Liver cell implants: A long road ^†

The potential indications for liver cell transplantation are numerous. They include rescue from acute liver failure until native liver regeneration occurs, the long-term supplementation of liver function in patients with chronic liver disease, and the treatment of inborn errors of metabolism. The list of approaches that might be applied to cell transplantation is similarly substantial. The source and type of liver cells to be used need to be defined, as do the suitability of various possible sites for implantation. Techniques to encourage implantation, survival, function and perhaps proliferation of the implanted cells need to be optimized. The state of the art in the clinic is at a crossroads: many different approaches have been tried in small trials, which have almost universally been uncontrolled. The results of these trials show promise, but they have not defined effective clinical protocols and have also emphasized just how much basic cell biology research needs to be performed to create robust cell therapies.

Until recently, interest in liver cell transplantation was concentrated on mature hepatocytes, which are capable of replacing the majority of the liver's synthetic and detoxification functions. The original observation made by Mito et al.1 that hepatocytes implanted into the spleen would survive and indeed proliferate was followed by similar observations at various sites (eg, at subcutaneous and intrapulmonary sites and under the renal capsule). Gupta and coworkers2 adopted the dipeptidyl peptidase IV–negative rat strain; into this strain, otherwise syngeneic dipeptidyl peptidase IV–positive hepatocytes could be transplanted, and they subsequently could be identified by histochemistry for this membrane-expressed enzyme. This led to a series of articles identifying the process by which cells injected into the portal circulation (which was generally accessed after injection of cells into the splenic pulp) could implant within their natural home, the liver parenchyma. They described the processes by which the implanted hepatocytes exit the sinusoidal circulation through the endothelium, integrate themselves within the hepatic cords, and establish cell-to-cell contacts with the host hepatocytes that are apparently identical to those between the native host cells. The process of migration into the hepatic cords can be modulated by, for example, the administration of cyclophosphamide.3 However, although these demonstrations are exciting, they pose more questions and problems requiring answers before effective clinical applications. These include (1) difficulties in accessing the hepatic cords when the normal sinusoidal barriers are distorted and thickened by connective tissue in cirrhosis and (2) problems in ensuring that implanted cells thrive when they are injected into a toxic milieu of disorganized necrotic parenchyma in patients with acute liver failure. The work implies that the option of treating inborn errors of metabolism is likely to be more feasible, particularly for diseases of hepatic function in which the liver parenchyma remains structurally normal but there is an abnormality in the synthetic or enzymatic activity of the liver (eg, the urea cycle defects of defective coagulation factor synthesis). Even under these conditions, there remain difficulties to be overcome, including the creation of sufficient space for an adequate number of implanted hepatocytes to enter the liver parenchyma, the prevention of rejection, the long-term survival of implanted hepatocytes, and the induction of selective proliferation in comparison with native cells.

One approach to creating space is to damage the host liver parenchyma, and experimentally, a variety of approaches (eg, toxic damage through the administration of carbon tetrachloride and radiation damage) have been employed. These approaches also provide a simultaneous regenerative stimulus to the implanted cells.4, 5 An alternative is to look to other spaces, and the peritoneal cavity has its attractions. There is a large potential volume, and there is no clearly defined function that would be imperiled by cell implantation. Early studies of hepatocyte implantation in the peritoneal cavity were disappointing with only short-term and sparse survival.6, 7 However, the first report of a complete biochemical cure of an inborn error of metabolism followed the implantation of a mixed liver cell population into the peritoneal cavity of mice with histidinemia; the researchers exploited the concept that cotransplanted nonparenchymal cells can provide hepatocytes with the extracellular matrix that they require.8

The role of extracellular matrix in the survival of ectopic hepatocytes and indeed in the expression of the hepatocyte repertoire appears fundamental. Classic observations pointed out that in a monolayer culture, hepatocytes performed poorly and rapidly lost differentiated function9; a variety of supporting matrices, which ranged from collagen gel to complex tissue extracts such as Matrigel, were shown to prolong both survival and function in culture. The liver cell biologist H. O. Jauregui said the following of cultured liver cells: “If they look like a fried egg they don't work—they have to look like a brick.” The relationship between extracellular matrix (a complex mixture of collagen subtypes, laminins, fibronectin, tenascin, and other substances in the native liver) and cell shape and function is complex.10 The repetitive motifs on linear matrix proteins, which interact with cellular integrins, can impose a shape on cells, maintain the nearly cuboidal, polarized shape of the hepatocyte, and activate intracellular pathways driving differentiated function.10 Yet differentiated function can also be encouraged by other maneuvers that impose a nearly cuboidal conformation and close cell-to-cell interaction; furthermore, although the nonparenchymal cell population is regarded as the primary source of extracellular matrix proteins, hepatocytes can express extracellular matrix.11 These 2 features probably underlie the efficacy of hepatocyte spheroids (spherical aggregates of tens or even hundreds of cells) in maintaining the expression of differentiated function over time.12 A variety of approaches, including simple physical approaches, can generate such spheroid aggregations; for example, spontaneous aggregation can be allowed to take place when a suspension of cells is subjected to a rotary motion.13

In this issue of Liver Transplantation, Török et al.14 report a technique combining a biodegradable polymer scaffold, human hepatocytes, and an extracorporeal perfusion system that encourages the high-level expression of differentiated function by those human hepatocytes. The cells, after they are seeded within the matrix, aggregate as spheroids and perform both synthesis (protein synthesis) and detoxification functions over a period of days. The authors envisage this as a step to be followed by the intraperitoneal implantation of the cells within the peritoneum (ie, the induction of liver neotissue for implantation), which potentially has advantages such as a lesser degree of cell loss after implantation, preformed functional integrity, and regenerative potential. All these claims will, of course, need to be verified, and clearly the authors are engaged in that exercise. Others have addressed this issue over the years with intraperitoneal cell implants, including alginate-encapsulated cells, cells within synthetic semipermeable membranes, and other supporting scaffolds.15-17 No technique has yet emerged as a proven and valuable clinical approach, but the use of human cells rather than rodent cells is clearly one important step.

Before these approaches can be adopted, we will need to resolve the following questions: What is the long-term fate of implanted cells in the peritoneum? Will their growth be sustained? If indeed they are vital to the survival of an individual, will they experience a growth advantage analogous to that of normal hepatocytes in animal models such as the mutant fumarylacetoacetate hydrolase-deficient (tyrosinemic) mouse, in which there is preferential and progressive growth of the advantaged cells to repopulate the liver?18 Will a similar proliferation of advantaged cells in the peritoneum under some conditions create a benign intraperitoneal tumor after the initial matrix has undergone biodegradation, or will the implanted cells wither when that initial support has dissipated?

The availability of human hepatocytes also presents difficulties. Primary human hepatocytes scarcely proliferate in culture. Specialized culture media can encourage sparse colony formation after many days,19 but they are not the easy resource that would be required to provide a predictable supply of cells for implantation. Primary human hepatocytes generally become available as a short-lived resource in the form of an explanted liver, for which there is currently a proven and effective use in conventional liver transplantation. Organs too marginal for grafting are the obvious source, but issues similar to those rendering the organs useless to the surgeon reduce the viability of cells isolated from such organs for implantation protocols.20 Fetal cells or hepatoblasts from perinatal livers are a biologically feasible source because they maintain their proliferative potential, but they present social and ethical difficulties.21, 22 The well-differentiated hepatocyte-derived cell lines currently being exploited in bioartificial livers, though probably safe in extracorporeal circuits, are clearly inappropriate for implantation.23 As with other novel areas of therapy, the use of hepatocyte-like cells differentiated from human embryonic stem cells,24 induced pluripotent stem cells,25 and other sources such as the amniotic sac26 seems an extraordinarily attractive proposition. However, the tumorigenicity of many cell lines derived from human embryonic stem cells and the uncertain status of induced pluripotent stem cells, together with the regulatory process that needs to be surmounted,27, 28 mean that these resources for cell implantation will not be available for some years to come.