HSP90 promotes tumor associated macrophage differentiation during triple-negative breast cancer progression

Abstract

Tumor-associated macrophages (TAMs) originating from monocytes are crucial for cancer progression; however, the mechanism of TAM differentiation is unclear. We investigated factors involved in the differentiation of monocytes into TAMs within the tumor microenvironment of triple-negative breast cancer (TNBC). We screened 172 compounds and found that a heat shock protein 90 (HSP90) inhibitor blocked TNBC-induced monocyte-to-TAM differentiation in human monocytes THP-1. TNBC-derived conditional medium (CM) activated cell signaling pathways, including MAP kinase, AKT and STAT3, and increased the expression of TAM-related genes and proteins. These inductions were suppressed by HSP90 inhibition or by knockdown of HSP90 in TNBC. Additionally, we confirmed that TNBC secreted HSP90 extracellularly and that HSP90 itself promoted TAM differentiation. In a mouse tumor model, treatment with an HSP90 inhibitor suppressed tumor growth and reduced TAMs in the tumor microenvironment. Our findings demonstrate the role of HSP90 in TAM differentiation, suggesting HSP90 as a potential target for TNBC immunotherapy due to its regulatory role in monocyte-to-TAM differentiation.

Fig. 1

Triple-negative breast cancer cells induced monocyte-to-TAM polarization. (a) Morphological features are visualized by microscopy. Scale bar; 30 μm. (b) TAM-related mRNA expression was determined by real-time RT-PCR. Data are expressed as the ratio to 18 S RNA (n = 6). (c) TAM-associated protein expression incubated with CM from MDA-MB-231 was examined by FACS analysis (n = 4). Data are expressed as mean fluorescence intensity (MFI). (d) TAM-related mRNA expression of THP-1 after coculturing with MDA-MB-231 is determined by real-time RT-PCR (n = 6). *p < 0.05, ***p < 0.005, ****p < 0.001. (c) and (d) were analyzed by two-tailed unpaired Student’s t-test. (b) was analyzed with one-way ANOVA with Dunnett’s post hoc test. All error bars represent SEM.

Fig. 2

HSP90 inhibitor suppressed MDA-MB-231-induced polarization of THP-1 monocytes. (a) Summary of screening test for a kinase inhibitor kit. Data are expressed as the ratio of CD163 mRNA expression to control. (b) THP-1 cell viability after incubation with ganetespib or radicicol (n = 6). (c) Effect of radicicol and ganetespib on MDA-MB-231-induced polarization into TAM is determined by real-time RT-PCR (n = 6). (d) MDA-MB-231 induced TAM-associated protein expression gated on live cells with or without ganetespib (1 µM) by FACS analysis (n = 6). Data are expressed as MFI. *p < 0.05, ***p < 0.005, ****p < 0.001. (b), (c) and (d) were analyzed with one-way ANOVA with Dunnett’s post hoc test. All error bars represent SEM.

Fig. 3

HSP90 inhibitor suppressed MDA-MB-231-induced cell signaling activation. (a) Representative immunoblot with indicated antibodies. THP-1 cells were incubated with CM for 24 h in the presence or absence of HSP90 inhibitors (radicicol; 5 µM, ganetespib; 1 µM). (b) Representative immunoblot with indicated antibodies. THP-1 cells were incubated with CM for 24 h with different doses of ganetespib (0.05–1 µM). (c) Time course of cell signaling in THP-1 cells incubated with CM in the presence or absence of ganetespib (1 µM). Representative data from two or three independent experiments.

Fig. 4

HSP90 directly promoted monocyte-to-TAM differentiation. (a) Morphological features are visualized by microscopy. Scale bar; 10 μm. (b) TAM-related protein expression, after treatment with TGFβR2 and HSP90, is determined by FACS (n = 6). Data are expressed as MFI gated on live singlets. (c) TAM-related mRNA expression determined by real-time RT-PCR. THP-1 cells were incubated with HSP90 (12.5–50 µg/mL) for 5 d. Data are expressed as the ratio to 18 S RNA (n = 6). (d) Time course of cell signaling (WB) of HSP90-induced THP-1 polarization within 3 h. (e) M2 like macrophage-related expression of HSP90-derived THP-1 is determined by real-time RT-PCR (n = 6). *p < 0.05, **p < 0.01, ***p < 0.005, ****p < 0.001. (b), (c), and (e) were analyzed with one-way ANOVA with Dunnett’s post hoc test. All error bars represent SEM.

Fig. 5

Deficiency of HSP90 in MDA-MB-231 reduced TAM differentiation and cell signaling activation. (a) HSP90 concentrations in CM from MDA-MB-231, MDA-MB-453 and MCF7 determined by ELISA (n = 3). Total protein content was divided by cell number of each subtype. (b) HSP90 mRNA expression in control-siRNA (20 nM) or HSP90-siRNA (10 or 20 nM) transfected MDA-MB-231 determined by RT-PCR. Data are expressed as the ratio to 18 S RNA (n = 6). (c) Representative immunoblot with indicated antibodies. MDA-MB-231 cells was transfected with HSP90-siRNA or control-siRNA. (d) HSP90 concentration in CM from control- or HSP90-siRNA treated transfected MDA-MB-231 determined by ELISA (n = 5). Total protein content was divided by cell number of each subtype. (e) TAM-related mRNA expressions of THP-1 after incubation with control- or HSP90-siRNA transfected MDA-MB-231 CM was determined by real-time RT-PCR. Data are expressed as the ratio to 18 S RNA (n = 6). (f) Representative immunoblot with indicated antibodies. THP-1 monocytes were incubated with CM from control- or HSP90-siRNA transfected MDA-MB-231 for 6 h. *p < 0.05, ****p < 0.001. (a), (b), (d) and (e) were analyzed with one-way ANOVA with Dunnett’s post hoc test. All error bars represent SEM.

Fig. 6

HSP90 inhibitor suppressed tumor growth in murine syngeneic tumor model. (a) (left) Co-culture of 4T1 cells and BM monocytes. (right) TAM-related mRNA expression is determined by real-time RT-PCR. Bone marrow cells were cocultured with 4T1 in the presence of HSP90 inhibitor (0.01 or 0.1 µM) for 5 d. Data are expressed as the ratio to 18 S RNA (n = 6). (b) Experimental outline for (c), (d), and (f). (c) BALB/c mice were subcutaneously injected with 4T1 cells (4 × 10⁵ cells per mouse), followed by tail-vein injection of ganetespib twice a week (n = 9). Tumor volume was measured every 3 d. Representative of three independent experiments. (d) Tumor weights of the two groups of mice. (e) Representative western blot of 4T1 cell lysate with HSP90 antibody. (f) Immune cells in the tumor tissue were examined by FACS analysis. Quantification of the percentage of CD3⁺ CD4⁺ T cells among CD45⁺ cells (left) and PD-1 protein expression (MFI) of CD45⁺ CD3⁺ CD4⁺ T cells (right) in the tumor from ganetespib or control administered mice (n = 9). (g) Experimental outline for (h). (h) Immune cells in tumor tissue were examined by FACS analysis. Quantification of the percentage of CD45⁺ CD11b⁺ F4/80⁺ Ly6C⁻ macrophages among CD45⁺ cells (left) and CD45⁺ CD11b⁺ F4/80⁺ CD206⁺ TAM-like cells among CD45⁺ cells (right) in tumor tissues from ganetespib or control administered mice (n = 5). *p < 0.05, **p < 0.01, ***p < 0.005, ****p < 0.001. (c), (d), (f), and (h) were analyzed by two-tailed unpaired Student’s t-test. (a) was analyzed with one-way ANOVA with Dunnett’s post hoc test. All error bars represent SEM.