Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Apr 30.
Published in final edited form as: Methods Mol Biol. 2022;2373:57–68. doi: 10.1007/978-1-0716-1693-2_4

Bile Duct-On-a-Chip

Yu Du 1,5, William J Polacheck 2, Rebecca G Wells 1,3,4,5
PMCID: PMC9056007  NIHMSID: NIHMS1801069  PMID: 34520006

Abstract

Cholangiopathies affect the biliary tree via various pathophysiological mechanisms. Research on biliary physiology and pathology, however, is hampered by a lack of physiologically-relevant in vitro models. Conventional models, such as two-dimensional (2D) monolayers and organoids, fail to replicate the structural organization of the bile duct, and both the size of the duct and position of cells are difficult to manipulate in a controllable way. Here, we describe a bile duct-on-a-chip (BDOC) that phenocopies the open-ended tubular architecture of the bile duct in three dimensions which, when seeded with either a cholangiocyte cell line or primary cells, demonstrates barrier function similar to bile ducts in vivo. This device represents an in vitro platform to study the pathophysiology of the bile duct using cholangiocytes from a variety of sources.

Keywords: Microfluidics, organ-on-chip, cholangiopathy, bile duct epithelium, biliary model, 3D cell culture, organotypic model

1. Introduction

Cholangiopathies, which are characterized by inflammation and fibrosis of the bile ducts and, in many cases, progression to liver cirrhosis and liver failure, are often associated with impaired tight junction integrity [1–3]. Research on biliary physiology and pathology would benefit from the ability to manipulate tubular structures and test the impact of these manipulations on epithelial permeability, but these capabilities are lacking in current in vitro models. Even organoids cultured in 3D extracellular matrix (ECM), which have many of the functions and architectural features of the bile duct (including lumens, polarized cells, and bile salt transport), fail to recapitulate the tubular structure of bile ducts and cannot be perfused or easily assayed for permeability. Additionally, their heterogeneity in size even in a single experiment limits their usefulness for quantitative studies.

In recent years, microfluidic organs-on-chips have been used to model the physiology of cell-cell and cell-ECM junctions in tissues including the alveolar-capillary interface [4], the blood-brain barrier [5–7], and liver sinusoids [8], in a controllable way. These systems have allowed high-resolution imaging and a variety of quantitative analyses. The 3D structure, controllability, and ability to integrate matrix and other physical cues into the devices led us to explore the potential of applying organ-on-chip technology to the study of cholangiopathies. Here, we describe the technology we have developed for engineering a micro-bile duct with controllable architecture. The structure of the device is based on a device used to model the vasculature [9,10]. We demonstrate that a self-organized cholangiocyte-lined channel faithfully recapitulates the tubular structure and barrier function of the bile duct. This technology represents a novel in vitro model to investigate bile duct function and is adaptable to the use of cholangiocytes from a variety of sources. We have found it useful in studying the biology of the bile ducts under both perturbed and unperturbed states [11].

2. Materials

2.1. Mold fabrication

  1. SU-8 2002/2010/2100/2150 photoresist.

  2. Microposit S1813 photoresist.

  3. Propylene glycol monomethyl ether acetate (PGMEA).

  4. Isopropyl alcohol (IPA).

  5. Trichloro(1H,1H,2H,2H-perfluorooctyl)silane.

  6. Film transparency masks.

  7. Glass dish.

  8. 3 inches silicon wafers.

  9. Calipers.

  10. Blocking layer: Combine 70 mL SU-8 2010 and 30 mL S1813 in a dust free bottle. Cover with aluminum foil and shake on heated orbital shaker at 37°C ON. Allow solution to cool to room temperature and store for >6 h to remove bubbles from solution. Sealed solution can be stored at room temperature (20°C) for 2 months.

2.2. Device fabrication

  1. Biopsy punches (2 and 5 mm in diameter).

  2. Polydimethylsiloxane (PDMS; Sylgard 184).

  3. 24×40 mm coverslips.

  4. Transparent tape.

  5. 0.01% (v/v) poly-L-lysine solution: Dilute 1 ml of 0.1% stock solution into 9 ml dH2O; stable stored at RT for one year.

  6. 0.5% (v/v) glutaraldehyde solution: Dilute 200 μl 50% glutaraldehyde stock solution into 10 ml dH2O, shield from light; stable stored at RT for one year.

  7. 70% ethanol.

  8. Acupuncture needles (160 μm ×30 mm).

  9. Blocking solution I: 2% bovine serum albumin (BSA) dissolved in dH2O, filtered with 0.22 μm filter.

  10. Vacuum grease.

  11. dH2O, 10x DMEM, 1M NaOH, 250 mM HEPES, 5% (w/v) NaCO3.

  12. 2.5 ml collagen solution: Calculate the volumes of dH2O, 10x DMEM, NaOH, and collagen required for gels with a final concentration of 2.5 mg/ml collagen in 1x phosphate buffered saline (PBS), using a volume of 1 M NaOH equal to 0.023 times the volume of collagen added. On ice, mix together dH2O, 10x DMEM, and 1 M NaOH, NaCO3 then add collagen and pipette well to ensure even mixing. Centrifuge the collagen solution for 3 min at 2012xg at 4˚C to remove the bubbles (see Note 1).

  13. 100 μg/ml laminin solution: 100 μl of 1 mg/ml laminin stock solution diluted in 900 μl of PBS; stable stored at 4 ˚C for one month.

2.3. Cell culture

  1. A line of small intrahepatic cholangiocytes, originally isolated from normal mice (BALB/c) and immortalized by transfection with the SV40 large-T antigen [12].

  2. Cell culture medium made as Table 1 [13]; stable stored at −20˚C for 3 months.

  3. 70 μm sterile filter.

  4. 0.05% Trypsin-EDTA (1×).

Table 1.

Media components.

Component Volume (ml) Stock concentration
DMEM/F12 402.1
Pen-Strep 5 ×100
MEM vitamin solution 5 ×100
L-glutamine 5 200 mM (×100)
Bovine pituitary extract 1.1 14mg/ml (×467)
Dexamethasone 0.5 393 ug/ml (×1000)
3’,3’,5’-triiodo-L-thyromine 0.5 3.4 mg/ml (×1000)
Epidermal growth factor 0.5 25 μg/ml (×1000)
Forskolin 5 0.411 mg/ml (×1000)
Fungizone 2
FBS 50
Soybean trypsin inhibitor 5 5mg/ml (×100)
MEM non-essential amino acid 5 10 mM (×100)
Insulin-transferrin-selenium 5 ×100
Na pyruvate 5
Chemically defined lipid concentrate 5 ×100
Gentamicin 0.2 50 mg/ml (×2500)
Ethanolamine 0.13
Total volume 502
Open in a new tab

2.4. Fixation and staining

  1. 4% paraformaldehyde (PFA).

  2. 0.3% Triton X-100.

  3. 4′,6-diamidino-2-phenylindole (DAPI).

  4. Phalloidin (1:100).

  5. Blocking solution II: 2% BSA dissolved in PBS, filtered by 0.22 μm filter.

2.5. Imaging and permeability assay

  1. Dextran solution: fluorescent dextran (70, 10, and 4 kDa, labeled with fluorescein isothiocyanate (FITC)) in PBS at 20 μg/mL, shield from light; stable stored at 4˚C for one month.

  2. Hank’s balanced salt solution (HBSS) (with Ca, Mg).

2.6. Equipment

  1. AutoCAD software.

  2. Adobe illustrator.

  3. UV mask aligner.

  4. Dissection microscope.

  5. Spin coater.

  6. Programmable hotplate.

  7. Ultrasonic cleaner.

  8. Plasma etcher.

  9. Fume hood.

  10. EVOS FL Auto 2 Imaging System.

  11. SCTR Leica confocal microscope and Leica application suite.

  12. Olympus IX81 spin disk confocal microscope.

  13. Desiccator.

  14. Rocker (5–15 rpm).

  15. Nitrogen tank with air gun.

3. Methods

3.1 Silicon master design and fabrication

  1. Draw transparency mask using AutoDesk AutoCAD.

  2. Have a vendor or core facility print high-resolution (≥25k DPI) film transparency mask.

  3. Heat silicon wafer at 200°C for 30 min on hot plate to dehumidify. Allow the wafer to cool to room temperature prior to further processing.

  4. Treat the wafer with oxygen plasma for 5 min at full power to clean wafer surface.

  5. To form an adhesion layer that promotes binding of subsequent layers to the silicon wafer, center the wafer in the spin coater and dispense SU-8 2002 onto the center of the wafer. Pre-coat wafer by spinning at 500 rpm for 10 s, then ramp the spin rate to 2000 rpm for 60 s.

  6. To evaporate photoresist solvent, heat the wafer for 2 min at 95°C.

  7. Flood expose (do not use transparency mask) the wafer with an exposure dose of ~100 mJ cm−2 to cure the adhesion layer.

  8. Return the wafer to a hotplate at 95°C for 2 min.

  9. To form the first structural layer of the device, return the wafer to the spin coater, and dispense SU-8 2100 photoresist onto the center of the wafer. Again, spread the photoresist by spinning at 500 rpm for 60s before ramping to 2000 rpm for 60s .

  10. Heat the wafer to 65°C for 5 min, then increase the temperature to 95°C for 3.5 h. Turn off heat source and allow the wafer to slowly return to room temperature.

  11. Return the wafer to the spin coater for the blocking layer. Dispense blocking solution onto the center of the wafer. To spread the photoresist, spin the wafer at 500 rpm for 60 s, then ramp the spin rate to 1000 rpm and hold for 60 s to create an even blocking layer.

  12. Heat the wafer to 95°C for 30 min.

  13. Load the wafer and transparency mask into a mask aligner, align the mask to the wafer, and expose the wafer with a dose of ~600 mJ cm-2.

  14. Heat the wafer for 5 min at 65°C and 12 min at 95°C.

  15. Return the wafer to the spin coater for the second structural layer. Dispense SU-8 2150 onto the surface of the wafer. Spread the photoresist at 500 rpm for 60 s prior to ramping to 2000 rpm for 60 s.

  16. Heat the wafer for 5 min at 65°C and 80 min at 95°C.

  17. Load the wafer and second transparency mask into a mask aligner, align the mask to the previously exposed structures on the wafer, and expose the wafer with a dose of ~300 mJ cm-2.

  18. Heat the wafer for 5 min at 65°C and 12 min at 95°C.

  19. Return the wafer to the spin coater for the third and final structural layer. Dispense SU-8 2150 onto the surface of the wafer. Spread the photoresist at 500 rpm for 60 s prior to ramping to 1400 rpm for 60 s.

  20. Heat the wafer for 5 min at 65°C and 80 min at 95°C.

  21. Load the wafer and third transparency mask into a mask aligner, align the mask to the previously exposed structures on the wafer, and expose the wafer with a dose of ~300 mJ cm-2.

  22. Heat the wafer for 5 min at 65°C and 12 min at 95°C.

  23. Transfer wafer to a bath of PGMEA and agitate on an orbital shaker for 10 min. Wash with IPA and inspect features on a dissection microscope. If uncrosslinked features have not been fully removed, repeat PGMEA bath cycles until features are fully developed.

3.2. Cell isolation and culture

  1. Multiple cholangiocyte sources can be used in the device (see Note 4). The small cholangiocyte cell line should be used at a passage number not exceeding 9. Culture cells in 100 mm petri dishes. To maintain the cells, wash with 10 ml PBS after aspirating the culture medium. Add 1 ml trypsin per dish. Incubate in the incubator for 60 s to dissociate cells. Neutralize the trypsin with 1 ml culture medium, and transfer the cell suspension into a 15 ml tube. Centrifuge at 300xg for 5 min. Resuspend the cells in culture medium, and seed into new dishes in 1:10 ratio.

3.3. PDMS device preparation

  1. Mix the PDMS silicone elastomer base and curing agent thoroughly in a weight ratio of 10:1 (base:curing agent).

  2. Pour the mixed PDMS to the mold (25 g per mold).

  3. Degas the PDMS mixture until there are no visible bubbles using a vacuum desiccator.

  4. Cure the PDMS gel fully in an oven at 75°C for 1.5 h.

  5. Peel the PDMS gel off carefully from the mold.

  6. Cut the gel into bricks of equal size.

  7. Punch reservoir ports with a 5 mm biopsy punch, and side ports with a 2 mm punch.

  8. Clean the PDMS stamps with tape by pressing on all surfaces, and coverslips using a nitrogen gun.

  9. Plasma treat the PDMS and the coverslips for 45 s using a plasma etcher.

  10. Align the PDMS and the coverslip, press gently to ensure fully bonded (Fig. 1).

  11. Inject 0.0.1% (v/v) poly-L-lysine to the chamber through side port; incubate for 1 h at RT.

  12. Suck out the poly-L-lysine and incubate with 0.5% (v/v) glutaraldehyde for 20 min.

  13. Wash the devices with dH2O x3.

  14. Immerse the devices in dH2O, sonicate for 30 min, and immerse in dH2O overnight.

Fig. 1.

Fig. 1.

Open in a new tab

Flow chart showing fabrication process.

3.4. Collagen gel channel preparation

  1. Incubate 160 μm acupuncture needles in blocking solution I for 2 h (see Note 2).

  2. Rinse the devices with 70% ethanol for 30 min.

  3. Air dry the devices using nitrogen gun.

  4. Insert the needles through the channels reaching second reservoir port.

  5. Sterilize the devices under UV for 20 min.

  6. Place devices, rat tail collagen, sterile 10x DMEM, sterile dH2O, and sterile 1 M NaOH on ice.

  7. Precool the centrifuge to 4˚C.

  8. Inject the collagen solution through side ports to fill the chamber (see Note 3).

  9. Allow collagen to gel by incubation at 37˚C for 20 min.

  10. Add a drop of PBS to the tops of the side ports, and fill the reservoir ports with PBS. Incubate at 37˚C ON.

  11. Gently pull the needles out, and seal the hole at the end with vacuum grease.

  12. Add 100 μl laminin solution to the reservoir ports of each device and incubate on rocker at 37˚C for 4 h.

  13. Wash 3 times with PBS before seeding cells.

3.5. Cell seeding and culture

  1. The cholangiocyte cell line is ready to seed when it reaches 70% confluency. Use 0.5% trypsin to digest the cells and resuspend cells in culture medium to a final concentration of 1×106 cells/ml.

  2. To seed cells, use a 200 μl plastic pipette to remove roughly half of the collagen gel in each side port. Add 40 μl cell suspension to one reservoir port, and 30 μl to the other. Flip the dish and devices after 30 s, and incubate for 2 min to allow cells to adhere to the top surface of the channel. Flip the devices back, and incubate for 5 min to allow cells to adhere to the bottom of the channel (see Note 5). Scrape cells in the reservoir ports with 100 μl tips, and add fresh culture medium. Put two caps from 15 ml tubes in each 100 mm dish, and fill the caps with dH2O to keep the humidity, replenishing the dH2O daily. Replace the medium every day until cells reach optimal confluency (Fig. 2A, C).

Fig. 2.

Fig. 2.

Open in a new tab

A. Top view of a bile duct-on-a-chip. B. Representative immunofluorescent image of the cholangiocyte tube in the device, stained with F-actin (red) and nuclei (DAPI, blue). C. Representative bright field image of the cholangiocytes lining channel. D. Fluorescent FITC-dextran (70, 10, 4 kDa) in the channel, imaged after 15 min. Scale bar, 100 μm (B, C), 200 μm (D). Figure adapted from Du Y, Khandekar G, Llewellyn J, Polacheck W, Chen CS, Wells RG. A Bile Duct-on-a-Chip With Organ-Level Functions. Hepatology 2020;71:1350–1363. under the terms of the Creative Commons Attribution Non-Commercial License CC BY-NC.

3.6. Permeability testing

  1. When cells reach optimal confluency, replace the culture medium with phenol-free culture medium, and incubate for 1 h on the rocker at 37˚C, 5% CO2.

  2. Preset the incubator of the EVOS FL Auto 2 Imaging System to 37˚C, humidity and 5% CO2.

  3. Aspirate the culture medium. Add 70 kDa fluorescent dextran solution to the left reservoir port, and keep the devices in the microscope incubator. Start imaging after 15 min.

  4. Wash the cholangiocyte channel with phenol-free HBSS (with Ca, Mg) twice.

  5. Repeat the same process for 10 kDa and 4 kDa dextran, sequentially (Fig. 2D).

  6. Fluorescent images can be taken sequentially with the EVOS FL Auto 2 Imaging System .

3.7. Staining

  1. Fix cholangiocytes in channels with 4% paraformaldehyde (PFA) at 37˚C for 20 min with rocking.

  2. Rinse cells 3 times with PBS.

  3. Permeabilize the cells using 0.3% Triton X-100 at 4˚C with rocking overnight.

  4. Block with blocking solution II at 4˚C with rocking overnight.

  5. Incubate with rhodamine diluted in 1:100, together with DAPI (1:5000) diluted in blocking solution II, at 4°C overnight on the rocker (see Note 6).

  6. Rinse 3 times with PBS for 5 min each on the rocker, followed by an overnight rinse.

4. Notes

  1. Pipette very slowly to avoid bubble formation in the gel. Solution color should be light pink. Make the collagen solution fresh before use, keep on ice, and use within 30 mins.

  2. Sonicate the needles in 70% ethanol for 30 min, and rinse with dH2O before use.

  3. Use about 80 μl collagen solution for each device. Mix each independently to avoid leaving collagen solution on ice for too long, which will lead to fibrous collagen gel.

  4. To use primary mouse cholangiocytes, isolate from extrahepatic bile ducts of wild-type adult mice (BALB/c) by the method described [13]. For primary extrahepatic cholangiocyte seeding, scrape the collagen gel off the plate gently so that it floats. Add 1 ml of collagenase solution per 60 mm dish to the top of the floating collagen gel, and incubate for 30 min at 37˚C, 5% CO2. Transfer cell sheets into a 15 ml tube. Spin down at 300xg for 5 min, discard the supernatant and resuspend in a similar volume of DMEM. Spin again at 300xg for 5 min. Remove supernatant. Resuspend pellet in 3 ml of 0.5% trypsin, and incubate 5 min at 37˚C, 5% CO2. Pipette up and down to break up the cholangiocyte sheets. Neutralize the trypsin with 5 ml culture medium, and filter the cell suspension through a 70 μm filter. Spin down at 300xg for 5 min and remove supernatant. Resuspend the cells in culture medium to a final concentration of 1×106 cells/ml.

  5. Observe the status of cell adhesion, and adjust incubation time accordingly to obtain optimal cell density (about 50% of channel area covered, with cells uniformly distributed throughout the collagen channel).

  6. If primary antibodies are to be used, they should be diluted together with DAPI in blocking solution II, incubated overnight at 4°C on the rocker. Rinse 3 times with PBS for 5 min each on the rocker, followed by an overnight rinse. Secondary antibodies were used at 1:400 in blocking solution II, and incubated overnight at 4°C on the rocker. Rinse 3 times with PBS for 5 min each on the rocker, followed by an overnight rinse on the rocker.

Acknowledgements

We are grateful to the UPenn Cell and Developmental Biology Microscopy Core and the UPenn NIDDK Center for Molecular Studies in Digestive and Liver Disease (NIH-P30-DK050306) for assistance with imaging. This work was supported by the Center for Engineering MechanoBiology (CEMB), an NSF Science and Technology Center, under grant agreement CMMI: 15-48571 and by NIDDK R01 DK119290 (to R.G.W.). It was carried out, in part, at the Singh Center for Nanotechnology, part of the National Nanotechnology Coordinated Infrastructure Program, which is supported by the National Science Foundation grant NNCI-1542153. The small cholangiocyte cell line was generously provided by Gianfranco Alpini (Texas A&M Health Science Center College of Medicine and Baylor Scott & White Digestive Disease Research Center).

References

  • 1.Matsumoto K, Imasato M, Yamazaki Y, Tanaka H, Watanabe M, Eguchi H, Nagano H, Hikita H, Tatsumi T, Takehara T, Tamura A, Tsukita S (2014) Claudin 2 deficiency reduces bile flow and increases susceptibility to cholesterol gallstone disease in mice. Gastroenterology 147 (5):1134–1145.e1110. doi: 10.1053/j.gastro.2014.07.033 [DOI] [PubMed] [Google Scholar]
  • 2.Herr KJ, Tsang YH, Ong JW, Li Q, Yap LL, Yu W, Yin H, Bogorad RL, Dahlman JE, Chan YG, Bay BH, Singaraja R, Anderson DG, Koteliansky V, Viasnoff V, Thiery JP (2014) Loss of alpha-catenin elicits a cholestatic response and impairs liver regeneration. Scientific reports 4:6835. doi: 10.1038/srep06835 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Nakagawa H, Hikiba Y, Hirata Y, Font-Burgada J, Sakamoto K, Hayakawa Y, Taniguchi K, Umemura A, Kinoshita H, Sakitani K, Nishikawa Y, Hirano K, Ikenoue T, Ijichi H, Dhar D, Shibata W, Akanuma M, Koike K, Karin M, Maeda S (2014) Loss of liver E-cadherin induces sclerosing cholangitis and promotes carcinogenesis. Proceedings of the National Academy of Sciences of the United States of America 111 (3):1090–1095. doi: 10.1073/pnas.1322731111 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Huh D, Matthews BD, Mammoto A, Montoya-Zavala M, Hsin HY, Ingber DE (2010) Reconstituting organ-level lung functions on a chip. Science (New York, NY) 328 (5986):1662–1668. doi: 10.1126/science.1188302 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Griep LM, Wolbers F, de Wagenaar B, ter Braak PM, Weksler BB, Romero IA, Couraud PO, Vermes I, van der Meer AD, van den Berg A (2013) BBB on chip: microfluidic platform to mechanically and biochemically modulate blood-brain barrier function. Biomedical microdevices 15 (1):145–150. doi: 10.1007/s10544-012-9699-7 [DOI] [PubMed] [Google Scholar]
  • 6.Brown JA, Codreanu SG, Shi M, Sherrod SD, Markov DA, Neely MD, Britt CM, Hoilett OS, Reiserer RS, Samson PC, McCawley LJ, Webb DJ, Bowman AB, McLean JA, Wikswo JP (2016) Metabolic consequences of inflammatory disruption of the blood-brain barrier in an organ-on-chip model of the human neurovascular unit. Journal of neuroinflammation 13 (1):306. doi: 10.1186/s12974-016-0760-y [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Booth R, Kim H (2012) Characterization of a microfluidic in vitro model of the blood-brain barrier (muBBB). Lab on a chip 12 (10):1784–1792. doi: 10.1039/c2lc40094d [DOI] [PubMed] [Google Scholar]
  • 8.Du Y, Li N, Yang H, Luo C, Gong Y, Tong C, Gao Y, Lü S, Long M (2017) Mimicking liver sinusoidal structures and functions using a 3D-configured microfluidic chip. Lab on a chip 17 (5):782–794. doi: 10.1039/C6LC01374K [DOI] [PubMed] [Google Scholar]
  • 9.Polacheck WJ, Kutys ML, Tefft JB, Chen CS (2019) Microfabricated blood vessels for modeling the vascular transport barrier. Nature protocols 14 (5):1425–1454. doi: 10.1038/s41596-019-0144-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Polacheck WJ, Kutys ML, Yang J, Eyckmans J, Wu Y, Vasavada H, Hirschi KK, Chen CS (2017) A non-canonical Notch complex regulates adherens junctions and vascular barrier function. Nature 552 (7684):258–262. doi: 10.1038/nature24998 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Du Y, Khandekar G, Llewellyn J, Polacheck W, Chen CS, Wells RG (2020) A Bile Duct-on-a-Chip With Organ-Level Functions. Hepatology (Baltimore, Md) 71 (4):1350–1363. doi: 10.1002/hep.30918 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Ueno Y, Alpini G, Yahagi K, Kanno N, Moritoki Y, Fukushima K, Glaser S, LeSage G, Shimosegawa T (2003) Evaluation of differential gene expression by microarray analysis in small and large cholangiocytes isolated from normal mice. Liver International 23 (6):449–459. doi: 10.1111/j.1478-3231.2003.00876.x [DOI] [PubMed] [Google Scholar]
  • 13.Karjoo S, Wells RG (2014) Isolation of neonatal extrahepatic cholangiocytes. Journal of visualized experiments : JoVE (88):51621. doi: 10.3791/51621 [DOI] [PMC free article] [PubMed]

RESOURCES