Fig 2: BQU57 blocks TNBC growth in vitro and in vivo. a Western blot and quantification of GTP-bound RALB in MDA-MB-231 cells after BQU57 (50μM) or DMSO treatment for 24 h. b Quantification of MDA-MB-231 and MVT1 cell viability upon treatment with varying doses of BQU57 for 72 h. c Quantification of MDA-MB-231 and MVT1 growth in low adhesion (GILA) conditions upon treatment with varying doses of BQU57 for 5 days. d Quantification of MDA-MB-231 orthotopic mammary tumor volume over time [n = 9 DMSO; n = 10 BQU57 treatment (50mg/kg by i.p. injection, M-F)]. e Representative H&E images (left) and quantification (right) of spontaneous lung metastases in MDA-MB-231 tumor bearing mice following treatment (arrows point to metastatic lesions). Scale bars = 500μm. f Quantification of subcutaneous TNBC PDX tumor volume over time. [n = 9 DMSO; n = 9 BQU57 treatment (50mg/kg, M-F)]. g Quantification of MDA-MB-231 cell viability upon treatment with BQU57 (100μM) or DMSO in combination with various doses of paclitaxel for 72 h. h RALA expression, but not RALB expression, is predictive of BC patient response to chemotherapy. Data from ROC Plotter. Data are presented as mean ± SEM; (*), P < 0.05

Fig 3: RALA is elevated in TNBC and is prognostic of overall survival in ER-negative disease. a Western blots demonstrating RALA and RALB expression in a panel of BC cell lines denoted by molecular subtype. b RALA and RALB mRNA expression in a panel of 46 BC cell lines categorized according to molecular subtype. Data from the Broad Institute Cancer Cell Line Encyclopedia (CCLE), (*) P < 0.05. c Box and whisker plots of RALA and RALB expression in normal breast tissue (n = 61), BC subtypes other than TNBC (n = 300) , and TNBC (n = 49). Data from TCGA Research Network. (*) P < 0.05. d Box and whisker plots of RALA and RALB expression in normal breast tissue (n = 144), BC subtypes other than TNBC (n = 1725), and TNBC (n = 250). Data from METABRIC. (*) P < 0.05. e Left: Kaplan-Meier analysis segregates all BC patients by RALA (upper quartile vs lower three-quartiles) and RALB (spilt along the median) expression in METABRIC and TCGA datasets wherein high RALA is prognostic of worse disease specific survival (DSS) in the METABRIC (P = 0.0079) and overall survival in the TCGA (P = 0.0003) cohorts and low RALB is prognostic of worse DSS in the METABRIC cohort (P = 0.0311), but is not prognostic in the TCGA cohort (P = 0.9477). Right: Within the TNBC population of the METABRIC cohort, RALA (upper tenth percentile vs lower 90th percentail) is prognostic of worse DSS TNBC patients (P = 0.0544) while RALB (spilt along the median) is not prognostic (P = 0.9297). Significance determined by log-rank. f Top: Representative high and low RALA immunostaining of BC patient samples. Bottom: Kaplan-Meier analysis segregating all BC patients (P = 0.0417) and TNBC patients (P = 0.0440) by RALA H-score (upper tertile vs lower two tertiles) where high RALA is prognostic of worse overall survival (Scale bars = 60μm)

Fig 4: RALA supports, while RALB opposes, in vitro measures of aggressive cancer phenotypes in TNBC cell lines. a Representative images and quantification of MDA-MB-231 Ctrl, RALA-KO, and RALB-KO cell migration over 6 h (scale bar = 100μm). b Representative images and quantification of MDA-MB-468 siRNA control (siCtrl), siRALA, and siRALB cell migration over 6 h (scale bar = 100μm). c Representative images and quantification of MVT1 shCtrl and shRALA cell migration over 6 h (scale bar = 200μm). d Representative images and quantification of MDA-MD-231 shCtrl and shRALA scratch assay over 6 h (scale bar = 500μm). e Representative images and quantification of MDA-MB-231 Ctrl, RALA-KO, and RALB-KO cell invasion through matrigel coated transwell inserts over 24 h (scale bar = 200μm). f Representative images and quantification of MDA-MB-231 Ctrl, RALA-KO, and RALB-KO cell invasion as three dimensional spheroids over 5 days (scale bar = 400μm). g Quantification of viability for MDA-MB-231 Ctrl, RALA-KO, and RALB-KO cells ± siCtrl or siRALA/B transient knockdown over 72 h. h Quantification of viability for MDA-MB-468 siCtrl, siRALA and siRALB cells over 72 h. i Quantification of BrdU incorporation for MDA-MB-231 Ctrl, RALA-KO, and RALB-KO cells over 72 h. j Quantification of Annexin V positivity for MDA-MB-231 Ctrl, RALA-KO, and RALB-KO cells. k Quantification of growth in low adhesion (GILA) conditions for MDA-MB-231 Ctrl, RALA-KO, and RALB-KO cells ± siCtrl or siRALA/B transient knockdown over 5 days. h Quantification of growth in low adhesion (GILA) for MDA-MB-468 siCtrl, siRALA and siRALB cells over 5 days. Data are presented as mean ± SEM; (*), P < 0.05

Fig 5: RALA, not RALB, is pro-tumorigenic in preclinical in vivo models of TNBC utilizing short hairpin knockdown of RALA/B. a Western blots demonstrating RALA and RALB expression in MDA-MB-231 shRNA control (shCtrl), shRALA, and shRALB cells. b Comparison of MDA-MB-231 shCtrl (n = 5) and shRALA (n = 5) orthotopic mammary tumor growth. c Comparison of MDA-MB-231 shCtrl (n = 7) and shRALB (n = 8) orthotopic mammary tumor growth. d Representative Ki67 immunostaining images and H-score quantification of MDA-MB-231 shCtrl (n = 4) and shRALA (n = 4) orthotopic mammary tumors. e Representative cleaved caspase 3 (CC3) immunostaining images and H-score quantification of MDA-MB-231 shCtrl (n = 4) and shRALA (n = 4) orthotopic mammary tumors. f Quantification of Rala mRNA expression in MVT1 shCtrl and shRALA cells by qRT-PCR (left) and growth of orthotopic mammary tumors (right; n =6 per group). Inset shows western blot confirmation of RALA protein expression in MVT1 shCtrl and shRALA cells. Data are presented as mean ± SEM; (*), P < 0.05; scale bars = 20μm