Supplementary Figure 4. Nuclear p85β expression in Stage III TNBC patients and paxalisib- treated 4T1 TNBC mice.

Supplementary Figure 2. Dose de-escalation of the PI3K-mTOR inhibitor, paxalisib, overcomes toxicity burden in combination treatments.

upplementary Table S2. The 23 antibody CODEX panel.

Supplementary Results to accompany manuscript

Supplementary Figure 1. PI3K-mTOR blockade inhibits proliferation of TNBC cells

Supplementary Figure 7. Associations between PIK3R2 gene expression and survival probability amongst TNBC patient cohorts.

Supplementary Figure 6. EZH2 co-localizes with NF-κB to activate pro-oncogenic gene expression in TNBC.

Supplementary Figure 3. NanoString nCounter analysis highlighting distinct gene expression profiles in Paxalisib-treated tumors.

Supplementary Table S1. List of the PI3K/mTOR inhibitors

Background The nucleocytoplasmic shuttling of ERK5 has gained recent attention as a regulator of its diverse roles in cancer progression but the exact mechanisms for this shuttling are still under investigation. Methods Using in vitro, in vivo and in silico studies, we investigated the roles of shorter ERK5 isoforms in regulating the nucleocytoplasmic shuttling of active phosphorylated-ERK5 (pERK5). Retrospective cohorts of primary and metastatic breast cancer cases were used to evaluate the association of the subcellular localization of pERK5 with clinicopathological features. Results Extranuclear localization of pERK5 was observed during cell migration in vitro and at the invasive fronts of metastatic tumors in vivo . The nuclear and extranuclear cell fractions contained different isoforms of pERK5, which are encoded by splice variants expressed in breast and other cancers in the TCGA data. One isoform, isoform-3, lacks the C-terminal transcriptional domain and the nuclear localization signal. The co-expression of isoform-3 and full-length ERK5 associated with high epithelial-to-mesenchymal transition (EMT) and poor patient survival. Experimentally, expressing isoform-3 with full-length ERK5 in breast cancer cells increased cell migration, drove EMT and led to tamoxifen resistance. In breast cancer patient samples, pERK5 showed variable subcellular localizations where its extranuclear localization associated with aggressive clinicopathological features, metastasis, and poor survival. Conclusion Our studies support a model of ERK5 nucleocytoplasmic shuttling driven by splice variants in an interplay between mesenchymal and epithelial states during metastasis. Using ERK5 as a biomarker and a therapeutic target should account for its splicing and context-dependent biological functions. Graphical Abstract ERK5 isoform-3 expression deploys active ERK5 (pERK5) outside the nucleus to facilitate EMT and cell migration. In cells dominantly expressing isoform-1, pERK5 shuttles to the nucleus to drive cell expansion.

Almost half of patients with triple-negative breast cancer develop distant metastases, heralding unfavorable outcomes. Here, we provide novel insights into the contribution of the PI3K-mTOR pathway to the triple-negative breast cancer phenotypes that promote growth, migration, metastasis, and therapy resistance. Specifically, we demonstrate that dual targeting of PI3K and mTOR but not PI3K alone inhibits cancer cell proliferation and migration in vitro. Dual PI3K-mTOR inhibition with paxalisib not only promotes a favorable mesenchymal-to-epithelial phenotype but also inhibits signatures associated with metastasis-initiating cells, including the highly aggressive cancer stem cell phenotype, persister cancer cell phenotype (p65, FOXQ1, NRF2, and NNMT), and a cancer drug resistance signature (ABCB5, SNAIL, and ALDH1). In vivo, paxalisib overcomes immunotherapy resistance to reduce primary tumor burden, circulating tumor cells, and direct and indirect indicators of metastasis with a favorable toxicity profile. Gene expression and spatial analyses show that paxalisib profoundly affects the immune microenvironment in tumors, reducing adaptive immune phenotypes associated with immunotherapy resistance (exhausted T cells and regulatory T cells) and protumor innate immune populations such as mast cells. PI3K-mTOR blockade acts upstream of EZH2, impacting both the classic repressive catalytic p85β-EZH2-H27ME3 and active EZH2-NF-κB pathways. Our data suggest that dual targeting of the PI3K-mTOR pathway disrupts both the catalytic and noncatalytic axes of EZH2 to inhibit metastasis and enhance cancer immune visibility, potentially increasing the utility of immunotherapy in resistant individuals.

Supplementary Figure 5. PI3K-mTOR inhibition targets the p85β:H3K27Me3 switch.

Supplementary Table S3. Inhibition of human PI3K isoenzymes in vitro by paxalisib in fluorescence polarization biochemical assays.

Supplementary Figure 2: Maximum tolerated doses of gemcitabine, Chk1i and EGFR-directed RIT in vivo. (A) Balb/c nude mice bearing PANC-1 subcutaneous xenografts (60 mm3 in volume) were treated with 50 mg/kg or 100 mg/kg of gemcitabine on days 1, 4, 7 and 10 administered intravenously combined with 15 mg/kg Chk1i, administered as two doses per day (7.5 mg/kg per dose) on days 1, 4, 7 and 10 at 3 hours before and after gemcitabine administration. Mice were treated with gemcitabine and Chk1i combinations (gem + Chk1i) alone or in combination with 6 or 9 MBq/20g (300 or 450 MBq/kg) of 177Lu-anti-EGFR mAb. Mice were monitored for 14 days to determine acute toxicities as judged by weight, posture, movement, eating and drinking. Tumor volume was also monitored during this short time as an indication of efficacy. (B) The tumor growth curves were used to measure the growth rate (k = day-1) from exponential growth equations and shown in the bar graph in panel (5 mice per group, error bar is the standard error of the mean, SEM). *** p < 0.001 in One-way ANOVA in GraphPad® Prism comparing treatments involving anti-EGFR RIT and those not involving this RIT (p<0.001). No statistical differences were observed amongst treatments involving the anti-EGFR RIT.

Background Genome-wide association studies have identified a breast cancer risk locus at 19q13.31. The candidate causal variants at this locus are located in the first exon of KCNN4. KCNN4, which regulates membrane potential and Ca 2+ signaling, is a good candidate for drug repositioning because its inhibitor, Senicapoc, has been shown to be well tolerated in Phase-II and -III clinical trials for asthma and sickle cell anemia. Methods We evaluated public mRNA expression data to determine whether the allele at 19q13.31 associated with increased breast cancer risk was associated with KCNN4 expression. We also used immunohistochemistry to evaluate the relationship between KCNN4 protein expression and breast cancer survival. We then used Senicapoc in two murine mammary tumor models to determine if it would delay tumor development. We also treated mice bearing 4T1 mammary tumors with Senicapoc, by subcutaneous injection and by oral gavage. Finally we used gene editing to make deletions within Kcnn4 in 4T1 to determine whether Senicapoc had off-target effects on tumor growth. Results Analysis of the Genotype-Tissue Expression Project showed that the allele at 19q13.31 associated with increased breast cancer risk is associated with increased KCNN4 expression, suggesting that inhibiting KCNN4 might reduce breast cancer risk. Using immunohistochemistry in a large breast cancer cohort, we found that membrane and cytoplasmic expression is a marker of poor prognosis in triple negative breast cancer. We then tested the efficacy of Senicapoc to prevent and treat breast cancer. This showed that it delays the development of mammary tumors in two murine models, and slows growth of a syngeneic (4T1) model of triple negative breast cancer. Senicapoc monotherapy showed similar efficacy to anthracycline/taxane-based chemotherapy in these studies, with a stronger effect when they were combined. Conclusions These results provide a rationale for clinical testing of Senicapoc for treating, and even preventing, breast cancer.

Supplementary Figure 4: Efficacy of anti-EGFR directed RIT in combination therapy in vivo against PDAC model from the BxPC-3 ATCCTM cell lines. BxPC-3 cells expressing luciferase were used to inoculate balb/c nude mice subcutaneously at 4 weeks of age. Treatments were initiated when tumors reached 60 mm3 as outlined in diagram in A and in Figure 2. Single agent treatments included: EGFR control: 50 μg of unlabeled anti-EGFR mAb; Chk1i: 15 mg/kg of the Chk1 inhibitor PF-477736 administered as two 7.5 mg/kg injections per day subcutaneously on days 1, 4, 7 and 10; Gem: 50 mg/kg of gemcitabine administered intravenously on days 1, 4, 7 and 10; RIT: 177Lu-anti-EGFR mAb (50 μg with 6 MBq radioactivity per 20 g mouse) injected intravenously on day 2 only. For combinations, treatments were performed as described for the single agents where Chk1i was administered 3 hours before and after gemcitabine on days 1, 4, 7 and 10 and RIT was administered on day 2 only. (B) Tumor growth curves presented as the change in tumor volume compared to day 0 (% change +SEM, n = 5 mice per treatment group). (C) Representative images of live bioluminescence imaging using luciferin performed on day 1 (prior to treatment) and day 14 (96 hours after treatments). (D) Quantification and representative images of live bioluminescence imaging using caspase-3/7 substrate (Z-DEVD-Luciferase) performed on day 7 after treatment initiation. Data shown is the average of caspase-3/7 activation (+SEM, n = 5 per treatment group). *** p < 0.001 in One-way ANOVA in GraphPad® Prism.

Supplementary Figure 1: Sensitization of PANC-1 cells to gemcitabine and EGFR-directed RIT by Chk1 inhibition. (A) Adherent cultures of the PDAC PANC-1 cell line were treated in the absence or presence of escalating doses of gemcitabine, in the absence of presence of Chk1i alone (PF-477736, 180 nM) or the combinations. Chk1i was added either 16 hours after gemcitabine or concurrently. Cells were collected 24-96 hours after treatment for standard cell cycle analysis by DNA content using FACS. (B) PANC-1 cells were treated with escalating concentrations of 177Lu-anti-EGFR mAb (RIT) to achieve the specified radiation doses (0-4 Gy) over 72 hours of incubation. RIT was performed alone or in combination with Chk1i (180 nM) and then clonogenic survival was determined by standard assays. Chk1i alone at 180 nM did not have any effect on clonogenic survival (data not shown). (C) PANC-1 cells were left untreated (vehicle control, data not shown) or incubated with anti-EGFR mAb (unlabeled, control which was not different from the vehicle control), 177Lu-DOTA, 177Lu-labelled mAb with irrelevant specificity (Sal5, raised against Salmonella antigen) or 177Lu-anti-EGFR mAb (2 Gy over 72 hours). Cells were washed 3 hours after incubation to remove unbound material and left untreated or treated with Chk1i alone (180 nM), gemcitabine alone (40 ng/mL) or the combination (Chk1i+gemcitabine) before standard clonogenic survival assays.

Detailed methods of: EGFR immunohistochemistry of PDAC tissue microarrays Patient derived xenografts (PDXs) and cell lines (PDCLs) generation Cell culture Treatments in vitro Histological and immunoblotting analyses of PDXs post-treatment

Supplementary Figure 3: EGFR expression in the PANC-1 and BxPC-3. Cell cultures were used for standard immunoblot and flow cytometry analysis for EGFR expression. The anti-EGFR mAb clone 225 was used for staining and anti-mouse IgG antibody conjugated with HRP was used for immunoblot or conjugated with Alexa488 for flow cytometry. Anti-tubulin was used to confirm equal loading for immunoblot. Both PDAC cell lines express EGFR.