PriCells: Isolation of Human Primary Hepatocytes                                             

A number of factors are important in ensuring a functional hepatocyte preparation with a high yield.

1. Perfusion of the tissue by various media is critical in the isolation of high-quality hepatocytes. It is essential that the perfused tissue has only one cut surface and that the capsule surrounding the tissue is otherwise intact. Careful placement of the cannulae is crucial in achieving the optimum perfusion of the tissue sample.

2. It is important to monitor pH and temperature. The advantage of the bicarbonate buffering system reported here is that oxygenation of the perfusate with a mixture of oxygen and carbon dioxide (95%/5%) is combined with a very high buffering capacity. A disadvantage of the system is the formation of microbubbles. It is important not to allow bubbles to enter the tissue as this will prohibit the perfusion of the tissue. A small drop in temperature of 1–2°C in perfusion temperature can prolong the perfusion time and increase the cellular damage caused by the perfusion. The temperature at the liver should be monitored and maintained at 37°C.

3. The size of the liver wedge biopsy sample to be perfused should be considered. Smaller human liver samples up to 5 g gave better yields than samples over 5 g. Furthermore, these workers have indicated that the quality of the collagenase used for the perfusion is important. Collagenase with a high specific activity provides a better yield of viable hepatocytes. This increase in yield was thought to be caused by a reduced perfusion time, which, in turn, leads to a reduction in cell destruction. Other workers have shown that hepatocyte yield and viability can be increased if either hyaluronidase (0.05%) or dispase (500 U/mL) is added to the hepatocyte dissociation medium. However, for the preparation of rat hepatocytes, hyaluronidase is thought to be detrimental.

4. The use of a calcium chelator to remove calcium ions that are essential for junc- tional complex desmosome integrity is recommended. Perfusion with EGTA brings about cleavage of the junctional complexes. It has been suggested that perfusion without a chelator reduces hepatocyte yield.

5. Trypan blue exclusion or leakage of certain cytoplasmic enzymes are relatively quick and simple assays of cell viability. Care must be taken with the trypan blue assay because it has been reported that a decrease in media pH can lead to increased dye uptake in rat hepatocytes, which is reversible. This phenomena has not been reported for human hepatocytes. Furthermore, it should be noted that staining intensity can vary depending on the batch of trypan blue or media components. Protein binds trypan blue strongly and, thus, there should be no protein in the medium when analyzing cell viability by this method, other- wise a falsely high value for viability will be obtained.

6. A review detailing a number of methods for the isolation of hepatocytes has recently been completed.

7. Confluent monolayers have been established with seeding densities of 4–18×10⁴ cells/cm². In general, seeding densities should be sufficient to allow cell growth and to enable the maintenance of cell–cell contacts. Generally, cell densities of 10–15×10⁴ cells/cm2 should be sufficient.

8. The attachment efficiency (number of cells attached/number of cells seeded) should be determined and be in the region of 70–80%. Attachment periods of up to 20 h for human hepatocytes have been reported. Human hepatocytes prepared from the livers of older donors require increased attachment time.

9. The influence of various matrix components on hepatocytes in culture has been widely studied. In the liver, the hepatocyte is surrounded by a matrix that contains collagen, laminin, fibronectin, and heparan sulfate proteoglycan. An ideal cell matrix should contain some of these components, be able to support hepatocyte morphology over long periods, allow formation of the monolayer and intercellular contacts, protect cells from sheer forces of fluids, and allow free exchange of salts within the media. However, in general, these organic matrices do not delay the occurrence of phenotypic changes. Moreover, the substrata can enhance alteration of liver gene expression.

Hepatocytes in culture have been shown to have the ability to secrete several types of collagen and noncollagenous glycoproteins. The use of plastic culture plates coated with an extracellular matrix such as collagen, fibronectin, laminin, collagen gels, collagen gels embedded on a nylon mesh, or a laminin-rich matrix termed Matrigel® has been shown to improve the efficiency of hepatocyte attachment, increasing the duration of functional activity and the morphology of the cells in culture. The use of attachment factors in animal hepatocyte culture is extensively reviewed. Human hepatocytes in culture have been reported to attach well to extra cellular matrices such as uncoated plastic, soluble rat tail or calf skin collagen type I, fibronectin, collagen gel immobilized hepatocytes, and Matrigel.

10. The effects of various hepatocyte culture media and supplements on the maintenance of hepatocyte function have been the subject of a number of reviews. Many different media have been used and little work has been published comparing media effects on animal hepatocytes. In one recent study, for example, WME was the best of the commercially available media for rat hepatocytes, but that Chee’s medium offered advan- tages. However, other authors find little difference in the experimental effects of culture medium for human hepatocytes in short-term culture (1–2 wk). As with matrices, the systems employed for human hepatocyte culture are very diverse and tend to be extensions of conditions used for the culture of animal hepatocytes. Media that have been used for human hepatocyte culture include Weymouth’s 752, minimal essential media mixed with media 199 (75% 25% v/v), Ham’s F12, Leibovitz L-15, WME, and ISOMs.

Once the basic media has been selected, a number of media supplements can be added. These supplements range from natural compounds including hormones, minerals, and vitamins to nonphysiological factors such as dimethyl sulfoxide (DMSO).

a. Serum—composition unknown but contains proteins, growth factors.

b. Hormones—insulin, glucagon, glucocorticoids, thyroxine, sex steroids.

c. Growth supplements—selenium, zinc, EGF, HGF.

d. Nonphysiological factors—DMSO, ammonium chloride, metyrapone

The role of each of these supplements is, at best, unclear, even for the well-evaluated rat hepatocyte culture. However, a number of workers have established long-term human hepatocyte cultures by the use of a number of different medium additives.

11. Studies of the functionality of human hepatocytes in culture are also limited when compared to data on animal hepatocytes. However, workers have shown that gluconeogenesis and glycolysis could be stimulated in human hepatocytes by physiological concentrations of glucagon and insulin. The gluconeogenic rate in human hepatocytes was similar to that estimated for the fasted human liver, whereas basal glycolysis is higher in cultured human hepatocytes than in vivo.

Urea synthesis from ammonia has been used as an indicator of mitochondrial function. Under basal conditions human hepatocytes produce 2–4 nmoles urea/mg cell protein/minute. Hepatocytes can be stimulated to produce significantly more urea, up to 11 nmoles/min/mg. The maximal rate of urea genesis decreased by 50% after one day in culture but was then stable for several days thereafter.

Protein production by human hepatocytes has been used as a measure of cell function. Human hepatocytes in culture have been shown to secrete albumin, α-antitrypsin, α-antichymotrypsin, α-acid glycoprotein, serum amyloid A, fibronectin, hepatoglobulin, and α2-macroglobulin. The ability of cultured human hepatocytes to secrete albumin has been used by many workers as a measure of culture viability. However, a number of other methods have also been reported including measurement of ATP content and LDH leakage.

12. A primary objective of human hepatocyte cultures is to provide models to study the potential disposition of xenobiotics prior to administration to humans. The cytochromes P450 (CYP450) present in cultured human hepatocytes are thought to be more stable than those in rat hepatocytes. Studies have shown that CYP450 may be maintained for a number of days in primary culture. In addition, human hepatocytes co-cultured with nonparenchymal cells maintain a relatively stable cytochrome P450 content for longer periods of time. Cytokines produced during inflammation or infection have been shown to affect the expression of the major human CYP450 enzymes in hepatocyte cultures. The interleukins -1B, -4, -6, tumor necrosis factor-α, and interferon-α and -γ in addition to nitric oxide have been shown to downregulate specific isozymes of CYP450 in human hepatocyte cultures. The observed in vitro effects of cytokines on drug-metabolizing enzymes may be important clinically with respect to chronic hepatitis patients who receive high doses of interferon as well as antiviral compounds.

Investigations and evaluation of the drug-metabolizing activities present in human hepatocyte cultures has demonstrated that these cells in culture appear to be a good model for predicting drug metabolism in vivo. Human hepatocyte cultures provide an opportunity to study both the routes and rates of metabolism of new drugs and are a key in vitro model to improve the drug development process. Several workers have shown that the metabolic routes and rates of a number of drugs in human hepatocyte cultures are similar to those observed in vivo. Furthermore, human hepatocyte cultures have also been used to help identify species differences in the metabolism of a number of drugs.

Hepatocyte cultures have been used extensively to study CYP450 expression and regulation. The CYP450 can be induced by certain compounds in human hepatocyte cultures. Incubation of human hepatocyte cultures in the presence of either phenobarbital (1.5–32 mM), rifampacin (50 μM), 3-methylcholanthrene (25–50 μM), benzanthracene (12.5 μM), and ethanol (200 mM) induce both the mRNA and protein of specific cytochrome CYP450 isozymes. Furthermore, the human hepatocyte culture system has been used successfully to examine the potential of therapeutic agents such omeprazole, lansoprazole, and pantoprazole to induce cytochromes P450 in vitro. Whether the induction observed in closed in vitro systems can be extrapolated to humans is a key question currently under investigation. However, the contribution of pharmacokinetics to the in vivo disposition of compounds shown to affect the regulation of enzymes in vitro should be carefully considered.

The human hepatocyte culture has also been used to model a number of drug–drug interactions resulting from changes in the metabolism compounds caused by either induction or inhibition of the enzymes responsible for their metabolism.

The main focus of many researchers has been the cytochrome CYP450 drug- metabolizing enzymes. However, the conjugating enzymes have also been shown to be altered in human hepatocyte culture. UDP glucuronyltransferase increased during culture whereas sulfotransferase activity decreased. Glutathione content remained relatively stable. Again, the phase II enzyme content of the cells is affected by medium composition and other culture conditions.

In addition to metabolism studies human hepatocytes in culture have been evaluated as screens for hepatotoxicity or to elucidate mechanisms of toxicity.

The Cryopreservation of Human Hepatocytes: Owing to the infrequent supply of human material, methods of cryopreserving hepatocytes prior to culture have been developed. Numerous researchers have published results showing that cryopreserved hepatocytes retain their metabolic capabilities. Human hepatocytes have been cryopreserved in mixtures of buffers containing a cryoprotectant. DMSO generally appears to be the cryoprotectant of choice. Hepatocytes are then frozen in a controlled manner and then ideally stored at <–150°C. At present, cyropreserved human hepatocytes tend to have poor attachment rates (approx 50%) in culture.

Immortalized Cell Lines: In the absence of freshly isolated or cryopreserved native hepatocytes, immortalized hepatic cell lines such as the human-derived hepatoma cell line HEP G2 have also been utilized. These cell types have been shown to express both phase I and II drug-metabolizing activity albeit lower than that expressed in normal liver cells. Cell lines have also been used with some success for the study of drug metabolism, induction, or cytotoxicity.