Fig 2: NRBP1 regulates Rac1/Cdc42 activation in TNBC via P-Rex1.A NRBP1 overexpression enhances activation of Rac1 and Cdc42. Results of PAK-1 PBD pull-down assays for active Rac1/Cdc42 using MDA-MB-231_EcoR cells expressing mycBioID2-NRBP1 or the empty vector. Cell lysates and PAK-1 PBD pulldown assays were Western blotted as indicated. Densitometry results of Rac1/Cdc42 activation are shown in Supplementary Fig. 4A. Data are representative of n = 3 independent experiments. B Stable knockdown of NRBP1 reduces activation of Rac1 and Cdc42. Vector control and NRBP1 knockdown MDA-MB-231_HM cells were treated with doxycycline prior to lysis. Cell lysates and PAK-1 PBD pulldown assays were subjected to Western blotting as indicated. Densitometry results of Rac1/Cdc42 activation are shown in Supplementary Fig. 4B. Data are representative of n = 3 independent experiments. C, D NRBP1-mediated activation of Rac1 and Cdc42 is P-Rex1-dependent. MDA-MB-231-shNRBP1 cells were complemented with shRNA-resistant NRBP1 in the presence or absence of siRNA-mediated P-Rex1 knockdown. Parental cells+/− P-Rex1 knockdown were also included as controls. Cell lysates and PAK-1 PBD pulldown assays were then Western blotted as indicated. Confirmation of P-Rex1 knockdown was obtained by Western blotting of P-Rex1 IPs. Representative blots are shown in (C), and densitometry results of Rac1/Cdc42 activation are shown in (D). The PAK1-PBD band is obtained as a cross-reacting signal when blotting with the Rac1 antibody and serves as a loading control for the PAK1- PBD protein. Data are representative of n = 3 independent experiments. Rac1/Cdc42 activation was normalized to total Rac1/Cdc42, which was normalized to β-actin. Data are expressed relative to parental cells with si-Control, which was arbitrarily set at 100%. Error bars represent the standard error of the mean from n = 3 independent experiments. NS indicates p > 0.05, *p < 0.05, **p < 0.01 by two-way ANOVA with Tukey’s multiple comparisons test. E Association of NRBP1 with P-Rex1, Rac1 and Cdc42. MFM-223 cells were transfected with Flag-tagged NRBP1. Anti-Flag IPs were then Western blotted as indicated.

Fig 3: Characterization of the NRBP1 interactome by BioID-MS.A Schematic workflow of BioID-MS. A promiscuous biotin ligase (BirA*) is fused to the protein of interest (the ‘bait’ protein, here NRBP1) and the fusion protein expressed in cells. Addition of excess biotin results in biotinylation of endogenous proteins in the proximity of the bait. After affinity pulldown using streptavidin agarose beads, the biotinylated proteins are identified by MS. The figure was adapted from “BioID Assay”, using BioRender.com (2022). B The NRBP1 interactome in TNBC. A volcano plot representation of the MS dataset with cut-off of adjusted p-value ≤ 0.05 and Log2 fold change ≥ 1. Data are generated from n = 3 independent experiments. Labelled proteins represent previously-identified interactors or novel ones associated with cytoskeletal or redox regulation. C Functional pathways associated with NRBP1 interactors. The graph shows the top ten functions of interactors identified by “Metascape” software. D Confirmation of NRBP1-P-Rex1 interaction. Flag-tagged NRBP1 or vector control were transfected into MFM-223 cells. Total cell lysate (TCL) and Flag IPs were subjected to Western blotting as indicated.

Fig 4: The P-Rex1/Rac1 axis plays essential roles in cellular functions regulated by NRBP1.A Western blotting validation of NRBP1 knockdown, NRBP1 rescue and P-Rex1 knockdown. MDA-MB-231 cells with shRNA-mediated NRBP1 knockdown were complemented with expression of shRNA-resistant NRBP1 in the presence or absence of P-Rex1 knockdown. Parental cells+/− P-Rex1 knockdown were also included as controls. Total cell lysates or P-Rex1 IPs were Western blotted as indicated. B Rescue of cell migration by NRBP1 is P-Rex1 dependent. Cells validated in (A) were subjected to transwell assays. C Expression of Rac1_G12V. MDA-MB-231 cells with shRNA-mediated NRBP1 knockdown were programmed to express GFP-Rac1_G12V. Cell lysates were Western blotted as indicated. D Effect of active Rac1 on monolayer proliferation of NRBP1-depleted cells. Cells validated in (C) were subjected to MTS assays. Data were obtained at Day 4. E Effect of active Rac1 on invasion of NRBP1-depleted cells. Cells validated in (C) were subjected to transwell assays. For B, D and E, error bars represent the standard error of the mean from n = 3 independent experiments. NS indicates p > 0.05, *p < 0.05, **p < 0.01, ***p < 0.001 by two-way ANOVA with Tukey’s multiple comparisons test.

Fig 5: NRBP1 knockdown inhibits TNBC cell growth.A Confirmation of inducible knockdown. Western blotting of lysates from MDA-MB-231_HM cells (top panel) and MFM-223 cells (bottom panel) expressing different doxycycline-inducible NRBP1 shRNAs. Cells were treated with doxycycline or PBS control for 48 h (MDA-MB-231_HM cells) or 72 h (MFM-223 cells) prior to lysis. B NRBP1 knockdown reduces colony formation by MDA-MB-231_HM cells. Left panel, images of plates at Day 11. Right panel, quantification of colony size. C NRBP1 knockdown reduces colony formation by MFM-223 cells. Graph shows quantification of colony size at Day 23. Representative images are shown in Supplementary Fig. 2B. Error bars represent the standard error of the mean from n = 3 independent experiments. NS indicates p > 0.05, *p < 0.05, **p < 0.01 by one-way ANOVA with Tukey’s multiple comparisons test. D NRBP1 represents a breast cancer context-dependent fitness gene. Results for CRISPR-Cas9 screens across different breast cancer cell lines, extracted from the DepMap portal. A loss of fitness score < 0 indicates a statistically significant decrease in proliferation after NRBP1 knockdown in comparison with controls and therefore increased dependency of the cell line for NRBP1 [12].