Combinatorial chemistry is a technology for creating a multitude of different compounds by reacting different combinations of interchangeable chemical "building blocks." The compounds are then screened for their ability to carry out a specified function, most commonly to act as drugs to treat a disease. Combinatorial chemistry allows the rapid synthesis and testing of many related compounds, greatly speeding the pace of drug discovery. Automated synthesis and screening systems are key to this approach.
The Combinatorial Approach
There are two general approaches for finding the correct answer to a question (besides asking someone who knows). One way is to learn everything relevant to the topic and then to use your knowledge to arrive logically at the answer. Scientists, and most other people, almost always use this method. A second approach is to keep guessing until you've guessed right! This seems like a foolhardy strategy, and usually is. What if it took a million guesses before you stumbled upon the right answer? But what if you could make a million or a billion guesses all at once? Through combinatorial chemistry, scientists can make and test millions, billions, or even quadrillions (1015) of guesses about which chemical compound might have a desirable function, such as the ability to bind to a specific molecule, or to serve as a drug.
Many chemicals are pieced together through combinations of smaller building blocks. For example, benzene is a chemical consisting of six carbon atoms connected in an aromatic ring structure, with a hydrogen atom bound to each carbon. Substituting one of the hydrogens with a hydroxyl (-OH) group forms the chemical phenol. Substituting a methyl (-CH3) group instead forms toluene, and substituting an amino (-NH2) group forms aniline. Because of their different "functional groups," or side groups, all of these compounds have very different physical and chemical properties. More variations can be synthesized by substituting additional side groups with more than one of the hydrogens. By substituting one of just these three groups (or by not adding any groups) for any of the six hydrogens in a benzene ring, there are 46, or 4,096, possible combinations (the number of different compounds is much smaller, because benzene is symmetrical, and many of the combinations represent equivalent structures).
Side groups can also be placed onto other side groups. For example, a single chlorine atom can substitute for one of the hydrogens of the methyl group in toluene to form benzyl chloride. By using a moderately sized collection of side groups, placing them onto a "scaffold" molecule that is more complex than benzene (such as cholesterol, which has three six-carbon rings and a five-carbon ring), and by using additional levels of side groups, combinatorial chemists can synthesize vast numbers of distinct but related compounds.
Although the utility of combinatorial chemistry was not fully appreciated by scientists until the 1980s, nature uses this strategy over and over. Genes, after all, are composed of different combinations of only four different nucleotides, and just twenty different amino acids form the building blocks of all proteins. In the immune systems of mammals, B lymphocytes use an elaborate scheme for mutating and combining different segments of antibody genes to generate a diverse pool of antibody molecules that can recognize and bind a wide array of alien molecules that enter the body with a pathogen infection.
Combinatorial chemistry is most often used to synthesize "small molecules," in contrast to macromolecules such as DNA, RNA, proteins, and polysaccharides, which are polymers containing long chains of monomer subunits. Because of their enormous size, macromolecules cannot easily enter cells, which is an important requirement for compounds intended for use as drugs.
In many cases the combinatorial chemist is looking for a compound that will bind tightly and specifically to a cellular molecule, such as the catalytic, or "active," site on the inside of an enzyme . Small molecules can fit into the holes and crevasses leading to the active site. By binding the enzyme, the synthetic compound may prevent it from binding to its natural substrate or from carrying out its catalytic reaction. Defective enzymes that resist normal cellular restraints on their activities are responsible for many diseases, including certain cancers. Chemical inhibitors of such rogue enzymes hold promise as powerful drugs. Alternatively, binding of a small molecule to an enzyme could enhance the enzyme's normal activity. Such molecules have potential as drugs for diseases caused by insufficient activity of a crucial enzyme.
For two molecules to bind to one another, they must have a proper fit, like a key in a lock. The fit depends on the shapes of the two molecules as well as on the chemical interactions between them. For example, two positively charged side groups will repel each other, but negative and positive groups can attract. Not surprisingly, a synthetic compound that binds a particular molecule often has chemical properties and a shape mimicking the natural ligand for the molecule. Such compounds are termed analogues.
When the drug target is known, its structure can be used as a template to create analogues with complementary shapes. Alternatively, if an analogue is already in hand that binds the target but has undesirable properties (such as weak binding, poor solubility, or serious side effects), this structure can be used as a starting point. Even without such clues, the speed and automation of the combinatorial approach makes it feasible to randomly synthesize and test millions of compounds.
A library of a billion or more different molecules is only useful if the molecules can be quickly and economically screened for the desired function. "High-throughput" techniques have been developed that automate most of the steps required to combine the molecules with their targets and evaluate the extent of any reaction.
Typically, the molecules are arrayed on a solid surface and the target is added. Unbound target is washed away. Fluorescent tags are often added to the target, to allow easy (and automated) visualization of the results. Robotic systems controlled by computers can react and evaluate billions of separate compounds in the time it would take a human to screen a dozen. One such approach is used in DNA microarrays, in which thousands of genes from a DNA library are attached to a solid base. These are reacted with messenger RNAs from a cell, and the results are visualized fluorescently.
In addition to its use in drug development, combinatorial chemistry can be applied to other areas of biomedical research, such as the design of molecules for diagnosing medical conditions. Compounds for these applications can be larger than pharmaceutical compounds, and do not have to be designed to enter the body. Using a novel combinatorial chemistry method called in vitro selection or SELEX (systematic evolution of ligands by exponential enrichment), a short DNA or RNA molecule (termed an oligonucleotide) with a desired property, such as the ability to specifically recognize and bind a molecule associated with a particular disease, can be selected in a single experiment from a library containing approximately 1015 different compounds. First, a library of oligonucleotides is created in a machine called an oligonucleotide synthesizer. This apparatus can make oligonucleotides with either a defined or random sequence.
Oligonucleotides for SELEX are designed to have a central region containing random sequence and outer, flanking regions with defined sequences. These defined sequences will be used as primer-binding sites for the polymerase chain reaction (PCR ). The oligonucleotide library is prepared as a mixture, usually containing about 1014 to 1015 different sequences. These specialized oligonucleotides, termed aptamers, are then exposed to target ligand molecules, which are typically attached to a solid support, such as a filter membrane. The unbound aptamers are then washed away, leaving only the rare aptamers that can bind the ligand adhering to the filter. These aptamers can then be recovered from the filter by washing it with a solution that disrupts the binding.
These binding candidate aptamers represent a minuscule fraction of the original library. Some may bind the target ligand tightly, but others may bind weakly. Since all the aptamers have defined primer-binding sites on the ends, this much-reduced population can now be amplified exponentially by PCR. After amplification, the aptamers can be subjected to another round of ligand binding, now using more stringent washing conditions, in which only the tightest-binding molecules will stay bound. These high-affinity binders can be recovered again subjected to still more cycles of PCR amplification, binding, washing, and recovery, until the population of aptamers consists exclusively of very tightly binding molecules.
For some applications, these molecules are useful directly. They can also be studied to design non-DNA molecules that have similar shapes but that will have more potential as drugs.
see also DNA Libraries; DNA Microarrays; High-Throughput Screening.
Paul J. Muhlrad
Alberts, Bruce, et al. Molecular Biology of the Cell, 4th ed. New York: Garland Science, 2002.
Borman, Stu. "Combinatorial Chemistry." Chemical and Engineering News 75, Feb. 24, 1997.
Combinatorial chemistry is a sophisticated set of techniques used to synthesize, purify, analyze, and screen large numbers of chemical compounds, far faster and cheaper than was previously possible. The direct precursor of combinatorial chemistry was the solid-phase synthesis of polypeptides developed by American biochemist Robert Bruce Merrifield in the 1960s, followed by the advances in laboratory automation since then. Initial development of the field has been led by the pharmaceutical industry in the search for new drugs, but its applications are spreading into other fields of chemistry. Other terms associated with this field are parallel array synthesis and high-throughput chemistry.
Whereas classical synthetic chemistry involves the stepwise synthesis and purification of a single compound at a time, combinatorial chemistry makes it possible to synthesize thousands of different molecules in a relatively short amount of time, usually without the intermediate separation of compounds involved in the synthetic pathway, and with a high degree of automation. Such procedures result in the production of new compounds faster and in greater numbers than is possible with standard synthetic methods. The first and still the most common type of combinatorial synthesis involves attaching a molecular species onto a macroscopic substrate such as a plastic bead and performing one or several well-characterized chemical reactions on the species. After each reaction, the product mixture can be split among several reaction containers and then recombined after the reaction (a procedure called mix and split ), or else carried out in parallel containers. The resulting mixture of compounds is referred to as a molecular library and can contain many thousands of individual compounds. The analysis, or screening, of these libraries to identify the compounds of interest, along with their subsequent isolation and identification, can be completed by a variety of methods. One example is iterative deconvolution ; it involves the successive identification of each of the units backward along the chain of synthesized units. Another, called positional scanning, requires the multiple synthesis of a library, each time varying the location of a known unit along the chain and comparing the activities of the resulting libraries. More recent advances in library screening involve the "tagging" of a substrate with tiny radio frequency transmitters or unique two-dimensional barcodes. Another important recent advance by researchers allows combinatorial syntheses to be carried out in solution, which further extends the scope and utility of this field.
Although the initial applications of combinatorial and high-throughput chemistry have occurred in the pharmaceutical field, the same techniques are now being used successfully to aid in the discovery of new catalysts, polymers, and high temperature superconductors.
see also Chemical Informatics.