3D pancreatic carcinoma spheroids induce a matrix-rich, chemoresistant phenotype offering a better model for drug testing

Abstract

Background: Pancreatic ductal adenocarcinoma (PDAC) is the fourth most common cause of cancer related death. It is lethal in nearly all patients, due to an almost complete chemoresistance. Most if not all drugs that pass preclinical tests successfully, fail miserably in the patient. This raises the question whether traditional 2D cell culture is the correct tool for drug screening. The objective of this study is to develop a simple, high-throughput 3D model of human PDAC cell lines, and to explore mechanisms underlying the transition from 2D to 3D that might be responsible for chemoresistance.

Methods: Several established human PDAC and a KPC mouse cell lines were tested, whereby Panc-1 was studied in more detail. 3D spheroid formation was facilitated with methylcellulose. Spheroids were studied morphologically, electron microscopically and by qRT-PCR for selected matrix genes, related factors and miRNA. Metabolic studies were performed, and a panel of novel drugs was tested against gemcitabine.

Results: Comparing 3D to 2D cell culture, matrix proteins were significantly increased as were lumican, SNED1, DARP32, and miR-146a. Cell metabolism in 3D was shifted towards glycolysis. All drugs tested were less effective in 3D, except for allicin, MT100 and AX, which demonstrated effect.

Conclusions: We developed a high-throughput 3D cell culture drug screening system for pancreatic cancer, which displays a strongly increased chemoresistance. Features associated to the 3D cell model are increased expression of matrix proteins and miRNA as well as stromal markers such as PPP1R1B and SNED1. This is supporting the concept of cell adhesion mediated drug resistance.

Figure 1

Spheroid development. [A] Different PDAC cell lines grown in 3D culture for 4 days. According to the grading system by Sipos et al.[6], Panc-1, BXPC3 and ASPC-1 are poorly differentiated and carry both mutant KRAS and p53 (Panc-1) or either p53 (BXPC3) or KRAS (ASPC-1), whereas Capan-1 is a well differentiated PDAC cell line. A previously established pancreatic stellate cell line (PSC) is also included as a non-transformed control cell line. [B] Development of a single representative Panc-1 spheroid, photographed from day 2 to day 8 by counting with the Boyden chamber the cell number of trypsinized spheroids and taking pictures of spheroids at fixed time points. [C] Cell counts from 2D and 3D cultures at different time points. Bars correspond to 500 μm.

Figure 2

Morphological analysis of 3D cultured Panc-1 cells. [A] Hematoxilin Eosin staining and [B-D] electron microscopy analysis of a central 7 days spheroid section. C and D show details of the same section at EM. In C the arrow indicates the presence of a lumen. In D tight junction structures are indicated by arrows. [E] E-cadherin expression in both 2D and 3D culture assessed by RT-PCR. Data are calculated as expression ratio 3D/2D. [F] Western blotting shows E-cadherin protein expression in 2D and 3D culture on day 4 and 7.

Figure 3

Metabolical and physiological analyses comparing 2D with 3D culture. [A] Energy metabolism: lactate accumulation measurements in 3D compared to 2D culture. Data are expressed as ratio μmoles lactate/ μg protein. [B] Western immunoblotting showing HIF1a protein expression both in 2D and 3D cultures at day 4 and 7. [C] mRNA expression of target genes downstream HIF1a in 2D and 3D cultures. Real Time-PCR data are calculated as expression ratio 3D/2D. A representative experiment out of three is shown. [D] Western blotting shows Glut1 protein expression both in 2D and 3D cultures on day 4 and 7. GAPDH is used as loading control.

Figure 4

Analyses of chemoresistance related genes. [A] Time course of mRNA expression in 2D and 3D cultures of drug resistance-involved genes. Real Time-PCR data are calculated as expression ratio 3D/2D. [B] Time course expression of drug resistance-relevant miRNAs in 2D and 3D cultures. Real Time-PCR data are calculated as expression ratio 3D/2D. A representative experiment out of three is shown.

Figure 5

Analyses of extracellular matrix related genes. [A] mRNA expression of ECM relevant genes in 2D and 3D cultures. Real Time-PCR data are calculated as expression ratio 3D/2D. A representative experiment is shown [B] Collagen I and fibronectin I staining of a central section from 7 day spheroids (right). One representative picture is shown. Secondary antibody alone staining is used as negative control (CTR, left). [C] ECM-relevant time course of selected miRNA expression in 2D and 3D cultures. Real Time-PCR data are calculated as expression ratio 3D/2D. A representative experiment out of three is shown.

Figure 6

Increased chemoresistance against gemcitabine in 3D culture. Cell viability after Gemcitabine treatment of different PDAC cell lines grown in 2D and 3D culture. A Pancreatic Stellate Cell line (PSC) is included as non-transformed control cell line. Data are plotted as percentage of untreated control cells. A representative experiment out of three is shown.

Figure 7

Comparison of chemoresistance between 2D and 3D culture using multiple cytotoxic compounds. Histogram summarizing the results from viability assays performed on 2D and 3D Panc-1 cell cultures. Different drugs were used at the indicated concentrations. Data are plotted as percentage of the respective untreated control (CTR) and each drug was tested three times in octuplets. Gem: gemcitabine. All: allicin. AX: AXP-107-11.

Figure 8

Spheroids from mouse pancreatic cancer. [A] PDAC cells from a mouse bearing mutated Kras and Trp53 in the pancreas (KPC cells) are compared to the human PDAC Panc-1 cell line in their ability to form spheroids when grown under 3D culture conditions. [B] Drug assay performed on KPC cells. Different compounds are used at the indicated concentrations and cell viability of 2D versus 3D culture is compared. Data are plotted as percentage of the respective untreated control (CTR). A representative experiment out of three is shown.

Figure 9

3D spheroid model. Representation of the characteristics of the 3D spheroid model in relation to the holistic hallmarks of cancer according to Hanahan and Weinberg [59] observed in the reductionist pancreatic cancer monospheroid model.