Abstracts of papers presented at the ISLAR(International Symposium on LaboratoryAutomation and Robotics) 2001

The 19th International Symposium on Laboratory Automation and Robotics provided presentations on state-of-the-artdevelopments in laboratory automation and robotics. The symposium programme included papers and posters on allaspects of the technology. These comprised: managing laboratory automation (drug discovery); bioanalytical analysis;managing laboratory automation in drug discovery development and QC laboratory; functional genomics strategies andhigh throughput screening; advanced integration strategies; method development and global methods transfer;compound handling and logistics; combinatorial chemistry and automated synthesis ; high throughput LC-MS-MS;increasing eæ ciency in dissolution testing; lead optimization; strategies for UHTS; increasing throughput for ADMEtoxicology; data management/data handling and bioinformatics; using contract laboratories to increase productivity;assay miniaturization; process optimization; compliance and automation Ð the regulatory perspective; novel highthroughput screening technologies; compliance and automation Ð the industry perspective. Several discussion sessionswere included and activated, and provided interactive communication on a wider range of subjects.

Although the programme was very comprehensive, the Symposium was designed to provide time for both formal andinformal exchange of information. The technical presentations were organized into concurrent sessions with groupedpapers on related topics.

Abstracts for each paper and each poster are included here. Full presentations of several of these papers will appear inlater editions of this journal.

Designing custom automation systems to � t theneeds of early development

Adam M. Fermier, The R. W. Johnson PharmaceuticalResearch Institute, Raritan, NJ, USACo-authors: John Troisi, Ramon L. Rodriguez, and James V.Weber, Michael P. Graham and Erin C. Herita

The pharmaceutical industry is under extreme pressureto reduce the costs of bringing drugs to market. Thus, toachieve this goal the entire drug development processrequires continued ® ne tuning and the introduction ofnew technologies. Drug discovery has been movingthrough this change over the past decade. Drug devel-opment is now looking at that same daunting task ofincreasing throughput while maintaining complianceand headcount restraints. Custom automation systemsprovides one opportunity to attain this goal.

We have been investigating and building custom auto-mation into our drug development process. For example,we have modi® ed a Zymark Benchmate for weighingsmall quantities (1± 50 mg) of dry powders. The additionof weighing dry powders from a stock tube enables manyadvantages and increases the ¯ exibility of the Benchmate.One application we have been using this robot for is toweigh individual vials of drug substance into HPLC vialsfor subsequent accelerated solid state degradation studies.This typically requires about 20± 40 weighings for a singlestudy. A typical degradation program would have adozen studies for a single compound. The weighing robotnot only saves time but also oå ers some intrinsic safetybecause the analyst is not directly exposed to thematerial.

We have also designed a custom automated system toconduct accelerated degradation studies. These studies

require precise control of sampling and temperature. Atypical degradation study is usually conducted over aperiod of weeks. To compare studies in a relative mannerfor excipient compatibility studies, these times andtemperatures must be maintained. It is an ideal applica-tion for automation because of its repetitive nature. Thesystem allows for 16 reactions to be conducted in parallel.The software keeps track of temperature and times ofsampling.

In this presentation we will review criteria we have foundto be successful in designing the instrumentation andidentifying processes that would be ideal to automate.

Bringing Caco-2 studies from moderate- to high-throughput screening

Dawn Alderman, Neurogen Corporation, Branford, CT, USA

High throughput screening (HTS) technologies havebeen widely used in many areas of drug discovery.Human Caco-2 cells provide an in vitro model ofdrug bioavailability. They are derived from a colorectaladenocarcinoma cell line and exhibit functionalcharacteristics of the lower small intestinal tract.They diå erentiate as a monolayer with apical andbasal directionality that mimics the intestinalbarrier. Involvement of P-glycoprotein and other activetransport systems expressed endogenously in¯ uence drugabsorption.

Currently, most Caco-2 studies have low to moderatethroughput due to assay complexity. Many diå erentassay formats exist and are commercially available.Becton Dickinson provides a 3-day assay system and a10-day assay protocol. The 3-day system includes a

Journal of Automated Methods & Management in Chemistry

Vol. 24, No. 3 (May–June 2002) pp. 61–98

Journal of Automated Methods & Management in Chemistry ISSN 1463± 9246 print/ISSN 1464± 5068 online # 2002 ISLARhttp://www.tandf.co.uk/journals

61

media kit, vital to the growth and diå erentiation of thecells used within a 3-day period. The 10-day protocolsuggests that the cells can be seeded and used after a 10-day growth period without the requirement of theenhanced media necessary for the 3-day system. The10-day assay media that we currently use has an anti-biotic in it to protect against the potential bacterialcontamination that may occur with the cells seeded inthe plate for 10 days. The antibiotic condition may aå ectthe data when comparing with the 3-day system. Neuro-gen is currently evaluating these assays for any potentialdiå erences.

The Caco-2 model does not take into consideration ® rst-pass metabolism. The detection method also creates asigni® cant bottleneck. Currently LC/MS/MS is beingused. By using HTS technologies, such as laboratoryautomation and miniaturization, we have shown thatthe throughput of the Caco-2 assay can be signi® cantlyincreased. The introduction of the TECAN GENESIShas allowed the automation of the drug addition to theCaco-2 assay plates. It has also allowed for automatedsampling. The TECAN GENESIS that is currently inour facility has 1 ml Te on-coated syringes, and a 21-plate deck design. The programs being used have beenwritten in-house using the GENESIS software.

We have evaluated the 3-day assay system and arecurrently evaluating the 10-day protocol, in 24- and96-well formats. We have found that our data are ofgood quality and are reproducible in-house. We are stillevaluating many aspects of the 96-well assay format.Eå ective cell density and the volumes used in each assayformat have been assessed. We began by investigating acouple of cell densities. We have evaluated 12.5 K/well(Becton Dickinson protocol) and 25 K/well. Well vol-umes have also been assessed. Initially 100 ml in the apicalports and 250 ml in the basal ports was used. However,this condition is associated with a suspected wickingproblem. The sample under the plate lid as well as thevolume moving up the interior wall of the basal compart-ments has been observed. This is a possible suspect forinterwell contamination. We are further developing thisassay and will evaluate this phenomenon. We haveattempted to cut the volumes from 100 ml in the apicalcompartment to 80 ml. The basal well we have cut thevolume from 250 to 200 ml.

There are many assay conditions that have been pub-lished that vary between laboratories. Many publicationssuggest that a pH gradient from 6.5 (apical compart-ments) to 7.4 (basal compartments) is the best model,representing the conditions of the lower small intestine.Another assay condition that varies between laboratoriesis what concentration of drug to use. High concentrationsof drug create the risk of saturating the P-glycoproteintransporter as well as the other active transport mechan-isms. We are currently evaluating our Caco-2 studies inmodi® ed HBSS buå er at 5 mm drug concentration usingthe pH gradient. Again, cell density is a factor that variesbetween laboratories. Becton Dickinson recommends acell density in a 24-well plate to be 200 K/well and in a96-well plate to be 12.5 K/well. We are currently usingthis recommendation for both assay formats discussed.

Currently, samples are evaluated for Papp (apparentpermeability) using this published equation:

Papp…10¡6cm s¡1† ˆ …Vd=A†…dC=dT†:

This is a rate distance measurement, where Vd is thevolume of the donor compartment. A is the surface areaof the monolayer, dC is the change in concentration anddT is the change in time.

The Caco-2 cell model mimics the dynamics of thehuman lower small intestinal tract. The potential for thismodel to move towards a high-throughpu t format oå ersthe high-throughput compound screening in the future,which is essential to modern pharmaceutical develop-ment. Optimizing the current cell density and volumespeci® cations serve to strengthen this assay format, and iscurrently a major hindrance in its development.

Rational approach for selecting extraction par-ameters for homogenizer-base d tablet assaymethods

Alger Salt, GlaxoSmithKline, Research Triangle Park, NC,USA

This poster will present a rational approach for choosingextraction parameters, speci® cally homogenizer timesand speeds for tablet test methods that employ theTPW II workstation or similar homogenization-base dsample extraction procedures. This approach is basedon the assumption that extractions can be modelled usinga ® rst-order rate equation. The extraction pro® le, a curveshowing the amount of drug dissolved versus time, can becharacterized by a single number k, the ® rst-order rateconstant.

The amount of drug extracted at a given time can beestimated or predicted if the rate constant is known.Conversely, the amount of time required to reach a givenamount of drug dissolved can be estimated. Theseassumptions allow one to model the process and predicta combination of parameters to ensure that the tablet iscompletely dispersed and that the analyte is extractedquantitatively. A formal protocol can then be designedand performed to validate the extraction process. Theposter will describe this approach and illustrate it with acase study.

Introduction of new technology into the pharma-ceutical quality-control environment: ‘cuttingcosts without cutting corners’

Andy Boughey, AstraZeneca, UK

Ever increasing demands at the drug-discovery level inrecent years have lead to signi® cant advances in auto-mation. This in turn has lead to increasing numbers ofdrug candidates and correspondingly more new productsbeing manufactured, with a subsequent impact on theQuality Assurance Department.

Automation plays a key role in the future of qualitycontrol. It is recognized that signi® cant cost savings canbe made with the introduction of automation. However,there are equally important bene® ts to be gained withrespect to cGMP.

Abstracts of papers presented at the 2001 ISLAR

62

This presentation will discuss the pressures being placedon quality control and the bene® ts and concerns ofintroducing new technologies.

It is easy to support the introduction of new technologywhen the bene® ts can be expressed in terms of dollarssaved, but what is the true value of improved regulatorycompliance?

High-throughput analysis of free amino acids inbiological matrices

Arthur Rugg, Cereon Genomics LLC, Cambridge, MA, USACo-authors: Shaoxia Yu and Lily Li

Traditionally, chromogenic or ¯ uorescent derivatization(FD) coupled with high-pressure chromatography(usually 20± 30 min/separation) is used for amino acidanalysis due the amino acids’ lack of UV absorbance. Todetect an alternated change in the levels of free aminoacids through the screening of thousands of biologicalsamples, the analysis of 20 amino acids using an HPLC/FD approach is tedious and time consuming. Therefore,a high-speed 1.5 min per sample LC/MS/MS method hasbeen developed, validated and implemented to analyse20 free amino acidsÐ in biological matrices.

Multitasking, scalability and � exibility: the Stac-cato

TMautomated workstation

Matt Boeckeler, Neurogen Corporation, Branford, CT, USA

With drug-discovery technology advancing and chang-ing as rapidly as it is, there is a growing need forautomated equipment to have the capability to changewith these changes in technology. The use of single-workstation approaches to liquid handling is decreasingdue to the increasing need for a system that can performvarious liquid-handling tasks on the same platform.

The StaccatoTM automated workstation is an extremelyversatile liquid handler with the capabilities of handlingplate formats ranging from 96- to 1536-well plates. The` Drag and Drop’ architecture of the software allows for¯ exibility and ease in process development. With the useof CLARATM 2001 software, method development andsystem recon® guration become less intensive. Its scal-ability lends itself to supporting changes in platforms asemerging technologies arise.

The StaccatoTM system can be equipped or con® guredwith many components and options to ® t your liquid-handling needs. Using the SciClone ALH as the baseliquid handler, the system can come equipped with adisposable tip Rapid head, or a small volume ® xed tiphead with 96 or 384 cannula arrays. To augment theliquid-handling capabilities of the SciClone, up to fourseparate CavroTM bulk-dispensing manifolds can bemounted on the head for reagent additions. Liquidhandling can be further enhanced by the addition of aneight tip, independent Z-axis probe used for variousliquid-handling applications. Integration of the Auto-stack with the SciClone gives an exceptionally largestorage capacity for unattended operation. The use ofon-board barcode readers in conjunction with theCLARATM Data Manager, gives the system the ability

to track all compound and assay plates entering thesystem.

Reformatting chemical libraries with the StaccatoTM

system is very eæ cient. With its large storage capacityand ¯ exible con® gurations, reformatting can be donethrough a variety of methods. Using the Rapid headcon® guration, 300 96-well plates can be reformatted to75 384-well plates in one 8-h day. Using the small volume® xed-tip con® guration, one can convert 200 96-wellplates to 50 384-well plates in one 8-h day. Although® xed tip con® gurations yield less throughput due toimplementation of wash protocols, there are no consum-able costs. It also gives you the ability to reformat furtherfrom 384- to 1536-well plates, with an even smallervolume 384-cannula array designed for 1536 applica-tions.

The independent Z-axis probes provide the ability tocherry pick select compounds, assemble serial dilutionplates, and manipulate compounds and reagents forADMET studies. Performing parallel processing of platetypes using the ` on the ¯ y’ array swap capabilities, theStaccatoTM is capable of reformatting from 96- to 384-well plates using the 96-cannula array. After dropping oåthe 96 array, it can then pick up the 384 array andfurther dilute or deliver compounds directly to assayplates for execution with no human intervention.

The presented pipetting specs are based on ® ve trays runwith three transfers performed per tray. A 0.5 ml wettransfer produces excellent precision with an average%CV of 6.4, and an accuracy range of 1.06± 7.3%. A0.5-ml dry transfer yields an average precision %CV of8.6, and an accuracy range of 0.03± 10.1%. Dry transfersof 1 ml yield a precision average %CV of 5.8 and anaccuracy range of ± 1.5 to 4.4%. Volumes > 1 ml yieldsimilar results to the 1-ml numbers.

System use at Neurogen Corporation entails taking inchemical entities from our existing archive or from ourhigh-speed synthesis group and performing compounddilution procedures, or reformatting processes on theplates and transferring the compounds directly into assayplates, all within the same method. On this platform, thesystem can also build dose± response plates, preparecompounds for interdepartmental support or execute avariety of assays.

After compounds are in the proper format and concen-tration, with the use of the CavroTM dispensers, multiplereagent addition to assay plates can be accomplishedwith the proper scheduling of processes. In essence, thesystem can take compounds from their starting formatand concentrations, manipulate the plates to accommo-date assay-speci ® c plate formats and conditions, thenproceed to add the appropriate reagents via the CavroTM

dispensers, bringing the assay from compound delivery,through reagent addition. Dispensing by the CavroTM

syringes for volumes of 10± 2500 ml all yield similar results.Across plate precision yields an average%CV of 3.8 withan accuracy range of 1.1± 5.5%. Non-contact dispensingof reagents eliminates cross-well contamination and withmultiple manifolds, various reagents can be used.

Replication of compound data generated by the 384-cannula array at a dry transfer volume of 0.5 ml is quite

Abstracts of papers presented at the 2001 ISLAR

63

good. Dry dispensing of such a small volume is not easilyachieved, but obtainable with acceptable %CV usingparticular pipetting methods. Using the same pipettingmethods and comparing data to proven control methodsfor drug delivery yields very acceptable %CVs andcorrelation.

CLARATM 2001 Software is used for scheduling methodsand integrating peripherals into the existing system. Itcommunicates with adapter modules that translate com-mands between CLARATM and the ICP of the periph-eral module, triggering program execution of theinstrument. This method of integration makes the scal-ability of the system much easier and more eæ cient thanprevious conventional methods. Upon process comple-tion, the ICP sends a message back to CLARATM

through the adapter to con® rm completion, or sendnotice of an error in execution. The system’ s openarchitecture, in conjunction with the CLARATM soft-ware, allows for ¯ exibility and ease when modifying orimplementing new assay platforms and technologies.

The CLARATM Data Manager gives you the ability tomanipulate and ® lter incoming data with the use of theData Manager ’ s ZyMap. Data can be con® gured intotext ® les, Excel sheets or directly transported to adatabase via ODBC connections. However, the currentversion of the Data Manager is unable to catch criticaleventsÐ events that would trigger the data to be sent tothe database, such as an ` end of run’ event. This inabilitydoes not yet allow the direct triggers to export data to adatabase. A temporary remedy supplied by Zymark is adeveloper ’ s kit (a collection of APIs) that gives you theability to build your own Data Manager using prede® nedtemplates. Using this method could require extensivetime and programming on the user’ s end depending onthe process and data being captured. However, anupdated version of the Data Manager is anticipated tobe released at the end of October that would alleviate thisproblem.

Although the StaccatoTM system has many bene® ts andcapabilities for liquid-handling needs, there were someimplementation issues along the way. The biggest setbackexperienced was with limitations of the Data Manager.One of the top reasons for the selection of the Zymarksystem over other vendors ’ was the ease of databaseconnectivity and data export. Another issue was in thedesign of the ® xed tip arrays. Great care must be takennot to damage their Te on coating. Any scarring orscraping of the material will cause variation in pipettingprocedures especially at small volumes. Service for themost part is good; however, the redesign and complexityof the system left the need for more time in training oftechnicians to come up to speed in the overall workings ofthe system.

Overall, Neurogens experience with the StaccatoTM

system has been a good one. All systems have theirbene® ts and drawbacks. The most important thing isthe cooperation and speed of assistance received whenproblems do arise and eæ ciency in coming up withsolutions. In this aspect, Zymark has shown dedicationand commitment to their products and customers.

Update from PQRI’s Blend Uniformity WorkingGroup: balancing workload with batch homogene-ity assurance

Garth Boehm, Purepac Pharmaceutical Co., Elizabeth, NJ, USA

The Product Quality Research Institute (PQRI) is aconsortium between industry, academia, and FDA,which aims to provide a scienti® c basis for the develop-ment of regulatory policy. One of PQRI’ s ® rst initiativeswas to form a Blend Uniformity Working Group(BUWG) charged with providing a scienti® c basis forcontinued development of FDA policy on BU testing.The BUWG sought wider industry input through asurvey of current BU testing practices, and a publicWorkshop. Using this input together with the experienceof the BUWG members and colleagues, a draft proposalhas been developed and re® ned. The proposal provides aguide for appropriate testing during scale-up and valida-tion, and for subsequent routine manufacture. Theproposal is based around initially establishing the rela-tionship between BU testing and strati® ed testing of theresulting dosage units. The type and amount of testingrecommended during routine manufacture depends onthe outcome and robustness determined from the valida-tion testing. The BUWG is currently seeking data fromindustry to challenge the proposal with real data toestablish the validity of the suggested approach. Thenext step will be to seek public comment on the BUWGproposal.

Views from the past: HTS in the Molecular Bio-chemistry Department, Glaxo Wellcome, Inc.

Brent Butler, GlaxoSmithKline, Research Triangle Park, NC,USACo-authors: Cole Harris and Steve Blanchard

The pharmaceutical industry saw a tremendous increasein screening throughput in the 1990s with the intro-duction of advanced robotics and liquid-handlingsystems, as well as homogeneous assay formats conduciveto higher density assay plates. During this time, therehave been numerous debates on the advantages anddisadvantages of centralized versus decentralized screen-ing sites. There have also been discussions about the useof large, integrated robotic systems compared with smallbench-top robotics.

Glaxo Wellcome Research Triangle Park has screened ina decentralized manner using both large automationplatforms and small bench-top systems. In addition, thescreens have migrated from 96- to 384-well plates. Thistalk will focus on the lessons learned from the screeningeå ort in the Molecular Biochemistry Department ofGlaxo Wellcome.

Fully automated screening of intracellular cal-cium using a novel detector

Brian Rasnow, Amgen, Inc., Thousand Oaks, CA, USACo-authors: Jack Lile, Karen Kearns, James Treanor and PeterGrandsard

We have constructed a fully automated high-throughput ,cell-based screening system that measures aequorin lumi-

Abstracts of papers presented at the 2001 ISLAR

64

nescence kinetics as an indicator of intracellular calciumconcentration. The system uses a modi® ed ZymarkRapidplate96 pipetting station to add drug candidatesand known modulators to aequorin-expressing cellsgrown in microtitre plates. Simultaneous with compoundaddition, custom optics and a high-performance chilledCCD camera added to the Rapidplate96 take an imagesequence of the plate bottom. Compounds that aå ectcalcium entry or release from intracellular stores can beanalysed using this system. Agonists and antagonists canbe detected in the same screen by imaging both duringcompound addition and during a subsequent addition ofknown agonists.

The 96-channel imaging ¯ ash luminometre is integratedwith a Beckman CO2 incubator, Zymark Twister, tipstore, ORCA robot, and other peripherals to performfully automated screening of up to 72 microtitre plateswith a cycle time of several minutes per plate (about10 000 high-content, cell-based assays/system per day).The system is controlled with Amgen’ s automation soft-ware (` Synchronicity ’ ). Analogue and digital signal pro-cessing are performed in real time to remove artefactsfrom cosmic rays and bad pixels, and normalize anysystematic eå ects from geometrical and biological hetero-geneity across the plates. This reduces the about 1.5 MB/plate of raw data to about 10 kB of time series, which canbe further compressed and the compounds classi® ed usingprincipal component analysis and other methods.

We have validated the system for ion channel and GPCRHTS and dose± response characterization, and it hasscreened thousands of plates. We will discuss and contrastits design and performance with other devices that havecomparable capabilities, and explore some of the dataand information processing challenges presented by high-throughput, high-content screening systems.

Automated production of medicinal agents fromplant matrices

Bruce E. Richter, Dionex Corporation, Salt Lake City, UT,USACo-author: Richard E. Carlson

Interest in the pharmacologically active compoundsfound in plant tissues is growing. The extraction tech-niques normally used to remove these compounds fromplant tissues require long periods and copious amounts ofsolvents. Accelerated solvent extraction (ASE) has beenproven eå ective in removing target compounds from avariety of plant tissues. Using ASE, the extraction ofcompounds from medicinal plants is completed in about15 min using only 20 ml solvent. Extraction of the targetcompounds may be only a part of the isolation process.Interfering compounds such as waxes, pigments andtannins can be coextracted.

These interferences must be removed before the extractscan be subjected to screening techniques. Solid-phaseextraction (SPE), liquid± liquid extraction (LLE) andpreparative liquid chromatography are the techniquesmost often used for removing interferences from plantextracts. However, these steps are separate and notcoupled with the extraction process.

This presentation will discuss the coupling of ASE withliquid-handling apparatus to produce plant extracts freefrom interfering coextracted compounds. This combinedsystem produces natural product extracts that are readyfor biological or chromatographic assay. The comparisonof results from this procedure to those obtained withoutautomation will be discussed.

Determination of cytotoxic e ects of di´ erentcompounds using a multiparametric cell-basedcytotoxicity assay

Chandrasekaran Vasudevan, Cellomics, Inc., Pittsburgh, PA,USACo-authors: Oleg Lapets, Nancy Kerkvliet, Greg LaRocca,Rachael Krally, Lisa Lemmler, Patricia Petrosko, Megan Weiss,Joseph Zock and Richik Ghosh

In vitro screening of active compounds for toxic eå ectsearlier in the drug-discovery process will help identifyproblems and reduce the high failure rate of currentclinical candidates. Cell-based toxicity assays are crucialin this eå ort, but are limited by the result generated,usually cell death. Thus, we have developed a moresensitive multiparameter cell-based cytotoxicity assaythat quanti® es early changes to key aspects of cellularphysiology that can lead to cellular toxicity.

This assay uses our ArrayScan1HCS System HighContent Screening (HCS) platform that simultaneouslymeasures changes in nuclear morphology, cell perme-ability and lysosomal physiology for individual cells, andalso changes in cell density in microplate wells. Dose±response and time-course data demonstrating the assay’ smultiparametric nature will be presented for samplecompounds that aå ect the cellular targets in variouscultured cell lines and also rat primary hepatocytes.Our data show the assay’ s ability to capture cellularreactions to compounds and correlate multiple toxicolo-gical indicators at a single-cell level. This assay representsa distinct advantage for HCS as a screening tool early inthe drug-discovery process.

Development of a high-throughput, homogeneous,cell-based assay for screening inhibitors of mul-tiple drug-resistance pumps on FLIPR or FLEX-Station systems

Christopher Silva, Molecular Devices Corporation, Sunnyvale,CA, USACo-authors: Kelly J. Cassutt, Anne T. Ferguson and Jesse Twu

Multiple drug resistance (MDR) pumps expressed on thesurface of normal cells are involved in eliminating toxiccompounds generated by or exogenously introduced tothe cells. In cancer chemotherapy, these pumps maybecome over-expressed by the tumour cells and renderthe drug treatment ineå ective following relapse. MDRpumps are also normally expressed at the blood-brainbarrier (BBB) and as such give the brain ` sanctuary ’ fromchemotherapeutic drugs.

Our goal at Molecular Devices is to provide a tool toscreen libraries for compounds that inhibit the MDRpumps. Such compound leads could be important for

Abstracts of papers presented at the 2001 ISLAR

65

inhibiting MDR, which might otherwise lead to failure ofthe therapeutic regimen. The compound leads identi® edmight also help facilitate the absorption or retention ofother useful therapeutics for disease treatment.

The assay is designed primarily for screening inhibitorsagainst the two most common MDR pumps, P-glycopro-teins (P-gps) and multidrug resistance-associated proteins(MRPs). Both classes of pumps are known to expel avariety of mechanistically and structurally unrelatedcytotoxic drugs such as anthracyclines, tacanes, vincaalkaloids, and epipodophyllotoxin s as well as some ¯ uor-escence substrates or indicators from the cells. For a givenassay the compound library is prepared in a 96- or 384-well plate and allowed to incubate with cells at roomtemperature for 15 min or longer. The microplate isplaced in the FLIPR or FLEXStation System followedby addition of the ¯ uorescence indicator by the instru-ment. The kinetic readout of the results can then becollected in a real-time mode. The results are expressed asthe change in relative ¯ uorescence units (RFU) overtime, usually 5-8 min after the start of the experiment.Inhibition of the MDR pumps is indicated by theincrease in RFU above the baseline control.

Using the cell line MCF-7/ADR* (resistant to adriamy-cin), we showed that the baseline change in countswithout inhibitor is nearly zero while 30 mm cyclosporinA (a known inhibitor to P-gp) increased the RFU to> 5000. These results indicate that the P-gp expressionlevel is high enough to keep dye completely out of thecells and the MDR pumps were inhibited eå ectively bycyclosporin A. Various tumour and non-tumour cell linesincluding the parental MCF-7, Caco-2, and T-cell lineCEM were also tested and showed parallel results. Verylittle or no DMSO interference to the assay was observed.

We have developed a robust and reproducible homo-geneous cell-based assay for high-throughput screening ofcompound libraries for inhibitors of MDR pumps. Theease of use inherent in the format with no wash stepinvolved along with the rich kinetic data previouslyunavailable to the large-scale MDR studies indicate thatthe FLIPR MDR Assay Kit will be a time-saving andcost-eå ective tool in HTS screening of inhibitors of MDRpumps.

Custom hit-picking robotic system

Claude Dufresne, Merck Research Laboratories, Rahway, NJ,USACo-authors: Christopher Napolitano and Keith Silverman

An in-house integrated automated ` cherry picking’system will be described. It is designed to provide backupsamples of natural products’ ` hits’ to biologists perform-ing primary screening. The system uses a TomtecMegaStor for high-capacity source plate storage. Thesystem is controlled by custom software, written in VisualBasic 6.0, and makes use of ActiveX controls for each ofthe system components.

Direct analysis of basic drugs in cell culture lysateusing online extraction LC/MS/MS

Claude R. Mallet, Waters Corporation, Milford, MA, USACo-authors: Je¡ Mazzeo and Uwe D. Neue

During the last 10 years, pharmaceutical companies haveconstantly pushed for shorter analysis time to breach the1000 analyses per day barrier. With this demand forhigh-speed analysis, new techniques such as 96-wellplates, fast gradients or ultrahigh- ¯ ow chromatographyare showing promising results.

We have focused on online extraction techniques forhigh-throughput analysis. In a previous study we inves-tigated the potential of online extraction for the analysisof basic and acidic drugs in rat plasma. Recently, wehave turned our attention to other types of matrices, e.g.the study of toxicity of drugs in cell cultures used in pre-and post-discovery phases. Several chemical and physicallysate methods were evaluated. The extracts were in-jected onto an extraction column at high ¯ ow rate (i.e.4 ml min¡1) [1± 3] to remove macromolecular compoundssuch as proteins, but trap smaller analytes on the head ofthe column. Several con® gurations for direct injection arepossible. In the simplest con® guration, the extractioncolumn is connected directly to the MS/MS system.Other versions are con® gured with a single or a dualextraction column coupled to an analytical column. It isoften necessary to split the ¯ ow. However, in cases wheresensitivity is low, this option is not recommended. Foreæ cient high-speed analysis, the use of a second pumpand a 10-port valve is also a good choice. One line (high¯ ow rate) can be dedicated to the extraction column,while the other (low ¯ ow rate) drives the analyticalcolumn and the mass spectrometer.

A three-valve con® guration using two extraction columnswas used for the analysis of basic drugs in cell culturelysate. The online analysis was performed on an OasisHLB extraction column (2:1 £ 30 mm, 25 mm) using aWaters Alliance 2790TM in the gradient mode and a 515stand-alone pump in the isocratic mode. The extractedanalytes were forward ¯ ushed into an XTerra column(2:1 £ 30 mm, 3.5 mm), which was added to provideadditional separation power. The drugs were quanti® edusing a MicroMass UltimaTM triple-quadrupole massspectrometer equipped with an electrospray source andset in multiple reaction-monitoring mode (MRM).

References

1. Xia, Y. Q., Whigan, D. B., Powell, M. L. and Jemal, M., 2000,Rapid Commun. Mass Spectrom., 14, 105.

2. Ayrton, J., Dear, G. J., Leavens, W. J., Mallett, D. N. andPlumb, R. S., 2000, J. Chromatogr. A, 828, 199.

3. Eeckhout, N. V., Perez, J. C., Clearboudt, J., 2000, Vandeputte,R. and Van Peteghen, C. Rapid Commun. Mass Spectrom., 14, 280.

Abstracts of papers presented at the 2001 ISLAR

66

A look to the future: screening in the SystemsBiology Department, GlaxoSmithKline , Inc.

Cole O. Harris, GlaxoSmithKline, Research Triangle Park, NC,USACo-authors: Brent T. Butler, Steven G. Blanchard and David C.Morris

Screening at the newly formed GlaxoSmithKline Re-search Triangle Park site will be heavily dependent onoptimizing the successful processes incorporated over theprevious 5 years in the Molecular Biochemistry Depart-ment at Glaxo Wellcome. This will require expansion ofthe developing approaches to research aided by successfulutilization of existing automation equipment. GlaxoS-mithKline will compartmentalize these processes to ex-ploit the bene® ts of speci® c strategies, further de® ning theprocess of moving chemistry towards hit identi® cation.This discussion will focus on accommodating a largerinfrastructure that provides greater target numbers andlarger libraries, as well as complete follow-up screeningusing the automation resources available.

Distributing mammalian cell lines for HTS usingthe Hydra-96 Microdispenser

Courtney T. Ward, MDS Pharma Services, Bothell, WA, USA

Reliable and consistent distribution of cells into micro-plates is a key aspect of the performance of cell-basedscreening assays. Typically, this has been done using ahand-held multichannel pipettor in a biosafety cabinet, atask that is both labor-intensive and slow. We have foundthat this step can be accelerated signi® cantly using a 96-channel pipettor such as the Hydra-96 Microdispenserwhose physical pro® le permits the use of the instrumentin a standard tissue-culture hood. This device consists ofan array of 96 glass syringes (290 ml) with ® xed needleswhere the plungers move in unison under microprocessorcontrol. Using four mammalian cell lines, we comparedthe pipetting performance of the Hydra-96, a hand-heldpipettor, and a disposable tip-based 96-channel pipettingworkstation. The results indicate that the well-to-welldispense precision and cell viability are equivalent for allthree devices.

Automated o² ine solid-phase extraction for polarnew chemical entities using the Zymark RapidTrace

TMsystems

Daksha Desai-Krieger, R. W. Johnson PRI, Springhouse, PA,USA

A generic automated o¥ine SPE method using theZymark Rapid TraceTM systems was developed for threevery polar new chemical entities (NCEs) undergoingdevelopment at RWJPRI, namely a neuroaminidaseinhibitor, a cephalosporin and a thrombin inhibitor.The polar characteristics of these compounds made itimpossible to carry out liquid± liquid extraction as asample clean-up step for these compounds.

Our goal was to obtain simple automated SPE methodsas sample clean-up for these polar compounds frombiological matrices and develop quantitative LC-MS/

MS assays for these compounds. Challenges encounteredduring assay development and details of the method willbe presented. Transfer of these automated o¥ine assaysto online automated SPE and/or o¥ine automated 96-well SPE will also be discussed.

Development and validation of an automated pool-ing process for a compliant preparation of bloodplasma pools destined for molecular testing

Danuta Kierek-Jaszczuk, Cangene Corporation, Winnipeg,Manitoba, CanadaCo-authors: Mike Labossiere, Beverly Belanger, Patti Hosler andLeeAnne Macaulay

Blood-borne viruses such as HCV, for example, pose adanger of being transmitted from aå ected donor(s) ontothe recipients of the blood/plasma or thereof derivedpharmaceuticals. The use of nucleic acid testing/nucleo-tide-ampli ® cation testing (NAT) was shown to shortenthe ` window period’ , i.e. a period when donors are in astate of active but serologically undetectable infection.When applied to the plasma pools assembled from thedonations destined for the manufacture of the plasma-based therapeutics, HCV-speci® c NAT would decreasethe risks of HCV transmission onto the recipients of suchtherapeutics if virus-contaminated donation(s) are iden-ti® ed and removed from the fractionation pool.

Accordingly, an automated plasma-pooling process andpooling-driven back-tracing algorithm have been devel-oped and validated at Cangene to support the produc-tion of WinRho and VariZig therapeutics bearing lesserrisk for HCV transmission. The validated MicrolabATplus 2 automated liquid handler has been used forthe preparation of the plasma mega (master) pools in athree-stage process where ® nal mega pools were as-sembled from the intermediary secondary pools andthese, in turn, from the subpools. The automated bar-code-reading function of the pooling instrument, in acombination with a suitable barcode-labelling system ofthe individual plasma donation- and pool-holding con-tainers, allowed for the positive identi® cation of both theindividual donations and assembled pools. The pipettingreports generated by the instrument facilitated a fulltraceability of all pipetting events whereas gravimetricin-process veri® cation assured pipetting accuracy.

A pooling scheme-driven back-tracing algorithm has alsobeen developed and validated for singling out the indi-vidual contaminated donation(s). This algorithm in-volved four-stage back-testing and was based on theassumption that all NAT-positive test results are true-positives. It allowed for identi® cation of viral contamina-tion of a single donation by testing 26 samples, i.e. one® nal pool, ® ve secondary pools, 12 subpools and eightindividual donations. The developed and validated pro-cesses allow for cGMP-compliant preparation of theplasma pools and regulatory required back-tracing ofthe viral contamination to individual plasma dona-tion(s).

Abstracts of papers presented at the 2001 ISLAR

67

Extending the reach of miniaturization technologyacross the screening laboratory

David A. Dunn, Pharmacopeia, Inc.Co-authors: James R. Beasley, Paul McCoy, Barbara Strohl,Robert Swanson, Ti¡any Walker and Jingchun Yang

Assay miniaturization and the implementation of high-density 1536-microwell screening is vital to our need toincrease the eæ ciency of primary screening and earlyphase lead discovery. To serve this need, we havedeveloped a 1536 uHTS platform that employs microwellplates, novel ¯ uid handling and optical detection tech-nologies. Full use of this platform requires the ability todevelop assays eæ ciently for a variety of biological targetsin a straightforward and expeditious manner. Having aportfolio of miniaturizable assay technologies that can beeasily developed into ultrahigh-throughpu t screens forany member of a class of biological targets wouldenhance and extend the utility of this technology. Wehave evaluated a variety of assay technologies that aresuitable for uHTS of kinase and GPCR targets. Thispresentation will discuss the relative merits of thesetechnologies in assembling a portfolio of uHTS assaysfor screening these target classes.

Prequali � cation of pharmaceutical leads

David Casebier, ArQule, Inc., Woburn, MA, USA

Reliable predictive models and increased throughput ofADME screens currently enable the design, constructionand pro® ling of chemical libraries, providing con® dencefor series selection in hit-to-lead discovery. Predictivemodels for metabolism, absorption, BBB as well asdiversity measurements assist in library design andreagent selection. Robotic syntheses then generate spa-tially addressed, high-quality libraries, which in turn arepro® led for ADME characteristics. Integration of thesetechnologies allows for the prioritization of lead seriesusing data indicative of developmental survivability.

Automation of identi� cation and screening ofgenomic targets related to psychiatric disease

David M. Evans, Psychiatric Genomics, Inc., Gaithersburg,MD, USA

Psychiatric Genomics, Inc., is interested in discoveringthe underlying genetic causes of psychiatric diseases andidentifying therapeutics for them. Many of these diseasesare multigeneic in nature and have not previously beenwell characterized owing to the diæ culty in obtaining theappropriate tissue samples and the lack of technologiesthat could detect the changes in gene expression pro® leswithin the brain regions.

In recent years, several tools and technologies have beendeveloped that allow multiple parameters to be tested ina single experiment. Micro-array technology allows thediscrimination of changes in many genes at one time, andby comparing ` normal’ versus diseased tissue, it is poss-ible to identify gene expression patterns quickly that arealtered due to the disease. Having identi® ed the diseasegene pro® le, it is important to be able to develop screensto examine whether small chemical compounds can

revert the gene pro® le of the diseased state back towards` normal’ .

This presentation will describe some of the approachesbeing taken by Psychiatric Genomics, Inc., to automateeach of the steps in the drug-discovery process using thenew paradigm of automated drug discoveryÐ ` Multi-Parameter High Throughput Screening’ (MPHTSTM).

Comparison of ELISA between automation andmanual testing for measuring ARANESPTM inrat serum

De Chen, Amgen, Inc., Thousand Oaks, CA, USACo-authors: Yan Wang, Monica Zordich, Han Gunn and SharonBaughman

This presentation compares the results of measuringARANESPTM in rat serum, obtained manually or viaa fully automated immunoassay system. ARANESPTMis a new erythropoietic protein currently used in clinicaltrials for the treatment of anaemia related to chronicrenal failure.

The Tecan Genesis workstation 200 was developed toperform the ARANESPTM quantitative ELISA. It wasequipped with eight pipetting ceramic needles capable ofpipetting 5± 1000 ml, a shaking incubator, a Columbusmicroplate washer, microplate hotels and a microplatereader. A robotic arm transported microplates and lidsbetween the peripherals on the workstation deck.

Seven independent studies were performed on variousdays by the TECAN and manually. Standards, qualitycontrols and known samples were loaded in triplicate onthe plates in each assay. The data generated showed nosigni® cant diå erence. ANOVA analysis con® rmed thisconclusion.

Fully automated systems can be introduced to quantita-tive immunoassay for protein measurement in serum.Technical advances in methodologies, robotics and com-puterization will lead to signi® cant enhancements in thecapacity of immunoassay, resulting in signi® cant cost andeæ ciency gains.

LC-MS/MS assay for the quantitation of RWJ-270201 using the Zymark Rapid Trace SPE work-station

Denise Preston, R. W. Johnson PRI, Spring House, PA, USACo-authors: Daksha Desai-Krieger, Virginia Scott and R. JohnStubbs

A simple, automated SPE extraction and LC-MS/MSassay was developed for the extraction and quantitationof two NCEs, RWJ-270201 and RWJ-270204, beingevaluated as viral neuroaminidase inhibitors intendedfor use in the treatment of in¯ uenza A and B. A solid-phase extraction method was developed for these analo-gues using Zymark Rapid Trace SPE workstations.

Extraction of plasma samples was performed using ISTPhenyl SPE cartridges (200 mg 3 ml¡1) conditioned withmethanol and water. Sample clean-up was achieved byrinsing cartridges with methyl-t-butyl ether. The com-pounds of interest were eluted using ethanol:water (90:10

Abstracts of papers presented at the 2001 ISLAR

68

v/v) and subsequently evaporated under nitrogen todryness. The residue obtained was reconstituted beforeinjection on an LC-MS/MS in the positive-ion APCImode.

The developed assay demonstrated good precision andaccuracy and was linear over a broad curve range. Themethod developed for these analogues was routinely usedto support preclinical pharmacokinetic studies and waslater adapted for use with 96-well plate technology insupport of clinical studies.

Fully automated protein precipitation LC/MS/MSassay using a 96-well plate technology and trackrobotic system

Scott Paul Depee, GlaxoSmithKline, Research Triangle Park,NC, USACo-author: Brent T. Butler

The need arose in our department to develop a faster andsafer yet user-friendly means of sample preparationbefore sample analysis. With shorter mass spectrometermethod run times, the rate-limiting factor was thesample-processing procedure. To expedite this pro-cedure, automation through robotics and 96-well platetechnology was implemented. Not only is this roboticprotein precipitation assay process faster than manualpreparation, but also the precision and accuracy arecomparable. In addition, our staå has less exposure topotential pathogens in the samples, therefore increasingsafety. This robotic system is operated by very user-friendly software that enables the processing of sampleseither in a fully or semi-automated manner. The systemcan manage 96 samples in < 1:5 h.

Automated o² ine solid-phase extraction for polarnew chemical entities using the Zymark RapidTrace systems

Daksha Desai-Krieger, R. W. Johnson PRI, Springhouse, PA,USA

A generic automated o¥ine SPE method using theZymark Rapid Trace( systems was developed for threevery polar new chemical entities (NCEs) undergoingdevelopment at RWJPRI, namely a neuroaminidaseinhibitor, a cephalosporin and a thrombin inhibitor.The polar characteristics of these compounds made itimpossible to carry out liquid± liquid extraction as asample clean-up step for these compounds. Our goalwas to obtain simple automated SPE methods as sampleclean-up for these polar compounds from biologicalmatrices and develop quantitative LC-MS/MS assaysfor these compounds. Challenges encountered duringassay development and details of the method will bepresented. Transfer of these automated o¥ine assays toonline automated SPE and/or o¥ine automated 96-wellSPE will also be discussed.

New technique for a high-throughput solubilityassay

Dima Voloboy, pION, Inc., Woburn, MA, USACo-authors: Chau M. Du, M. Stra¡ord, K. Tsinman and A.Avdeef

We have successfully adapted the classical shake-¯ askmethod to the 96-well microtitre plate format to obtainnear shake-¯ ask-quality results at high-throughputspeeds. The new instrument uses a robotic ¯ uidics systemand a parallel detection system employing a microtitreplate UV (190-500 nm) spectrophotometer . Samples areintroduced as DMSO stock solutions.

The essence of the approach is based on making UVconcentration reference solutions from moleculesÐ espe-cially ionizable moleculesÐ whose UV spectroscopicproperties are unknown at the start of the assay. Thiscan be achieved in either of two ways, with or withoutcosolvents.

A unique computational method was developed to ex-tract the aqueous intrinsic solubilities of drug moleculesfrom data distorted by DMSO drug binding or drug±drug aggregation reactions.

An improved method for determining concentration byUV spectrophotometry was derived. It uses a novel peak-shape algorithm for adjusting weights in a whole spec-trum-weighted regression analysis, matching spectra ofreference solutions (of known concentration, under con-ditions avoiding or suppressing precipitation) to solutionscontaining an analyte of unknown concentration (due toprecipitation).

A quality-assessment scheme has been developed aroundthe standard drugs diclofenac, indomethacin, ¯ urbipro-fen, chlorpromazine, phenazopyridine, verapamil, pirox-icam and griseofulvin.

The new solubility method, which allows determinationat one pH or as many as 96 pHs, has a limit of detectionof about 0.1 mg ml¡1, as demonstrated by measurementsof the intrinsic solubilities of terfenadine and tamoxifen.

About 200 assays may be performed per day on theinstrument.

Automated solution-phase synthesis work� ow atSchering AG

Christoph M. Huwe, Schering AG, Berlin, Germany

At Schering we have established a ¯ exible, integratedsolution-phase synthesis, puri® cation and reformattingwork¯ ow based on Chemspeed- and Zymark-automatedsynthesizers, parallel normal-phase chromatography aswell as reversed-phase HPLC-MS puri® cation equip-ment, and a highly automated Zymark reformattingrobot. In this presentation, the special features of thechosen equipment, the ideas behind the system designand the respective work ows will be discussed.

Abstracts of papers presented at the 2001 ISLAR

69

Intelligent automation of HPLC method develop-ment

Douglas R. Myers, Intelligent Laboratory Solutions, Naperville,IL, USACo-author: Je¡rey D. DeCicco

Developing new HPLC methods is a time-consuming andknowledge-intensive task. We have developed an onlinesoftware package called ChromSmart that automatesHPLC method development for both normal and re-versed-phase chromatography for chiral and non-chiralsample mixtures. Unlike other software packages thatassist with o¥ine HPLC modelling, ChromSmart usesonline arti® cial intelligence technology to capture andimplement the method-development knowledge of expertchromatographers . ChromSmart contains a real-timeintelligent experiment planner and rule-based enginethat execute the HPLC methods, monitor results anddynamically update the development plan. A completeaudit trail of the decision-making and results is gener-ated.

ChromSmart comes with a core set of chromatographyknowledge that automates method development. Basedon user preferences, such as retention times and resolu-tion, ChromSmart develops a method-developmentstrategy. This strategy is executed and can dynamicallychange based on information from online, real-timeexperimental results. Features include automatedequilibration, detection and cleaning of retainedcomponents, gradient and isocratic optimization, peakpurity analysis, and peak tracking. When ChromSmart® nds a method that meets or exceeds the user’ s criteria, itnoti® es the user and continues to develop anothermethod for the next sample. ChromSmart operates24 £ 7 £ 365 signi® cantly increasing productivity ofequipment and personnel.

Laboratory automation using Zymark robots andUV � breoptics from Delphian Technology

Je¡rey B. Medwid, Wyeth-Ayerst Research, Pearl River, NY,USACo-authors: Jerry Edgar, Syed Rahman and Scott Keenan

In the fast-paced pharmaceutical business, quicker turn-around times and larger sample loads face every analy-tical laboratory manager. Unfortunately, this must beaccomplished with no signi® cant increase in analystresources. Laboratory automation is the only answer.Our research automation laboratory contains severalZymark robots including the Multidose DissolutionWorkstations and Tablet Processing Workstations . Thelatest addition to our laboratory is a UV Fiber OpticDissolution Workstation manufactured by DelphianTechnology. This presentation will highlight some ofour recent work accomplished using these workstationsand some of our other initiatives in automation.

‘Cradle to grave’ tracking of synthetic combina-torial libraries

Phil Small, Tripos Receptor Research, Bude, UK

A process-integrating design, synthesis and analysis ofcombinatorial libraries has been implemented at TriposReceptor Research. Crucial to the success of this processhas been the development of a proprietary informaticssystem. This has so far been developed to manage reagentinventory, track samples and record data from synthesisand analysis to provide a valuable database of compoundinformation.

The combination of Tripos’ proprietary design softwarewith automated synthesis and informatics leads to theproduction of drug-like libraries with well-de® ned purityand full synthetic history.

This presentation will cover what were considered theimportant aspects involved in setting up and managingan automated chemistry facility.

Imaging technology in ultrahigh-throughpu tscreening: overcoming the bottleneck of platereader throughput

E. Michael August, Boehringer Ingelheim Pharmaceuticals, Inc.,Ridge¢eld, CT, USACo-authors: Lori Patnaude and Carol A. Homon

Screening throughput is limited by its slowest step, andthus a major goal of uHTS must be to identify andovercome these bottlenecks. With the advent of high-density liquid-handling devices capable of deliveringliquid to 384 or more wells at rates approaching 1plate min¡1, the bottleneck has shifted from reagentdelivery to plate reading. Conventional plate readersrequire several minutes to obtain data from an entireplate.

Imaging technology allows the detection of an entireplate simultaneously. Coupled with rapid, sensitive im-age-analysis software, such a system enables data captureto keep pace with reagent delivery. We present herescreening results from a recent enzymatic Del® a uHTScampaign comparing the Wallac Viewlux Imagingsystem with the LJL Analyst. Several hundred plates inboth 96- and 384-well formats were read on both readersand the data were analysed with a variety of statisticalparameters.

Similar comparisons of other assay formats will bediscussed. The inherent ¯ exibility of imaging systems ishighly compatible with assay miniaturization and thetrend to higher-density plate formats. However, we mustalso realize that the bottleneck in the overall screeningprocess has not been removed, but surely shifted toanother step.

Finally the paperless and compliant laboratory

Ed Halpin, VelQuest Corporation, Hopkinton, MA, USA

Compliance and cost have long been at odds in themodern regulated laboratory. Focusing on one meantsacri® cing the other. Regulatory compliance today is a

Abstracts of papers presented at the 2001 ISLAR

70

labour-intensive, paper-based operation. The ` paperless’laboratory has been a goal of the pharmaceutical in-dustry for over the past decade. The con¯ uence oftechnology and innovation has reached a point wherethe paperless laboratory is now economically and tech-nologically practical. Good manufacturing practices and21CFR Part 11 de® ne the requirements to attain apaperless, compliant laboratory.

The solution would enable direct electronic data capturefrom human observations, instruments and networkedPCs replacing paper notebooks, ® les and control sheets.The solution would enhance the collection of data andmetadata by dynamically linking it to the procedure bywhich it was collected. Reviews, audits, investigationsand approvals would be electronic processes, replacingthe current paper systems. Redundant checking would bereplaced by ` review by exception’ and resources reallo-cated towards the real goal of bringing new humantherapies to the market more rapidly. The solution mustalso complement and exchange data with other labora-tory applications such as Laboratory Information Man-agement Systems (LIMS) and Chromatography DataSystems (CDS).

Generating predictive models from high-through-put pro� ling

Dragos Horvath, 128 rue Danton, Rueil Malmaison, FranceCo-author: Michael Entzeroth

Current drug development is characterized by a strongbelief in high-throughpu t methods, both in combinatorialchemistry and screening. The trigger for constantlyincreased performance is the need the industry to adjustto the race against time in order to secure shares on themarket place. Over many years the industry has tried tokeep pace by the constant introduction of new tech-nologies. The associated costs for drug developmentincreased over the last years. Aggregated R&D expendi-tures for new chemical entities (NCE) peaked in 1996close to US$500m.1 Recently, Lehman Brothers esti-mated the total costs for a launched product to accountup to US$635m,2 while for drugs entering clinical trialsbetween 1972 and 1982 these costs were estimated to addup to US$312m.3 Research-based companies will invest$26.4 billion in R&D in 2000, a 10.1 percent increaseover 1999.4 The number New Chemical Entities ap-proved by the FDA has not changed signi® cantly overthe last years (1997: 39, 1998: 30, 1999: 35), indicatingthat the success rate is not related to the initial eå orts.

The portion of IND applications that fail has beenestimated to approximately 20% in phase II, 60% inphase II and most importantly 87% in phase I.5 83% ofthe drugs for which INDs were ® led between 1964 and1989 were dropped before reaching NDA status.6;7 Forthe pharmaceutical companies each dropout during thedevelopment is associated with huge costs that reduce thepro® t derived from NCEs that reached the market. Thefailure of the candidates can be primarily attributed toeæ cacy, safety and economic concerns. The predominantreason for failure, however, are inappropriate pharma-cokinetic properties of the drug candidate (39.4%),

followed by adverse reactions and toxicity that add upto 21% of the causes for withdrawal from development.

In most of the cases the lack of a complete picture of thedrug’ s properties is not available at the time the decisionto select a respective candidate is taken. In part this isdue to the current process of drug discovery whereinformation is accumulated stepwise rather than in ade® ned parallel manner. The choice of the candidatetaken forward is often dominated by emphasizing theaspect of potency rather than addressing early on selec-tivity and drug property issues. On the other hand, thecompanies have accumulated vast amounts of in-vitro andin-vivo data that may be used to re® ne the decision matrixeither by retro- or prospective interpretation of existingresults. As a consequence and in order to take theserendipity out of such fundamental decisions, manypharmaceutical companies have started broad rangepro® ling (i.e. testing the drugs in parallel against a panelof diå erent in-vitro model, both in pharmacology andADME/PK) in the preclinical phase and have putemphasis on the development and automation of second-ary tests. High throughput pro® ling using robotics andautomated workstations allows to accumulate today abroader knowledge on the hits identi® ed in the primaryscreening. Both pharmacological Ð speci® city and selec-tivity versus other targets, such as enzymes and recep-torsÐ and pharmaceutical properties Ð physicochemicalparameters or metabolic and in-vitro safety aspectsÐ ofthe candidates are evaluated in depth.8

Through the progress in linking of high-throughputpharmacological , physicochemical and ` in-silico’ methodswith in-vitro approaches, progress is being made inpredicting drug properties.8 The goal to predict in-vitroand, ® nally, in-vivo properties of drugs has becomereasonable by linking the chemical structure via mol-ecular descriptors, area descriptors or ComPharm ® eldsto certain in-vitro pro® les that again can be correlated tothe corresponding in-vivo eå ects (® gure 1).

An example for pharmacophore ® ngerprints are theFuzzy Bipolar Pharmacophore Autocorrelograms(FBPA9) monitoring the numbers of atom pairs withineach of the 252 ˆ 21 £ 12 categories that can be de® nedin terms of the 21 combinations of six pharmacophoricfeatures a,b2{Hydrophobicity, Aromaticity, HydrogenBond Acceptor & Donor, Cation, Anion} times 12considered distance ranges ¢2{(3..4), (4..5), . . . ,(14..15)} [A® ]. For similarity searches and structuralcomparisons, these ® ngerprints are extremely powerfulin combination with ComPharm ® eld descriptors10 thatare taken as the intensities of the empirically de® ned` pharmacophore ® elds’ generated by the atoms of theconsidered compound at the space points occupied by areference structure, in a con® guration corresponding toits optimal alignment with respect to this reference.

The data mining approach described in this report is part

of the BioPrintTM project11 that makes use of a broadspectrum pro® ling of marketed drugs and candidates inmore than 90 diå erent in-vitro models. The marketeddrugs, included in the test collection, comprise more than1,500 compounds that are available in the US pharma-cies today. Important are also the drug candidates

Abstracts of papers presented at the 2001 ISLAR

71

included which have failed to make it all way through thediå erent phases of preclinical and clinical drug develop-ment. The models and predictive tools generated open anew approach for candidate selection and early discoveryactivities, such as target identi® cation or library designand will result in the long term in substantial cost savingsdue to reduced failure rates and time spent on theresearch and development process. In-silico, linear modelsto describe solubility and permeability as well as predic-tions in pharmacology based on predictive neighborhoodbehavior are presented.

Linear ADME models

The interest in an in-silico logD prediction model residesboth in its applicability in library design and its furtheruse as a calculated molecular descriptor in other struc-ture-activity relationships, if the activity under study isexpected to relate to the overall lipophilicity of com-

pounds. By contrast to logP models that mostly rely onincremental group contributions and/or molecular de-scriptors of the average polarity of the solvent-accessiblesurface (an aspect which is in our model successfullycaptured in terms of PTA descriptors), a logD modelneeds to account for proteolytic eå ects in the aqueousbuå er of pH ˆ 7.4 used for experimental determinations.In our models, proteolytic ionization of acids/bases isimplicitly accounted for by the pharmacophore descrip-tors relying on rule-based assignments of a positive/negative charge status to ionizable groups. Furthermore,bipolar pharmacophore elements or ® eld overlap termssuccessfully account for potential pKa shifts due toneighboring groups of the ionizable center.

Linear apparent permeability models

The main challenge in predicting the apparent perme-ability of compounds through a Caco-2 cell monolayer is

In Vivo Effects

�ADME/PK

�Adverse effects

�Therapeutic effects

In Vivo Effects

�ADME/PK

�Adverse effects

�Therapeutic effects

Molecular Descriptors Molecular

Descriptors Chemical Structure

N

F

F

N

NS

O

CH3

Chemical Structure

N

F

F

N

NS

O

CH3

N

F

F

N

NS

O

CH3

N

NS

O

CH3

In Vitro Profile

In Vitro tests

In Vitro Profile

In Vitro tests

PharmacophorePharmacophoreFingerprintsFingerprints

Area DescriptorsArea DescriptorsComPharmComPharm FieldsFields

Receptors/EnzymesReceptors/EnzymesCellular TestsCellular Tests

In VitroIn Vitro ADMEADME--TTpropertiesproperties

PharmacophorePharmacophoreFingerprintsFingerprints

Area DescriptorsArea DescriptorsComPharmComPharm FieldsFields

Receptors/EnzymesReceptors/EnzymesCellular TestsCellular Tests

In VitroIn Vitro ADMEADME--TTpropertiesproperties

Figure 1. Predictive modelling of the properties of compounds correlates chemical structures via molecular descriptors with ADME/ëK andtherapeutic e¡ects.

-4

-3

-2

-1

0

1

2

3

4

5

6

-4 -3 -2 -1 0 1 2 3 4 5 6

Expt. LogD

Pre

d. L

og

D

Figure 2. Predicted versus experimental log D. The properties of compounds from a diverse library of combinatorial compounds (squares)and medicinal chemistry (circles) were calculated from a model based on a training set of 320 entities (rhomboids).

Abstracts of papers presented at the 2001 ISLAR

72

the in-silico recognition of molecules which are trans-ported by one of the active in- or e¥ux mechanisms.This will enable the introduction of speci® c correctionfactors explaining the diå erence between the actualmembrane crossing rate and the rate expected on thebasis of purely passive transmembrane diå usion. Whilethe latter aspect of predicting the passive diå usion rateshas been tackled in various publications, little progresshas been made to understand the problem of transported/e¥uxed compounds, routinely discarded as ` expectable’outliers in previous work. By contrast, the herein de-scribed modeling eå ort focused on pharmacophore de-scriptors in order to search for potential pharmacophoremotifs that may characterize e¥ux and/or transporta-tion, while descriptors of the molecular surface propertiesand the generic calculated logD descriptors were sup-posed to account for passive membrane crossing proper-ties. The calculated logD (or, even better, theexperimental logD values) are relevant variables of thelinear log(PA± B) apical-to-basa l apparent permeabilitymodel. A positive coeæ cient for the linear contribution,together with a negative term in logD2 characterize anoptimal logD range maximizing the passive membranecrossing rate. This makes physical sense, since too hydro-philic compounds never leave the aqueous phase, whiletoo hydrophobic ones tend to accumulate in the mem-branes.

In spite of the important number of bipolar pharmaco-phore elements entering the model, a series of heavilye¥uxed compounds are nevertheless mispredicted by themodel and appear as ` false permeable’ . Most of theoutliers in the marked area are shown (® gure 3, circledarea) to be indeed e¥uxed molecules, since their meas-ured basal-to-apica l apparent permeabilities were sig-ni® cantly higher that the corresponding apical-to-basa lvalues.

It can be therefore concluded that either (a) e¥uxcannot be understood in terms of bipolar pharmacophoredescriptors only or (b) e¥ux might in principle becharacterized in terms of such descriptors, but the

number of descriptors that would be required to enterthe model is much larger than the number of examples ofe¥uxed compounds currently included in the data setÐnote that multiple e¥ux/transportation mechanisms withpotentially diå erent pharmacophore characteristics co-exist in Caco-2 cells, and that enough compoundsexemplifying each one of them would be required tosolve the problem.

The introduction of ComPharm descriptors overridingthe limitations of bipolar pharmacophore terms is indeedseen (® gure 3) to signi® cantly improve the permeabilityprediction of the outliers from the previous model. These® eld descriptors were obtained on the basis of super-imposition models of BioPrintTM compounds againstsome of the most heavily e¥uxed/transported substancesencountered in this set.

Solubility model

A categorical solubility model based on a linear equationhas been developed (® gure 4) in order to predict thesolubility class of the compounds: (1) Low: S < 10 mM;(2) Medium: 10 mM < S < 100 mM; (3) High:S > 100 mM. This model (not shown here) included 12explaining variables, such as logD and logD2 and variousPharmacophore Type Areas (PTA) and bipolar pharma-cophore elements that predicts drug solubility from itsstructural features. With encouraging results it has alsobeen compared with other solubility prediction softwarepackages and shown its advantages with 86% of thedrugs predicted in the correct solubility class.

Predictive neighborhood behaviour (PNB) models

By contrast with linear approaches that require a speci® cequation for each of the properties that are to bepredicted in terms of molecular descriptors, PNB modelsrely on the measured properties of related compounds(e.g. nearest neighbors in descriptors space) from the

-2

-1.5

-1

-0.5

0

0.5

1

1.5

2

2.5

-2 -1 0 1 2

Expt. LogPa-b

Pre

d. L

og

Pa-

b

Figure 3. Improvement of the permeability prediction of e¥uxed compounds due to the use of ComPharm ¢eld descriptors in the model.Training set (rhombs); test set (squares).

Abstracts of papers presented at the 2001 ISLAR

73

BioPrint set in order to provide an estimation of theproperty value of a new molecule. The PNB algorithmworks in principle like a ` fuzzy data base querying tool’ inthe sense that if molecule X is not included in theBioPrint database, the BioPrint compounds that are mostsimilar to X are retrieved and used for the evaluation ofthe properties of X. The property of the unknownmolecule is computed as a weighted average of theexperimentally determined properties PN(x) of its near-est neighbors N(X) from the BioPrint set, where theweighing factors are decreasing functions of the mol-ecular dissimilarity scores that de® ne how ` near’ aneighbor N(X) is with respect to X. These dissimilarityscores are Dice scores calculated on the basis of FBPA,PTA and EFO descriptors, the intervening empiricalparameters being speci® cally ® ne-tuned in order to maxi-mize the predictive power with respect to a particularproperty. Figure 5 shows the example in the case ofSCH23390, a D1 receptor ligand, which was not in-

cluded in the original data set, however, tested later forcomparison.

Conclusion

During the last decade technologies such as ultrahigh-throughput screening and combinatorial chemistry havesigni® cantly contributed to the advances in drug dis-covery. The increase in R&D costs associated and thebottlenecks further down the development as well as thehigh dropout rates have increased the demand for analternative approach. BioPrintTM combines moleculardescriptors, high-throughput pro® ling with in-vivo drugpro® les. It takes advantage of data mining technologiesto generate predictive models that link the three com-ponents. These models help to build a eå ective strategythat will foster drug discovery in the new millenium.

Expt. Class -> H M L Tot.Pred Pred. Class:H 265 19 2 286 Pred. Class:M 15 15 12 42 Pred. Class:L 0 0 12 12 Tot.ExptClass 280 34 26 340

Predicted in correct class (green)=(265+15+12)/340=86%

Predicted in neighboring class (yellow)=(15+19+12+0)/340=13.5%

Predicted in WRONG class(red)=2/340=0.5% Figure 4. Comparison of the predicted solubility class (vertical) of compounds versus the experimental class (horizontal) using the

BioPrintTM solubility model.

Figure 5. Receptor pro¢le of SCH23390, 10 ·m. (upper) Predicted pro¢le (each bar represents the percentage inhibition in apharmacological model ‡ RMS prediction error) based on the activity of its nearest neighbours; (lower) experimental data in the sameassay panel.

Abstracts of papers presented at the 2001 ISLAR

74

References

1. Kettler, H., 1999, J. Comm. Biotech., 6, 60.2. Kettler, H., 1999, Updating the Costs of a New Chemical Entity

(London: OHE).3. DiMasi, J. D., Hansen, R., Grabowski H. G. and Lasagna, L.,

1991, J. Health Econ., 10, 107.4. PhRMA., Pharmaceutical Industry Pro¢le 2000, 20.5. Kennedy, T, 1997, Drug Disc. Today, 2, 436± 444.6. DiMasi, J. A., 1995, Clin. Pharmacol. Ther., 58, 1.7. Prentis, R. A. and Walker, S. R., 1986, Br. J. Clin. Pharmacol.,

21, 437.8. Smith, D. A. and van deWaterbeemd, H., 1999, Curr. Opin. Chem.

Biol., 3, 373.9. Horvath, D., 2001, in A. K. Ghose and V. N. Viswanadhan

(eds), Combinatorial Library Design and EvaluationöPrinciples, SoftwareTools and Applications in Drug Discovery (New York: MarcelDekker), 429.

10. Horvath, D., 2001, in M. Diudea (ed.), QSPR/QSAR Studies byMolecular Descriptors (New York: Nova), 395.

11. Entzeroth, M., Chapelain, B., Guilbert, J. and Hamon, V.,2000, JALA, 5, 69.

Powder-dispensing capabilities to the ZymarkBenchmate system

Erin C. Heritage, R. W. Johnson Pharmaceutical ResearchInstitute, Raritan, NJ, USA

One of the most fundamental routines for a chemist isthe preparation of samples from dry powdered samples.There have been many expensive attempts to automatethis process. We demonstrate powder dispensing usingan existing robot workstation and some commondisposables.

Powder dispensing was accomplished by building adispensing station on an integrated weighing robotÐthe Zymark Benchmate. The operating principle of thestation is based on applying a vacuum to an airlineconnected to a pipette tip. The system is designed totransfer dry powder from four-dram vials to 16 £ 100 testtubes. The user can specify target weight and replicationsfor each compound through a custom-designed VisualBasic program.

The sample vial is basically carried to the ® xed locationof the pipette tip, placing the tip into the powder. Thevacuum is applied to the airline and the vial removed.The vacuum holds a small amount of sample powder onthe pipette tip. A test tube is then moved into positionbelow the pipette tip, the vacuum is released and thepowder is dispensed into the 16 £ 100 test tube. Whenthe target weight is reached, sample preparation (e.g.dilution to concentration and mixing) can then beaccomplished on the same workstation.

uHTS of phosphatases using 2D-FIDA anisotropydetection

Flora Tang, Sugen, Inc., South San Francisco, CA, USA

Various parameters of ¯ uorescence measurement, such astranslational diå usion, brightness and rotation, havebeen widely recognized as readout fully amenable tohomogenous, miniaturized uHTS. A novel detectionsystem using Evotec’ s FCSplus reader, which combinesthe ¯ uorescence-intensity distribution analysis (FIDA)anisotropy detection with confocal microscopy, provides

a very sensitive and ultrahigh-throughpu t assay format.The collaboration between Sugen and Evotec OAI isfocused on phosphatases and has demonstrated successfuladaptation of 2D-FIDA anisotropy detection for phos-phatase assays at the 1 ml/well level. A discussion of thisassay format and the uHTS screening results on theautomated EVOscreen Mark II system for several phos-phatases will be covered.

Considerations in dissolution automation

G. Bryan Crist, VanKen Technology Group, Cary, NC, USA

Among the future challenges for the pharmaceuticallaboratory exists the necessity to automate dissolutionmethods to allow maximum throughput with optimumcompliance. This presentation will provide suggestionsfor determining the level of automation needed to obtainpeak performance as well as providing insight intoregulatory compliance issues. Associated topics to bediscussed include: novel dosage forms today and provid-ing analytical challenges in the future, alternate methodsof analysis, the dissolution apparatus of the future and theinvolvement of new analytical approaches. Instrumenta-tion ¯ exibility will be required in the dissolution labora-tory of the future to utilize fully pooled samplingtechniques, online UV, online HPLC, capillary electro-phoresis and even atomic absorption for nutritionalanalysis.

Semi-automation to total automation concepts includingin situ methods of analysis will also be discussed. As novelconcepts evolve each day, engineers working with scien-tists will lead to more eæ cient, better-integrated dissolu-tion instrumentation in the future.

Update from PQRI’s Blend Uniformity WorkingGroup: balancing workload with batch homogene-ity assurance

Garth Boehm, Purepac Pharmaceutical Co., Elizabeth, NJ, USA

The Product Quality Research Institute (PQRI) is aconsortium between industry, academia, and FDA thataims to provide a scienti® c basis for the development ofregulatory policy. One of PQRI’ s ® rst initiatives was toform a Blend Uniformity Working Group (BUWG)charged with providing a scienti® c basis for continueddevelopment of FDA policy on BU testing. The BUWGsought wider industry input through a survey of currentBU testing practices and a public workshop. Using thisinput together with the experience of the BUWG mem-bers and colleagues, a draft proposal was developed andre® ned. It provides a guide for appropriate testing duringscale-up and validation, and for subsequent routinemanufacture. The proposal is based around initiallyestablishing the relationship between BU testing andstrati® ed testing of the resulting dosage units. The typeand amount of testing recommended during routinemanufacture depends on the outcome and robustnessdetermined from the validation testing. The BUWG iscurrently seeking data from industry to challenge theproposal with real data to establish the validity of thesuggested approach. The next step will be to seek publiccomment on the BUWG proposal.

Abstracts of papers presented at the 2001 ISLAR

75

Potential impacts of implementing automateddissolution

Gerard Schneider, LEC Consulting, Blairstown, NJ, USA

The Sotax AT-70 Smart oå ers the ability to automatefully both USP Apparatus 1 and 2 dissolution methods.This system, used in conjunction with other eæ ciency-driven technologies such as premanufactured mediumand online spectrophotometri c analysis of samples, canprofoundly reduce the resources required for conven-tional dissolution testing while enhancing the consistencyof results. Most importantly, these eæ ciency gains can beachieved within the parameters of a compliance-con-scious industry.

The eå ects of implementing such technology will bediscussed. The discussion will address the eå ect thateæ cient, automated dissolution testing has on the prod-uct development area, where companies are focussed ondelivering products to market in increasingly less time, aswell as marketed product stability and release labora-tories where QC managers are asked to reduce sampleturnaround times without increasing resources. Examplesof eæ ciency gains will be given in both cases.

Assay automation: let humans do what they aregood at and take robotics to the limit

Gladys Range, Human Genome Sciences, Rockville, MD, USA

In the automation of any process, there are basic stepsthat when carefully suited and applied to a particularlaboratory assay or work¯ ow, serve as a tool for eå ectiveautomation. Biological validation, time and high-throughput screening assays are, by nature, very diå er-ent, and are designed with diå erent criteria in mind. Abiological assay is a complex system designed to yieldbiological activity using qualitative measurements. Issuesof cost, amount and quality of reagents, although import-ant, do not play a paramount role in development. Thebiological assay is usually a manual assay that requiressigni® cant change before it is automation ready.

High-throughput assays in contrast are speci® c, fast,reproducible, low standard deviations, low cost andusually can be automated as much as mechanically andhumanly possible.

Key elements in assay automation are sample sourcing,screening design, robotics’ hardware and data manage-ment. These elements need individual careful analysis, amaster plan and follow-up. A challenge in automation isto enable individual laboratory personnel to contributeto the ® ne ongoing activity of building a database bydoing their work without stopping to write in a labora-tory book or even to enter information in a computer.These in return will prevent redundancy and will im-prove accuracy and completeness.

Second-generation laboratory robotics and liquid hand-lers are now designed to handle very high-throughputs ,very dense formats and very small-volume applications.Likewise, the recent development of sophisticated ¯ uor-escence systems operating at longer wavelengths, usingtime-resolve signals, and ® breoptics’ technology reduceread time and facilitate moving away from less desired

chemistries and separation steps. While some of the verylarge complex robotics systems could automate thecomplete sequence of operations comprising an assay, ahigher level of productivity is achieved by paying carefulattention to the work¯ ow. Determining the optimumassay conditions, standardizing some of the operationsand selectively introducing automation bring in the bestresults and bene® ts.

Improving the quality and speed of automatedsample transfers

Glenn Smith, GlaxoSmithKline, Research Triangle Park, NC,USACo-authors: Jimmy Bruner, Charles Buckner and Bob Biddle-combe

The quantitative transfer of biological ¯ uids is themost critical aspect of automated sample preparation.The sampling of plasma, serum, blood or urine representsthe most diæ cult and time-consuming step in mostbioanalytical extraction procedures. We will presenttwo practical solutions that address these separateproblems:

. automating ` diæ cult’ sample transfers with higherquality;

. signi® cantly increasing the processing speed of most` simple’ sample transfers.

First, most laboratories automate sampling on a mul-tiple-tip liquid-handling workstation such as a PackardMultiPROBE1 or Tecan GenesisTM. Today, samples aremost commonly transferred from various size tubes to adeep-well plate or a 96-well solid-phase extraction (SPE)block. Despite their widespread use, these systems cannotperform sampling with suæ cient precision and accuracyfor certain applications (e.g. samples containing clots orother insoluble matter; samples with low or insuæ cientvolumes).

We will show how the integration of an analyticalbalance on an x-y-z liquid-handling deck can easilyimprove the overall accuracy of assay data. The applica-tion also allows the workstation to process samples thatwould otherwise need to be transferred manually.Samples are aspirated in parallel then dispensed one ata time, allowing the balance to capture the exact weight(ˆ> volume) of each sample that was added to a deep-well or SPE plate. The data are used to correct auto-matically for any diå erences in sample weights, usuallyby pasting the results directly in the ` dilution factor’column of an analytical system’ s worklist.

Second, for most other applications in which reliabletransfers are easily achieved without gravimetric con-® rmation, the speed of sample processing remains a rate-limiting step. Most systems require about 20± 45 min totransfer 96 samples. This time is signi® cant and limitingcompared with the 1± 2 min required for other steps suchas the addition of a reagent using a 96-channel pipetting

station (e.g. Zymark RapidPlateTM-96, Tomtec Quadra

96TM, Apricot PP550TM).

We will detail how custom sample racks can be used tospeed up tube-to-plate sample transfers dramatically.

Abstracts of papers presented at the 2001 ISLAR

76

These novel racks have a standard microplate footprintand can be used on any 96-channel pipetting station. Theracks, which consist of complimentary pairs, each hold 48sample tubes arranged in an alternating (staggered)sequence. With these racks, 48 samples are transferredsimultaneously and 96 are transferred in as little as 2 min.The racks work with most brands of 13-mm outerdiameter cryotubes and 11-mm outer diameter micro-centrifuge tubes commonly used in bioanalytical labora-tories.

With these two inexpensive workstation modi® cations wecan now automate 100% of our applications and sig-ni® cantly speed up the majority of these assays.

Identi� cation of cytokine-regulated genes associ-ated with infertility and obesity by DNA gene-chipmicro-arrays

Grace Wong, Serono Reproductive Biology Institute, Randolph,MA, USACo-authors: H. Yarovoi, Q. Chen, S. Nataraja, Q. Xu, C. Liu,W. Mesadie J. Lai, J. Straaubhaar, J. Strickler, M. Tepper,M. Dreano, G. Garotta, S. Fumero, T. Wells, A. Eshkol and S.Arkinstall

Although the sequence of the human genome has beendetermined, the regulation and function of many genesremain largely unknown. Cytokines are tightly regulatedsoluble proteins transiently produced by cells in responseto immunological stimulation or disease manifestation.The expression of these proteins has been linked to manydiseases including cancer, AIDS, obesity, autoimmunity,immunode® ciency and infertility, suggesting that cyto-kines, their receptors, as well as cytokine-induced genesare potential targets for new drug development. Micro-arrays such as gene chips and protein arrays are high-throughput technologies for drug discovery in the post-genomic era.

We employed DNA transcriptional pro® ling micro-ar-rays to detect genes under control of cytokines in selectedexperimental models. Hence, in ovarian granulosa cells,TNF stimulates DNA synthesis but inhibits FSH-inducedoestrogen production. These results suggest that identi® -cation of TNF inducible ` master control genes’ mayreveal new targets for infertility and possibly otherdiseases associated with oestrogen. TNF induces morethan 70 genes in rat ovaries in vivo, including genes forcytokines, kinases, receptors, transcription factors andenzymes. In contrast to its inhibitory eå ect on oestrogenproduction in the ovary, TNF stimulates production ofthis sex steroid in human pre-adipocytes and adipocytesderived from more than 10 patients. Interestingly, hu-man pre-adipocytes/adipocytes themselves produce TNFas well as other cytokines and our data suggest a positiveautocrine/paracrine mechanism for control of oestrogenproduction by fat cells. In addition to TNF, 16 cytokinesincluding LT-a, IL-1, IL-6, IL-11, LIF and oncostatinM all stimulate oestrogen production in human fat cells.

Other cytokines such as TGF-­ 1, TGF-­ 2, VEGF,Rantes, IL-4, IL-10, IL-12, IL-18, CD40 ligand andTNF-related proteins such as LT-b, TWEAK, RANKLigand, APRIL and BAFF have no inducing eå ect.These results indicate that TNF has diå erential and

cell-speci® c actions on key endocrine mechanisms. Iden-ti® cation of genes induced or inhibited by TNF and othercytokines may lead to the discovery of novel targets fordrug development.

CrystalScreen : a novel microplate for automatedprotein crystallizatio n

Gunther Knebel, Greiner Bio-One GmbH, Frickenhausen,GermanyCo-authors: Lajos Nyaè rsik1, Patrick Umbach, Martin Horn,Thomas Przewieslik, Wolfram Saenger, Hans Lehrach, PeterOpfermann and Holger Eickho¡

The three-dimensional structure of any protein plays akey role in the process of understanding the exactfunction of these macromolecules. One proven approachto structural information of proteins is based on X-raydiå raction of single crystals. In the past, automatedcrystallization was restricted because a reliable hardwareplatform and well-suited microplates for high-throughputscreening were missing.

The Max-Planck-Institut e for Molecular Genetics(MPIMG), the Protein Structure Factory (PSF) andGreiner Bio-One have collaborated to develop a unique96-well protein crystallization microplate (Crystal-ScreenTM) with a standardized microplate footprint forhigh-throughput applications. Each of the 96 motherliquor wells corresponds to three crystallization wells.This allows checkerboard screening with up to 288crystallization options per plate to investigate optimalcrystal growth followed by 3D-structure analysis. Incombination with a preformed lid, this microplate en-ables high-throughput multichannel screening of proteinsin automated systems.

The system in use at MPIMG and PSF allows simul-taneous sitting-drop and hanging-drop vapour diå usioncrystallization experiments at reduced costs. A hugestorage system for 10 000 crystallization microplates anda pipetting device based on solenoid inkjet technologyenables one to set up a complete plate with 96 crystal-lization conditions in < 3 min. All crystallization wellsare inspected in regular intervals with an automatedcamera-based detection system to inspect the crystalgrowth.

Seamless integration of information in a pharma-ceutical development environment: integratingtechnology from LabWare, NuGenesis, VelQuestand Waters

Guy Talbot, Purdue Research Center, Ardsley, NY, USA

This presentation will describe Purdue Pharma’ s ap-proach to an integrated solution between LaboratoryInformation Management Systems (LIMS), Chromato-graphy Data Systems (CDS), Process Management andCompliance Systems (PMC) and Scienti® c Data Man-agement Systems (SDMS). It will describe the vendor-selection process and how Purdue Pharma created anenvironment to produce a best-of-breed information sol-ution to creating the paperless laboratory. Workingtowards the creation of a paperless laboratory has the

Abstracts of papers presented at the 2001 ISLAR

77

potential to result in signi® cant bene® ts including re-source liberation and an electronic compliance platformfor process management and data review.

This paper will describe the process of working with fourvendors to create a complementary data-managementsolution to Purdue Pharma’ s needs. The process oflogging in samples, creating worklists, executing analy-tical methods, capturing data electronically, storingelectronic records, and submitting reports will be re-viewed. After data creation, electronic data review willbe discussed enabling electronic ` instant replay’ of la-boratory information.

The primary bene® ts of this complementary solutionapproach hold the following possibilities.

. Increased capacity by liberating valuable resourcesfrom rework loops and the manual data review pro-cess.

. Leveraging the approach to validation of the overallsolution saving time associated with multiple com-puter validation plans and execution.

. Promise of reduced time to market for new drugs byusing a paperless environment where possible.

This presentation will describe the challenges, successesand experiences surrounding the task of integratingcomplementary information systems in a pharmaceuticaldevelopment environment.

Using laboratory automation to prepare com-pounds precisely for HTS

Haissam Abdelhamid, Purdue Pharma LP, Cranbury, NJ, USA

Pharmaceutical companies are constantly expandingtheir compound inventory and reorganizing their screen-ing plates in an eå ort to ensure new chemical entities canbe identi® ed as rapidly as possible. Whether these newcompounds come from libraries (focussed or diverse) ormedicinal chemistry eå orts, robotics has become anessential tool when preparing compounds for high-throughput screening (HTS). Using robotics to automateroutine laboratory tasks not only increases the rate atwhich these activities can be completed, but also reducesincidents of human error. In our laboratories, we haveintegrated several robotic systems to aid in compounddissolution and HTS plate preparation. These enhance-ments have saved time and increased the reliability of ourscreening inventory, particularly when coupled with ourin-house data-management and inventory tracking soft-ware. Examples of laboratory automation used at Purdueover that past year will be presented.

DNA micro-array fabrication and processing:automation in the laboratory

Katrine Verdun, BIOGEM, Division of Biology, University ofCalifornia San Diego, La Jolla, CA, USACo-authors: Richard Rouse and Gary Hardiman

The con¯ uence of robotics, biotechnology, computersciences and the completion of genome sequencing eå ortsfor several organisms have resulted in revolutionarychanges in how biomedical research is carried out. It is

now possible to fabricate high-density arrays of speci® edDNA sequences that include every known gene of anorganism on a single glass slide. Labelled RNA or DNAtargets (such as mRNAs obtained from cells, tissues ororganisms under diå erent conditions) can be analysed byhybridization on the array. Diå erences in the levels ofexpression for thousands of genes can thus be assessed allat the same time in a single, simple experiment. Geno-mics, informatics and automation are playing increas-ingly important roles as discovery tools in the basicbiological sciences, and as diagnostic and rational ther-apeutic aids in the clinical arena. We discuss the use ofautomation to increase productivity in micro-array fab-rication and describe how automated procedures increasethe quality of results in micro-array experimentation.

Use of the Bio-Tek Precision 2000TM

automatedpipetting system for micro-array sample prepara-tion

Paul Held, Bio-Tek Instruments, Winooski, VT, USACo-author: Gavin Picket

The production of micro-arrays requires the spotting oflarge numbers of unique DNA fragments onto severaldiå erent substrates. While several commercially availableinstruments have automated this spotting task, thesample preparation, culture propagation and mainten-ance of the DNA library are often performed manuallywith multichannel pipettes in 8 £ 12, 96-well, or 16 £ 24,384-well formats. Manual multichannel pipettes, whilemore eæ cient than single-channel pipettes, still representa large amount of pipetting with many opportunities forpipetting errors. Here we describe the use of the Precision2000TM automated pipetting system to carry out many ofthe necessary pipetting tasks required for the preparationof samples for micro-array spotting. These steps include:the propagation of plasmid libraries, PCR reactionpreparation, treatment of PCR products for agarose gelelectrophoresis and the reconstitution of lyophilizedsamples before micro-array spotting. The Precision2000 has a completely con® gurable six-station platformto hold the required pipette tips, reagent troughs andmicroplates (96 and 384 well) for ¯ uid transfer. Theplatform is removable, allowing for multi-user friendli-ness, easy cleaning and set-up of the instrument. Theeight-channel pipette arm moves up and down as well asside to side, while the platform moves front to back toprovide complete access to all locations on the workplatform and complete con® gurability. The pipette armuses a proprietary technology to pick up and seal reliablyany standard tip with individual free-¯ oating barrels thatcompensate for tips out of position. An optional rapiddispense eight-channel manifold, which uses a precisebidirectional syringe pump to dispense accurately andrapidly ¯ uids from a large unpressurized reservoir, is alsoavailable. The Precision 2000 has a built-in micropro-cessor that controls all movements. The ¯ exible software,both onboard and PC-based, provides complete pro-gramming capabilities. For more complete automation,robotics interfaces can be developed using ActiveX1

software commands. The Precision 2000’ s small size, witha 15 £ 21-inch footprint and a height of 16 inches, allows

Abstracts of papers presented at the 2001 ISLAR

78

it to be used almost anywhere including most biologicalsafety cabinets or chemical fume hoods.

Transposition and validation of a manual methodof sample preparation on the TPWII

Heè leé ne Brillard, Novartis Pharma, Orleè ans la Source, FranceCo-authors: Laurent Frances, Florence Dupas, Samir Haddouchiand Olivier Garinot

The Analytical Development Laboratory NOVARTISthat works conjointly with the Pharmaceutical Develop-ment Department had recently acquired the ZymarkTablet Processing Workstation, TPWII.

Within this laboratory, analytical methods are developedto perform manual analysis. Consequently, we developeda procedure to transfer successfully a manual analyticalmethod to this robotic system. First, the whole par-ameters of the TPWII that have an in¯ uence on sample

preparation have been determined and, second, the mostimportant ones have been pointed out in order that theybe validated. This has led to the establishment of atransfer protocol, which is in two parts:

. development of the automatic method (study andoptimization of all parameters);

. validation of the automatic method.

Accelerating knowledge transfer for improvedlead-candidate selection

John P. Helfrich, NuGenesis Technologies Corporation, West-borough, MA, USA

Today’ s modern drug-discovery and development re-search groups are expected to improve the eæ ciency ofdelivering NCEs to the clinic. The new high-throughputprocesses are dramatically increasing the raw number ofdata sets that must be interpreted for decision supportacross a global research team eå ort. The NuGenesisSDMS database serves as the centralized repository foranalytical reports, compound presentation or summarydocuments, project reports, and instrumental raw data.This database can then be extracted to work in conjunc-tion with your LIMS, Enterprise Data ManagementSystem and/or local specialized visualization and statis-tical data software products. If your data sources save orprint, the NuGenesis SDMS platform can get it, auto-matically save it and allow fast eæ cient utilization acrossthe entire enterprise. After all, the data become informa-tion that when eå ectively communicated turns intocritical-path knowledge for decision support in drugdiscovery.

Maximum return (strategic investments for high-throughput screening)

James LaRocque, Wyeth-Ayerst Research, Pearl River, NY,USA

Advanced technologies theoretically enable modern HTSlaboratories to screen expanding libraries quickly againsta drastically longer list of targets while reducing reagentcosts and maintaining more rigorous quality of results.Even in large pharmaceutical corporations, however, thecollision of ® scal reality and the cost of state-of-the-ar tHTS technologies make getting the most out of capitalinvestments a critical skill for success. Wyeth-AyerstResearch has made cost-eå ective investments in PackardCCS PlateTraks and Wallac CCD imagers, while alsocontinually upgrading our Zymark-integrate d systemswith new components and detectors.

This combination of equipment provides maximumthroughput for 384-well formats and supports the transi-tion to 1536-well formats with relatively modular unitsthat can be programmed for multiple tasks. While themanagement of individual screening projects is delegatedto individual scientists, the complexity of the automationinfrastructure oå ers ample opportunity for the develop-ment of specialty skills, creating a well-rounded HTSstaå . By practising both individual initiative and co-operativity, a relatively small group is empowered to

Abstracts of papers presented at the 2001 ISLAR

79

achieve eå ective sample plate replication, 384- and 1536-well HTS, and rate-based hit characterization.

Automated LC/MS analysis of biomolecules usingProMass

Je¡ Whitney, Novatia, Princeton, NJ, USACo-authors: Mark E. Hail and David J. Detlefsen

The recent increase in genomic and proteomic discoveryhas increased the need for high-throughput automatedtools for biomolecule characterization. The ¯ ood of newprotein targets will ultimately demand more eæ cienttools for the evaluation of expressed proteins for drugdiscovery and structural biology studies. ProMass is anautomated biomolecule deconvolution and reportingprogram used to process LC/ESI/MS data or singleESI mass spectra to derive molecular mass information.We recently integrated ProMass with the ThermoFinni-gan Xcalibur data system to create a version of ProMassknown as ProMassXcal iTM. ProMassXcal i processes en-tire Xcalibur sample sequences, deconvolutes mass spec-tra from LC/MS data and produces web-based reports.

ProMassXcali uses a novel cross-platform deconvolutionalgorithm known as ZNovaTM. ZNova incorporates aunique charge-state scoring method that assigns thecharge states of all signals in the ESI mass spectra andtransforms the input ESI mass spectra to produce zero-charge mass spectra (i.e. molecular mass information).ZNova incorporates signal-processing techniques andunique logic that allow application to low charge statespectra and data of low signal-to-noise ratio. As a result,ZNova can be used to process data from a wide variety ofbiomolecules including large proteins, oligonucleotides,peptides, etc. In this poster, an overview of the Pro-MassXcali/ZNova system will be presented along withselected applications which highlight its utility in a high-throughput environment.

Automated metabolite con� rmation and identi� -cation using LC/MS and intelligent chemometrics

Je¡ Whitney, Novatia, Princeton, NJ, USA

In recent years, drug-discovery researchers have placedgreater emphasis on obtaining qualitative measures ofdrug candidate ` quality’ by measuring ADME/Toxtendencies earlier in the discovery process. This has ledto the development of in-vitro assay methodologies formeasuring metabolic stability, cell permeability, solubi-lity, toxicity, etc. The challenge today is to automatefully all aspects of these assays by integrating intelligentdata analysis and interpretation tools into one simple-to-use solution.

This talk will focus on the analytical methods andintelligent data analysis approaches we have employedto assess rapidly metabolic stability and metaboliteidenti® cation. With the use of SmartLCMSTM tech-nology, we will demonstrate the rapid assessment ofmetabolic stability of a compound followed by automaticdetailed re-analysis using on-the-¯ y intelligent data-analysis techniques. In addition to up-front intelligentautomation, we will illustrate the preliminary results of

our MetLabTM software suite for back-end chemometric-based data processing techniques to con® rm automati-cally expected and unknown metabolites.

Development of a high-throughput biochemicalpro� ling platform

Je¡rey Murray, Paradigm Genetics, Inc., Research TrianglePark, NC, USACo-authors: Ioana Popa-Burke and Chris Beecher

Paradigm Genetics, Inc., has industrialized the process ofgene-function discovery for human health, nutrition,crop production and industrial products. The companyhas designed the GeneFunction FactoryTM, an indus-trial-scale laboratory that explores the function of genesin organisms by integrating state-of-the-ar t sequencingtechnology with phenotype, metabolite and gene-expres-sion pro® ling to collect hundreds of data points through asingle technology platform.

The Metabolic Pro® ling group is responsible for monitor-ing changes in the biochemical pro® les of organisms thatoccur over the course of development in response to stressor induced genetic modi® cations. This is accomplished byusing LC-TOF-MS, GC-TOF-MS and ICP-MS instru-ments. We describe here the innovative high-throughpu tprocesses developed for the cataloguing, storing in a dryenvironment, grinding, dispensing and extraction/deri-vatization of samples using custom robots, as well asproprietary instruments such as the ` Mash-A-Matic and` Buster’ .

Role of automation and robotics in high-through-put ADME pro� ling of potential drug candidates

Kelly M. Jenkins, Bristol-Myers Squibb Co., San Diego, CA,USACo-authors: Robyn A. Rourick, Reginald Angeles, MarianneTeopaco-Quintos and Daniel B. Kassel

Early determinations of pharmaceutical properties canserve as predictors of a compound’ s likely developmentalsuccess. Our laboratory has implemented high-through -put ADME assays that address absorption, metabolismand physicochemical properties in an eå ort to minimizediscovery to market attrition. While trying to meet thethroughput demands of parallel synthesis, we establishedan integrated solution for ADME assays which incorpo-rates a SAGIANTM core system for the determination ofboth metabolic stability in human liver microsomes(HLM) and cytochrome P450 (CYP) inhibition. Thisautomated solution has allowed an increase in capacity,throughput and reliability for both ADME assays.

The HLM assay uses a MultimekTM 96-channel pipettorfor liquid handling. The analysis plates are transferredo¥ine for ® nal analysis using high-throughput parallelLC/MS. The CYP inhibition method uses a combinationof liquid handlers and a ¯ uorescence plate reader toperform a single concentration pro® le assay for 88compounds. CYP inhibition is measured for bothCYP3A4 and 2D6 isozymes.

This system represents a fully integrated approach insupport of high-throughput ADME evaluation in a drug-

Abstracts of papers presented at the 2001 ISLAR

80

discovery environment. The core system concept createsa plug-and-play approach that combines a series ofmodular stations to build a robotic system, which is¯ exible, upgradable and easily recon® gurable whenassays change or are newly developed. The applicationof these strategies as a means of assessing metabolicstability and CYP inhibition of our combinatorial li-braries is discussed.

Novel homogeneous FRET-based assay for endo-peptidase inhibitors: assay development, steady-state kinetics, high-throughput screening andminiaturization

Joe Bradley, P¢zer Global R&D, Sandwich, UKCo-authors: Chris Chambers, Helen Boyd, Emma Faure, JoeBradley, Simon Dales, Neil Benson, Wilma Keighley andAndreas Sewing

The growth in compound numbers and subsequent needfor increases in throughput and reduction in cost are keydrivers of developments in the ® eld of high-throughputscreening. Increasingly, there is a move towards ¯ uores-cence-based assay technologies, which are ideally suitedfor screening because they are versatile, homogeneousand amenable to both automation and miniaturization.

We developed a novel FRET-based assay to identifyinhibitors of a metallo-endopepidase . This assay uses anovel substrate with a low Km, thus making it cost-eå ective for high-throughpu t screening. We have success-fully transferred and validated this assay onto ourRobolab linear track-screening platform. A full HTScampaign was subsequently conducted in 384-well for-mat, achieving a maximum throughput of 75 000 datapoints within 24 h.

This poster will illustrate the HTS process from assaydevelopment to high-throughput screening using state-of-the-art technology.

Accelerating knowledge transfer for improvedlead-candidate selection

John Helfrich, NuGenesis Technologies Corporation, Westbor-ough, MA, USA

Managing and tracking data through the discovery pro-cess requires a compilation of many diå erent types ofanalytical, biological and image output. This includes thecollection, storage and management of relevant scienti® cinformation about lead candidates, as well as immediateaccess to this information for complete compound docu-ment creation. Data collaboration across the discoveryarena, even across the entire enterprise, is critical tomaking crucial go/no go decisions about subsequent leadcandidate development.

The NuGenesis1 Scienti® c Data Management System(SDMS) allows discovery scientists electronically to view,share, reuse and access data within the laboratory, as wellas throughout the enterprise. You can easily capturelaboratory and report data produced in any Windows-based software application. These data can be cataloguedautomatically and accurately for easy retrieval. Immedi-ate access to this electronic information is possible from

anywhere around the world using Web technology. Thispresentation will focus on the principles of good datamanagement that allow for pharmaceutical and biotechdiscovery facilities to leverage data as an asset, protectvaluable intellectual property and ensure the accurateand easy creation of complete compound documentsthroughout the drug-discovery and development process.

Elutri-Zone MS: a new SPE system for rapidsample processing and LC/MS analysis

John Janiszewski, P¢zer Global R&D, Groton, CT, USACo-authors: R. M. Olech, R. A. Pranis, J. R. Jacobson, C. A.Perman, B. A. Boman, J. Soldo, R. Speziale, T. W. Astle,M. J. Cole, J. S. Janiszewski and K. W. Whalen

To increase sample throughput, many recent LC/MSmethods have described techniques to shorten and/orsimplify the chromatography step(s) before MS analysis.We describe a discrete solid-phase extraction step usingEmpore SPE membrane to capture the analyte, followedby an elution step that features the rapid quantitativetransfer of the elutri zone (pure zone) into the MS. Thisapplication was designed primarily for bioanalyticalsupport of high-throughpu t ADME screening.

An SPE Card having the outer dimensions of a microtitreplate was developed by molding a plastic frame around asandwich of 3M Empore sorbent (0.5 mm thickness,8 mm particles) and micro® bre support material.Ninety-six discrete elution zones (7 mm diameter, 9 mmcentres) were welded into the sheet. The SPE Card wasdesigned to ® t in a modi® ed cell harvester (Tomtec) thatfacilitates SPE processing. The harvester performs sor-bent activation, load and wash steps. Once the analytehas been retained, an interference wash is completed andthe plate transferred to an elution device (Elutrix,Tomtec). Each well was eluted directly into the MSusing ¯ ow rates ranging from 1.0 to 3.0 ml min¡1. Asingle HPLC pump delivered eluent composed of 5 mmAmmonium Acetate combined with 50± 90% methanolor acetonitrile.

The SPE Card and Harvester protocol allow rapid (2±5 min) o¥ine sample clean-up followed by direct elutioninto the MS detector. This protocol results in theconcentration of the sample and reduces liquid-handlingsteps as dry plates are transferred directly to the Elutrixfor LC/MS processing. Owing to the small particle size ofthe Empore sorbent, the SPE Card can mimic theperformance of discrete HPLC columns, similar to thoseused in high-throughput sample analysis or in onlineclean-up protocols. Relevant parameters and preliminaryresults will be described here.

Combined gravimetric and colorimetric methodfor the calibration and performance testing ofliquid-handling workstations in a GLP laboratory

John R. Alianti, GlaxoSmithKline, Research Triangle Park,NC, USACo-author: Glenn Smith

Good laboratory practices (GLPs) require thatappropriate calibration and performance testing be con-

Abstracts of papers presented at the 2001 ISLAR

81

ducted on all pipetting devices, including automatedliquid-handling systems. Routine evaluation of theequipment must be performed to assess both the accuracyand reproducibility (precision) of liquid transfers.Additionally, the testing methods and acceptancecriteria must adhere to established company-widepolicies (SOPs). These methods must be sensitive aswell as being robust and practical. Likewise, the datagenerated must be easily analysed, interpreted anddocumented.

Pipetting accuracy is determined by weighing 96-wellplates (in replicates of three) before and after dispensinga speci® ed volume of liquid. System accuracy at variousvolumes is determined and the results used to calibratethe workstation as appropriate (e.g. Zymark RapidPlateSyringe Calibration Factors, Tecan GENESIS LiquidClasses, Packard MultiPROBE II Performance Files).The between-tip precision, individual-tip precision andindividual-tip accuracies are all ascertained by sub-sequent colorimetric testing using a photometric micro-plate reader. A preformatted Microsoft Excel spreadsheetcontaining embedded formulas automatically analysesthe data and generates a report.

The procedure developed combines the simplicity ofgravimetric measurement with the speed of colorimetricmicroplate reading. The process is generic and can beused to assess or compare the accuracy and precision ofany workstation. Automated data analysis and standar-dized report generation facilitate the process and ensureregulatory compliance.

Implementing automated sample preparation forGLP and non-GLP bioanalysis to increase produc-tivity in a contract research organization

John R. Kagel, Charles River Discovery and DevelopmentServices, Worcester, MA, USACo-authors: Larry E. Elvebak, Brian E. Lilley, Jakal M. Aminand James A. Jersey

Client services for analysis and bioanalysis must: providereliable results, meet increasingly aggressive timelines,meet increasingly lower unit costs and be performedunder appropriate regulatory conditions. Automationwas used successfully to satisfy these criteria regardingsample preparation for high-throughput LC/MS/MSGLP and non-GLP bioanalysis in a CRO. The ® rstphase in automating sample preparation involved pro-cessing samples in a 96-well format using 96-tip parallelpipetting workstations (e.g. Tomtec) in an open-accessenvironment. The impact of this implementation was toincrease productivity by at least twofold. The secondphase in automating sample preparation used a liquidhandler (e.g. Tecan Genesis) for the preparation ofcalibration standards and automated reformatting anddilution of samples from tubes or vials into 96-well plates.In addition, calibration and compliance issues related toautomated sample preparation will be discussed.

Octave system: a new automated instrument forthe analysis of molecular interactions

Robert J. Kaiser, Prolinx, Inc., Bothell, WA, USACo-authors: Karin A. Hughes, Guisheng Li, Deborah D. Lucas,Kevin P. Lund, Chris Pershing, Douglas A. Spicer, Mark L.Stolowitz, Jean P. Wiley, Steve Bailey, Michael Baum, ErikEngstrom, Scott MacInnes, Rich Ward, Craig Yamamoto,George de la Torre, Lynn Hilt, Charles J. Hutchings, ArturoKozel, Terry A. Weeks, Dwight Bartholomew, Rick Carr, KerenDeng, Jerry Elkind, Tom Martin and John Quinn

Surface plasmon resonance (SPR) technology is a power-ful, label-free method for the analysis of molecular inter-actions. Prolinx1, Inc., has developed a new automatedinstrument, the Octave Molecular Interaction AnalysisSystem, that allows this powerful analytical platform tobecome ubiquitous in life-science research and drug-discovery laboratories. This technological breakthroughis the result of combining the miniaturized Texas Instru-ments SpreetaTM 2000 SPR sensor with the ProlinxVersalinxTM Chemical Aæ nity Tools. These technologieshave made possible the development of an instrument ofmoderate cost that incorporates eight independent sen-sors operating in parallel. The sensor surfaces can bereadily and eæ ciently modi® ed with molecular targets,and exhibit low non-speci ® c binding. Samples are intro-duced to the sensors from standard microwell plates usingan integrated liquid-handling robot. This new instru-ment will signi® cantly increase the throughput of SPR-based molecular interaction analysis in basic biologicalscience and drug-discovery applications.

Increased throughput e³ ciency with dual-columntechnology in LC/MS/MS sample analysis

Kathryn B. O’Mara, GlaxoSmithKline, Research TrianglePark, NC, USACo-authors: Lisa St John-Williams and John A. Dunn

Recent advances in automation have signi® cantly re-duced the amount of time it takes to prepare clinicalsamples for LC/MS/MS analysis. As a result, the eæ -ciency bottleneck has shifted from sample preparationtime to analytical run time on the mass spectrometer. Toreduce the mass spectrometry analysis time, the use ofdual-column technology was evaluated for three single-column validated LC/MS/MS methods.

For each of the three methods, the HPLC system wasenhanced with switching valves and the appropriateprogramming to allow the unit to function in the dual-column mode. A semi-automated liquid handler was usedto perform either protein precipitation or solid-phaseextraction using 96-well technology. A triple quadrupolemass spectrometers was used to acquire data using mul-tiple reaction monitoring.

The three quantitative LC/MS/MS methods using dual-column technology are currently being used in a high-throughput laboratory. Each method has been validatedto demonstrate acceptable accuracy and precision. Ro-bustness has also been demonstrated by the analysis ofthousands of samples in a GLP environment. For eachmethod, the analytical run time has been reduced by

Abstracts of papers presented at the 2001 ISLAR

82

50% and instrument eæ ciency and sample throughputhave been increased twofold.

Rapid, e³ cient method for determining metabolicstability

Kelly Johnson, Waters Corporation, Milford, MA, USACo-authors: John Erve, Andre Dandeneau and Beverly Kenney

Modern drug discovery has been transformed by theautomation of research. The resulting explosion of datain the discovery pipeline, combined with the pressure toreduce costs and speed up drug-discovery cycles, providesa strong demand for fast and selective analytical methodsto produce quality data.

In vitro metabolic stability assays provide a rapid estima-tion of new chemical entities in the discovery phase ofpharmaceutical development. These multistep assaysroutinely incorporate robotic liquid-handling systems toautomate the incubation step of test compounds withhuman liver microsomes in 96-well plates. Metabolicstability is subsequently measured by LC/MS as theamount of substrate metabolized (expressed as the per-centage remaining of the initial substrate). The resultinginformation is crucial in the selection process in determin-ing a compound’ s potential ` drugability ’ .

Given the large number of samples generated by theliquid-handling systems in metabolic stability assays,employing automation and higher throughput in everystep of the assay has become a necessity. This presenta-tion will demonstrate the use of high-throughput LC/MSmethods to analyse samples produced in a metabolicstability assay. We will develop LC/MS methods thatsigni® cantly increase overall throughput without sacri ® -cing data quality. By incorporating such factors asparallel sample processing, high sample capacity, alter-nating column regeneration and smaller diameter col-umns to reduce cycle times, we can analyse the largenumber of samples produced and ultimately help ex-pedite the drug-discovery process.

Filter-immobilize d arti� cial membrane (� lter-IAM) permeability assay: a new high-throughputin vitro drug-absorption model

Konstantin Tsinman, pION, Inc., Woburn, MA, USACo-authors: A. Avdeef, D. Voloboy, M. Stra¡ord and B. Kenney

The assessment of passive transport properties of over 20drug and natural product molecules was made using thein vitro absorption model based on ® lter-immobilizedarti® cial membranes (® lter-IAM), assembled from phos-phatidylcholine in dodecane, in buå er solutions atpH 7.4.

Several of the compounds were lactones extracted fromthe roots of the kava-kava plant. Experiments weredesigned to test the eå ects of stirring during assays andthe eå ects of varying the assay times.

The highly mobile kava lactones permeated in the orderdihydromethisticin > yangonin > kavain > methisticin >desmethoxyyangonin . Other molecules in the studyranked: phenazopyridine > testosterone > propranolol >

ketoconazole > piroxicam > caå eine > metoprolol (pro-posed BCS internal standard) > terbutaline.

Stirring during assay signi® cantly increased the observedpermeabilities for highly mobile molecules.

In addition to permeability measurements, membraneretention of compounds was determined. Yangonin,desmethoxyyangonin , ketoconazole and phenazopyridinewere > 60% retained by the arti® cial membranes con-taining phospholipids.

The in¯ uence of hydrogen bonding was explored bydetermining permeabilities using ® lters coated with do-decane free of phospholipids. The membrane transport ofphenazopyridine (strong hydrogen-bond donor) is abouttwice as fast and retention is about twice reduced in theinert lipid membranes compared with phospholipid-based membranes.

In the ® lter-IAM method, concentrations were deter-mined by microtitre plate UV (190± 500 nm) spectro-photometry and by LC/MS. Higher-throughput wasachieved with direct UV by the use of 96-well microtitreplate formats and with LC/MS by the use of cassettedosing (5-in-1).

Seamless integration of information in a pharma-ceutical development environment: integratingtechnology from LabWare, NuGenesis, VelQuestand Waters

Kurt Roinestad, Purdue Pharma LP, Ardsley, NY, USA

This presentation will describe Purdue Pharma’ s ap-proach to an integrated solution between LaboratoryInformation Management Systems (LIMS), Chromato-graphy Data Systems (CDS), Process Management andCompliance Systems (PMC), and Scienti® c Data Man-agement Systems (SDMS). It will describe the vendor-selection process and how Purdue Pharma created anenvironment to produce a best-of-breed information sol-ution to creating the paperless laboratory. Workingtowards the creation of a paperless laboratory has thepotential to result in signi® cant bene® ts including re-source liberation and an electronic compliance platformfor process management and data review.

This paper will describe the process of working with fourvendors to create a complementary data managementsolution to Purdue Pharma’ s needs. The process oflogging in samples, creating work-lists, executing analy-tical methods, capturing data electronically, storingelectronic records and submitting reports will be re-viewed. After data creation, electronic data review willbe discussed enabling electronic ` instant replay’ of la-boratory information.

The primary bene® ts of this complementary solutionapproach hold the possibilities to do the following.

. Increased capacity by liberating valuable resourcesfrom rework loops and the manual data reviewprocess.

. Leveraging the approach to validation of the overallsolution saving time associated with multiplecomputer validation plans and execution.

Abstracts of papers presented at the 2001 ISLAR

83

. Promise of reduced time to market for new drugs byusing a paperless environment where possible.

This presentation will describe the challenges, successesand experiences surrounding the task of integratingcomplementary information systems in a pharmaceuticaldevelopment environment.

High-throughput robotic workstation for per-forming gene-expressio n assays

Michael A. Kuziora, Gene Logic, Inc., Gaithersburg, MD, USA

Metabolism and growth are dependent on a highlyorchestrated interplay of a variety of proteins withincells. These proteins function as enzymes in metabolicpathways, signalling molecules for communication be-tween and within cells, and as various cellular structuralcomponents. In a disease state, cells often modulate theamounts of speci® c proteins and may produce newproteins not normally found in a particular type of cell.

The alterations in the proteome observed in the diseasedstate most often result from changes in gene expressionlevels within a cell. Gene Logic believes that pharmaceu-tical companies can reduce the time, risk and cost associ-ated with drug discovery if they know the expressionlevels of genes that play roles in the disease-associatedpathways. Such knowledge may help them discover drugtargets, screen drug leads, and predict toxic and phar-macological responses to drug leads.

Using the Aå ymetrix, Inc. GeneChip1 micro-array plat-form, Gene Logic has measured the expression levels ofthousands of genes from a diverse range of normal anddiseased tissues to create a reference gene expressiondatabase called the Gene Express1 Suite. The bioinfor-matic analysis tools incorporated in the Gene Express Suitefacilitate the identi® cation of distinct sets of genes whoseexpression is consistently altered in a particular diseasestate. The expression patterns of these gene sets become amolecular ® ngerprint of the disease and thus not onlyre¯ ect the disease status of metabolic pathways, but alsocould also serve to indicate the eå ect of potential treat-ments when aå ected cells are exposed to a potential drug.

A primary goal of a pharmaceutical company is todiscover drugs that restore the normal functioning ofthe disease-aå ected pathways. In one scenario, a phar-maceutical company could use the information in theGene Express database to identify a small set of genesthat indicate a disease state. The expression of this geneset is then monitored when an appropriate cultured cellline that represents a disease state of interest is exposed tochemicals from a compound library. A potential drugcan be identi® ed for further characterization if, for ex-ample, it is found to restore expression levels of the geneset to normal levels.

Unfortunately, the use of large micro-arrays to measuregene expression of a few genes in a high-throughputscreen is not economically feasible at this time. Wetherefore investigated the use of alternative method-ologies to measure gene expression. Bayer Corporation’ sQuantigeneTM assay, which uses branched DNA (b-DNA) for signal ampli® cation, and provides a sensitive

and reproducible method for measuring mRNA levels ofa small number of genes. The assay’ s 96-well plate formatand simple handling procedures makes it highlyamenable to high-throughput robotics. This talk willdescribe a Zymark StaccatoTM Workstation designed toperform the hybridization, ampli® cation and signal-detection steps of the Quantigene assay.

Multivariate calibration of a UV-VIS � breopticprobe used for direct measurements in a ZymarkXP-based automated dissolution system

Lars A. Svensson, AstraZeneca, Molnda, SwedenCo-authors: Rimstedt Eva and Svensk-Ankarberg Anna

Correlation between drug dissolution results obtained byboth liquid chromatography (LC) and ultraviolet-visua l(UV-VIS) spectrophotometry has been performed bymultivariate calibration.

The dissolution of tablets or capsules, stored and exposedto harsh conditions (high humidity/elevated tempera-tures) at diå erent degrees, was carried out by an auto-mated system based on the Zymark XP robot. Thesystem had the capability of withdrawing samples to besubsequently prepared into LC vials at the same instanceas an UV-VIS spectrum was recorded via a ® breopticprobe. The same probe was used in all vessels and couldbe attached directly adjacent to the ® lter tip for thesample withdrawal on one of the robotic hands.

The subsequent multivariate calibration was performedwith Simca-P 8.0 (Umetrics AB, Sweden). Most of thecalibration is explained in two components, with the ® rstmainly corresponding to the UV spectrum and thesecond taking into consideration the baseline oå set.

High-throughput chemistry: an integration ofchemistry, automation and informatics

Li Chen, Ho¡mann-La Roche, Inc., Nutley, NJ, USA

One of many impacts that combinatorial chemistry hasbeen made in the last decade is to stimulate intellectualcreativity to invent new and more eæ cient ways ofmaking new compounds. Although much eå ort has beeninvestigated into developing synthetic methods, auto-mated synthesizers and analytical tools, the high produc-tivity of compound library synthesis cannot be achievedwithout the integration of chemistry, automation andinformatics into an eæ cient process. I will emphasizewhat we have learned in how to establish an infrastruc-ture that integrates functional modules containing adiverse set of tools for high-throughput organic synthesisapplications. This modular approach provides the maxi-mum eæ ciency of combinatorial tool functions to supportmultiple project teams for drug lead generation, explora-tion and optimization.

The agenda includes the following.

. HTOS process design.

. Pro- and post-synthesi s automation.

. Synthetic chemistry method.

. HT analysis and puri® cation.

. Data management and chemo-informatics.

Abstracts of papers presented at the 2001 ISLAR

84

Living ChipTM

: a nano� uidic platform for ultra-high-throughput, massively parallel synthesis,storage and screening (MPS3)

John R. Linton, BioTrove, Inc., Cambridge MA, USA

The Living ChipTM is a nanotitre plate consisting of auniform and addressable through-hole array. Thethough-hole s are nominally 300 mm square and 500 mmdeep, giving each well a 50-nl volume. Proprietary coat-ings make the surfaces of the plates hydrophobic and theinterior of the wells hydrophilic. This allows the samplesto be held in the wells by surface forces and preventssample contamination from adjacent wells. Ten thou-sand-well chips are currently in use and 100 000-wellchips are in production.

Massively parallel mixing of assay components takesplace when chips containing diå erent components arestacked such that the through-holes align. The chips areimaged using a CCD array for readouts such as absor-bance, ¯ uorescence and luminescence. Detection is per-formed in a transmission geometry, taking advantage ofthe bottomless wells. These key features and the sophis-ticated robotics built by BioTrove allow for the rapid

acquisition of biochemical information, on the order of107 measurements per day. Combining nanolitre reactionvolumes and a simple interface to microtitre plates, theLiving ChipTM format conserves compound libraries,increases analytical capabilities and decreases costs. Theplatform has applications in all aspects of drug discoveryfrom materials’ handling and storage, target ID andvalidation, to lead discovery and optimization.

Cell culture in the Living ChipTM using yeast (S.cerevisiae) and bacteria (E. coli) shows similar growthcharacteristics to cells grown in bulk. Additionally, E.coli cells expressing GFP inoculated into isolated channelsshow no crosstalk.

Diverse biological libraries can be introduced to the chipusing a simple dip loading procedure. This technique wasused by BioTrove collaborator Genofocus to dip loadcells expressing a lipase enzyme library into a LivingChipTM. The cells were grown up overnight and thenstacked with a chip containing a tagged substrate so thatwells containing clones with increased lipase activityyielded a greater ¯ uorescence signal. The ` hits’ wereharvested by a puå of air from a microsolenoid valvepositioned above the well, into a microtitre plate waiting

125-300um

>500um

hydrophilic hydrophobic

Figure 1. The Living ChipTM: a nanotiter plate. The surfaces are hydrophobically coated and the wells are hydrophilic to contain andisolate samples.

5 cm square

Figure 2. 10 000 well chips (above) are currently in use and 100 000 well chips are in production.

Abstracts of papers presented at the 2001 ISLAR

85

below. Kinetic analysis con® rmed the discovery of alipase mutant with a 20-fold higher activity relative tothe wild-type. This screening consisted of 30 000 assaysperformed by one person in less than 2 days, and wouldhave taken several scientists weeks using conventionalmeans.

Next step in miniaturization : submicrolitre assaysin 96-well formats

M. J. Wildey, R. W. Johnson Pharmaceutical ResearchInstitute, Raritan, NJ, USACo-author: C. Fleming

Some of the challenges and goals in many of today’ sscreening laboratories are to investigate, evaluate andvalidate new techniques that will enable the reduction incost, provide an increase in eæ ciency and strengthen thequality of screening data. Towards the end of reducingcost, we have been a Beta test site for a novel 96-wellsubmicrolitre assay system called Arteas. The architec-ture of Arteas oå ers a solution to the problem ofevaporation in small-volume assays.

We have successfully performed a ¯ uorescent-based en-zyme assay in Arteas with a reaction volume of 400 nl.We successfully reproduced published control IC50s androutinely generate Z’ -factors in the range 0.6± 0.7. Wewill show these data and variability around some testsperformed with actual library compounds. Translation ofthis device to production screening will be dependentupon the successful integration of nanoliquid handlingonto our robotic platforms. Initially, we will integratethis new liquid-handling option onto our Fast Trackrobotic platform. Progress on this translation will beaddressed.

Automated compound dilution and presentationfor determination of IC50 data

Malcolm Willson, Systems Research, GlaxoSmithKline, Steve-nage, UKCo-authors: David Brown and David Hayes

Programme-targeted SAR screening within Systems Re-search at Stevenage calls for a process able to progresscompound activity determination (80 to > 320 com-pounds/week) in real time, enabling ongoing chemistryto continue based on known biological activity.

Eæ cient compound dilution and generation of assayplates is required to meet weekly turn around times forIC50 and cross-screening data generation.

Comparison of manual across-plate dilution, verticaldilution down a series of plates and Z-dilutions within aseries of 384-well plates will be discussed.

Use of the twin head Biomek FX uses the ¯ exibility ofboth the 96- and 384-tip heads, allowing Z-dilutions, avery eæ cient method for automated generation of IC50s.By having space on the stock plates for several extracompounds, a selected range of ` standards ’ coveringseveral assays can be included to enable assay perform-ance and automation QC to be monitored for each set ofcompounds, along with suæ cient control/blanks forstatistical analysis of assay performance.

Cloning of assay templates reduces IT resource require-ments for data analysis as templates can be cloned ratherthan written from scratch.

If compound numbers increase, the system can beconverted to run vertical dilutions down a series of plates(diå erent dilution per plate). However, this only works ifthe assays are stable enough for the increased length ofassay due to the higher plate numbers.

Conversion to 1536 plates doing Z-dilutions from 384-well stock plates is also an option that would be possiblewhen liquid-handling devices can cope with smallervolumes for assays.

Lead generation and optimization: integrated datamining and informatics in drug discovery

Charles J. Manly, Discovery Technologies, Neurogen Corpora-tion, Branford, CT, USA

Drug discovery today includes considerable focus oflaboratory automation and other resources on high-throughput technologies, but lead generation and opti-mization to clinical candidates continues to be a lengthyand costly process. The real bene® t of today’ s tech-nologies is beyond the exploitation of each individually.Only recently have signi® cant eå orts focused on eå ec-tively integrating these complex discovery disciplines torealize their larger potential. Informatics, computationalchemistry, virtual screening and data mining play a largerole in this integration and in increasing the eæ ciency ofthe drug-discovery process.

Figure 3. Isolated wells: all wells loaded with media. Cellexpressing GFP innoculated into wells show no crosstalk afterculturing.

Abstracts of papers presented at the 2001 ISLAR

86

Survival skills for managing robotics and automa-tion: how to outwit, outplay, outlast

Maria DeGuzman, A¡ymetrix, Inc., Santa Clara, CA, USA

Our experience in automating a genomics’ laboratory forsingle nucleotide polymorphism (SNP) detection hasshown us that automation projects are not often com-pleted in the planned or expected manner. With auto-mation playing an ever-increasing role in the researchand development laboratory, we will discuss why someprojects succeed while others fail. For instance, what arethe real bottlenecks to address? Does experience make adiå erence? Is management expectation realistic? Whatare the advantages and disadvantages of custom orcommercial automation? What are the roles of thescientists, engineers, and laboratory managers in suchprojects?

The recent completion of our high-throughput screeningproject has enabled us to look back objectively andanswer these questions for our genomics’ application.

The high-throughput screening project at Aå ymetrixbegan in 1999. The initial goals of the project were toscreen for SNPs across 40 unrelated individuals in anautomated fashion. In the initial stages of this project, theprocess), which encompassed sample preparation toscanning, was done manually. This provided an outputof about 2.3 Mb of sequence a week for every 4.5 people.With the collaborative eå ort of the executive manage-ment group, the high-throughput screening laboratory,the engineering group and the bioinformatics group, weset out to accomplish this task.

Unfortunately, as the project continued, con¯ ictsand diæ culties arose. Each group had its own expecta-tions for the project that were not necessarily commu-nicated with the others eå ectively. This led to manyproblems including delays, miscommunications, unrealis-tic expectations, frustration and blame being placed onone another. In addition, another facet compounded thealready diæ cult process: reality. We found several in-stances where reality interfered in the development of theproject including budget, time, resources and processbottlenecks. We also saw evidence of con¯ icts such aswhen we needed to decide between 96- or 384-well plates,tubes or plates, custom or commercial or semi-automa-tion, or even con icts between the scientists and theengineers.

Fortunately, we realized many of these issues andmanaged them eå ectively. We also proceeded to perfectthe process in other ways using protocol enhancementsand microscaling improvements. This increased ourthroughput to 1.4 Mb per day for every two people whilereducing reagent costs.

At the conclusion of this project, approximately 25 000genes were screened covering 8.2 Mb of the genome(including genes, regulatory regions and STSs). Sincemost of the projects were tested against 40 individuals, wescreened 24 million bases of dsDNA and identi® ed 15 682SNPs that have been deposited to the SNP publicdatabase. In addition, much of the software, automationand protocols developed during this project have beenincorporated into other internal research laboratories.

Abstracts of papers presented at the 2001 ISLAR

87

Stand alone extractor for semi-automated methoddevelopment and validation

Maria Styslo-Zalasik, R. W. Johnson Pharmaceutical ResearchInstitute, Raritan, NJ, USACo-author: Kathleen Cirillo-Penn

In the past 10 years, laboratory automation has becomeincreasingly important for routine chemical analysistesting in the pharmaceutical industry. The use ofautomatization or semi-automated laboratory equipmentcan reduce the laboratory ¯ ow-through time of thesamples in the QC area. Likewise, automated and semi-automated equipment can be used as a tool for researchduring method development and validation.

This presentation will illustrate how semi-automatedsample preparation can aå ord rapid optimization ofmethod parameters during early-stage method develop-ment and validation. The potential dissolution rate andsolubility issues with high sample concentrations can beavoided through the evaluation of data from parameteroptimization of the sample preparation using semi-auto-mated instrumentation. Examples for the use of a standalone extractor in an assay/purity method validation willbe presented.

New automated instrument for the characteriza -tion of biomolecular interactions

Mark L. Stolowitz, Prolinx, Inc., Bothell, WA, USA

Surface plasmon-resonanc e technology is a powerful,label-free method for the analysis of biomolecular inter-actions. However, the use of this technology is limited bythe cost, throughput and complexity of existing instru-mentation and chemistries. Prolinx has developed a newinstrument that will address these limitations and allowthis powerful analytical platform to become ubiquitous inlife science research and drug-discovery laboratories.

This technological breakthrough is the result of combin-ing the Texas Instruments Spreeta1 2000 chip withProlinx’ s VersalinxTM Chemical Aæ nity Tools. Thesetechnologies enabled the development of an instrumentof moderate cost that incorporates eight parallel sensorsurfaces that can be eæ ciently modi® ed with moleculartargets and exhibit low non-speci ® c binding. The pre-sentation will encompass technology, design and applica-tions of this new instrument.

Automation of an infectivity assay for the quanti-tative analysis of virus in biopharmaceutica lproducts

Mervyn Cadette, GlaxoSmithKline, Beckenham, UK

Relative to chemical assays in a pharmaceutical devel-opment environment, biological assays are relatively ` lowthroughput ’ in comparison. However, the complexity ofassays associated with biological products is often of adiå erent magnitude. Biological assays become attractiveautomation candidates based upon their complexitycoupled with high demand for their use in biopharma-ceutical analysis.

The Tissue Culture Infectious Dose 50% (TCID50)assay was used for the quantitation of viral titres usedin vaccine products.

These assays required sterility, therefore requiring allequipment to be housed in Class II microbiological safetycabinets. The TCID50 assay was programmed on aTecan RSP 200 using eight disposable-tip liquid-hand-ling probes in conjunction with a robotic manipulatorarm. Culturing of plates containing tissue culture, tosupport the TCID50 assay, was automated using aZymark Twister coupled to a Multidrop dispenser.

Scheduling and liquid-handling software were combinedto execute this automated assay. The use of the ` joblist’function allowed the operator to select from a database ofprede® ned dilutions within the Logic pipetting software.

A formal comparison of the automated TCID50 assayagainst the manual TCID50 was performed and provedcomparability.

Integrated approach to high-throughput sampleprocessing, characterizatio n and puri� cation

Michael L. Moore, GlaxoSmithKline, King of Prussia, PA,USA

The extension of high-throughput synthesis in support oflead optimization has imposed increasingly stringentrequirements on compound characterization, purityand accurate concentration determination. We devel-oped an integrated and highly eæ cient process forcompound analysis, puri® cation and sample processingwith a capacity of 100 000 compounds/year at a 20-mgscale. Puri® cation is driven by ultrahigh-throughpu tanalytical LC/MS, which minimizes the number offractions collected and analysed. Custom software witha browser-based front end is employed to track samplesand provide the required data to robotic workstations.

Managing laboratory automation in the postge-nomic era

Michael R. Kozlowski, Axiom Biosciences, San Diego, CA,USA

The completion of the cloning of the human genome hasprovided the drug-discovery community with a wealth ofpotential drug targets. At the same time, it has producedchallenges to the way in which drug discovery is done.Formerly, the mandate of drug discovery was to screen arelatively small number of well-validated targets againstan immense number of compounds. Now, drug-discoveryscientists are faced with processing a very large number ofminimally validated targets. Most targets were formerlysingle proteins, which are amenable to screening inhighly reductionist systems. Now it is clear that mostbiological processes, including those contributing topathological states, must be the result of complex inter-actions between proteins. A way must be found to addressthis new level of complexity. In addition to thesechallenges, the availability of the entire sequence of thehuman genome raises expectations for the rapid intro-duction of more, and better, drugs.

Abstracts of papers presented at the 2001 ISLAR

88

These challenges demand new ways of thinking abouthow we carry out drug discovery, and about drug-discovery automation, from assay development to AD-MET pro® ling. This talk will discuss some of these newapproaches.

High-throughput screening inhibition assays toevaluate the interaction of P� zer proprietary com-pounds with cytochromes P450

Michael West, P¢zer Global R&D, Groton, CT, USACo-authors: Larry Cohen, Al¢n Vaz and Shawn Harriman

A ¯ uorescence-based drug interaction assay using recom-binant CYPs and a cocktail of CYP-speci® c probes inhuman liver microsomes was assessed as higher through-put methods for evaluating the potential for inhibition ofCYP1A2-, CYP2C9-, CYP2C19-, CYP2D6- andCYP3A4-mediated metabolism. Comparisons of IC50obtained with the ¯ uorogenic and conventional drugprobes in recombinant CYPs were similar for CYP1A2,2C9, 2C19 and 2D6, but not for CYP3A4.

Additionally, using a single-point estimated IC50 ap-proach, compounds that were shown to be inhibitorsusing conventional drug probes with human liver micro-somes were also classi® ed as inhibitors in recombinantCYPs using the ¯ uorescent probes. For the cocktailapproach, it was shown that the CYP-speci® c reactionswere not altered in the presence of multiple probes asindicated by no distinguishable eå ect on Km or Vmax. Asexpected from this result, the IC50s generated in thecocktail incubations were in good agreement to thoseobtained from individual incubations.

GigaMatrixTM

ultrahigh-throughpu t screeningplatform

Mike La¡erty, Diversa, San Diego, CA, USA

The myriad of microbes inhabiting this planet representa tremendous repository of biomolecules for pharmaceu-tical, agricultural, industrial and chemical applications.Diversa Corporation has the unique capability ofaccessing this microbial diversity by taking a culture-independent, recombinant approach to the discoveryof novel proteins and small molecules. Diversa ’ s discoveryprogramme uses genes and gene pathways captured fromnucleic acids extracted directly from the environment,which are then constructed into complex, 109-memberenvironmental libraries. These libraries often containup to 5000 diå erent microbial genomes and, thus,require high-throughput screening methods to cover theirdiversity eå ectively.

Diversa has developed GigaMatrixTM, a new ultrahigh-throughput screening platform. GigaMatrix plates have> 100 000 bottomless wells in the same footprint as amicrotitre plate. The platform is automated and capableof screening 1 billion clones per day. Less equipment timeand manpower are required and assay costs are drama-tically reduced as compared with traditional microtitreplate-based screening. The power of the GigaMatrixplatform to discover novel enzymes, small molecules,

protein therapeutics and other bioactive molecules willbe presented.

Building an electronic data-handling, compliant,foundation for the QA analytical laboratories atWestborough

Mike Stroz, AstraZeneca, Westboro, MA, USA

A signi® cant portion of QA laboratory activities arededicated to compliance and data handling in a paper-based system. AZ Westborough is implementing anelectronic environment in the laboratories to ensure 21CFR Part 11 compliance, improve eæ ciency and reducecosts. The presentation will discuss the applicationsselected, and the validation and system architecturebeing installed to achieve these goals.

Online SPE by column switching: another form ofbioanalytical laboratory automation

Min Chang, Abbott Labs, Abbott Park, IL, USACo-authors: Huong Mai, Anita Shen, Brendan Swaine, Qin Jiand Tawakol El-Shourbagy

Time event-controlled column switching valves havebeen available to analytical scientists since the mid-1970s. Earlier uses of the column-switching valve in-cluded removing late elution peaks from the HPLCrun, autosampler sharing, two-dimensional HPLC se-paration (heart cut), inline ® lter/guard column regenera-tion, online concentration, unattended column selectionand fractions’ collection. Although there is at least onecolumn-switching valve in an HPLC system (the injectorvalve), the idea of the addition of another valve has notbeen widely accepted. Analytical scientists have turnedaway from the technology possibly due to the lack ofruggedness of the ® rst-generation air-actuated valve andother additional pieces of HPLC equipment includingcolumns.

Recently, with the availability of commercial onlinesolid-phase extraction system(s), the column-switchingtechnique has become an acceptable option for bioana-lytical sample preparation. At Abbott Laboratory, wehave developed several online solid-phase extractionHPLC methods using automated valves, an internalreversed-phase guard cartridge and HPLC equipmentby Shimadzu and Agilent. Internal standard forti® edplasma were clari® ed by either centrifugation or ® lteredbefore the HPLC injection. These methods have beenused successfully to analyse two Abbott compounds andtheir metabolites. This application of automated column-switching valves has provided an alternative to o¥inesolid-phase extraction and liquid± liquid extraction andmade it possible to select the best technology to increaseassay throughput and sensitivity.

Abstracts of papers presented at the 2001 ISLAR

89

Automation and compliance in quality controllaboratories

Muhammad Albarakeh, Barr Laboratories, Inc., Pomona, NY,USACo-authors: Richard Ashley and Timothy Breuninger

Current laboratory demands related to the high through-put of samples, the need to keep costs and expensesminimal, yet keep the release function ¯ owing to preventproduct backorder, have necessitated the use of labora-tory automation and robotics systems. As a result of thisautomation, huge amounts of samples and related dataare processed and generated.

Without a properly controlled and validated system, thevery tools implemented to process high amounts ofsamples can bring your release function to a halt. Over-views on how to prevent cGMP logs jam and reach a` zero backorder ’ release of product will be presented.

Methods’ development for monoclonal antibodyscreening: how to think like a robot

Nanci E. Donacki, MedImmune, Inc., Gaithersburg, MD, USA

A single fusion for the development of monoclonalantibodies will often generated 20 or more microplatesthat need to be screened. ELISA (enzyme-linked im-munosorbent assay) is the most common method forscreening newly developed monoclonal antibodies forantibody production and speci® city. The method,although speci® c for the antibodies, is highly repetitivefor each step and can easily be automated. However, theway the assay is performed at the benchtop is not alwaysthe best way for an automated system to run the assay.Tips and techniques for transferring an assay from bench-top to automation will be presented.

Lean manufacturing and six sigma: an evaluationof the impact of these concepts on laboratoryautomation

Nigel North, Pharmaceutical Development, GlaxoSmithKline,Ware, UK

The concepts of lean manufacturing and six sigma arenow beginning to be applied to pharmaceutical manu-facturing processes. Lean manufacturing has been suc-cessfully applied for many years in the automotive andaerospace industries with the key goal of reducing waste.Six sigma is a more recent concept involving reducingvariation in manufacturing processes which has beenimplemented in the semiconductor industry resulting insigni® cant cost savings. The combination of lean manu-facturing and six sigma provides a powerful combinationof principles to deliver robust manufacturing processeswith the elimination of non-value-added activities. Theeå ect of these concepts on how we approach automationin the laboratory will be examined together with provid-ing some perspectives on new technology that will berequired to meet these challenges.

Applications of the Zymark Staccato for in vitrodrug metabolism studies

Heather Sulkowski, Boehringer Ingelheim Pharmaceuticals, Inc.,Ridge¢eld, CT, USACo-authors: J. Richard Mount¢eld, Donald Tweedie and DraneèO’Brien

Advances in genomics, combinatorial chemistry andpharmacology screening have led to new challengeswithin drug metabolism due to large numbers of com-pounds requiring in vitro evaluation. To meet thesedemands within Boehringer Ingelheim Pharmaceuticals,Inc., the Zymark Staccato was evaluated as a robotic toolfor conducting multiple in vitro assays. The criteria for anautomated system included the following.

. A 96-well head allowing for maximum samplethroughput.

. Thermal block to maintain constant temperature.

. Flexible deck layout.

. Open-access capability

As part of the evaluation process, assays were comparedusing both manual and automated methods. The resultsare described in this presentation

Automation of several key in vitro metabolism assays hasbeen achieved. The results indicate that the automatedassays allow for a higher throughput while maintaining ahigh degree of precision. The work will be extended toadditional in vitro assays, and the strategy of comparingmanual versus automated processes will be continued aspart of the validation procedures.

Analytical method validation: an automated soft-ware approach

Patricia A. Fowler, Waters Corporation, Milford, MA, USACo-author: Michael Swartz

Method validation is a tedious process performed todetermine if an analytical method meets the require-ments for its intended purpose. In the regulated labora-tory, method validation may take many days to performthe necessary analytical tests. Data reduction and thestatistical analysis of results performed can be a verytime-consuming process. There is also a greater poss-ibility of introducing error when calculations are per-formed manually. With the use of automated software toperform these calculations, method validation is muchfaster and easier, with less chance for error.

In this poster, we will show that an analytical methodwas validated using automated software. Chromato-graphic results were directly accessed from a relationaldatabase bypassing manual intervention. Statistical cal-culations were performed automatically and a reportgenerated showing the results of the analyses from theStudent, Cochran, Dixon and Fisher tests. Graphs weregenerated representing the results of the statistical analy-sis. In addition, we will show that the data reduction andstatistical calculations necessary to validate the method,complete with the necessary documentation and reportgeneration, are completed in signi® cantly less time.

Abstracts of papers presented at the 2001 ISLAR

90

Streamlining the dissolution test: automatedHPLC techniques

Patricia A. Fowler, Waters Corporation, Milford, MA, USACo-authors: Kelly A. Johnson, Michael E. Swartz and CharlesH. Frasier

The fast pace of the pharmaceutical industry requireslaboratories to reduce the analytical burden of theirtest procedures and increase productivity while stillsatisfying regulatory compliance. There are several ways -to meet these challenges in the dissolution laboratory.By automating the dissolution process, pharmaceuticallaboratories eliminate the slight variations that mayoccur in manual methods, insuring reproducibledata, higher throughput and cost reduction. Vali-dated single-source software control of the entiresystem, as well as dissolution data acquisition, calcu-lations and reporting, can further streamline the work¯ ow while maintaining FDA compliance with 21 CFRPart 11.

Automated online HPLC dissolution systems canpool dissolution samples signi® cantly to save time.Similarly, automated sampling at shorter intervalsand analysis of a large number of samples by onlineHPLC may provide a more complete solution forthe decision-making process in the early stages ofdrug development. In addition, these systems must becapable of handling increasingly complex formulationssuch as multiple actives and widely varying dosage levels,as well as diå erent media types such as buå ers andsurfactants.

Our poster shows applications used to help reduce theanalytical burden and increase the sample throughputand productivity in the laboratory.

Homogeneous high-throughput SNP assay directfrom genomic DNA

Patrick B. Cahill, Genome Therapeutics Corporation, Waltham,MA, USACo-authors: Dina Weymouth, Cori Gustafson, James Hurley,Michele Bakis, Veena Kamath, Doug Smith and Lynn Doucette-Stamm

Genome Therapeutics Corp has developed ahigh-throughpu t SNP assay based on the ability ofsome DNA polymerases to proofread the sequenceas they extend. This Exo-Proofreading SNP assaycan be accomplished, from sample addition to detection,in one tube. In addition, this SNP assay has thecapability to be performed directly on genomicDNA. Results will be presented for this assay on multipleSNPs screened on large populations. The results willbe compared with ASO and RFLP data with dis-crepancies resolved by sequencing. This assay providesa new, highly robust and fast method for large-scale SNPscreening required for human disease and pharmacoge-netic studies.

Development of a lead identi� cation platform forkinase drug discovery

Paul Gallant, Millennium Pharmaceuticals, Inc., Cambridge,MA, USA

The goal of any drug-discovery process is to eliminatequickly ` poor’ leads while rapidly advancing those withthe highest potential. Millennium has used a combina-tion of novel and mature technologies to establish asystem to progress novel kinase targets rapidly throughassay development, high-throughput screening, hit vali-dation and hit characterization. By selecting focusedtechnologies, automation, secondary screens and IT, anintegrated process has been constructed which is capableof handling multiple targets quickly while identifying thecompounds with the highest potential for lead optimiza-tion. This presentation will discuss the platforms chosenfor assay development, high-throughput screening, hitcon® rmation and potency/selectivity determinations. Aprocess will be described to show how using a standardwell-de® ned system can eå ectively move multiple kinasetargets rapidly into the ` Lead Optimization’ stage.

Optimization and automation of bioanalyticalmethods

Peter D. Bryan, Forest Laboratories, Abbott Park, IL, USACo-authors: Min S. Chang and Anita Shen

Now that LC-MS has become the technique of choice inthe GLP bioanalytical laboratory, the work¯ ow bottle-neck has shifted from the separation ± detection to thesample-processing portion of the bioanalytical method.To address the sample-processing bottleneck, eå orts havebeen concentrated on the conversion and validation ofexisting liquid± liquid extraction (LLE) to solid-phaseextraction (SPE) methods. Profound increases in samplethroughput have been achieved using LC-MS withautomated SPE in the 96-well format over manualLLE extraction HPLC methods. Validation of auto-mated 96-well format SPE methods have shown theirequivalence to manual LLE extraction. Automated 96-well format SPE is also much less tedious and allows theanalyst to concentrate more on data analysis and com-pliance issues. For routine analysis, transfer from clinicaltubes to the 96-well format has been accomplished usingeither the Biomek 2000 or the Hamilton MicroLab AT.Automated SPE is then performed on the Biomek 2000.

Additionally, semi-automated method development for96-well format SPE has been automated using theBiomek 2000.

How small is enough?

Peter Grandsard, Amgen, Inc., Thousand Oaks, CA, USACo-authors: Jim Petersen, Brian Rasnow, Mike Johnson, ChuckLi, Les Walling and Doug Overland

A signi® cant technical and organizational challengefacing the biopharmaceutical industry is the incorpora-tion of miniaturization technologies into their business.The successful application of these technologies requiresanswers to the following questions. What is this newminiaturization technology? Which technologies should

Abstracts of papers presented at the 2001 ISLAR

91

be applied to which work¯ ows? Why? When should theybe implemented? Should a technology be implementedearly, through some sort of technology access programme(TAP), or should one wait until it becomes commerciallyavailable? Our views on these matters will be presented.

In collaboration with other Amgen R&D groups, theResearch & Automation Technologies department hasbeen actively cultivating critical collaborators withknowledge of submicrolitre liquid-handling, microma-chining and detection technologies. For example, ourTAP with Caliper, Inc., has increased our institutionalunderstanding of miniaturization applications in thedrug-discovery process ¯ ow, most particularly in smallmolecule screening and DNA analysis.

Miniaturization at Amgen has invoked a cascade of newprojects or at least studies. The goals of these projects areto ® nd solutions for interfacing any miniaturized plat-form to the ` macro-world’ , such that the advantages ofminiaturization are not undone by integration or certaininterfaces. Operations that need to undergo changes tobene® t fully from miniaturization include the storage andretrieval of small molecules, detection and compoundhandling during screening, and information manage-ment. We will discuss some of our ® ndings and solutions.

Automation and tracking of combinatoriallibraries

Phil Small, Tripos Receptor Research, Bude, UK

A process-integrating design, synthesis and analysis ofcombinatorial libraries has been implemented at TriposReceptor Research. Crucial to the success of this processhas been the development of a proprietary informaticssystem. This has so far been developed to manage reagentinventory, track samples, and record data from synthesisand analysis to provide a valuable database of compoundinformation.

The combination of Tripos’ s proprietary design softwarewith automated synthesis and informatics leads to theproduction of drug-like libraries with well-de® ned purityand full synthetic history.

This presentation will cover what were considered theimportant aspects involved in setting up and managingan automated chemistry facility.

Automating pH solubility and stability determina-tions conducted in support of early development

Phil Waters, GlaxoSmithKline, Research Triangle Park, NC,USACo-authors: James Ormand, David Igo and Pingyun Chen

Evaluation of the equilibrium solubility and chemicalstability of compounds as a function of pH aids ourunderstanding of the bioavailability of a drug candidate.Additionally, these data lead to rapid development offormulations for use in preclinical and clinical studies.Conducting these measurements often requires a signi® -cant amount of human labour and time. Reducing thelabour and time burden is expected to increase thefrequency with which these measurements can be ap-

plied, thus increasing sample throughput and improvingthe quality of decisions impacting compound selectionand evaluation.

This presentation will describe sample-handling method-ologies and automated instrumentation developed toconduct solubility measurements as a function of pH.The use of these methodologies and instrumentation hasimproved both throughput and eæ ciency while main-taining the accuracy that can be achieved with manualmethods. The accuracy and precision of the system willbe illustrated using data on model compounds as well asdrug candidates. Finally, we will describe the strategyand process used to prepare samples for stability deter-minations at user-de® ned pHs, and how this process isdone in concert with the solubility determinations.

Development of novel micro-array technology forcell-based assays

Quiyang Zhang, Cytoplex Biosciences, Plano, TX, USACo-author: Alex Freeman

Cell-based assays are increasingly used for high-through -put screening in drug-discovery programmes becauseof their high information content. For the implementa-tion of miniaturized cell assay, micro-arraying is anessential part of the technology. Initially, we had per-formed cell-adhesion studies and evaluated ® ve diå erentsubstratesÐ polystyrene, polycarbonate , silicon, glass andPDMSÐ and found that with polylysine coating, theglass, silicon and PDMS substrates served well for cellattachment. However, the conventional well-less spottingand assaying method was diæ cult to implement inproducing a predetermined array pattern of cells.

To overcome the above diæ culty, we further developedboth bottomed and bottomless well arrays to address theneed for cell culturing and assays. PDMS and glass wereused as the bottom surfaces for both bottomless wellsilicon substrates and bottomed wells. Novel liquid distri-bution arrays were fabricated in silicon to facilitate initialcoating and liquid exchange from the entire array. Thearrays were implemented for culture of adherent andnon-adherent cells, immunochemical assays, and auto-matic liquid transfer all in high-density format. Thecurrent array platform will be also useful in cell libraryconstruction, combinatorial library synthesis, and con-tinuous homogeneous and heterogeneous assays.

Adaptive powder-dispensin g system

Rajesh K. Maheshwari, Schering-Plough Research Institute,Union, NJ, USACo-authors: Annaniy Berenshteyn, Gary Kowalski and JosephNorgard

This poster is about a robotic powder dispensing systemwe have developed. It can dispense on average10 mg § 5% from any of the 288 dispensers on itscarousel, into a sealed 24/96-well deep-well microtitreplate. The powders are kept in a sealed environment andare diå erentiated for dispensing purposes by theirdispensing parameters, which are stored by in aMicrosoft Access database. The database entry is pointed

Abstracts of papers presented at the 2001 ISLAR

92

to by a barcode label on the dispenserÐ hence, adaptivedispensing.

The user interface for this multiprocessor system consistsof 250 000 lines of Multithreaded/Multitasking VisualC‡‡ code. An Access database holds dispensingparameters ’ information for each powder. A DynamicQuery Screen has been designed to make it easy for achemist to view the database selectively.

Such a system can ® nd other uses such as distributingcompounds in the compound distribution centre, dispen-sing lyophalized microorganisms, dispensing resins forcombinatorial chemistry, and any other applicationwhen powder is to be delivered precisely in a controlledenvironment.

Automated system for compound library screen-ing by MTS cell-viability assays

Randall Engler, Kendro Lab Products, Newtown, CT, USACo-authors: C. Elliot and R. Moody

Cell proliferation, cytotoxicity and other viability assaysplay an important role the drug-discovery process. Manyscreening programmes target anticancer compounds,requiring in-vitro characterization of eæ cacy. Addition-ally, all potential therapeutic compounds must be char-acterized for their cytotoxicity regardless of the target(ADME/Tox).

Cytokinetics has developed a system to automate theprocess of reagent addition, incubation and readingof cell viability assays using the MTS method ofPromega, Inc. The integrated system includes hardwarefrom a variety of sources and software developed byCytokinetics.

Control of environmental conditions including CO2,temperature and humidity are important factors inaccurate and reproducible cell-based assays. The systemdescribed here includes the Heraeus1 Cytomat1 6000series incubator from Kendro for the incubation of cells.

Automation development for high-throughputphage library screening

Randy Yen, Genentech, Inc., San Francisco, CA, USACo-authors: Sherry Yeh and Suki Hyare

Phage display technology is a powerful tool in genomicand drug development. It allows scientists to screenquickly billions of peptides, antibodies and cellularproteins for binding to a target. It is a vital tool in studiesaimed at identifying molecules that bind to a speci® ctarget and at improving particular features of pre-exist-ing molecules. At Genentech, we automated the bindingassays for the speci® c target to screening the phagelibrary. By running fully automated 384-well assay onthe robotic system, we screened thousands of clones in amultiple target in 1 day and ® nished each library in ashort time.

Comparison and contrast of di erent dissolution-sampling systems using the Agilent 8453 UV-VISin a QC laboratory

Reem Malki, Andrx Pharmaceutical, Fort Lauderdale, FL, USACo-author: Michael K. Michaels

An integral part of an eæ cient QC laboratory is thoroughplanning. With a new facility about to open and manydiverse projects in the pipeline, the proper approach to asuccessful implementation is to think, research, plan,budget, set up, test, comply and train.

We knew as a laboratory that we could not function withonly one type of dissolution system. Our research led us totwo automated options. The ® rst, a peristaltic pumpsampling system using a single-¯ ow cell approach, oå ersthe detection of the sample. The second, a syringe pumpsampling system using a multi¯ ow cell approach, pro-vided us the options to detect, detect and collect, andcollect and dilute the sample. Each of these dissolutionset-ups is paired with an HP8453 UV-VIS while remain-ing 21 CFR Part 11 compliant.

Cross-functional team approach to implementingand managing laboratory automation in the qual-ity control laboratory

Timothy Reilly, Novartis Pharmaceutical Corporation, Su¡ern,NY, USA

Poorly de® ned project goals and an inadequate infra-structure have hindered previous attempts to initiateautomation projects in the Novartis Quality Controllaboratory. This haphazard approach has been replacedwith well-de® ned project expectations and agreementamong the functional groups that make up the projectinfrastructure to support aggressively the automationproject goals.

New concept for automation of dissolution testsfor biowaiver studies

Rolf Rolli, Sotax Ltd, Allschwil, Switzerland

New FDA regulations allow biowaivers for Class I drugsbased on solubility, permeability and dissolution testing.BCS-based biowaivers can be requested for signi® cantpost-approva l changes (e.g. Level 3 changes in com-ponents and compositions) to a rapidly dissolving im-mediate release product containing a highly soluble,highly permeable drug substance, provided that dissolu-tion remains rapid for the post-change product and bothpre-/post-change products exhibit similar dissolutionpro® les. Dissolution tests have to be performed at pHs1, 4.5 and 6.8. With this regulation, automation ofdissolution testing gains further importance with respectto pH changes.

A fully automated test system with the necessary softwareis described. This solution oå ers one the capability to runup to 15 USP 2 tests in series. With such a system, allsteps are fully automated, from medium preparation, totablet input and up to the printout of the reports. Thesystem includes a very eæ cient cleaning device thatprevents any carry-over between tests. Tests requiring

Abstracts of papers presented at the 2001 ISLAR

93

baskets are handled with the Basket-Station . This systemallows up to 10 USP 1 tests to be analysed. With thisautomated dissolution concept, biowaiver studies areexecuted rapidly allowing for accelerated drug develop-ment and SUPAC changes.

Integration of reagent management and computa-tionally biased combinatorial synthesis

Scott M. Harris, DuPont Pharmaceuticals Research Labora-tories, San Diego, CA, USA

One of the key aspects to successful high-throughput drugdiscovery is the ability to integrate several functions.These include medicinal chemistry, computationa l de-sign, high-throughput synthesis and puri® cation. A po-tential bottleneck in this process is a lack of establishedchemistries linked with readily available commercial andproprietary building blocks. Early in the process, it iscritical that a protocol exists that allows for planning andsynthesizing libraries, developing initial structure ± activ-ity relationships and validating screening leads.

It is also important to have access to novel buildingblocks and reagents for preparing targeted or focusedlibraries. DuPont Pharmaceuticals has developed asimple custom Web-based monomer request system thatuses a JAVA GUI that allows for tracking of orders andreport generation. The data are housed in a ReagentInventory Tracking System using Oracle and the physi-cal reagents are stored in a controlled ventilated cabinetsystem. The typical throughput and average turn aroundtime will be discussed. The overall bene® ts of thisprogramme are lower costs, better inventory controland an increase in eæ ciency in library production andchemistry.

Automation of chemistry in a core DNA-sequen-cing facility

Shawn Hallowell, P¢zer Global R&D, Groton, CT, USACo-authors: Suzanne P. Williams, Yevette C. Clancy, MelissaT. Cronan, Shawn E. Hallowell, Michael Polchanino¡, LanceRyley and Janice L. Palmer

Conversion of our core DNA sequencing laboratory tothe 3700 DNA analyser (AB) and sequencer for BioLIMSsequence analysis and database software (AB, genecodes)has greatly reduced the amount of manual interventionrequired for generating, analysing and maintaining DNAsequence data.

The set-up of the DNA sequencing reaction is the nextprocess that has been selected for improvement. Cur-rently, requests for sequencing from the TherapeuticArea laboratories are (1) submitted online, (2) thesample information is transferred to the 3700 plate recordand (3) the addition of DNA and primers to the thermalcycle plate is performed manually.

Manual chemistry requires up to 4 h of hands-on time perday depending on the number of samples. While a liquidhandler working oå a ® le can perform the hit-pickingrequired for core terminator chemistry, the set-up ofthe source racks is time-consuming and a possible sourceof errors. The poster will present the use of Websubmission, individually bar-coded sample tubes and aliquid handler for automating chemistry set-up in a coreDNA sequencing facility.

Micro-array process industrializatio n

Shane Weber, Millennium Pharmaceuticals, Cambridge, MA,USA

Millennium’ s nylon cDNA micro-array platform is suc-cessfully ® lling our drug- and marker-discovery plat-forms. This platform is eå ective because of itssensitivity, gene content, throughput eæ ciency and cost.The platform’ s eå ectiveness results from the industria-lized process ¯ ow of liquid handling, printing andinformation tracking. Eå ectiveness is constantly assessedby measurements of quality and eæ ciency. The process¯ ow and quality assessment will be reviewed.

Mass-directed puri� cation of combichemlibraries

William R. Hall, GlaxoSmithKline, Research Triangle Park,NC, USACo-authors: Melissa Lindsay and Dean Phelps

The Library Puri® cation Group at Research TrianglePark is responsible for puri® cation of both large andsmall libraries from discovery and targeted research.We use mass-directed preparative HPLC to purifylibrary samples with masses of 5± 100 mg per well. Oursystem incorporates at-column dilution, separate wastecollection for each sample and UV veri® cation of peakcollection. Eæ cient evaporation of solvents, robotic li-quid handling and reformatting, and fast QC on puri® edsamples allows for high-throughput of puri® ed libraries.

Abstracts of papers presented at the 2001 ISLAR

94

Establishment and management of the Bristol-Myers Squibb Automated Screening Core

Steven F. Innaimo, Bristol Myers Squibb PRI, Wallingford,CT, USA

As part of the evolution and industrialization of the BMSLead Discovery organization, a core group of specializedscientists focused solely on automated screening wasassembled. The mission of the Automated Screening Coreis to adapt and prosecute high-throughput screen cam-paigns on fully automated robotic screening systems.

This talk will cover the implementation of the BMSautomated screening infrastructure, the rationale forthe establishment of the Automated Screening Coreand the organizational impact of such a group.

Determining feasibility and parameter values forcompound orders based on transfer amounts,liquid-handler capabilities and container-volumecapacities

Steven Ho¡man, Bristol-Myers Squibb PRI, Princeton, NJ,USACo-author: Mark F. Russo

An algorithm is described for determining the feasibilityand calculating parameter values for one or morecompound processing orders. Feasibility is determinedprimarily by applying a series of inequality constraintsthat result in a valid target concentration range withinwhich orders can be processed. These inequality con-straints are obtained from initial compound amounts,source and destination container-volume capacities, theliquid-handler transfer volume range and the amount ofcompound to be processed. If feasible, calculated par-ameter values include the dilution, transfer and top-oåvolumes necessary to ful® l the order. Sample codeimplementing the algorithm is given in Microsoft VisualBasic1.

High-throughput screen to detect inhibitors of the3 0-5 0-HCMV exonuclease

Steven J. Conrad, Pharmacia Corporation, Kalamazoo, MI,USACo-authors: Thomas W. Pitts, Nancee Oien, Robert A. Anstadt,Roger A. Poorman, Peter A. Wells, Michael W. Wathen andYoshihiko Yagi

We carried out a ¯ uorescence polarization-based (FP)high-throughpu t screen (HTS) to detect inhibitors of the3 0-5 0-exonuclease activity of the human cytomegalovirusDNA polymerase (the HCMV exonuclease). We used a23-mer ssDNA substrate labelled with TAMRA at the 3 0

end. The emitted light from TAMRA in an intactsubstrate molecule was highly polarized (about240 mP), but, upon substrate digestion with the HCMVexonuclease, the emission from TAMRA free in solutionexhibited a lower polarization (about 40 mP).

The assay was designed to run in a 384-well format witha ® nal volume of 50 ml. The Km of the HCMV exonu-clease for the substrate was determined to be 20 nm. Inthe assay, 50 nm substrate was digested by 0:01 £ stock

HCMV exonuclease in 50 mm HEPES, pH 7.6, whichproduced a 50% reduction in mP in 8.3 min. Wescreened 117 843 compounds at 10 mm (2.5% DMSO).The screen identi® ed 647 compounds that inhibited theHCMV exonuclease3 30%.

High-throughput protein function screening forimmune system target validation

Stewart D. Chipman, Immunex, Corporation, Seattle, WA, USA

Validation of the biochemical function of a novel gene/protein in the pathophysiology of a disease is the rate-limiting step in novel target validation for drug discov-ery. Many groups have applied the technology andmethodology developed for small molecule high-through-put screening and clinical testing to assays that test fornovel protein function. This presentation will addresshow we have integrated gene discovery and expressionanalysis, protein production and inventory capability,cell-based protein function assays, laboratory automationtools and data management systems into an integratedtarget discovery and validation platform for immunesystem targets. Topics to be discussed are automatedand multiplexed cytokine detection, cognate liganddetection, calcium mobilization, high-throughput opticalimaging, large-scale cell preparation, large-scale proteinexpression and puri® cation, workstation style laboratoryautomation and automated data handling.

Laboratory automation: a cost-e ´ ective solutionto increasing laboratory workloads

Stephen Green, Forest Laboratories Ireland Ltd, Dublin, Ireland

Forest Laboratories, Inc., markets Celexa, an SSRIantidepressant, in the USA. The bulk tablets are pro-duced at the manufacturing site in Dublin, Ireland.Market projections in 1999 indicated a 250% increasein volumes to be produced at the Dublin site. Severallaboratory strategies to cope with this increased produc-tion were assessed and the use of automated roboticsystems was shown to be the most cost-eå ective solution.

The introduction of the automation into the laboratoryhad several phases. Initially, there were physical changesrequired to the laboratory layout to accommodate therobots. Second, there was the quali® cation of the roboticsystems and the training of the initial users. Finally, therewas the development and validation of several analyticalmethods, which was performed in conjunction with theZymark MTOV group.

Characterization and e´ ective use of the molecu-lar properties of reaction vessel surfaces inhigh-throughput assay development

Sven Erik Rasmussen, Nalge Nunc International, Naperville, IL,USACo-author: Nalge Nunc

An understanding of the molecular surface of a reactionvessel is critical to achieving the sensitivity and speci® cityfor high-throughput or high-content molecular and cell-based assays. We developed and characterized several

Abstracts of papers presented at the 2001 ISLAR

95

molecular surfaces for reaction vessels that can be eå ec-tively used in genomic and proteomic assays. Thesesurfaces, in conjunction with the physical properties ofthe substrate material (optical or thermal properties, forexample) along with format (MicroWell, slide, or chip),are compared with regard to assay performance.

Through XPS analysis (also known as ESCA), the mol-ecular and atomic compositions of various MicroWellsurfaces have been de® ned. These data will be used todemonstrate the speci® c utilization of the surface inmolecular and cell-based assays. For example, a hydro-philic, highly charged surface, MultiSorp, preferentiallybinds complex phospholipid molecules while excludingother molecules, including glycoproteins such as IgG.Another ` high-binding ’ hydrophilic surface, MaxiSorp,binds a signi® cant amount of glycoproteins such asimmunoglobins. These surfaces and others such asthe Nunclon Delta cell culture surface and non-treatedsterile polystyrene surfaces are molecularly de® ned byESCA techniques accompanied by speci® c proceduralexamples.

Well geometry, round, square, columnar or ` shallow’ ,in uences assay performance, particularly when an ac-tive surface is used. Additional parameters come intoeå ect when well volumes are miniaturized as in a 1536-well plate. The eå ect of a molecular surface changesagain as the assay is further miniaturized to a two-dimensional format as in a slide or chip and the in¯ uenceof the substrate material may increase.

Also important to the development of a speci® c, sensitiveand eæ cient assay are the physical characteristics ofvarious substrate materials. With regard to optical prop-erties, the ¯ uorescent and re¯ ective properties of solidblack and solid white as well as black or white opticalbottom plates (OBP) were compared by measuringsignal-to-noise ratios, light cross talk and sensitivity. Dataand practical applications will be discussed.

New polymer substrate for DNA micro-arrays

Svend Erik Rasmussen, Nalge Nunc International, Naperville,IL, USACo-authors: T. Kristensen, K. HolmstrÖm and L. Pedersen

The quality and surface properties of the micro-arraysubstrates are fundamental factors for a successful DNAmicro-array system. At present, coated glass-slides(amine-, alehyde- or epoxy-coated) are the dominatingsubstrates on the market. However, an often-encounteredproblem using these substrates is a non-homogenoussurface, which can lead to problems with spot uniformityand morphology. Nunc A/S has developed a pre-acti-vated, but uncoated, NucleoLinkTM polymer MicroAr-ray Slide that can be used, without any furthermodi® cation, for creating DNA micro-arrays. The Nu-cleoLinkTM polymer MicroArray Slide can be used forboth covalent attachment of DNA as well as immobiliza-tion of DNA by electrostatic forces and/or ionic bonds.

Studies of auto¯ uorescence properties, DNA-binding ca-pacity and DNA-hybridization eæ ciency have beenperformed and compared with conventional glass sub-

strates. Data from oligonucleotide arrays as well as cDNAarrays will be presented.

Flexibility paradigm in laboratory automationmanagement

T. C. Ramaraj, Roche Discovery Technologies, Ho¡mann-LaRoche, Nutley, NJ, USA

Considerable developments and rapid changes havetaken place in high-throughput screening and compoundmanagement in the last 5 years. This is a result of aconcerted eå ort between the pharmaceutical industry,laboratory automation companies, instrument companiesand manufacturers of consumable laboratory ware. Theevolution of new automation technologies has fuelled thegrowth and prospects for various modes of automation invirtually all phases of drug-discovery research. Here, wewill try to highlight our experiences in managing acomplex HTS and compound management infrastruc-ture at Hoå mann-La Roche.

The infrastructure consists of automation platforms thathave evolved over several years. All require proper careand maintenance for sustained performance. Some re-cently installed con® gurations require continued valida-tion to achieve expected levels of performance andreliability to match ever-increasing throughput require-ments. Promising new automated screening technologiesrequire careful evaluation. Validation of these new tech-nologies also require allocation of signi® cant resources formeeting demanding requirements of miniaturization,speed, ¯ exibility, cost eæ ciency and increased perform-ance levels. It is no easy task to steer winning method-ologies to their fullest potential and at the same time takea risk in exploring unproven avenues and technologies.

To maintain a balanced perspective among availablechoices requires ¯ exibility in approach, quick implemen-tation strategies, a diverse pool of expertise and talentthat can work as a team, contribute as individuals, sharethe successes as well as failures, learn from the mistakes,revise the expectations and still meet corporate goals andobjectives every year. The presentation will also includeexamples from some of our most recent integrationexperiences, the upgrade of existing systems, the imple-mentation of new detection modes in existing instrumentsand our experience with latest 384-well pipetting systems.

Variety of automation tools for drug metabolism

Thomas Lloyd, DuPont Pharmaceutical Co., Newark, DE, USA

Automation continues to be developed for new diversi® edapplications within drug metabolism. The choice of anautomated tool can vary considerably depending on thespeci® c requirements of the application. An overview ofthe automated tools that have been introduced atDuPont over the last 5 years will be presented alongwith speci® c application examples. An evaluation isoå ered for how the diå erent tools have proliferatedthroughout this function.

Automation application areas include bioanalyticalsample preparation techniques including solid-phaseextraction, liquid± liquid extraction, protein precipitation

Abstracts of papers presented at the 2001 ISLAR

96

and online approaches for in vivo samples, in vitro samplepreparation processes, sample handling, data handlingand training. Application areas include many functionswithin drug metabolism ranging from discovery throughclinical trials.

Characterization and e´ ective use of the molecu-lar properties of reaction vessel surfaces toachieve optimal assay performance

Tom Cummins, Nalge Nunc International, Rochester, NY, USACo-authors: Janne N. Knudsen, Lena B. Larsen, Svend ErikRasmussen, Tom Cummins and Joseph Granchelli

An understanding of the molecular surface of a reactionvessel is critical to achieving the sensitivity and speci® cityfor high-throughput or high-content molecular- and cell-based assays. We developed and characterized severalmolecular surfaces for reaction vessels that can be usedeå ectively in genomic and proteomic assays. Thesesurfaces, in conjunction with the physical properties ofthe substrate material (optical or thermal properties, forexample) along with the format, are compared withregard to assay performance.

Through ESCA analysis, the molecular and atomiccompositions of various MicroWell1 surfaces have beende® ned. These data will be used to demonstrate thespeci® c use of the surface in molecular and cell-basedassays. For example, a hydrophilic, highly charged sur-face preferentially binds complex phospholipid moleculeswhile excluding other molecules, including glycoproteinssuch as IgG. Another ` high-binding ’ hydrophilic surface,MaxiSorpTM, binds a signi® cant amount of glycoproteinssuch as immunoglobins.

Well geometry, round, square, columnar or ` shallow’ ,in uences assay performance, particularly when an ac-tive surface is used. Additional parameters come intoeå ect when well volumes are miniaturized as in a 1536-well plate.

Also important to the development of a speci® c, sensitiveand eæ cient assay are the physical characteristics ofvarious substrate materials. With regard to optical prop-erties, the ¯ uorescent and re¯ ective properties of solidblack and solid white as well as black or white opticalbottom plates (OBP) were compared by measuringsignal-to-noise ratios, light cross-talk and sensitivity.Data and practical applications will be discussed.

Intelligent automated dissolution system! Achallenge worth striving for

Umesh V. Banakar, Pharm-Assist International, Carmel, IN,USACo-author: Manoj Bagool

Dissolution testing requirements have been extended toall oral dosage forms over the past decade. Additionally,the dissolution regulations for pharmaceutical formula-tions have been tightened world-wide. Furthermore, thepast decade has also witnessed an increase in popularityof modi® ed release pharmaceutical s that inherently re-quire extended dissolution protocols. The increasedemphasis on correlating dissolution and bioavailability

data points to the desirability of more frequent samplingpoints to validate a dissolution pro® le. As a result, thesedevelopments have signi® cantly increased the require-ments of equipment, personnel, data processing andvalidation, which translates into increased productiontime for oral dosage forms. Thus, automation in dissolu-tion testing has become a necessity in order to handlethese demands.

Automation in dissolution testing is not new to thepharmaceutical industry. Simultaneous to the develop-ment in pharmaceutical technology over the past decade,there have been advances in automated dissolutiontesting, although at a slower pace. The virtual explosionin the diverse and yet speci® c requirements associatedwith a plethora of dosage form types and their respectivefunctions have curtailed the outright all-encompassingdevelopment in automated dissolution testing. While theadvantages of automated dissolution testing are unequi-vocal, the limitations of automated dissolution testingsystems cannot be refuted. There is a constant quest toimprove an existing automated system that meets speci® crequirements, either a unit function of the test or toaccomplish a speci® c expectation of a dosage form, thuslimiting the universality of the automated system.

Robert Kennedy once said, ` Some people watch thingshappen and ask the question, why; and there are thosethat dream of things and ask the question, why not! ’ It isbeyond doubt that an automated dissolution testingsystem that addresses the current limitations of theexisting ones and yet is ¯ exible and universal enough isthe need of the day. Going beyond the minimumexpectations of such an automated dissolution test intoits utility during early drug development and above all insimultaneous prediction of bioavailability, can be adream worth striving for. A vision for such an automateddissolution test system will be presented.

Development of a radiofrequency identi� cation(RFID) data carrier interface with an analyticalbalance

Je¡rey Veitch, GlaxoSmithKline, Ware, UKCo-authors: Tony Allcock and James Ormand

The use of process instrumentation such as analyticalbalances within the laboratory is commonplace and theneed for capturing related process measurement data areoften a necessity. Quite often these data need to beexported to databases and invariably to laboratoryinformation management systems (LIMS). Laboratoriesregularly use process equipment such as balances, pHmeters, etc. that are not attached to a dedicated PC. AtGlaxoSmithKline, we have developed a way of capturingprocess data (e.g. weight values) directly to radiofre-quency identi® cation (RFID) smart labels, where theselabels are used as data carriers.

This evaluation has proven to be a success and ispotentially the stepping stone towards a more speci® cuse of RFID within GlaxoSmithKline.

Abstracts of papers presented at the 2001 ISLAR

97

System integration for genomics using biorobotworkstations

Achim Wehren, QIAGEN Hilden, Germany, and Valencia, CA,USACo-authors: Fred Siegmanand Carola Schade

Automated robotic workstations are frequently used asstand-alone systems for sample puri® cation, reaction set-up or sample re-array in the ® eld of genomics. However,a variety of external instruments often must be integratedto provide full automation of several sequential tasks or toincrease the range of applications that can be performed.

QIAGEN1 Instruments provides complete solutionsto automate molecular biology and liquid-handlingapplications by integrating BioRobot workstationswith a wide range of accessory instruments, includingthermal cyclers, spectrophotometer s and the BioRobotRapidPlate1, a 96-channel pipetting system with cap-abilities for both 96- and 384-well pipetting. Theseintegrated systems provide rapid and fully automatedprocessing for applications such as DNA template nor-malization, PCR, sequencing reaction set-up and sampletransfer tasks.

BioRobot1 3000 extended-arm systems are designedto accommodate complementary instruments on boththe left and right sides. Additional integration ofthe BioRobot TwisterTM increases system storage ca-pacity for microplates, blocks and disposable tips, andallows the integration of accessory instruments with bothextended-arm and standard BioRobot workstation con-® gurations.

BioRobot1 8000 robotic workstations can be integratedwith accessory instruments using the BioRobot Twister IIexternal arm for walkaway front-end processing, e.g.from plasmid puri® cation to ® nal reaction set-up. Thelarge size of the BioRobot 8000 worktable and the highstorage capacity of the Twister II robotic arm allowunattended operation over extended periods.

QIAsoftTM, the BioRobot Operating System, commu-nicates with external instrument software, controls pro-cesses performed on the BioRobot and coordinates theactions of the BioRobot and external instruments.

Dispensing precision for the SciCloneTM

auto-mated liquid-handling workstation for 96-channelpipetting

Rudy Willebrords, Janssen Research Foundation, Beerse,Belgium

One of the ® rst steps in drug discovery involves identi® ca-tion of novel compounds that interfere with therapeuti-cally relevant biological processes. Identi® cation of ` lead’compounds in all therapeutic areas included in a drug-discovery programme requires labour-intensive evalua-tion of numerous samples in a battery of therapy-targetedbiological assays. To accelerate the identi® cation of leadcompounds, JRF has developed an automated high-throughput screening (HTS) based on the unattendedoperation of a custom Zymark-tracked robot system.Automation of enzymatic and cellular assays was realizedwith this system adapted to the handling of microtitreplates. The microtitre-plate technology is the basis of ourscreening. All compounds within our chemical library arestored and distributed in micronic tube racks or micro-titre plates for screening. An eæ cient in-house-developedmainframe-based laboratory information managementsystem supported all screening activities. Our experienceat JRF has shown that the preparation of test compoundsand making serial dilutions have been rate-limiting stepsin the overall screening process. To increase compoundthroughput, it was necessary both to optimize therobotized assays and automate the compound supplyprocesses. In HTS applications, one of the primaryrequirements is highly accurate and precise pipettingof microlitre volumes of samples into microplates.The SciCloneTM is an automated liquid-handling work-station capable of both 96- and 384-channel high-preci-sion pipetting. For high-throughput applications, theSciCloneTM instrumentation can pipette a variety ofliquid solutions with a high degree of accuracy andprecision between microplates (interplate variability)and tip-to-tip (intraplate variability) within a singleplate. The focus of this presentation is to review theliquid-handling performance of the SciCloneTM systemas a multipurpose instrument for pipetting aqueous ororganic solutions, and virus suspensions into 96- and 384-well microplates. The capabilities of the system and theresulting bene® ts for our screening activities will bedescribed.

Abstracts of papers presented at the 2001 ISLAR

98

Submit your manuscripts athttp://www.hindawi.com

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Inorganic ChemistryInternational Journal of

Hindawi Publishing Corporation http://www.hindawi.com Volume 2014

International Journal ofPhotoenergy

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Carbohydrate Chemistry

International Journal of

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Journal of

Chemistry

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Advances in

Physical Chemistry

Hindawi Publishing Corporationhttp://www.hindawi.com

Analytical Methods in Chemistry

Journal of

Volume 2014

Bioinorganic Chemistry and ApplicationsHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

SpectroscopyInternational Journal of

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

The Scientific World JournalHindawi Publishing Corporation http://www.hindawi.com Volume 2014

Medicinal ChemistryInternational Journal of

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Chromatography Research International

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Applied ChemistryJournal of

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Theoretical ChemistryJournal of

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Journal of

Spectroscopy

Analytical ChemistryInternational Journal of

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Journal of

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Quantum Chemistry

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Organic Chemistry International

ElectrochemistryInternational Journal of

Hindawi Publishing Corporation http://www.hindawi.com Volume 2014

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

CatalystsJournal of