Micropatterned Hepatocyte Co-Culture Platform (HepatoPac®) – an in vitro liver model

Home / Sponsored Spotlight / Micropatterned Hepatocyte Co-Culture Platform (HepatoPac®) – an in vitro liver model

Sponsored Spotlight

Micropatterned Hepatocyte Co-Culture Platform (HepatoPac®) – an in vitro liver model

by Katherine Sydney Cook and Okey Ukairo, Hepregen Corporation
Liver models for studying DILI

Drug induced liver injury (DILI) is a major health problem in the United States and accounts for the majority of clinical holds and postmarketing use restrictions by the FDA (Olson et al., 2000; Khetani et al., 2008). The current models for liver toxicity testing in preclinical development seem inadequate. Currently, there is significant interest in the selection of compounds deprived of hepatic liability early during the drug development process. The gold standard for preclinical toxicological evaluation of substances is whole rodent models; however, species-specific variations between rodents and humans can be significant, especially in liver-specific metabolic pathways (i.e. CYP450). This severely limits the utility of animal models for predicting human‐specific responses (Olson et al., 2000).

Several in vitro models such as liver slices, microsomes, cell lines, and isolated primary hepatocytes have been developed to complement in vivo studies to address the lack of translatability from animal data to human responses. However, many of these have limitations. Liver slices have limited viability; cell-free microsomes lack the dynamic gene expression and intact cellular machinery required for drug toxicity assessment; and carcinoma-derived cell lines and immortalized hepatocytes display an abnormal repertoire and levels of liver-specific functions.

Isolated primary hepatocytes are widely considered to be the best choice for in vitro drug/chemical screening applications because they maintain an intact cytoarchitecture. However, the precipitous loss of many liver-specific functions and the decline in cell viability within several days under conventional culture conditions (hepatocyte monolayers and sandwich culture of hepatocytes) hinder the use of these systems for accurate prediction of human liver toxicity. Therefore, there is an urgent need for more predictive liver models.

A new liver model: Hepregen’s HepatoPac® MicroPatterned Platform

Hepregen®, a company in Medford, Massachusetts has developed a micropatterned hepatocyte co-culture platform (MPCCs) that displays phenotypic stability for several weeks in vitro as assessed by major categories of liver-specific functions and gene-expression. In MPCCs, human hepatocytes are organized into islands of optimized dimensions surrounded by supportive stromal cells. Hepatocytes in MPCCs retain their in vivo-like morphology, display functional bile canaliculi, express liver genes, metabolize compounds using active Phase I and Phase II drug metabolism enzymes, and secrete diverse liver‐specific products for up to four weeks in vitro. Repeat dosing regimens in MPCCs in a test set of 45 compounds (35 DILI positive and 10 DILI negative com­pounds) demonstrated that improved sensitivity for DILI prediction can be achieved without loss of specificity (Xu et al., 2008).

Figure 1. Patented HepatoPac micropatterned islands are 500 microns in diameter with a spacing of 1200 microns between islands. Copyright Hepregen Corporation 2015.

Micropatterned co-cultures identify compounds with different mechanisms of toxicity and can be used as a predictive tool in early drug discovery

Liver toxicity is produced by several mechanisms. A predictive in vitro liver platform is one that has high concordance with human liver toxicity and is able to represent the different mechanisms of hepatotoxicity. Hepatocytes in MPCCs are able to retain their in vivo-like morphology and retain sustained hepatic functions for several weeks in vitro. The various mechanisms of hepatotoxicity (reactive metabolite toxicity, biliary transporter toxicity, innate immune mediated toxicity, etc.) are detectable using human MPCCs.

Reactive metabolite-mediated hepatotoxicity

Many types of drugs can be bioactivated leading to forma­tion of reactive metabolites. This metabolic activation (which is catalyzed by liver enzymes) is often the initial event in many chemically-induced toxicities. In a study of compounds withdrawn from the US market due to hepatotoxicity, there was evidence of reactive metab­olite formation in five out of six drugs that were withdrawn (Walgren et al., 2005). Human MPCCs show high expression of Phase l and ll enzymes for several weeks in vitro (Khetani & Bhatia, 2008), thereby possessing the necessary machinery to convert drugs to bioactive species.

The toxicities of cyclophosphamide, troglitazone, isoniazid, and diclofenac (all compounds converted to reactivemetabolites) have been observed in human MPCCs (Ukairo et al., 2013). Using acetaminophen (APAP) as a model toxicant and mimicking cellular glutathione (GSH) depletion by addition of L-Buthionine Sulfoximine (BSO) to cultures, the utility of the human MPCCs for detecting reactive metabolite toxicities can be identified (Figure 2). APAP is primar­ily metabolized by glucuronidation and sulfation. However, its hepatotoxicity is mediated by its toxic metabolite, N-acetyl-p-benzoquinone imine (NAPQI), which is generated by liver cytochrome P450s (CYP2E1 and CYP1A2) and is detoxified by conjugation with hepatic glutathione (GSH). Application of APAP alone to HepatoPac co-cultures (up to 20mM concentration) had little or no adverse effect on cellular ATP content (Figure 2). However, when co-incubated with 200μM BSO, APAP caused concentration-dependent depletion of cellular ATP levels between the ranges of 1.25mM to 20mM. For instance, ATP levels dropped to less than 50% of control levels at 10mM APAP in the presence of BSO. Thus, MPCCs can be used for in vitro screening of bioactivated compounds for selection of safer alternatives before in vivo tests are initiated.

Figure 2. Human HepatoPac cultures detect reactive metabolite toxicity as shown above where acetaminophen (APAP) toxicity is enhanced with cellular glutathione (GSH) depletion. Addition of 200 μ M L-buthionine (S,R)-sulfoximine (BSO), an inhibitor of GSH synthesis, to human MPCCs potentiated acetaminophen-induced ATP depletion in these cultures.

Cholestatic drug induced liver injury

A common mechanism of cholestasis is the inhibition of bile salt excretory pump (BSEP) leading to disruption of bile acid secretion. MPCCs are characterized by extensive canalicular networks (“bile pockets”) (Figure 3), and longer-term retention of transporter functions and can be used to successfully identify cholestasis-mediated hepatotoxins. Treatment of human MPCCs with the cholestasis inducing hepatotoxin, Troglitazone, for 5 days adversely affected hepatocyte health in these cultures. ATP and albumin levels were decreased to less than 10% of control values (Figure 4). Furthermore, troglitazone inhibited the biliary excretion index (BEI) of radiolabeled taurocholate in human MPCCs at low micromolar concentrations (IC50 ~ 7.5uM) indicating an inhibitory effect on the bile salt excretory pump (BSEP). Troglitazone was withdrawn from the market due to severe hepatotoxicity in patients. In human MPCCs, troglitazone was much more acutely toxic than its structural analog, rosiglitazone (Avandia), a drug approved by the FDA (Khetani & Bhatia, 2008). The ability to distinguish between toxic troglitazone and the non- hepatotoxic rosiglitazone shows that human MPCCs can be used reliably to detect hepatotoxic adverse reactions such as cholestasis. The hepatotoxicity of other cholestasis inducing hepatotoxins such as Bosentan and Sitaxsentan has been observed in human MPCCs as well (data not shown).

Figure 3: Human HepatoPac co-cultures form extensive canalicular networks as shown by Multidrug Resistance-associated Proteins 2 (MRP-2) transport of a fluorescent dye (CDCFDA) into the bile canaliculi between hepatocytes. Figure 3also shows robust immunostaining of MRP-2 protein in the canalicular domain.

Figure 4: Exposure of Human MPCCs to increasing concentrations of troglitazone (over 5 days) resulted in down regulation of albumin secretion and decreases in cellular ATP content.

Mitochondrial toxicity

Human MPCCs can be used to successfully identify mitochondrial hepatotoxins. When MPCCs (rat or human) were exposed to increasing concentrations (up to 100- fold Cmax) of Fialuridine, dose-dependent decreases in cellular ATP and mitochondrial activity were observed in human MPCCs but not rat MPCCs. Fialuridine (FIAU) is a nucleoside analog originally developed for the treatment of HBV infections but later withdrawn from the clinic due severe liver toxicity leading to death. FIAU toxicity was not predicted in preclinical animal studies. The putative mechanism of hepatotoxicity of FIAU is related to the incorporation and accumulation of FIAU into mitochondrial DNA, leading to disruption of mitochondrial Polγ activity and ultimately to reduced metabolic capacity and cell death. The ability to detect FIAU toxicity in human but not rat MPCCs indicates that MPCCs can be used to detect species differences in hepatotoxicity caused by mitochondrial impairment.

Figure 5: Treatment of MPCCs with increasing concentrations of FIAU (4 doses over 9 days) caused concentration- dependent decreases in cellular ATP and mitochondrial activity in human MPCCs but not rat MPCCs. Mitochondrial activity was evaluated using the MTT assay.

Immune-mediated toxicity

Hepregen has also developed an in vitro model consisting of a co-culture of Kupffer macrophages, primary hepatocytes, and murine 3T3 J2 Fibroblasts called HepatoMune™. Kupffer cells in the HepatoMune model respond to modulators such as lipopolysaccharide (LPS), and remain viable for several days in this triculture format. Proof-of-concept utility of HepatoMune co-cultures has been demonstrated in the evaluation of innate immune-mediated toxicity using trova­floxacin as a model compound. When co-administered with LPS, trovafloxacin toxicity was potentiated in HepatoMune co-cultures. The LPS-induced exacerbation of trovafloxacin toxicity was reversed by pre-treatment of the co-cultures with pentoxifylline (an inhibitor of TNF-α transcription) (data not shown) consistent with literature evidence that trovafloxacin toxicity is, in part, mediated by TNF-α. Treatment of HepatoMune co-cultures with LPS and Levofloxacin (a non-toxic analog of Trovafloxacin) had no adverse effect on the co-cultures.

Figure 6. Trovafloxacin (TVX) toxicity is potentiated in LPS-treated Human HepatoMune. (A) LPS-stimulation of HepatoMune cultures exacerbated TVX-induced toxicity, marked by the leftward shift in the ATP dose-response curves. (B) Co-administration of LPS and levofloxacin had no adverse effect on hepatocyte health.


Drug Induced Liver Injury (DILI) is often identified extremely late in drug development, when enormous amounts of time and money have been spent on a potential drug. Predictive in vitro tools such as human MPCCs, when deployed early in drug development, may help reduce the number of late stage failures thereby saving money and time.

Author Information

Katherine Sydney Cook
Hepregen Corporation
200 Boston Ave. Suite 1500
Medford, MA 02155 USA
Email: ksydneycook@hepregen.com

Okey Ukairo, Ph.D.
Hepregen Corporation
200 Boston Ave., Suite 1500
Medford, MA 02155 USA
Email: oukairo@hepregen.com

Posted: March 21, 2015