High level of polarized engraftment of porcine intrahepatic cholangiocyte organoids in decellularized liver scaffolds

Abstract In Europe alone, each year 5500 people require a life‐saving liver transplantation, but 18% die before receiving one due to the shortage of donor organs. Whole organ engineering, utilizing decellularized liver scaffolds repopulated with autologous cells, is an attractive alternative to increase the pool of available organs for transplantation. The development of this technology is hampered by a lack of a suitable large‐animal model representative of the human physiology and a reliable and continuous cell source. We have generated porcine intrahepatic cholangiocyte organoids from adult stem cells and demonstrate that these cultures remained stable over multiple passages whilst retaining the ability to differentiate into hepatocyte‐ and cholangiocyte‐like cells. Recellularization onto porcine scaffolds was efficient and the organoids homogeneously differentiated, even showing polarization. Our porcine intrahepatic cholangiocyte system, combined with porcine liver scaffold paves the way for developing whole liver engineering in a relevant large‐animal model.

liver-ECM maintains hepatocyte function, possibly due to the intact three-dimensional structure. 13,20 Thus, far recellularization of decellularized liver ECM with reimplantation has been performed in rodents, 21,22 providing proof-of-principle, but a larger animal model is needed in preclinical-studies. [23][24][25][26] Pigs present a very suitable model in this respect due to a physiology that is very close to human, 27 and decellularized porcine scaffolds can be repopulated at a low degree with rat/human liver organoids. 15,28,29 This low engraftment was possibly caused by the species differences between donor cells in the decellularized ECMscaffold, as ideally autologous cells should be used. 22 Even more importantly, porcine progenitor cells can help advance this preclinical research towards human patients as they represent a more human-relevant animal model. The lack of established porcine ICOs impeded this research so far.
Therefore, the aim of this study was to establish ICO culture from porcine adult stem cells and repopulate porcine decellularized liver scaffolds with these organoids. This would pave the way for whole organ bioengineering that can be used as an alternative for organ donation in the future.

| De-and recellularization
The process of porcine liver decelluarization and recellularization is schematically depicted in Figure 3A.

| Decellularization of porcine livers
Decellularization of porcine livers was performed according to the previously published protocol by Pla-Palacin et al. 31 In short, whole livers are dissected from pig cadavers and portal vein and hepatic artery are cannulated and subsequently perfused with alternating charges of distilled water and decellularization solution containing Triton X-100. The decellularized livers were then cut into circular discs of 5 mm diameter and 200 μm height.

| Recellularization of liver scaffolds
Before reseeding, the scaffolds were incubated for 1 h at 37°C with 200 μl EM with 25 ng/ml BMP7 in a humidified incubator at 37°C, 5% CO 2 and 21% oxygen. After the incubation the medium was removed, and the scaffolds were placed in an incubator at 37°C for 1 h to dry. Each scaffold was seeded with 200 000 cells as organoid fragments in 10 μl EM with the addition of 25 ng/ml BMP7 in a 96well plate. After 1 h incubation in an incubator at 37°C, 250 μl EM with 25 ng/ml BMP7 was added to each well. The EM was refreshed twice a day. After 2 days, the scaffolds were transferred to a 24 well plate and the medium was changed to DM. The DM was refreshed every 2-3 days until the end of differentiation at Day 5, which has proven to be the best timepoint based on preliminary experiments (data not shown).

| Metabolic activity
To test the metabolic activity of the organoids grown in expansion Blue assay was performed on all four donors in passage (p2) and the intensity was measured at Days 0, 2, 6 and 8 after plating.

TA B L E 1 Primer information for porcine specific qRT-PCR
marker Ki-67 for all conditions (EM, DM, discs); for tight junction protein 1 (ZO1) in DM and discs; and hepatocyte nuclear factor 4 alpha (HNF4A), keratin 18 (K18). For each antibody, the antigen retrieval method, antibody dilution and incubation times are summarized in Table 2. Slides were analysed using a Leica SPEII fluorescent microscope.

| Organoid expansion
The cells derived from porcine liver tissue samples were cultured in Matrigel™ and biliary duct fragments could be observed in the hydrogel scaffold (see Figure 1B). Over the next days of culture in EM media the formation of cyst-like spherical structures and their growth could be observed and their morphology stayed stable until at least passage 6 (see Figure 1B). Organoid growth was also shown by metabolic activity measurement (see Figure 1A), where activity significantly increased from day to day up to 451% compared with the start of culture (p < 0.0001). In Figure 1C, gene expression (values in  S1). In contrast, the organoids stained strongly for the ductular marker K19 and the proliferation marker Ki67 stained roughly 30 percent of the cells within the organoid showing continued expansion ( Figure 1D).

| Organoid differentiation
After differentiation of the porcine ICOs for 5 days, the grand mean of the albumin level had significantly increased by 71 mg/L (p < 0.0001) in DM compared with EM (see Figure 2A)

| Decellularized discs
Successful decellularization of livers was confirmed by the lack of nuclei and DNA present (see Figure 3C), and exemplary images of native versus decellularized liver scaffolds can be seen in Figure 3C as well. The organoids seeded onto these scaffolds showed lower expression of stemness markers LGR5 compared with EM control as well as lower expression of HNF1B (see Figure 3B). All hepatic markers were expressed by organoids seeded onto scaffolds. As seen by similar expression levels of HNF4A, ALB, TF, TTR, FAH and CYP3A22 between ICO's on scaffolds and in DM, maturation of the hepatocyte-like cells was at a comparable level.
The immunofluorescent stainings showed that the organoids infiltrated the scaffold with a high efficiency and proceeded to express several different markers ( Figure 3D). Albumin showed a positive staining in the vast majority of the cells seeded onto scaffolds. can be seen in Figure 3E.

| DISCUSS ION
In this study, we show that ICOs can be successfully grown and dif- Notably, immunofluorescent staining for K18 was very high and was observed in proximity of HNF4A positive cells, which further confirms the transformation towards hepatocyte-like cells. 40  fate. 37 The organoids had been fragmented before seeding onto the scaffold, but nevertheless proceeded to self-organize as is evident by the polarization seen in MDR1 and ZO1 stainings. 35 The formation of canaliculi as shown here by K19 and MDR1 positive staining has previously been unsuccessful when human foetal hepatocytes have been used to repopulate decellularized porcine scaffolds even after 13 days of perfusion time. 15 It is possible that this is caused by cross-species differences between scaffold and cell source.
Although differentiation was largely successful, the resulting cells do not exhibit the same characteristics as native hepatocytes or cholangiocytes as seen for example in the localized expression of HNF4A or the lack of liver transaminase production of some donors and further optimization/maturation is necessary. Additionally, animal derived components such as Matrigel™ and Wnt cannot be used in a clinical setting and must be replaced by alternatives such as other ECM-mimicking hydrogels (cellulose nanofibril, 34 gelPEG, 41 Gelatin-methacrylol or other hydrogels 42 ) and Wnt-surrogates. 43 Additionally, medium compositions could be further optimized in future studies. In the present study, proliferation of organoids in expansion medium is mediated by R-spondin that activates Wnt signalling and is therefore present in EM medium but absent in DM medium. 44 Here, the combination of hepatocyte growth factor and dexamethasone drives maturation towards hepatocyte-like cells, 45 as do FGF19 and BMP7. 46 Maturation could be enhanced by adding other compounds or ECM components during the differentiation such as Oncostatin M or fibronectin. 47,48 To increase both cholangiocyte and hepatocyte differentiation, addition of collagen type I to the culture system have been proven to be of advantage. 44,48 Maturation of bile duct structures is achieved in the current study by utilizing the Notch inhibitor DAPT in DM medium, but biliary differentiation could be increased by addition of laminin and collagen IV. 48 In this study, only small scaffold pieces were utilized to allow the cells to repopulate the ECM by diffusion which was very successful as seen by the engraftment percentage. However, in order to achieve complete population, it would be beneficial to use perfused liver scaffolds that make use of the native vascular network in the decellularized organ. 18,28,31 Thereby, it is of advantage to use fragmented organoids that are capable of self-organization

| CON CLUS ION
The aim of this work was to establish ICOs from porcine adult stem cells, which has been successful as they did not only show stable liver stem cell characteristics over multiple passages, but also differentiated into hepatocyte-and cholangiocytes-like cells. This differentiation was also possible on decellularized porcine liver scaffolds and the cells repopulated the liver discs, which has been very effective, and polarization even showed formation of canaliculi. This is an important step towards bioengineering whole liver tissue in a rel-  Figure 3A was created with BioRe nder.com.

FU N D I N G I N FO R M ATI O N
This work has received funding from the European Union's Horizon

research and innovation programme under the Marie
Sklodowska-Curie grant agreement No 64268.

CO N FLI C T O F I NTE R E S T
The author(s) declare no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

DATA AVA I L A B I L I T Y S TAT E M E N T
Data relevant to the study is included in the article. Any additional or detailed data can be requested by directly emailing the corresponding author.