Diabetes is endemic in industrializedsociety. A lucrative market is waiting forcompanies that can find an effectiveremedy, but traditional HTS approachesare often ineffective in uncovering leadsfor the key targets. This, then, is a taleof two necessities: to find a moreeffective diabetes drug, and a noveldiscovery strategy to assist the search.Now, a new technology dubbed‘breakaway tethering’ is being used touncover molecules that block a keyenzyme in the diabetes diseasemechanism [1].

Establishing a targetDiabetes is a common disorder in whichblood sugar is ineffectively processed.Type 2 diabetes is the most prolificform, representing ∼90% of cases. Theprimary cause is either insufficientsecretion or action of insulin, whoserole is to facilitate the passage ofglucose into cells. When insulin is scarceor ineffective, extracellular glucoseaccumulates, starving cells of theirenergy source and causing a diversity ofsymptoms that can include blindnessand heart disease. Type 2 diabetes canoften be controlled by careful lifestylemanagement, but insulin injections ormedication often become necessary.Unfortunately, diabetes drugs oftensuffer from low efficacy as monotherapyand can cause side-effects; therefore,there is a large medical need for aneffective treatment.

Protein tyrosine phosphatase-1B(PTP1B) is a well-validated target intype 2 diabetes drug research. Thisenzyme modulates the signallingcascade that is activated upon insulinbinding to its cell-surface receptor.

Specifically, PTP1B dephosphorylatestyrosine residues on the receptor,stopping insulin action. Knocking outPTP1B in mice improves insulinsensitivity over control mice [2]. Byselectively blocking PTP1B in humans,therefore, this halting mechanism mightprolong the effects of insulin. The PTPfamily of enzymes is large, however,and all are highly specific for thecharged phosphotyrosine residue.Finding a selective small-moleculeinhibitor of PTP1B is thus provingdifficult. Numerous PTP1B candidates

are undergoing trials, but big successeshave yet to emerge.

The tethering techniqueDrug discovery often uses HTStechnology to screen an enzyme activesite against thousands of molecules tofind those that bind with highestaffinity. For some targets, such ascertain protein–protein interactions,small-molecule binders can be difficultto find. Using molecular fragments isone way around the problem, but thefragments usually have low binding

DDT Vol. 8, No. 13 July 2003 updatenews

1359-6446/03/$ – see front matter ©2003 Elsevier Science Ltd. All rights reserved. PII: S1359-6446(03)02770-3 561www.drugdiscoverytoday.com

A tale of two necessities: breakawaytechnology versus diabetesMatt Brown, matt.brown@elsevier.com

Figure 1. An overview of thebreakaway tethering technique.A distal amino acid is mutatedto cysteine, which is thenattacked by a cleavableextender molecule. A moietywith affinity for the active site(shown in red) sterically protectsthe active-site cysteine ofprotein tyrosine phosphatase-1B(PTP1B) from alkylation by theextender. Once the protein hasbeen modified, the extender iscleaved, re-exposing the activesite and preparing the terminalportion of the extender for thiolexchange. Screening of a libraryof sulfur-containing compoundswill then reveal groups that areable to bind with the nearbyactive site and would thereforemake interesting leads for drugdevelopment. Figure providedby Daniel Erlanson of Sunesis Pharmaceuticals(http://www.sunesis.com/).

Drug Discovery Today

SS

SHSHProtein

Br

SProtein

SHSProtein

SProtein

Modify protein with extender

Use tethering to identify novel pharmacophores

Cleave extender

Active site

Distal cysteine

SH

affinities and can be difficult to pick outagainst ‘background noise’. To tacklethis problem, Sunesis Pharmaceuticals(http://www.sunesis.com/) havedeveloped a ‘tethering’ strategy, whichuses simple, yet elegant, chemistry [3]and relies upon the formation of adisulfide bond between the ligand anda cysteine residue on the protein; thismust be within or close to the activesite and can be natural or introducedby mutation. The protein is thenscreened against a library of sulfur-containing compounds under partiallyreducing conditions. Rapid thiolexchange occurs at the cysteine site aseach ligand forms a transient disulfidebond before being reduced. However,after disulfide formation, a few ligandswill show affinity for the nearby activesite. Even if this affinity is weak, thedisulfide bond will be entropicallystabilized and the tethered compoundcan be identified by MS. This techniqueis particularly good at finding low- andmoderate-affinity binders that routineHTS would miss. Stitching togetherseveral such low-affinity binders mightthen lead to a higher-affinity andselective drug.

With PTP1B, however, the active siteis deep and highly conserved;

introducing cysteine mutations into theactive site would probably disrupt theprotein’s function and lead toinaccurate binding studies. Aningenious way around this problem isto make the cysteine modification awayfrom the active site. A bridgingstructure can then be used to reachround to the active site (Fig. 1). Thebridge terminates with a thiol group,enabling screening with a library ofsulfur-containing compounds. Afterscreening a library of 15,000 disulfide-containing fragments, a high-potencycompound was discovered that ischemically distinct from previous PTP1Binhibitors. The binding mode, asassessed by X-ray crystallography(Fig. 2), was also found to differ fromthose of other PTP1B inhibitors.

Principle proven, partnershipspossibleThe breakaway tethering technique ispart of a suite of complementarytechnologies developed by Sunesis. Theoriginal methodology and other spin-offs have yielded several successes,including a highly potent molecule thatblocks the IL-2/IL-2Rα protein–proteininteraction. According to DanielErlanson, first author of the paper, the

technique holds great promise forfinding inhibitors of proteins that areintangible by traditional HTS. ‘Besidesthe published examples, this approachhas rapidly yielded compounds withpotencies as high as single-digit-nanomolar against importanttherapeutic targets in in-houseprograms’, he said. The novelty of theirapproach is generating interest amongpharmaceutical companies. Sunesis areactively pursuing partnerships todevelop the diabetes drug, and otherinhibitors generated by their technique.

Rob Hooft van Huijsduijnen (Serono;http://www.serono.com/) an expert onPTP1B inhibitors who is not connectedwith this study, is impressed by thetechnology as a proof-of principle. Hecommented that covalently linkingproteins to the enzyme substrate alsohelps co-crystallization and rationaldrug design. However, he was lessenthusiastic about the inhibitorsthemselves. ‘The compounds shown donot differ much from some of thecompetition,’ he added. ‘The road todrugs suitable for the clinic is still verylong.’

References1 Erlanson, D.A. et al. (2003) Discovery of a

new phosphotyrosine mimetic for PTP1Busing breakaway tethering. J. Am. Chem. Soc.125, 5602–5603

2 Elchebly, M. et al. (1999) Increased insulinsensitivity and obesity resistance in micelacking the protein tyrosine phosphatase-1Bgene. Science 283, 1544–1548

3 Erlanson, D.A. et al. (2000) Site-directedligand discovery. Proc. Natl. Acad. Sci. U. S. A.97, 9367–9372

update news DDT Vol. 8, No. 13 July 2003

562 www.drugdiscoverytoday.com

Figure 2. X-raycrystallographicstructure of a novelpTyr mimetic boundto protein tyrosinephosphatase-1B(PTP1B) at 2.4 Åresolution. PDB code1NWL. Figureprovided by DanielErlanson of SunesisPharmaceuticals(http://www.sunesis.com/).

Access Drug Discovery Todayonline at:

http://www.drugdiscoverytoday.com