Several studies have provided evidence that this technology offers a valuable platform for liver bioengineering through the repopulation of an acellular liver with appropriate fresh cells.
The first report addressed the methodology for the decellularization of rodent livers[1
]. Livers were cannulated through the inferior vena cava, with the portal vein severed and the superior vena cava clamped. The decellularization process began with rinsing of the liver with 100 mL of phosphate buffered saline (PBS) to clear the blood followed by perfusion of three 300 mL isotonic solutions of 1%, 2%, and 3% Triton X-100 at a rate of 5 mL/min. This was followed by perfusion of a 300 mL PBS solution containing 0.1% sodium monododecyl sulfate (SDS) and a 300 mL PBS wash. The disruption of the lipid membranes cleared most of the cellular components of the organ except for intact nuclear cages containing DNA, which was further removed by a solution of SDS. Hematoxylin and eosin staining of the intact decellularized liver showed a fine web of matrix remaining in the acellularized liver, which was further analyzed by immunohistochemical staining of collagen IV and laminin. The stains showed the presence of collagen within the matrix and that laminin was present within the basement membrane of the vessels. After the decellularization process, the scaffold remained intact and strong enough to maintain further cannulation for the perfusion of cells. 106
cells of the rat liver progenitor cell line WB344 in Roswell Park Memorial Institute medium were infused into the decellularized liver through the cannulated inferior vena cava. Further histological analysis of the center of the intact recellularized scaffold indicated that the intrahepatic vasculature was able to traffic cells from the inferior vena cava. This report demonstrated the necessary process of using SDS in the decellularization process to truly remove any cellular components, specifically DNA.
Another similar subsequent report that used a similar decellularization method showed vascular patency through portal vein dye[2
]. The decellularization process was performed by sequential perfusion of different concentrations of detergents through the portal vein at a flow rate of 1 mL/min. The livers were perfused for 72 h with SDS in distilled H2
O: 0.01% SDS for 24 h, 0.1% SDS for 24 h, and 1% SDS for 24 h. The livers were then perfused with distilled H2
O for 15 min and with 1% Triton X-100 for 30 min to cleanse the livers of any remaining SDS. After rinsing the decellularized livers with PBS for 1 h, only the median lobe was sterilized in 0.1% peracetic acid in PBS for 3 h and kept for recellularization after further extensive PBS washing. The decellularized scaffolds were histologically analyzed to demonstrate that the scaffolds were acellular and functionally similar to an intact normal liver, in order for recellularization to be possible. Histological analysis showed that there were no nuclei or cytoplasmic staining in the decellularized liver compared to a normal rat liver. Immunohistochemical analysis of four extracellular matrix (ECM) proteins (collagen type I, collagen type IV, fibronectin and laminin-β1) showed that the structural and basement membrane components of ECM remained intact similarly to the normal liver. DNA analysis of the decellularized scaffold showed that less than 3% of residual DNA remained after the decellularization process. They also reported intact functional vascular beds and microvasculature through the perfusion of the Allura Red dye. The dye flowed through the vasculature just as expected in a functioning liver. The acellular translucent scaffold was then infused with rat-derived hepatocytes through perfusion of the portal vein at 15 mL/min. The perfusion system consisted of a peristaltic pump, bubble trap, and oxygenator from a donors-after-cardiac-death organ resuscitation perfusion system. They introduced approximately 12.5 million cells during each of the four steps in ten-minute intervals, which showed superior engraftment efficiency when compared to a single-step infusion. The recellularized grafts were maintained in a perfusion chamber for up to 2 wk in vitro
, with histological staining of the recellularized sections at 4 h, 1 d, 2 d, and 5 d of perfusion. At 4 h, the majority of the cells remained in and around the vessels; however, at 1 d and 2 d, the cells leave the vessels and become distributed throughout the matrix.
It should be emphasized that this is the first report that contains data showing the level of function exhibited by the hepatocytes grown on the decellularized matrix. They report that hepatocyte viability was maintained during culture and that cell death was kept to a minimum. They were also able to determine that the cells migrated beyond the matrix barrier to reach decellularized sinusoidal spaces through scanning electron microscopy (SEM) and histological analysis. They also determined that albumin synthesis was not increased in the recellularized matrix compared to an intact liver; however, urea synthesis was significantly higher in the recellularized liver than the hepatocyte sandwich during culture. The analysis of the expression of drug metabolism enzymes showed that the levels of Cyp2c11, Gstm2, Ugt1a1, and Cyp1a1 that were expressed in the recellularized grafts were similar to those of the sandwich hepatocyte cultures. The recellularized liver grafts were then transplanted into recipient rats that underwent unilateral nephrectomy for auxiliary liver graft transplantation. The recellularized liver grafts were perfused quickly with blood and the appropriate efflux occurred only after 5 min. The graft was maintained in vivo for 8 h, and then harvested for further Tdt-mediated dUTP Nick-End Labeling staining analysis. This staining demonstrated that the cells were minimally damaged and further histological staining showed that the hepatocytes reserved normal morphology and parenchymal positions.
While it is extremely important to have these previous promising reports on liver decellularization and recellularization, the ultimate necessary technology that needs to be expanded upon is the decellularization and recellularization of whole organs-specifically human organs, and subsequently human derived cell lines-in order to create a transplant graft for possible human functioning. The report by Baptista et al[3
] demonstrated the potential for the colonization of human hepatocyte progenitors on a decellularized liver matrix. This is one of the first reports to show the decellularization and recellularization process with a whole liver instead of thin slices or lobes of the liver, as well as the first report to recellularize successfully with human liver cells. They attempted to decellularize whole livers from multiple species as well, including mice, rats, ferrets, rabbits, and adult pigs.
All of the dissected livers were cannulated with different gauged cannulas, depending on the species, through the inferior vena cava and the portal vein, which were then hooked up to a Masterflex peristaltic pump in preparation for decellularization. There was approximately 40 times the volume of the liver perfused with distilled water at a flow rate of 5 mL/min. The decellularization process was performed by perfusion of approximately 50 times the volume of the liver with 1% Triton-X 100/0.1% Ammonium Hydroxide. The approximate perfusion times for the decellularization process were 1 h for mice, 2 h for ferret, 3 h for rat, and 24 h for pig livers. It was visibly clear after the perfusion period that the parenchyma became transparent and the vascular tree was visible under low magnification microscopy (Figure ).
Figure 1 Gross and microscopic anatomy of acellular ferret livers. Upper row: The liver on the left is almost entirely decellularized, however it remains a segment still cellular (interrupted line); on the left, instead, the liver is fully acellular as expression (more ...)
Spectrophotometric and agarose gel DNA analysis showed the removal of approximately 97% of the DNA from the tissue, indicating efficiency of the decellularization process. SEM was performed to determine that that ultrastructure was preserved. The SEM analysis showed that reticular collagen fibers that support the hepatic tissue were present and the “portal triad” structures remained intact, as well as the lack of any cells. Histological analysis of acellular ECM was performed to further characterize the scaffold composition. The staining showed that there was no cellular nuclear material or any other cellular material present. The staining also showed that collagen layers with vascular channels were present, along with collagen fibers, elastin fibers, and glycosaminoglycans (Figure ).
Movat-Pentachrome staining of acellular liver sections shows yellow staining for collagen and dark staining for elastin surrounding the vascular structures.
Quantification of the ECM components showed higher levels of collagen and glycosaminoglycans in the decellularized scaffold compared to native tissue, which can be explained by the absence of cellular components, while there was no difference in elastin presence. The localization of the specific extracellular matrix proteins collagen I, collagen III, collagen IV, laminin, and fibronectin were all observed around the vascular structures, specifically denser around the larger vessels, and the parenchymal areas of the acellular liver, as well as the fresh tissue. Vascular preservation and patency was demonstrated by the ability of the network of vascular remnants to retain labeled dextran that had a similar molecular weight to that of blood proteins.
The recellularization methods used in this report show that perfusion through the vena cava or the portal vein (preferred) both allow the green fluorescent protein-labeled MS1 endothelial cells to line the vascular network, including the larger vessels to the capillary sized vessels. Portal vein-seeded endothelial cells were primarily deposited in the periportal regions of the liver lobule while the vena cava-seeded endothelial cells were primarily concentrated in the regions of the central veins and in smaller branches and vessels. Through fluorescent microscopy and transmission electron microscopy they were able to determine that the lumens of the acellular vascular remnants could be colonized by endothelial cells that were able to actively spread and cover the vessel basement membrane while forming appropriate cell-cell junctions. They also determined that the surface of the vascular lumen was non-thrombogenic, which was confirmed by the lower quantification of platelets in the bioscaffold compared to the fresh tissue. The reseeding experiments performed in this report utilized the coseeding of human umbilical vein endothelial cells and freshly isolated human fetal liver cell’s, while using similar recellularization protocols previously mentioned. Immunohistochemical analysis was used to assess the proliferation and analyze the presence of hepatocytic lineage markers. Staining of Ki67 to assess proliferation showed a high number of positive cells throughout the bioscaffold, which was 3 times higher than the number of apoptotic cells present. The staining also showed that the hepatocytic markers α-fetoprotein, CYP2A, and CYP3A were expressed in the parenchyma. Cytokeratin 19 was strongly seen throughout the bioscaffold in biliary tubular structures while clusters of albumin-expressing hepatocytes were distributed in the parenchyma. The small amount of coexpression of these specific markers implies that there are specific niches within the bioscaffold for bile duct and hepatocytes. Immunohistochemical staining also detected CK19+/CK18-/ALB-tubular structures and clusters of ALB+/CK18+ cells in the parenchyma, which suggests that the bioscaffold is able to support the differentiation of the fetal hepatoblasts into biliary or hepatocytic lineages. The ability of cells with immunophenotypes consistent with hepatocytes, cholangiocytes, and endothelial cells to form discrete pockets in the bioscaffold suggests that some of the micro-architectural “blueprint” was retained within the scaffold. This suggests that not only does the bioscaffold provide a three-dimensional vascularized scaffold (previously described) but it also retains the necessary environmental cues, further explained by the retention of the glucosaminoglycans that serve as active binding sites for growth factors that regulate cell phenotype, for progenitor hepatic and endothelial cells to grow, differentiate, and maintain functionality.
A related study reports on a refined decellularization procedure. This study demonstrated the ability of liver progenitor cells to differentiate to both the hepatocyte and cholangiocyte lineages while seeded on the decellularized scaffold[4
]. The strategy for recellularizing the bioscaffold was aimed at creating a more rapid and efficient differentiation of the stem cells using tissue-specific extracts enriched in extracellular matrix and a hormonally specific defined medium using associated growth factors and cytokines. They reseeded the scaffold with human hepatic stem cells in a hormonally defined medium specific for adult liver cells. The stem cell markers were expressed in the cells after the reseeding process and the cells differentiated into mature functional parenchymal cells in approximately one week. These cells remained viable and presented stable mature phenotypes for more than 8 wk.
Similar results have been obtained by other groups[5,6
]; however in all the above reported investigations liver ECM was produced from rodent livers. Instead, Barakat et al[7
] recently developed a method to decellularize porcine livers, which were eventually repopulated with human cells[8,9
]. The goal was to produce a clinically relevant model of liver bioengineering. Livers from Yorkshire pigs were decellularized with SDS. The ECM of the posterior segment of the right liver lobe was used as scaffold for cell seeding. Fetal hepatocytes co-cultured with fetal stellate cells were expanded, collected, resuspended in appropriate medium supplemented with hepatocyte growth factor and seeded within the ECM. The so-obtained constructs were perfused for 3 d, 7 d and 13 d. During perfusion, pH, PO2
, lactate, glucose, urea nitrogen and albumin were measured to assess metabolic and synthetic functions. Of note, some constructs were implanted in vivo
and perfused for 2 h to determine the behavior of the matrix in vivo
and its ability to withstand the shear stress produce by the blood flow in physiologic conditions. Results were encouraging. Liver organoids showed active metabolism and preserved capability to synthesize albumin, and were able to sustain blood pressure without harm. Notably, immunohistochemical analysis revealed cell differentiation into mature hepatocytes. This latter finding provides evidence that ECM is essential in that it supports cells and may drive the differentiation of progenitor cells into an organ-specific phenotype[10
]. Badylak’s group confirmed this information in an elegant model of liver hepatectomy in rats[11
], in which he demonstrated that liver ECM implanted in intact and amputated livers enhances hepatocyte proliferation and ultimately liver regeneration.
While the primary goal for the majority of the research pertaining to decellularizing and recellularizing an organ is the functional transplantation of a bioengineered organ into a recipient host, there are the possibilities of using this technology in in vitro
studies for advanced preclinical drug development[12
]. This report provided a 60-min rapid natural decellularization method for a 3-dimensional scaffold prepared from a rat liver that maintained the microvascular system and was able to withstand fluid flowing through all three hepatic circular systems. The method utilized two thirty-minute perfusion periods; a 1% Triton-X 100 solution followed by a 1% SDS solution. The development of a novel in vitro
3-dimensional model that closely represents the in vivo
liver could present the potential for toxicity testing of key compounds in preclinical drug developments since the liver is the main metabolizing organ that is usually the target of toxicity.