Expanding the Chemical Approach
Combinatorial Chemistry: Combinatorial chemistry can be defined as the rapid synthesis or computer simulation of a large number of different but structurally related molecules. This methodology has its roots in research conducted in the 1960’s on solid-state peptide synthesis. The field gained wide recognition in the mid-1990’s, again largely for its role in the mass synthesis of peptide and oligonucleotide libraries. At about the same time, technological improvements in high-throughput screening (HTS) technologies that included robotics and highly parallel synthesis platforms led to the widespread availability of large sets of compounds for biological screening. Combinatorial chemistry has been strongly embraced by the pharmaceutical industry for synthesizing lead compounds in the drug discovery process. However, this approach was not initially adopted by many agrichemical industries for herbicide synthesis, mainly because the amounts of chemicals produced are too small for whole plant screening. This situation has changed in recent years as companies adopt miniaturized HTS systems and in vitro assays that require much smaller quantities of chemicals.
QSAR and Related Modeling Methods: QSAR (quantitative structure-activity relationship) is a computer modeling tool in which a chemical structure is quantitatively correlated with biological or chemical activity. For example, biological activity can be expressed as the concentration of a chemical required to cause a certain biological response. For herbicides, an LD50 value (the concentration that is lethal to 50% of a population of target plants) is often determined experimentally. Such values are then used to form a mathematical relationship, or quantitative structure-activity relationship, which is subsequently used to predict the biological response of other, related chemical structures. Recent advances in computing power and modeling software have allowed rapid advances in the power and precision of QSAR predictions, mostly in the pharmaceutical industry. This and related technologies are the basis of the term ‘designer drugs,’ which originally described drugs that were synthesized as analogs of existing drugs. More recently, the term has taken on the connotation of illicit or street drugs.
QSAR-based predictions are often not as straightforward as they may seem. For example, very small changes in a chemical structure may cause unpredictable effects on overall chemical behaviors like solubility, reactivity, and interactions with a biological target site. This phenomenon is termed the Structure-Activity Relationship (SAR) paradox. It is often more productive to use QSAR to detect trends of biological activity, rather than predict specific biological effects based on specific chemical configurations.
For herbicides, QSAR was used to synthesize and develop the sugar beet herbicide metamitron (Bayer AG) and the rice compound bromobutide (Sumitomo Chemical Company). In a post hoc analysis, QSAR successfully predicted the toxicity of various triazine herbicides based on their chemical and structural properties.
CoMFA: CoMFA (comparative molecular field analysis) methodology is a three-dimensional refinement of QSAR in which the shapes and other properties of molecules are spatially related to specific molecular features such as substituent groups. This technique allows molecular modifications to be made to lead compounds as based on their actual chemistry, in attempts to improve their biological activity. CoMSIA (comparative molecular similarity indices analysis) is a related three-dimensional design tool used to model the interactions between small molecules and proteins. Both techniques are widely used in the pharmaceutical industry and are being adopted to direct the design of potential herbicides.