Fig. 7

Pharmaceutical inhibition of eIF4A3 blocks bone-metastasis of BC cells in vivo. a IHC analysis (left) and quantification (right) of EIF4A3 expression in 20 normal breast tissues and 331 clinical BC tissues, including 295 primary BC tissues (237 non-BM/BC and 58 BM/BC) and 36 bone-metastatic BC tissues (at bone). b Kaplan–Meier analysis of bone metastasis-free survival curves in patients with BM/BC with low vs high expression of EIF4A3 (n = 58; P < 0.001, log-rank test). c Left: Two representative specimens are shown. Scale bars, 10 μm. Right: Percentages of specimens showing low or high EIF4A3 expression in relative to the levels of circIKBKB, nuclear-p65, M-CSF and GM-CSF. d Left and middle: BLI and μCT images (left) and histological (H&E and TRAP) images (middle) of bone lesions from mice, which received vehicle or eIF4A3-IN-2 treatment started at 2 days after intracardial injection of SCP2 cells. Right: Quantification of osteolytic sites and TRAP⁺ osteoclasts along the bone-tumor interface of metastases from experiment in left and middle panel. Scale bar, 50 μm. e Left and middle: BLI and μCT images (left) and histological (H&E and TRAP) images (middle) of bone lesions from mice, which vehicle or eIF4A3-IN-2 was started when bone-metastatic tumors formed. Right: quantification of osteolytic sites and TRAP⁺ osteoclasts along the bone-tumor interface of metastases from experiment in left and middle panel. f Model: EIF4A3-mediated circIKBKB overexpression activated NF-κB/bone remodeling factors signaling, resulting in inducing formation of bone pre-metastatic niche and breast cancer metastasis, and treatment with EIF4A3 inhibitor eIF4A3-IN-2 might serve as a promising approach to inhibit BC bone-metastasis. Each error bar represents the mean ± SD of three independent experiments. * P < 0.05, ** P < 0.01, *** P < 0.001