ASSAY and Drug Development TechnologiesVolume 1, Number 3, 2003© Mary Ann Liebert, Inc.

Literature Search and Review

J. Fraser Glickman

In each issue of ASSAY and Drug Development Technologies, our Literature Editor, J. Fraser Glickman,Ph.D., selects several significant papers covering timely and pertinent topics that will keep our readers up-to-date on the ever-changing field of assay and drug development technologies. Dr. Glickman will provide rel-evant commentary on each of the cited abstracts.

The Matrix

489

Xia Y, Gopal KV, Gross GW: Differential acuteeffects of fluoxetine on frontal and auditory cortexnetworks in vitro. Brain Res 2003;973:151–160.

Abstract: Primary cultures of neuronal networks grown onmicroelectrode arrays were used to quantify acute effectsof fluoxetine (Prozac) on spontaneous spike and burst ac-tivity. For frontal cortex cultures, fluoxetine showed con-sistent inhibitory effects and terminated activity at 10–16mM. IC50 (mean 6 SE) for spike rates was 5.4 6 0.7 mM(n 5 15). For auditory cortex cultures, fluoxetine causedexcitation at 1–10 mM, initial inhibition at 15 mM, and ac-tivity cessation at 20–25 mM. The spike rate IC50 was15.9 6 1.0 mM (n 5 11). Fluoxetine did not change the ac-tion potential waveform shape. However, at high concen-trations, it caused total cessation of spike activity on allchannels. The inhibition caused by fluoxetine was re-

versible for both tissues. Based on the results, we concludethat cultures showed repeatable, concentration-dependentsensitivities to fluoxetine but demonstrated tissue-specificresponses for frontal and auditory cortex networks. Theseresponses may not be due to the interference with serotoninreuptake, but may be due to a secondary effect on ionicchannels.

Commentary: Assays for drug activity can range frommeasuring simple biomolecular binding interactions totests that measure the response of whole animals. As mea-surement systems become more sophisticated, researcherscan develop assays that are intermediate in multicellularcomplexity between cellular assays and whole organismassays. This article presents an example of how neuronalnetworks from the mouse embryo auditory cortex can becultured onto micro-electrode arrays and applied to drug

� �

Neurofilament antibody stained 5-week-oldneuronal network cultured on a 64-electroderecording matrix. Transparent indium–tin ox-ide conductors are 8 mm wide and terminatein four rows and 16 columns. Inserts on theright panel show greater morphological de-tails. Reprinted from Xia Y, Gopal KV, GrossGW: Differential acute effects of fluoxetineon frontal and auditory cortex networks invitro. Brain Res 973:151–160. © 2003.

Literature Search and Review490

analysis. As the neuronal networks develop, spontaneousneuro-transmission in a large population of cells or tis-sue can be measured and controlled. In this paper, the ef-fect of Prozac was studied, suggesting that the Prozac maywork not only through suppression of serotonin re-uptakebut also through side effects on ion channels. The ques-

tion remains as to whether these assays are a good modelfor the human; however, such assays are very imagina-tive and seem to suggest opportunities in developing morecomplex multicellular interaction systems. Additionally,the discussion section gives a useful review on the phar-macology of Prozac and serotonin. See figure.

A Bullseye on the Target

Oda Y, Owa T, Sato T, Boucher B, Daniels S, Ya-manaka H, Shinohara Y, Yokoi A, Kuromitsu J, Na-gasu T: Quantitative chemical proteomics for identi-fying candidate drug targets. Anal Chem 2003;75:2159–2165.

Abstract: We have developed a systematic strategy fordrug target identification. This consists of the followingsequential steps: (1) enrichment of total binding proteinsusing two differential affinity matrixes upon which are im-mobilized positive and negative chemical structures fordrug activity, respectively; (2) covalent labeling of theproteins with a new cleavable isotope-coded affinity tag(ICAT) reagent, followed by proteolysis of the com-bined proteins; (3) isolation, identification, and relativequantification of the tagged peptides by liquid chro-matography-mass spectrometry; (4) array-based tran-scription profiling to select candidate proteins; and (5)confirmation of direct interaction between the activity-associated structure and the selected proteins by usingsurface plasmon resonance. We present a typical appli-cation to identify the primary binding protein of a novelclass of anticancer agents exemplified by E7070. Ourresults suggest that this approach provides a new aspectof quantitative proteomics to find specific binding pro-teins from protein mixture and should be applicable toa wide variety of biologically active small moleculeswith unidentified target proteins.

Commentary: The advent of new bio-analytical tech-nologies and refinement of classical approaches of affin-ity chromatography should lead to a more systematic andrapid identification of the target receptors for newly iden-tified drug candidates. This paper describes a systematicapproach using compound-affinity chromatography, two-dimensional gel electrophoresis, mass spectroscopy withICAT labels, surface plasmon resonance (BiaCore), andtranscription profiling to rapidly identify the intracellu-lar targets of a phase II drug candidate (E7070). ICAT,or isotope-encoded affinity tags (Applied Biosystems),are protein labeling reagents that are isotopically dis-

tinct but chemically identical, thus enabling the nor-malization in mass spectroscopy, of two independent ex-periments. In this study, ICAT was used to distinguishbetween those proteins from a cell that bound specifi-cally to the drug, from those proteins that bound to boththe drug and its inactive analog (unspecific binders). Al-though many target candidates were identified, bindingto the target protein could be validated with surfaceplasmon resonance, and the overall best was malate de-hydrogenase. Interestingly, E7070 did not inhibit the en-zyme assay, probably because of its poor solubility inthe assay buffer, but was shown to compete with co-fac-tor binding in a BiaCore experiment. The approachseems to be broadly applicable, and has the potential tobe efficient and relatively routine in pharmaceutical re-search. See figure.

Cells

Chemical affinity matrix 1 Chemical affinity matrix 2

Binding proteins Binding proteins

Separate, detect & quantify proteins

MixLabel with tag BLabel with tag A

Ch

emic

al p

rob

e 1

Ch

emic

al p

rob

e 2

proteins proteins

Schematic representation of the procedure for chemical quantitative pro-teomics. Binding proteins to chemical affinity matrix 1 (i.e., E7070-type)or chemical affinity matrix 2 (i.e., amide-type) were labeled with differ-ent tags (Cy3 and Cy5 or ICAT reagents), combined, digested, separated,and analyzed as described in text. Reprinted with permission from OdaY. Owa T, Sato T, Boucher B, Daniels S, Yamanaka H, Shinohara Y,Yokoi A, Kuromitsu J, Nagasu T: Quantitative Chemical Proteomics forIdentifying Candidate Drug Targets. Anal Chem 75:2159–2165. © 2003American Chemical Society.

Do Heo W, Meyer T: Switch-of-function mutantsbased on morphology classification of ras superfam-ily small GTPases. Cell 2003;113:315–328.

Abstract: Signaling proteins from the same family canhave markedly different roles in a given cellular context.Here, we show that expression of 100 constitutively activehuman small GTPases induced cell morphologies that fellinto nine distinct classes. We developed an algorithm forpairs of classes that predicted amino acid positions that canbe exchanged to create mutants with switched functional-ity. The algorithm was validated by creating switch-of-function mutants for Rac1, CDC42, H-Ras, RalA, Rap2B,and R-Ras3. Contrary to expectations, the relevant residueswere mostly outside known interaction surfaces and werestructurally far apart from each other. Our study shows thatspecificity in protein families can be explored by combin-ing genome-wide experimental functional classificationwith the creation of switch-of-function mutants.

Commentary: Small GTPases are ubiquitous intracel-lular signaling proteins with a variety of critical roles incell growth and differentiation. Many of these, such asp21 ras, have been important targets for therapeutic in-tervention. This intriguing paper demonstrates that withover 100 small GTPases, which fall into several families,the cellular consequences of activation and overexpres-sion can be divided into 9 distinct morphologies. Classi-cal approaches to change the function of the GTPases bymutating critical residues known to be involved in inter-action with downstream or upstream effector moleculeshave been ineffective in switching the GTPase function;however, a novel approach using evolutionary logic re-sulted in the ability to engineer GTPases that resulted inphenotype swapping. Amazingly, the mutations identifieddo not lie in any known effector interaction domains, andthus suggest there are many other functional regions ofthese molecules, which have yet to be explored. See fig-ure.

Literature Search and Review 491

The Many Faces of GTPases

A B C DNo change Rounded Filopodia Lamellipodia Stress fibers

F G H I J Shrunk Multiple Polar Eyelashes Local spread

K

Multiple Local spread

Eyelashes

Polar

LamellipodiaFilopodia

RoundedStress fibersMultiple

Shrunk

Cell morphology changes induced by con-stitutively active small GTPases can begrouped into nine classes. (A–J) Two ex-amples of different small GTPases areshown for each of the morphology classes.The classification of each small GTPasewas based on visual inspection of at leastthree separate experiments with NIH3T3cells as well as on control experiments us-ing HeLa cells and Swiss 3T3 cells. Themorphologies were examined with TRITC-phalloidin staining after fixation and per-meabilization, as described in the Experi-mental Procedures section. Scale bar 5 20mm. (K) Sequence homologies between hu-man small GTPases and the induced cellmorphologies. Homologies were analayzedwith a Clustal W MSA algorithm and rep-resented in an evolutionary tree plot. Eachmorphology class is marked by differentcolors. Reprinted with permission from DoHeo W, Meyer T: Switch-of-Function Mu-tants Based on Morphology Classificationof Ras Superfamily Small GTPases. Cell113:315–328. © 2003.

Literature Search and Review492

A Picture Is Worth a Thousand Words

Rudin M, Weissleder R: Molecular imaging in drugdiscovery and development. Nat Rev Drug Discov2003;2:123–131.

Abstract: Imaging sciences have grown exponentiallyduring the past three decades, and many techniques, suchas magnetic resonance imaging, nuclear tomographicimaging, and X-ray computed tomography, have becomeindispensable in clinical use. Advances in imaging tech-nologies and imaging probes for humans and for smallanimals are now extending the applications of imagingfurther into drug discovery and development, and havethe potential to considerably accelerate the process. Thisreview summarizes some of the recent developments inconventional and molecular imaging, and highlights theirimpact on drug discovery.

Commentary: It is becoming evermore possible to visu-alize biochemical events in whole animals, allowing fora more detailed and predictive understanding on howcompounds and targets behave in vivo. This is a verynice review presenting the new technologies that allowfor a molecular image analysis of the behavior of drugsin vivo. The techniques, applications, advantages, anddisadvantages of technologies such as MRI, CT, PET,BLI, FMT, and SPECT are presented. The possibilitiesand examples of the measurement of binding and recep-tor occupancy and protease enzymes, in processes suchas angiogenesis and cellular tracking are summarized.Interestingly, the point is raised that although quite ef-fective as biomarkers for disease, these systems are prob-ably underutilized because they have not yet been ap-proved by regulatory agencies.

A New Orphan Adopted

Jones CE, Holden S, Tenaillon L, Bhatia U, SeuwenK, Tranter P, Turner J, Kettle R, Bouhelal R, Charl-ton S, Nirmala NR, Jarai G, Finan P: Expression andcharacterization of a 5-oxo-6E,8Z,11Z,14Z-eicosate-traenoic acid receptor highly expressed on humaneosinophils and neutrophils. Mol Pharmacol2003;63:471–477.

Abstract: Using a bioinformatics approach, we haveisolated a novel G-protein-coupled receptor (GPCR),R527, and have demonstrated that this receptor showsno significant homology to previously deorphanizedGPCRs. Quantitative reverse transcription-polymerasechain reaction analysis of the expression of GPCR R527indicated a very high level of mRNA expression ineosinophils, with high expression also detected in neu-trophils and lung macrophages. Stable cell lines weregenerated expressing this receptor together with the G-protein a-subunit G a(16). These cells were used toscreen an agonist collection in a calcium mobilizationassay and 5-oxo-6E,8Z,11Z,14Z-eicosatetraenoic acid(5-oxo-ETE) was identified as a putative ligand. 5(S)-Hydroxyperoxy-6E,8Z,11Z,14Z-eicosatetraenoic acidwas also shown to activate the receptor, whereas theleukotrienes LTB4, LTC4, LTD4, and LTE4 failed toelicit a response. In cyclic AMP (cAMP) assays, per-tussis toxin reversed the inhibitory effects of 5-oxo-ETEon cAMP production, indicating that the receptor is Ga(i)-coupled. The GPCR R527 shows pharmacologicalproperties similar to those of the previously described5-oxo-ETE receptor expressed on eosinophils, neu-

trophils, and monocytes. These cell types show chemo-tactic responses to 5-oxo-ETE, and this eicosanoid hasbeen proposed to play a key role in the inflammatoryresponse. The molecular identification of a receptorbinding 5-oxo-ETE will expand our understanding ofthe physiological role of this mediator and may providenew therapeutic opportunities.

Commentary: GPCRs are the most accepted drug tar-gets, and efforts are continuing to identify the sequenceand function of new GPCRs through the process of “de-orphaning.” In this paper, a new receptor for theeicosanoid, 5-oxo-ETE, which is a strong promoter ofchemotaxis in eosinophils (those cells involved in the al-lergic response), is described. In this study, the groupused genomic data mining to identify a sequence encod-ing a putative seven-transmembrane receptor. The se-quence was amplified from human neutrophils usingPCR, and stable cell lines expressing the protein weregenerated. Two thousand known GPCR agonists weretested in a high-throughput FLIPR screen—which mea-sured calcium mobilization in response to the agonists.Of the two lipids identified in this screen, 5-oxo-ETE wassignificantly more potent. This agonist was also shown toincrease cAMP formation in the same cell line, which wasreversible by pertussis toxin. The native expression of thereceptor was shown to be most abundant in eosinophils,human liver, and kidney. The paper presents a nice ex-ample of how high-throughput screening technologiessuch as FLIPR, data mining, and PCR can be used syn-ergistically to “deorphanize” receptors.