What Are Primary Human Hepatocytes and Their Key Uses?

Primary human hepatocytes are liver cells isolated directly from human liver tissue and kept alive in laboratory culture. They are the closest thing researchers have to studying a working human liver outside of the body, retaining the full set of metabolic enzymes, transport proteins, and signaling pathways that make the liver the body’s central chemical processing plant. That makes them indispensable for drug development, toxicology screening, disease modeling, and even experimental cell therapies. But working with these cells is far from straightforward, and the gap between their scientific value and their practical limitations shapes much of modern liver research.

Where Primary Human Hepatocytes Come From

The liver tissue used to isolate primary human hepatocytes comes from two main situations: surgical resections and organ transplantation. When a patient has part of their liver removed to treat a tumor or another condition, the leftover tissue that would otherwise be discarded can be used. Similarly, when a patient receives a new liver through transplantation, the diseased liver that was taken out, or donor tissue that was deemed unsuitable for transplant, becomes a source of cells. Researchers have successfully isolated hepatocytes from livers affected by a wide range of conditions, including alcoholic liver disease, biliary cirrhosis, cystic fibrosis, autoimmune hepatitis, and non-alcoholic steatohepatitis, as well as from tissue removed during surgery for colorectal cancer metastases and benign liver conditions.¹ Hepatocytes can also be obtained from explanted livers taken from patients with decompensated cirrhosis or certain metabolic liver diseases, offering a resource for pharmaceutical research even when the tissue itself is severely compromised.²

The classic method for getting hepatocytes out of liver tissue is a two-step process involving an enzyme called collagenase, which breaks down the connective scaffold holding the cells together. This perfusion-based approach, adapted from a technique first developed for rat livers, remains the standard because it yields high numbers of intact, functional cells.³ However, perfusion requires relatively large, intact pieces of tissue with visible blood vessels to push fluid through, which limits how small a tissue sample can be. A newer non-perfusion method, where thin slices of liver tissue are digested directly, has shown it can recover roughly 1.17 million viable cells per gram of tissue with about 80% viability, and works even with fatty liver samples.⁴ That flexibility matters, because many research-grade tissue samples are small surgical leftovers rather than whole organs.

Why They Are Considered the Gold Standard for Drug Metabolism

Your liver is responsible for breaking down the vast majority of drugs you take, and the enzymes that do this work belong to a large family often referred to collectively as cytochrome P450 enzymes. Primary human hepatocytes express these enzymes at levels that closely match what is found in a living human liver, which is exactly why drug companies rely on them. When researchers need to know how a new drug candidate will be metabolized in a real person, hepatocytes from human donors are the most reliable in vitro model available.

The alternative cell models fall short in important ways. HepG2 cells, a widely used liver cancer cell line, have extremely low levels of the phase I metabolic enzymes that handle the initial chemical transformation of most drugs. While HepG2 cells can be useful for studying how drug-metabolizing enzymes are regulated, they are a poor stand-in for predicting actual human drug metabolism.⁵ Stem cell-derived hepatocytes and HepaRG cells (another cancer-derived line sometimes marketed as a closer approximation) also lag behind. A head-to-head comparison of gene expression across these models found that HepG2 and stem cell-derived hepatocytes clustered together statistically, both showing a low cytochrome P450 profile that was distant from the expression pattern seen in primary hepatocytes from multiple donors.⁶ In practical terms, if you test a drug’s breakdown in HepG2 cells and assume the result applies to humans, you could badly misjudge how quickly or slowly a patient will clear that drug.

Detailed protein measurements in cultured primary hepatocytes have pinpointed which specific enzymes and transporters are present and at what quantities. Among the cytochrome P450 enzymes, CYP2C9 tends to be the most abundant, while some enzymes like CYP2B6 and CYP3A5 can be so scarce in certain donors that they fall below the detection limit.⁷ That kind of donor-specific detail is exactly what makes primary hepatocytes valuable: they capture the real-world variation in how different people handle drugs.

Drug Transport Studies

Metabolism is only part of the story. Before a drug can be broken down by liver enzymes, it first has to get inside the liver cell. A set of transporter proteins embedded in the hepatocyte membrane actively pulls certain compounds out of the bloodstream and into the cell. If a drug relies on these uptake transporters, understanding their activity is essential for predicting how the drug behaves in the body. Cryopreserved human hepatocytes, which can be thawed and used on demand, are now routinely used to estimate how quickly the liver takes up a given compound and which specific transporter is responsible.⁸

Researchers have developed rapid screening methods using selective substrates and inhibitors for the six main uptake transporters found in human hepatocytes, allowing them to characterize a batch of cryopreserved cells quickly and confirm which transporters are functioning before running an experiment.⁹ This quality-control step matters, because not every batch of frozen hepatocytes performs equally well, and a weak batch could lead to misleading results about how a drug enters the liver.

Predicting Drug-Induced Liver Injury

One of the most consequential uses of primary hepatocytes is screening drug candidates for liver toxicity before they reach patients. Drug-induced liver injury remains a leading cause of drug withdrawals from the market and late-stage clinical trial failures, so any system that can flag a toxic compound early saves both lives and enormous development costs.

Traditional two-dimensional cultures of primary hepatocytes, where cells are plated in a flat layer, have been used for decades for this purpose. But three-dimensional culture formats, where hepatocytes are allowed to form small spherical clusters called spheroids or microtissues, have proven more sensitive. A study testing 110 drugs with known clinical safety profiles found that 3D liver microtissues were better at correctly identifying hepatotoxic compounds than flat-plated primary hepatocytes, while maintaining similar accuracy in correctly clearing safe compounds.¹⁰ A separate platform screening 122 clinical drugs confirmed that a collagen-based 3D primary hepatocyte model outperformed all previously reported in vitro models in its ability to predict liver toxicity.¹¹ The advantage comes from the fact that three-dimensional arrangements better mimic how hepatocytes behave in an actual liver, maintaining higher enzyme activity and better cell-to-cell communication over longer periods.

The De-Differentiation Problem and How Culture Systems Address It

The biggest practical headache with primary human hepatocytes is that they start losing their liver-like properties almost immediately after being placed in a standard flat culture dish. Within days, key metabolic enzymes decline, gene expression shifts, and the cells begin to look and act less like hepatocytes. This process, called de-differentiation, severely limits how long experiments can run and how reliable the results are in late time points.¹²

Three-dimensional spheroid cultures have emerged as one of the most effective solutions. Hepatocytes grown as spheroids remain functionally stable for at least five weeks, continuing to produce albumin and urea and store glycogen in ways that flat cultures cannot sustain.¹³ A metabolomics study confirmed that spheroid cultures kept their metabolic patterns stable over multiple weeks, while flat cultures from the same donors deteriorated rapidly.¹⁴

Microfluidic bioreactors, sometimes called “liver-on-a-chip” devices, take a different approach. These tiny platforms flow culture medium over or around hepatocytes, mimicking the constant blood flow that liver cells experience in the body. In one comparison, hepatocytes grown in a microfluidic biochip for 13 days maintained the ability to metabolize drugs via CYP3A4, a critically important enzyme, while the same cells in a standard dish had virtually no detectable CYP3A4 activity by that point. The metabolic ratio for one test drug was roughly 5,000-fold higher in the biochip after 13 days.¹⁵ Another microfluidic device has been used to predict how quickly the liver would clear specific marketed drugs, using cryopreserved hepatocytes from a single donor.¹⁶ These engineering-driven platforms are steadily closing the gap between what a dish of cells can tell us and what actually happens inside a living person.

Modeling Infectious Liver Disease

Hepatitis B virus infects human liver cells with a stubborn selectivity that makes it difficult to study in most laboratory systems. Primary human hepatocytes are considered the closest model to actual infection because the virus interacts with the same receptors and triggers the same cellular responses it would encounter in a patient’s liver. Research using hepatocytes has revealed that HBV particles activate a specific branch of the innate immune system, the toll-like receptor 2 pathway, immediately upon infection. This triggers the release of inflammatory signaling molecules and activates several downstream pathways, a finding that was confirmed by showing that antibodies blocking the receptor could neutralize the response.¹⁷

A persistent challenge in HBV research is that standard hepatocyte monocultures lose their ability to support viral replication within a couple of weeks. Co-culture systems, where hepatocytes are paired with other liver cell types like hepatic stellate cells, have extended the window for productive HBV infection without needing artificial chemical supplements to keep the virus replicating.¹⁸ These engineered co-culture platforms, where hepatocytes are arranged in micropatterned configurations alongside stromal cells, have supported prolonged infection and have become tools for testing antiviral strategies.

Hepatocyte Transplantation

Beyond the laboratory bench, primary human hepatocytes have been transplanted directly into patients as a treatment strategy. The idea is straightforward: infuse healthy hepatocytes into a patient whose liver is failing or whose liver cells carry a genetic defect, and let the transplanted cells take over some of the missing function. Clinical experience over three decades has shown that the procedure is safe, and results have been most encouraging in patients with liver-based metabolic disorders, conditions where the liver’s structure is intact but a single enzyme or protein is missing or defective.¹⁹

Hepatocyte transplantation has also shown promise as a temporary bridge in acute liver failure, buying time for a patient’s own liver to regenerate or until a donor organ becomes available. However, the therapy has not yet reached the point of being a cure for any condition. The transplanted cells often engraft poorly, meaning they don’t integrate well enough into the recipient’s liver to provide lasting correction. The hostile environment in a severely damaged liver, full of dying and inflamed tissue, makes engraftment particularly difficult.¹⁹ On top of that, the supply of high-quality donor cells depends on the same scarce pool of unused donor organs that whole-liver transplant programs draw from.²⁰

Donor-to-Donor Variability

No two batches of primary human hepatocytes behave identically, because no two human livers are identical. Genetic differences in drug-metabolizing enzymes, the donor’s age, sex, medication history, and the state of the liver at the time of collection all influence how the cells perform once isolated. One study measuring drug metabolism in cryopreserved hepatocytes from multiple donors found that variability was particularly pronounced for certain metabolic pathways, such as sulfation reactions.²¹

This variability is a double-edged sword. On one hand, it makes standardization difficult and means that experiments need to be run across multiple donors to draw generalizable conclusions. On the other hand, it is exactly the kind of variation that exists in real patient populations, and capturing it can be scientifically valuable. A large-scale study using hepatocytes from fifty donors to map how individual differences shape the response to chemical stress found that cells from patients with existing liver disease tended to be less reactive to chemical challenges, likely because they were already under a higher baseline level of stress.²² The presence of cancer elsewhere in the body, interestingly, did not appear to influence the hepatocytes’ sensitivity in any obvious way. Findings like these help researchers decide which donor characteristics need to be accounted for in study design and which can be safely ignored.

Why Animal Hepatocytes Are Not Interchangeable

Drug companies sometimes use hepatocytes from rats, dogs, or monkeys during early-stage research, but cross-species differences in liver metabolism are large enough that animal cells cannot substitute for human ones when predicting what will happen in patients. A direct comparison of rat, pig, and human hepatocytes maintained in the same culture conditions found that rat cells had dramatically higher activity for certain enzymes. For example, rat hepatocytes showed about 28 times the activity of human cells for one key enzyme and roughly 94 times the activity for another. Pig hepatocytes fell in between, and their patterns of testosterone metabolism were closer to human than rat patterns were.²³

Transporter proteins show a similar species gap. The rate at which rat hepatocytes pumped out certain waste compounds was roughly four times faster than in human cells, a difference traced to species-specific variations in a transporter on the bile-facing side of the cell membrane. Meanwhile, human and dog hepatocytes showed efficient activity of a different efflux transporter that was barely functional in rat and monkey cells.²⁴ For arsenic metabolism, dog, rat, and monkey hepatocytes were considerably more efficient at converting inorganic arsenic into its methylated forms than human hepatocytes were, meaning that toxicity predictions based on animal data could substantially underestimate the risk to people.²⁵ These differences underscore why regulators and pharmaceutical companies treat human hepatocyte data as the definitive standard for predicting human drug behavior.

Cryopreservation and Supply Logistics

Fresh hepatocytes are ideal but impractical for most purposes. Liver tissue becomes available unpredictably, surgical schedules do not align with experimental timelines, and shipping live cells across continents is difficult. Cryopreservation, freezing cells in a carefully controlled way for later use, has transformed the field by making hepatocytes available on demand. The standard approach uses a chemical called DMSO to protect cells during freezing, but survival after thawing is often mediocre. Adding a sugar called trehalose to the freezing solution improved post-thaw viability from roughly 47% to about 63%, and more than doubled the proportion of cells that successfully attached to a culture surface afterward.²⁶ The thawed cells treated with trehalose also produced more albumin and showed less membrane damage, indicating better overall health.

Even with improved freezing protocols, cryopreserved hepatocytes do not perform as well as freshly isolated cells, and not every batch survives the freeze-thaw cycle equally. Researchers routinely characterize each batch for enzyme activity, transporter function, and viability before using it in experiments, and commercial suppliers typically provide quality-control data with each lot. China has published what appears to be the first national set of guidelines specifically covering quality control requirements for primary human hepatocytes, addressing technical specifications from isolation through storage and transport. The existence of such standards reflects how widely hepatocytes are now used and how critical consistency has become for regulatory-grade data.

Emerging Platforms Built on Primary Hepatocytes

The field is not standing still. Humanized mouse liver models, where a mouse’s own liver cells are replaced with transplanted human hepatocytes, now offer a way to study human drug metabolism in a living animal. In these systems, the human cells express drug-metabolizing enzymes and transporters at levels comparable to or higher than those seen in cultured primary hepatocytes. The human cells maintained albumin secretion and key enzyme activity for at least four weeks, and they responded to known enzyme-inducing drugs in the expected human-specific patterns.²⁷ These chimeric animals offer something that neither a dish of cells nor a standard lab mouse can: human liver metabolism operating within a whole-body context, with blood circulation, immune responses, and organ crosstalk all present.

Organoid technology represents another frontier. Researchers can coax primary hepatocytes to form self-organizing three-dimensional structures that can be expanded and maintained far longer than traditional cultures. While organoids still require further maturation steps to recover full metabolic function, they offer the tantalizing possibility of expanding a small number of donor cells into a much larger supply, potentially easing the chronic shortage that limits the field. Gene editing of hepatocyte-derived cells has also been explored for correcting metabolic liver diseases, with proof-of-concept work showing that defective hepatocytes from patients can be edited outside the body and then potentially returned, though clinical translation remains early-stage.