European Urology Focus E2F1 Signalling is Predictive of Chemoresistance and Lymphogenic Metastasis in Penile Cancer: A Pilot Functional Study Reveals New Prognostic Biomarkers  Ferdinand Fenner, Deborah Goody, Chris Protzel, Andreas Erbersdobler, Christin Richter, Juliane M. Hartz, Carsten M. Naumann, Holger Kalthoff, Ottmar Herchenröder, Oliver W. Hakenberg, Brigitte M. Pützer  European Urology Focus  DOI: 10.1016/j.euf.2017.02.009 Copyright © 2017 European Association of Urology Terms and Conditions

Fig. 1 E2F1 status in human penile cancer (PC) cell lines. (A) Protein expression levels of components of the RB/E2F1 signalling pathway in PC cell lines (Ki-PeCa-P1, Ki-PeCa-P2, Ki-PeCa-L2, and Ki-PeCa-L3) and healthy primary foreskin fibroblasts (VH-6) were analysed using western blotting. Actin served as a loading control. (B) PC cell invasion was assessed in a Boyden chamber assay (left). Cells were seeded onto transwell filters coated with Matrigel for 48h, and invading cells were stained and counted. Fold changes were calculated relative to Ki-PeCa-L2 cells (set as 1). The proliferation kinetics of PC cell lines were measured using a viability assay at the times indicated (right). Each data point represents the mean number of viable cells for triplicate assays. (C) Ki-PeCa-P1 and Ki-PeCa-P2 were transiently transfected with an E2F1 overexpression plasmid (left). After 48h, invading cells were examined and quantified. Fold changes were calculated relative to wild-type cells (WT) and cells transfected with an empty vector control (ctrl; set as 1). Inserts show western blot analyses for transfected cells, with actin serving as a loading control. Proliferation kinetics were measured as described above (right). (D) Lymph node metastases–derived Ki-PeCa-L2 and Ki-PeCa-L3 cells were stably transduced with lentiviral vectors expressing either sh.E2F1 (sh.E2F1#1 or sh.E2F1#2) or scrambled RNA (sh.scr) and subjected to cell invasion assays (left panel). Fold changes were calculated relative to control (sh.scr) cells (set as 1). Protein expression is indicated by western blots, with actin serving as a loading control. Cell proliferation assays after E2F1 knockdown by sh.RNAs (right panels) were performed as for (C). Protein expression levels were assessed via western blotting using antibodies against the proteins depicted. Actin was used to verify equal sample loading. All bar graphs are presented as mean±standard deviation. Statistical significance was determined using Student t test. * p<0.05. OD=optical density. European Urology Focus DOI: (10.1016/j.euf.2017.02.009) Copyright © 2017 European Association of Urology Terms and Conditions

Fig. 2 E2F1 knockdown renders metastases-derived penile cancer (PC) cells chemosensitive. Cells were stably transduced with lentiviral vectors expressing either sh.E2F1 (sh.E2F1#1 or sh.E2F1#2) or scrambled RNA (sh.scr) and treated with chemotherapy. (A) Depletion of E2F1 in Ki-PeCa-L2 and Ki-PeCa-L3 cells was confirmed by western blotting with actin as control. The percentage of viable (B) Ki-PeCa-L2 and (C) Ki-PeCa-L3 cells treated with cisplatin (left panels) and paclitaxel (right panels) at the concentrations shown for 24h was measured. Bar graphs denote the mean±standard deviation (* p<0.05). Control sh.scr had no influence on cell viability when compared to untreated cells (data not shown). European Urology Focus DOI: (10.1016/j.euf.2017.02.009) Copyright © 2017 European Association of Urology Terms and Conditions

Fig. 3 E2F1 stimulates tubule formation of human dermal lymphatic endothelial cells (HDLECs) via regulation of lymphangiogenic factors. (A) Endogenous levels of VEGFR-3 and VEGF-C in primary and metastases-derived PC cell lines. (B) Analysis of VEGFR-3/VEGF-C protein levels and epithelial-mesenchymal transition markers in the PC cell lines indicated after overexpression or inhibition of E2F1. (C) HDLECs seeded on Matrigel were cultured with conditioned media from cancer cells overexpressing E2F1 (upper panels) or treated with sh.E2F1 (lower panels). Representative images of capillary tube formation (TF) monitored via phase contrast microscopy and quantified using ImageJ. Protein expression levels were assessed as stated above. Bar graphs are presented as mean±standard deviation. Statistical significance was determined using Student t test. * p<0.05. European Urology Focus DOI: (10.1016/j.euf.2017.02.009) Copyright © 2017 European Association of Urology Terms and Conditions

Fig. 4 High levels of E2F1 correlate with metastasis in penile cancer (PC) patients. (A) Scatter plot showing relative expression levels of E2F1 in nonmetastatic tumours (NM), primary tumours that metastasised (PT), and lymph node metastases (M) from PC patients (n=8 per group). RNA was isolated from formalin-fixed, paraffin-embedded samples and E2F1 expression levels were assessed by qRT-PCR with 18S as a normalisation control. Mean expression levels are shown with the median. Statistical significance was determined by Mann-Whitney U test. * p<0.05; *** p<0.0005. (B) Haematoxylin and eosin staining (upper images) of specimens of primary penile squamous cell carcinoma and corresponding immunohistochemistry with an antibody against E2F1 (lower images). The left panel shows a well-differentiated tumour with an E2F1 Quick score of 3, and the right a poorly differentiated tumour with an E2F1 Quick score of 7. Original magnification for all photographs: 200×. European Urology Focus DOI: (10.1016/j.euf.2017.02.009) Copyright © 2017 European Association of Urology Terms and Conditions