A method for preparing a large number of chemical compounds, commonly known as a combinatorial library, which are then screened to identify compounds having a desired function, such as a particular biological or catalytic activity. Combinatorial synthesis is an aspect of combinatorial chemistry, which allows for the simultaneous generation and rapid testing for a desired property of large numbers of chemically related compounds. One could regard combinatorial chemistry as the scientist's attempt to mimic the natural principles of random mutation and selection of the fittest. Combinatorial chemistry has already become an invaluable tool in the areas of molecular recognition, materials science, drug discovery and optimization, and catalyst development. See also: Combinatorial chemistry; Organic synthesis
Combinatorial synthesis was developed to prepare libraries of organic compounds containing from a few dozen to several million members simultaneously, in contrast to traditional organic synthesis where one target compound is prepared at a time in one reaction (Fig. 1). Thus, a target compound AB, for example, would be prepared by coupling of the substrates A and B in a traditional (orthodox) synthesis and would be isolated after reaction processing (workup) and purification (such as crystallization, chromatography, or distillation). Combinatorial synthesis offers the potential to prepare every combination of substrates, type A1–m and type B1–n, providing a set of compounds, A(1–m)B(1–n). There are two main approaches to combinatorial synthesis: parallel synthesis and split synthesis (also known as split-and-pool or mix-and-split).
In parallel synthesis, each compound is prepared separately in an individual reaction vessel. In order to accomplish parallel combinatorial synthesis, large numbers of reactions must be carried out simultaneously and the resulting products must be isolated and purified very quickly. Recently, various automated devices, such as weighing, reaction, filtration, and chromatography machines, were developed for high-throughput parallel combinatorial synthesis. Another approach employs insoluble reagents that can be removed by filtration from a solution-phase reaction mixture. Acids, bases, reducing reagents, oxidation reagents, or various catalysts are attached to insoluble supports (such as silica-gel or polymers) in this approach. Functional groups bound to insoluble supports are also applied to scavenge unreacted starting materials in solution.
Transformation of a substrate anchored to an insoluble support by treatment with a solution of reagents, generally called solid-phase synthesis, is a frequently used protocol for combinatorial synthesis. A starting substrate is first attached to a solid support by an appropriate linker group and undergoes various synthetic transformations to furnish a product on the solid support which is readily separated from the unreacted reagents and coproducts in the solution by filtration. The resulting solid-supported product can then be used in subsequent transformation steps. Solid-phase peptide synthesis is a typical example. The final products, as well as the synthetic intermediates, are detached from the solid support on demand. An example of the solid-phase parallel synthesis giving a 3 × 3 = 9-membered library is shown in Fig. 2. This library can be screened for optimum “hit” candidates before or after release of the supported products; however, the preferred method for biological screening is to test the compounds in solution after cleavage from the solid support. A1–3 and B1–3 could be amino acids to give dipeptides, nucleotides to give dinucleotides, or sugars to give disaccharides. See also: Nucleotide; Peptide
In both protocols—solution-phase synthesis with solid-supported reagents and solid-phase synthesis with solution-phase reagents—various polymers, glass surfaces, and silica gel are typical solid supports; cross-linked polystyrene beads having appropriate functional groups (such as NH2, Cl, OH, and COOH) are the most frequently used.
Polystyrene beads bearing a reactive functional group are also used as a starting support in split synthesis. The beads range from 10 to 200 micrometers in diameter. For example, 1 g of beads having a 100-μm diameter contains about 1 million beads. The split synthesis is illustrated in Fig. 3 for a library of 3 × 3 × 3 = 27 possible combinations of trimeric products. Three steps of solid-phase synthesis to attach A, B, and C to the starting beads (step 1) are carried out individually to give the supported A, B, and C. The resulting three kinds of beads are suspended in a solvent, mixed by shaking, and filtered to give one portion of a mixture. The pooled mixture of beads is partitioned to three portions where each portion of beads contains A, B, and C in a statistical ratio (1:1:1). Similar manipulations (step 2) afford three portions of beads where each portion contains every combination of nine compounds (from AA to CC) in a statistical ratio. The three portions of a mixture consisting of nine compounds are subjected to coupling with X, Y, and Z (step 3) to give a nine-membered mixture having a terminal group X, a nine-membered mixture having a terminal group Y, and a nine-membered mixture having a terminal group Z, respectively. Finally, the resulting beads are all mixed together to give a library consisting of 27 compounds bound to the solid support, where three sets of three coupling reactions (3 × 3 = 9 manipulations) afford the cube of three (33 = 27) compounds. In this manner, split synthesis can offer an enormous number of combinatorial products. For example, linear coupling of 10 amino acids five times (10 × 5 = 50 manipulations) can furnish 105 pentapeptides. Since one bead of a library prepared by split synthesis must have one compound, the library is also referred to as the one-bead–one-compound library. Identification of the most active compound from the one-bead–one-compound library has gained in popularity with the advent of more efficient methods for one-bead spectroscopic analysis.