Chemical & Engineering News

February 24, 1997

Copyright © 1997 by the American Chemical Society


Researchers continue to refine techniques for identifying potential drugs in "libraries" of small organic molecules

Stu Borman
C&EN Washington

Combinatorial chemistry--a set of techniques for creating a multiplicity of compounds and then testing them for activity--has been widely adopted by large and small drug discovery companies alike over the past few years.

The idea of combinatorial chemistry is to form large "libraries" of molecules en masse--instead of synthesizing compounds one by one, as has been done traditionally--and to identify the most promising "lead" pharmaceutical compounds by high-throughput screening of the libraries. The concept has swept through the drug and biotechnology industries like a tidal wave, leaving few, if any, companies untouched. It has been the focus of much academic research as well.

According to David K. Stone, head of biotechnology research at the investment bank Cowen & Co., New York City, "Virtually every organization engaged in pharmaceutical research and development is using some form of combinatorial technology, or claims to be, and a host of entrepreneurial companies have sprung up to supply the industry's demands for libraries, high-throughput screening, synthesis methods, hardware, software, and analytical tools." No drugs discovered combinatorially have yet been approved for marketing, although several are currently in clinical trials.

A synthetic organic chemist in the pharmaceutical industry used to be able to make about 100 compounds per year, at a cost of thousands of dollars per compound, says Mark L. Peterson, vice president at Advanced ChemTech Inc., Louisville. But the exploding number of potential drug targets emerging from molecular biology research such as the Human Genome Project, as well as the continuing drive to minimize health care costs, has created a need to synthesize more compounds and test them more rapidly for bioactivity. Combinatorial chemistry provides this faster and cheaper route to new medicines, says Peterson, whose company markets automated systems for solution- and solid-phase synthesis. He spoke at the fourth annual conference on Exploiting Molecular Diversity: Small Molecule Libraries for Drug Discovery, held in Coronado, Calif., earlier this month and organized by Cambridge Healthtech Institute, Newton Upper Falls, Mass.

Panlabs chemistry associates (from left) Devin Hendricks, Jennifer Poulos, and William McFee weigh reagents for combinatorial synthesis using weighing station specially engineered for high throughput.

"These are heady days for chemists involved in combinatorial chemistry," says Anthony W. Czarnik, senior director of chemistry at Irori Quantum Microchemistry, La Jolla, Calif. "Large pharmaceutical companies are, to a site, developing internal groups in this area. Perhaps more amazingly, biotech companies once comfortable with mainly biology staffs are hiring combinatorial chemists as the absolute requirement for chemistry in drug discovery becomes more apparent. Furthermore, the joy of discovering and inventing really useful new techniques has energized the community in a way and at a rate this generation of medicinal chemists has not experienced before."

Combinatorial chemistry was first conceived about 15 years ago--although it wasn't called that until the early 1990s. Initially, the field focused primarily on the synthesis of peptide and oligonucleotide libraries. H. Mario Geysen, distinguished research scientist at Glaxo Wellcome Inc., Research Triangle Park, N.C., helped jump-start the field in 1984 when his group developed a technique for synthesizing peptides on pin-shaped solid supports. At the Coronado conference, Geysen reported on his group's recent development of an encoding strategy in which molecular tags are attached to beads or linker groups used in solid-phase synthesis. After the products have been assayed, the tags are cleaved and determined by mass spectrometry (MS) to identify potential lead compounds.

In 1985, Richard A. Houghten, president of Torrey Pines Institute for Molecular Studies and chief scientific officer of Houghten Pharmaceuticals Inc., both in San Diego, developed a technique in which tiny mesh packets, or" tea bags," act as reaction chambers and filtration devices for solid-phase parallel peptide synthesis. Houghten later also discovered positional scanning, a technique for rapidly assessing the biological activity contributed by different functional groups in a compound. Houghten Pharmaceuticals is one of several companies with a combinatorial drug in clinical trials--a cytokine-regulating agent for cancer pain, obesity, and non-insulin-dependent diabetes.

The field's original predominant focus on peptide and oligonucleotide libraries began to change about five years ago, when groups such as those of chemistry professor Jonathan A. Ellman at the University of California, Berkeley, and organic chemist Sheila Hobbs DeWitt at Parke-Davis Pharmaceutical Research, Ann Arbor, Mich., developed the first combinatorial techniques for producing small organic molecules with molecular weights of up to 500 or so - the class of compounds from which drugs are most often found. The approach conceived by DeWitt and colleagues provides the basis for automated synthesizers marketed by a Parke-Davis spin-off, Diversomer Technologies Inc., Ann Arbor, where DeWitt is now vice president of technical development.

Researchers at Pharmacopeia Inc. process samples for encoded, split-synthesis combinatorial chemistry. Christine Kowalewski (above) uses a microscope and display to select beads for tag reading. Phil Ezzo (upper right) uses vacuum drying apparatus to prepare samples for assays. And Sue Zhang uses gas chromatographs equipped with autosamplers and electron capture detectors to decode tags.

An emerging trend in the combinatorial field, according to Peter L. Myers, chief scientific officer of the combinatorial drug discovery company CombiChem in San Diego, is a move away from very large and complex mixtures of compounds with an inevitable redundancy of information content. Instead, Myers sees an increased focus on libraries in which components are prepared individually ("one compound per well"), have relatively high purity, and are focused more rationally on the biological targets at hand.

Such a trend is exemplified in the philosophy of the new company PharmaQuest, San Diego. President Dale S. Dhanoa says PharmaQuest's approach involves "synthesis of diverse druglike molecules as pure, single compounds that can be rapidly identified in high-throughput screening as initial lead compounds."

The trend toward more targeted libraries is also seen in work by senior scientist Eric J. Martin and coworkers at Chiron Corp., Emeryville, Calif. "To minimize redundancy and improve screening efficiency," says Martin, "we initially developed computational techniques to design libraries with maximum structural diversity. We found lots of potent ligands, but in analyzing the properties of the libraries and hits, we concluded that many factors besides diversity impact good library design for drug discovery." Martin's group has developed a "sequential D-optimal design" algorithm that facilitates the design of libraries biased to accommodate the requirements of specific receptors and the physicochemical properties of orally available drugs.

Computer programs that achieve similar goals are being developed by several companies. "The focusing of libraries interactively for lead optimization is what many perceive as the next hurdle in exploiting combinatorial approaches for drug discovery," says Scott D. Kahn, director of life sciences marketing at Molecular Simulations Inc., San Diego. The company's software-based approach uses both structure-activity information from analog data and structural information from protein crystallography and nuclear magnetic resonance spectroscopy to help focus libraries.

ChemSpace, by Tripos, St. Louis, selects compounds for synthesis in terms of their collective diversity, their molecular properties, and how druglike their structures are. Chemical Design Ltd., Chipping Norton, Oxfordshire, England, has developed ChemIdea, a module for the Chem-X molecular modeling package that uses biological activity data to identify areas of unexplored "property space" and to suggest substituents likely to further enhance activity. MDL Information Systems, San Leandro, Calif., makes a number of programs directed at combinatorial applications, including MDL Central Library, which helps optimize combinatorial reactions to identify promising library members.

Researchers at ArQule Inc., Medford, Mass., are using computational analyses, crystallographic data, and diversity assessment techniques to produce combinatorial libraries designed to provide maximal structure-activity relationship (SAR) data for use in lead optimization. OntoDiversity--a program marketed by Ontogen Corp., Carlsbad, Calif.--helps design combinatorial libraries by computing the structural transformations undergone by the "reactive fragments" used to construct them. And Alanex Corp., San Diego, uses its LiBrain software to design maximally diverse exploratory libraries and to guide medicinal chemists through the succession of optimization steps required to produce viable drug candidates. The trend toward" smarter" libraries is very clear.

One of the most persistent debates in the field of combinatorial chemistry involves the relative advantages and disadvantages of solid-phase versus solution-phase combinatorial synthesis. Combinatorial compounds are created either by solution-phase synthesis or by producing compounds bound covalently to solid-phase particles. Solid-phase synthesis makes it easier to conduct multistep reactions and to drive reactions to completion, because excess reagents can be added and then easily washed away after each reaction step. But a much wider range of organic reactions is available for solution-phase synthesis, the technology used traditionally by most synthetic organic chemists, and products in solution can be more easily identified and characterized.

One key factor in favor of solid-phase synthesis, says Czarnik, is that it makes it possible to use split synthesis, a technique developed in 1982 by Árpád Furka, organic chemistry professor at Eötvös University, Budapest, Hungary, and currently senior research fellow at Advanced ChemTech. Split synthesis produces large support-bound libraries in which each solid-phase particle holds a single compound, or soluble libraries produced by cleavage of compounds from the solid support.

"To make 10,000 compounds via a three-step solution-phase method requires 10,000 beakers, vials, or wells, each used three times," says Czarnik." There's no avoiding this physical requirement. This necessitates at least 30,000 liquid-handling steps, and more likely double that many. Making the same 10,000 compounds on solid support in three steps can be accomplished using as few as the cube root of 10,000 beakers--about 22 beakers, each used three times. Some small multiple of 22 liquid-handling steps will be required."

Therefore, says Czarnik, "making lots of compounds using solution-phase methods is automation-intensive. The same project on solid phase can be accomplished without liquid-handling automation at all." But he believes that both solution-phase and solid-phase techniques "are useful--period. Each platform has strengths and weaknesses, depending on the application."

Chemistry professor George Barany of the University of Minnesota, Minneapolis, who specializes in solid-phase synthesis, says: "I still think for convenience that solid phase is going to be the way to go. The field is still on a learning curve. But certainly you can't do split synthesis in solution, and split synthesis gives you orders of magnitude more diversity. This requires on-resin screening techniques or cleavable linkers. The issue is finding the right linkers and the right supports to use."

Stephen W. Kaldor, head of combinatorial chemistry at Eli Lilly & Co., Indianapolis, and coworkers have developed a solution-phase "covalent scavenger" technology that retains some of the advantages of solid-phase chemistry while greatly reducing some of its disadvantages. The reaction product is formed by solution-phase synthesis, and unreacted starting material is then selectively removed from solution via covalent bond formation to a solid-phase-supported electrophile or nucleophile. Kaldor points out that a similar technique was reported much earlier for removal of reactive impurities from natural product mixtures by chemistry professor Jean Fréchet and coworkers at Louis Pasteur University, Strasbourg, France [Tetrahedron Lett., 21, 617 (1980)].

"The concept is very generalizable, and we have used it with success on dozens of drug discovery projects at Lilly," says Kaldor. For instance, his group used it to optimize an existing noncombinatorial lead compound, resulting in the discovery of LY334370, a selective serotonin receptor agonist that is currently in clinical trials as a central nervous system agent for treatment of migraine.

The combinatorial optimization procedure took less than six months, from identification of the initial lead to discovery of the clinical candidate--a process that would more typically take one or more years. Kaldor's group recently reported the use of covalent scavengers for expedited amine alkylations and acylations [Tetrahedron Lett., 37, 7193 (1996)] and for the discovery of novel leads in a whole-cell antirhinovirus assay [Bioorg. Med. Chem. Lett., 6, 3041 (1996)].

Related techniques have been developed independently by at least three other groups--those of John C. Hodges, director of exploratory chemistry at Parke-Davis Pharmaceutical Research; Daniel L. Flynn at the Combinatorial Chemistry/ Parallel Medicinal Chemistry Discovery Unit of Monsanto/Searle, St. Louis; and Robert W. Armstrong, chemistry professor at the University of California, Los Angeles, and director of chemistry at Amgen Inc., Thousand Oaks, Calif. Hodges and Flynn presented their techniques in January at the 2nd Winter Conference on Medicinal & Bioorganic Chemistry in Steamboat Springs, Colo.

After solution-phase synthesis of one molecule at a time, says Hodges, "the strategy for purification, if any, is trial and error. Chromatography is commonly a first resort since it usually works. It is time-consuming and expensive, however."

Hodges' technique, called polymer-supported quench, is designed to purify solutions at each stage of multistep synthesis, using either polymer-bound reagents or liquid-liquid extraction. "In rare cases we must do both," he says, "but that is still preferable to chromatography when running parallel arrays of 100 or more reactions." He estimates the cost of a single purification by the polymer-supported quench method to be about one-tenth that of a single purification by flash chromatography (a manual column chromatography technique commonly used by synthetic organic chemists) because of savings in solvent and labor.

Flynn and coworkers at Monsanto/Searle have also found that, after solution-phase synthesis, excesses of reagents with certain types of molecular functionality can be sequestered by incubating them with resins having complementary reactive groups. Simple filtration then affords solution-phase products free of the sequestered reactants.

Reagents that are not susceptible to purification in this manner can be made so by tagging. The reagents are encoded with artificial functional group tags of known identity. After solution-phase synthesis, Flynn and coworkers incubate the reaction mixtures with resins containing functional groups complementary to those of the reagent tags. If conditions are right, the resins sequester not only the original reagents but also their by-products.

Armstrong and coworkers at UCLA and Amgen have also developed a technique called resin capture, a solution-phase cleanup method similar to the covalent scavenger and polymer-supported quench concepts. The technique involves solution-phase chemistry for the first few steps in a synthetic sequence, capture of a solution-phase intermediate on a resin, subsequent solid-phase transformations, and then release of the product back into solution.

In addition, the researchers are developing high-yield combinatorial solution-phase syntheses based on multicomponent condensation reactions such as the Ugi reaction, conceived by Ivar Ugi of the Institute of Organic Chemistry & Biochemistry at the Technical University of Munich, Germany. The Ugi reaction is a high-yield process in which a ketone, aldehyde, or Schiff base is combined with an amine, an isonitrile, and an acid to yield a variety of products, depending on choice of reactants.

Repligen Director of Chemistry Alan R. Jacobson and his coworkers in Needham, Mass., are also using the Ugi reaction in solution--in this case to identify inhibitors of protein-protein and protein-carbohydrate interactions. They recently carried out parallel syntheses and assayed the products directly, without the need for cleanup, in a high-throughput screen for antagonists of protein-glycosaminoglycan binding. The screen yielded lead compounds for the development of tumor growth inhibitors and agents that block angiogenesis (new blood vessel growth).

Researchers at Isis Pharmaceuticals, Carlsbad, Calif., have developed SPSAF (solution-phase simultaneous addition of functionalities), a high-yield technique that permits the generation of libraries containing roughly equimolar amounts of all components--which can be difficult to achieve with solid-phase synthesis. The group recently identified a series of highly potent antimicrobial leads by using SPSAF to generate modified polyazaphanes.

Chemistry professor Dennis P. Curran of the University of Pittsburgh has developed a strategy involving use of polyfluorinated tags to facilitate solution-phase purifications [Science, 275, 823 (1997)]. In Curran's "fluorous synthesis" approach, organic molecules are made soluble in fluorocarbon solvents so they can be easily separated from excess reactants and by-products present after solution synthesis.

A starting material for combinatorial synthesis is first derivatized with a perfluorinated group. After solution synthesis has been carried out, a water-immiscible fluorocarbon solvent is added to the solution. Because the product is carrying the perfluorinated group, it is extracted preferentially into the fluorocarbon solvent. The perfluorinated group can then be removed to recover the purified product.

Curran believes the analysis of phases of products, reagents, and by-products could become standard practice in library synthesis. "In a future six-step synthesis," he says, "you might do the first three steps on a small-molecule substrate, then switch to the fluorous phase for the next two steps, and finally do the last one on the solid phase. The desired products would then zigzag through various phases in a planned path that was dictated by the reactions that you were doing, the kinds of reagents that were available, and the kinds of by-products and impurities that you expected at each stage."

Fluorous synthesis is conceptually similar to a method developed two years ago by chemistry professor Kim D. Janda and coworkers at Scripps Research Institute, La Jolla, Calif., in which the polymer polyethylene glycol monomethyl ether is used as a foundation for combinatorial assembly. The polymer is soluble in a variety of aqueous and organic solvents and can be precipitated out of solution by crystallization for purification purposes after a synthesis.

However, Curran says fluorous synthesis differs from Janda's technique at the purification, identification, and analysis stages. "He works largely with polymer techniques at these stages, while we work with small-molecule techniques," Curran explains. "As an example, we can analyze our compounds by HPLC [high-performance liquid chromatography] or characterize them by MS without ever cleaving the fluorous label. You can't do this easily with polymers."

Curran adds: "Our fluorous technique and Janda's polyethylene glycol method nicely illustrate the point that ... while there might seem to be a huge gulf between solid-phase and solution methods, there is in fact a continuum that is just beginning to be filled in. Janda's work is somewhere in the middle, but toward the 'solid' side. Ours is more over on the' solution' end. It gets hard to pigeonhole techniques like the polyethylene glycol method, fluorous synthesis, resin capture, covalent extractions--and therein lies their power."

Kaldor agrees that although solution- phase and solid-phase combinatorial chemistry might sometimes seem to represent two extremes, "it is not necessarily an either-or situation. Some of the best combinatorial chemistry methods that are currently under investigation involve hybrid approaches that use the best features of both."

Director of Medicinal Chemistry Mark J. Suto and coworkers at Signal Pharmaceuticals, San Diego, have found ion-exchange resins useful as reagents to remove reaction by-products and eliminate aqueous workup in solution-phase synthesis. For example, they use resin to absorb excess reagent in reactions of amines with acid chlorides to produce amides. The resin eliminates an extraction procedure that would otherwise be necessary to clean up the solution of amide product. Reactions in the presence of a weakly basic resin have produced products that were more than 97% pure, compared with an average 61% purity in the absence of resin.

However, the Signal Pharmaceuticals scientists also employ solid-phase synthesis and have developed a linker that makes it possible to add diversity to a molecule as it is cleaved from a solid support. One disadvantage of solid-phase synthesis is that a hydroxyl, amine, carboxyl, or other polar group must be present on a molecule to be able to attach it to a solid support. This is a potentially undesirable constraint on the structure of compounds synthesized on solid phase, because products retain the polar group even after they are cleaved from the support.

Several groups, including that of Ellman and coworkers, have devised "traceless" linkers that avoid this problem, because the linkers are removed completely from products during the cleavage process. For example, Ellman's group recently reported an acylsulfonamide linker that can be displaced by various nucleophiles to add diversity to a library.

Suto's group has developed an inert sulfur traceless linker that can likewise be substituted by a wide variety of nucleophiles to introduce additional diversity. The researchers currently are using this chemistry to prepare small heterocycle-based libraries on a solid support.

Many other companies avoid using either solution-phase or solid-phase combinatorial chemistry exclusively. For example, researchers at Chiroscience Ltd., Cambridge, England, are successfully using an ambidextrous solid- or solution-phase strategy to generate and optimize lead compounds with activity as matrix metalloproteinase inhibitors. These compounds have potential applications for treatment of cancer and inflammatory diseases.

Chemists at Arris Pharmaceutical Corp., South San Francisco, have developed an inexpensive and simple technique for carrying out aqueous extractions on libraries made by solution-phase synthesis. The desired product of a reaction is purified by eluting the reaction mixture in a commercial sample prep column, which selectively retains by-products.

But the company also has developed some innovative techniques for solid-phase synthesis, including react-and-release chemistry, in which reagents used to cleave products from the solid support are incorporated into the product. Arris researchers also have developed a sequential displacement strategy that makes it possible to isolate several different compounds from each solid-phase reaction chamber.

"Substoichiometric amounts of different cleaving reagents are sequentially reacted with a compound that is synthesized on a react-and-release type resin, and each product is individually eluted," explains Douglas A. Livingston, director of combinatorial chemistry at Arris. "The number of compounds available from a piece of equipment is routinely increased up to 10-fold using this technique."

Considerable effort is also being devoted to the development of analytical chemistry techniques capable of keeping pace with the demands of combinatorial chemistry.

In pharmaceutical companies, notes analytical scientist Robyn A. Rourick of Bristol-Myers Squibb Co., Wallingford, Conn., it's important to obtain stability, metabolic, pharmacokinetic, and toxicological data on drug candidates as soon as possible to aid in the selection, optimization, and overall development of new drugs. But the advent of larger numbers of combinatorial drug candidates--compared with the number that used to be produced by conventional synthesis--makes it extremely difficult for analytical research groups to keep up. Such groups "cannot simply increase resources 10-fold to assist in the evaluation of new drug candidates using traditional protocols," she says.

As a result, Rourick and coworkers are developing new analytical procedures for combinatorial drug profiling. They find that parallel processing by LC/ MS eliminates the need for repetitive analyses of single compounds. Traditional analysis protocols require scale-up, extraction, separation, and fractionation of complex mixtures, followed by spectroscopic analysis of each individual fraction. LC/MS profiling permits all these steps to be accomplished by a single instrument, resulting in increased efficiency and productivity.

For example, metabolic and stability data can be obtained by subjecting a combinatorial mixture to forced degradation and in vitro metabolism conditions and then analyzing the treated sample with LC/MS. Metabolites and degradation products will retain most of the substructures of the compounds from which they are derived, and the chemical ancestry of these components can be established by LC/MS/MS. Rourick and coworkers use this method to develop profiles of each library compound that can be referenced throughout the drug development cycle.

Cheryl D. Garr, project manager of synthetic and combinatorial chemistry at Panlabs Inc., Bothell, Wash., and coworkers have developed methods for high-throughput sample analysis of Optiverse screening libraries, which are marketed to other companies for use in their drug discovery programs. They have developed an LC/MS technique for library analysis in collaboration with Perkin-Elmer Sciex, Toronto, and a preparative-scale HPLC process for purifying the libraries in conjunction with the Biotage division of Dyax Corp., Charlottesville, Va. High-throughput HPLC instrumentation developed by companies such as Biotage are capable of purifying hundreds of samples per day.

Automated tools for combinatorial chemistry include Chiron's multiwash head system (far left), which generates large numbers of compounds simultaneously, and a flash chromatography purification workstation, used by chemistry associate Julie Stripes of Panlabs.

In combinatorial chemistry, as in opera, it isn't over until the fat lady sings--meaning usually that a lead compound is in the drug development pipeline. The field has now progressed to a stage at which several drugs are in clinical trials, and others are well on the way.

For example, researchers at Pharmacopeia Inc., Princeton, N.J., are using encoded combinatorial libraries synthesized on solid phase to identify aspartyl protease inhibitors. Potential leads have emerged from libraries containing a total of about 32,000 compounds that were constructed around a statine core, using encoded split synthesis.

The encoding technique is based on a strategy--conceived by chemistry professor W. Clark Still and coworkers at Columbia University--in which inert halogenated compounds are used to record the chemical reaction history of each support bead. The tags are analyzed by capillary gas chromatography to reveal the identity of active compounds in the library.

"Pharmacopeia has optimized this method by automating the tag-reading process and combining the technology with high-throughput screening of targets of pharmaceutical interest," says research scientist Theodore O. Johnson Jr." Using four dual-channel gas chromatography instruments with electron capture detectors and autosamplers, between 1,000 and 1,500 tag decodes are routinely performed each month" in Pharmacopeia's research laboratories.

Johnson and coworkers have screened the encoded libraries against the aspartyl proteases plasmepsin and cathepsin D. Plasmepsin II is a malarial enzyme involved in the digestion of hemoglobin, and cathepsin D is an aspartyl protease involved in tumor metastasis and possibly Alzheimer's disease. The strategy yielded a family of selective plasmepsin II inhibitors, including one that inhibits plasmepsin II at the 100-nM level. The researchers also found compounds selective for cathepsin D, but they were not as potent.

Novartis' CGP64222, a lead compound for suppression of human immunodeficiency virus replication that was identified combinatorially, is shown docked to the binding site of TAR RNA.

Aspartyl protease inhibitors are also being sought by graduate student Ellen Kick and coworkers in Ellman's group at UC Berkeley, working with pharmaceutical chemistry professor Irwin D. Kuntz Jr. and coworkers at the University of California, San Francisco. The researchers constructed two 1,000-compound libraries. Dock, Kuntz's structure-based design program, was used to help select the building blocks that went into one of these libraries. The study yielded compounds that inhibit cathepsin D at low nanomolar levels.

Research scientist Eduard R. Felder and coworkers at Novartis (formerly with Ciba-Geigy's pharmaceuticals division), Basel, Switzerland, have identified a highly active and proteolytically stable hybrid peptoid-peptide oligomer, called CGP64222. (Peptoids are peptide analogs.) CGP64222 inhibits a key interaction in the human immunodeficiency virus (HIV) life cycle, thus suppressing viral replication.

TO SIDEBAR: 'Evolution' used to isolate high-affinity ligands

One of the first steps in HIV gene activation is the binding of HIV's Tat protein to TAR RNA, the Tat-responsive element of HIV. Starting from a pool of more than 3 million compounds synthesized on solid phase, the Novartis group was able to identify CGP64222 as an effective inhibitor of this interaction. The oligomer blocks formation of the Tat/TAR complex in vitro at nanomolar concentrations and inhibits HIV replication in human lymphocytes. Studies conducted in conjunction with researchers from the MRC Laboratory of Molecular Biology, Cambridge, England, show that CGP64222 binds TAR by a mechanism similar to that used by larger Tat-derived polypeptides. Felder and coworkers are currently studying simplified analogs of CGP64222 in an effort to identify compounds with enhanced in vivo efficacy.

Small-molecule (nonpeptide) inhibitors of Tat/TAR complex formation have also been identified by Research Associate Houng-Yau Mei and coworkers at Parke-Davis Pharmaceutical Research. The inhibitors were found by high-throughput screening of the corporate compound library.

At Versicor, South San Francisco, mechanism-based design elements are incorporated into the design of libraries. Screening of a 1,200-member quinazolinedione library has enabled Versicor scientists to identify several novel antimicrobial lead compounds that inhibit theß-lactamase class of bacterial enzymes. A focused combinatorial chemistry-driven SAR study based on these leads is currently in progress.

Molecumetics Ltd., Bellevue, Wash., is using its Smart Library technology to identify and optimize oral inhibitors of the clotting enzymes thrombin, Factor VIIa, and Factor Xa, and the transcription factor NF-KB. A Smart Library is a collection of nonpeptide small molecules designed to interact with therapeutic target proteins. The company is currently testing the thrombin inhibitor MOL-144 as an anticoagulant for treatment of cardiovascular diseases, and the NF-KB inhibitor MOL-294 as a potential asthma treatment.

Researchers also are continuing to attempt using combinatorial strategies for a nonbiological application--identifying inorganic catalysts (C&EN, Nov. 4, 1996, page 37). A recent example is work by chemistry professor Craig L. Hill and graduate student Robin Damico Gall of Emory University, Atlanta, in which polyoxometalates were prepared by parallel synthesis and evaluated for their ability to catalyze the aerobic oxidation of tetrahydrothiophene, an analog of the chemical warfare agent mustard gas [J. Mol. Catal., 114, 103 (1996)]. Previous techniques developed for the aerobic catalytic oxidation of thioethers such as tetrahydrothiophene require high temperatures and pressures, whereas one of the catalysts identified by Hill and Gall works under milder conditions (95 °C and 1.52 atm).

But most of the big guns in combinatorial chemistry remain focused on drug development. Over lunch, one attendee at the Coronado meeting expressed the view that the plethora of strategies being presented seemed overwhelming and somewhat confusing. "How can you tell which techniques work well?" she asked.

All of them work, to one degree or another, replied Houghten. They are all potentially valuable tools to help speed the identification of new drugs. If the drugs that result are safe and effective ones, that's all that really matters, he said.

ChemCenter Pubs Div. Home Page
Copyright © 1997 by the American Chemical Society.