Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Jan 1.
Published in final edited form as: Biotechnol Bioeng. 2020 Aug 28:10.1002/bit.27546. doi: 10.1002/bit.27546

Development of liver microtissues with functional biliary ductular network

Ehab O A Hafiz 1,2,3, Beyza Bulutoglu 1,2, Soheir S Mansy 3, Yibin Chen 1,2, Hoda Abu-Taleb 4, Somia A M Soliman 5, Ali A F El-Hindawi 5, Martin L Yarmush 1,2, Basak E Uygun 1,2
PMCID: PMC7775340  NIHMSID: NIHMS1632184  PMID: 32856740

Abstract

Liver tissue engineering aims to create transplantable liver grafts that can serve as substitutes for donor’s livers. One major challenge in creating a fully functional liver tissue has been to recreate the biliary drainage in an engineered liver construct through integration of bile canaliculi (BC) with the biliary ductular network that would enable the clearance of bile from the hepatocytes to the host duodenum. In this study, we show the formation of such a hepatic microtissue by coculturing rat primary hepatocytes with cholangiocytes and stromal cells. Our results indicate that within the spheroids, hepatocytes maintained viability and function for up to 7 days. Viable hepatocytes became polarized by forming BC with the presence of tight junctions. Morphologically, hepatocytes formed the core of the spheroids, while cholangiocytes resided at the periphery forming a monolayer microcysts and tubular structures extending outward. The spheroids were subsequently cultured in clusters to create a higher order ductular network resembling hepatic lobule. The cholangiocytes formed functional biliary ductular channels in between hepatic spheroids that were able to collect, transport, and secrete bile. Our results constitute the first step to recreate hepatic building blocks with biliary drainage for repopulating the whole liver scaffolds to be used as transplantable liver grafts.

Keywords: biliary ductular network, cholangiocytes, hepatocytes, liver tissue engineering, spheroids

1 |. INTRODUCTION

Liver diseases represent the eighth leading cause of mortality in the USA, whereby severe conditions, such as cirrhosis, cause over 1 million deaths per year worldwide (Everhart & Ruhl, 2009; Mohdad et al., 2014). The only available treatment of end-stage liver disease requires orthotopic liver transplantation which is limited by the severe shortage of high-quality donor organs. To address the gap in available donor organs, there have been advances in creating engineered liver grafts as donor liver substitutes using the whole organ engineering approach (Acun, Oganesyan, & Uygun, 2019; Uygun et al., 2010). Development of fully functional liver substitutes has proven to be complicated, as liver is one of the most complex organs, owing to its challenging functional and structural organization composed of different networks such as vasculature and biliary (Mazza, Al-Akkad, Rombouts, & Pinzani, 2018; Shafiee & Atala, 2017).

The repopulation of the intrahepatic bile ducts (IHBDs), which is a critical component of the nonparenchymal population, has been largely neglected in the engineered liver grafts until recently (Chen, Devalliere, Bulutoglu, Yarmush, & Uygun, 2019). IHBD constitute the biliary network formed by the cholangiocytes, or bile duct epithelial cells, whose primary role is to modify and transport the bile formed by the hepatocytes. Bile is a yellow–green basic solution containing bile pigments, cholesterol neutral fats, phospholipids, and various electrolytes. It emulsifies lipid during the digestion process (Hofmann, 1999). Hepatocytes secrete bile into the bile canaliculi (BC) which connect to bile ductules and ducts; and subsequently, bile is removed from the liver via the extrahepatic bile duct and stored in the gallbladder (Muriel, 2017). If bile is not directed into the bile ducts for removal, some of its components may become toxic (Fickert & Wagner, 2017); hence, it is critical that the bile removal should be addressed in an engineered transplantable liver graft for long-term efficacy and functionality.

One major challenge in recreating the biliary drainage in an engineered liver construct has been the integration of BC, formed in-between polarized hepatocytes in culture, with the biliary ductular network, that is formed by the growth and rearrangement of biliary epithelial cells, which subsequently enables the clearance of bile from the hepatocytes. In a three-dimensional (3D) environment, such as spheroid cultures, hepatocytes quickly assume cellular polarity with BC forming and extending into the interior of the spheroid and produce bile (Godoy et al., 2013; Sharma et al., 2019). However, in the absence of a functional ductular network, it is not possible to collect the bile produced by the hepatocytes. Moreover, the addition of nonparenchymal cells, have been shown to result in cellular organization and hepatic phenotype that closely resemble the in vivo microenvironment present in the liver (Bell et al., 2016). Primary cholangiocytes form 3D structures, such as cysts and ducts, when embedded in a 3D gel (Rizki-Safitri et al., 2018). When the cells are encapsulated in decellularized liver matrix, a complex network of ductules form spontaneously over a period of 7 days that show mature biliary functions (Lewis et al., 2018). A mix culture of rat hepatoblasts, biliary epithelial cells, and mouse embryonic fibroblasts embedded in collagen gel on a polydimethylsiloxane membrane yielded tubular biliary formations that were able to extract florescent dye secreted by hepatocytes (Rizki-Safitri, Shinohara, Tanaka, & Sakai, 2020). When cholangiocytes are mixed with hepatocytes in a 3D culture model, Katsuda, Kojima, Ochiya, and Sakai (2013) showed biliary ductular formation within spheroids of rat fetal liver cells and biliary epithelial cells. No study so far has demonstrated biliary flow out of the engineered liver tissues resembling in vivo microarchitecture that would allow the collection of the bile away from the hepatocytes.

In this study, we aim to develop hepatic microtissue building blocks that have the integrated BC and the biliary ductular network. We cocultured rat primary hepatocytes with cholangiocytes and stromal cells in a spheroid model as the building blocks that are clustered together to form a higher order structure. We examined the multicell type spheroids morphologically, and ultrastructurally; and demonstrated that the multispheroid clusters could be formed through the integration of biliary ductules. This study serves as the first step to recreate hepatic building blocks with biliary drainage that can be scaled up and used as transplantable liver grafts.

2 |. MATERIALS AND METHODS

2.1 |. Materials

Cell culture medium, medium supplements, and other reagents were obtained from Sigma-Aldrich, Thermo Fisher Scientific, and Electron Microscopy Science unless stated otherwise.

2.2 |. Cells

Freshly isolated primary rat hepatocytes were obtained from Cell Resource Core at Massachusetts General Hospital. The viability was over 90% as evaluated by Trypan Blue exclusion and the cells were used immediately after isolation for spheroid cultures. A parallel monolayer culture in collagen sandwich configuration was performed to evaluate the viability of the cells.

Cryopreserved normal rat cholangiocytes (NRC; courtesy of Nicholas Larusso Cholangiopathies Laboratory of Mayo Clinic) is a cell line established from intrahepatic bile ducts (IHBDs) (Vroman & LaRusso, 1996) and were grown in NRC medium; Dulbecco’s modified Eagle’s medium (DMEM)/nutrient mixture F-12 supplemented with 5% fetal bovine serum (FBS), 0.01 U/ml insulin, 25 ng/ml epidermal growth factor, 14 ng/ml glucagon, 0.393 μg/ml dexamethasone, 30 μg/ml bovine pituitary extract, 3.4 μg/ml 3′5-triiodo-l-thyronine, 4.11 μg/ml forskolin, 100 U/ml penicillin, and 100 μg/ml streptomycin. Cells were trypsinized after reaching 80% confluency and resuspended to be used in spheroid cultures.

Murine 3T3-J2 fibroblasts (courtesy of Howard Green Laboratories, Harvard Medical School) were maintained in T175 tissue culture flasks in DMEM (Gibco) supplemented with 10% FBS and 2% penicillin and streptomycin. After reaching confluence, 3T3-J2 fibroblasts were growth arrested by treatment with 12 μg/ml of mitomycin-C for 2.5 h. Subsequently, the cells were washed with phosphate-buffered saline (PBS), trypsinized, and used in spheroid cultures.

2.3 |. Spheroid preparation

Spheroids were generated using the liquid-overlay technique. Each well of a 96-well plate (Corning) was coated with 1.5% agarose, and 100 μl cell suspension was added per well. The plate containing the cell suspension was centrifuged at 100 g for 3 min to pellet the cells. The plate was then placed in 37°C and 5% CO2 for up to 3–4 days to allow for spontaneous self-aggregation. When the spheroids were sufficiently compact with no visible single cells, usually on Day 3 postseeding, 50% of the media was changed every other day.

Three types of spheroids were prepared containing (1) hepatocytes only (H), (2) hepatocytes with cholangiocytes (HC), or (3) hepatocytes with cholangiocytes and 3T3-J2 fibroblasts (HCF). Hepatocyte only spheroids were prepared using 1500, 750, or 500 viable cells. HC spheroids were prepared by mixing cholangiocytes and hepatocytes at 1:4 ratio for a total number of 1500 and 750 cells per spheroid. HCF spheroids were prepared using a cholangiocyte:-fibroblast:hepatocyte ratio of 1:2:2. The total number of cells was 500 cells per spheroid.

The cell culture medium was DMEM (high glucose 4.5 g/L) or Williams E medium (medium glucose 2 g/L) supplemented with 10% FBS (Gibco), 7 ng/ml glucagon (Bedford Laboratories), 7.5 μg/ml hydrocortisone (Pharmacia Corporation), 0.25 U/ml insulin (Eli Lilly), 20 ng/ml epidermal growth factor, 200 U/ml penicillin, and 200 μg/ml streptomycin (Gibco).

2.4 |. Assessment of spheroid size and viability

Spheroids were visualized daily by either bright field or phase-contrast microscopy (EVOS FL Cell Imaging System; Thermo Fisher Scientific) and images were captured. The size of spheroids was determined through geometric mean diameter measurements using ImageJ (NIH). The spheroids were referred to in text by the initial number of cultured cells per spheroid. The spheroid viability was determined using the Presto Blue assay which is a cell-permeable resazurin-based solution and is reduced to highly florescent resorufin by aerobic respiration of metabolically active cells. Viable cells continuously convert resazurin to resorufin increasing the overall fluorescence and the color of the surrounding media, thus indicating cell viability. After complete spheroid formation (3 days postseeding), all spheroids from one 96-well plate were collected and transferred to agarose coated well of a 48-well plate. The spheroids were incubated in 250 μl media containing 10% Presto-Blue Cell Viability Reagent (Invitrogen) for 90 min. The fluorescence of the media was measured in triplicate using Biotek Synergy 2 Multi-Detection Microplate Reader at 530/590 nm. The florescence measurement was normalized to the initial number of cells to be able to compare the different numbers used to form the spheroids.

2.5 |. Functional analysis of spheroids

The hepatic function of spheroids was determined through measurement of albumin production and urea secretion. Multiple spheroids were collected and incubated in the wells of a 48-well plate for up to 7 days. The media was sampled every other day. The albumin level was determined by a direct competitive Enzyme-Linked Immunosorbent Assay method, and the urea concentration was determined using a commercially available kit (Stanbio Laboratory). The results of both albumin and urea assays were normalized to the number of hepatocytes per day.

The BC formation in hepatocyte plate cultures and the biliary flow pathway within the spheroids were visually assessed using carboxy-(5-(and-6)-carboxy-2′,7′-dichlorofluorescein diacetate (CDFDA) and indocyanine green (ICG) dye, respectively. In hepatocyte plate cultures, media was aspirated followed by washing with PBS, then cells were incubated in 2 μM solution of CDFDA (Invitrogen and Thermo Fisher Scientific) at 37°C for 5 min. As nonfluorogenic CDFDA is internalized by hepatocytes, hydrolyzed by intracellular esterases, and turn into fluorogenic carboxyfluorescein and is excreted from the canalicular membrane into BC by active multidrug resistance-associated protein 2 transporter (MRP2; Kudo et al., 2004). All phase contrast and fluorescent images were taken using EVOS FL Cell Imaging System (Thermo Fisher Scientific).

Individual spheroids were allowed to form for 3 days after which they were collected in nonadherent round-bottom plates and further incubated in suspension for additional 3 days to form spheroid clusters. Spheroid clusters were incubated at 37°C for 40 min with 0.025 mg/ml ICG in PBS and 4′,6-diamidino-2-phenylindole (DAPI) nuclear stain during which they were imaged at 10-min intervals for up to 40 min. All fluorescent confocal images were acquired using Nikon A1R confocal microscope system. Images were processed with Nikon NIS-Elements Advanced Research and ImageJ software.

2.6 |. Histology and immunofluorescence

The cellular arrangement within the spheroids was assessed via histology and immunohistology using agarose cell block technique (Mansy, 2004). The spheroids were fixed in 4% paraformaldehyde (Sigma-Aldrich) for 1 h and then suspended in melted agarose. Fixed spheroids in agarose were centrifuged at 1500 rpm for 7 min and agarose containing the spheroid pellet was allowed to cool for 30 min at 4°C. The solidified agarose block was then cut at the tip where the spheroids were held together and fixed one more time in buffered 4% paraformaldehyde for 2 h. The samples were processed in tissue processor (Tissue Tek VIP-2), then embedded in the paraffin block. Four-micron sections were cut and stained with hematoxylin and eosin (H&E).

Immunofluorescent staining was performed for HCF spheroids. The spheroids were fixed, paraffin-embedded, and sectioned as described above. Paraffin-embedded sections were incubated at 60°C for 15 min, deparaffinized by incubation in xylene, and then rehydrated by serial incubation in descending grades of ethanol (100%, 95%, 70%, and 50%). Sections were rinsed in deionized water and stored in PBS followed by antigen retrieval in citrate buffer at 90°C. Cooled slides were rinsed in PBS and the sections were permeabilized for 5 min in 0.2% Triton-X 100 followed by blocking with 1% bovine serum albumin (Sigma-Aldrich). The sections were incubated with the primary antibodies at 1:200 overnight at 4°C in a humidified chamber and Alexa Fluor conjugated secondary antibodies at 1:500 in the dark for 1 h at room temperature in humidified chamber. The following primary antibodies (Abcam) were used: Chicken polyclonal anti-albumin (ab106582) as the hepatocyte marker, mouse monoclonal anti-cytokeratin 7 (ab9021) as the cholangiocyte marker, and rabbit monoclonal anti-vimentin (ab92547) as the fibroblast marker. The corresponding secondary antibodies (Abcam) were goat anti-chicken immunoglobulin Y (IgY) H&L (Alexa Fluor 488; ab150169), goat anti-mouse IgG H&L (Alexa Fluor 647; ab150115), and goat anti-rabbit IgG H&L (Alexa Fluor 568; ab175471), respectively. Slides were counter-stained and mounted with Fluoroshield mounting medium containing DAPI (Abcam). We tested hepatocytes and the other cell types in plate cultures with all the target antibody candidates to ensure the selected target antibody is appropriate, highly specific, and does not cross label control cells. As a negative control, cultured cells were tested as described but with the primary antibody omitted.

2.7 |. Scanning and transmission electron microscopy imaging

Spheroids formed in concave microwells were fixed in buffered 4% glutaraldehyde for 2 h at 4°C, and gently washed twice for 1 h each in an equal volume mixture of 0.2 M sodium cacodylate and 0.4 M sucrose at 4°C. For secondary fixation, spheroids were immersed in 2% osmium tetroxide for 1 h at 4°C, washed with distilled water. The fixed spheroids were subsequently dehydrated in ascending grades of ethyl alcohol (30%, 50%, 70%, 90%, and 100%), each for 5 min except for 100% ethanol which was repeated three times for 10 min each at room temperature. The spheroids were examined using scanning electron microscopy (SEM; Inspect S; FEI Company) provided by Electron Microscopy Laboratory at TBRI.

Transmission electron microscopy (TEM) was conducted to confirm differentiating cell types in HC and HCF spheroids at the MGH TEM core supported by NIH/NINDS P30NS045776. The spheroids were prepared as described above with further embedding in Epon capsule, sectioning with ultramicrotomy (Leica Ultracut R) at 70 A° thickness and double staining with uranyl acetate and lead citrate. Contrasted sections were examined by TEM (JEOL JEM-1011 with AMTv601 software).

2.8 |. Statistical analysis

At least four cultures from different hepatocyte isolations were evaluated. All statistical analyses were performed using SPSS software for Windows, version 24.0. Numerical values were expressed as the mean ± standard error of four independent experiments except where otherwise noted. Statistical comparisons between the three groups were determined by one-way analysis of variance. The Bonferroni post hoc test was used for multiple comparisons with the appropriated H 750/H 500. Numerical values of probability smaller than .05 were considered as rejection of the null hypothesis and statistically significant. Statistically significant differences in viability were defined on the basis of H 750 two-tailed Mann–Whitney (U-tests between) using p smaller than .05 as a threshold.

3 |. RESULTS

3.1 |. Characterization of HC spheroids

Before formation of the spheroids, we evaluated the proliferative and functional capacity of cholangiocytes and hepatocytes, separately. We showed that NRC proliferated in culture both in 2D and 3D and were positive for cytokeratin 7 (Figure S1). In 3D, the cells formed aggregates and developed cystic structures over a period of 7 days; however, these structures were not robust and were prone to disintegration. We also show that primary rat hepatocytes formed cuboidal shaped monolayers in collagen sandwich cultures, express albumin, and were highly polarized confirmed by the presence of BC (Figure S2).

Individual spheroids of primary rat hepatocytes and NRC were allowed to form for 72 h in the wells of a 96-well plate (Figure 1). We tested two different spheroid sizes by varying the initial cell number to determine the optimal size for highest hepatocyte viability and function over a period of 7 days. H spheroids continued to decrease in size reaching a diameter of 247.8 ± 3.1 μm and 274.4 ± 4.0 μm by Day 7 for spheroids that initially had 750 and 1500 hepatocytes, respectively (Figure 2a). On the other hand, the size of HC spheroids did not change over the 7-day culture period, diameters plateauing at 284.0 ± 1.1 μm and 321.2 ± 5.8 μm for spheroids that initially had 750 and 1500 cells, respectively (Figure 2a). The results of the Presto Blue assay showed that spheroid viability increased with time for both H and HC spheroids (p < .001). However, this increase was the highest for spheroids made with 750 cells. The measured viability on a given day in culture was consistently higher in HC spheroids than in H spheroids although not statistically significant except for spheroids with 750 cells on Day 3 (Figure 2b).

FIGURE 1.

FIGURE 1

Open in a new tab

Schematic representation of the spheroid formation. H, hepatocytes only; HC, hepatocytes with cholangiocytes; HCF, hepatocytes with cholangiocytes and fibroblasts

FIGURE 2.

FIGURE 2

Open in a new tab

Morphology and physical characterization of hepatocyte (H) and hepatocyte–cholangiocyte (HC) spheroids. (a) Diameter and (b) viability of spheroids as a function of initial cell number and culture time. (c) Microscopic appearance of spheroids with 750 cells on Days 3, 5, and 7 (left: H, right: HC spheroids). (d) Scanning electron micrographs of H (left) and HC (right) spheroids with 750 cells on Day 5. (e) Histology (hematoxylin and eosin) of spheroids with 750 cells on Day 5, arrowhead points hydropic change of hepatocytes, and asterisks indicate cystic and tubular structures formed by cholangiocytes. Scale bars: (c) 200 μm, (d) 100 μm, and (e) 100 μm (top) and 50 μm (bottom), respectively. *p < .01, #p < .001

Morphologically, H spheroids became more compact with a smooth outer surface as a function of culture time observed by phase-contrast microscopy and SEM. Interestingly, HC spheroids were surrounded by cystic-like buds outgrowing from the surface of the spheroid possibly with an epithelial lining (Figure 2c). SEM imaging revealed that HC spheroids were decorated with tube-like buds in addition to cystic growths (Figure 2d, right). Further investigation showed that tubular growths on the surface of the spheroids had clear openings with channel-like structures (Figure S3). Histological analysis of HC spheroids with 750 cells on culture Day 5 showed that cholangiocytes form microcysts and tubular structures at the periphery of the spheroids (Figure 2e, asterisks), while hepatocytes populate the center of the spheroids. The hepatocyte rich center in H and HC spheroids indicated ballooning and hydropic degeneration as well as scattered apoptotic bodies and pyknotic nuclei (Figure 2e).

We assessed the hepatic function of spheroids by measuring albumin and urea secreted into the media. While there were detectable levels of albumin secreted by H and HC spheroids over the 7-day culture period, the differences based on the cell number and the presence of cholangiocytes were not statistically significant except on Day 7 when HC spheroids with 750 cells showed a significant increase over H spheroids with 750 cells (p < .01), indicating suboptimal conditions for hepatocyte viability and function (Figure S4a). There was significantly higher urea secretion by spheroids containing 750 cells than by those containing 1500 cells (p < .001). HC spheroids with 750 cells produced 1.5-fold urea on Day 7 compared to H spheroids with 750 cells (Figure S4b).

TEM analysis revealed intercellular interactions within the HC spheroids. Viable hepatocytes within the HC spheroids were found to become polarized by forming BC (yellow arrows) with the presence of intercellular spaces (Figure 3a). Neighboring cholangiocytes were found to form tight junctions securing continuous intercellular spaces decorated with microvilli and extending to the surface of the spheroid (Figure 3b). They also formed interdigitating membranous connections with the adjacent cells bounding larger bile collecting space in between (Figure 3c, boxed inset). HC spheroids displayed canalicular ductular junctions that are bound by hepatocyte on one side and biliary cell at the other side (Figure 3d). Cholangiocytes were detected to form small ductular spaces (D) that harbored a membrane-bound axoneme, characteristic of the primary cilium, and were composed of microtubules that measured 190 nm on average diameter (Figure 3d, inset). Tight junctions were seen securing adjacent cells together (yellow arrowheads) and the luminal spaces that have formed between the cells show microvilli (red arrows). Hepatocytes were found to have cytoplasmic appendices with lipid vacuoles possibly to be excreted into the intercellular space (bile collecting space) which was bound by a hepatocyte on one side and a biliary cell at the other side (Figure 3d, yellow arrow). However, the presence of fat droplets also indicated cellular stress (Figure 3d), consistent with histology and functional analysis results. We also confirmed ballooning and hydropic degeneration in HC spheroids in electron microscopy (EM) images of the inner hepatocytes. While the cytosol had abundant lipid vacuoles, rough endoplasmic reticulum, and mitochondria, many mitochondria showed various stages of degenerative changes. Some showed condensation figures and matrix deposition of electron-dense precipitates. Others were markedly swollen (ballooned) and showed structural damage with fragmented and disoriented cristae and rupture of their membrane with liberated cristae in the cytoplasm in the form of small vesicles (Figure S5).

FIGURE 3.

FIGURE 3

Open in a new tab

Ultrastructural evaluation of hepatocyte–cholangiocyte (HC) spheroids on Day 5. Transmission electron micrographs indicating intercellular interactions within the spheroids. (a) Bile canalicular spaces (yellow arrows) form between adjacent hepatocytes with intercellular spaces (red arrows). (b) The tangential cut shows continuous intercellular spaces (red arrows) between cholangiocytes reaching from the center to the surface of the spheroid. Tight junctions are seen securing adjacent cells together (yellow arrowhead). Magnification: ×10,000. (c) Tongue-like extensions projected by cholangiocytes to form cystic structures around the HC spheroids contain numerous microvilli and showing interdigitating membranous connections between adjacent cholangiocytes (red box) (magnification: ×10,000 to ×25,000). (d) A widefield view of a canalicular ductular junction on the surface of the spheroid. Hepatocyte shows cytoplasmic subapical appendix with lipoid vacuole (F) possibly to be excreted into the intercellular space (collecting bile space) which is bound by hepatocyte on one side and biliary cell at the other side. The luminal spaces that have formed between the cells show microvilli (red arrows). Two small ductular spaces (D) are bound by cholangiocyte extensions contain small primary cilium and a few microvilli (inset). Magnification: ×10,000, ×15,000 (inset). Yellow arrowheads: tight junctions, Red arrows: microvilli; yellow arrows: bile collecting space. C, cholangiocyte; F, fat droplet; H, hepatocyte

3.2 |. Characterization of HCF spheroids

In an effort to improve hepatocyte viability and metabolic activity of the cells within the spheroids, we included growth-arrested 3T3-J2 fibroblasts and generated spheroids that contain three cell types: hepatocytes, cholangiocytes, and fibroblasts at a ratio of 2:1:2 (Figure 4a). The initial number of cells incorporated within the spheroids was 500. The size of the HCF spheroids was around 200 μm in diameter, similar to H spheroids beyond Day 5. HCF spheroids had consistently higher levels of daily metabolic activity indicating higher viability than hepatocyte only spheroids over a 7-day culture period (p < .05 for albumin and p < .001 for urea; Figure 4b, Figure S4a,b).

FIGURE 4.

FIGURE 4

Open in a new tab

Hepatocyte, cholangiocyte, and fibroblast (HCF) containing spheroids. (a) Microscopic appearance of spheroids on Days 3, 5, and 7 (left: hepatocyte only [H], right: HCF spheroids). (b) Size and viability as a function of culture time compared to H spheroids. (a) Scale bar = 200 μm. *p < 0.001, #p < .01, †p < .05

H&E staining of the HCF spheroids on Day 5 showed that the hepatocytes within the core had healthy morphology with well-spread nuclei and cobblestone architecture. The spheroids showed cystic or tubular formations which continually extended towards the periphery (Figure 5a, asterisk). Unlike HC spheroids, the appearance of the cystic or tubular spaces on the periphery of the HCF spheroids is delayed a few days which is attributed to the lower number of initially cultured cholangiocytes per spheroid. Therefore, it takes a longer time for HCF spheroid to develop full-blown cystic growths (Figure S6). The multicellular architecture of the spheroids was further examined by immunostaining. These spaces were enclosed mostly by cholangiocytes such that CK7 positive cholangiocytes lined the peripheral wall of the formed spaces within the spheroid (Figure 5b). Vimentin positive 3T3-J2 fibroblasts were detected both in the periphery as well as within the parenchymal cell core of the spheroids (Figure 5b). TEM images of HCF spheroids showed successful integration of fibroblasts within spheroids alongside hepatocytes and cholangiocytes (Figure 5c). Cholangiocytes around the periphery of the spheroid formed tubular structures by extending the tongue plasma membrane with numerous microvilli similar to those observed in HC spheroids (Figure 5d). Hepatocytes within the center of the spheroids appeared healthy and polarized, forming BC bound by tight junctions in close proximity of bile collecting spaces formed by cholangiocytes (Figure 5e). A more detailed image revealed another canalicular space bound by multiple tight junctions within the center of the spheroid (Figure 5f).

FIGURE 5.

FIGURE 5

Open in a new tab

Histological and ultrastructural evaluation of hepatocyte (H), cholangiocyte (C), and fibroblast (F) spheroids. (a) Histology (hematoxylin and eosin [H&E]) and (b) immunostaining for albumin (green), cytokeratin 7 (red), vimentin (yellow), and nuclei (blue) imaged via confocal florescence. Transmission electron micrographs of (c) a section that shows all three cell types well integrated within the spheroid; the red arrows indicate nuclei for the three cell types H, C, and F (×5000) and (d) a cut section of a tubular structure (asterisk) formed by a cholangiocyte at spheroid periphery showing numerous microvilli along the outer cell membrane. The arrowhead points to the cholangiocyte nucleus (×6000). (e) A section from the center of the spheroid that shows bile canaliculi bound by tight junctions (yellow arrows) in close vicinity of a bile collecting space (blue arrow) (×6000). (f) Canalicular space formed between adjacent cells and showing multiple tight junctions (yellow arrowheads) (×40,000). Morphological analyses were done on Day 5. Scale bars (a) 100 μm and (b) 50 μm, respectively. DAPI, 4′,6-diamidino-2-phenylindole

3.3 |. Biliary flow pathway in HCF spheroids

We tested the biliary transport function of the HCF spheroids by monitoring the ICG uptake and release by the spheroids. To achieve this, HCF spheroids with 500 cells initially were collected in round-bottomed wells and cultured together for an additional 3 days (Figure 6a). During the additional culture period of multiple HCF spheroids, we observed cord-like structures growing at the periphery of some spheroids fusing with the adjacent spheroids within spheroid clusters (Figure 6b). These cord-like structures resembled tubular growths that displayed multiple side openings (fenestrations; Figure S7) which indicate that the periphery of the spheroids was populated with cholangiocytes.

FIGURE 6.

FIGURE 6

Open in a new tab

Biliary flow in HCF spheroid clusters. (a) Schematics of HCF spheroid cluster formation. (b) Phase-contrast images of HCF clusters on Days 0 and 3 of culture. Red arrowhead indicates duct-like extension between the spheroids. (c) Time-lapse images, red arrows point to exit sites of the ICG. (d) Duct-like connections within the spheroid cluster. Light upright microscopic (left) and confocal microscopy (right) images of indocyanine green transport in HCF spheroids. ICG fluorescence detected inside a channel (white arrow) connecting two HCF spheroids. Scale bars (c) 1 unit = 43.45 μm and (d) 200 and 100 μm (magnification ×20). DAPI, 4′,6-diamidino-2-phenylindole; HCF, hepatocyte cholangiocyte, and fibroblast; ICG, indocyanine green

The spheroid clusters were incubated with ICG to visualize the biliary flow pathway and continuously monitored via confocal fluorescence microscopy for up to 45 min after which the fluorescence reaches a plateau. Time-lapse imaging revealed rapid uptake of ICG by hepatocytes in the core of the spheroids by 10 min. The cells on the periphery of the spheroids remained negative for green ICG fluorescence. These cells were likely cholangiocytes, not hepatocytes based on histology and EM analysis, and did not take up ICG. By 30 min, ICG fluorescence was detected outside the clusters at specific sites pointing to its release through fenestrations (Figure 6c). The accumulation sites outside the clusters are marked with red arrows (Figure 6c). A bridge connecting two individual spheroids also showed ICG florescence indicating that it has a channel-like structure lined with cells and is able to collect ICG transported through the biliary flow pathway (Figure 6d).

4 |. DISCUSSION

There is an increasing demand for donor organs suitable for liver transplantation. To address the donor organ shortage, temporary procedures are developed to bridge the patients until transplantation, such as extracorporeal bioartificial livers, which still cannot compensate for the significant decline in the patients’ quality of life (Kumar, Tripathi, & Jain, 2011). Another approach is hepatocyte transplantation, which is restricted by the low liver-engraftment rate and low survival rate of the transplanted hepatocytes (Ibars et al., 2016). Thus, the ultimate solution to the donor shortage is engineering an equivalent liver tissue graft that can be transplanted into patients. Recent advances in whole liver engineering hold the promise to create transplantable liver grafts by employing the native organ’s extracellular matrix (ECM) as the template. In the whole liver engineering approach, the decellularized organ matrix is repopulated with healthy liver cells such as primary or stem-cell-derived hepatocytes and nonparenchymal cells, especially with endothelial cells (Acun et al., 2019; Wang et al., 2017). The repopulation of the IHBDs, which is a critical component of the nonparenchymal population, has not been addressed in the engineered liver grafts until recently (Chen et al., 2019). The integration of BC to the bile ducts within engineered liver tissues is yet to be shown. In this work, we studied the formation of hepatic microtissues by coculturing rat primary hepatocytes with cholangiocytes and stromal cells in a spheroid model and demonstrated that a functional biliary drainage network could be engineered in multicell-type spheroid clusters (Figures 1 and 6).

We have prepared two different sizes of HC spheroids by using two different initial number of cells. As expected, we found that the viability of spheroids made using 750 cells was higher than that of those made using 1500 cells, most likely due to higher oxygen gradients that could be achieved over smaller diameters. Compared to spheroids composed of H, HC spheroids also demonstrated higher viability as shown in Figure 2b. However, the increase in viability over time was observed both in H and HC spheroids, and, hence, it is unlikely that the viability increase in HC spheroids is solely due to the cholangiocytes. Unlike H spheroids, HC spheroids also exhibited channel-like structures on their surface indicative of biliary structure formation by the cholangiocytes (Figure 2d,e). Ultrastructurally, we have shown that ductular canalicular junctions developed within the spheroids. However, morphologically, hepatocytes showed characteristics of cellular stress such as ballooning and pyknotic nuclei even in spheroids with smaller sizes (Figure 2e, Figure S4).

To improve the morphology and the viability of the hepatocytes within the spheroids, fibroblasts were incorporated within the spheroids (HCF) and the spheroid size was reduced to 200 μm. Fibroblasts constitute the connective tissue and secrete the ECM proteins, and lay down the structural framework of tissues (Alberts et al., 2002; Williams, 1998). It was previously shown that primary hepatocytes cultured on growth-arrested 3T3-J2 fibroblast-feeder layers become polarized with stable morphology for 5 weeks of culture (Cho, Berthiaume, Tilles, & Yarmush, 2008; Cho et al., 2010). Addition of the growth-arrested 3T3-J2 fibroblasts into the HC spheroids significantly increased the viability of the spheroids as evident by higher metabolic activity levels compared to H spheroids demonstrated in Figure 4b and improved the morphology of the hepatocytes compared to HC spheroids, whereby fat accumulation was observed in the latter indicative of cellular stress (Figure 3; Masarone et al., 2018; Parafati, Kirby, Khorasanizadeh, Rastinejad, & Malany, 2018). Cholangiocyte proliferation and morphogenesis are influenced by the presence of the fibroblasts most likely due to the ECM that is secreted by the stromal cells in the engineered micro-tissues. We have observed that cholangiocyte only spheroids cultures result in aggregates with tubular and cystic structures that are fragile and prone to disintegration. In HCF spheroids, the 3D structures formed by the cholangiocytes were robust, strengthened by paracrine signaling by mature hepatocytes (Auth, 2001; Raynaud, Carpentier, Antoniou, & Lemaigre, 2011), and by ECM signaling by fibroblasts which induce and support the formation of tubular-like networks by cholangiocytes (Chiarini, Takiya, Borojevic, & Monteiro, 2006; Hashimoto et al., 2008).

Morphological examination of the spheroids showed that in HC and HCF spheroids, hepatocytes constitute the core and cholangiocytes line the periphery of the spheroids (Figures 3 and 5). This is consistent with findings in literature such that when hepatic spheroids have been prepared by mixing the hepatocytes with nonparenchymal cells, stellate and Kupffer cells were detected within the spheroid while cholangiocytes were located on the periphery of the spheroid (Bell et al., 2016). This spontaneous reorganization of the cells is most likely as a consequence of adhesion-dependent cell sorting in long-term culture (Dunn, Yarmush, Koebe, & Tompkins, 1989). The spheroid model provides a 3D environment that resembles the in vivo 3D liver structure to recreate hepatocyte morphology, cell–cell interactions, BC formation, and enhance hepatic functions such as activation of the bile acid biosynthesis pathway (Bell et al., 2016). In this study, ultrastructural evaluation of the spheroids revealed that hepatocytes formed BC bound by tight junctions, and cholangiocytes formed ductular spaces decorated with numerous microvilli around the spheroids similar to in vivo microarchitecture (Figures 3 and 5c–f). In addition, vimentin staining of HCF spheroids showed successful incorporation of fibroblasts as well (Figure 5b).

After we observed that cholangiocytes formed cysts and tubular structures on the periphery of the spheroids, we allowed the HCF spheroids to grow in close proximity. These tubular structures extended further and made connections with neighboring spheroids to form interconnected bile duct networks with an open lumen that was able to collect, transport, and accumulate ICG fluorescent probe (Figure 6), indicating the presence of BC and ductular canalicular junctions. ICG cholangiography is a commonly used practice to assess biliary duct perfusion and determine the hepatic biliary function in vivo (Ishizawa et al., 2009). ICG is taken up by hepatocytes via the transporter organic anion transport protein, which is exclusively expressed in the basolateral membrane of hepatocytes, and then excreted into the BC by MRP2 (Cusin et al., 2016; Ho et al., 2012). ICG in the liver can then be detected in the biliary ducts enabling visual detection of the biliary flow network, a technique that we took advantage of in this study. Similarly, Katsuda et al. (2013) demonstrated the formation of functional biliary ducts within the spheroids containing fetal hepatoblasts on a single spheroid scale using the fluorescent diacetate dye. To our knowledge, this study presents the first example of an in vitro microliver tissue platform composed of three different cell types, hepatocytes, cholangiocytes, and fibroblasts, demonstrating a functional biliary network interconnected with the BC structure in spheroid clusters.

Our long-term goal is to perform in vivo implantation of the microtissues to evaluate their ultimate functional performance. The work reported here represents the primary step to recreate hepatic building blocks that incorporate cellular architecture enabling biliary drainage. Therefore, we focused our efforts on the integration of the BC between hepatocytes with the biliary ductules formed by the cholangiocytes in the clusters of spheroids. To this end, we prepared structures that are around 500 μm in diameter through clustering of individual spheroids and analyzed them morphologically and ultrastructural while verifying the viability and the functionality of the biliary network in vitro. To test the functional capacity of the building blocks and their higher-order structures in vivo, we plan to create much larger structures (>1 cm) which would be more amenable to implantation, but at the same time would present additional challenges. One of the major issues we anticipate in engineering large-scale tissue structures is the lack of oxygen and other nutrients available for the cells seeded in the core of these structures which would potentially be resolved by recreating a vascular network within the cluster for blood circulation. Unfortunately, this is a major effort and is the topic of our current studies. Furthermore, pharmacological studies to evaluate the biliary content and permeability of the bile ducts (Du et al., 2020) would provide further evidence for the functionality of the bile ducts formed in between the spheroids. To our knowledge, this study presents the first example of an in vitro microliver tissue platform composed of three different cell types, hepatocytes, cholangiocytes, and fibroblasts, demonstrating a functional biliary network interconnected with the BC structure in spheroid clusters.

5 |. CONCLUSION

In this study, we report the formation of hepatic microtissues by coculturing rat primary HC and stromal cells in a spheroid model. We demonstrated that the multispheroid clusters could be formed through integration of biliary ductules and the biliary network that is a functional biliary flow pathway for the collection and transportation of bile. This study is the first step to recreate hepatic building blocks with biliary drainage to be used as transplantable liver grafts. Future work will involve the incorporation of other nonparenchymal cells, such as endothelial and stellate, within the spheroids for complete hepatic architecture and sustenance of viability of large tissue masses.

Supplementary Material

supplemental figures

ACKNOWLEDGMENTS

This study was supported by the National Institutes of Health (R01DK084053 Martin L. Yarmush and Basak E. Uygun), Shriners Hospitals for Children (Beyza Bulutoglu), and the Shriners Hospitals for Children in Boston Genomics and Proteomics and Translational Regenerative Medicine Special Shared Facilities. The Egyptian Ministry of Higher Education and Scientific Research supported this study through a joint supervision program (Ehab O. A. Hafiz). The authors would like to thank Dr. Nicholas F. LaRusso for the generous gift of Normal Rat Cholangiocytes.

Funding information

National Institute of Diabetes and Digestive and Kidney Diseases, Grant/Award Number: R01DK084053; Shriners Hospitals for Children, Grant/Award Number: 84311; Egyptian Ministry of Higher Education and Scientific Research

Footnotes

CONFLICT OF INTEREST

Basak E. Uygun has a financial interest in Organ Solutions, LLC, that is reviewed and arranged by MGH and Partners HealthCare in accordance with their conflict of interest policies. Other authors declare that there are no conflict of interests.

SUPPORTING INFORMATION

Additional supporting information may be found online in the Supporting Information section.

REFERENCES

  1. Acun A, Oganesyan R, & Uygun BE (2019). Liver bioengineering: Promise, pitfalls, and hurdles to overcome. Current Transplantation Reports, 6(2), 119–126. 10.1007/s40472-019-00236-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Alberts B, Johnson A, Lewis J, Raff M, Roberts K, & Walter P (2002). Fibroblasts and their transformations: The connective-tissue cell, Molecular biology of the cell (4th ed.). New York, NY: Garland Science. [Google Scholar]
  3. Auth M (2001). Morphogenesis of primary human biliary epithelial cells: Induction in high-density culture or by coculture with autologous human hepatocytes. Hepatology, 33(3), 519–529. 10.1053/jhep.2001.22703 [DOI] [PubMed] [Google Scholar]
  4. Bell CC, Hendriks DFG, Moro SML, Ellis E, Walsh J, Renblom A, … Ingelman-Sundberg M (2016). Characterization of primary human hepatocyte spheroids as a model system for drug-induced liver injury, liver function and disease. Scientific Reports, 6, 25187 10.1038/srep25187 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Chen Y, Devalliere J, Bulutoglu B, Yarmush ML, & Uygun BE (2019). Repopulation of intrahepatic bile ducts in engineered rat liver grafts. Technology, 7(1–2), 46–55. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Chiarini LB, Takiya CM, Borojevic R, Monteiro ANA (2006). Long-term culture of cholangiocytes from liver fibro-granulomatous lesions. BMC Gastroenterology, 6(1). 10.1186/1471-230x-6-13 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Cho CH, Berthiaume F, Tilles AW, & Yarmush ML (2008). A new technique for primary hepatocyte expansion in vitro. Biotechnology and Bioengineering, 101(2), 345–356. 10.1002/bit.21911 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Cho CH, Park J, Tilles AW, Berthiaume F, Toner M, & Yarmush ML (2010). Layered patterning of hepatocytes in co-culture systems using microfabricated stencils. Biotechniques, 48(1), 47–52. 10.2144/000113317 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Du Y, Khandekar G, Llewellyn J, Polacheck W, Chen CS, & Wells RG (2020). A bile duct-on-a-chip with organ-level functions. Hepatology, 71(4), 1350–1363. 10.1002/hep.30918 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Dunn JC, Yarmush ML, Koebe HG, & Tompkins RG (1989). Hepatocyte function and extracellular matrix geometry: long-term culture in a sandwich configuration. FASEB Journal, 3(2), 174–177. [DOI] [PubMed] [Google Scholar]
  11. Everhart JE, & Ruhl CE (2009). Burden of digestive diseases in the United States Part III: Liver, biliary tract, and pancreas. Gastroenterology, 136(4), 1134–1144. 10.1053/j.gastro.2009.02.038 [DOI] [PubMed] [Google Scholar]
  12. Fickert P, & Wagner M (2017). Biliary bile acids in hepatobiliary injury—What is the link? Journal of Hepatology, 67(3), 619–631. 10.1016/j.jhep.2017.04.026 [DOI] [PubMed] [Google Scholar]
  13. Godoy P, Hewitt NJ, Albrecht U, Andersen ME, Ansari N, Bhattacharya S, … Hengstler JG (2013). Recent advances in 2D and 3D in vitro systems using primary hepatocytes, alternative hepatocyte sources and nonparenchymal liver cells and their use in investigating mechanisms of hepatotoxicity, cell signaling and ADME. Archives of Toxicology, 87(8), 1315–1530. 10.1007/s00204-013-1078-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Hashimoto W, Sudo R, Fukasawa K, Ikeda M, Mitaka T, Tanishita K (2008). Ductular Network Formation by Rat Biliary Epithelial Cells in the Dynamical Culture with Collagen Gel and Dimethylsulfoxide Stimulation. The American Journal of Pathology, 173(2), 494–506. 10.2353/ajpath.2008.071024 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Hofmann AF (1999). The continuing importance of bile acids in liver and intestinal disease. Archives of Internal Medicine, 159(22), 2647–2658. 10.1001/archinte.159.22.2647 [DOI] [PubMed] [Google Scholar]
  16. Ibars EP, Cortes M, Tolosa L, Gómez-Lechón MJ, López S, Castell JV, & Mir J (2016). Hepatocyte transplantation program: Lessons learned and future strategies. World Journal of Gastroenterology, 22(2), 874–886. 10.3748/wjg.v22.i2.874 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Ishizawa T, Tamura S, Masuda K, Aoki T, Hasegawa K, Imamura H, …Kokudo N (2009). Intraoperative fluorescent cholangiography using indocyanine green: A biliary road map for safe surgery. Journal of the American College of Surgeons, 208(1), e1–e4. 10.1016/j.jamcollsurg.2008.09.024 [DOI] [PubMed] [Google Scholar]
  18. Katsuda T, Kojima N, Ochiya T, & Sakai Y (2013). Biliary epithelial cells play an essential role in the reconstruction of hepatic tissue with a functional bile ductular network. Tissue Engineering. Part A, 19(21–22), 2402–2411. 10.1089/ten.TEA.2013.0021 [DOI] [PubMed] [Google Scholar]
  19. Kumar A, Tripathi A, & Jain S (2011). Extracorporeal bioartificial liver for treating acute liver diseases. The Journal of Extra-Corporeal Technology, 43(4), 195–206. [PMC free article] [PubMed] [Google Scholar]
  20. Lewis PL, Su J, Yan M, Meng F, Glaser SS, Alpini GD, … Shah RN (2018). Complex bile duct network formation within liver decellularized extracellular matrix hydrogels. Scientific Reports, 8(1), 12220 10.1038/s41598-018-30433-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Mansy SS (2004). Agarose cell block: Innovated technique for the processing of urine cytology for electron microscopy examination. Ultrastructural Pathology, 28(1), 15–21. [PubMed] [Google Scholar]
  22. Masarone M, Rosato V, Dallio M, Gravina AG, Aglitti A, Loguercio C, …Persico M (2018). Role of oxidative stress in pathophysiology of nonalcoholic fatty liver disease. Oxidative Medicine and Cellular Longevity, 2018, 1–14. 10.1155/2018/9547613 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Mazza G, Al-Akkad W, Rombouts K, & Pinzani M (2018). Liver tissue engineering: From implantable tissue to whole organ engineering. Hepatology Communications, 2(2), 131–141. 10.1002/hep4.1136 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Mohdad AA, Lopez AD, Shahraz S, Lozano R, Mokdad AH, Murray CJL, & Naghavi M (2014). Liver cirrhosis mortality in 187 countries between 1980 and 2010: A systematic analysis. BMC Medicine, 12(145), 1–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Muriel P (2017). Chapter 1—The liver: General aspects and epidemiology Liver Pathophysiology: Therapies and Antioxidants, (3–22). London: Elsevier Inc. [Google Scholar]
  26. Parafati M, Kirby RJ, Khorasanizadeh S, Rastinejad F, & Malany S (2018). A nonalcoholic fatty liver disease model in human induced pluripotent stem cell-derived hepatocytes, created byendoplasmic reticulum stress-induced steatosis. Disease Models & Mechanisms, 11(9), 1–15. 10.1242/dmm.033530 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Raynaud P, Carpentier R, Antoniou A, Lemaigre FP (2011). Biliary differentiation and bile duct morphogenesis in development and disease. The International Journal of Biochemistry & Cell Biology, 43(2), 245–256. 10.1016/j.biocel.2009.07.020 [DOI] [PubMed] [Google Scholar]
  28. Rizki-Safitri A, Shinohara M, Miura Y, Danoy M, Tanaka M, Miyajima A, & Sakai Y (2018). Efficient functional cyst formation of biliary epithelial cells using microwells for potential bile duct organisation in vitro. Scientific Reports, 8(1), 549 10.1038/s41598-018-29464-w [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Rizki-Safitri A, Shinohara M, Tanaka M, & Sakai Y (2020). Tubular bile duct structure mimicking bile duct morphogenesis for prospective in vitro liver metabolite recovery. Journal of Biological Engineering, 14, 11 10.1186/s13036-020-0230-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Shafiee A, & Atala A (2017). Tissue engineering: Toward a new era of medicine. Annual Review of Medicine, 68, 29–40. 10.1146/annurev-med-102715-092331 [DOI] [PubMed] [Google Scholar]
  31. Sharma VR, Shrivastava A, Gallet B, Karepina E, Charbonnier P, Chevallet M, … Deniaud A (2019). Canalicular domain structure and function in matrix-free hepatic spheroids. Biomaterials Science, 8(1), 485–496. 10.1039/c9bm01143a [DOI] [PubMed] [Google Scholar]
  32. Uygun BE, Soto-Gutierrez A, Yagi H, Izamis ML, Guzzardi MA, Shulman C, … Uygun K (2010). Organ reengineering through development of a transplantable recellularized liver graft using decellularized liver matrix. Nature Medicine, 16(7), 814–820. 10.1038/nm.2170 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Vroman B, & LaRusso NF (1996). Development and characterization of polarized primary cultures of rat intrahepatic bile duct epithelial cells. Laboratory Investigation, 74(1), 303–313. [PubMed] [Google Scholar]
  34. Wang Y, Nicolas CT, Chen HS, Ross JJ, De Lorenzo SB, & Nyberg SL (2017). Recent advances in decellularization and recellularization for tissue-engineered liver grafts. Cells Tissues Organs, 203(4), 203–214. 10.1159/000452761 [DOI] [PubMed] [Google Scholar]
  35. Williams IR (1998). Fibroblasts In Delves PJ (Ed.), Encyclopedia of Immunology (2nd ed., pp. 905–909). United Kingdom: Elsevier. [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

supplemental figures
NIHMS1632184-supplement-supplemental_figures.pdf (29.1MB, pdf)

RESOURCES