Supplementary Figure 3: Angiogenesis in vitro and in vivo is unaffected by MSK1 levels in breast cancer cells.

(a) mRNA levels fold change of angiogenesis-associated genes ANGPT1, ANGPT2, ANGPTL4 and VEGF in DBM, cells upon MSK1 downregulation, n = 4. Two-tailed Wald test from a linear model in which group was included as explanatory variable. (b) mRNA levels fold change of angiogenesis-associated genes ANGPT1, ANGPT2, ANGPTL4 and VEGF in DBM, cells upon MSK1 knockout, n = 4. Two-tailed Student’s t-test. (c) Quantification of in vitro endothelial HUVEC cell tube formation assay upon treatment with conditioned medium derived from shControl, shMSK1#1 and shMSK1#2 DBM cells, n = 3. Multiple t-test across points comparing each condition to control (d) Quantification of in vitro endothelial HUVEC cell tube formation assay upon treatment with conditioned medium derived from wild-type (WT) and MSK1 knockout (KO) DBM cells, n = 3. Multiple t-test across points comparing each condition to control (e) Quantification of number of CD31 positive vessels in metastatic lesions formed by shControl (n = 12), shMSK1#1 (n = 13) and shMSK1#2 (n = 18) DBM cells (left). Representative images of CD31 staining (right). Scale bar, 50 µm. Two-tailed Wald test from a linear model in which group was included as explanatory variable. Experiment was performed once (f) Quantification of EdU-positive cells in latent lesions formed by DBM shControl (n = 19 ROIs from 3 limbs) and shMSK1#2 cells (n = 19 ROI from 3 limbs) and metastatic lesions formed by DBM shControl (n = 22 ROIs from 2 limbs) and shMSK1#2 (n = 11 ROI from 2 limbs). Two-tailed Mann–Whitney test. (g) Representative image of EdU staining in lesions during latency or metastatic phase. Experiment was performed once. Panels (a-d) show data as mean ± SEM from three biologically independent samples each of which is an average from three technical replicates and panel (e-g) shows data as whisker plots: mid-line, median; box, 25 – 75 percentile; whisker, min-max. ns (nonsignificant) P > 0.05, * P ≤ 0.05, ** P ≤ 0.01. Schaffer method was used for P-value correction when three or more groups are compared. Statistics source data and P values are provided in Supplementary Table 9.