Fig 1: H2O2-induced oxidative stress alters RPE polarized secretion of angiogenic regulatory factors. Bar graphs show differences in apical and basolateral secretion of VEGF, PEDF, CFH, and PTN before (Pre-treatment, black) and after incubation with (A) 800 µM H2O2 for 24 h (Post-H2O2, light grey), or (B) 1 mM tBH for 6 h (Post-tBH, light grey). Dark grey bars show protective effects of quercetin pre-treatment prior to oxidative stress with H2O2 or tBH. N = 9. Results were analyzed by paired 2-way ANOVA. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ns = not significant. There was no significant interaction between polarity and treatment effects for PTN, but p-values for post-hoc comparisons are still shown.
Fig 2: Human iPSC-RPE polarized secretome is more apical than basolateral. (A) Volcano plot from data set 2 showing proteins with at least twofold change and p < 0.05, in blue (more basolateral) or red (more apical). N = 4 biologic replicates. The 3 most differentially secreted proteins in both the apical and basolateral compartments, as well as proteins of interest as described in the results section, are labeled by number on the graph: 1. RNF7, 2. TIMP3, 3. UBR, 4. PTPN22, 5. CFI, 6. VEGF, 7. C1QTNF5, 8. PTN, 9. HTRA1, 10. CFH, 11. PEDF, 12. RPL12, 13. HMGN2, 14. CRYAB, 15. PSMA1. (B) Dendrogram from data set 2 showing molecular functions that were differentially represented in apical vs basolateral samples (p < 0.05, adjusted for false discovery rate). Analysis was performed using Advaita Bio ipathwayguide12,13.
Fig 3: Cancer cells and endothelial cells produce PTN in the TME. (A) t-distributed stochastic neighbor embedding (tSNE) plot from scRNAseq of a patient’s primary tumor, BC-P1 CID4471, showing the different cell types within the TME. (B) tSNE plot showing the expression of PTN in cell types from A. Data were accessed from the online portal https://singlecell.broadinstitute.org/ (Wu et al., 2021). (C) Breast cancer patient primary tumor FFPE sections were stained by multiplex IHC for PTN (red), panCK (cancer cells; green), CD31 (endothelial cells; white), and DAPI (nucleus; blue). Images are representative of two biological replicates. Scale bar, 100 µm. White arrows point at colocalization of the PTN signal with panCK and CD31. (D) Representative images of dual RNA FISH of PTN (red) and IHC of CD31 (endothelial cells; green), PyMT (cancer cells; white) in MMTV-PyMT primary tumor (n = 4). Scale bar, 100 µm. White arrows point at colocalization of the PTN signal with PyMT and CD31. (E) Representative images of dual RNA FISH of PTN (red; a, b), PTN receptor RPTPß/? (red; i, j) and IHC of CD31 (white; a, d), PyMT (green; a, c, i, k), CD45 (green; e, g), podoplanin (white; e, h) in MMTV-PyMT lung metastases (n = 4). Scale bar, 100 µm.
Fig 4: PTN promotes an inflamed TME by recruiting neutrophils. (A) Heat map highlighting changes in neutrophil associated genes in MMTV-PyMT primary tumors treated with C44 or 3B10 (n = 3 each). (B) CXCL5 expression in normal mammary fat pad (n = 2) or Met-1 (MMTV-PyMT–derived) tumors treated with C44 (n = 8) or 3B10 (n = 6) was determined by cytokine array. Unpaired t tailed test; **, P < 0.01. (C) MMTV-PyMT tumors from mice treated with C44 (n = 12) or 3B10 (n = 9) were stained for neutrophils (Ly6G+ cells) by immunofluorescence (IF). Representative images are shown at 20× magnification. Scale bar, 100 µm. (D) Quantification of C. Unpaired t tailed test; *, P < 0.05. (E) CIBERSORT analysis using RNA sequencing data of 4T1 tumors (n = 3 each) shows a decrease in neutrophil infiltration with 3B10 treatment. Analysis was performed at the online portal http://timer.cistrome.org/. (F) CXCL5 expression in 4T1 lungs from mice treated with C44 or 3B10 (n = 4 each) was determined by cytokine array. Unpaired t tailed test; *, P < 0.05. (G) 4T1 metastatic lungs from mice treated with either C44 (n = 6) or 3B10 (n = 5) were stained for neutrophils (Ly6G+ cells) by IF. Representative images are shown at 20× magnification. Scale bar, 100 µm. (H) Quantification of Ly6G+ cells from G. Unpaired t tailed test; *, P < 0.05. (I) Representative H&E images of lung metastases in WT BALB/c mice bearing 4T1 tumors. Scale bar, 1 mm. WT BALB/c mice were treated with either isotype control (IgG), 3B10, Ly6G, or Ly6G+3B10 (n = 8 mice each). (J) Quantification of I in terms of metastatic area/total lung area. Kruskal–Wallis test; *, P < 0.05. (K) Representative H&E images of lung metastases in WT C57BL/6 mice bearing E0771-LG tumors. WT C57BL/6 mice were treated with isotype control (IgG; n = 8), 3B10 (n = 10), anti-Ly6G (n = 9), or anti-Ly6G+3B10 (n = 9). (L) Quantification of I in terms of metastatic area/total lung area. Kruskal–Wallis test; *, P < 0.05.
Fig 5: PTN expression within different cell types of TME. (A) Violin plot showing PTN expression in different cell types within the TME was generated using breast cancer patient scRNAseq data (BC-P1 CID4471) from a previous study (Wu et al., 2021). (B) tSNE plot showing the different cell types identified from scRNAseq of late-stage PyMT tumors. Data accessed from the previous study (Valdés-Mora et al., 2021). (C) tSNE plot showing the expression of PTN in different cells from B. (D) Violin plot, generated from scRNAseq of MMTV-PyMT primary tumors, showing PTN expression by different cell types within the TME. (E) Representative image of RNA FISH of PTN (red) on a lung bearing metastatic lesion from a MMTV-PyMT mice (n = 4). Scale bar, 1 mm. Boxes 1 and 2 indicate metastatic lesion and normal lung area, respectively. Magnified images at 20× of metastatic lesion and normal lung area are shown in 1.i, 1.ii, 1.iii and 2.i, 2.ii, 2.iii, respectively. Scale bar, 100 µm.
Supplier Page from Thermo Fisher Scientific for Human Pleiotrophin/PTN ELISA Kit