Fig 2: NF-?B pathway is responsible for Morgana dependent cancer cell metastasis. a, b Luciferase assays of AP-1 (left) and NF-?B (right) activity in a MDA-MB-231 infected with an empty vector (EMPTY) or shRNAs targeting Morgana (shMORG1 and shMORG2) and b MCF7 infected with an empty vector (EMPTY) or overexpressing Morgana (MORGANA). Cells were transfected with an AP-1 or NF-?B luciferase reporter or a control vector. c, d Gene expression analysis by Real-time PCR of NF-?B target genes in c MDA-MB-231 EMPTY or shMORG1 and shMORG2, and d MCF7 EMPTY or MORGANA, treated or not with 10nM TNFa for 4 h. e The top 10 gene ontology (GO) terms by enrichment P-value among the down-regulated genes obtained in a microarray analysis of MDA-MB-231 cells silenced for Morgana compared to control cells. The stars indicate signatures NF-?? related. f Immunoblotting of I?Ba, Morgana and Vinculin in MDA-MB-231 cells EMPTY or shMORG1 and shMORG2 in combination or not with I?Ba shRNA. g Luciferase assays of NF-?B activity in MDA-MB-231 cells described in f transfected with a NF-?B luciferase reporter or control vector. h Representative pictures of lungs of NSG mice after 4 weeks from the intravenous injection of MDA-MB-231 cells described in f. i Quantification of metastasis colonies in the lungs of mice described in h (n = 6 NSG mice per group). j Tumor weight 5 weeks after subcutaneous injection of cells described in f (n = 3 NSG mice per group). k, l Representative haematoxylin and eosin stained sections k and percentage of lung metastatic area l. Each data point represents one mouse (n = 3 NSG mice per group). Data are the results of three independent experiments. Bars in graphs represent standard errors (**p < 0.01; ***p < 0.001)

Fig 3: Morgana-NF-?B axis in breast cancer cells induces neutrophil recruitment in primary tumor. a Tumor weight 30 days after subcutaneous injection of 4T1 cells EMPTY or shMORG1 and shMORG2 infected or not with I?Ba shRNA (n = 3 BALB/c mice per group). b Percentage of immune cells (gated on CD45+ cells) detected in the primary tumor 30 days after injection of 4T1 described in a (n = 3 BALB/c mice per group). c Percentage of neutrophils detected in the primary tumor of each mouse 30 days after injection of 4T1 cells described in a. d Immunoblotting of Morgana, I?Ba and Vinculin in 4T1 cells described in a. e Luciferase assays of NF-?B activity in 4T1 cells described in a transfected with a NF-?B luciferase reporter or control vector. f Tumor weight 10 days after subcutaneous injection of 4T1 cells EMPTY or shMORG1 and shMORG2 (n = 3 BALB/c mice per group). g Percentage of immune system cells (gated on CD45+ cells) detected in the primary tumor 10 days after injection of 4T1 EMPTY or shMORG1 and shMORG2 (n = 3 BALB/c mice per group). h Percentage of neutrophils detected in the primary tumor of each mouse 10 days after injection of 4T1. Bars in graphs represent standard errors (*p < 0.05; **p < 0.01; ***p < 0.001)

Fig 4: Schematic model based on our study showing an essential role for Morgana in connecting IKK complex and its substrate I?Ba. The consequent activation of NF-?B in the tumor induces the production of several cytokines able to modulate the recruitment of cells of the immune system promoting metastatic spreading of tumor cells

Fig 5: Morgana is essential for TNBC cell invasion and in vivo metastasis formation. a Immunoblotting of Morgana and Vinculin, as loading control, in MCF7, MDA-MB-231, BT549 and MCF10A cell lines. b Immunoblotting of Morgana and Vinculin in MDA-MB-231 cells infected with empty vector (EMPTY) or two independent Morgana shRNAs (shMORG1, shMORG2). c Growth curves of MDA-MB-231. d Quantification of invasion assays performed on MDA-MB-231 EMPTY or shMORG1 and shMORG2. e Quantification of soft agar colony formation by MDA-MB-231 EMPTY or shMORG1 and shMORG2. f–h Representative haematoxylin and eosin stained sections f, quantification of g pulmonary metastases and h metastatic burden, 4 weeks after tail vein injection of MDA-MB-231 EMPTY or shMORG1 and shMORG2 (n = 6 NSG mice per group). i Metastasis quantification in distant organs in mice injected with MDA-MB-231 EMPTY or shMORG1 and shMORG2 (n = 3 NSG mice per group). j Tumor weight 5 weeks after subcutaneous injection of MDA-MB-231 EMPTY or shMORG1 and shMORG2 (n = 6 NSG mice per group). k–m Representative haematoxylin and eosin stained sections k, quantification of l pulmonary metastases and m metastatic burden of MDA-MB-231 EMPTY or shMORG1 and shMORG2 (n = 6 NSG mice per group). Data are the results of at least three independent experiments. Bars in graphs represent standard errors (*p < 0.05; **p < 0.01; ***p < 0.001)