Fig 2: TP73 is prognostic, aberrantly upregulated in the PDAC tumor organoids, and preferentially expressed in the squamous subtype of PDAC.(A) Survival curves of patients stratified by TP73 expression. Log-rank (Mantel-Cox) test used to assess the median survival values and p-values. (B) Expression of TP73 in human organoids derived from normal pancreata (Normal) or PDAC tumors (Tumor) (Tiriac et al., 2018). P-value calculated using unpaired t-test. ****, P ≤ 0.0001. Mean ± SEM shown. (C) Expression of transcription factors in the basal-like vs progenitor subtype of tumors. Transcription factors are ranked by their mean log2 fold-change in expression levels in the basal-like vs. progenitor subtype tumors. TP63, a master regulator of the squamous subtype PDAC (Somerville et al., 2018) is shown as a reference, and the rank of each gene is written inside parentheses. (D) Western blot analysis of TP73 in PDAC cell lines using a pan-TP73 antibody. VINCULIN (VIN) shown as a loading control. Bands corresponding to the molecular weight of the longest p73 isoform (i.e. p73-TAα) indicated with an arrow. Asterisk points to antibody cross-reactivity with p63-dN in cell lines expressing high levels of p63-dN. (E) (Top) Schematic of the human TP73 gene. Exons (white boxes), UTRs (gray boxes), and introns (thin lines) are depicted. Exons are numbered. Transactivation domain (TAD) is indicated with black box. 3’ spliced isoforms are indicated in Greek letters. (Bottom) TP73 isoform expression in PDAC cell lines from CCLE (Ghandi et al., 2019). Isoforms expressed from P1 or P2 promoter are referred to as TP73-TA or TP73-dN, respectively. Isoforms lacking exon1 but retaining the transactivation domain are noted as dExon1. 3’ spliced isoforms with variations at the 3’ UTR are marked with a prime (‘) symbol (i.e. dExon1-ε’ and dExon1-ζ’). Isoform expressed from an alternative promoter other than P1 or P2 is noted as dN’ (i.e. TP73- dN’α). For the heatmap, cell lines are listed in descending order (from left to right) of total TP73 isoform expression. TP73-TA and dExon1 are collectively considered as TA isoforms, and TP73-dN and TP73- dN’ are collectively considered as dN isoforms. Expression of each isoform is normalized within a cell line (i.e. % expression = tpm value of isoform A / sum of tpm values of all isoforms * 100).

Fig 3: p73-TA elicits p53-like responses in a dose-dependent manner.(A) Western blot analysis of p73 following lentiviral delivery of TP73-TAα cDNA or CRISPRa-based induction of p73-TA in ASPC1, CFPAC1, and PANC1 cell lines, compared to the empty vector (Empty) or scramble sgRNA (Scramble) controls, respectively. VINCULIN (VIN) shown as a loading control. (B) Proliferation of cell lines mentioned in (A) measured by CellTiterGlo Luminescent Cell Viability Assay; luminescence is measured in relative light units (RLU). For each sample, every data point is normalized to the RLU value on day 0 and represented as fold-change. P-values compared to cDNA Empty sample calculated by two-way ANOVA with Sidak’s multiple comparisons test. *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001; ****, P ≤ 0.0001. Mean ± SEM shown (n=3). (C-D) RNA-seq analysis of cell lines mentioned in (A). (C) Gene ontology (GO) analysis using Metascape (Zhou et al., 2019), interrogating significantly upregulated genes following overexpression of p73-TAα using cDNA. Terms are ranked by their p-values; TP53 and anti-proliferative/apoptotic terms are highlighted in blue. (D) Comparison between cDNA and CRISPRa induced samples of 72 genes that were categorized to the gene ontology terms highlighted in (C). Heatmap shows log2 fold-change in the expression levels of genes when p73-TA is overexpressed compared to the controls. TP73 and example TP53 target genes are labeled (Zhan et al., 1998; Hermeking et al., 1997; El-Deiry et al., 1993). See also Materials and Methods and Supplementary Table 2.

Fig 4: Detached endfeet, caused by loss of Myh10, impact cortical architecture.(A) Cartoon schematic of an RGC (green) and the local niche where the endfeet reside made up of BM (black), fibroblasts (brown), vasculature (red), CR cells (grey), and interneurons (red). Box indicates MZ region of images in B, E, G. (B) Labeling of all nuclei with DAPI (blue) in E16.5 Myh10 Ctrl and cKO brains. Boundary between pia and MZ marked by yellow dashed line. White bar represents MZ thickness measurements in D. (C) Quantification of DAPI cells in the MZ. n = 5 Ctrl, 6 cKO from 2 litters. (D) Quantification of MZ thickness. Each data point represents the average of 3 measurements per section and 3 sections per brain. n = 6 per genotype from 2 litters. (E) Staining with Calretinin (white) to label CR cells and interneurons and DAPI (blue) to mark all nuclei in E16.5 Myh10 Ctrl and cKO brains. Yellow bar represents where Calretinin thickness was measured for F. (F) Quantification of Calretinin thickness. n = 3 per genotype from 1 litter (G) P73 staining (green) of CR cell nuclei and DAPI (blue). Asterisks denote vasculature. White dotted line marks boundary between pia and MZ. (H) Quantification of P73+ cells in MZ. n = 6 per genotype from 2 litters Error bars: SD. (C, D, F, H). Student unpaired, two-tailed t test (C, D, F, H). Data points color-coded by litter (C, D, F, H). ns p > 0.05, *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001. Scale bar: (B, E, G), 50 μm. Data underlying graphs included in S4 Data. BM, basement membrane; CR cells, Cajal–Retzius cells; MZ, marginal zone; RGC, radial glial cell.

Fig 5: p73-TA promotes cell migration and invasion in PDAC cells.(A) GO analysis using Metascape (Zhou et al., 2019), interrogating significantly upregulated genes following activation of p73-TA in a panel of PDAC cell lines. Terms are ranked by their p-values; TP53-related terms are highlighted in blue and cell invasion/migration-related terms are highlighted in magenta. (B) Scratch assays following activation of p73-TA compared to the control (Scramble). Quantification of % wound confluence plotted at the indicated time points post-scratch and representative images are shown; scale bar indicates 400 µm. P-values compared to Scramble calculated using two-way ANOVA with Sidak’s multiple comparisons test. Mean ± SEM shown (n=3). *, P ≤ 0.05; **, P ≤ 0.01; ****, P ≤ 0.0001. (C) Transwell assays following activation of p73-TA compared to the control (Scramble). Bar charts show % invasion calculated by number of cells migrated through Matrigel-coated membrane (Invasion) vs. uncoated membrane (Migration). Representative images shown; scale bar indicates 200 µm. P-values calculated using unpaired t-test. Mean ± SEM shown (n=3). ns, P > 0.05; *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001. (D) 3D Matrigel colony formation assays following activation of p73-TA in SUIT2 cells compared to the control (Scramble). Bar charts show colony number and size. P-values calculated using unpaired t-test. Mean ± SEM shown (n=3). *, P ≤ 0.05. Representative images shown; scale bar indicates 200 µm. (E) Representative images of phalloidin staining (green) to visualize F-actin following p73-TA activation compared to the control (Scramble); scale bar indicates 10 µm. Nuclei stained with DAPI (blue). See also Supplementary Figure 6. Bar charts show quantification of % cells presenting actin polymerization. P-values calculated using unpaired t-test. Mean ± SEM shown (n=3). ns, P > 0.05.