The State of Combinatorial Chemistry

Combinatorial chemistry refers to a technique by which large chemical libraries can be generated by connecting together a diverse set of appropriate chemical building blocks in a systematic way. The ability to generate such large, chemically diverse libraries, either in a combinatorial fashion or by any other high throughput parallel synthetic method, combined with high throughput screening techniques, provides an immensely powerful tool for drug lead discovery and optimization. Initial efforts in combinatorial research have focused on oligomeric libraries such as peptide, N-alkyl glycine, polycarbamate and polyureas. However, due to the poor oral bioavailability and rapid clearance associated with these oligomeric compounds, these libraries have not proven particularly attractive as a source of drug leads. Recent efforts have been directed towards the synthesis of more drug-like small molecule libraries, mostly on solid supports but also in solution.

Small substituted heterocycles are an exceptionally diverse class of compounds, showing a wide spectrum of biological activity. These include antibacterials, antifungals (thiazolidinones, b-lactams), antihistamines and anti-inflammatory compounds (thiazolidinones), antihypertensives (dihydropyridines, pyrrolidines), antiepileptics (hydantoins), anxiolytics and reverse transcriptase inhibitors (benzodiazepines). The usefulness of heterocycles as scaffolds for library generation has not gone unnoticed. A wide variety of heterocyclic compound libraries have been synthesized by solid phase methods (Thompson and Ellman, Chem. Rev., 1996, 96, 555-600; Fruchtel and Jung Angew., Chem. Int. Ed. Engl., 1996, 35, 17-42). These include benzodiazepins, pyrrolidines, hydantoins, 1,4-dihydropyridines, isoquinolinones, diketopiperazine, benzylpiperazines, quinolones, dihydro- and tetrahydroisoquinolines, 4-thiazolidinones, b-lactams, benzisothiazolones, pyrroles and imidazoles. Some solution phase heterocylic libraries have also been generated (Holmes, C. P. et. al., J. Org. Chem.,1995, 60, 7328-7333). Many biologically active compounds have been discovered from these libraries, including a potent tyrosine kinase inhibitor from a benzodiazepin library, a neurokinin-2 receptor antagonist from a diketopiprazine library, and an angiotensin-converting enzyme inhibitor from a pyrrolidine library, to name a few.

Given the large number of potential heterocyclic templates, how can we evaluate the most appropriate scaffolds for the problem at hand? How can we make use of the pseudo dilution effect on solid phase to carry out heterocyclic chemistry that is difficult or impossible to do in solution? Given the demonstrated usefulness of NMR and IR methods on resin bound compounds, would solid phase be a good medium to investigate heterocyclic reaction mechanisms?

Finally, the use of combinatorial chemistry is not limited to medicinal chemistry. For example, a pyrrolidinemethanol library has been generated by Liu and Ellman for asymmetric catalyst development and optimization (J. Org. Chem., 1995, 60, 7712-7713). To what extent would such non-therapeutically oriented heterocyclic combinatorial libraries find utility in other areas of science?

Comments were added using this form during the conference sessions.

jim snyder, accepted 10/Jun/1996

Is it reasonable to expect that combinatorial libraries will provide genuine drug candidates, or, in the short term must we be satisfied with leads only?

Michal Lebl, accepted 03/Jul/1996

Just a comment about the existence of 'Diversity Lovers' Forum' which is a part of the home page of the journal 'Molecular Diversity' (http://vesta.pd.com). You are welcome to submit your comments and questions there. You can also check the list of articles published in this field.