Development of a polarized pancreatic ductular cell epithelium for physiological studies

Abstract

Pancreatic ductular epithelial cells comprise the majority of duct cells in pancreas, control cystic fibrosis transmembrane conductance regulator (CFTR)-dependent bicarbonate ([Formula: see text]) secretion, but are difficult to grow as a polarized monolayer. Using NIH-3T3-J2 fibroblast feeder cells and a Rho-associated kinase inhibitor, we produced well-differentiated and polarized porcine pancreatic ductular epithelial cells. Cells grown on semipermeable filters at the air-liquid interface developed typical epithelial cell morphology and stable transepithelial resistance and expressed epithelial cell markers (zona occludens-1 and β-catenin), duct cell markers (SOX-9 and CFTR), but no acinar (amylase) or islet cell (chromogranin) markers. Polarized cells were studied in Ussing chambers bathed in Krebs-Ringer [Formula: see text] solution at 37°C gassed with 5% CO₂ to measure short-circuit currents ( I_sc). Ratiometric measurement of extracellular pH was performed with fluorescent SNARF-conjugated dextran at 5% CO₂. Cells demonstrated a baseline I_sc (12.2 ± 3.2 μA/cm²) that increased significantly in response to apical forskolin-IBMX (∆ I_sc: 35.4 ± 3.8 μA/cm², P < 0.001) or basolateral secretin (∆ I_sc: 31.4 ± 2.5 μA/cm², P < 0.001), both of which increase cellular levels of cAMP. Subsequent addition of apical GlyH-101, a CFTR inhibitor, decreased the current (∆ I_sc: 20.4 ± 3.8 μA/cm², P < 0.01). Extracellular pH and [Formula: see text] concentration increased significantly after forskolin-IBMX (pH: 7.18 ± 0.23 vs. 7.53 ± 0.19; [Formula: see text] concentration, 14.5 ± 5.9 vs. 31.8 ± 13.4 mM; P < 0.05 for both). We demonstrate the development of a polarized pancreatic ductular epithelial cell epithelium with CFTR-dependent [Formula: see text] secretion in response to secretin and cAMP. This model is highly relevant, as porcine pancreas physiology is very similar to humans and pancreatic damage in the cystic fibrosis pig model recapitulates that of humans. NEW & NOTEWORTHY Pancreas ductular epithelial cells control cystic fibrosis transmembrane conductance regulator (CFTR)-dependent bicarbonate secretion. Their function is critical because when CFTR is deficient in cystic fibrosis bicarbonate secretion is lost and the pancreas is damaged. Mechanisms that control pancreatic bicarbonate secretion are incompletely understood. We generated well-differentiated and polarized porcine pancreatic ductular epithelial cells and demonstrated feasibility of bicarbonate secretion. This novel method will advance our understanding of pancreas physiology and mechanisms of bicarbonate secretion.

Keywords: bicarbonate secretion; cystic fibrosis; duct cells; pancreas.

Fig. 1.

Isolation of acinar-ductular complexes from pancreas. Sample of pancreas was removed and minced (1), followed by collagenase digestion for 15 min (2); sample was then filtered, supernatant was discarded, and pellet was resuspended (3) and plated in a petri dish overnight (4). Acinar-ductular complexes (arrows) were identified by swelling (5) and hand-picked into pipettes and transferred into 24-well plates (6); remainder was saved for islet cell cultures.

Fig. 2.

Acinar-ductular complexes: confocal (A, B, and D–H) and direct (C and I) microscopy images of acinar-ductular complexes showing amylase for acinar cells, Sex-determining region Y-box-9 (SOX-9) for ductular cells, and Hoechst dye for nuclei days 1 (A–D) and 2 (E–I) after isolation. Images are representative of 4 separate experiments in each group from 4 different animals.

Fig. 3.

Pancreatic ductular epithelial cell cultures. Acinar-ductular complexes were observed in culture for ~7 days (1); once growth of typical epithelial cells was identified, cells were treated with F medium that contained ROCK inhibitor and fibroblast feeder cells (2); cell growth was continuously monitored under microscopy, and cells that grew poorly or contaminated with fibroblasts were discarded (3). Once consistent epithelial growth was established, cells were removed gently and plated on filters (4); 2 days later, apical medium was removed to grow cells at the air-liquid interface (ALI) supplemented with differentiation medium (5). Cells grown under these conditions developed stable transepithelial resistance (TER) 5–7 days after plating.

Fig. 4.

Growth and expansion of pancreatic ductular epithelial cell cultures: microscopic images of successful (A–D) and failed (E and F) pancreatic ductular culture on day 2 (A and E), day 5 (B and F), and day 20 (C and D). Images are representative of 4 separate experiments in each group from 4 different animals.

Fig. 5.

Pancreatic ductular cells show typical duct epithelial cell characteristics. A and B: cells at passage 2 (A) or 6 (B) were plated on semipermeable filters; 2 days later apical medium was removed to grow cells at the air-liquid interface supplemented with differentiation medium, and the transepithelial resistance (TER) was measured daily with an ohmmeter. Images are representative of 6 separate experiments in each group. C–I: immunofluorescent images of pancreatic ductular epithelial cell cultures show β-catenin (C), ZO-1 (D and G), Hoechst dye (E and H), SOX-9 (F), and ZO-1 and SOX-9 (I, merged). Images are representative of 4 separate experiments from 4 different animals.

Fig. 6.

Pancreatic ductular cells express cystic fibrosis (CF) transmembrane conductance regulator (CFTR) and no acinar or islet markers. A and B: cells were lysed, and RNA was subjected to qRT-PCR for CFTR or chromogranin. Actin was the housekeeping gene. Negative control for CFTR was CF airway epithelial cells; positive controls for chromogranin were islet cells identified the day after isolation. Each sample was run in triplicate. Cells expressed CFTR mRNA (A) and no chromogranin A (B). n = 3. *P < 0.01. C–E: immunofluorescent images of pancreatic ductular epithelial cell cultures showing no amylase (C and D) with day 1 acinar-ductular cultures used as positive control (E). C and E: amylase. D: Hoechst dye. Images are representative of 4 separate experiments from 4 different animals.

Fig. 7.

Pancreatic ductular epithelial cells have CFTR-mediated, cAMP-stimulated $HC{O}_3^{-}$ secretion. A and B: Ussing chamber experiments. Representative short-circuit current (I_sc) tracings from porcine pancreatic ductular epithelial cells mounted in Ussing chambers. Cells showed no response in I_sc to apical amiloride (100 µM) or apical DIDS (100 µM) but a significant increase to forskolin 10 µM and IBMX 100 µM (A and B) or secretin 20 nM (B) (P < 0.001), inhibited by GlyH (100 µM) (P < 0.01). Experiments were conducted in Krebs-Ringer $HC{O}_3^{-}$ solution, at 37°C and gassed with 5% CO₂. Data are representative of 6 independent experiments from 6 different animals. C–E: luminal pH measurements. C: pH calibration curve used to calculate pH values. The empirical pK_a of SNARF-1 dextran at 37°C was 7.23. D: a baseline pH value was obtained (time = 0), then forskolin was added to the basolateral medium to a final concentration of 10 µM. Alkalization was evident at 10 min and maintained for at least 2 h (P = 0.03). E: pH values from B obtained under 5% CO₂ were converted to [$HC{O}_3^{-}$] with the Henderson-Hasselbalch equation. After 2 h, apical [$HC{O}_3^{-}$] was significantly doubled compared with baseline (P = 0.02). Results are representative of 3 separate experiments from 3 different animals. A, amiloride; D, DIDS; F+I, forskolin+IBMX; G, GlyH; S, secretin.