Fig 2: CILK1 is critical for breast cancer cell proliferation. (A) Quantification of cell cycle percentage from flow cytometry analyses in MDA-MB-231 (upper) and BT549 (bottom) cells. (B) CCK8 assay was performed in vector control and CILK1-knockdown MDA-MB-231, BT549, T47D and BT474 cells. (C) Xenograft tumor growth curves were shown, mice were injected with control MDA-MB-231 cells (n=7) or CILK1-knockdown clones (sh-1, n = 5 and sh-2, n = 7). (D) Tumor weight was shown. (E) The photo of tumors was shown. (F) CCK8 assay was done to determine the effect of CILK1-overexpression on the proliferation of MCF12a (upper) and MCF7 (bottom) cells. (G) Relative cell proliferation index in DMSO- or paclitaxel-treated control or CILK1-overexpression cells was calculated based on crystal violet staining. (H) Caspase-3 and cleaved-caspase-3 levels were detected in MCF12a and MCF7 cells after treated with paclitaxel (5 nM) or DMSO for 48 hours. Error bars represent mean ± SD, 3 independent experiments in triplicate were performed, data were analyzed by unpaired Student t-test in Prism. *p < 0.05, **p < 0.01, ***p < 0.001.

Fig 3: CILK1 directly phosphorylates ERK1. (A) Volcano plots of the quantitative TMT‐based proteomic analysis, which indicated MAPK3 as one of the differential phospho-proteins. The data of CILK1-knockdown or CILK1-C30-treatment were shown respectively. Blue and red dots represented significantly down- and up-regulated phospho-proteins (p < 0.05). (B) Western blot was performed to detect the expression levels of ERK1/2 and phospho-ERK1/2 in sgRNA (left) or shRNA (right)-mediated CILK1-knockdown 231R cells. (C) Detection of the expression levels of ERK1/2 and phospho-ERK1/2 in CILK1-C28/30 treated cells. (D) Western blot was done to detect the expression levels of CILK1, ERK1/2 and phospho-ERK1/2 in control and CILK1-knockdown BT549 and MDA-MB-231 cells, which treated with vehicle or trametinib (10 nM) for 30 min. (E) Western blot was performed to detect the expression levels of CILK1, ERK1/2 and phospho-ERK1/2 in control and CILK1-overexpression MCF7 cells, which treated with vehicle or trametinib (10 nM) for 30 min. (F) CILK1 kinase assay was done using purified ERK1 or ERK2 protein as substrate. (G) Phosphorylation site of ERK1 protein catalyzed by CILK1 was detected by mass spectrum. (H) Inhibitory IC50 values of paclitaxel were measured by CCK8 assay in cells pre-treated with GDC-0994 (100 nM) for 48 h, calculated by GraphPad. (I) CCK8 assay was done in vector control or CILK1-overexpression MCF7 cells, treated with DMSO or GDC-0994 for 72 h. Statistical data (mean ± SD) were shown. Error bars represent mean ± SD, 3 independent experiments in triplicate were performed, data were analyzed by unpaired Student t-test in Prism. *p < 0.05, **p < 0.01, ***p < 0.001.

Fig 4: Pharmacological inhibition of CILK1 in TNBC mice models. (A) NSG mice bearing PDX tumors were administered with CILK1-C30 (10 mg/kg, i.p) every two days, for a total of five injections; or administered with paclitaxel (10 mg/kg, i.p) every three days, for a total of four injections; or administered (i.p) with vehicle control. Growth curves of PDX tumors during 11 days of treatment with vehicle control, or paclitaxel (PTX), or CILK1-C30 as well as PTX/CILK1-C30 combination were shown. (B) Balb/c mice bearing MDA-MB-231-derived tumors were administered with CILK1-C30 (10 mg/kg, i.p) every two days, for a total of four injections; or administered with paclitaxel (10 mg/kg, i.p) every three days, for a total of three injections; or administered (i.p) with vehicle control. Growth curves of xenograft tumors during 8 days of treatment with vehicle control, or paclitaxel (PTX), or CILK1-C30 as well as PTX/CILK1-C30 combination were shown. (C) Immunoblot analysis and quantification of the indicated proteins from MDA-MB-231-derived tumors. Quantification of proteins was presented in normalized against Actin. Data are presented as mean ± SD. *p < 0.05, **p < 0.01, ***p < 0.001. (D) Representative images and quantification of the IHC staining of Ki-67 and cleaved caspase-3 (CC3) in MDA-MB-231-derived tumors. Data are presented as mean ± SD. **p < 0.01 by t-test. (E) Quantification and representative immunohistochemistry images of the indicated proteins in MDA-MB-231-derived tumors. Scale bar represents 40 μm. Results are mean ± SD of at least nine mice per group. *p < 0.05, **p < 0.01, ***p < 0.001.

Fig 5: Elevated CILK1 expression confers resistance to chemotherapy. (A) Scheme of the paclitaxel treatment timeline and time-point of tumor tissue collection in clinical breast cancer patients (top). IHC staining of CILK1 before and after paclitaxel treatment in one typical breast cancer patient (left); IHC-based quantification of CILK1 expression in tumor tissues of breast cancer patients (n = 12) before and after PTX treatment (right). PTX: Paclitaxel. (B) IHC staining of phospho-CILK1 (Tyr-159) before and after paclitaxel treatment in one typical breast cancer patients (left); IHC-based quantification of phospho-CILK1 (Tyr-159) expression in tumor tissues of breast cancer patients (n = 12) before and after PTX treatment (right). (C) CILK1 levels were analyzed in MDA-MB-231 and 231R cells by western blot (upper) and PCR (bottom) in response to paclitaxel (10 nM) or DMSO treatment for 24 hours. (D) Quantification of cell cycle percentage in control and CILK1-silencing 231R cells by flow cytometry. (E) CCK8 assay was performed to determine the effect of shRNA-mediated CILK1-knockdown on the proliferation of 231R cells. (F) Annexin V-FITC flow cytometry assay was used to measure the pro-apoptotic effect of paclitaxel in vector control and CILK1-silencing 231R and BT549 cells. Error bars represent mean ± SD, and the experiments were repeated three times. A representative experiment was shown, *p < 0.05, **p < 0.01, ***p < 0.001.