Mast Cells Retard Tumor Growth in Ovarian Cancer: Insights from a Mouse Model

Abstract

Ovarian cancer has the highest mortality rate among female reproductive tract malignancies. A complex network, including the interaction between tumor and immune cells, regulates the tumor microenvironment, survival, and growth. The role of mast cells (MCs) in ovarian tumor pathophysiology is poorly understood. We aimed to understand the effect of MCs on tumor cell migration and growth using in vitro and in vivo approaches. Wound healing assays using human tumor cell lines (SK-OV-3, OVCAR-3) and human MCs (HMC-1) were conducted. Murine ID8 tumor cells were injected into C57BL6/J wildtype (WT) and MC-deficient C57BL/6-Kit^W-sh/W-sh (Kit^W-sh) mice. Reconstitution of Kit^W-sh was performed by the transfer of WT bone marrow-derived MCs (BMMCs). Tumor development was recorded by high-frequency ultrasonography. In vitro, we observed a diminished migration of human ovarian tumor cells upon direct or indirect MC contact. In vivo, application of ID8 cells into Kit^W-sh mice resulted in significantly increased tumor growth compared to C57BL6/J mice. Injection of BMMCs into Kit^W-sh mice reconstituted MCs and restored tumor growth. Our data show that MCs have a suppressive effect on ovarian tumor growth and may serve as a new therapeutic target.

Keywords: mast cell-deficiency; mast cells; mouse model; ovarian carcinoma; ultrasonography.

Figure 1

MCs are present in human ovarian cancer samples in vivo. Representative toluidine blue-stained sections of four (i–iv) human ovarian cancer samples (n = 11) showing MCs (arrow) within the tumor tissue (magnification 20×, scale bar = 100 µm). Square shows higher magnification (magnification 100×).

Figure 2

MCs reduce the migration of human ovarian cancer cells in vitro. Overgrown area of SK-OV-3 (A) or OVCAR-3 (B) cells co-cultured with HMC-1 cells in different cell-to-cell ratios (1:1, 1:5) with or without transwell at 24 h and 48 h (n = 3). Results are presented as mean with SD. Statistical analysis was performed using a two-way ANOVA followed by Bonferroni’s multiple comparison test (* p < 0.05, ** p < 0.01).

Figure 3

MC presence in an ID8 cell-derived tumor. (A) Tumor volumes of ID8-derived tumors in C57BL/6J mice (n = 7) and PBS-treated C57BL/6J mice (n = 3) from day 0 until day 98 after s.c. ID8 cell injection. (B) Representative grey-scale images of in vivo tumor measurements (length, width) via ultrasonography in B-Mode at day 98. (C) Representative toluidine blue-stained section of an ID8 tumor showing MCs (purple) within the tumor tissue (scale bar = 100 µm). Results are presented as means with SD. Statistical analysis was performed using a two-way ANOVA followed by Bonferroni´s multiple comparison test (* p < 0.05, *** p < 0.001, **** p < 0.0001).

Figure 4

Significantly increased tumor growth in MC-deficient Kit^W-sh mice compared to WT. (A) Tumor volume in C57BL/6J (n = 7) and Kit^W-sh mice (n = 5) from day 0 until day 98 after tumor cell injection. (B) Representative grey-scale images of the ID8-derived tumors at day 77 in C57BL/6J and Kit^W-sh mice. Results are presented as means with SD. Statistical analysis was performed using a two-way ANOVA followed by Bonferroni´s multiple comparison test (* p < 0.05, *** p < 0.001).

Figure 5

Diminished tumor growth in BMMC-reconstituted Kit^W-sh mice. (A) Tumor volume in C57BL/6J (n = 12), Kit^W-sh (n = 5), and BMMC-reconstituted Kit^W-sh (n = 5) mice from day 0 until day 98 after tumor cell injection. (B) Representative grey-scale images of ID8-derived tumors at days 28, 70, and 98 in C57BL/6J, Kit^W-sh, and BMMC-treated Kit^W-sh mice. Results are presented as means with SD. Statistical analysis was performed using a two-way ANOVA followed by Bonferroni´s multiple comparison test (*^{, #} p < 0.05, ^## p < 0.01, *** p < 0.001; *, comparison of ID8 in C57BL/6J and ID8 in Kit^W-sh; #, comparison of ID8 in C57BL/6J and BMMC + ID8 in Kit^W-sh).

Figure 6

Absence or presence of MCs affects the tumor environment. Visualization of molecules analyzed with RT-PCR of tumor tissue of ID8-injected C57BL/6J (n = 9) and Kit^W-sh (n = 4), and BMMC-reconstituted Kit^W-sh mice (n = 3). Results are presented as heatmap showing medians. Statistical analysis was performed using a two-way ANOVA followed by Bonferroni´s multiple comparison test (* p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001).