CCL2 blockade combined with PD-1/P-selectin immunomodulators impedes breast cancer brain metastasis

Abstract

Over the last two decades, the diagnosis and treatment of breast cancer patients have improved considerably. However, brain metastases remain a major clinical challenge and a leading cause of mortality. Thus, a better understanding of the pathways involved in the metastatic cascade is essential. To this end, we have investigated the reciprocal effects of astrocytes and breast cancer cells, employing traditional 2D cell culture and our unique 3D multicellular tumouroid models. Our findings revealed that astrocytes enhance the proliferation, migration and invasion of breast cancer cells, suggesting a supportive role for astrocytes in breast cancer outgrowth to the brain. Elucidating the key players in astrocyte-breast cancer cells crosstalk, we found that CCL2 is highly expressed in breast cancer brain metastases tissue sections from both patients and mice. Our in vitro and in vivo models further confirmed that CCL2 has a functional role in brain metastasis. Given their aggressive nature, we sought additional immune checkpoints for rationale combination therapy. Among the promising candidates were the adhesion molecule P-selectin, which we have recently shown to play a key role in the crosstalk with microglia cells and the co-inhibitory receptor PD-1, the main target of currently approved immunotherapies. Finally, combining CCL2 inhibition with immunomodulators targeting either PD-1/PD-L1 or P-selectin/P-Selectin Ligand-1 axes in our human 3D tumouroid models and in vivo presented more favourable outcomes than each monotherapy. Taken together, we propose that CCL2-CCR2/CCR4 is a key pathway promoting breast cancer brain metastases and a promising target for an immunotherapeutic combination approach.

Keywords: 3D cancer models; CCL2; P-selectin; PD-1; astrocytes; breast cancer brain metastasis.

Figure 1

Astrocytes and breast cancer cells engage in reciprocal interactions. (A and B) Proliferation rates after 96 h of (A) iRFP-labelled murine 4T1 cells in the presence or absence of murine astrocytes (n = 3) and (B) iRFP-labelled human MDA-MB-231 cells in the presence or absence of human astrocytes (n = 3). Data are represented as fold-change to time 0. Statistical significance was determined using one-way ANOVA and Dunnett’s multiple comparisons test. (C and D) Proliferation rates after 190 h of (C) murine astrocytes in the presence or absence of 4T1 conditioned medium (n = 4) and (D) human astrocytes in the presence or absence of MDA-MB-231 conditioned medium (n = 3). Data are represented as fold-change to time 0. (E and F) Quantification of tumour spheroids area after 48 h of (E) mCherry-labelled 4T1 cells in the presence or absence of murine astrocytes (n = 5) and (F) mCherry-labelled MDA-MB-231 cells in the presaence or absence of human astrocytes (n = 4). Data are presented as fold-change to time 0. (G and H) Representative images of tumour spheroids from E and F, respectively. Scale bar = 400 µm. (I) Quantification of mCherry-labelled MDA-MB-231 trans-endothelial cells and/or trans-astrocytic migration, presented by the MDA-MB-231 mCherry signal recorded on the surface level of the Transwell® membrane. Reduction in the fluorescent signal indicates the transmigration of the cells through the barrier (n = 1). (J) Trans-epithelial electrical resistance (TEER) values of a barrier consisting of induced pluripotent stem cell (iPSC)-derived brain microvascular endothelial cells (BMEC) and human pericytes, with or without human astrocytes and mCherry-labelled MDA-MB-231 measured 20 h after the addition of MDA-MB-231 cells. Data represented as % of time 0. Statistical significance was determined using one-way ANOVA and Tukey’s multiple comparisons test. (K–L) Illustration of the blood–brain barrier (BBB) models used in I and J, respectively. This image was created with BioRender.com. All data are expressed as mean ± standard error; Unless otherwise stated, statistical significance was determined using an unpaired Student’s t-test. CM = conditioned medium; hAstro = human astrocytes; mAstro = murine astrocytes.

Figure 2

CCL2–CCR2/4 axis is upregulated in human and murine breast cancer brain metastasis models and patient samples. (A and B) Cytokine arrays of media collected following 72 h of cell culturing from mono-cultures and co-cultures of (A) Murine 4T1 cells and murine astrocytes (n = 1) and (B) Human MDA-MB-231 cells and human astrocytes (n = 1). Membrane negative controls were subtracted, and data were normalized to membrane reference spots and expressed per million cells. The 10 most upregulated cytokines are presented for each cell line, determined by the fold-change in the co-culture compared to the average of the mono-cultures. Green indicates low expression, and red indicates high expression. (C and D) CCL2–CCR2/CCR4 (red), Hoechst (blue) staining of brain metastases, primary breast cancer, adjacent/normal breast and normal brain in (C) murine tissue sections and (D) human tissue sections. Scale bars = 100 µm. Representative images of selected fields (the highest expressing fields are presented). (E–G) Quantification of the immunostained murine tissues from C. Data are expressed as mean ± standard error of the mean (SEM). Statistical significance was determined using one-way ANOVA and Tukey’s multiple comparisons tests. (H–J) Quantification of the immunostained human tissues from D (H and I, n = 3 and J, n = 1–5 fields per sample). Data are expressed as mean ± SEM. Statistical significance was determined using one-way ANOVA and Tukey’s multiple comparisons tests for H and I and unpaired Student’s t-test for J. Dots on the graph in J represent technical repeats. (K–M) mRNA expression analysis of (K) CCL2, (L) CCR2 and (M) CCR4 on human primary breast tumours that metastasized to the brain, compared with primary tumours that metastasized to other organs. Data are presented as violin plots, showing median and quartiles. Statistical significance was determined using a one-tailed unpaired Student t-test. ANG = Angiogenin; Ang1 = Angiopoietin-1; Ang2 = Angiopoietin-2; BC = breast cancer; BCBM = breast cancer brain metastases; CFD = Complement Factor D; CHI3L1 = Chitinase 3-like 1; FLT3L = Flt-3 ligand; hAstro = human astrocytes; LCN2 = Lipocalin-2; mAstro = murine astrocytes; PTX2 = Pentraxin 2.

Figure 3

The CCL2–CCR2/CCR4 axis is upregulated in breast cancer brain metastases in vitro models. (A and B) CCL2 secretion measured by ELISA of media collected after 72 h of cell culturing, from mono-cultures and co-cultures of (A) Murine 4T1 cells and murine astrocytes (n = 3) and (B) human MDA-MB-231 cells and human astrocytes (n = 3). Data expressed as % of astrocyte secretion. (C and D) CCL2 mRNA expression measured by reverse transcription qPCR of RNA extracted from cancer cells and astrocytes: (C) 4T1 cells cultured in astrocyte medium (AM), astrocyte starvation medium (AM 0%, 0% fetal bovine serum) or murine astrocyte conditioned medium (CM); and murine astrocytes cultured in AM, AM 0% or 4T1 CM (one representative of n = 3; dots on the graph represent technical repeats). Data are expressed as fold-change compared to murine astrocytes in AM. (D) mCherry-labelled MDA-MB-231 cells and iRFP-labelled human astrocytes after sorting the cells by their fluorescence labels from either mono- or co-cultures (n = 5). Data are expressed as fold-change compared with the human astrocytes from monoculture. (E and F) Fluorescence-activated cell sorting (FACS) analysis of murine astrocytes for the intracellular CCL2 expression following exposure to 4T1 serum-free medium (SFM), 4T1 SFM supplemented with lipopolysaccharide (LPS) or 4T1 CM. The analysis was performed for two types of astrocytic populations: (E) Activated/reactive astrocytes [gated for GLAST negative (−), GFAP positive (+)] and (F) non-GFAP-activated astrocytes (gated for GLAST+, GFAP−). (G and H) FACS analysis of 4T1 cells in murine astrocyte medium (n = 3) of (G) CCR4 expression and (H) CCR2 expression. (I and J) FACS analysis of MDA-MB-231 cells in human astrocyte medium (n = 3) of (I) CCR4 expression and (J) CCR2 expression. (K–N) Quantification of mean fluorescence intensity of cells from G–J. All data are expressed as mean ± standard deviation. Unless otherwise stated, statistical significance was determined using one-way ANOVA and Tukey’s multiple comparisons tests. hAstro = human astrocytes; mAstro = murine astrocytes.

Figure 4

The CCL2–CCR2/CCR4 axis is important for breast cancer proliferation and invasion. (A and B) Proliferation rates after 160 h of (A) mCherry-labelled 4T1 cells in the presence or absence of murine astrocytes and CCL2 inhibitor bindarit (n = 3) and (B) mCherry-labelled MDA-MB-231 cells in the presence or absence of human astrocytes and CCL2 inhibitor bindarit (n = 3). Data are represented as fold-change to time 0. (C and D) Proliferation rates of (C) iRFP-labelled 4T1, shCCR4 or negative control sequence (shNC) cells in the presence or absence of murine astrocytes after 94 h (one representative of n = 2; dots on the graph represent technical repeats) and (D) iRFP-labelled MDA-MB-231, shCCR4 or shNC cells in the presence or absence of human astrocytes after 72 h (one representative of n = 2; dots on the graph represent technical repeats). Data are represented as fold-change to time 0. (E and F) Illustrations of the blood–brain barrier (BBB) models used in G and in H and I (respectively). This image was created with BioRender.com. (G) Trans-epithelial electrical resistance (TEER) values of a barrier consisting of induced pluripotent stem cell (iPSC)-derived brain microvascular endothelial cells (BMEC) and mCherry-labelled MDA-MB-231 cells, with or without the addition of recombinant CCL2 protein, measured 16 h after the addition of MDA-MB-231 cells (n = 3). Statistical significance was determined using an unpaired Student t-test. (H) TEER values of a barrier consisting of iPSC-derived BMEC, human pericytes, human astrocytes and mCherry-labelled MDA-MB-231 cells, with or without the addition of CCL2 inhibitor bindarit, measured 20 h after the addition of MDA-MB-231 cells (n = 3). Statistical significance was determined using an unpaired Student t-test. (I) Immunostaining of a barrier consisting of iPSC-derived BMEC, human pericytes, human astrocytes and mCherry-labelled MDA-MB-231 cells, with or without the addition of CCL2 inhibitor bindarit or recombinant CCL2 protein, measured 20 h after the addition of MDA-MB-231 cells. DAPI (cyan), tight junction marker ZO1 (green) and mCherry-labelled MDA-MB-231 cells (red). Scale bar = 20 µm. (J and K) Quantification of tumour spheroids area after 48 h of (J) mCherry-labelled 4T1 cells in the presence or absence of murine astrocytes and CCL2 inhibitor bindarit (one representative of n = 2; dots on the graph represent technical repeats, data are presented as fold-change to time 0) and (K) mCherry-labelled MDA-MB-231 cells in the presence or absence of human astrocytes and CCL2 inhibitor bindarit (n = 3) (data are presented as % of control). (L and M) Representative images of tumour spheroids from J and K (respectively). Scale bars = 500 µm. All data are expressed as mean ± standard deviation. Unless otherwise stated, statistical significance was determined using one-way ANOVA and Tukey’s multiple comparisons tests. mAstro = murine astrocytes; hAstro = human astrocytes.

Figure 5

P-selectin–P-selectin ligand and PD-1–PD-L1 axes are promising candidates for immunotherapeutic combination therapy together with CCL2 inhibition. (A–D) mRNA expression analysis of (A) P-selectin (SELP), (B) P-selectin ligand 1 (PSGL-1), (C) Programmed death-ligand 1 (PD-L1) and (D) Programmed cell death protein 1 (PD-1) on human primary breast tumours that metastasized to the brain, compared with those that metastasized to other organs (data from GSE12276). Data are presented as violin plots, showing the median and quartiles. Statistical significance was determined using a one-tailed non-paired Student’s t-test. (E and F) Quantification of tumour spheroids area after 120 h of mCherry-labelled MDA-MB-231 cells in the presence of human astrocytes, human microglia and peripheral blood mononuclear cells (PBMCs) treated with (E) bindarit and/or SELPi (one representative of n = 3; dots on the graph represent technical repeats) and (F) bindarit and/or PD-L1i (n = 4–5). Data are expressed as mean ± standard deviation. Statistical significance was determined using one-way ANOVA and Tukey’s multiple comparisons tests. (G) Representative images of spheroids from E and F. Scale bar = 300 µm. (H–J) Patient-derived breast cancer liver metastasis organoids co-cultured with human astrocytes, human microglia and PBMCs treated with bindarit and/or SELPi, PD-L1i and Pembrolizumab. (H) Confocal imaging of a representative control organoid stained for CK14 antibody (magenta), EdU-based proliferation marker (green) and DAPI (blue). Scale bar = 50 µm. (I) Quantification of CK14 intensity of immunostained organoids (n = 5 fields for each group). Data are presented as a box-and-whisker plot, with a line at the median and an error bar representing minimal and maximal values. Statistical significance was determined using one-way ANOVA and Tukey’s multiple comparisons tests. (J) Representative confocal images of the CK14 quantified organoids from H. CK14 (magenta), EdU-based proliferation marker (green). Scale bar = 200 µm. PD-L1i = programmed death-ligand 1 inhibitor; SELPi = p-selectin inhibitor.

Figure 6

Combination therapy of bindarit and either P-selectin inhibitor or αPD-1 antibody leads to favourable outcomes in orthotopic breast cancer brain metastases mouse model. (A) Timeline (days) of primary tumour inoculation, tumour resection, metastases induction (intracranial injection), treatments and follow-up. (B and C) Kaplan–Meier survival curves of the different treatment groups. Statistical significance was determined using a survival analysis of each group compared to the control group, with further adjustment of P-values using a Holm–Šídák analysis. Number of mice for survival analysis: control n = 4, bindarit n = 6, SELPi n = 7, αPD-1 n = 6, bindarit + SELPi n = 5, bindarit + αPD-1 n = 6. (D and E) Primary tumour recurrence and extracranial metastases development inhibition are presented as the kinetics of development over time within the groups. Number of mice for relapse analysis: control n = 11, bindarit n = 14, SELPi n = 14, αPD-1 n = 14, bindarit + SELPi n = 14, bindarit + αPD-1 n = 17. (F and G) Quantification of brain metastases size 10 days post intracranial injection. Data are expressed as mean ± standard error of the mean; statistical significance was determined using one-way ANOVA and Tukey’s multiple comparisons tests. Number of mice for tumour volume analysis: control n = 5, bindarit n = 7, SELPi n = 6, αPD-1 n = 6, bindarit + SELPi n = 7, bindarit + αPD-1 n = 5. (H) Body weight change is presented as the % of change compared with the initial weight prior to primary tumour inoculation. Data are expressed as mean ± standard error of the mean. αPD-1 = anti-programmed death-1 antibody; SELPi = P-selectin inhibitor.

Figure 7

Combination therapy of bindarit and either P-selectin inhibitor or αPD-1 antibody leads to reduced tumour proliferation, increased immune activation and reduced immune suppression. GFAP (red) + CCL2 (green), CCR2 (red), CCR4 (red), Iba1 (red), CD8 (red) + Caspase 3 (green), Ki67 (red), CD31 (red), PD-1 (red), PD-L1 (red), PSGL-1 (red) + SELP (green), Hoechst (blue) staining in brain metastases tissue sections from control and treated mice. Scale bars = 100 µm. Representative images of selected fields. αPD-1 = anti-programmed death-1 antibody; SELPi = p-selectin inhibitor.

Figure 8

Summary model. Illustration showing proposed immunotherapeutic combination approach—inhibiting breast cancer brain metastases interactions with the brain milieu. Inhibition of the three axes CCL2–CCR2/CCR4, SELP–PSGL-1 and PD-1–PD-L1 can prevent breast cancer brain metastases progression. By combining these treatments, favourable long-term effects could be achieved. This image was created with BioRender.com.