1 
Title: Cholangiocyte organoids can repair bile ducts after transplantation in human liver. 1 
Authors: Fotios Sampaziotis1,2,3,*, Daniele Muraro1, Olivia C. Tysoe1,4, Stephen Sawiak5, 2 
Timothy E. Beach4, Edmund M. Godfrey6, Sara S. Upponi6, Teresa Brevini1, Brandon T. 3 
Wesley1, Jose Garcia-Bernardo7, Krishnaa Mahbubani4, Giovanni Canu1, Richard Gieseck III8, 4 
Natalie L. Berntsen9,10,11, Victoria L. Mulcahy2,12, Keziah Crick13, Corrina Fear13, Sharayne 5 
Robinson13, Lisa Swift13, Laure Gambardella1,2, Johannes Bargehr1,2,14, Daniel Ortmann1, 6 
Stephanie E. Brown1, Anna Osnato1, Michael P. Murphy15, Gareth Corbett16, William T. H. 7 
Gelson2,3, George F. Mells2,3,12, Peter Humphreys1, Susan E. Davies17, Irum Amin4,13, Paul 8 
Gibbs4,13, Sanjay Sinha1,2, Sarah A. Teichmann7,18, Andrew J Butler4,13, Teik Choon See6, 9 
Espen Melum9,10,11,19,20, Christopher J. E. Watson4,13,21,22, Kourosh Saeb-Parsy4,13, †, Ludovic 10 
Vallier1,4†* 11 
 12 
Affiliations:  13 
1. Wellcome and MRC Cambridge Stem Cell Institute. 14 
2. Department of Medicine, University of Cambridge. 15 
3. Cambridge Liver Unit, Cambridge University Hospitals NHS Foundation Trust. 16 
4. Department of Surgery, University of Cambridge and NIHR Cambridge Biomedical 17 
Research centre, Cambridge, UK 18 
5.  University of Cambridge, Department of Clinical Neurosciences, University of Cambridge, 19 
Cambridge, UK. 20 
6. Department of Radiology, Cambridge University Hospitals NHS Foundation Trust. 21 
7. Wellcome Sanger Institute, Wellcome Genome Campus, Hinxton, Cambridge CB10 1SA, 22 
UK. 23 
8. Immunopathogenesis Section, Laboratory of Parasitic Diseases, National Institute of Allergy 24 
and Infectious Diseases, NIH, Bethesda, MD 20852, USA 25 
 2 
9. Norwegian PSC Research Center, Department of Transplantation Medicine, Division of 1 
Surgery, Inflammatory Diseases and Transplantation, Oslo University Hospital Rikshospitalet, 2 
Oslo, Norway. 3 
10. Research Institute of Internal Medicine, Division of Surgery, Inflammatory Diseases and 4 
Transplantation, Oslo University Hospital, Oslo, Norway. 5 
11. Institute of Clinical Medicine, Faculty of Medicine, University of Oslo, Oslo, Norway. 6 
12. Academic Department of Medical Genetics, University of Cambridge, Cambridge, UK. 7 
13. Department of Surgery, Cambridge University Hospitals NHS Foundation Trust, 8 
Cambridge, UK. 9 
14. Division of Cardiovascular Medicine, University of Cambridge, ACCI Level 6, Box 110, 10 
Addenbrooke's Hospital, Hills Road, Cambridge CB2 0QQ, UK. 11 
15. MRC Mitochondrial Biology Unit, University of Cambridge, Cambridge, UK. 12 
16. Cambridge University Hospitals NHS Foundation Trust 13 
17. Department of Histopathology, Cambridge University Hospitals NHS Foundation Trust, 14 
Cambridge, UK. 15 
18. Cavendish Laboratory, JJ Thomson Ave, Cambridge CB3 0HE, UK. 16 
19. Section for Gastroenterology, Department of Transplantation Medicine, Division of 17 
Surgery, Inflammatory Diseases and Transplantation, Oslo University Hospital 18 
Rikshospitalet, Oslo, Norway. 19 
20. Hybrid Technology Hub-Centre of Excellence, Institute of Basic Medical Sciences, 20 
Faculty of Medicine, University of Oslo, Oslo, Norway. 21 
21. National Institute of Health Research (NIHR) Cambridge Biomedical Research Centre, 22 
and the NIHR Blood and Transplant Research Unit (BTRU) at the 23 
22. University of Cambridge in collaboration with Newcastle University and in partnership 24 
with NHS Blood and Transplant (NHSBT), Cambridge, UK 25 
 26 
† These authors share senior authorship. 27 
 3 
*Correspondence to: Fotios Sampaziotis, fs347@cam.ac.uk; Ludovic Vallier, 1 
lv225@cam.ac.uk.  2 
 3 
 4 
One sentence summary: 5 
Single-cell RNA sequencing analyses combined with a novel model for cell transplantation in 6 
human livers reveal that intra- and extra-hepatic cholangiocytes are interchangeable for 7 
regenerative medicine applications. 8 
  9 
 4 
Abstract: 1 
Organoid technology holds great promise for regenerative medicine but has not yet been 2 
applied to humans. Here, we address this challenge in the context of cholangiocyte organoids 3 
and cholangiopathies, which represent a leading indication for liver transplantation. Using 4 
single-cell RNA sequencing we show that primary human cholangiocytes display 5 
transcriptional diversity which is lost in organoid culture. However, cholangiocyte organoids 6 
remain plastic and resume their in vivo signatures when transplanted back in the biliary tree. 7 
We then utilize a new model of cell engraftment in human livers undergoing ex vivo 8 
normothermic perfusion to demonstrate that this property allows extrahepatic organoids to 9 
repair human intrahepatic ducts after transplantation. Our results provide proof-of-principle 10 
that cholangiocyte organoids can be used to repair human biliary epithelium. 11 
  12 
 5 
Main text: 1 
Organoids have a unique potential for tissue repair as they retain key functions and 2 
characteristics of their tissue of origin. Nevertheless, their ability to repair native epithelia and 3 
restore their complexity has not been established in humans, while organoid engraftment and 4 
survival in vivo has only been demonstrated in a limited number of animal studies (1). The bile 5 
duct epithelium presents an archetypal and clinically important system for addressing this 6 
challenge and for developing proof-of-concept studies in human. Indeed, disorders of the 7 
biliary system, which transfers bile from the liver to the duodenum, account for 70% of 8 
paediatric and up to a third of adult liver transplantation (2). This results in a pressing need for 9 
therapeutic alternatives, such as cell-based therapy. Furthermore, organoids suitable for 10 
regenerative medicine applications can be easily derived from biliary epithelial cells, known 11 
as cholangiocytes (3). Finally, the bile ducts also recapitulate the epithelial diversity found in 12 
other hollow-lumen organs (4). Indeed, different regions along the biliary tree display distinct 13 
transcriptional profiles and functional properties, such as the chemical modification of bile (5, 14 
6), as well as variation in disease susceptibility between the intrahepatic ducts, extrahepatic 15 
ducts and the gallbladder. Nevertheless, the impact of this regional variation on the 16 
characteristics and regenerative potential of the organoids derived from different regions of 17 
the biliary tree remains to be characterized. To address these questions and demonstrate the 18 
value of organoids for regenerative medicine in humans, we first characterize cholangiocyte 19 
diversity in vivo using single-cell transcriptomics and confirm that different regions of the 20 
human biliary tree contain cells with distinct transcriptional profiles. We then show that 21 
cholangiocytes lose these differences in organoid culture and become interchangeable, but 22 
their regional identity can be restored in vitro by environmental stimuli. We subsequently use 23 
a biliary injury mouse model and a novel model for cell transplantation in human organs 24 
undergoing ex vivo normothermic perfusion to prove that this plasticity allows cholangiocytes 25 
from one region to repair a different region of the biliary tree paving the way for cell-based 26 
therapy using organoids. 27 
 6 
To characterize the cellular composition of the human biliary epithelium, cholangiocytes from 1 
different regions (Intrahepatic Bile Ducts (IHD): 5 patients, 7295 cells; Common Bile Duct 2 
(CBD): 3 patients, 3006 cells; Gallbladder (GB): 3 patients, 3702 cells) were isolated using 3 
magnetic bead sorting and their transcriptome was determined using droplet encapsulation 4 
single-cell RNA sequencing (scRNAseq) (Fig. 1A-B, Fig. S1A-C). The isolated cells 5 
expressed key cholangiocyte markers, including KRT7, KRT19, SOX9, and GGT (Fig. S2A). 6 
The transcriptomes of all three biliary cell populations shared a core transcriptional profile, 7 
illustrated by their proximity in UMAP space and high connectivity in Partition-based Graph 8 
Abstraction (PAGA) analysis when compared to different liver cell types, such as stellate cells 9 
and liver sinusoidal endothelial cells (LSECs, Fig. S2B-S2E). However, more detailed analysis 10 
after sub-clustering of cholangiocytes revealed non-overlapping expression modules of the 11 
three populations (Fig. 1B). This suggests that, despite their similarities, cholangiocytes from 12 
different regions exhibit unique gene expression signatures (6). Accordingly, Differentially 13 
Expressed Genes (DEG) analysis (Data S1) identified known region-specific markers, 14 
including aquaporins (7), mucins (8), FGF19 (9), SOX17 (10) in the extrahepatic biliary tree, 15 
JAG1 (11), TACSTD2 (12) and YAP target genes in intrahepatic cholangiocytes (13, 14), as 16 
well as novel markers including DCDC2, TFF1-3, SLC15A1 (Fig. 1C-1D, Fig. S3A-S3D). Most 17 
of these genes correspond to functional markers such as bile acid receptors or channels 18 
modifying bile composition (Fig. S3C). Thus, the transcriptional divergence among 19 
cholangiocytes from different regions could reflect adaptation to their microenvironment, such 20 
as variation in bile properties along the biliary tree (15). Accordingly, cholangiocytes from 21 
anatomically adjacent and hence environmentally similar regions (e.g. intrahepatic and 22 
common bile duct vs. gallbladder) displayed higher transcriptional similarity. This was 23 
illustrated by PAGA analysis (Fig. S3E-S3F), in agreement with results from diffusion 24 
pseudotime (DPT) and single-cell consensus clustering (SC3) analyses (Fig. S4-S5). These 25 
results point towards a progressive change in the expression of region-specific markers (Fig. 26 
1E, Data S2), and a gradual transition in the transcriptional signature of cholangiocytes from 27 
adjacent regions (Fig. S4A-S4C) rather than distinct subpopulations (Fig. 1E, S4-S5). This 28 
 7 
gradient in gene expression is likely to support adjustment of the cells to environmental 1 
conditions, such as the gradual change in bile composition from the intrahepatic ducts to the 2 
gallbladder. In conclusion, our results show that the human biliary epithelium is comprised of 3 
cholangiocytes displaying a gradual shift in their transcriptional profile along the biliary tree, 4 
which is likely imposed by region-specific microenvironments. 5 
We subsequently used this single-cell map of the human biliary tree as a framework to 6 
characterise cholangiocyte organoids. To this end, a fraction of the primary cholangiocytes 7 
isolated for scRNAseq from each region (IHD, CDB, GB) were propagated as organoids using 8 
our established conditions (3, 16). The resulting organoids expressed cholangiocyte markers 9 
(KRT7, KRT19, SOX9, HNF1B, CFTR; Fig. S6A-S6B); displayed comparable functionality 10 
(ALP, GGT activity; Fig. S6C-S6D) and similar expansion potential regardless of their region 11 
of origin (Fig. S6E). To further explore these similarities, we performed scRNAseq on these 12 
organoids (2 lines per region; GB:  5859 cells; CBD 5321 cells; IHD 6641 cells; Fig. S1A-C). 13 
UMAP and PCA analyses demonstrated that organoids exhibited overlapping transcriptomic 14 
profiles (Fig. 2A, Fig. S7A-S7D) indicating that cholangiocytes grown in vitro assume a similar 15 
transcriptional signature independent of their region of origin. Of note, regressing cell cycle-16 
related genes did not change these observations excluding that a common “proliferation” 17 
signature could mask differences between organoids of different spatial origins (Fig. S7A-18 
S7C, S7E). Furthermore, we did not detect any cells co-expressing known somatic stem cell 19 
markers (LGR5, PROM1, TACSTD2, NCAM), excluding the possibility that organoid 20 
similarities reflect a common progenitor/stem cell identity (Fig. S7F). 21 
We then compared organoids from different regions with primary cholangiocytes to explore if 22 
these similarities corresponded to loss of their original regional identity in vitro (Fig. 2A). 23 
Organoids and primary cells following cell cycle regression shared a core transcriptional profile 24 
reflecting their common cholangiocyte nature, which was illustrated by their proximity in UMAP 25 
space and high PAGA connectivity when compared to different liver cell types, such as stellate 26 
cells and LSECs (Fig. S7C-D). However, DEG analyses highlighted downregulation of region-27 
 8 
specific markers, such as SLC13A1 and SLC26A3 (Fig. 2B, Fig. S7G); while Gene Ontology 1 
(GO) and Gene Set Enrichment Analyses (GSEA) identified these DEGs as factors facilitating 2 
the adaptation of cholangiocytes to their respective microenvironments, e.g. bile acid vs. 3 
culture medium processing genes (Fig. S8A-S8C). Furthermore, we confirmed upregulation 4 
of YAP target genes (Data S3) in organoids,  in accordance with previous reports (14). 5 
Consequently, primary cholangiocytes propagated as organoids adapt to their new 6 
microenvironment by maintaining their core transcriptional signature, while losing the 7 
expression of markers specific to their region of origin. 8 
To explore the mechanisms controlling cholangiocyte identity, we decided to add bile in our 9 
culture conditions as the principal determinant of the cholangiocyte microenvironment. 10 
Different organoids (IHD, CBD, GB) were treated with human gallbladder bile for 72 hours and 11 
then characterized using scRNAseq (Fig. 2A, S1A-S1C) (GB: 3815 cells; CBD 3224 cells; 12 
IHD 3653 cells). UMAP and PCA revealed that treated organoids assumed a new overlapping 13 
gene expression profile (Fig. 2A, S9A) confirming a shared capacity to adapt to exposure to 14 
bile. Importantly, PAGA and DEG analyses showed that this transcriptional profile was shifted 15 
towards a gallbladder identity (Fig. 2B, Fig. S9B-S9C). To characterise the factors controlling 16 
this transition, we interrogated differentially expressed genes in bile-treated organoids. GO, 17 
GSEA and UMAP analyses (Fig. S9D-S9F) confirmed the induction of region-specific markers 18 
(SOX17, MUC13, FGF19; Fig. 2B, S9F) and revealed upregulation of bile acid receptor 19 
pathways and downstream targets (NR1H4/FXR, NR1I2, NR0B2, SLC51A, FGF19, ABCA1, 20 
PPARG; Fig. 2B, S9D-S9F). Of note, these results were validated through activation and 21 
inhibition of the Farnesoid X receptor (FXR), using chenodeoxycholic acid and z-22 
guggulsterone respectively (Fig. 2C-2D), thereby confirming that regardless of their origin, 23 
cholangiocytes grown in vitro can respond and adapt to environmental stimuli. Together, these 24 
results suggest that cholangiocyte organoids could assume different regional identities when 25 
instructed by the appropriate niche factors.  26 
 9 
To validate cholangiocyte plasticity and explore its functional implications, we decided to 1 
assess if organoids from one region of the biliary tree could repair a different region following 2 
transplantation. For this, we induced cholangiopathy in immunodeficient mice using 4,4’-3 
methylenedianiline (MDA) (17) (Fig. 3A-3B, S10, S11) and attempted to rescue the phenotype 4 
with intraductal delivery (18) of human gallbladder organoids expressing Red Fluorescent 5 
Protein-expressing (RFP). Control animals receiving carrier medium without cells lost weight 6 
(Fig. S10A) and died within 3 weeks (Fig. 3B, Table S1), developing cholestasis (Fig. S10B) 7 
and cholangiopathy demonstrated by IF (Fig. S10C), histology (Fig. S10D) and Magnetic 8 
Resonance Cholangio-Pancreatography (MRCP) (Fig. 3C, S11A-S11C, Movie S1-S6). On 9 
the contrary, animals receiving organoids were electively culled at the end of the experiment 10 
and survived for up to 3 months with resolution of cholangiopathy and normal serum 11 
biochemistry (Fig. 3B-3C, S10A-S10B, S11B-S11C, Movie S3-S4, S6). The transplanted 12 
gallbladder cholangiocytes engrafted in various size intrahepatic ducts (Fig. 3D, S12A-C, 13 
Movie S7-S9) corresponding to ~25-55% of the regenerated biliary epithelium (Fig. S12C), 14 
and assumed an intrahepatic identity by losing gallbladder (SOX17) and expressing 15 
intrahepatic markers (SOX4, DCDC2, BICC1) (Fig. 3D, Fig. S12A-S12B). Core biliary 16 
markers (KRT7, KRT19, CFTR) were also expressed (Fig. S12A), while we observed YAP 17 
activation both in engrafted and native cells (Fig. S12B, S12E) in accordance with previous 18 
reports (13). Of note, we never observed expression of other hepatic lineage markers such as 19 
albumin indicating that cholangiocyte organoid plasticity is likely to be limited to their biliary 20 
lineage (Fig. S12A). Furthermore, the engrafted cells expressed proliferation markers (Fig. 21 
S12B, S12D) at similar levels to native mouse cholangiocytes; while abnormal growth or 22 
tumour formation was never noticed in all the analyses performed (Fig. 3C, 3D, S10D, S12A-23 
S12B), including T1 weighed body MR imaging at the end of the experiment (Movie S1, S3). 24 
Thus, organoid transplantation provides the healthy cells required to repair the damaged 25 
epithelium and rescue acute injury. 26 
 10 
To ensure that animal rescue and resolution of cholangiopathy was specific to cholangiocyte 1 
organoids, we repeated the experiment using Mesenchymal Stem Cells (MSCs), as a different 2 
cell type known to provide anti-inflammatory effects following transplantation through 3 
paracrine signalling (Fig. S13A-S13C). This experiment allowed us to explore if organoids are 4 
essential for duct regeneration and animal rescue; and if some of the observed effects could 5 
be attributed to paracrine signals which are not unique to our cells. In sum, MSCs failed to 6 
engraft (Fig. S13C) and rescue the transplanted animals, which exhibited no difference in 7 
survival compared to controls (P>0.05; Fig. S13A) and no resolution of cholestasis on serum 8 
biochemistry (Fig. S13B). Consequently, cholangiopathy resolution is specific to the 9 
engraftment of cholangiocyte organoids; and although additional therapeutic effects of our 10 
cells through growth factor and cytokine secretion cannot be completely excluded, these 11 
effects are unique to cholangiocyte organoids. 12 
We then explored if organoid culture is required to ‘unlock’ the cells’ plasticity or if this reflects 13 
an inherent property of primary cholangiocytes. To achieve this, we transplanted primary 14 
gallbladder cholangiocytes (Fig. S13A-S13C) directly post isolation without in vitro culture. 15 
Under these conditions very few primary cholangiocytes engrafted in the mouse bile ducts 16 
(Fig. S13C) most likely due to the cumulative stress of isolation and transplantation; and failed 17 
to rescue the animals or resolve cholestasis (Fig. S13A, S13B). Nonetheless, the engrafted 18 
cells expressed intrahepatic markers and lost expression of gallbladder markers (Fig. S13C). 19 
In conclusion, cholangiocyte plasticity is not limited to organoids grown in vitro; however, 20 
organoid culture is necessary for the cholangiocytes to recover from the stress of isolation and 21 
for large scale expansion providing the cell numbers required for engraftment and biliary 22 
repair. 23 
Finally, to ensure that our results are not specific to the intrahepatic compartment or 24 
gallbladder organoids, we used our established methodology (3) to transplant common bile 25 
duct-derived cholangiocyte organoids in the gallbladder of immunocompromised mice. The 26 
engrafted cells exhibited loss of common bile duct makers and upregulation of gallbladder 27 
 11 
markers (Fig. S14), confirming that our previous findings apply to different compartments of 1 
the biliary tree and to organoids of different origin. Taken together, these results establish that 2 
cholangiocytes from different regions of the biliary tree are interchangeable and suggest that 3 
extrahepatic cells can be used to repair acute intrahepatic duct injury. 4 
Cell transplantation experiments in mouse models are extremely useful but are not always 5 
predictive of therapeutic outcome (19). Furthermore, the mouse liver microenvironment is 6 
different to human, raising the possibility that our results may not translate between species. 7 
To address these challenges, we developed a new model for cell-based therapy in human 8 
utilizing ex vivo organ perfusion (20). Ex-vivo Normothermic Perfusion (NMP) was developed 9 
to improve organ preservation and reduce ischaemia-reperfusion injury by circulating warm 10 
oxygenated blood through liver grafts prior to transplantation. Importantly, the biliary tree is 11 
particularly susceptible to ischaemia which results in duct damage (21, 22). Low bile pH (< 12 
7.5) during NMP is used as a predictor of this type of cholangiopathy (23).  13 
To assess the therapeutic potential of our cells for repairing human bile ducts, RFP gallbladder 14 
organoids were injected in the intrahepatic ducts of deceased transplant donor livers (n=3) 15 
with a bile pH<7.5 at the start of the experiment, signifying ischaemic duct injury. The organs 16 
were perfused with oxygenated blood and nutrients at normal body temperature (20); Fig. 4A-17 
4B, S15A) for up to 100 hours in order to maintain a near-physiological microenvironment. 18 
Importantly, the organoids were delivered in a terminal branch of the intrahepatic ducts under 19 
fluoroscopic guidance to minimize the area of distribution of the cells and maximize cell density 20 
(Fig. S15B). At the end of the experiment, ultrasound imaging revealed no evidence of duct 21 
dilatation or obstruction (Fig. S15C), while RFP-expressing cells were not detected in the 22 
perfusate by flow cytometry, confirming that the injected cells remained in the biliary 23 
compartment (Fig. 4C). More importantly, the transplanted organoids engrafted in the 24 
intrahepatic biliary tree (Fig. 4D, S16A), with RFP cells regenerating ~40-85% of the injected 25 
ducts (Fig. 16B); and expressing key biliary markers (KRT7, KRT19, CFTR, GGT). 26 
Furthermore, engrafted gallbladder organoids exhibited loss of gallbladder (SOX17) and 27 
 12 
upregulation of intrahepatic (SOX4, BICC1, DCDC2) markers without differentiation to other 1 
hepatic lineages (Fig. 4D, S15D, S16A-S16B). Thus, at the end of the experiment, the injected 2 
ducts consisted of a mixture of native and transplanted cholangiocytes (Fig. S16A-S16B), with 3 
multiple transition points between donor and recipient cells and no evidence of cholangiopathy 4 
(Fig 4D, S15D, S16A). 5 
Conversely, control ducts not receiving cells demonstrated evidence of ischaemic injury with 6 
loss of epithelial continuity and sloughing of cells in the duct lumen (Fig. 4D).  We 7 
subsequently characterised the impact of engraftment on organ function. Physiologically, 8 
cholangiocytes modify the composition and pH of bile through water transfer and bicarbonate 9 
secretion (6). Therefore, we compared the bile from organoid-injected vs. carrier-injected 10 
ducts. Accordingly, bile aspirated from ducts injected with cells exhibited higher pH and 11 
volume (Fig. 4E) confirming that transplanted cholangiocytes retain their function to modify 12 
bile composition. Together, these results provide the first proof-of-principle that perfused 13 
organs can be used to ascertain functional engraftment of human cells and validate our mouse 14 
data by showing that cholangiocytes are interchangeable for transplantation in human organs.  15 
Our results show that the biliary epithelium is composed of cholangiocytes with diverse 16 
transcriptional profiles which are determined by their local environment. This diversity is lost 17 
in organoid culture due to the lack of niche stimuli. However, organoids can adapt 18 
appropriately to local environmental cues both in vitro and following transplantation, restore 19 
the expression of region-specific markers and assume different regional identities. Thus, 20 
organoids from a single region could potentially repair the entirety of the biliary tree. This 21 
plasticity could have significant implications for regenerative medicine. Indeed, although 22 
autologous cell-based therapy potentially avoids the need for immunosuppression its 23 
application for primary organoids is limited by the impact of disease on the epithelium. 24 
However, cholangiopathies belong to a family of localising diseases, affecting predominantly 25 
specific regions of an organ (24). Consequently, our results provide proof-of-concept that 26 
cholangiocytes from spared regions, such as the gallbladder, could be used for autologous 27 
 13 
cell-based therapy to repair human intrahepatic bile ducts, which constitute the most common 1 
site of injury in cholangiopathies. Moreover, our novel model for cell engraftment in human 2 
perfused organs paves the road for the use of ex vivo cell-based therapy to improve graft 3 
function prior to transplantation, which could ultimately increase the number of useable organs 4 
and reduce pressure on the transplant waiting list. In this context, quality controlled and readily 5 
available allogeneic cholangiocyte organoids from a cell bank could be used routinely in the 6 
future to prevent ischaemic cholangiopathy in organs at risk of biliary injury (e.g. low bile pH), 7 
since the organ recipients will receive immunosuppression as part of their standard care.  8 
Importantly, our results provide proof-of-principle for the transplantation of organoids in human 9 
organs which could expedite regulatory approval and fast-tack first-in-man trials. Ultimately, 10 
the same approach could also be applied to a variety of ex vivo perfused organs and cell types 11 
to validate functional cell engraftment, demonstrate safety, improve cell transplantation 12 
technique and efficacy and accelerate clinical translation of new cell-based therapies.  13 
  14 
 14 
References and notes 1 
1.  J. Drost, H. Clevers, Translational applications of adult stem cell-derived organoids. 2 
Dev. 144, 968–975 (2017). 3 
2.  R. H. Squires, V. Ng, R. Romero, U. Ekong, W. Hardikar, S. Emre, G. V Mazariegos, 4 
Evaluation of the pediatric patient for liver transplantation: 2014 practice guideline by 5 
the American Association for the Study of Liver Diseases, American Society of 6 
Transplantation and the North American Society for Pediatric Gastroenterology, 7 
Hepatology and Nutrition. Hepatology. 60, 362–98 (2014). 8 
3.  F. Sampaziotis, A. W. A. W. Justin, O. C. O. C. Tysoe, S. Sawiak, E. M. Godfrey, S. 9 
S. S. S. Upponi, R. L. R. L. Gieseck, M. C. M. C. De Brito, N. L. N. L. Berntsen, M. J. 10 
M. J. Gómez-Vázquez, D. Ortmann, L. Yiangou, A. Ross, J. Bargehr, A. Bertero, M. 11 
C. F. M. C. F. Zonneveld, M. T. M. T. Pedersen, M. Pawlowski, L. Valestrand, P. 12 
Madrigal, N. Georgakopoulos, N. Pirmadjid, G. M. G. M. Skeldon, J. Casey, W. Shu, 13 
P. M. P. M. Materek, K. E. K. E. Snijders, S. E. Brown, C. A. C. A. Rimland, I. 14 
Simonic, S. E. S. E. Davies, K. B. K. B. Jensen, M. Zilbauer, W. T. H. W. T. H. 15 
Gelson, G. J. G. J. Alexander, S. Sinha, N. R. F. N. R. F. Hannan, T. A. T. A. Wynn, 16 
T. H. T. H. Karlsen, E. Melum, A. E. A. E. Markaki, K. Saeb-Parsy, L. Vallier, 17 
Reconstruction of the mouse extrahepatic biliary tree using primary human 18 
extrahepatic cholangiocyte organoids. Nat. Med. 23, 954–963 (2017). 19 
4.  K. Parikh, A. Antanaviciute, D. Fawkner-Corbett, M. Jagielowicz, A. Aulicino, C. 20 
Lagerholm, S. Davis, J. Kinchen, H. H. Chen, N. K. Alham, N. Ashley, E. Johnson, P. 21 
Hublitz, L. Bao, J. Lukomska, R. S. Andev, E. Björklund, B. M. Kessler, R. Fischer, R. 22 
Goldin, H. Koohy, A. Simmons, Colonic epithelial cell diversity in health and 23 
inflammatory bowel disease. Nature. 567, 49–55 (2019). 24 
5.  C. A. Rimland, S. G. Tilson, C. M. Morell, R. A. Tomaz, W.-Y. Lu, S. E. Adams, N. 25 
Georgakopoulos, F. Otaizo-Carrasquero, T. G. Myers, J. R. Ferdinand, R. L. Gieseck, 26 
 15 
F. Sampaziotis, O. C. Tysoe, B. Wesley, D. Muraro, G. C. Oniscu, N. R. Hannan, S. J. 1 
Forbes, K. Saeb-Parsy, T. A. Wynn, L. Vallier, Regional differences in human biliary 2 
tissues and corresponding in vitro derived organoids. Hepatology (2020), 3 
doi:10.1002/hep.31252. 4 
6.  J. H. Tabibian, A. I. Masyuk, T. V. Masyuk, S. P. O’Hara, N. F. LaRusso, Physiology 5 
of cholangiocytes. Compr. Physiol. 3, 541–565 (2013). 6 
7.  A. I. Masyuk, N. F. LaRusso, Aquaporins in the hepatobiliary system. Hepatology. 43, 7 
S75–S81 (2006). 8 
8.  K.-S. Yoo, H. S. Choi, D. W. Jun, H. L. Lee, O. Y. Lee, B. C. Yoon, K. G. Lee, S. S. 9 
Paik, Y. S. Kim, J. Lee, MUC Expression in Gallbladder Epithelial Tissues in 10 
Cholesterol-Associated Gallbladder Disease. Gut Liver. 10, 851–8 (2016). 11 
9.  S. J. L. B. Zweers, K. A. C. Booij, M. Komuta, T. Roskams, D. J. Gouma, P. L. M. 12 
Jansen, F. G. Schaap, The human gallbladder secretes fibroblast growth factor 19 13 
into bile: Towards defining the role of fibroblast growth factor 19 in the enterobiliary 14 
tract. Hepatology. 55, 575–583 (2012). 15 
10.  Y. Zong, B. Z. Stanger, Molecular mechanisms of bile duct development. Int. J. 16 
Biochem. Cell Biol. 43, 257–64 (2011). 17 
11.  K. Si-Tayeb, F. P. Lemaigre, S. A. Duncan, Organogenesis and Development of the 18 
Liver. Dev. Cell. 18, 175–189 (2010). 19 
12.  N. Aizarani, A. Saviano, Sagar, L. Mailly, S. Durand, J. S. Herman, P. Pessaux, T. F. 20 
Baumert, D. Grün, A human liver cell atlas reveals heterogeneity and epithelial 21 
progenitors. Nature. 572, 199–204 (2019). 22 
13.  B. J. Pepe-Mooney, M. T. Dill, A. Alemany, J. Ordovas-Montanes, Y. Matsushita, A. 23 
Rao, A. Sen, M. Miyazaki, S. Anakk, P. A. Dawson, N. Ono, A. K. Shalek, A. van 24 
Oudenaarden, F. D. Camargo, Single-Cell Analysis of the Liver Epithelium Reveals 25 
 16 
Dynamic Heterogeneity and an Essential Role for YAP in Homeostasis and 1 
Regeneration. Cell Stem Cell. 25, 23-38.e8 (2019). 2 
14.  L. Planas-Paz, T. Sun, M. Pikiolek, N. R. Cochran, S. Bergling, V. Orsini, Z. Yang, F. 3 
Sigoillot, J. Jetzer, M. Syed, M. Neri, S. Schuierer, L. Morelli, P. S. Hoppe, W. 4 
Schwarzer, C. M. Cobos, J. L. Alford, L. Zhang, R. Cuttat, A. Waldt, N. Carballido-5 
Perrig, F. Nigsch, B. Kinzel, T. B. Nicholson, Y. Yang, X. Mao, L. M. Terracciano, C. 6 
Russ, J. S. Reece-Hoyes, C. Gubser Keller, A. W. Sailer, T. Bouwmeester, L. E. 7 
Greenbaum, J. J. Lugus, F. Cong, G. McAllister, G. R. Hoffman, G. Roma, J. S. 8 
Tchorz, YAP, but Not RSPO-LGR4/5, Signaling in Biliary Epithelial Cells Promotes a 9 
Ductular Reaction in Response to Liver Injury. Cell Stem Cell. 25, 39-53.e10 (2019). 10 
15.  A. Lanzini, BILE. Encycl. Food Sci. Nutr., 471–478 (2003). 11 
16.  O. C. Tysoe, A. W. Justin, T. Brevini, S. E. Chen, K. T. Mahbubani, A. K. Frank, H. 12 
Zedira, E. Melum, K. Saeb-Parsy, A. E. Markaki, L. Vallier, F. Sampaziotis, Isolation 13 
and propagation of primary human cholangiocyte organoids for the generation of 14 
bioengineered biliary tissue. Nat. Protoc. 14, 1884–1925 (2019). 15 
17.  M.-O. Lee, J. H. Kim, B.-I. Yoon, S.-B. Kwon, K.-S. Kang, M. Kim, G. Kong, J.-W. 16 
Hwang, B.-H. Lee, H.-L. Kim, J.-Y. Yi, J.-S. Park, H. Chung, Time- and Dose-based 17 
Gene Expression Profiles Produced by a Bile-duct–damaging Chemical, 4,4′-18 
methylene Dianiline, in Mouse Liver in an Acute Phase. Toxicol. Pathol. 36, 660–673 19 
(2008). 20 
18.  N. L. Berntsen, B. Fosby, L. Valestrand, C. Tan, H. M. Reims, E. Schrumpf, T. H. 21 
Karlsen, P. D. Line, E. Melum, Establishment of a surgical bile duct injection 22 
technique giving direct access to the bile ducts for studies of the murine biliary tree. 23 
Am. J. Physiol. - Gastrointest. Liver Physiol. 314, G349–G359 (2018). 24 
19.  C. Arber, M. K. Brenner, P. Reddy, Mouse models in bone marrow transplantation 25 
and adoptive cellular therapy. Semin. Hematol. 50, 131–144 (2013). 26 
 17 
20.  D. Nasralla, C. C. Coussios, H. Mergental, M. Z. Akhtar, A. J. Butler, C. D. L. Ceresa, 1 
V. Chiocchia, S. J. Dutton, J. C. García-Valdecasas, N. Heaton, C. Imber, W. Jassem, 2 
I. Jochmans, J. Karani, S. R. Knight, P. Kocabayoglu, M. Malagò, D. Mirza, P. J. 3 
Morris, A. Pallan, A. Paul, M. Pavel, M. T. P. R. Perera, J. Pirenne, R. Ravikumar, L. 4 
Russell, S. Upponi, C. J. E. Watson, A. Weissenbacher, R. J. Ploeg, P. J. Friend, A 5 
randomized trial of normothermic preservation in liver transplantation. Nature. 557, 6 
50–56 (2018). 7 
21.  A. I. Skaro, C. L. Jay, T. B. Baker, E. Wang, S. Pasricha, V. Lyuksemburg, J. A. 8 
Martin, J. M. Feinglass, L. B. Preczewski, M. M. Abecassis, The impact of ischemic 9 
cholangiopathy in liver transplantation using donors after cardiac death: The untold 10 
story. Surgery. 146, 543–553 (2009). 11 
22.  C. K. Enestvedt, S. Malik, P. P. Reese, A. Maskin, P. S. Yoo, S. A. Fayek, P. Abt, K. 12 
M. Olthoff, A. Shaked, Biliary complications adversely affect patient and graft survival 13 
after liver retransplantation. Liver Transpl. 19, 965–72 (2013). 14 
23.  C. J. E. Watson, V. Kosmoliaptsis, C. Pley, L. Randle, C. Fear, K. Crick, A. E. 15 
Gimson, M. Allison, S. Upponi, R. Brais, I. Jochmans, A. J. Butler, Observations on 16 
the ex situ perfusion of livers for transplantation. Am. J. Transplant. 18, 2005–2020 17 
(2018). 18 
24.  R. G. Farmer, W. A. Hawk, R. B. Turnbull, Farmer, R. G., Hawk, W. A. and Turnbull, 19 
R. B. (1975) ‘Clinical patterns in Crohn’s disease: a statistical study of 615 cases.’, 20 
Gastroenterology, 68(4 Pt 1), pp. 627–35. Available at: 21 
http://www.ncbi.nlm.nih.gov/pubmed/1123132 (Accessed: 11 September 201. 22 
Gastroenterology. 68, 627–35 (1975). 23 
25.  S. A. MacParland, J. C. Liu, X. Z. Ma, B. T. Innes, A. M. Bartczak, B. K. Gage, J. 24 
Manuel, N. Khuu, J. Echeverri, I. Linares, R. Gupta, M. L. Cheng, L. Y. Liu, D. Camat, 25 
S. W. Chung, R. K. Seliga, Z. Shao, E. Lee, S. Ogawa, M. Ogawa, M. D. Wilson, J. E. 26 
 18 
Fish, M. Selzner, A. Ghanekar, D. Grant, P. Greig, G. Sapisochin, N. Selzner, N. 1 
Winegarden, O. Adeyi, G. Keller, G. D. Bader, I. D. McGilvray, Single cell RNA 2 
sequencing of human liver reveals distinct intrahepatic macrophage populations. Nat. 3 
Commun. 9, 4383 (2018). 4 
26.  D. J. McCarthy, K. R. Campbell, A. T. L. Lun, Q. F. Wills, Scater: pre-processing, 5 
quality control, normalization and visualization of single-cell RNA-seq data in R. 6 
Bioinformatics. 33, 1179–1186 (2017). 7 
27.  A. Butler, P. Hoffman, P. Smibert, E. Papalexi, R. Satija, Integrating single-cell 8 
transcriptomic data across different conditions, technologies, and species. Nat. 9 
Biotechnol. 36, 411–420 (2018). 10 
28.  L. Haghverdi, A. T. L. Lun, M. D. Morgan, J. C. Marioni, Batch effects in single-cell 11 
RNA-sequencing data are corrected by matching mutual nearest neighbors. Nat. 12 
Biotechnol. 36, 421–427 (2018). 13 
29.  F. A. Wolf, P. Angerer, F. J. Theis, SCANPY: Large-scale single-cell gene expression 14 
data analysis. Genome Biol. 19 (2018), doi:10.1186/s13059-017-1382-0. 15 
30.  L. Haghverdi, M. Büttner, F. A. Wolf, F. Buettner, F. J. Theis, Diffusion pseudotime 16 
robustly reconstructs lineage branching. Nat. Methods. 13, 845–848 (2016). 17 
31.  X. Qiu, Q. Mao, Y. Tang, L. Wang, R. Chawla, H. A. Pliner, C. Trapnell, Reversed 18 
graph embedding resolves complex single-cell trajectories. Nat. Methods. 14, 979–19 
982 (2017). 20 
32.  V. Y. Kiselev, K. Kirschner, M. T. Schaub, T. Andrews, A. Yiu, T. Chandra, K. N. 21 
Natarajan, W. Reik, M. Barahona, A. R. Green, M. Hemberg, SC3: Consensus 22 
clustering of single-cell RNA-seq data. Nat. Methods. 14, 483–486 (2017). 23 
33.  L. Zappia, A. Oshlack, Clustering trees: a visualization for evaluating clusterings at 24 
multiple resolutions. Gigascience. 7 (2018), doi:10.1093/gigascience/giy083. 25 
 19 
34.  A. Subramanian, P. Tamayo, V. K. Mootha, S. Mukherjee, B. L. Ebert, M. A. Gillette, 1 
A. Paulovich, S. L. Pomeroy, T. R. Golub, E. S. Lander, J. P. Mesirov, Gene set 2 
enrichment analysis: A knowledge-based approach for interpreting genome-wide 3 
expression profiles. Proc. Natl. Acad. Sci. U. S. A. 102, 15545–15550 (2005). 4 
35.  M. V. Kuleshov, M. R. Jones, A. D. Rouillard, N. F. Fernandez, Q. Duan, Z. Wang, S. 5 
Koplev, S. L. Jenkins, K. M. Jagodnik, A. Lachmann, M. G. McDermott, C. D. 6 
Monteiro, G. W. Gundersen, A. Ma’ayan, Enrichr: a comprehensive gene set 7 
enrichment analysis web server 2016 update. Nucleic Acids Res. 44, W90–W97 8 
(2016). 9 
36.  F. Sampaziotis, M. C. De Brito, P. Madrigal, A. Bertero, K. Saeb-Parsy, F. a C. F. A. 10 
C. Soares, E. Schrumpf, E. Melum, T. H. T. H. Karlsen, J. A. A. Bradley, W. T. H. W. 11 
T. H. Gelson, S. Davies, A. Baker, A. Kaser, G. J. G. J. Alexander, N. R. F. N. R. F. 12 
Hannan, L. Vallier, M. Cardoso de Brito, P. Madrigal, A. Bertero, K. Saeb-Parsy, F. a 13 
C. F. A. C. Soares, E. Schrumpf, E. Melum, T. H. T. H. Karlsen, J. A. A. Bradley, W. 14 
T. H. W. T. H. Gelson, S. Davies, A. Baker, A. Kaser, G. J. G. J. Alexander, N. R. F. 15 
N. R. F. Hannan, L. Vallier, Cholangiocytes derived from human induced pluripotent 16 
stem cells for disease modeling and drug validation. Nat. Biotechnol. 33, 845–852 17 
(2015). 18 
37.  F. Sampaziotis, M. C. de Brito, I. Geti, A. Bertero, N. R. Hannan, L. Vallier, Directed 19 
differentiation of human induced pluripotent stem cells into functional cholangiocyte-20 
like cells. Nat. Protoc. 12, 814–827 (2017). 21 
38.  R. Ravikumar, W. Jassem, H. Mergental, N. Heaton, D. Mirza, M. T. P. R. Perera, A. 22 
Quaglia, D. Holroyd, T. Vogel, C. C. Coussios, P. J. Friend, Liver Transplantation 23 
After Ex Vivo Normothermic Machine Preservation: A Phase 1 (First-in-Man) Clinical 24 
Trial. Am. J. Transplant. 16, 1779–1787 (2016). 25 
 26 
 20 
Acknowledgements: 1 
The authors would like to thank A. Petrunkina and the NIHR Cambridge BRC Cell Phenotyping 2 
Hub for their help with flow cytometry and cell sorting; K Burling and the MRC MDU Mouse 3 
Biochemistry Laboratory [MRC_MC_UU_12012/5] for processing mouse serum samples; the 4 
Cambridge Biorepository for Translational Medicine for the provision of human tissue used in 5 
the study, Pedro Madrigal and Chichau Miau for bioinformatic support; Carlos Costa for his 6 
help with the NMP experiments; Kate Reid and Rachel Clarke for their help with radiological 7 
imaging during NMP experiments; Peter Humphreys, Darran Clements and Simon McCallum 8 
for their help with confocal imaging. The monoclonal antibody TROMA-III developed by R 9 
Kemler was obtained from the Developmental Studies Hybridoma Bank, created by the NICHD 10 
of the NIH and maintained at The University of Iowa, Department of Biology, Iowa City, IA 11 
52242.  12 
Funding: F.S. was supported by an NIHR Clinical Lectureship, the Academy of Medical 13 
Sciences Starter Grant for Clinical Lecturers, the Addenbrooke’s Charitable Trust and the 14 
Rosetrees Trust. T.B. was supported an EASL Juan Rodes fellowship. The L.V. lab is funded 15 
by the ERC advanced grant New-Chol, the Cambridge University Hospitals National Institute 16 
for Health Research Biomedical Research Centre and the core support grant from the 17 
Wellcome Trust and Medical Research Council of the Wellcome–Medical Research Council 18 
Cambridge Stem Cell Institute. L.G. and Sa.S were supported by a BHF Senior Research 19 
Fellowship (FS/18/46/33663).  20 
Author contributions: F.S. conceived and designed the study, performed experiments, 21 
acquired, interpreted and analysed the data, developed and validated the protocols described, 22 
performed bioinformatic analysis, generated the figures, wrote and edited the manuscript. 23 
D.M. Performed bioinformatics analysis. O.C.T. contributed to cell culture, harvesting and 24 
processing of tissue, single-cell RNA sequencing (scRNAseq) experiments and animal 25 
experiments. T.E.B. contributed to animal experiments. St.S. performed the Magnetic 26 
Resonance Imaging (MRI) experiments. E.M.G. performed the 3D reconstruction of the mouse 27 
 21 
biliary tree from MRI images. E.M.G. and S.S.U. reviewed and reported the MRI images. T.B. 1 
contributed to tissue culture, QPCR, immunofluorescence, tissue clarification and wholemount 2 
staining experiments. B.T.W. contributing to tissue dissociation and scRNAseq experiments. 3 
J.G.B. contributed to scRNAseq experiments. K.M. contributed to primary tissue harvesting 4 
tissue. Gi.C. contributed to scRNAseq experiments, flow cytometry and magnetic associated 5 
cell sorting. R.L.G. contributed to animal experiments, IF, and tissue histology. N.L.B. 6 
contributed to animal experiments. V.L.M. contributed to harvesting primary tissue. K.C., C.F., 7 
S.R, L.S. contributed to ex-vivo normothermic perfusion (NMP) experiments. J.B. contributed 8 
to imaging. L.G. contributed to tissue clarification and wholemount staining experiments as 9 
well as critical review of this data. D.O. and A.O. contributed to flow cytometry analyses and 10 
contributed through critical revision of the manuscript for important intellectual content. P. H. 11 
contributed to 3D reconstruction and rendering of immunofluorescence images. S.E.B. 12 
contributed to experimental design and critical revision of the manuscript for important 13 
intellectual content. Ga.C provided primary human samples. S. E. D. reviewed and reported 14 
the histology images and contributed through critical revision of the manuscript for important 15 
intellectual content. I.A. contributed to bile aspiration using microdialysis catheters. A.J.B., 16 
C.J.E.W. contributed to NMP experiments and critical revision of the manuscript for important 17 
intellectual content. T.C.S performed fluoroscopic cannulation of peripheral ducts in NMP 18 
experiments and contributed through critical revision of the manuscript for important 19 
intellectual content. M.P.M., W.T.H.G, G.F.M, Sa.S, S.T., P.G., E.M. contributed through 20 
critical revision of the manuscript for important intellectual content. K.S.-P. provided primary 21 
tissue, performed animal and ex-vivo normothermic perfusion experiments, contributed to the 22 
design and concept of study, interpreted the data and edited the manuscript. L.V. Designed 23 
and conceived the study, interpreted the data, wrote and edited the manuscript. All the authors 24 
approved the manuscript. 25 
 22 
Competing interests: FS, KSP and LV are founders and shareholders of BILITECH. LV is a 1 
founder and shareholder of DEFINIGEN. The remaining authors have no competing interests 2 
to disclose. 3 
Data and materials availability: All data is available in the main text or the supplementary 4 
materials. Single-cell RNA sequencing data are available on ArrayExpress. Accession 5 
number: E-MTAB-8495 6 
Supplementary Materials: 7 
Materials and Methods 8 
Figures S1 to S16 9 
Tables S1 to S3 10 
Movies S1 to S9 11 
Data S1 to S3 12 
References (25, 26, 35–38, 27–34)  13 
 23 
 1 
 2 
Fig. 1. Transcriptional profiling of primary cholangiocytes. (A) Schematic representation 3 
of the methodology used for single cell RNA sequencing (scRNAseq). (B) UMAP plot (7295 4 
primary cells, n=10 individuals) illustrating distinct primary cholangiocyte populations in 5 
different regions of the biliary tree. (C-D) Immunofluorescence images (C) and UMAP 6 
 24 
representation of normalized gene expression (D) of primary cholangiocytes illustrating 1 
differential expression of representative region markers. Scale bars: 50μm. (E) Heatmap of 2 
top 100 differentially expressed genes (DEGs) in pseudotime (Data S2) demonstrating a 3 
gradual transition in the transcriptional profile of cholangiocytes between different regions of 4 
the biliary tree. PRI, Primary; IHD, IntraHepatic Ducts; CBD, Common Bile Duct; GB, 5 
Gallbladder; P, Pangreas; D, Duodenum. 6 
  7 
 25 
 1 
 2 
Fig. 2. Cholangiocyte Organoid (CO) identity is controlled by niche stimuli. (A) UMAP 3 
(35,603 cells) of primary cholangiocytes and their corresponding organoids before and after 4 
gallbladder bile treatment, illustrating similarities between different region organoids and 5 
changes in their signature in response to bile. PRI, Primary; IHD, IntraHepatic Ducts; CBD, 6 
Common Bile Duct; GB, Gallbladder; ORG, Organoids; BTO, Bile-Treated Organoids. (B) 7 
Heatmap of top 100 Differentially Expressed Genes (DEGs) between primary regions, 8 
organoids and BTOs (Data S1-S2), illustrating that organoids lose regional differences and 9 
upregulate culture-related genes, but re-acquire gallbladder markers following bile treatment. 10 
(C-D) QPCR (C) (n=4 samples per group; center line, median; box, interquartile range (IQR); 11 
whiskers, range; housekeeping gene, HMBS; #P>0.05, **P<0.01, ***P<0.001, ****P<0.0001); 12 
 26 
and immunofluorescence (D) demonstrating upregulation of gallbladder markers and bile acid 1 
target genes following treatment with chenodeoxycholic acid (CDA), in the absence of the FXR 2 
inhibitor Z-GS.  Z-GS, Z-guggulsterone. Scale bars, 50μm.  3 
  4 
 27 
 1 
Fig. 3. Cholangiocyte organoids (COs) rescue cholangiopathy following transplantation 2 
and assume an identity corresponding to the site of engraftment. (A) Experimental 3 
outline schematic. (B) Kaplan-Meier curve (Table S1: number of animals at risk) 4 
 28 
demonstrating animal rescue following gallbladder organoids injection; P=0.0018(**), log-rank 1 
test. (C) Magnetic Resonance Cholangiopancreatography (MRCP) demonstrating rescue of 2 
cholangiopathy following organoid injection. (D) Immunofluorescence demonstrating 3 
engraftment of Red Fluorescent Protein (RFP)-expressing gallbladder organoids in portal 4 
triads, with upregulation of intrahepatic (SOX4) markers. Scale bars; yellow, 50μm; white, 5 
100μm. PV, portal vein. 6 
  7 
 29 
 1 
  2 
 30 
 1 
Fig. 4. Cholangiocyte organoids (COs) engraft in a human liver receiving Normothermic 2 
Perfusion (NMP) and improve bile properties. (A) Schematic representation of the 3 
technique for organoid injection and (B) photograph of the NMP circuit used. BD, Bile Duct; 4 
GB, Gallbladder; HA, Hepatic Artery; PV, Portal Vein; IVC, Inferior Vena Cava; L, Liver RFP, 5 
Red Fluorescent Protein; P, pump; O, oxygenator; PRC, Packed Red Cells. (C) Flow 6 
cytometry revealing absence of RFP cells in the perfusate. (D) Immunofluorescence revealing 7 
engraftment of RFP gallbladder organoids with upregulation of intrahepatic (SOX4) and loss 8 
of gallbladder (SOX17) markers. Scale bars, 50μm. (E) Organoid injection improves bile pH 9 
and choleresis. ***P<0.001. N=3 NMP livers. Each measurement is represented by a different 10 
data point, each organ is represented by a different symbol. 11 
 12 
 13 
  14 
 31 
 1 
 2 
 3 
Supplementary Materials for 4 
 5 
Cholangiocyte organoids can repair bile ducts after transplantation in human 6 
liver 7 
 8 
Fotios Sampaziotis*, Daniele Muraro, Olivia C. Tysoe , Stephen Sawiak, Timothy E. Beach, 9 
Edmund M. Godfrey, Sara S. Upponi, Teresa Brevini, Brandon T. Wesley, Jose Garcia-10 
Bernando, Krishnaa Mahbubani, Giovanni Canu, Richard Gieseck III, Natalie L. Berntsen, 11 
Victoria L. Mulcahy, Keziah Crick, Corrina Fear, Sharayne Robinson, Lisa Swift, Laure 12 
Gambardella, Johannes Bargehr, Daniel Ortmann, Stephanie E. Brown, Anna Osnato, 13 
Michael Murphy, Gareth Corbett, William T. H. Gelson, George F. Mells, Peter Humphreys, 14 
Susan E. Davies,  Irum Amin, Paul Gibbs, Sanjay Sinha, Sarah Teichmann, Andrew J Butler, 15 
Teik Choon See, Espen Melum, Christopher J. E. Watson, Kourosh Saeb-Parsy†, Ludovic 16 
Vallier†* 17 
†These authors share senior authorship  18 
*Correspondence to: fs347@cam.ac.uk; lv225@cam.ac.uk 19 
 20 
 21 
This PDF file includes: 22 
 23 
Materials and Methods 24 
Figs. S1 to S16 25 
Tables S1 to S3 26 
Captions for Movies S1 to S9 27 
Captions for Data S1 to S3 28 
 29 
Other Supplementary Materials for this manuscript include the following:  30 
 31 
Movies S1 to S9 32 
Movie S1: Movie S1 Carrier1 T1.mp4 33 
Movie S2: Movie S2 Carrier1 T2.mp4 34 
Movie S3: Movie S3 Cells1 T1.mp4 35 
Movie S4: Movie S4 Cells1 T2.mp4 36 
Movie S5: Movie S5 MR 3D reconstruction Carrier.mp4 37 
Movie S6: Movie S6 MR 3D reconstruction Cells.mp4 38 
Movie S7: Movie S7 Z-stack KRT19(GREEN)-RFP(RED).avi 39 
Movie S8: Movie S8 3D reconstruction KRT19(GREEN)-RFP(RED).mp4 40 
Movie S9: Movie S9 3D rendering KRT19(GREEN)-RFP(RED).mp4 41 
 42 
Data S1 to S3 43 
Data S1: Data S1 Differentially Expressed Genes in Primary Cholangiocytes.xlsx 44 
 32 
Data S2: Data S2 Differentially Expressed Genes in Pseudotime.xlsx 1 
Data S3: Data S2 Differentially Expressed Genes between organoids and primary 2 
cholangiocytes.xlsx 3 
4 
 33 
Materials and Methods 1 
 2 
Ethical approval 3 
Gallbladder, bile duct, liver biopsy and bile samples were obtained from deceased organ 4 
donors (National Research Ethics Committee East of England – Cambridge South 5 
15/EE/0152). Human livers retrieved for transplantation but subsequently declined were used 6 
for ex vivo administration of cholangiocytes (National Research Ethics Committee East of 7 
England – Cambridge East 14/EE/0137). All human tissue was used after obtaining informed 8 
consent for use in research.  9 
 10 
Tissue collection 11 
Gallbladder, bile duct, liver biopsies and bile were obtained under sterile conditions from 12 
deceased transplant organ donors as rapidly as possible after cessation of circulation. Tissue 13 
samples, and liver retrieved for transplantation but subsequently declined, were transferred to 14 
the laboratory at 4°C in University of Wisconsin (UW®) organ presentation solution.  15 
 16 
Tissue dissociation 17 
Resected tissue (gallbladder, extrahepatic ducts and liver) was transferred to the lab as 18 
described above and processed immediately after resection. Gallbladder and extrahepatic bile 19 
duct samples were drained of bile and the organ lumen was exposed through a longitudinal 20 
incision. Liver samples were divided into 1cm2 cubes prior to processing. All samples were 21 
washed twice with warm PBS with Ca2+Mg2+ +EDTA (0.5mM), followed by enzymatic 22 
digestion with using Liberase (0.2 Wünsch/ml) in an incubated shaker at 37oC and 200 RPM 23 
for 30 minutes. DNAse I (2000 U/ml) was added to the solution to prevent cell clumping and 24 
increase viability. Liver samples were dissociated further using the Miltenyi Biotec 25 
GentleMACS tissue dissociator and GentleMacs Tissue Dissociation C Tubes. For the 26 
gallbladder and extrahepatic duct samples, gentle mechanical scrapping of the lumen was 27 
adequate to release the epithelial cells following enzymatic digestion. All cell suspensions were 28 
filtered through 70um filters to remove debris and remaining tissue, washed with PBS 29 
containing 1% BSA (W/V) and centrifuged at 400g, for 5mins in a refrigerated centrifuge 30 
maintaining a temperature of 4oC. The cells were resuspended in Miltenyi Biotec red blood cell 31 
(RBC) lysis and incubated for 10 minutes at room temperature (RT). The Miltenyi Biotec 32 
Debris Removal solution kit was used according to the manufacturer’s instructions to remove 33 
remaining debris and dead cells. For liver samples, the resulting cell suspensions were 34 
centrifuged at 50g for 5 minutes (4oC) to pellet the hepatocyte fraction, the supernatant was 35 
collected and cholangiocytes were isolated as described below. 36 
 37 
Cell isolation 38 
Following tissue dissociation to single cells, cholangiocytes were isolated with Magnetic 39 
Associated Cell Sorting (MACS) using the Miltenyi Biotech autoMACS Pro separator and 40 
CD326 (EpCAM) MicroBeads according to the manufacturer’s instructions. The resulting cells 41 
were counted, centrifuged at 444g for 5 minutes, resuspended to a concentration of 1000 cells/ 42 
μL and stored on ice. 43 
 44 
10x Single Cell Library Making Process 45 
GEM-RTs (Gel-beads-in-emulsion which barcode the ploy adenylated mRNAs, followed 46 
by Reverse Transcription) were broken and Silane magnetic beads are used to purify first stand 47 
cDNA from the GEM-RT mixture and the cDNA was then amplified via PCR. Enzymatic 48 
fragmentation, end-repair and A-tailing were followed by size selection (using SPRISelect 49 
reagent). An adapter was ligated to the fragments and following a clean-up step, index PCR 50 
 34 
took place. After a further round of size selection with SRISelect, completed libraries were 1 
quantified, (Agilent Bioanalyser and qPCR) and diluted for running on an Illumina sequencing 2 
instrument (HS4000). 3 
 4 
Processing and normalization of 10X data 5 
The results from the sequencing runs were checked manually to confirm that the overall 6 
yield and quality were as expected. The data from the instrument were converted to fastq 7 
format, the input format required by the 10X software cellranger, and aligned using the human 8 
reference GRCh36-1.2.0 available from 10X. The dataset was augmented by integrating counts 9 
of a cluster of cholangiocytes from a published dataset (cluster 17 in MacParland SA et al, 10 
2018) (17). Cells were annotated as part of different origins, these being primary tissue (PRI), 11 
untreated organoids (ORG), treated organoids (ORGT). Each origin comprises three regions: 12 
intrahepatic duct (IHD), common bile duct (CBD), gallbladder (GB). The number of cells in 13 
each origin and region are reported in Figure S1C. Genes with read counts > 0 in at least 3 14 
cells from each batch in at least one origin were maintained for downstream analysis. Low 15 
quality cells were removed based on the percentage of UMI mapping to the mitochondrial 16 
genome and the number of genes detected by determining outliers (3 median-absolute-17 
deviations) with the routine isOutlier in the package scater (18). Cholangiocytes were isolated 18 
by retaining cells expressing at least one of the biliary markers EPCAM, KRT7, KRT19 (with 19 
number of counts > 3). Normalization, identification of highly variable genes and cell cycle 20 
regression (regressing out the difference between the G2M and S phase scores) were performed 21 
with the Seurat package (19). We employed the routine fastMNN in scran for batch correction 22 
(20). Batch corrected samples are shown in figure 2A. Small clusters derived by applying the 23 
Louvain method for community detection and characterized by cells which were outliers in the 24 
percentage of UMI mapping to the mitochondrial genome and the number of genes detected 25 
were filtered out. 26 
 27 
Analysis of normalized 10X data 28 
The normalized data were clustered using the Louvain method in the Scanpy package (21) 29 
by selecting a resolution which generated 3 clusters and with 10 random initialisations. 30 
Similarity between Louvain clusters and origin annotations was assessed using the Adjusted 31 
Rand Index (ARI) and the Adjusted Mutual Information (AMI). Both measures lie in the 32 
interval [0,1], where a value close to 0 indicates random labelling and exactly 1 means that the 33 
two partitions are identical. The average value calculated on the different partitions obtained 34 
by random initializations was > 0.95 for both measures, indicating a high correspondence 35 
between origins and clusters (Fig. S5A). The same analysis performed on regions showed poor 36 
matching between regions and clusters, suggesting similarity in the transcriptional profile of 37 
cells located in different regions (Fig. S5A-B). Transcriptional similarity was quantified at 38 
origin and region resolution by estimating the connectivity of data manifold partitions within 39 
the partition-based graph abstraction (PAGA) framework. At the origin resolution, this analysis 40 
notably highlighted higher transcriptional similarity between treated organoids and primary 41 
tissue than between untreated organoids and primary tissue (Fig. S9B). Interestingly, at the 42 
region resolution we identified higher transcriptional similarity between adjacent locations in 43 
primary tissues, with intrahepatic duct and gallbladder having the lowest connectivity value. 44 
This association between connectivity and anatomical location, together with the similarity of 45 
cells located in different regions, suggested a gradual variation in the transcriptional profile of 46 
cells in primary tissue that could be represented as a pseudo-spatial dimension. In this view, 47 
we analyzed the primary tissue by applying two methods for pseudo-temporal (or pseudo-48 
spatial) ordering: diffusion pseudo-time (22) and Monocle 2 (23). In Monocle 2 differential 49 
expression in pseudotime was calculated using the differential GeneTest routine. Both methods 50 
 35 
confirmed an association between transcriptional similarity and anatomical location, as 1 
highlighted by the density plot in Figure S4B and allowed the representation of regional 2 
markers along a pseudo-spatial dimension (Fig. S4C). Since the majority of cells had a 3 
diffusion pseudotime value >0.65 the density plot if figure S4B is shown in the range [0.65,0.9] 4 
to improve visualization and avoid overcrowding. We then analyzed each region individually 5 
in organoids (treated and untreated) and primary tissue to identify potential subpopulations of 6 
cells. Due to the relatively small sample sizes, we applied the clustering method SC3, whose 7 
high accuracy and robustness is derived combining multiple clustering solutions through a 8 
consensus approach (24). SC3 allows the user to pre-define the number of clusters. Because of 9 
the arbitrariness of this choice we varied the number of clusters between 1 and 10, calculated 10 
the stability of clusters across resolutions (SC3 stability index) and built a clustering tree 11 
showing how cells move as the clustering resolution is increased (package clustree), (25). As 12 
shown in Figure S5C, no stable sub-trees were formed within each region, indicating absence 13 
of stable clusters defining subpopulations of cells.  14 
Regional markers and differentially expressed genes were identified by applying the 15 
Wilcoxon-Rank-Sum test (p-value<0.01, |log2 fold change| > 1) in Scanpy. Gene set, gene 16 
ontology and pathway enrichment were performed using the packages GSEA (26) and Enrichr 17 
(27).  18 
 19 
Data availability 20 
10X raw data (fastq files) have been deposited in the repository ArrayExpress with the 21 
accession number E-MTAB-8495 22 
 23 
Organoid derivation and culture 24 
A portion of the cells isolated for scRNAseq was cultured and propagated as organoids 25 
using our established methodology (11, 12). Cells were cultured under the same conditions 26 
irrespective of their region of origin. 27 
 28 
Immunofluorescence, RNA extraction and Quantitative Real Time PCR 29 
IF, RNA extraction and QPCR were performed as previously described (11, 12, 28, 29). 30 
A complete list of the primary and secondary antibodies used is provided in table S2. A 31 
complete list of the primers used is provided in table S3. 32 
All QPCR data are presented as the median, interquartile range (IQR) and range 33 
(minimum to maximum) of four independent lines unless otherwise stated. Values are relative 34 
to the housekeeping gene Hydroxymethylbilane Synthase (HMBS). 35 
All IF images were acquired using a Zeiss Axiovert 200M inverted microscope or a Zeiss 36 
LSM 700 confocal microscope. Imagej 1.48k software (Wayne Rasband, NIHR, USA, 37 
http://imagej.nih.gov/ij) was used for image processing. IF images are representative of 3 38 
different experiments. 39 
 40 
GGT activity 41 
GGT activity was measured in triplicate using the MaxDiscovery™ gamma-Glutamyl 42 
Transferase (GGT) Enzymatic Assay Kit (Bioo scientific) based on the manufacturer’s 43 
instructions. Error bars represent SD.  44 
 45 
Alkaline Phosphatase staining 46 
Alkaline phosphatase was carried out using the BCIP/NBT Color Development Substrate 47 
(5-bromo-4-chloro-3-indolyl-phosphate/nitro blue tetrazolium) (Promega) according to the 48 
manufacturer’s instructions. 49 
 50 
 36 
Flow cytometry analyses 1 
Flow cytometry analyses were performed as previously described (11, 12, 28, 29). 2 
 3 
Bile acid treatment 4 
Organoids were incubated for 72 hours with 10μM CDA (Sigma, C9377-5G) in the 5 
presence or absence of 10μM Z-GS (Santa Cruz, sc-204414). 6 
 7 
Animal experiments 8 
All animal experiments were performed in accordance with UK Home Office regulations 9 
(UK Home Office Project License number PPL 70/8702). Immunodeficient NSG mice 10 
(NOD.Cg-Prkdcscid Il2rgtm1Wjl/SzJ), which lack B, T and NK lymphocytes, were bred in 11 
house, and food and water were available ad libitum before and after procedures. Male animals 12 
aged 4–8 weeks were used. Animals were assigned randomly to treatment and control groups. 13 
Experiments were performed blinded, and where this was not possible (e.g., due to performance 14 
of a surgical procedure), data were analysed blinded to the identity of the experimental groups. 15 
Littermate animals were used as controls. 16 
 17 
Cell delivery  18 
Cholangiocytes were delivered into the liver retrogradely through the extrahepatic biliary 19 
tree (14). In brief, a fine bore cannula was placed and secured in the gallbladder. To divert the 20 
infusion into the liver, the distal common bile duct was occluded with a clamp. The cells were 21 
infused through the cannula in the gallbladder in a total volume of 1μl/g of total body weight, 22 
at a maximum speed of 1μl/second.  23 
 24 
MDA administration 25 
Cholangiopathy was induced through intraperitoneal (IP) administration of 4,4′-26 
methylene dianiline (MDA) on 3 occasions 7, 5, and 3 days prior to cell delivery at a 27 
concentration of 50 μg/g of total body weight. An additional dose of MDA was administered 28 
directly into the extrahepatic biliary tree prior to cell delivery as described above.  29 
 30 
Blood sample collection 31 
Blood was taken using a 23g needle directly from the inferior vena cava under terminal 32 
anesthesia at the time the animals were electively culled and transferred into 1.5ml Eppendorf 33 
tubes for further processing. 34 
 35 
Blood sample processing 36 
The blood samples were routinely processed by the University of Cambridge Core 37 
biochemical assay laboratory (CBAL). All of the sample analysis was performed on a Siemens 38 
Dimension EXL analyzer using reagents and assay protocols supplied by Siemens. 39 
 40 
Tissue collection 41 
Tissue for sectioning and staining was collected at the end of all animal experiments when 42 
the animals were culled, unless otherwise stated. The animals were culled due to due to animal 43 
welfare reasons (weight loss, jaundice and clinical deterioration) or electively 3 months after 44 
transplantation. Timepoints are indicated on the relevant Kaplan-Meier curves (Fig. 3B; Fig. 45 
S13A). 46 
 47 
Cryosectioning 48 
Excised tissue was fixed in 4% PFA, immersed in sucrose solution overnight, mounted in 49 
optimal cutting temperature (OCT) compound and stored at -80°C until sectioning. Sections 50 
 37 
were cut to a thickness of 6-10µm using a cryostat microtome and mounted on microscopy 1 
slides for further analysis. 2 
 3 
Haematoxylin and Eosin (H&E) Staining 4 
H&E staining was performed by the histology service of Addenbrooke’s hospital or using 5 
Sigma-Aldrich reagents according to the manufacturer’s instructions. Briefly, tissue sections 6 
were hydrated, treated with Meyer’s Haematoxylin solution for 5 minutes (Sigma-Aldrich), 7 
washed with warm tap water for 15 minutes, placed in distilled water for 30-60 seconds and 8 
treated with eosin solution (Sigma-Aldrich) for 30-60 seconds. The sections were subsequently 9 
dehydrated and mounted using the Eukitt® quick-hardening mounting medium (Sigma-10 
Aldrich).  11 
 12 
Histology 13 
Histology sections were reviewed by an independent histopathologist with a special 14 
interest in hepatobiliary histology (SD). 15 
 16 
Quantification of transplanted cells in mouse liver 17 
For each animal 3 random sections were analyzed, with different lobes being assessed. A 18 
total of 49,846 cells were analyzed, approximately 10,000 cells per animal. 19 
 20 
MR imaging 21 
Magnetic resonance cholangio-pancreatography was performed after sacrifice of the 22 
animals. MRCP was performed at 9.4T using a Bruker BioSpec 94/20 system (Bruker, 23 
Ettlingen, Germany). For higher signal to noise ratio to give improved visualisation of the 24 
biliary ducts a two-dimensional sequence was used with slightly varied parameters (24 spaced 25 
echoes at 11ms intervals to give an effective echo time of 110ms; repetition time 5741ms; 26 
matrix size of 256×256; field of view of 4.33×5.35cm2 yielding a planar resolution of 27 
170×200µm2). Slices were acquired coronally through the liver and gall bladder with a 28 
thickness of 0.6mm. For this acquisition, a volume coil was used to reduce the impact of 29 
radiofrequency inhomogeneity. 30 
To examine the biliary tree, images were prepared by maximum intensity projections. 31 
Structural imaging to rule out neoplastic growths was performed using a T1-weighted 3D 32 
FLASH (fast low-angle shot) sequence with a flip angle of 25°, repetition time of 14ms and an 33 
echo time of 7ms. The matrix was 512×256×256 with a field of view of 5.12×2.56×2.56cm3 34 
for a final isotropic resolution of 100 µm. 35 
Volume rendered images of the biliary tree were generated from source data using Osirix 36 
software. The region of interest was segmented from the remaining data manually. 37 
The MRCP images were reviewed by 2 independent radiologists with a special interest in 38 
hepatobiliary radiology (EMG, SU).  39 
 40 
Ex vivo normothermic perfusion of donor livers 41 
The metra (OrganOx, Oxford, UK) normothermic liver perfusion device was used for ex 42 
vivo perfusion of human livers as previously described (15, 30). The machine, which is 43 
clinically used for preservation of livers for transplantation (15) enables prolonged automated 44 
organ preservation by perfusing it with ABO-blood group-compatible normothermic 45 
oxygenated blood. The perfusion device incorporates online blood gas measurement, as well 46 
as software-controlled algorithms to maintain pH, PO2 and PCO2 (within physiological limits), 47 
temperature and mean arterial pressure within physiological normal limits. In brief, the hepatic 48 
artery, portal vein, inferior vena cava and bile duct were cannulated, connected to the device 49 
and perfusion commenced.  50 
 38 
 1 
Bile duct cannulation 2 
Cannulation of the bile duct was achieved by inserting two 4 Fr sheaths into the common 3 
bile duct under fluoroscopy guidance, followed by cannulation of the left and right hepatic 4 
ducts and subsequently segment 3 and segment 5 ducts respectively, using two 2.7 Fr 5 
microcatheters via the sheaths. Peripheral placement of the microcatheters was confirmed by 6 
cholangiogram with small amount of ionic contrast medium. Cells were injected into segment 7 
3 and carrier was injected into segment 5.  8 
 9 
Cell delivery 10 
RFP-expressing organoids were mechanically dissociated to a mixture of small clumps 11 
and single cells and approximately 10x106 RFP-expressing cells were administered in a 12 
peripheral duct of segment 3 with a distribution area of ~2cm3, which was cannulated under 13 
fluoroscopic guidance to maximize cell delivery (see Bile duct cannulation section) (Fig. 14 
S15B). Carrier medium was delivered in a peripheral branch of segment 5 using the same 15 
technique and the organ was maintained on NMP for up to 100 hours.  16 
 17 
Quantification of transplanted cells in human livers  18 
3 human livers injected with RFP-labelled gallbladder organoids were analysed. Sections 19 
were obtained from the area of the distribution of the cells (~2cm3). 5 sections per liver and a 20 
total of 4,463 cells were analysed. 21 
 22 
Bile aspiration  23 
Bile duct cannulation was performed as described in the relevant section. Following 24 
cannulation, 2 microfluidic catheters (CMA Microdialysis Catheter, Harvard  Biosience Inc, 25 
USA) were placed into the respective segmental ducts using a guide wire exchange technique.  26 
The inner and outer shaft of the catheter and the inlet and outlet tubing are made of 27 
polyurethane and the membrane composed of polyarylethersulphone with a membrane pore 28 
size of 100kDa and outer diameter of 0.4mm. The inlet tubing for each catheter was connected 29 
to a portable battery driven CMA 107 Microdialysis Pump (Harvard Biosience Inc, USA) and 30 
the pump was set to aspirate at a rate of 1µl/min. 31 
 32 
Bile volume and pH measurements 33 
Measurements were performed in n=3 different livers. A minimum of 2 repeat 34 
measurements were performed for each liver increasing to 3 where possible, as previously 35 
described (27). Bile volume was normalised over the volume of the bile ducts producing it, 36 
which corresponds to the volume of distribution of the cells or the carrier in the control arm. 37 
This was calculated using the volume of the contrast medium required to delineate these ducts 38 
on cholangiogram. Please note all catheters were primed prior to volume measurements. 39 
 40 
Ultrasound imaging 41 
The liver was imaged ex-vivo in a normothermic perfusion device using a Hitachi Aloka 42 
Arrieta V70 and a 10Mhz hand-held probe. Images were obtained in axial and sagittal planes 43 
and assessment of the portal vein, hepatic veins and their major branches was carried out. The 44 
intrahepatic bile ducts were also assessed, with particular attention to segment 3 where the 45 
organoids had been instilled, and a control area in segment 5 receiving carrier. 46 
 47 
Statistical analysis 48 
All statistical analyses were performed using GraphPad Prism 6. For small sample sizes 49 
where descriptive statistics are not appropriate, individual data points were plotted. For 50 
 39 
comparison between 2 mean values a 2-sided Student’s t-test was used to calculate statistical 1 
significance. The normal distribution of our values was confirmed using the D'Agostino & 2 
Pearson omnibus normality test where appropriate. Variance between samples was tested using 3 
the Brown-Forsythe test. For comparing multiple groups to a reference group one-way 4 
ANOVA followed by Dunnett’s test was used between groups with equal variance, while the 5 
Kruskal-Wallis test followed by Dunn’s test was applied for groups with unequal variance. 6 
Survival was compared using log-rank (Mantel-Cox) tests. Where the number of replicates (n) 7 
is given this refers to organoid lines or number of different animals unless otherwise stated.  8 
For animal experiments, group sizes were estimated based on previous study variance. 9 
Final animal group sizes were chosen to allow elective culling at different time point while 10 
maintaining n > 4 animals surviving past 30 days to ensure reproducibility. No statistical 11 
methods were used to calculate sample size. No formal randomization method was used to 12 
assign animals to study groups. However, littermate animals from a cage were randomly 13 
assigned to experimental or control groups by a technician not involved in the study. No 14 
animals were excluded from the analysis. Blinding was used for radiology imaging.  15 
 16 
 17 
  18 
 40 
 1 
 2 
 3 
Fig. S1. 4 
Characteristics and quality control of single cell RNA sequencing samples. (A) UMAP 5 
plot of all sequenced samples and 1 publicly available intrahepatic cholangiocyte dataset (PRI 6 
IHD 5; from MacParland SA et al, 2018, cluster 17). Each patient and cell line are distinguished 7 
by a unique color and marker combination. (B) Number of genes and percentage of 8 
mitochondrial genes detected per cell. (C) Number of cells isolated from each region PRI, 9 
Primary; IHD, IntraHepatic Ducts; CBD, Common Bile Duct; GB, Gallbladder; ORG, 10 
Organoids; BTO, Bile-treated organoids. 11 
 12 
  13 
 41 
 1 
Fig. S2. 2 
Single cell RNA sequencing characterization of primary cholangiocytes. (A) UMAP plots 3 
demonstrating the expression of key cholangiocyte markers by the isolated cells, confirming 4 
their biliary identity. (B) UMAP plot of primary cholangiocytes compared to stellate and liver 5 
sinusoidal endothelial cells (LSECs) illustrating overlap between different region 6 
cholangiocytes when compared to a different cell type, which reflects a shared core biliary 7 
signature. (C) UMAP plots illustrating the expression of LSEC and stellate cell markers, 8 
confirming the cells’ identity. (D-E) PAGA connectivity plot (D) and corresponding 9 
connectivity values (E) demonstrating a higher degree of transcriptional similarity between 10 
cholangiocytes from different regions compared to different cell types, confirming the shared 11 
core transcriptional signature of the cells. IHD, IntraHepatic Ducts; CBD, Common Bile Duct; 12 
GB, Gallbladder. 13 
 14 
  15 
 42 
 1 
 43 
Fig. S3. 1 
Characterization of the transcriptional signature of cholangiocytes from different regions 2 
of the biliary tree. (A) Heatmap of top 100 Differentially Expressed Genes (DEGs) in 3 
cholangiocytes isolated from distinct regions of the biliary tree revealing transcriptional 4 
diversity in the primary biliary epithelium. IHD, IntraHepatic Ducts; CBD, Common Bile Duct; 5 
GB, Gallbladder (Data S1). (B) UMAP plots confirming the expression of previously 6 
described markers in IHDs. (C) Gene Ontology (GO) analysis on DEGs between biliary tree 7 
regions using EnrichR illustrating enrichment of cholangiocyte-to-niche interaction markers, 8 
such as bile processing and modifying genes. (D) Gene Set Enrichment Analyses on DEGs 9 
between biliary tree regions identifying differences in the expression of YAP target genes, 10 
P<0.001. (E-F) PAGA connectivity plot (E) and corresponding connectivity values (F) 11 
demonstrating a higher degree of transcriptional similarity between adjacent regions of the 12 
biliary tree. Connectivity values illustrated in (E) are multiplied by 100.  13 
14 
 44 
  1 
 45 
Fig. S4. 1 
Pseudotime analysis of primary cholangiocytes. (A) Cell trajectory in pseudotime using 2 
Monocle; (B) Density plot of pseudo-time coordinates and (C) Gene expression in pseudotime 3 
of representative region markers indicating a gradual transition in transcriptional profile 4 
between cholangiocyte populations from adjacent regions. IHD: Intrahepatic Ducts, CBD: 5 
Common Bile Duct, GB: Gallbladder 6 
  7 
 46 
 1 
 47 
Fig. S5. 1 
Characterization of cluster stability. (A) Adjusted Rand Index (ARI) and the Adjusted 2 
Mutual Information (AMI) confirming that primary cholangiocytes, organoids, and bile-treated 3 
organoids constitute distinct populations by illustrating a high correspondence between 4 
Louvain clusters and cell type (primary, organoids, bile-treated organoids) annotations 5 
(average value > 0.95 for both measures) vs. poor correspondence between Louvain clusters 6 
and region (intrahepatic ducts, common bile duct, gallbladder) annotations (average value<0.3 7 
for both measures). (B) UMAP plot of Louvain clusters demonstrating poor matching between 8 
regions and clusters. The plot corresponds to the UMAP plot in Fig. 1B illustrating different 9 
regions. (C-D) Clustering trees derived from SC3 clusters by varying the pre-defined number 10 
of clusters k from 1 to 10 (see Methods) for a positive control comprising of stellate cells and 11 
LSECs (C) vs. cholangiocytes from different regions and corresponding cholangiocyte 12 
organoids (D). Cluster stability across different clustering resolutions confirms the presence of 13 
different populations (stellate vs. LSECs) in the positive control (C); while the absence of well-14 
defined cholangiocyte subpopulations in each anatomical region or between organoids from 15 
different regions is demonstrated by the lack of stable clusters in (D). 16 
  17 
 48 
 1 
 2 
Fig. S6. 3 
Characterization of cholangiocyte organoids from different regions of the biliary tree. (A) 4 
Immunofluorescence and (B) QPCR analysis of cholangiocyte organoids derived from 5 
different regions of the biliary tree demonstrating uniform expression of key biliary markers. 6 
n=4 samples per group; center line, median; box, interquartile range (IQR); whiskers, range; 7 
housekeeping gene, HMBS; #P>0.05#; scale bars, 50μm. (C-D) Organoids from different 8 
regions demonstrate Alkaline Phosphatase (ALP) (C) and GGT (Gamma-glutamyltransferase) 9 
(D) function. Scale bars, 100μm. (E) Growth curves illustrating comparable expansion 10 
potential between organoids from different regions. #, P>0.05. IHD, IntraHepatic Ducts; CBD, 11 
Common Bile Duct; GB, Gallbladder; ORG, Organoids; Primary, Primary CBD 12 
cholangiocytes. 13 
 14 
  15 
 49 
 1 
Fig. S7. 2 
Single-cell RNA sequencing characterization of cholangiocyte organoids from different 3 
regions of the biliary tree. (A) PCA (unregressed, 24.8%; cell cycle regression, 21.8% of 4 
variance) and (B) UMAP representation demonstrating overlap in the transcriptional profile of 5 
different region organoids before and after cell cycle regression, confirming that cell cycle 6 
genes are not responsible for these similarities. (C) UMAP plot demonstrating that organoids 7 
and primary cholangiocytes irrespective of region occupy adjacent and overlapping spaces 8 
when compared to different cell types, illustrating a shared cholangiocyte transcriptional 9 
signature between biliary cells in vivo and in vitro. (D) PAGA connectivity plot demonstrating 10 
a higher degree of transcriptional similarity between cholangiocytes in vivo (PRI, Primary) and 11 
in vitro (ORG, organoids) compared to different cell types, confirming the shared core 12 
transcriptional signature of the cells. Respective connectivity values multiplied by 100 are 13 
illustrated on the plot. IHD, IntraHepatic Ducts; CBD, Common Bile Duct; GB, Gallbladder; 14 
LSECs, Liver Sinusoidal Endothelial Cells. (E) UMAP representation following regression of 15 
 50 
cell cycle genes illustrating that the similarities between cholangiocyte organoids are preserved 1 
despite cell-cycle regression and therefore they are not attributable to a common ‘proliferation’ 2 
signature. (F) UMAP representation of cells co-expressing somatic stem cell markers 3 
(normalized expression>1), illustrating that similarities between organoids are not attributable 4 
to a common ‘stem cell’ signature. (G) UMAP representation of normalized gene expression 5 
values showing that organoids lose differences in the expression of region marks in culture. 6 
  7 
 51 
 1 
 2 
 3 
Fig. S8. 4 
Gene ontology (GO) analyses on cholangiocyte organoids. (A-B) GO analysis on 5 
differentially expressed genes between primary cholangiocytes and organoids using EnrichR 6 
demonstrating that genes upregulated in primary tissue (A) are related to cholangiocyte-to-7 
niche interaction, such as bile processing genes; while genes upregulated in organoids (B) 8 
reflect adaptation to cell culture conditions such as insulin, pyruvate and cytokine processing 9 
genes. (C) Gene Set Enrichment Analyses on DEGs between primary cells and organoids 10 
identifying differences in the expression of bile acid processing genes, P= 0.035. IHD, 11 
IntraHepatic Ducts; CBD, Common Bile Duct; GB, Gallbladder; ORG, Organoids. 12 
 13 
 14 
  15 
 52 
 1 
 53 
 1 
Fig. S9. 2 
Characterization of bile-treated organoids. (A) PCA analysis (16.8% of variance) showing 3 
overlap between organoids, primary cholangiocytes and bile-treated organoids irrespective of 4 
region suggesting a shared core transcriptional profile between all cells. (B) PAGA 5 
connectivity plot demonstrating that bile-treated organoids (BTO) shift their transcriptional 6 
profile towards primary gallbladder cholangiocytes. (C) Connectivity values corresponding to 7 
the PAGA connectivity plot in panel (B) IHD, IntraHepatic Ducts; CBD, Common Bile Duct; 8 
GB, Gallbladder; ORG, Organoids; BTO, Bile-treated organoids; PRI, Primary. (D-E) GSEA 9 
(D) and GO analysis using EnrichR (E) on differentially expressed genes in organoids before 10 
and after treatment with bile showing enrichment in bile processing genes and in particular bile 11 
acid nuclear receptors and their downstream targets. P=0.012. (F) UMAP representation of 12 
normalized gene expression values illustrating upregulation of gallbladder markers and bile 13 
acid downstream targets following treatment of organoids with gallbladder bile. 14 
 15 
  16 
 54 
 1 
 2 
  3 
 55 
Fig. S10. 1 
Gallbladder organoids rescue an acute cholangiopathy mouse model following 2 
transplantation. (A) Weight curve of animals treated with MDA (not transplanted) vs. animals 3 
injected with organoids following toxin treatment, demonstrating that injected animals recover 4 
and gain weight; n=5 animals in each arm. (B) Serum biochemistry demonstrating resolution 5 
of cholestasis following organoid injection; *P<0.05, #P>0.05, Kruskal-Wallis test. (C) 6 
Immunofluorescence images of MDA treated animals not transplanted with cells (toxin 7 
injection) vs. untreated controls (no injection) illustrating biliary injury following MDA 8 
administration. The images are complementary to Fig. 3D. (D) Histology (Heamatoxylin & 9 
Eosin and Elastic Picro Sirius Red) illustrating resolution of cholangiopathy following 10 
organoid injection. Asterisks: Bile ducts.  11 
  12 
 56 
 1 
Fig. S11.  2 
Gallbladder organoids regenerate the biliary tree of an acute cholangiopathy mouse 3 
model following transplantation. (A) Magnetic Resonance Cholangiopancreatography 4 
(MRCP) demonstrating biliary injury with loss of bile duct signal (white), immediately after 5 
toxin injection. The white dashed line outlines the liver margins. The image is complementary 6 
to Fig. 3C. Scale bars, 5mm. (B) 3D reconstruction of MRCP images demonstrating biliary 7 
injury with loss of bile duct signal in MDA-treated animals receiving carrier (not transplanted); 8 
vs. duct reconstruction in MDA-treated animals receiving organoid injections; vs. healthy 9 
animals. Scale bars, 5mm. (C) Quantification of bile duct signal on MRCP normalized over 10 
total liver volume in not transplanted vs. transplanted vs. healthy animals, demonstrating 11 
resolution of cholangiopathy following organoid injection; #, P>0.05; *, P<0.05; **, P<0.01; 12 
one-way ANOVA.  13 
 57 
 1 
 58 
Fig. S12.  1 
Gallbladder organoids regenerate the biliary epithelium of an acute cholangiopathy 2 
mouse model following transplantation. (A-B) Immunofluorescence analysis demonstrating 3 
engraftment, expression of key biliary markers, loss of gallbladder markers, expression of 4 
intrahepatic markers, absence of markers of other hepatic lineages (A); and expression of 5 
human specific markers, proliferation markers and active YAP (B) in human Red Fluorescent 6 
Protein (RFP) expressing cells following transplantation in immunocompromised mice with 7 
cholangiopathy. Scale bars; (A), 50μm; (B), 50μm (yellow), 100μm (white). The images are 8 
complementary to Fig. 3. (C) Quantification of human gallbladder-derived RFP-expressing 9 
cells in the bile ducts of transplanted vs. not transplanted animals; ** P<0.01; Mann-Whitney 10 
test. The data corresponds to 5 different animals and 3 random sections per animal. Each 11 
section is represented by a data point, while each animal is represented by a different symbol. 12 
(D-E) Quantification of the ratio of cells expressing proliferation markers (Ki67, D) and YAP 13 
downstream targets (CYR61, E) in ducts regenerated from engrafted human RFP-expressing 14 
cells vs. native mouse bile ducts in the same animals; # P>0.05; Mann-Whitney test.  15 
  16 
 59 
 1 
Fig. S13.  2 
Primary human cholangiocytes and mesenchymal stem cells fail to rescue mice with acute 3 
cholangiopathy following transplantation. (A) Kaplan-Meier curve of mice with MDA-4 
induced cholangiopathy receiving directly isolated human primary gallbladder cholangiocytes 5 
and human mesenchymal stem cells (MSCs) vs. carrier medium (carrier) demonstrating no 6 
statistically significant difference in survival between the three groups; P>0.05, log-rank test. 7 
(B) Serum biochemistry at the end of the experiment demonstrating persistent cholestasis in 8 
animals receiving primary gallbladder cholangiocytes, MSCs or carrier medium compared to 9 
healthy controls; *P<0.05, ***P<0.001, #P>0.05, one-way ANOVA. (C) Staining for human 10 
markers following cell transplantation reveals lack of engraftment of MSCs; while primary 11 
gallbladder cholangiocytes exhibit low level engraftment, which was not adequate to repair the 12 
damaged bile duct epithelium (white arrowheads). Engrafted primary gallbladder 13 
cholangiocytes lose gallbladder markers and upregulate intrahepatic markers. Scale bars; 14 
white, 100μm; yellow, 10 μm.  15 
 60 
 1 
Fig. S14.  2 
Transplantation of human common bile duct organoids in mouse gallbladder. 3 
Immunofluorescence analysis demonstrating expression of gallbladder markers and loss of 4 
common bile duct markers following transplantation of cholangiocyte organoids derived from 5 
human common bile duct in the gallbladder of immunocompromised mice. Scale bars; white, 6 
100μm; yellow, 10μm. 7 
  8 
 61 
 1 
 62 
Fig. S15. 1 
Administration of gallbladder organoids in human livers receiving Normothermic 2 
Perfusion (NMP). (A) Photograph of a human liver on NMP demonstrating anatomical 3 
landmarks, as well as the bile duct catheter used for administration of the Red Fluorescent 4 
Protein (RFP) expressing organoids. PV, portal vein; IVC, inferior vena cava; HA, hepatic 5 
artery; BD, Bile duct; GB, gallbladder; L, Liver. (B) Fluoroscopic images of peripheral duct 6 
cannulation. The position of the biliary catheters used for the injection of cells or carrier in the 7 
peripheral ducts of liver segments 3 and 5 respectively is shown in the top image. A 8 
cholangiogram of segment 3 following catheter placement, illustrating the peripheral position 9 
of the catheter and the area of distribution of injected the cells is shown in the bottom image. 10 
A magnified and contrast enhanced image is provided in the insert.  Black arrow, sheath; red 11 
arrow, catheter tip; white arrow, cholangiogram. (C) Ultrasound imaging of the injected area 12 
of the liver revealing no duct dilation or any other abnormality at the end of the experiment. 13 
(D) Immunofluorescence analysis demonstrating engraftment, expression of key biliary 14 
markers, loss of gallbladder markers, expression of intrahepatic markers and loss of markers 15 
of other lineages in human Red Fluorescent Protein (RFP) expressing cells following 16 
transplantation in NMP human livers. Scale bars, 50μm. The images are complementary to Fig. 17 
4. 18 
  19 
 63 
 1 
Fig. S16 2 
Engraftment of gallbladder organoids in human livers receiving Normothermic 3 
Perfusion (NMP). (A) Immunofluorescence analysis demonstrating engraftment of human 4 
Red Fluorescent Protein (RFP) expressing cells following transplantation in NMP human 5 
livers. Scale bars, 100μm. The images are complementary to Fig. 4, S15. (B) Quantification of 6 
gallbladder-derived RFP-expressing cells in injected vs. not injected human bile ducts; **** 7 
P<0.0001, Mann-Whitney test. The data corresponds to 3 different livers and 5 random sections 8 
per liver. Each section is represented by a data point, while each organ is represented by a 9 
different symbol. 10 
  11 
 64 
Table S1. 1 
Table of the number of animals at risk corresponding to the Kaplan-Meier curve in Fig. 3B. 2 
 3 
 4 
  Number of animals at risk 
Days Organoids Carrier 
0 5 5 
5 5 5 
8 5 4 
16 5 3 
17 5 2 
18 5 1 
59 5 0 
92 4 0 
 5 
  6 
 65 
Table S2. 1 
Table of antibodies used. 2 
 3 
Antibody Provider 
Catalogue 
number Dilution Species 
Anti-FGF19 Santa Cruz sc-390621 1:100 Mouse 
Anti-FGF19 Abcam ab225942 1:100 Rabbit 
Anti-TFF2 
R&D 
systems MAB4077 1:50 Mouse 
Anti-DCDC2 Santa Cruz sc-166051 1:100 Mouse 
Anti-human albumin 
R&D 
systems MAB1455 1:50 Mouse 
Anti-SOX4 Abcam ab86809 1:50 Rabbit 
Anti-SOX17 
R&D 
systems AF1924 1:100 Goat 
Anti-RFP Abcam ab62341 1:100 Rabbit 
Anti-RFP Rockland  200-101-379 1:200 Goat 
Anti-KRT19 DSHB TROMA III 1:100 Rat 
Anti-KRT19 Abcam ab7754 1:100 Mouse 
Anti-KRT19 Abcam ab52625 1:100 Rabbit 
Anti-KRT7 DAKO GA61961-2 1:100 Mouse 
Anti-KRT7 Abcam ab68459 1:100 Rabbit 
Anti-αSMA DAKO GA61161-2 1:100 Mouse 
HNF1B (c-20) 
SANTA 
CRUZ sc-7411 1:100 Goat 
GAMMA-GLUTAMYL TRANSPEPTIDASE 
(GGT) 
Abcam ab55138 
1:100 Mouse 
CYSTIC FIBROSIS TRANSMEMBRANE 
CONDUCTANCE REGULATOR (CFTR) 
SANTA 
CRUZ sc-10747 1:100 Rabbit 
ALEXA FLUOR DONKEY ANTI-Rabbit 568 A10042 INVITROGEN 1:1000 Donkey 
ALEXA FLUOR DONKEY ANTI-Rabbit 488 A21206 INVITROGEN 1:1000 Donkey 
ALEXA FLUOR DONKEY ANTI-Rabbit 647 A31573 INVITROGEN 1:1000 Donkey 
ALEXA FLUOR DONKEY ANTI-goat 568 A11057 INVITROGEN 1:1000 Donkey 
ALEXA FLUOR DONKEY ANTI-goat 488 A11055 INVITROGEN 1:1000 Donkey 
ALEXA FLUOR DONKEY ANTI-goat 647 A21447 INVITROGEN 1:1000 Donkey 
ALEXA FLUOR DONKEY ANTI-mouse 568 A10037 INVITROGEN 1:1000 Donkey 
ALEXA FLUOR DONKEY ANTI-mouse 488 A21202 INVITROGEN 1:1000 Donkey 
ALEXA FLUOR DONKEY ANTI-mouse 647 A31571 INVITROGEN 1:1000 Donkey 
 4 
  5 
 66 
Table S3 1 
Table of QPCR primers used.  2 
 3 
 4 
Gene  Primer sequence (5’ à 3’)  
HNF1B  F  TCACAGATACCAGCAGCATCAGT  
   R  GGGCATCACCAGGCTTGTA  
PBGD  F  GGAGCCATGTCTGGTAACGG  
 R  CCACGCGAATCACTCTCATCT  
SOX9 F CTCTGGAGACTTCTGAACGAGAG 
 R CCTTGAAGATGGCGTTGGGG 
CK19 F  ACGACCATCCAGGACCTGCGG 
 R  TCCCACTTGGCCCCTCAGCGTA 
CK7 F  GATTGCTGGCCTTCGGGGT 
 R  TCATCACAGAGATATTCACGGCTC 
GGT F  GTGAGAGCAGTTGGCTGTGC 
 R  GTTGAACTCTGCTGTGGGGC 
CFTR F  AGTTGCAGATGAGGTTGGGC 
 R  AAAGAGCTTCACCCTGTCGG 
SOX4 F AGCGACAAGATCCCTTTCATTC 
 R CGTTGCCGGACTTCACCTT 
TFF2 F CCCATAACAGGACGAACTGC 
 R GCACTGATCCGACTCTTGCT 
SOX17 F CGCACGGAATTTGAACAGTA 
 R GGATCAGGGACCTGTCACAC 
FGF19 F ATGCAGGGGCTGCTTCAGTA 
 R AGCCATCTGGGCGGATCT 
 5 
  6 
 67 
Movie S1. 1 
T1 weighted Magnetic Resonance Imaging (MRI) of a control mouse, receiving MDA followed 2 
by injection of carrier medium without organoids in the biliary tree.  3 
 4 
Movie S2. 5 
T2 weighted MRI/ Magnetic Resonance CholangioPancreatography (MRCP) of a control 6 
mouse receiving MDA followed by injection of carrier medium without organoids in the biliary 7 
tree demonstrating the presence of cholangiopathy. The MRCP sequence corresponds to the 8 
reconstructed MRCP image in Fig. 3C (not transplanted panel). 9 
 10 
Movie S3. 11 
T1 weighted Magnetic Resonance Imaging (MRI) of a mouse receiving MDA followed by 12 
injection of organoids in the biliary tree. The images were acquired 90 days after the injection 13 
of organoids demonstrating normal liver anatomy with no formation of tumors. 14 
 15 
Movie S4. 16 
T2 weighted MRI/ Magnetic Resonance CholangioPancreatography (MRCP) of a mouse 17 
receiving MDA followed by injection of organoids in the biliary tree demonstrating resolution 18 
of cholangiopathy. The MRCP sequence corresponds to the reconstructed MRCP image in Fig. 19 
3C (transplanted panel). 20 
 21 
Movie S5 22 
MRI-based 3D reconstruction of the biliary tree of a control mouse receiving MDA followed 23 
by injection of carrier medium without organoids in the biliary tree demonstrating the presence 24 
of cholangiopathy with loss of bile duct signal. The bile ducts were reconstructed from T2 25 
weighted MR images. 26 
 27 
Movie S6 28 
MRI-based 3D reconstruction of the biliary tree of a mouse receiving MDA followed by 29 
injection of organoids in the biliary tree demonstrating resolution of cholangiopathy. The bile 30 
ducts were reconstructed from T2 weighted MR images. 31 
 32 
Movie S7 33 
Z-stack of native and regenerated RFP-expressing bile ducts in the liver of an animal receiving 34 
MDA followed by injection of RFP-expressing human gallbladder organoids in the biliary tree. 35 
KRT19 is shown in green. RFP is shown in red. The movie is complementary to movies S8 36 
and S9. 37 
 38 
Movie S8 39 
3D reconstruction illustrating native and regenerated bile ducts in the liver of an animal 40 
receiving MDA followed by injection of RFP-expressing human gallbladder organoids in the 41 
biliary tree. Native ducts, KRT19 positive/ RFP negative; regenerated ducts, KRT19 positive/ 42 
RFP positive. The bile ducts were reconstructed from the RFP and KRT19 43 
immunofluorescence images used to generate movie S7. KRT19 is shown in green, RFP is 44 
shown in red. The movie is complementary to movies S7 and S9. 45 
 46 
Movie S9 47 
 68 
3D rendering illustrating native and regenerated bile ducts in the liver of an animal receiving 1 
MDA followed by injection of RFP-expressing human gallbladder organoids in the biliary tree. 2 
Native ducts, KRT19 positive/ RFP negative; regenerated ducts, KRT19 positive/ RFP 3 
positive.  The bile ducts were reconstructed from the RFP and KRT19 immunofluorescence 4 
images used to generate movie S7 and S8. KRT19 is shown in green, RFP is shown in red. The 5 
movie is complementary to movies S7 and S8. 6 
 7 
Data S1. (separate file) 8 
Table of differentially expressed genes between different regions of the biliary tree. IHD, 9 
Intrahepatic ducts; CBD, Common Bile Duct; GB, Gallbladder. The table corresponds to genes 10 
with a log2 fold change > 1 and an adjusted P value < 0.001. 11 
 12 
Data S2. (separate file) 13 
Table of differentially expressed genes in pseudotime in primary cholangiocytes with an 14 
adjusted P value<0.001. 15 
 16 
Data S3. (separate file) 17 
Table of differentially expressed genes upregulated in organoids or organoids treated with bile 18 
versus primary cholangiocytes. ORG, organoids; ORGT, Bile treated organoids. The table 19 
corresponds to genes with a log2 fold change > 1 and an adjusted P value < 0.001. 20 
 21