Personalized PDAC chip with functional endothelial barrier for tumour biomarker detection: A platform for precision medicine applications

Abstract

Pancreatic ductal adenocarcinoma (PDAC) is a highly aggressive cancer characterised by poor survival rates and an increasing global incidence. Advances in the staging and categorization of pancreatic tumours, along with the discovery of functional mutations, have made precision treatments possible, which may lead to better clinical results. To further improve customized treatment approaches, in vitro models that can be used for functional drug sensitivity testing and precisely mimic the disease at the organ level are required. In this study, we present a workflow for creating a personalized PDAC chip utilising primary tumour-derived human pancreatic organoids (hPOs) and Human Umbilical Vein Endothelial Cells (HUVECs) to simulate the vascular barrier and tumour interactions within a PDMS-free organ-on-a-chip system. The patient PDAC tissue, expanded as tumour hPOs, could be cultured as adherent cells on the chip for more than 50 days, allowing continuous monitoring of cell viability through outflows from tumour and endothelial channels. Our findings demonstrate a gradual increase in cell density and cell turnover in the pancreatic tumor channel. Tumour-specific biomarkers, including CA-19.9, TIMP-1, Osteopontin, MIC-1, ICAM-1 and sAXL were consistently detected in the PDAC chip outflows. Comparative analyses between tissue culture plates and microfluidic conditions revealed significant differences in biomarker secretion patterns, highlighting the advantages of the microfluidics approach. This PDAC chip provides a stable, reproducible tumour model system with a functional endothelial cell barrier, suitable for drug sensitivity and secretory biomarker studies, thus serving as a platform for functional precision medicine application and multi-organ chip development.

Pancreatic ductal adenocarcinoma (PDAC) organ-on-a-chip for biomarker detection.

Fig. 1

PDAC chip setup scheme. (A) An illustration of the chip cross section with primary pancreatic ductal adenocarcinoma (PDAC) tumour cells in the upper channel and endothelial (HUVEC) cells in the bottom channel separated by a porous PET membrane. A brightfield image of the PDAC chip channel. (B) A diagram showing the flow direction in upper and bottom channels of the chip with the cell seeding membrane area highlighted. (C, D) A schematic of the microfluidic setup with the chip and outflow collection tubes in the incubator.

Fig. 2

Generation and characterisation of human pancreatic tumour organoids (hPOs). (A) Tumour hPOs were isolated from the donor material obtained during a tumour resection procedure. (B) An allele fraction heat map showing pathogenic mutations identified by targeted DNA sequencing in the original donor tissue and tumour hPOs expanded to passage five. (C) Microscopy images of the corresponding tumour hPOs. A bright-field image of a gel dome containing tumour hPOs and immunofluorescent images of a whole-mount (CK19) and cryo section preparations showing the expression of epithelial (CDH1, CK19), pancreatic progenitor (PDX1, SOX9) and tumour (MUC5AC, GATA6) markers. Scale bars – 100 μm.

Fig. 3

Functional characterisation of the PDAC chip. (A) A diagram depicting cell seeding and cultivation timeline for the PDAC chip. (B) Bright-field images of the dissociated hPOs seeded in the tissue culture plate well and on the chip membrane. (C) A series of bright-field images of the hPOs on the PDAC chip taken from the same channel at different time points. (D) Immunofluorescent phenotyping of the dissociated tumour hPOs on the chip membrane fixed at D18. (E) Tumour hPO cells on the chip after more than 50 days of culture labelled with anti-CK19 antibody. (F) EpCAM-positive cells detected in the tumour channel outflow of the PDAC chip. (G) HUVEC cell morphology and immunophenotyping on the plate and in the chip. (H) Apparent permeability (P_app) values for the endothelial/epithelial barrier of the PDAC chip under various flow rate conditions obtained using cascade blue (n = 3 channels). A fluorescent image of the cascade blue visible in the bottom channel of a PDAC chip under 4 μl/min flow rate. Scale bars: (B) – 400 μm; (D,E) – 100 μm; (G) – 400 μm for bright-field images, 100 μm for fluorescent; (H) – 1000 μm. ∗P < 0.05, ∗∗∗p < 0.001 (n = 3-6).

Fig. 4

Cell viability on the PDAC chip. LDH signal detected in the chip outflows (A) and conditioned media from the tissue culture wells (B) in tumour hPO (red line) and HUVEC (blue line) samples, displayed as optical density (OD) values (the dotted line represents LDH signal derived from the respective fresh media). Fractalkine concentration values detected in the PDAC chip outflows (C) and conditioned media from the tissue culture wells (D). (E) Combined values of the fractalkine concentration levels detected in both pancreatic tumour and HUVEC channel outflows of the PDAC chip, measured at different time points one week apart. (F) Representative immunofluorescence images of Caspase 3 staining performed on pancreatic tumour and HUVEC cells on the chip and in the tissue culture wells. (G) Quantification of the Caspase 3 staining in dissociated tumour hPOs and HUVEC in the well and on the chip. Number of replicates (n): (A) n = 4 replicates from two individual channels for each condition. (B) n = 2 wells for each condition; (C) n = 9 measurements at different time points from a total of 5 channels across 2 chips for each condition; (E) n = 4 channels from 2 chips; (G, left panel) high magnification images from 3 individual PDAC and HUVEC channels. (G, right panel) n = 3 high magnification images from 3 different PDAC/HUVEC wells; Scale bars – 100 μm. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗p < 0.001.

Fig. 5

PDAC biomarker analysis. Concentration means of various PDAC biomarkers detected at different time points (T1 and T2) in tumour hPO (blue lines) and HUVEC (red lines) channel outflows in two different chip setups. (A) Chip #1 was cultured at a 2 μl/min flow rate and outflows collected at T1 = D11, and T2 = D18. (B) Chip #2 was cultured at a 4 μl/min flow rate, outflows collected at T1 = D22, and T2 = D29. (C) Biomarker concentration in the conditioned medium from the tumour hPOs (red bars) and HUVECs (blue bars) cultivated as adherent cells in tissue culture wells. (D) A diagram showing the outflow collection time points (T1 and T2) for each chip. (E) The changes in biomarker concentration in both tumour and HUVEC outflows combined, expressed as the ratio between concentration detected on T2 vs. T1. Number of replicates (n): (A) n = 1 channel outflow at T1 and n = 2 channel outflows at T2 for each condition; (B) n = 3 channel outflows at T1 and T2 for each condition; (C) conditioned media from n = 3 wells analysed for each condition. (E) n = 4 channels across two chips. ∗P < 0.05,∗∗P < 0.01, ∗∗∗P < 0.001.