Microreactor Microfluidic Systems with Human Microsomes and Hepatocytes for use

Biomedical Microdevices7:2,117–125,2005C 2005Springer Science+Business Media,Inc.Manufactured in TheNetherlands.Microreactor Microfluidic Systems with Human Microsomes and Hepatocytes for use in Metabolite StudiesJeanna C.Zguris,1Laura J.Itle,2Daniel Hayes,3andMichael V.Pishko4,∗1Department of Chemical Engineering,2Department of Chemical Engineering and The Huck Institute for theLife Sciences,3Department of Material Science,and4Departments of Chemical Engineering,Chemistry,and MaterialsScience&Engineering,The Pennsylvania State University,UniversityPark,PA16802-4420,USAE-mail:mpishko@Abstract.In the area of drug discovery,high-speed synthesis has increased the number of drug candidates produced.These poten-tial drugs need to be evaluated for their adsorption,distribution, metabolism,elimination,and toxicology(ADMET)properties as early in the drug development stage as possible.Previously,a poten-tial drug’s ADMET properties have been found out by using mono-layer cell cultures and live animals.These methods can be costly, time-intensive,and impractical for screening the large amount of potential drugs created by combinatorial chemistry.A quick,small, inexpensive,and highly parallel device would be desirable to deter-mine a drug candidate’s properties(i.e.,metabolism of the drug). Here we fabricate a microfluidic device entrapping human micro-somes within poly(ethylene)glycol hydrogels thereby generating an in situ microreactor to assess a drug candidate’s metabolic proper-ties that can be coupled to analysis equipment.We show that mi-crosomes can be entrapped without the loss of enzymatic activity during photopolymerization.Additionally,a microreactor utilizing hepatocytes was also created for comparison with the microsome microreactor.Key Words.microsomes,hydrogels,microreactor,cytochrome P450IntroductionAdsorption,distribution,metabolism,elimination,and toxicology(ADMET)studies are a key area of drug development in which potential drugs are evaluated in an attempt to determine adverse in vivo reactions prior to human trials.Despite rigorous testing,approximately 75,000–135,000deaths due to adverse drug reactions oc-cur annually in the United States,making it the sixth lead-ing cause of death(Hodgson,2001).In addition to limit-ing the number of deaths due to adverse drug reactions, a distinct economic advantage exists for pharmaceutical companies that determine ADMET properties early in the development process(Myers and Baker,2001).An esti-mated90%of drugs are dropped from development in the late clinical stage;eliminating any number of these dead-end drug candidates earlier in the process would allow for the reinvestment of research moneys into more productive avenues(Contag,2002).We have devised a microreactor with microsomes en-trapped in a polymer matrix of poly(ethylene)glycol that can be used to determine the metabolic products of potential drug candidates.Derived from the ultrasonic homogenization of human liver,microsomes retain the metabolic properties of the liver,the organ largely re-sponsible for drug metabolism(Pearce,et al.,1996). Microsomes contain a high percentage of cytochrome P450,group of hemeproteins(Khan,2003)that com-pose an enzyme superfamily subdivided by enzyme struc-ture(Chauret et al.,1999).The cytochrome P450family plays a major role in the oxidative metabolism of xeno-biotics,including toxins,exogenous chemicals,and drugs (Pearce et al.,1996;Chauret et al.,1999;V ondracek et al., 2001).Cytochrome P450metabolism of potential drug can-didates is becoming a routine aspect of high through-put drug screening.For example,the inhibition of Cy-tochrome P450activity can indicate an adverse drug-drug interaction(Crespi et al.,1997).In vitro drug metabolism and cytochrome P450activity related to drug exposure is measured in a variety of biological systems with in-creasing orders of structure and organization.These sys-tems include microsomes,supersomes,hepatocytes,and liver slices(Brandon et al.,2003).The production of drug metabolites can vary depending on the system used,with microsomes and supersomes producing Phase I metabo-lites due to Cytochrome P450,while hepatocytes and liver slices can produce Phase I(hydrolysis,oxidation, and reduction)and Phase II(conjugation)metabolites in in vitro environments(Brandon,2003;DeGraaf et al.,2003).Despite incomplete metabolism of some drug candidates,microsomes are an inexpensive,easy to use *Corresponding author.117118Zguris et al.alternative for initial testing of a drug’s metabolism.In this case human microsomes are used,which will pro-vide us with information regarding the metabolites that will be produced in humans(Sivapathasundaram et al., 2003).The immobilization of microsomes in a three-dimensional matrix provides several advantages over co-valent linkage or physical adsorption including increased storage stability,the ability to separate products from the incubation step,and the ability to reuse and recycle the microsomes for later tests(Sakai-Kato et al.,2002).It has been shown by Sakai-Kato and co-workers that micro-somes can be entrapped within a sol-gel matrix(Sakai-Kato et al.,2002).Microsomes can also be entrapped in a poly(ethylene)glycol matrix,which has several advan-tages over sol-gel technology.These advantages include biocompatibility of the matrix,non-toxic crosslinking so-lutions,faster polymerization time,and ease of patterning (Honiger et al.,1995;Cruise et al.,1998;Alcantar et al., 2000;Elisseeff et al.,2000;Revzin et al.,2001;Burdick and Anseth,2002;Koh et al.,2002,2003;Liu,2002;Liu et al.,2002;Revzin,2003).Previously,we have encap-sulated and patterned enzymes and whole cells for use in sensing applications(Revzin et al.,2001;Koh et al., 2002,2003).However,microsomes harness the advan-tages of both enzyme and whole cell encapsulation.Mi-crosomes contain multiple enzymes,leading to a variety of products,but without needing a tightly regulated external environment.Here we describe the fabrication of a poly(ethylene) glycol microreactor containing microsomes in microflu-idic channels for easy sample delivery.Because it is con-tained within a microfluidic device,this microreactor has the ability to be coupled with a variety of downstream de-tection devices for drug metabolites.However,this method relies on the ability of drug candidates to be transferred through the hydrogel matrix and for drug products to be returned to the process stream.To this end,we charac-terized the mass transfer by calculating the mesh size of PEG hydrogels.Additionally,we compare the activ-ity of encapsulated microsomes with microsomes in so-lution,hepatocytes in solution,and encapsulated hepa-tocytes by utilizing afluorescent substrate,EFC and its fluorescent product HFC(Roberts et al.,1995;Kent et al., 2004).Experimental SectionMicrosomesGLP human liver microsomes(CellzDirect,Austin TX) were used.For imaging,the microsome membrane was stained with3,3 -dioctadecyloxacarbocyanine perchlorate (DiOC)(Molecular Probes,Eugene,OR)dissolved in dimethyl sulfoxide(DMSO)(VWR International,West Chester,PA).Cell cultureSV-40transformed murine hepatocytes were obtained from American Type Culture Collection(Manassas, V A).Cells were incubated at37◦C in5%CO2and 95%humidified air in Dulbecco’s Modified Eagle’s Medium(DMEM,Sigma,St.Louis,MO)with4.5g/L glucose,200nM dexamethasone,4%fetal bovine serum(FBS),and1%antibiotic/antimycotic solution (Sigma,St.Louis,MO).Hepatocytes were grown to confluence in75cm2polystyrene tissue cultureflasks and confluent cells were subcultured every2to3days by trypsinization with0.25%(w/v)trypsin and0.13%(w/v) EDTA.Forfluorescent whole cell imaging,cells were stained with Cell Tracker Orange CMTMR(5-(and-6)-(((4-chloromethyl)benzyl)amino)tetramethylrhodamine) (Molecular Probes,Eugene,OR).Preparation of PEG hydrogel spheresHydrogel spheres were prepared by following a pre-viously described procedure(Russell et al.,1999).A poly(ethylene)glycol diacrylate(PEG-DA,Sigma,St. Louis,MO)precursor solution was extruded through a21-gauge needle into a graduated cylinder of mineral oil.The precursor droplets were photopolymerized with a365nm, 300mW/cm2light source(EFOS Ultracure100ss Plus, UV spot lamp,Mississauga,Ontario).The spheres were collected and washed repeatedly with water.For mass transfer experiments,spheres containing 100nmfluorescently particles(Molecular Probes,Eu-gene,OR)and spheres that did not contain any particles with the composition of700µL of PEG-DA,300µL of phosphate buffer saline solution(0.1M,pH7.3)and 10µL of Darocur1173.For spheres that were created mi-crosome activity measurements,a precursor solution with 700µL of PEG-DA,200µL of phosphate buffered saline solution(0.1M,pH7.4),100µL of glucose-6-phosphate dehydrogenase at0.07KU/mL,20µL of microsomes at 20mg protein/mL and10µL of2-Hydroxy-2-methyl-1-phenyl-1-propanone(Darocur1173,Ciba,Tarrytown, NY)was prepared.To entrap hepatocytes in PEG spheres, a cell suspension of1.0×106to1.5×106cells/mL in90%phosphate buffered saline,10%PEG-DA(MW 575),and1%2-hydroxy-2-methyl-1-phenyl-1-propanone (Darocur1173,Ciba,Tarrytown,NY)photoinitiator was used.Microreactor Microfluidic Systems with Human Microsomes and Hepatocytes119Cytochrome P450activity in microsomes and mammalian cellsFor the determination of cytochrome P450in microsomes, suspension or entrapped microsomes were used.For sus-pension studies,solutions were made with3000µL of phosphate buffer saline solution(0.1M,pH7.4),100µL of glucose-6-phosphate dehydrogenase at0.07KU/mL, 20µL of microsomes at20mg protein/mL,100µL of glucose-6-phosphate(0.57M,Sigma,Sigma,St.Louis, MO),and100µL of nicotinamide adenine dinucleotide phosphate oxidized sodium salt(NADPH)(VWR Inter-national,West Chester,PA).The NADPH,glucose-6-phosphate and glucose-6-phosphate dehydrogenase were added for the NADPH regenerating system that is needed with the use of microsomes.Microsome containing spheres were generated as previously described and sus-pended in2mL PBS with100µL of glucose-6-phosphate (0.57M,Sigma,Sigma,St.Louis,MO)and100µL of NADPH.7-ethoxy-4-trifluoromethylcoumarin(1g/L, EFC)(Molecular Probes,Eugene,OR)were added to microsomes or sphere suspensions in varied amounts. Fluorescent emission spectra of the EFC or the prod-uct of the reaction,7-hydroxy-4-trifluoromethyl coumarin (HFC)(DeLuca et al.,1988;Buters et al.,1993)were ob-tained using afluorescence spectrometer(QM-1,Photon Technology International).For the determination of cytochrome P450activity in mammalian hepatocytes,cell suspensions or entrapped hepatocytes were used.For suspension studies,adherent cultures were detached via trypsinization,pelleted,and re-suspended at a density of1.0×106to1.5×106cells/mL in sterile0.1M phosphate buffer saline solution(PBS),pH 7.4.Cell containing spheres were generated as previously described and suspended in PBS.10µL of1g/L7-ethoxy-4-trifluoromethylcoumarin(EFC)(Molecular Probes,Eu-gene,OR)were added to cell or sphere suspensions.Flu-orescent emission spectra of the EFC or the HFC were obtained using afluorescence spectrometer(QM-1,Pho-ton Technology International).Preparation of PEG hydrogel microstructuresThe hydrogel microstructures were made using a simi-lar method as previously reported(Revzin et al.,2001).In short,the hydrogel arrays were patterned photolithograph-ically with poly(ethylene)glycol diacrylate(PEG-DA) (MW575,Sigma,Milwaukee,WI)on glass substrates.To prepare the substrates for the arrays,an oxidized surface was created by an acid wash(6N sulfuric acid,Aldrich Chemical Co.,Milwaukee,WI)followed by a base wash (0.1M sodium hydroxide,Sigma,St.Louis,MO)of the glass followed by treatment with3-(trichlorosilyl)propyl methacrylate(Sigma,St.Louis,MO)to form a self as-sembled monolayer(SAM)with pendant methacrylate groups.A solution containing PEG-DA(MW575,Sigma, St.Louis,MO),0.1M phosphate buffered saline, and a photoinitiator,2-Hydroxy-2-methyl-1-phenyl-1-propanone(Darocur1173,Ciba,Tarrytown,NY)was prepared.For microsome containing arrays,the con-centration of components was700µL of PEG-DA, 200µL of phosphate buffer saline solution(0.1M, pH7.3),100µL of glucose-6-phosphate dehydrogenase at0.07KU/mL,20µL of microsomes at20mg pro-tein/mL that were previously exposed to DIOC and10µL of2-Hydroxy-2-methyl-1-phenyl-1-propanone.To gen-erate cell-containing hydrogel arrays,a cell suspension of1.0×106to1.5×106cells/mL in PBS was ing the cell suspension,a10%PEG-DA solution with1%photoinitiator,2-Hydroxy-2-methyl-1-phenyl-1-propanone(Darocur1173,Ciba,Tarrytown,NY),was prepared.The microsome or cell-containing precursor solution was subsequently placed onto the methacrylated substrate. The precursor solution was then covered with a photomask and exposed to ultraviolet light at365nm,300mW/cm2 (EFOS Ultracure100ss Plus,UV spot lamp,Mississauga, Ontario)region for1–2seconds.Areas exposed to light cross-linked via a free radical mechanism and the result-ing hydrogel network isfixed in place while the remaining unreacted monomer is washed away with water.Arrays were imaged on an Axiovert Zeiss200Mfluorescent mi-croscope with a mercury light source(Zeiss,Thornwood, NY).Preparation of microfluidic channelsThe masters had negative patterns that were prepared us-ing photolithography with SU-850negative photoresist (Microlithography Chemical Corp,Newton MA).The patterns were made on glass slides and silica wafers. The pattern was placed on the middle of the substrate utilizing a chrome sodalime photomask made by Ad-vanced Reproductions(Andover,MA).To create the poly(dimethylsiloxane)(PDMS)channels for the genera-tion of microreactors within microfluidic channels,mas-ters were placed in petri dishes and covered with a10:1 mixture of PDMS prepolymer to curing agent(Dow Corn-ing Sylgard184,Midland,MI).The resulting degassed mixture was poured into the Teflon mold with the master and cured at60◦C for at least2hours.After curing,the PDMS replica was removed from the master and the in-let and outlet ports were placed by piercing the replica with a blunted syringe needle through the backside of the network.PDMS channels and glass coverslips were treated with a115V high frequency generator model120Zguris et al.BD-10AS(Electro-Technic Products,Inc,Chicago,IL) to irreversibly seal the channel to the glass.To create microreactors within the microfluidic net-work,polymer precursor solution previously described containing either microsomes or hepatocytes were pumped into the microfluidic channels.A photomask was positioned over the microfluidic channel,and the polymer precursor solution was exposed to ultraviolet light as de-scribed above.Unreacted polymer precursor was rinsed from the channel using water.Arrays within microfluidic networks were imaged using an Axiovert Zeiss200Mflu-orescent microscope with a mercury light source(Zeiss, Thornwood,NY).Results and DiscussionFabrication of microsome containing hydrogel arrays The process of making hydrogel microstructures with mi-crosomes has been developed.The microsomes are bought with a20mg/mL protein,which includes a large amount of tissue debris.The large debris can be seen in Fig-ure1(a).The large tissue debris could block the microflu-idic channel and alter the mass transfer of the analyte to the microsomes in the hydrogel.Depending on the chan-nel width it could be critical that the larger tissue debris was removed.This was accomplished by centrifuging the solution,removing and keeping thefluid and throwing out the pellet.After thisfiltering technique most of the large debris is removed.This can be shown by Figure1(b), there is sufficiently less debris after thisfiltering technique.Thefiltered or unfiltered solution of microsomes can be entrapped in hydrogel microstructures.We have shown that the microsomes in the unfiltered state can be en-trapped with a polymer matrix and stay intact.This is shown in Figure2,where the lipid membranes are stained with DiOC remainingfluorescently active.The hydrogels contain a7:3ratio of PEG:PBS precursor solution.Mass transfer to microsomes in PEG hydrogels Microsomes must be permanentlyfixed within the hy-drogel or free microsomes could disrupt any analysis of products.PEG hydrogels containingfluorescently tagged nanoparticles the same size as microsomes were moni-tored for2.5hours to insure that no particles leached into the surroundingfluid.In Figure3,thefluorescent inten-sity of the hydrogels is shown as a function of time.No substantial decrease influorescent intensity was shown, indicating that no particles leached from the hydrogels.To substantiate this result and to verify that drug candidates could reach microsomes entrapped in a hydrogel matrix, the mesh size of the gels was calculated.Additionally,the mesh size of the gel can be used to find the effective diffusivity(Peppas et al.,2000)of the potential drug;the effective diffusivity would allow the incubation time of the microsomes with the drug candi-date to be calculated.Mesh size was estimated by using the Peppas-Merrill equation.Though the Peppas-Merrill equation corrected the Flory-Rehner model(Peppas et al., 2000)for solvated systems,it still contains some limita-tions(Wu et al.,2004).A key limitation precludes the cal-culation of mesh size for polymer networks formed with dilute precursor solutions;in this case,the mesh size can-not be calculated mathematically for hydrogels entrapping hepatocytes.The determination of the mesh size has been deter-mined for other polymer systems(Mellott et al.,2001; Russell et al.,2001;Kim and Peppas,2002;Berger et al., 2003;Wang et al.,2003)by using the Peppas-Merrill equation(Peppas and Merrill,1976)(equation(1)).In the equation,v denotes the volume of the polymer af-ter crosslinking in either the relaxed state(subscript r) and in the swollen state(subscript s).V is the molar vol-ume of the medium,M N is the molecular weight of lin-ear polymer chains polymerized in the same conditions without crosslinks,andχis the Flory-Huggins interaction parameter.1c=2N−(v/V1)ln(1−v2,s)+v2,s+χ1v22,sv2,s(v2,s/v2,r)1/3−12(v2,s/v2,r)(1) N is the number of links between crosslinks,which is calculated by equation(2).In this equation,M r is the average molecular weight of the repeating unit.n=2M cM r(2)Equation(3)is used to determine r,the root mean squared end to end distance in the freely jointed state.In this equa-tion,l is the carbon-carbon bond length(1.54angstroms) and n is obtained from equation(2).(r2)1/2=l√n(3) Equation(4)is used to calculate the root mean square end to end distance of the polymer chain in the unperturbed state,where C n is the Flory characteristic ratio or rigidity factor of polymer.r o1/2=C n(r)1/2(4)The above value can be used to calculate the mesh size (ξ)in equation(5).ξ=υ−1/32,sr2o1/2(5)Microreactor Microfluidic Systems with Human Microsomes and Hepatocytes121Fig.1.Microsomes with lipids stained with DiOC(a)unfiltered microsome mixture originally purchased from CellzDirect at10×(b) Filtered microsome mixture at10×(c)Filtered microsome mixture at 40×magnification.Fig.2.Microsomes entrapped in PEG hydrogel microstructures(2.5×magnification).To calculate mesh size,hydrogel spheres were fabri-cated with and without nanoparticles.They are investi-gated in the relaxed,hydrated,and dry states.The av-erage mesh size was found to be11.4±1.6Angstroms (standard deviation,n=22)for a precursor solution of 700µL of PEG,300µL of water,10µL of Darocur 1173,and nanoparticles.The average mesh size of hy-drogels without nanoparticles was10.0±2.5(standard deviation,n=71)for a precursor solution of700µL of PEG,300µL of water,and10µL of Darocur1173. There is no significant difference between these two val-ues,indicating that100nm microsomes will notleach Fig.3.Leaching of the100nmfluorescently tagged particles from a hydrogel over time.122Zguris etal.Fig.4.Investigation of activity of microsomes of the average values between the wavelengths of 468.5±2.5nm (A)product emission of microsomes in solution over time and (B)product emission of microsomes in PEG sphere microstructures over time.from the gel and that the inclusion of nanoparticles in the hydrogels does not significantly alter hydrogel mesh size.Cytochrome P450activityThe cytochrome P450activity of microsomes in solu-tion was compared to the activity of the microsomes entrapped in the hydrogel microstructures,to ensure that encapsulation did not disrupt enzymatic activity.As shown in Figures 4(A)and (B),microsomes en-trapped in PEG maintained cytochrome P450activity de-spite undergoing the gelation process.The EFC emission peak increased over time in both microsomes in solution and in hydrogel spheres;no such increase was notice-able in a control sample containing only PEG spheres (Figure 4(C)).Similar experiments were conducted to determine the cytochrome P450activity of hepatocytes both in solution and in spheres (Figures 5(A)and (B)).These graphs show that for the same activity a larger amount ofcellsFig.5.Cytochrome P450activity of hepatocytes in free solution (A)and in PEG hydrogels (B)expressed in counts/second.or total protein are needed.This is due to the concen-trated nature of cytochrome P450present in microsomes and the lack of competitive pathways (Brandon,Raap et al.,2003).This indicates that the microreactor can be tai-lored to the needs of the user whether they want phase I metabolites (microsomes)or a combination of Phase I &II (hepatocytes).Fabrication of microreactors within PDMS microchannelsTo illustrate the feasibility of placing microsome or hep-atocyte microreactors within microfluidic channels for easy sample delivery,we generated arrays microsome and hepatocyte microreactors in microfluidic networks.Poly-mer precursor solutions containing either microsomes or hepatocytes were placed in the microfluidic networks.A chrome mask was positioned over the reactor chamber to generate patterns of hydrogels within the microfluidic channel.Arrays of microsomes are shown in Figure 6,while hepatocytes are shown in Figure 7.Microreactor Microfluidic Systems with Human Microsomes and Hepatocytes123(A)Fig.6.Images of microsomes in microfluidic channels.(A)Schematic drawing of hydrogel structures within in the microfluidic chamber.(B)2.5×Fluorescent image of the channel with hydrogel structures with microsomes stained with DIOC lipid stain.(C)20×Fluorescent image of one hydrogel structure in the channel with microsomes stained with DIOClipid.(A)(C)Fig.7.Array of hydrogels containing hepatocytes:(A)schematic drawing,(B)fluorescent,and (C)an enlarged image of one array element.124Zguris et al.ConclusionWe have demonstrated the ability to entrap microsomes in a poly(ethylene)glycol hydrogel microstructures without loss of activity during the gelation process.The activ-ity of cytochrome P450in microsomes in PEG structures was compared with the activity of hepatocytes in simi-lar microstructures.It was shown that fewer microsomes are needed than hepatocytes to generate the same level of activity in this system.Hydrogel arrays containing mi-crosomes and hepatocytes were created in microfluidic channels for easy sample delivery.Additionally,these de-vices can be coupled to mass spectrometers to create a device that requires little of the operator.Previously it has been shown that microfluidic devices have been cou-pled to MALDI-TOF MS(Brivio et al.,2002;Gustafsson et al.,2004)and electrospray ionization-mass spectrome-try(Kim and Knapp,2001;Chiou et al.,2002;Benetton et al.,2003;LeGac et al.,2004),which are both used to detect metabolite products currently. AcknowledgmentWe thank NASA(contract BIOTECH-01-0023-0131)for their kind support.ReferencesN.A.Alcantar,E.S.Aydil et al.,“Polyethylene glycol-coated biocom-patible surfaces,”J Biomed Mater Res51(3),343–351(2000).S.Benetton,J.Kameoka et al.,“Chip-based P450drug metabolism cou-pled to electrospray ionization-mass spectrometry detection,”Analyt-ical Chemistry:A-G.(2003).J.Berger,M.Reist et al.,“Structure and interactions in covalently and ionically crosslinked chitosan hydrogels for biomedical applications,”European Journal of Pharmaceutics and Biopharmaceutics(57),19–34(2003).E.F.A.Brandon,C.D.Raap et al.,“An update on in vitro test methods in human hepatic drug biotransformation research:Pros and cons.”Toxicology and Applied Pharmacology(189),233–246(2003).E.F.A.Brandon,C.D.Raap,I.B.Meijerman,H.Jos,and Jan H.M. Schellens,“An update on in vitro test methods in human hepatic drug biotransformation research:Pros and cons,”Toxicology and Applied Pharmacology189,233–246(2003).M.Brivio,R.H.Fokkens et al.,“Integrated microfluidic systems enabling (Bio)chemical reactions with on-line MALDI-TOF mass spectrome-try,”Anal Chem74(16),3972–3976(2002).J.A.Burdick and K.S.Anseth,“Photoencapsulation of osteoblasts in injectable RGD-modified PEG hydrogels for bone tissue engineering,”Biomaterials23(22),4315–4323(2002).J.T.Buters,C.D.Schiller et al.,“A highly sensitive tool for the assay of cyctochrome P450enzyme activity in rat,dog and man,”Biochemical Pharmacology46(9),1577–1584(1993).N.Chauret,N.Tremblay et al.,“Description of a96-well plate assay to measure cytochrome p4503a inhibition in human liver microsomesusing a selectivefluorescent probe,”Analytical Biochemistry276, 215–226(1999).C.-H.Chiou,G.-B.Lee et al.,“Micro devices integrated with microchan-nels and electrospray nozzles using PDMS casting techniques,”Sen-sors and Actuators B86,280–286(2002).P.R.Contag,“Whole-animal cellular and molecular imaging to acceler-ate drug development,”Drug Discovery Today7(10),555–562(2002).C.L.Crespi,ler et al.,“Microtiter plate assays for inhibition of human,drug-metabolizing cyctochromes P450,”Analytical Biochem-istry248,188–190(1997).G.M.Cruise,D.S.Scharp et al.,“Characterization of permeability and network structure of interfacially photopolymerized poly(ethylene glycol)diacrylate hydrogels,”Biomaterials19(14),1287–1294 (1998).I.A.M.DeGraaf,C.E.Van Meijeren et al.,“Comparison of in vitro prepa-rations for semi-quantitative prediction of in vivo drug metabolism,”Drug Metabolism and Disposition30(10),1129–1136(2003).J.G.DeLuca,G.R.Dysart et al.,“A direct,highly sensitive assay for cytochrome P-450catalyzed O-Deethylation using a novel coumarin analog,”Biochemical Pharmacology37(9),1731–1739(1988).J.Elisseeff,W.McIntosh et al.,“Photoencapsulation of chondrocytes in poly(ethylene oxide)-based semi-interpenetrating networks,”J Biomed Mater Res51(2),164–171(2000).M.Gustafsson,D.Hirschberg et al.,“Integrated sample prepartation and MALDI mass spectrometry on a microfluidic compact disk,”Anal Chem76,345–350(2004).J.Hodgson,“ADMET—Turning chemicals into drugs,”Nature Biotech-nology19,722–726(2001).J.Honiger,P.Balladur et al.,“Permeability and biocompatibility of a new hydrogel used for encapsulation of hepatocytes,”Biomaterials 16(10),753–759(1995).U.M.Kent,L.Pascual et al.,“Mechanistic studies with N-benzyl-1-aminobenzotriazole-inactivated CYP2B1:Differential effects on the metabolism of7-ethoxy-4-(trifluoromethyl)coumarin,testosterone and benzphetamine,”Archives of Biochemistry and Biophysics423, 277–287(2004).K.K.Khan,“Homotropic versus heterotopic cooperativity of cytochrome P450eryF:A substrate oxidation and spectral titration study,”Drug Metabolism and Disposition31(4),356–359(2003).B.Kim and N.A.Peppas,“Synthesis and characterization of pH-sensitive glycopolymer for oral drug delivery systems,”J.Biomater.Sci.Poly-mer Edn13(11),1271–1281(2002).J.-S.Kim and D.R.Knapp,“Microfabricated PDMS multichannel emit-ter for electrospray ionization mass spectrometry,”J Am Soc Mass Spectrom12,463–469(2001).W.G.Koh,L.J.Itle et al.,“Molding of hydrogel microstructures to cre-ate multiphenotype cell microarrays,”Anal Chem75(21),5783–5789 (2003).W.G.Koh, A.Revzin et al.,“Poly(ethylene glycol)hydrogel mi-crostructures encapsulating living cells,”Langmuir18(7),2459–2462 (2002).S.LeGac,J.Carlier et al.,“Monoliths for microfluidic devices in pro-teomics,”Journal of Chromatography B808,3–14(2004).V.Liu and S.N.Bhatia,“Three-dimensional photopatterning of hydro-gels containing living cells,”Biomedical Microdevices4(4),257–266 (2002).V.A.Liu,W.E.Jastromb et al.,“Engineering protein and cell adhesivity using PEO-terminated triblock polymers,”J Biomed Mater Res60(1), 126–134(2002).M.B.,Mellott,K.Searcy et al.,“Release of protein from highly cross-linked hydrogels of poly(ethylene glycol)diacrylate fabricated by UV polymerization,”Biomaterials22,929–941(2001).S.Myers and A.Baker”Drug discovery—an operating model for a new era,”Nature Biotechnology19,727–730(2001).。