Fig. 6: The R310A mutation that blocks Klk8-mediated hemicleavage of Activin-A abrogates its tumor-promoting activity.

A Schematic of the mutations in proActivin-A used to block hemicleavage by KLK8 (R310A), or to inhibit both KLK8 and furin (mS1) (left), and of the strategy to analyze their influence on Activin-A induced tumor growth (right). Created in BioRender. Bulliard, M. (2025) https://BioRender.com/f04n549. B Growth curves of syngeneic B16-F1 melanoma grafts expressing control lentivirus (Ctrl), or wild-type (WT) βA, or the R310A, or mS1 cleavage site mutant derivatives. Data show the average of ten mice per group ± SEM (Two-way ANOVA). C Activin-A Western blot of 2 representative tumor lysates per genotype at the experimental endpoint. D Changes in body (top) and heart weights (bottom) between experimental start- and endpoints. Data show the average of ten mice per group (One-way ANOVA). E Mature Activin-A detected by ELISA in plasma at the experimental endpoint. Data show the amount of mature Activin-A in pg/mL per mg of tumor in 3 to 5 tumors per genotype (One-way ANOVA). F Western blot analysis confirming the depletion of Activin-A in plasma by both the mS1 and R310A mutations. Two tumor-bearing mice were analyzed per genotype. G Schematic of Activin-A processing in the absence of KLK8 and its effect on tumor growth. Created in BioRender. Bulliard, M. (2025) https://BioRender.com/w96o575. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.