Temporally resolved proteomics identifies nidogen-2 as a cotarget in pancreatic cancer that modulates fibrosis and therapy response

Abstract

Pancreatic ductal adenocarcinoma (PDAC) is characterized by increasing fibrosis, which can enhance tumor progression and spread. Here, we undertook an unbiased temporal assessment of the matrisome of the highly metastatic KPC (Pdx1-Cre, LSL-Kras^G12D/+, LSL-Trp53^R172H/+) and poorly metastatic KP^flC (Pdx1-Cre, LSL-Kras^G12D/+, Trp53^fl/+) genetically engineered mouse models of pancreatic cancer using mass spectrometry proteomics. Our assessment at early-, mid-, and late-stage disease reveals an increased abundance of nidogen-2 (NID2) in the KPC model compared to KP^flC, with further validation showing that NID2 is primarily expressed by cancer-associated fibroblasts (CAFs). Using biomechanical assessments, second harmonic generation imaging, and birefringence analysis, we show that NID2 reduction by CRISPR interference (CRISPRi) in CAFs reduces stiffness and matrix remodeling in three-dimensional models, leading to impaired cancer cell invasion. Intravital imaging revealed improved vascular patency in live NID2-depleted tumors, with enhanced response to gemcitabine/Abraxane. In orthotopic models, NID2 CRISPRi tumors had less liver metastasis and increased survival, highlighting NID2 as a potential PDAC cotarget.

Fig. 1.. Temporal proteomics of pancreatic cancer GEMMs reveal NID2 as a potential target in aggressive disease.

(A) Representative images of nonneoplastic pancreas, early-, mid- and late-stage KPC tumors stained for hematoxylin and eosin (H&E) (top row), Picrosirius Red imaged by bright-field (second row) and polarized light (third row; matched to Picrosirius Red images), and SHG imaging (bottom row). Scale bars, 100 μm (H&E and Picrosirius Red bright-field and polarized light) and 25 μm (SHG). (B) Representative dark-field images of decellularized WT pancreas (top left) and late-stage KPC specimen (top right). Representative image of collagen IV IF (green) of decellularized WT pancreas (bottom). Scale bars, 500 μm (dark-field) and 100 μm (collagen IV). (C) Workflow for matrisome-enriched DIA LC-MS/MS analysis of WT, KP^flC, and KPC tumors at early-, mid-, and late-stage time points. m/z, mass/charge ratio; 0.5% SDC, sodium deoxycholate. (D) Classification of matrisome categories detected (33). Core matrisome (n = 123; purple) and matrisome-associated (n = 158; gray). Core matrisome category (purple colors); ECM glycoproteins (n = 83), collagens (n = 27), and proteoglycans (n = 13). Matrisome-associated category (gray colors); ECM regulators (n = 93), ECM-affiliated proteins (n = 46), and secreted factors (n = 19). (E) Principal components analysis of WT (green), KP^flC (blue), and KPC (red) tumors at early- (circle), mid- (star), and late-stage (square) time points. (F) Volcano plot of matrisome proteins comparing KP^flC and KPC mid-stage tumors. Two-sample t test with FDR = 0.05. y axis = −log(P value) and x axis = difference (fold change). ITIH3, Inter-Alpha-Trypsin Inhibitor Heavy Chain 3; LAMC1, Laminin Subunit Gamma 1; ADAM10, A Disintegrin And Metalloproteinase Domain 10; PLXNB2, Plexin B2; S100A16, S100 Calcium Binding Protein A16; MUC2, Mucin 2; PRSS1, Serine Protease 1; SERPINA3K, serine (or cysteine) peptidase inhibitor, clade A, member 3K; ECM1, Extracellular Matrix Protein 1; COL7A1, Collagen Type VII Alpha 1 Chain. Schematics were created with Biorender.com.

Fig. 2.. NID2 expression is highly expressed in CAFs and can be depleted via CRISPRi.

(A) Kaplan-Meier analysis of ICGC PDAC cohort (n = 267) based on NID2 mRNA. High NID2 (red; n = 67) compared with low NID2 (blue; n = 200). Log-rank test, P = 0.042. (B) NID2 Western blot for KPC cancer cells and KPC CAFs with quantification. One-sample t test, *P < 0.05. n = 3 independent repeats. (C) RT-qPCR NID2 for KPC cancer cells and KPC CAFs, exon 1/2 and exon 6/7. Two-way analysis of variance (ANOVA) with Sidak’s test, **P < 0.01 and ***P < 0.001. n = 6 repeats for KPC cancer cells and n = 3 repeats for CAFs, in triplicate. (D) NID2 Western blot for GFP-1 KRAB, B500 NID2 KRAB, and C500 NID2 KRAB CAFs with quantification. One-sample t test, *P < 0.05. n = 3 independent repeats. (E) RT-qPCR NID2 for GFP-1 KRAB, B500 NID2 KRAB, and C500 NID2 KRAB CAFs, exon 1/2 and exon 6/7. Two-way ANOVA with Dunnett’s test, *P < 0.05, **P < 0.01, and ***P < 0.001. n = 5 GFP-1 KRAB repeats, n = 8 B500 NID2 KRAB repeats, and n = 4 C500 NID2 KRAB repeats, in triplicate. (F) CAF cell-derived matrix (CDM) assay. Representative regions of interest (ROIs) of SHG maximum projection for GFP-1 KRAB, B500 NID2 KRAB, and C500 NID2 KRAB CAF CDMs (day 7). Scale bars, 100 μm. Quantification of normalized SHG maximum intensity. One-sample t test, **P < 0.01. n = 4 repeats all CAF lines, with 4 to 8 ROIs per well in triplicate. All data represented as means ± SEM. Schematics were created with Biorender.com.

Fig. 3.. NID2 reduction in CAFs reduces biomechanical properties in organotypic matrices, leading to altered cancer cell invasion.

(A) 3D organotypic contraction assay with CRISPRi CAFs. (B) Representative ROIs of NID2 IHC with NID2-positive cell (%) quantification. One-way ANOVA with Dunnett’s test, *P < 0.05 and **P < 0.01. n = 3 repeats, in triplicate. Scale bars, 50 μm. (C) Quantification at days 3, 6, 9, and 12 and representative images of matrix contraction (in square millimeter) at days 3 and 12. Scale bars, 1 cm. (D) Young’s modulus (in kilopascals) of matrices on day 12. One-way ANOVA with Dunnett’s test, **P < 0.01. n = 3 repeats, in triplicate. (E) Representative SHG maximum intensity with quantification. One-way ANOVA with Dunnett’s test, *P < 0.05. n = 3 repeats, in triplicate. Scale bars, 100 μm. (F) Representative ROIs of Picrosirius Red with coverage (%) quantification. One-way ANOVA with Dunnett’s test, *P < 0.05 and **P < 0.01. n = 3 repeats, in triplicate. Scale bars, 100 μm. (G) Representative ROIs of polarized light Picrosirius Red (matched to Picrosirius Red ROIs) with coverage (%) quantification. One-way ANOVA with Dunnett’s test; green, yellow, and red-orange birefringence proportion quantification. Two-way ANOVA with Dunnett’s test, *P < 0.05 and **P < 0.01. n = 3 repeats, in triplicate. Scale bars, 100 μm. (H) Organotypic invasion assay. Representative H&E ROIs of cancer cell invasion with quantification. One-way ANOVA with Dunnett’s test, ***P < 0.001. n = 5 GFP-1 KRAB, n = 7 B500 NID2 KRAB, and n = 8 C500 NID2 KRAB repeats, in triplicate. Scale bars, 100 μm. All data represented as means ± SEM. Schematics were created with Biorender.com.

Fig. 4.. NID2 reduction in CAFs leads to reduced tumor growth and fibrosis in subcutaneous model under gemcitabine/Abraxane treatment.

(A) Subcutaneous coinjection experiment using GFP-1 KRAB or B500 NID2 KRAB CAFs (75%) with cancer cells (25%). Mice were treated twice weekly with gemcitabine/Abraxane, beginning day 11. (B) Quantification of GFP-1 KRAB (purple) and B500 NID2 KRAB (blue) tumor growth (in cubic millimeters) over time and at day 23. Welch’s t test, *P < 0.05. n = 9 GFP-1 KRAB mice and n = 7 B500 NID2 KRAB mice. IP, intraperitoneal injection. (C) Representative images of day 23 tumors. Scale bars, 2 cm. (D) Representative ROIs of NID2 IHC with quantification of positive pixels (%). Mann-Whitney test, *P < 0.05. Scale bars, 100 μm. (E) Representative ROIs of NID1 IHC with quantification of positive pixels (%). Welch’s t test, ns: P > 0.05. Scale bars, 100 μm. (F) Representative ROIs of Picrosirius Red with quantification of coverage (%). Welch’s t test, *P < 0.05. Scale bars, 100 μm. (G) Representative ROIs of polarized light Picrosirius Red with quantification of coverage (%). Welch’s t test, **P < 0.01. Quantification of proportions of green, yellow, and red-orange birefringence. Two-way ANOVA with Sidak’s test, ***P < 0.001 for green and yellow birefringence. Scale bars, 100 μm. (H) Representative ROIs of CD31 IHC with quantification of positive pixels (%). Welch’s t test, ***P < 0.001. Scale bars, 100 μm. All data represented as means ± SEM. Schematics were created with Biorender.com.

Fig. 5.. NID2 reduction in CAFs leads to vascular changes in subcutaneous model under gemcitabine/Abraxane treatment shown via intravital imaging and quantum dots.

(A) Subcutaneous coinjection intravital imaging experiment using GFP-1 KRAB or B500 NID2 KRAB CAFs (75%) with eGFP-tagged cancer cells (25%). Quantum dots injected via tail vein. Mice were treated twice weekly with gemcitabine/Abraxane, beginning day 11. n = 7 GFP-1 KRAB mice and n = 8 B500 NID2 KRAB mice. (B) Representative images for GFP-1 KRAB and B500 NID2 KRAB tumors imaged live via multiphoton intravital microscopy. SHG of fibrillar collagen (purple), eGFP cancer cells (green), vasculature (quantum dots; red), Imaris surface of blood vessels (red), and merged image. Scale bars, 100 μm. (C) Representative images of the vasculature of GFP-1 KRAB and B500 NID2 KRAB analyzed using VesselVio. Vessel skeletonization (black skeleton; top), skeletonization with 3D vessel rendering (black skeleton with red rendering; middle), and 3D vessel rendering alone (red rendering; bottom). Scale bars, 100 μm. (D) Quantification of total vessel volume (normalized to z depth of image in micrometers). Welch’s t test, *P < 0.05. (E) Quantification of total vessel surface area (normalized to z depth of image in micrometers). Welch’s t test, *P < 0.05. (F) Quantification of mean vessel radius (in micrometers) for GFP-1 KRAB (purple) and B500 NID2 KRAB (blue) live tumors. Welch’s t test, *P < 0.05. All data represented as means ± SEM. Schematics were created with Biorender.com.

Fig. 6.. NID2 reduction in CAFs leads to increased median survival in an orthotopic coinjection model.

(A) Orthotopic coinjection survival experiment using GFP-1 KRAB or B500 NID2 KRAB CAFs (75%) with luciferase cancer cells (25%). Mice were treated twice weekly with gemcitabine/Abraxane upon detectable IVIS signal. (B) Representative IVIS signal image on day 7 of mouse bearing GFP-1 KRAB tumor and B500 NID2 KRAB tumor. IVIS signal with a Fstop1, exp30s. Color scale is radiance. (C) Kaplan-Meier survival curves of mice bearing orthotopic tumors with GFP-1 KRAB or B500 NID2 KRAB treated with vehicle (saline) or gemcitabine/Abraxane. n = 8 GFP-1 KRAB saline mice (median survival, 42 days; black), n = 8 B500 NID2 KRAB saline mice (median survival, 50 days; pink), n = 8 GFP-1 KRAB gemcitabine/Abraxane mice (median survival, 73.5 days; purple), and n = 10 B500 NID2 KRAB gemcitabine/Abraxane mice (median survival, 83 days; blue). Kaplan-Meier curves (GFP-1 KRAB saline compared to B500 NID2 KRAB saline and GFP-1 KRAB gemcitabine/Abraxane compared to B500 NID2 KRAB gemcitabine/Abraxane) were compared with a log-rank Mantel-Cox test. *P < 0.05 and **P < 0.01. Schematics were created with Biorender.com.

Fig. 7.. NID2 reduction in CAFs leads to reduced tumor fibrosis and liver metastasis in an orthotopic model.

(A) Orthotopic coinjection experiment using GFP-1 KRAB or B500 NID2 KRAB CAFs (75%) with luciferase cancer cells (25%). Vehicle mice culled day 44 (n = 8 GFP-1 KRAB and n = 6 B500 NID2 KRAB). Chemotherapy mice culled day 65 (n = 7 GFP-1 KRAB and n = 7 B500 NID2 KRAB). (B) IVIS signal (day 7) for GFP-1 KRAB and B500 NID2 KRAB mice with Fstop1, exp30s. Color scale is radiance. (C) Pancreatic tumor weight (in grams). Welch’s t test, ns: P > 0.05. (D) Number of visible liver metastases. Welch’s t test, *P < 0.05. (E) Representative livers at end point. Scale bars, 1 cm. (F) Liver metastasis coverage. Welch’s t test, ***P < 0.001. (G) H&E liver sections (top) and QuPath detection (bottom). Scale bars, 5 mm. (H) Representative ROIs of Picrosirius Red with coverage (%) quantification. Welch’s t test, *P < 0.05. Scale bars, 100 μm. (I) Representative ROIs of polarized light Picrosirius Red with coverage (%) quantification. Welch’s t test, ***P < 0.001. Scale bars, 100 μm. (J) Pancreatic tumor weight (in grams). Welch’s t test, **P < 0.01. (K) Number of visible liver metastases. Welch’s t test, *P < 0.05. (L) Representative livers at end point. Scale bars, 1 cm. (M) Liver metastasis coverage. Mann-Whitney test, **P < 0.01. (N) H&E liver sections (top) and QuPath detection (bottom). Scale bars, 5 mm. (O) Representative ROIs of Picrosirius Red with coverage (%) quantification. Welch’s t test, **P < 0.01. Scale bars, 100 μm. (P) Representative ROIs of polarized light Picrosirius Red with coverage (%) quantification. Welch’s t test, *P < 0.05. Scale bars, 100 μm. Schematics were created with Biorender.com.