fig. 4E, F). In addition to E0771 cells, SI-2 effectively suppressed 4T1 breast cancer progression in immune-intact female mice.

Tumor-infiltrating immune cells in 4T1 breast tumors were also assessed to determine whether SI-2 treatment changes the TIME. Treatment with 2.5 mg/kg SI-2 elevated the levels of CD4 + and NK cells in 4T1 breast tumors compared to the vehicle controls (Fig. 5A, C). However, CD8 + levels were not changed in 4T1 breast tumors by SI-2 treatment compared to vehicle controls (Fig. 5B). SI-2 treatment significantly elevated Ifng levels in 4T1 breast tumors compared with vehicle controls (Fig. 5D). Collectively, SI-2 treatment also recruited cytotoxic immune cells (CD4 + T cells and NK cells) and then generated a tumor-suppressing TIME to suppress 4T1 breast tumor progression in immune-intact female mice.

Our previous study revealed that SI-2 treatment reduced SRC-3 levels in E0771 cells in vitro [31]. Thus, we examined SRC-3 levels in E0771 breast tumors in C57BL/6J mice treated with SI-2 compared with vehicle. SI-2 treatment significantly reduced SRC-3 levels in E0771 breast tumors compared to the vehicle (Fig. 6A).

Next, we determined whether SRC-3 plays an essential role in E0771 breast cancer progression in immune-intact mice. We generated stable SRC-3 knockdown (KD) luciferase-labeled E0771 cells using a lentivirus expressing an shRNA against mouse SRC-3 and then determined the effect of the loss of SRC-3 function on luciferase-labeled E0771 breast tumor progression. The shRNA against SRC-3 significantly reduced SRC-3 levels in E0771 cells compared to parental E0771 cells and E0771 cells expressing a nontargeting (NT) shRNA (Fig. 6B). Afterward, SRC-3 KD luciferase-labeled E0771 cells were orthotopically injected into the mammary fat pads of recipient C57BL/6J female mice. As a control, luciferase-labeled E0771 cells expressing the NT shRNA were injected into the mammary fat pads of recipient C57BL/6J female mice. The luciferase activity of SRC-3 KD E0771 breast tumors was significantly lower than that of control E0771 breast tumors (Fig. 6C, D). After the final luciferase activity analysis, we harvested SRC-3 KD and control E0771 breast tumors and determined their volumes. The SRC-3 KD E0771 breast tumor volume was significantly lower than that of control E0771 breast tumors (Fig. 6E, F). SRC-3 played a critical role in E0771 breast cancer progression in immune-intact mice.

Next, we determined the composition of tumor-infiltrating immune cells in SRC-3 KD E0771 breast tumors. IHC analysis revealed that the levels of CD4 + T cells, CD8 + T cells, and NK cells were significantly elevated in SRC-3 KD E0771 breast tumors compared to the control E0771 breast tumors (Fig. 7A–C). Additionally, Ifng levels were significantly increased in SRC-3 KD E0771 breast tumors compared to the control E0771 breast tumors (Fig. 7D). These results indicate that SRC-3 has a critical role in generating a tumor-enhancing immune microenvironment for breast cancer progression.

We also determined the effect of SRC-3 KD on 4T1 breast tumor progression in syngeneic immune-intact BALB/cJ female mice. SRC-3 shRNA significantly reduced SRC-3 protein levels in 4T1 cells compared to NT shRNA control cells (Fig. 8A). SRC-3 KD effectively inhibited 4T1 breast tumor progression in BALB/cJ female mice compared to KD control 4T1 cancer cells (Fig. 8B and C). After the final luciferase activity analysis, we harvested SRC-3 KD and control 4T1 breast tumors and determined their volumes. The SRC-3 KD 4T1 breast tumor volume was significantly lower than that of control 4T1 breast tumors (Fig. 8D, E). Therefore, SRC-3 is required for 4T1 breast tumor progression in immune-intact BALB/cJ female mice.

Although SI-2 treatment activated the host immune system to generate an antitumor immune response that suppressed E0771 breast cancer progression, we sought to obtain direct evidence supporting the hypothesis that SI-2-mediated suppression of E0771 breast cancer progression specifically requires host immune system function. Therefore, we employed C57BL/6J-SCID mice that carry the severe combined immunodeficiency mutation due to a spontaneous mutation of the PRKDC gene on the C57BL/6J background and thus lack functional T or B cells to answer this question [32]. In contrast to its effects on immune-intact C57BL/6J mice, SI-2 (2.5 mg/kg) treatment did not suppress E0771 tumor growth in C57BL/6J-SCID mice compared to vehicle-treated mice (Fig. 9A, B). However, a high dose of SI-2 (10 mg/kg) suppressed E0771 tumor growth in C57BL/6J-SCID mice compared to vehicle-treated mice (Fig. 9C, D).

To define the host immune effect of SI-2-mediated suppression of 4T1 breast tumor progression in immune-intact mice, 4T1 cells were orthotopically injected into female SCID mice (8 weeks old). SI-2 (2.5 mg/kg) treatment did not suppress the growth of 4T1 breast tumor progression in SCID mice compared to vehicle treatment, similar to E0771 breast tumors (Fig. 9E, F). Therefore, the low dose of SI-2 (2.5 mg/kg)-mediated breast tumor suppression required the host’s immune system.

SRC-3 KD suppressed E0771 breast tumor progression in immune-intact mice (Fig. 6). We determined whether breast tumor suppression mediated by SRC-3 KD occurred in immune-deficient mice. SRC-3 KD E0771 and KD control E0771 cells were orthotopically injected into SCID female mice. Compared to immune-intact mice, however, SRC-3 KD E0771 cells developed robustly into breast tumors in SCID mice compared to control E0771 cells (Fig. 9G, H). In contrast to immune-intact BALB/cJ female mice, SRC-3 KD did not suppress 4T1 breast tumor progression in SCID mice compared to NT shRNA control mice (Fig. 9I, J). Therefore, the host immune system is required to suppress SRC-3 KD breast tumor progression.

We conclude that a low dose of SI-2-mediated inhibition of breast tumor growth and suppression of SRC-3 KD breast tumor progression is a host immune system-dependent process.

Cytokines mediate communication between immune and nonimmune cells to control innate and adaptive immune responses and suppress or enhance cancer progression [33]. To address whether SRC-3 inhibition changes the cytokine profile in breast tumors, E0771 tumors were harvested from tumor-bearing C57BL/6J mice on the 23rd day after SI-2 or vehicle treatment (Fig. 10A). Mouse cytokine profiling revealed that compared to vehicle controls, SI-2 treatment significantly increased the levels of complement component 5 (C5)/C5a and interleukin 1 receptor antagonist (Il-1ra) in E0771 breast tumors (Fig. 10B, C). Therefore, the elevation of Il-1ra by SI-2-mediated SRC-3 might suppress breast cancer progression because Il-1ra has tumor-suppressive activity [34]. Thus, we determined whether the increase in Il-1ra levels also suppressed the proliferation of E0771 breast cancer cells. E0771 cells were treated with different doses of recombinant human Il-1ra proteins. Compared to vehicle, Il-1ra effectively suppressed the proliferation of E0771 cells (Fig. 10D). As a control, we measured the effect of tissue metalloprotease inhibitor-1 (TIMP-1) on the proliferation of E0771 cells because SI-2 treatment did not significantly increase TIMP-1 expression in E0771 tumors compared to vehicle treatment (Fig. 10C). In contrast to Il-1ra, TIMP-1 did not suppress the proliferation of E0771 cells compared to vehicle-treated cells (Fig. 10D). Therefore, SI-2 increased Il-1ra levels in E0771 cancer cells to suppress their proliferation.

We next determined the cytokine profiles in SRC-3 KD and control E0771 breast tumors to address whether SRC-3 KD also changes the cytokine profile in breast tumors. Consistent with the cytokine array using SI-2-treated E0771 breast tumors, Il-1ra levels were increased in SRC-3 KD E0771 breast tumors compared to control breast tumors (Fig. 10E, F). Therefore, SRC-3 KD increased the expression of Il-1ra in breast tumors, leading to the suppression of E0771 breast cancer progression. Furthermore, Il-16 (2.7-fold) and C-X-C motif chemokine ligand 9 (Cxcl9, ninefold) levels were significantly increased in SRC-3 KD E0771 breast tumors compared to control breast tumors (Fig. 10E, F). IHC analysis revealed that Cxcl9 levels were elevated in SI-2-treated E0771 breast tumors (Fig. 10G) and SRC-3 KD E0771 breast tumors (Fig. 10H) compared with their control breast tumors. However, immune cells also produce Cxcl9, Il-1ra, and Il-16, while it is not clear whether SRC-3 KD breast tumors express these cytokines. To address this question, we cultured SRC-3 KD E0771 cells and control E0771 cells and then compared the levels of Cxcl9, Il-1ra, and Il-16 between them. RNA analysis revealed that SRC-3 KD enhanced the expression levels of Cxcl9, Il-1ra, and Il-16 in E0771 cells compared to control cells (Fig. 10I–K). Cxcl9 mainly regulates the recruitment of C-X-C motif chemokine receptor 3 (Cxc3R)-positive immune cells, such as CD4 + T cells, CD8 + T cells, natural killer (NK) cells, and macrophages [35]. Therefore, the targeting of SRC-3 by SI-2 and SRC-3 KD recruits Cxcr3-expressing cytotoxic immune cells, such as CD4 + , CD8 + , and CD56 + (NK cells), into breast cancers by increasing Cxcl9 and then generates an antitumor immune microenvironment to suppress breast cancer progression (Fig. 11) [36, 37].