Fig 2: RNF167 mediates K29-linked polyubiquitination and degradation of CASTOR1 in response to growth factors.a Kinetics of CASTOR1 protein level and activation status of AKT and mTORC1 following fetal bovine serum (FBS) deprivation in 293T cells. b Kinetics of CASTOR1 protein level and activation status of AKT and mTORC1 following arginine deprivation in 293T cells. c CASTOR1 protein level and activation status of AKT and mTORC1 following deprivation of FBS, arginine, or both or treatment with AKT inhibitor MK2206 in 293T cells. FBS or arginine deprivation or AKT inhibitor treatment was carried out for 24 h. d CASTOR1 ubiquitination status following deprivation of FBS, arginine, or both. e CASTOR1 was labeled by K29-linked polyubiquitination. An ubiquitin mutant K29 contained only the K29 lysine residue was sufficient to cause CASTOR1 polyubiquitination while mutation of K29 (K29R) abolished CASTOR1 polyubiquitination. f, g Ectopic expression of RNF167 increased (f), whereas knockdown of RNF167 decreased (g) CASTOR1 ubiquitination. h, i RNF167 knockdown increased (h), whereas RNF167 overexpression decreased (i) CASTOR1 protein level. j Schematic depiction of the K29-marked polyubiquitination and degradation of CASTOR1 protein by RNF167 in response to FBS. Blots in a–i are representatives of n = 3 independent experiments. Source data are provided in Source data file.

Fig 3: AKT-mediated phosphorylation and RNF167-mediated ubiquitination of CASTOR1 release mTORC1 inactivation by CASTOR1.a CASTOR1 bound to MIOS in a dose-dependent manner. b Overexpression of RNF167 decreased CASTOR1 protein level and activated mTORC1 with and without the presence of arginine. c Overexpression of a myristoylated constitutively active AKT1 (myr) but not the kinase-dead AKT1 mutant (K179M) reduced CASTOR1 protein level, decreased its binding to MIOS, and activated mTORC1. d AKT regulated mTORC1 activity by suppressing CASTOR1 was independent of TSC2. e–g CASTOR1 S14D had a weaker binding to MIOS shown by CASTOR1 co-immunoprecipitation (co-IP) of MIOS (e) and reversed MIOS co-IP of CASTOR1 (g), hence S14D had a less inhibitory effect on mTORC1 than WT and S14A had (e, g), and quantification of results from two independent experiments shown in f. h An illustration depicting that AKT phosphorylation and RNF167 ubiquitination of CASTOR1 reverse CASTOR1 inactivation of mTORC1. Blots in a–d and g are representatives of n = 3 independent experiments, and blots in e are representatives of n = 2 independent experiments. Source data are provided in Source data file.

Fig 4: RNF167-mediated ubiquitination and AKT1-mediated phosphorylation of CASTOR1 promotes breast cancer progression.a, b Weaker suppression of colony formation of ER+ (a) and HER2+ (b) breast cancer cells in softagar by CASTOR1 S14D than WT and S14A. c CASTOR1 silencing enhanced colony formation in softagar of T47D cells. d–g Overexpression of CASTOR1 S14D had a less suppressive effect on breast tumor growth and a lower extended animal survival rate than WT and S14A had in a breast cancer xenograft model; the tumor volumes at the indicated time point post-inoculation were measured (d); the tumor volumes of the last time point were compared (e), and the actual tumors (f) and the survival rates (g) are shown. h–j CASTOR1 knockdown promoted tumor growth and shortened animal survival rate. The tumor volumes at the indicated time point post-inoculation were measured (h); and the tumor volumes of the last time point (i) and the survival rates (j) are shown. The lower panels of a–c were quantifications of colony numbers from three independent experiments presented as mean ± SEM and analyzed by one-way ANOVA. For d, e, g, each mouse group contains 20 tumors (n = 20). For h, j, the Vector control, shRNA1, and shRNA2 groups contain 15, 11, and 10 tumors, respectively (n = 15, 11, and 10). For i, the Vector control, shRNA1, and shRNA2 group contain 6, 7, and 5 tumors, respectively (n = 6, 7, and 5). For e, i, the boundary closest to the zero indicates the 25th percentile, a line within the box means the median, and the boundary of the box farthest from zero marks the 75th percentile. Whiskers (error bars) above and below the box indicate the minima and maxima. d, e, h, i were presented as mean ± SEM and analyzed by two-sided Student’s t test. g, j was analyzed by two-sided Log-rank test. “*” and “***” denote P < 0.05 and P < 0.001, respectively. Scale bars: 200 µM (a–c). Source data are provided in Source data file.

Fig 5: High CASTOR1 protein level overrides arginine activation of mTORC1 in physiological conditions.a Response of mTORC1 activation to CASTOR1 overexpression in a dose-dependent manner with and without the presence of arginine in 293T cells. High level of CASTOR1 overrode arginine-mediated mTORC1 activation. b CASTOR1 protein expression levels in multiple cell types, including human lobar bronchial epithelial cells (HLBEC), human small airway epithelial cells (HSAEC), 293T, HeLa, and breast cancer cell lines MCF7 and T47D. c, d HeLa cells, which had almost no detectable CASTOR1 protein and a high level of constitutively activated mTORC1, was minimally responsive to arginine regulation of mTORC1, including arginine deprivation for 80 min (c) and re-stimulation for 10 min following arginine deprivation for 50 min (d). e MCF7 cells were more responsive than T47D cells to arginine-mediated mTORC1 activation, which was inversely correlated with their CASTOR1 protein levels (b). f Cells with high endogenous CASTOR1 protein levels including HSAEC and HLBEC (b) were not responsive to arginine-mediated mTORC1 activation. g CASTOR1 knockdown in T47D cells, which had a high level of endogenous CASTOR1 protein (b), activated mTORC1. h Summary of the relative endogenous CASTOR1 protein expression levels in different types of cells and their responsiveness to arginine regulation of mTORC1. Blots in a–g are representative of n = 3 independent experiments. Source data are provided in Source data file.