Chemico-Biological Interactions 168 (2007) 16–29

Human hepatocytes: Isolation, cryopreservationand applications in drug development

Albert P. Li a,b,∗a The ADMET Group LLC and In Vitro ADMET Laboratories LLC, 15235 Shady Grove Road,

Suite 303, Rockville, MD 20850, United Statesb Advanced Pharmaceutical Sciences Inc., 9221 Rumsey Road, Suite 8, Columbia, MD 21045, USA

Available online 9 January 2007

Abstract

The recent developments in the isolation, culturing, and cryopreservation of human hepatocytes, and the application of the cells indrug development are reviewed. Recent advances include the improvement of cryopreservation procedures to allow cell attachment,thereby extending the use of the cells to assays that requires prolong culturing such as enzyme induction studies. Applicationsof human hepatocytes in drug development include the evaluation of metabolic stability, metabolite profiling and identification,drug–drug interaction potential, and hepatotoxic potential. The use of intact human hepatocytes, because of the complete, undisruptedmetabolic pathways and cofactors, allows the development of data more relevant to humans in vivo than tissue fractions such ashuman liver microsomes. Incorporation of key in vivo factors with the intact hepatocytes in vitro may help predictive human in vivodrug properties. For instance, evaluation of drug metabolism and drug–drug interactions with intact human hepatocytes in 100%human serum may eliminate the need to determine in vivo intracellular concentrations for the extrapolation of in vitro data to in vivo.Co-culturing of hepatocytes and nonhepatic primary cells from other organs in the integrated discrete multiple organ co-culture

(IdMOC) may allow the evaluation of multiple organ interactions in drug metabolism and drug toxicity. In conclusion, humanhepatocytes represent a critical experimental model for drug development, allowing early evaluation of human drug properties toguide the design and selection of drug candidates with a high probability of clinical success.© 2007 Elsevier Ireland Ltd. All rights reserved.

opreserv

Keywords: Human hepatocytes; Hepatocyte isolation; Hepatocyte cryDrug toxicity; Toxicogenomics; IdMOC

1. Introduction

Primary cultures of hepatocytes represent an experi-mental tool that has been used extensively in biomedicalresearch, both in academia as well as for commercialpurposes such as drug development. The most excit-

ing advances are the successful isolation, culturing,and cryopreservation of hepatocytes from human livers.Human hepatocytes are used routinely in drug develop-

∗ Tel.: +1 410 869 9037; fax: +1 410 869 9560.E-mail address: lialbert@apsciences.com.

0009-2797/$ – see front matter © 2007 Elsevier Ireland Ltd. All rights reservdoi:10.1016/j.cbi.2007.01.001

ation; Drug development; Drug metabolism; Drug–drug interactions;

ment as an experimental model for the evaluation of keyhuman-specific drug properties such as metabolic fate,drug–drug interactions, and drug toxicity. The applica-tions range from the early screening the most appropriatenew chemical entities for further development, to thedetermination of key drug properties for New DrugApplications (e.g. drug–drug interactions) to U.S. FDA.

Successful application of human hepatocytes in drugdevelopment requires a thorough understanding of the

strengths and weaknesses of this valuable experimentalsystem. This review is an effort to present a compre-hensive review on the state-of-the-art of the isolation,cryopreservation, and applications of human hepatocytes

ed.

al Inter

if

2

ntprt

12345678

usobvwcl

upnptiiltlssiliewuyasi

A.P. Li / Chemico-Biologic

n drug development, with emphasis on the research per-ormed in our laboratories in the past 20 years.

. Human hepatocyte isolation

One of the major advances in human hepatocyte tech-ology is the availability of human livers for research. Inhe United States, livers procured but not used for trans-lantation are allowed to be used in research. The majoreasons that procured livers are not used for transplanta-ion are as follows:

. unavailability of a matched recipient;

. physical damage to the liver;

. pre-existing liver diseases;

. breach of sterility during the procurement process;

. high liver fat content;

. inappropriate age (too young or too old);

. inappropriate warm ischemic time;

. inappropriate cold storage time.

Livers that are procured for transplantation aresually flushed extensively with a cold preservationolution, with the University of Wisconsin (UW)r histidine–tryptophan–ketoglutarate (HTK) solutionseing the most common [1]. The rule of thumb is thatiable hepatocytes can be obtained from human liversith up to 24 h of cold preservation. There are, however,

ases that highly viable hepatocytes can be obtained fromivers stored beyond this 24-h period.

Hepatocyte isolation from human livers is nowniversally performed with a “two-step” collagenaserocedure developed by Berry and Friend [2]. Origi-ally developed for the isolation of rat hepatocytes, thisrocedure has been modified by various laboratories forhe isolation of hepatocytes from sevreal animal species,ncluding human (e.g. [3,4]). The procedure involves thenitial perfusion of the liver with a warm (37 ◦C) diva-ent ion-free, EGTA-containing, isotonic buffer (Step 1)o remove blood and to loosen cell–cell junctions, fol-owed by perfusion with a warm, isotonic, collagenaseolution (Step 2) to dissociate the liver parenchyma intoingle cells. In general, a higher amount of collagenases required for the isolation of hepatocytes from humanivers than that required for rat livers. As collagenases a mixture of proteases, its composition can affect itsffectiveness in the dissociation of the hepatocytes asell as its cytotoxicity. It is a common practice to eval-ate multiple lots of collagenase to select the one lot

ielding the highest number of viable hepatocytes fromliver. After digestion, the cells are harvested by low-

peed centrifugation. A density gradient such as Percolls commonly used to enrich for viable cells. The isolated

actions 168 (2007) 16–29 17

cells can be used in suspension for experiments requir-ing a relatively short time duration (hours), plated ontissue culture surfaces pretreated with attachment sub-strates (e.g. collagen; Matrigel) for longer term studies,or cryopreserved for future use.

The method of isolation of human hepatocytes froma human liver is by no means optimized. Currently, a so-called “good” yield of human hepatocytes from a humanliver is approximately 10–30 billion viable cells when awhole liver is perfused. Using an approximation of 1.5 kgas an average weight of a human liver, this leads to a yieldof or approximately 7–20 million hepatocytes per gramof liver (e.g. [5,6]), which is considerably less that thetotal number of hepatocytes (approximately 300 billion)in the human liver. It is to be noted that the yield ofhuman hepatocytes (in terms of number of hepatocytesper gram liver) is in general higher from smaller (e.g.,10 to 300 g) liver fragments then whole livers or lobes.

3. Cryopreservation

Hepatocytes, especially human hepatocytes, are nowroutinely used after they are cryopreserved [7,8]. Thegeneral procedures for hepatocyte cryopreservation havenot deviated extensively from the original procedures[9]. Via the use of equipments to control freezing rates(e.g. programmable control-rate freezer) and appropriatecryopreservation agents (e.g. dimethyl sulfoxide), hepa-tocytes now can be stored in liquid nitrogen (lower than−150 ◦C) for an extensively time period (years) withthe retention of high viability and drug metabolizingenzyme activity [7]. The most recent advancement ofhuman hepatocyte cryopreservation is the ability of thethawed hepatocytes to be plated as monolayer cultures(“plateable” hepatocytes) [10,11]. In our laboratory, weroutinely prepare cryopreserved human hepatocytes withmost of the lots having post-thaw viability of >90%(Table 1) and with approximately half of the lots yield-ing over 50% confluent monolayer cultures when platedonto collagen-coated plates (Table 1; Fig. 1). In ourlaboratory, we believe that the key to successful cryop-reservation is to ensure that the hepatocytes are isolatedfrom the human liver with minimal damages to theplasma membrane.

It is important to fully understand the propertiesof cryopreserved human hepatocytes as compared tofreshly isolated cells. The following are general con-clusions on the drug metabolizing enzyme activities of

cryopreserved human hepatocytes:

1. P450 isoform activities: Since the first publicationshowing similar P450 isoform activities between

18 A.P. Li / Chemico-Biological Interactions 168 (2007) 16–29

Table 1Post-thawed viability and attachment efficiency of human hepatocytes isolated and cryopreserved in APSciences Inc.

Date of preparation Lot Yield (cells/vial) Viability, % (trypan blue) Plateability Confluency (%)

8 September 2006 HU4019/20/21 5.4 × 106 89 Yes 7013 September 2006 HU4017/18/22 5.5 × 106 91 Yes 8014 September 2006 HU4026 5.9 × 106 91 No <1028 September 2006 HU4027 5.9 × 106 92 No 30-403 October 2006 HU4028 3.2 × 106 83 No 30-405 October 2006 HU4023 2.1 × 106 89 No 208 October 2006 HU4024/25/29 8.5 × 106 88 Yes 8018 October 2006 HU4030 3.5 × 106 88 Yes 5030 October 2006 HU4032 4.8 × 106 99 Yes 7030 October 2006 HU4034 5.8 × 106 97 Yes 85

ryoprese0%, wi

In September and October of 2006. Using proprietary procedures for chave post-thaw viability values (based on trypan blue exclusion) of >8cultures (with >50% confluency at 24-h after plating).

freshly and cryopreserved human hepatocytes [7],there are various independent confirmation of thisobservation (e.g. [10,13]). In general, the majorhuman P450 isoform activities: CYP1A2, CYP2A6,CYP2B6, CYP2C9, CYP2C19, CYP2D6, CYP2E1,and CYP3A4 are retained in cryopreserved humanhepatocytes. Direct comparison of activities betweenfreshly isolated and cryopreserved human hep-

atocytes from the same preparation showed nosubstantial differences [7].

2. UDP-dependendent glucuronosyl transferase (UGT)and sulfotransfase (ST) activities: UGT and ST are

Fig. 1. Photomicrograph of human hepatocytes cultured after cryop-reservation. Historically, human hepatocytes in general would losetheir ability to attach for monolayer culturing. One rule of thumb is thatapproximately 1 out of 10–20 lots (each lot representing hepatocytesfrom a single isolation) of cryopreserved human hepatocytes lots wouldbe “plateable”. In our laboratory, we have developed proprietory tech-nologies with which approximately 50% of the cryopreserved humanhepatocytes would be “plateable” as defined by their ability to formmonolayer cultures with over 50% confluency (photograph obtainedfrom CellzDirect Inc.).

rvation and post-thaw recovery, the human hepatocytes were found toth over 50% of the lots found to be “plateable” monolayer hepatocyte

two major Phase II metabolism pathways, leadingto conjugation of xenobiotics or metabolites withthe highly water-soluble glucose (UGT) and sulfate(ST). There is no apparent deterioration of thesePhase II conjugating pathways upon cryopreservation([13,14]).

3. Glutathione (GSH) level and glutathione transferaseactivities (GST): It is now well-established that whileGST activities are similar between freshly isolatedand cryopreserved human hepatocytes, intracellularGSH levels are drastically reduced upon cryopreser-vation and thawing [15].

4. Transporter activities: Transporter research is thelatest exciting area in drug metabolism. Trans-porter activities are now known to be involved indrug metabolism, drug–drug interactions, and drugtoxicity. Some lots of cryopreserved human hep-atocytes are known to retain transporter activitiesand therefore can be used for transporter research.The transporters that have been reported to beactive in human hepatocytes include human Na(+)-taurocholate cotransporting polypeptides (NTCP)and organic anion transporting polypeptides (OATP)[16,17].

The general consensus is that cryopreserved humanhepatocytes can be used routinely for the evaluationof drug metabolism ([7,8,14]). The strengths of thecryopreserved human hepatocytes over freshly iso-lated hepatocytes as an experimental model are asfollows:

1. Ease of experimentation: Unlike freshly isolated hep-atocytes which is dependent on the availability of ahuman liver for research, experimentation with cry-opreserved human hepatocytes can be planned.

al Inter

2

3

1

2

3

4d

fd

A.P. Li / Chemico-Biologic

. Repeat experimentation: Cryopreserved hepatocytesfrom a single donor (or combination of multipledonors) can be used at different times to allow theperformance of repeat studies. This is not possiblewith freshly isolated hepatocytes.

. Choice of donors: Experimentation with freshly iso-lated cells are performed with the liver from thedonor available at a specific point of time. Limitingthe choices of donors may lead to a prolonged wait-ing period for experimentation. Experimentation withcryopreserved human hepatocytes, however, allowsthe researcher to select hepatocytes from donorswith properties most appropriate to the experimentalobjectives.

The following are the limitations:

. Incubation time for metabolism studies: Most lots ofcryopreserved human hepatocytes cannot be culturedas attached, monolayer cultures. For these cells thatcannot be cultured, their applications will be limitedto those compatible with suspension cultures. Hepa-tocytes in suspension have a finite lifespan (hours),and probably should not be used for longer than 5 h.For hepatocytes that can attach, their viability is pro-longed (days). It is known, however, that activities forP450 isoforms such as CYP1A2 and CYP3A4 woulddecrease (approximately 50% per day) in culture. Itis probably most appropriate to use the “plateable”cryopreserved hepatocytes for not more than 3 daysin culture for metabolism studies.

. GSH conjugation: Because of the decreased intra-cellular GSH level in cryopreserved hepatocytes,GSH conjugation may be limited in this experi-mental system, especially when used as short-termincubations in suspension. It may be more appropri-ate to use freshly isolated or cultured cryopreservedhuman hepatocytes (e.g. for 24 h to allow recov-ery of intracellular GSH) for the evaluation of GSHconjugation.

. Transporter studies: Not all lots of human hepato-cytes retain intact transporter activities. It would beprudent to evaluate multiple lots of cryopreservedhuman hepatocytes to select the lots with active trans-porters.

. Application of human hepatocytes in drugevelopment

The key areas in drug development that can benefitrom the use of human hepatocytes are drug metabolism,rug–drug interaction, and drug toxicity evaluation.

actions 168 (2007) 16–29 19

4.1. Drug metabolism

As liver is the major organ for drug metabolism, hep-atocytes represent a valuable experimental system fordrug metabolism studies. Unlike subcellular fractions,intact hepatocytes contain all the drug metabolizingenzymes and cofactors at physiological levels. The keyexperiments in drug metabolism are as follows:

4.1.1. Metabolic stabilityEstimation of intrinsic hepatic clearance is an impor-

tant discipline in drug development. It allows theselection of drug candidates with the most appropriatemetabolic stability in vivo. In the past, metabolic sta-bility studies are performed predominantly with livermicrosomes supplemented with NADPH+ cofactors,with the assumption that Phase I oxidation representthe most important pathway for metabolic clearance.This is obviously not universally true as a large num-ber of structures are metabolized by non-microsomalenzymes (e.g. cytosolic dehydrogenases, mitochondrialmonoamine oxidases), and not always by Phase I oxida-tion (e.g. UGT and ST conjugating activities).

Intrinsic hepatic clearance studies with hepatocytesare performed via the incubation of the chemical in ques-tion with intact hepatocytes and the quantification of thechemical at various times after incubation. Intrinsic hep-atic clearance (CL) can be estimated using one of thefollowing three equations:

1. CL = {Vmax/(Km + [S])}/number of hepatocytes (or=(Vmax/Km)/number of hepatocytes, if [S]≪Km);

2. CL = 0.693/(t1/2 × hepatocyte concentration);3. CL = (initial concentration)/(AUC × hepatocyte con-

centration).

CL is in general expressed in the units of ml/min/million hepatocytes. CL estimation using equation 1requires the performance of the study with multiple con-centrations of the drug substrate, while CL estimationusing equation 2 or 3 requires a simpler procedure of theincubation of hepatocytes with a single concentrationof the drug substrate. In most laboratories, the simplerprocedure of incubation of hepatocytes with a single con-centration of the drug substrate is used with CL derivedusing equations 2 or 3. The results can further scaled upto hepatic intrinsic clearance per liver and per humanusing the following assumptions:

1. hepatocytes (1.2 million) per gram of human liver[18];

2. liver (1.8 kg) per 70 kg human body weight [19].

al Inter

20 A.P. Li / Chemico-Biologic

Hepatocytes are usually used as suspension cultures,with a cell concentration of 0.5–2.0 million cells/ml, andincubation times up to 5 h. Cryopreserved human hep-atocytes used as suspension cultures in a 96-well plate,using 150,000 cells/well, have been shown to be useful asa screening assay for human hepatic metabolic stabilityduring early drug discovery [48].

Due to the known decreases in drug metabolizingenzyme activities with primary cultures of hepatocytes, itis generally recommended that metabolic stability stud-ies to be performed with freshly isolated hepatocytes orthawed cryopreserved hepatocytes from fresh isolates. Ifprolonged incubation (e.g. days) is required, it is recom-mended that freshly isolated or plateable cryopreservedhepatocytes are allowed a short attachment period (e.g.4 h) and then used for incubation with the drug substrate.It is not advisable to use hepatocytes that have been cul-tured for over 48 h for the estimation of in vivo hepaticclearance (e.g. [20,21]).

An interesting advancement in the application ofintact hepatocytes in the evaluation of metabolic stabil-ity is the incubation of the test article and hepatocytes in100% human serum, therefore providing an experimen-tal condition similar to humans in vivo. For instance,it has been reported that in vivo hepatic clearance canbe predicted more accurately from data obtained withhepatocytes incubated in serum than from data obtainedin the absence of serum using rat hepatocytes [20] andhuman hepatocytes [21,22]. The combination of humanhepatocytes and 100% human serum may represent amore physiologically relevant experimental model. Iteliminates the need to consider plasma protein bindingand intracellular concentration of the chemical stud-ied for the estimation of in vivo metabolic rates. Thisapproach therefore is conceptually attractive and shouldbe carefully evaluated for universal applications. Onepuzzling observation made in our laboratory was thattestosterone, a naturally occurring steroid that is read-ily metabolized in human in vivo, was metabolizedby cryopreserved human hepatocytes in suspension inthe absence of serum, but was virtually not metab-olized in the presence of 100% serum (unpublishedobservation).

Metabolic stability studies are usually conducted dur-ing earlier stages of drug development to allow theselection of structures with the most appropriate stabil-ity for further development. Because of the presence ofall hepatic drug metabolizing enzymes and cofactors at

physiological levels, intact hepatocytes, especially cry-opreserved human hepatocytes represent a more relevantexperimental system than human liver microsomes forgeneral screening for metabolic stability. Human liver

actions 168 (2007) 16–29

microsomes represent a relevant system if microsomalenzymes are known to be responsible for the metabolismof the chemicals in question.

4.1.2. Metabolite profiling, metabolite identificationand species comparison

Metabolism is a determinant of metabolic stability asdiscussed above, and also a key aspect of drug toxicity.A chemical can be “activated” from a nontoxic parentmolecule to toxic metabolites. Conversely, a toxic parentchemical can be “detoxified” via biotransformation (e.g.by Phase II conjugation) to nontoxic metabolites.

Metabolite identification therefore is an importantpart of drug development. Knowing the structure of themetabolite, one can design structures with improvedmetabolic stability and toxicity, for instance, by elim-inating or blocking the critical sites of metabolism.

Early species–species comparison of metabolite pro-filing is also critical to drug development to allow theselection of the most appropriate animal species for pre-clinical drug development. The appropriate species willbe the ones with metabolite profiles similar to human.

Human liver microsomes are generally used inmetabolite profiling. As discussed earlier, this approachassumes that Phase I oxidation is the most importantpathway of the metabolism of the chemical entity inquestion. This is an appropriate approach if one is cer-tain that the predominant route of metabolism is PhaseI oxidation by microsomal pathways. If this is notwell-established, intact hepatocytes represent a more sci-entifically relevant experimental system.

The use of human liver microsomes versus hepa-tocytes in metabolite profiling can be illustrated withethynyl estradiol (EE2) [12]. EE2 is predominantlymetabolized by human in vivo via Phase II metabolisminto glucuronide and sulfate conjugates. However,liver microsomes produce primarily the 2-hydroxylationmetabolite. Human hepatocytes produce the conjugatemetabolites, similar to human in vivo. The results withEE2 illustrate the use the human hepatocytes to providean initial evaluation of the major metabolites, while theuse of liver microsomes allows the evaluation of specificPhase I oxidative metabolites.

4.2. Drug–drug interactions

A major advancement of the application of humanin vitro drug metabolism systems is its application in

drug–drug interaction (DDI) studies. U.S. FDA nowrequires in vitro DDI studies to be performed for NewDrug Applications (NDA) [23–25]. It is proposed byU.S. FDA that, if the in vitro DDI studies show no poten-

al Inter

tb

1

2

3

4

etvmpciFPmnm

4

sImIw

A.P. Li / Chemico-Biologic

ial interactions, clinical DDI studies are not required toe performed [25].

DDI studies involve the following major approaches:

. Determination of key drug metabolism pathways:Elucidation of the key pathway of metabolism of adrug would allow the estimation of its interaction withexisting drugs that are known to inhibit or induce thepathways involved.

. Evaluation of drug metabolizing enzyme inhibitorypotential: Evaluation of the ability of a drug to inhibitthe major human drug metabolizing enzymes (e.g.P450 isoforms) would allow the estimation of itsDDI potential with co-administered drugs that aresubstrates of the affected pathway. The result ofinhibitory DDI is the unwarranted increase in sys-temic burden of the affected drug which may lead tosevere toxicity (e.g. ketoconazole/terfenadine inter-actions [26]).

. Evaluation of P450 induction potential: Evaluation ofthe ability of a drug to induce the major human drugmetabolizing enzymes (e.g. P450 isoforms) wouldallow the estimation of its DDI potential with co-administered drugs that are substrates of the inducedpathway. The major consequence of inductive DDI isthe lowering of the systemic burden to the affecteddrug due to accelerated metabolic clearance, leadingto loss of efficacy (e.g. cyclosporine/rifampin inter-actions [27]).

.2.1.1. Metabolic pathway elucidationIn general, human liver microsomes and cDNA

xpressed microsomes are used for pathway elucida-ion studies. Human hepatocytes, however, can providealuable information which will complement livericrosomal data. One key experiment that is not always

erformed, but should be performed, for pathway elu-idation of the metabolism of a test article is thedentification of metabolites from intact hepatocytes.or instance, if the major metabolites are results ofhase II conjugation, then evaluation with human livericrosomes or genetically expressed microsomes are

ot necessary. This initial experiment is in fact recom-ended by FDA [25].

.2.1.2. P450 inhibitionAs for pathway elucidation, liver or cDNA micro-

omes are routinely used for P450 inhibition studies.

t is now realized that the use of intact hepatocytesay allow a more relevant estimation of in vivo effects.

ntact hepatocytes provide an intact cell membrane,hich allows the modeling of differential uptake of the

actions 168 (2007) 16–29 21

test article into the hepatocytes. Using cell free sys-tems such as liver microsomes, one usually estimatein vivo effect (e.g. by the ratio of plasma Cmax ([I])to the inhibitory constant, Ki, with [I]/Ki of <0.1 con-sidered as unlikely to yield DDI [25]) using plasmaconcentration assuming that it is equal to the concentra-tion of the inhibitor at the site of the inhibited enzyme.This may not be correct if the inhibitor has a low per-meability across the plasma membrane (leading to anintracellular concentration lower than plasma concen-tration), or if the inhibitor is actively uptaken (leadingto an intracellular concentration higher than the plasmaconcentration).

It was discussed earlier that intact hepatocytescultured in 100% human serum may provide physio-logically relevant data for metabolic clearance studies.The same approach has also been evaluated forinhibitory DDI studies [28]. Lu et al. applied a novelapproach using cryopreserved human hepatocytes sus-pended in human plasma for the evaluation of P450inhibitory effects. They proposed that this approachwould allow prediction of P450 inhibitory potential invivo without the need of estimating the inhibitor con-centrations at the enzyme active site or the Ki. Theyreported success in the estimation of the magnitudeof the clinical DDI of ketoconazole, an investiga-tional compound MLX, and several marketed drugs(theophylline, tolbutamide, omeprazole, desipramine,midazolam, alprazolam, cyclosporine, and loratadine),with a correlation coefficient (r2) of 0.992.

4.2.1.3. Enzyme inductionAs described earlier, induction of drug metabolizing

enzyme is a second major mechanism of pharmacoki-netic drug–drug interactions. P450 isoforms representthe most important drug metabolizing enzymes forenzyme induction studies.

Primary human hepatocytes have always been the“gold standard” for enzyme induction studies. As of thiswriting, the following statement apparently is true:

“All known human P450 inducers in vivo are inducersin human hepatocytes in vitro”

The reverse of this statement, i.e., if all in vitro induc-ers are in vivo inducers, is not yet established.

It has been long established that primary humanhepatocytes are known to be responsive to prototypi-cal P450 inducers such as omeprazole (for CYP1A2)

and rifampin (for CYP3A4) [29,30,49]. It is interest-ing that while other P450 isoforms such as CYP2A6,CYP2B6, CYP2C8, CYP2C9, CYP2C19 and CYP2E1are also inducible (e.g. [30]), there is not a known

22 A.P. Li / Chemico-Biological Interactions 168 (2007) 16–29

al Inter

iwiit3vtCitwdostet[go6afo

wooti[bctrhs

dfqoiaa

FiirI

A.P. Li / Chemico-Biologic

nducer which would selectively induce these isoformsithout also inducing CYP3A4. For instance, a potent

nducer of CYP2B6, phenobarbital, is also a potentnducer of CYP3A4 in human hepatocytes [30]. In con-rast, CYP1A2 inducers in general (e.g. omeprazole,-methylcholanthrene) would not induce CYP3A4 andice versa. For the investigation of P450 induction poten-ial of drug candidates, it may be adequate to evaluateYP1A2 and CYP3A4 induction. If CYP3A4 induction

s observed, then investigations on the induction poten-ial of the drug candidate towards other isoforms arearranted. If no CYP1A2 or CYP3A4 is observed, therug candidate is not likely to have induction effectsn the other P450 isoforms, and therefore can be con-idered not to be a potential P450 inducer. As ofhis writing, isoniazid is the only inducer that appar-ntly would induce CYP2E1 but with little effects onhe other isoforms, including CYP1A2 and CYP3A430]. It is to be noted while induction of CYP2E1ene expression can be readily detected, inductionf CYP2E1 activity, (e.g. measured as chloroxazone-hydroxylation activity) by inducers, including isoni-zid and ethanol, is not a reproducible phenomenonor human hepatocytes in culture (Li, unpublishedbservation).

A major advancement in the P450 induction studiesith human hepatocytes is the use of plateable cry-preserved human hepatocytes. The first observationf the responsiveness of cryopreserved human hepa-ocytes to P450 inducers made by Ruegg et al. [31]s now substantiated by results of several laboratories10,32,33], including ours (Fig. 3). It is now generallyelieved that there is little qualitative difference betweenryopreserved and freshly isolated human hepatocytesowards known P450 inducers. U.S. FDA has recentlyecommended the use of plateable cryopreserved humanepatocytes for the evaluation of P450 induction for DDItudies [25].

P450 activity measurement remains the “gold stan-ard” for P450 induction studies as activity is responsibleor DDI. Surrogate endpoints such as mRNA and proteinuantification are also used, especially for the evaluation

f induction potential for a compound which is also annhibitor of the activity of the induced P450 (e.g. riton-vir [34]). As of now, FDA considers activity as the mostcceptable endpoint for the definition of DDI potential

ig. 2. Heat map summarizing the toxicogenomics results of the treatment ofts less toxic analogs, rosiglitazone and pioglitazone. The results of genes than A and B, respectively. Genes are grouped by physiological mechanisms of aesponse is defined as >50% induction or suppression over solvent control. Mnc. (data reconstructed from [43]).

actions 168 (2007) 16–29 23

due to P450 induction [25]. A positive inducer has beendefined in practical terms by FDA as the following:

>2× basal activity (activity of solvent control) or >40%of positive control activity.

4.3. Transporter studies

The field on hepatic transporters has received greatattention lately, mainly due to the activities of sev-eral key leading scientists in this area who show thattransporter activities can be critical to drug absorption,metabolism, and drug–drug interactions. Hepatocytes invivo have both uptake and efflux transporters, with theuptake transporter responsible for the active uptake ofbiomolecules and xenobiotic, and the efflux transportersfor the bile excretion of bile salts and conjugated xeno-biotics.

Hepatocytes have been used in suspension to evalu-ate the activities of uptake transporters. For this type ofstudies, hepatocytes are incubated in suspension with thetransporter substrate, followed by centrifugation of thehepatocytes through silicon oil to remove extracellularsubstrates, thereby allowing accurate quantification ofintracellular or uptaken substrates.

As with drug metabolizing enzymes, it is now knownthat there are substantial species differences in hepatictransporters. Intact human hepatocytes therefore repre-sent a convenient experimental system for the evaluationof human transporter activities. Cryopreserved humanhepatocytes are now known to active uptake transporters[35], and can be cultured to expressed efflux transporters[36].

An elegant application of intact hepatocytes to evalu-ate transporter-mediated DDI was the study of Shitaraand coworkers [16]. The study was performed withcryopreserved human hepatocytes to evaluate if theknown DDI between cerivastatin (CER) and cyclosporinA (CsA) was due to drug metabolism or transporter-mediated hepatic uptake. They found that CER uptakewas inhibited by CsA with Ki values approximately 10Xlower than the estimated Ki values the inhibition of

CER metabolism. From these results, it was concludedthat the DDI between CER and CsA is mainly due tothe inhibition of transporter (at least partly OATP2)-mediated uptake in the liver. This study represents one

primary human hepatocytes with hepatotoxic drug, troglitazone andt are differentially suppressed and induced by troglitazone are shownction, and each line represents the results of a single gene. Significanticroarray used was the TOX-chip from Phase I Molecular Toxicology

al Inter

24 A.P. Li / Chemico-Biologic

of the first to define transporter-mediated DDI interac-tion with human hepatocytes. Several subsequent studieswere performed based on the methods developed in thispivotal study, providing critical information concerninghuman drug–transporter interactions [17,37,38].

While the study of transporter-mediate uptake canbe evaluated with hepatocytes in suspension, the studyof efflux transporter is more difficult. It is now knownthat the efflux transporters in hepatocytes are “internal-ized” in hepatocytes in suspension, and therefore cannotbe studied effectively (e.g. by evaluating the excretionof a compound into the extracellular medium). Hepato-cytes are found to form functional bile canaliculi whencultured between two layers of collagen gels (collagensandwich), an observation firstly observed in rat hepato-cytes [39,40] and later extended to cryopreserved humanhepatocytes [36]. It is to be noted that while most effluxtransporters are located in the liver canaliculi, there isalso evidence that the transporters that are responsi-ble for the uptake of some drugs may be involved inthe sinusoidal efflux of xenobiotics. For instance, therat Oatp-1, Oatp-2 have been identified as efflux trans-porters for reduced glutathione, and therefore may alsohave activities towards the efflux of certain drugs [41].

Human hepatocytes, used as either freshly isolatedor cryopreserved cells, therefore can be use effectivelyfor the evaluation of hepatic transporter functions. Animportant application of human hepatocytes will bethe evaluation of the complicated roles of drug uptake,metabolism and efflux on drug metabolism, drug–druginteractions, and drug toxicity.

4.4. Drug toxicity

An important application of human hepatocytes isto evaluate human-specific hepatotoxicity that may notbe possible with nonhuman preclinical animal species.Unanticipated drug toxicity, especially hepatotoxicity,has always been a major challenge of the drug indus-try [42]. As drug metabolism is known to be a criticalparameter for hepatotoxicity, and that species–speciesdifferences in drug metabolism is a well-established phe-nomenon, human hepatocytes may represent a relevantin vitro experimental system for the evaluation of humanhepatotoxic potential of drugs and drug candidates.

In vitro toxicity assays with human hepatocytes canbe applied in various during phases of drug development[44,50]. The two major applications are as follows:

4.4.1. Early screening of intrinsic toxicityHuman hepatocytes, especially cryopreserved hepa-

tocytes, can be used for rapid screening of new chemical

actions 168 (2007) 16–29

entities to allow the selection of structures with low hep-atotoxic liability for further development. The screeningassay can allow logical evaluation of structures respon-sible for toxicity (toxicophore) which, hopefully, can beseparated from structures for pharmacological activity(pharmacophore). Toxicity screening with human hep-atocytes (e.g. using ATP content as an endpoint [7])can be performed in 96-well or 384-well plate formats,requiring as little as 1500 cells/well, and would need arelatively small amount of test articles (e.g. 0.1 mg), andis rapid and quantitative.

4.4.2. Mechanistic evaluationsBesides screening, in vitro experimental systems such

as human hepatocytes can be used to define toxic mech-anisms. The defined experimental conditions and theavailability of reagents and approaches for multiple end-points allow one to define the key pathways involvedin a toxicological phenomenon. High-content endpointssuch as toxicogenomics [43,45] have been applied inhuman hepatocytes for the evaluation of toxic mech-anisms. Mechanistic understanding is critical to drugdevelopment. It allows a better understanding of humanhealth risks, defines potential risk factors, and the rela-tionship between efficacy and adverse effects.

4.4.3. ToxicogenomicsHigh-content assays such as toxicogenomics for gene

expression, proteomics for protein expression, metabo-nomics for natural metabolite synthesis, and cellomicsfor cell morphology changes, are relative novel tech-nologies with potential application in the definition ofdrug toxicity. It is anticipated that these high-contentassays, which allow one to evaluate the effect of a testsubstance on myriad events (e.g. expression of 20,000human genes), one has the potential to achieve the fol-lowing:

1. discovery of response patterns (response profiles) thatcan be used to define toxicity;

2. discovery of new biomarkers for specific types oftoxicity;

3. discovery of early indicators of chronic toxicity;4. discovery of indicators of idiosyncratic toxicity.

The high-content assays may be extremely valuableto the definition of toxicity difficult to evaluate with con-ventional approaches. For instance, prediction of chronic

toxicity in human that eludes detection in preclinicalnonhuman animals may be possible using human cells invitro (e.g. human hepatocytes for hepatotoxicity), pro-vided that the key events for chronic toxicity can be

al Inter

di

ctmcOimptcIicd

tat

Fteoro

A.P. Li / Chemico-Biologic

efined in the target cells after treatment with the toxicantn vitro.

Another potential valuable application of high-ontent assays is to the definition of idiosyncratic drugoxicity. Efforts to use a single theory (e.g. reactive

etabolite formation) to universally defined idiosyn-ratic drug toxicity have proven not yet to be useful.ne possible explanation for the illusiveness of the def-

nition of idiosyncratic toxicants is that there may beultiple factors involved [44]. These factors are pro-

osed to be the chemical properties of the toxicant, andhe genetic make up of the human victim, as well as theomplex environmental factors surrounding the victim.t is projected that the factor that can be readily controlleds the “chemical properties”. Therefore, a more practi-al approach to eliminate idiosyncratic drug may be theefinition of the common properties of the toxicants [44].

Towards this goal of defining the properties of drugshat would cause idiosyncratic hepatotoxiciy, we havepplied toxicogenomics to evaluate troglitazone, a drughat was withdrawn from the market due to its association

ig. 3. Schematic presentation of the integrated discrete multiple organ co-chat in the human body, there are multiple organs that are physically separatedntering the systemic circulation will have interactions with all the interconnf metabolites or the induction of reactive biomolecules (e.g. cytokines), willeduced to practice using a wells-in-a-chamber concept, with multiple wells irgans as physically discrete wells interconnected by an overlying medium (b

actions 168 (2007) 16–29 25

with idiosyncratic hepatotoxicity [43]. Trogltiazone wascompared to the relatively nontoxic structural analogs,rosiglitazone and pioglitazone [43]. Freshly isolatedhuman hepatocytes from three donors were treated invitro with the three chemicals, followed by extractionof mRNA and hybridization with cDNA microarrays.The gene expression heat map (using colors to clas-sify the magnitude of response) is presented in Fig. 2.Dramatic differences were observed in gene expressionprofiles between the toxic troglitazone and the non-toxic structural analogues. Troglitazone was found toselectively down regulate genes with protective func-tions (Fig. 2A) and up regulate genes with toxificationfunctions (Fig. 2B). Similar toxicogenomics results withtroglitazone were also obtained by the laboratory of War-ing and coworkers [45].

Toxicogenomics studies with troglitazone and its less

toxic analogs, rosiglitazone and pioglitazone providedclues for the differential hepatotoxicity of these struc-turally related chemicals. All three chemicals appearto induce genes for oxidative metabolism (e.g. P450

ulture (IdMOC) system. The IdMOC system is based on the conceptbut interconnected by the systemic circulation (top figure). A toxicantected organs. Results of the interactions, which can be the formationhave the potential to interact with the multiple organs. The concept isn a containing chamber, allowing the culturing of cells from differentottom figure). Figures reconstructed from [46]).

26 A.P. Li / Chemico-Biological Inter

Fig. 4. Photograph of an IdMOC plate based on the format of a standard96-well plate (IdMOC-96). The plate consists of six individual wellsinside a containing chamber. The wells are filled with colored liquids(left) to show that they are physically separated. The chambers are filled(right) to show the connection of the wells. Each chamber thereforerepresents a single experimental unit, allowing each plate to be used formultiple treatment conditions. For instance, one IdMOC-96 plate canbe used to evaluate the effect on six cell types for 15 test articles (usingone chamber as control) at a single concentration, two test articlesat four concentrations in duplicates, or a single test article at fourconcentrations in quadruplicates.

Fig. 5. Differential cytotoxicity of tamoxifen, an anticancer drug with knownorgans and a tumor cell line. The primary cells are: human hepatocytes (hcells (HAEC), representing the vascular endothelium; human astrocytes (astrotubule epithelial cells (RPTC), representing the kidney; human small airway ep(MCF-7). The results shown here illustrate an application of the IdMOC as awith minimal toxicity towards normal human tissues. Results are adopted fro

actions 168 (2007) 16–29

isoform 3A4) which may be responsible for the genera-tion of reactive metabolites [47]. Troglitazone, however,uniquely induced gene expression of the toxificationpathways such as apoptosis, and inflammation, and sup-pressed gene expressions of the gene responsible fordetoxification pathways such as acute phase proteins,phase II conjugation, and stress proteins [43,45]. It isprojected that the genes that are affected differentiallyby troglitazone and its less toxic structural analogs maybe used to predict the human hepatotoxic potential ofnew troglitazone structural analogs.

The study with troglitazone and its less toxic analogsillustrates an application of toxicogenomics in the devel-opment of hypothesis to allow subsequent mechanisticinvestigations. The combination of toxicogenomics andhuman cells such as human hepatocytes in vitro can be apowerful combination to aid the definition or predictionof human drug toxicity. The key is to develop an in vitrosystem to incorporate in vivo factors that are criticalto the manifestation of toxicity. One such system is theintegrated discrete multiple organ co-culture (IdMOC)system.

4.4.4. IdMOCOne major drawback of in vitro systems is that each

cell type is studied in isolation, while in the human body,

multiple organ toxicity, in IdMOC with primary cells from five majorepatocytes), representing the human liver; human aortic endothelialcytes), representing the central nervous system; human renal proximalithelial cells (SAEC), presenting the lung; the human breast carcinoma“tumor-bearing man” in the discovery of potential anticancer agents

m [46].

al Inter

tttfo

cbaabtTtoTmp

Twgeamcav

1

2

3

iiac(oet

A.P. Li / Chemico-Biologic

here may be multiple organ interactions that are criticalo drug toxicity. An example of multiple organ interac-ions is a drug which is firstly metabolized by the liver toorm metabolites which may cause toxicity in a distantrgan such as the heart.

To overcome this deficiency, the independent dis-rete multiple organ co-culture (IdMOC) system haseen developed in our laboratory [46]. The IdMOCllows the co-culturing of cells from different organss physically separated cultures that are interconnectedy an overlying medium, akin to the blood circula-ion connecting the multiple organs in the human body.he IdMOC models the multiple organ interaction in

he whole organism in vivo, allowing the evaluation ofrgan-specific effects a drug and its metabolites (Fig. 3).he IdMOC represents an improved in vitro experi-ental system for routine screening of ADMET drug

roperties.The IdMOC involves the “wells-in-a-well” concept.

he typical IdMOC plate consists of a chamber withinhich are several wells (Fig. 4). Cells of different ori-ins (e.g. from different organs) are initially cultured,ach in its specific medium, in the wells. When the cellsre established, the wells are flooded with an overlyingedium, thereby connecting all the wells. The multiple

ell types now can interact via the overlying medium,kin to the multiple organs in a human body interactingia the systemic circulation.

The IdMOC system can be used for the following:

. Differential cytotoxicity: Evaluation of the toxicityof a substance on different cell types (e.g. cellsfrom different organs) under virtually identical exper-imental conditions with multiple cell type inter-actions.

. Differential distribution: Evaluation of the differen-tial accumulation/distribution of a substance amongmultiple cell types.

. Multiple organ metabolism: Evaluate the ultimatemetabolic fate of a substance upon metabolism bycells representing multiple organs with metabolicfunctions (e.g. liver, kidney, lung).

We have recently evaluated the differential cytotox-city of tamoxifen using the IdMOC system [46] whichllustrates the application of IdMOC in the screening ofnticancer agents for their cytotoxicity towards cancerells (the desired property) and cells from normal tissues

the undesired properties) (Fig. 5). Further validationf the system is now in progress in our laboratory tovaluate the strengths and limitations of this novelechnology.

actions 168 (2007) 16–29 27

5. Conclusions

In the past decade, human hepatocytes have beenused increasingly in drug development. The experienceas of now is that relevant information can be obtainedwith this experimental system, including human-specific drug properties such as metabolic stabilityand fate, drug–drug metabolizing enzyme interactions,drug–transporter interactions, and drug toxicity. Theavailability of human hepatocytes for research has beendramatically increased due to the increased availabilityof commercial sources for the cells.

While it is generally agreed that isolated hepato-cytes retain most of the key liver properties (e.g. drugmetabolism), the value of the data obtained with hep-atocytes remains limited by the inherent inadequaciesof the in vitro systems (at least as of the current prac-tice), namely, the absence of key in vivo factors. Itis therefore of no surprise that the recent improve-ments on the use of intact hepatocytes is to incorporatekey in vivo factors (e.g. incubations of intact hepa-tocytes in 100% serum; IdMOC). Human hepatocytesrepresent a valuable experimental system in drug devel-opment. It is through the application of the experimentalsystem within its limitations, while continually devel-oping approaches to overcome these limitations, thatthis valuable resource can be utilized to its fullestpotential.

References

[1] X.N. Feng, X. Xu, S.S. Zheng, Current status and perspective ofliver preservation solutions, Hepatobiliary Pancreat. Dis. Int. 5(4) (2006) 490–494.

[2] M.N. Berry, D.S. Friend, High-yield preparation of isolated ratliver parenchymal cells: a biochemical and fine structural study,J. Cell Biol. 43 (3) (1969) 506–520.

[3] A.P. Li, M.A. Roque, D.J. Beck, D.L. Kaminski, Isolation andculturing of hepatocytes from human livers, J. Tissue Cult. Meth.14 (1992) 139–146.

[4] L. Pichard, E. Raulet, G. Fabre, J.B. Ferrini, J.C. Ourlin, P. Maurel,Human hepatocyte culture, Meth. Mol. Biol. 320 (2006) 283–293.

[5] K. Alexandrova, C. Griesel, M. Barthold, H.G. Heuft, M. Ott, M.Winkler, H. Schrem, M.P. Manns, T. Bredehorn, M. Net, M.M.Vidal, S. Kafert-Kasting, L. Arseniev, Large-scale isolation ofhuman hepatocytes for therapeutic application, Cell Transplant.14 (10) (2005) 845–853.

[6] R.R. Mitry, R.D. Hughes, A. Dhawan, Progress in human hepa-tocytes: isolation, culture and cryopreservation, Semin. Cell Dev.Biol. 13 (6) (2002) 463–467.

[7] A.P. Li, C. Lu, J.A. Brent, C. Pham, A. Fackett, C.E. Ruegg,

P.M. Silber, Cryopreserved human hepatocytes: characterizationof drug-metabolizing enzyme activities and applications in higherthroughput screening assays for hepatotoxicity, metabolic stabil-ity, and drug–drug interaction potential, Chem. Biol. Interact. 121(1) (1999) 17–35.

al Inter

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

[

28 A.P. Li / Chemico-Biologic

[8] J.G. Hengstler, D. Utesch, P. Steinberg, K.L. Platt, B. Diener, M.Ringel, N. Swales, T. Fischer, K. Biefang, M. Gerl, T. Bottger,F. Oesch, Cryopreserved primary hepatocytes as a constantlyavailable in vitro model for the evaluation of human and ani-mal drug metabolism and enzyme induction, Drug Metab. Rev.32 (1) (2000) 81–118.

[9] L.J. Loretz, A.P. Li, M.W. Flye, A.G. Wilson, Optimizationof cryopreservation procedures for rat and human hepatocytes,Xenobiotica 19 (5) (1989) 489–498.

10] M.J. Gomez-Lechon, A. Lahoz, N. Jimenez, J. Vincente Castell,M.T. Donato, Cryopreservation of rat, dog and human hepato-cytes: influence of preculture and cryoprotectants on recovery,cytochrome P450 activities and induction upon thawing, Xenobi-otica 36 (6) (2006) 457–472.

11] C. Terry, A. Dhawan, R.R. Mitry, S.C. Lehec, R.D. Hughes,Preincubation of rat and human hepatocytes with cytoprotectantsprior to cryopreservation can improve viability and function uponthawing, Liver Transplant. 12 (1) (2006) 165–177.

12] A.P. Li, N.R. Hartman, C. Lu, J.M. Collins, J.M. Strong, Effectsof cytochrome P450 inducers on 17alpha-ethinyloestradiol (EE2)conjugation by primary human hepatocytes, Br. J. Clin. Pharma-col. 48 (5) (1999) 733–742.

13] D.F. McGinnity, M.G. Soars, R.A. Urbanowicz, R.J. Riley, Eval-uation of fresh and cryopreserved hepatocytes as in vitro drugmetabolism tools for the prediction of metabolic clearance, DrugMetab. Dispos. 32 (11) (2004) 1247–1253.

14] A.P. Li, P.D. Gorycki, J.G. Hengstler, G.L. Kedderis, H.G. Koebe,R. Rahmani, G. de Sousas, J.M. Silva, P. Skett, Present status ofthe application of cryopreserved hepatocytes in the evaluation ofxenobiotics: consensus of an international expert panel, Chem.Biol. Interact. 1 (121(1)) (1999) 117–123.

15] A.K. Sohlenius-Sternbeck, S. Schmidt, Impaired glutathione-conjugating capacity by cryopreserved human and rat hepato-cytes, Xenobiotica 35 (7) (2005) 727–736.

16] Y. Shitara, T. Itoh, H. Sato, A.P. Li, Y. Sugiyama, Inhibi-tion of transporter-mediated hepatic uptake as a mechanism fordrug–drug interaction between cerivastatin and cyclosporin A, J.Pharmacol. Exp. Ther. 304 (2) (2003) 610–616.

17] K. Maeda, M. Kambara, Y. Tian, A.F. Hofmann, Y. Sugiyama,Uptake of ursodeoxycholate and its conjugates by human hep-atocytes: role of Na(+)-taurocholate cotransporting polypeptide(NTCP), organic anion transporting polypeptide (OATP) 1B1(OATP-C), and oatp1B3 (OATP8), Mol. Pharm. 3 (1) (2006)70–77.

18] K. Ito, T. Iwatsubo, S. Kanamitsu, Y. Nakajima, Y. Sugiyama,Quantitative prediction of in vivo drug clearance and drug inter-actions from in vitro data on metabolism, together with bindingand transport, Annu. Rev. Pharmacol. Toxicol. 38 (1998) 461–499.

19] D.G. Levitt, T.W. Schnider, Human physiologically based phar-macokinetic model for propofol, BMC Anesthesiol. 5 (1) (2005)4.

20] N. Blanchard, L. Richert, B. Notter, F. Delobel, P. David, P. Coas-solo, T. Lave, Impact of serum on clearance predictions obtainedfrom suspensions and primary cultures of rat hepatocytes, Eur. J.Pharm. Sci. 23 (2) (2004) 189–199.

21] N. Blanchard, E. Alexandre, C. Abadie, T. Lave, B. Heyd, G.

Mantion, D. Jaeck, L. Richert, P. Coassolo, Comparison of clear-ance predictions using primary cultures and suspensions of humanhepatocytes, Xenobiotica 35 (1) (2005) 1–15.

22] N. Blanchard, N.J. Hewitt, P. Silber, H. Jones, P. Coassolo, T.Lave, Prediction of hepatic clearance using cryopreserved human

[

actions 168 (2007) 16–29

hepatocytes: a comparison of serum and serum-free incubations,J. Pharm. Pharmacol. 58 (5) (2006) 633–641.

23] FDA, Drug Metabolism/Drug Interaction Studies in the DrugDevelopment Process: Studies In Vitro, FDA, 1997. http://www.fda.gov/cder/guidance/clin3.pdf.

24] FDA, In Vivo Drug Metabolism/Drug Interaction Studies—StudyDesign, Data Analysis, and Recommendations for Dosingand Labeling, FDA, 1999. http://www.fda.gov/cder/guidance/2635fnl.pdf.

25] FDA, Drug Interaction Studies–Study Design, Data Analy-sis, and Implications for Dosing and Labeling, FDA, 2006.http://www.fda.gov/cder/guidance/6695dft.htm.

26] P.K. Honig, D.C. Worham, K. Zamani, J.C. Mullin, D.P. Con-ner, L.R. Cantilena, The effect of fluconazole on the steady-statepharmacokinetics and electrocardiographic pharmacodynamicsof terfenadine in humans, Clin. Pharmacol. Ther. 53 (6) (1993)630–636.

27] N.J. Daniels, J.S. Dover, R.K. Schachter, Interaction betweencyclosporin and rifampicin, Lancet 2 (8403) (1984) 639.

28] C. Lu, G.T. Miwa, S.R. Prakash, L.S. Gan, S.K. Balani, A novelmodel for the prediction of drug–drug interactions in humansbased on in vitro cyp phenotypic data. Drug Metab. Dispos., inpress.

29] A.P. Li, P. Maurel, M.J. Gomez-Lechon, L.C. Cheng, M. Jurima-Romet, Preclinical evaluation of drug–drug interaction potential:present status of the application of primary human hepatocytesin the evaluation of cytochrome P450 induction, Chem. Biol.Interact. 107 (1–2) (1997) 5–16.

30] A. Madan, R.A. Graham, K.M. Carroll, D.R. Mudra, L.A. Burton,L.A. Krueger, A.D. Downey, M. Czerwinski, J. Forster, M.D.Ribadeneira, L.S. Gan, E.L. LeCluyse, K. Zech, P. Robertson Jr.,P. Koch, L. Antonian, G. Wagner, L. Yu, A. Parkinson, Effects ofprototypical microsomal enzyme inducers on cytochrome P450expression in cultured human hepatocytes, Drug Metab. Dispos.31 (4) (2003) 421–431.

31] C.E. Ruegg, P.M. Silber, R.A. Mughal, J. Ismail, C. Lu, D.C.Bode, A.P. Li, Cytochrome P450 induction and conjugationmetabolism in primary human hepatocytes after cryopreservation,In Vitro Toxicol. 10 (2) (1997) 217–222.

32] D. Roymans, P. Annaert, J. Van Houdt, A. Weygers, J. Noukens,C. Sensenhauser, J. Silva, C. Van Looveren, J. Hendrickx, G.Mannens, W. Meuldermans, Expression and induction potentialof cytochromes P450 in human cryopreserved hepatocytes, DrugMetab. Dispos. 33 (7) (2005) 1004–1016, Epub March 31, 2005.

33] S. Kafert-Kasting, K. Alexandrova, M. Barthold, B. Laube, G.Friedrich, L. Arseniev, J.G. Hengstler, Enzyme induction in cry-opreserved human hepatocyte cultures, Toxicology 220 (2-3)(2006) 117–125.

34] G. Luo, M. Cunningham, S. Kim, T. Burn, J. Lin, M. Sinz, G.Hamilton, C. Rizzo, S. Jolley, D. Gilbert, A. Downey, D. Mudra,R. Graham, K. Carroll, J. Xie, A. Madan, A. Parkinson, D. Christ,B. Selling, E. LeCluyse, L.S. Gan, CYP3A4 induction by drugs:correlation between a pregnane X receptor reporter gene assay andCYP3A4 expression in human hepatocytes, Drug Metab. Dispos.30 (7) (2002) 795–804.

35] Y. Shitara, A.P. Li, Y. Kato, C. Lu, K. Ito, T. Itoh, Y. Sugiyama,Function of uptake transporters for taurocholate and estradiol

17beta-d-glucuronide in cryopreserved human hepatocytes, DrugMetab. Pharmacokinet. 18 (1) (2003) 33–41.

36] Y.A. Bi, D. Kazolias, D.B. Duignan, Use of cryopreserved humanhepatocytes in sandwich culture to measure hepatobiliary trans-port, Drug Metab. Dispos. 34 (9) (2006) 1658–1665.

al Inter

[

[

[

[

[

[

[

[

[

[

[

[

A.P. Li / Chemico-Biologic

37] M. Hirano, K. Maeda, Y. Shitara, Y. Sugiyama, Contribution ofOATP2 (OATP1B1) and OATP8 (OATP1B3) to the hepatic uptakeof pitavastatin in humans, J. Pharmacol. Exp. Ther. 311 (1) (2004)139–146.

38] W. Yamashiro, K. Maeda, M. Hirouchi, Y. Adachi, Z. Hu, Y.Sugiyama, Involvement of transporters in the hepatic uptakeand biliary excretion of valsartan, a selective antagonist of theangiotensin II AT1-receptor, in humans, Drug Metab. Dispos. 34(7) (2006) 1247–1254.

39] M.A. Talamini, B. Kappus, A. Hubbard, Repolarization of hepa-tocytes in culture, Hepatology 25 (1) (1997) 167–172.

40] X. Liu, J.P. Chism, E.L. LeCluyse, K.R. Brouwer, K.L. Brouwer,Correlation of biliary excretion in sandwich-cultured rat hepa-tocytes and in vivo in rats, Drug Metab. Dispos. 27 (6) (1999)637–644.

41] N. Ballatori, C.L. Hammond, J.B. Cunningham, S.M. Krance, R.Marchan, Molecular mechanisms of reduced glutathione trans-port: role of the MRP/CFTR/ABCC and OATP/SLC21A familiesof membrane proteins, Toxicol. Appl. Pharmacol. 204 (3) (2005)238–255.

42] J.A. DiMasi, R.W. Hansen, H.G. Grabowski, The price of inno-vation: new estimates of drug development costs, J. Health Econ.22 (2) (2003) 151–185.

43] L.D. Kier, R. Neft, L. Tang, R. Suizu, T. Cook, K. Onsurez, K.

Tiegler, Y. Sakai, M. Ortiz, T. Nolan, U. Sankar, A.P. Li, Appli-cations of microarrays with toxicologically relevant genes (toxgenes) for the evaluation of chemical toxicants in Sprague Daw-ley rats in vivo and human hepatocytes in vitro, Mutat. Res. 549(1-2) (2004) 101–113.

[

[

actions 168 (2007) 16–29 29

44] A.P. Li, A review of the common properties of drugs with idiosyn-cratic hepatotoxicity and the “multiple determinant hypothesis”for the manifestation of idiosyncratic drug toxicity, Chem. Biol.Interact. 142 (1-2) (2002) 7–23.

45] M.J. Liguori, M.G. Anderson, S. Bukofzer, J. McKim, J.F. Pre-genzer, J. Retief, B.B. Spear, J.F. Waring, Microarray analysisin human hepatocytes suggests a mechanism for hepatotoxicityinduced by trovafloxacin, Hepatology 41 (1) (2005) 177–186.

46] A.P. Li, C. Bode, Y. Sakai, A novel in vitro system, the inte-grated discrete multiple organ cell culture (IdMOC) system, forthe evaluation of human drug toxicity: comparative cytotoxic-ity of tamoxifen towards normal human cells from five majororgans and MCF-7 adenocarcinoma breast cancer cells, Chem.Biol. Interact. 150 (1) (2004) 129–136.

47] K. Kassahun, P.G. Pearson, W. Tang, I. McIntosh, K. Leung,C. Elmore, D. Dean, R. Wang, G. Doss, T.A. Baillie, Futureof toxicology—metabolic activation and drug design: challengesand opportunities in chemical toxicology, Chem. Res. Toxicol. 19(7) (2006) 889–893.

48] D. Jouin, N. Blanchard, E. Alexandre, F. Delobel, P. David-Pierson, T. Lave, D. Jaeck, L. Richert, P. Coassolo, Cryopreservedhuman hepatocytes in suspension are a convenient high through-put tool for the prediction of metabolic clearance, Eur. J. Pharm.Biopharm. 63 (3) (2006) 347–355 [Epub 2006 Feb 23].

49] A.P. Li, A. Rasmussen, L. Xu, D.L. Kaminski, Rifampicin induc-tion of lidocaine metabolism in cultured human hepatocytes, J.Pharmacol. Exp. Ther. 274 (2) (1995) 673–677.

50] A.P. Li, An integrated, multidisciplinary approach for drug safetyassessment, Drug. Discov. Today 9 (16) (2004) 687–693.