The primary objective of this research is to determine a molecular level understanding of the biological mechanisms regulating folate-dependent enzymes, in particular dihydrofolate reductase (DHFR) and thymidylate synthase (TS). A secondary goal is to use this knowledge to facilitate the development of folate-dependent enzyme inhibitors as more effective anticancer drugs. To elucidate correlations between structure and biological activity, we will test our model relating antifolate potency and conformationl flexibility using the antifolate structures available in our data base to search for patterns of intra- and inter- molecular interactions and hydrogen bond directionality and compare these patterns with those observed for the enzyme- inhibitor complexes. Semiempirical molecular orbital and molecular mechanics calculations of the energetic and entropic ramification of these binding patterns will be carried out to assess the role of solvent and electrostatic effects. These goals will be achieved through (1) determination of the X-ray crystal structures of selected members from four antifolate classes (lipophilic diaminopyrimidines, soluble diamino s-triazines, lipophilic quinazolines, and side chain modified antifolate pteridines), (2) description of their molecular and electronic properties from empirical molecular mechanics and ab initio molecular orbital energy minimization techniques, and (3) testing of our models that describe the specific binding interactions at the active site of the enzyme-inhibitor complex. This study will provide data that can identify the stable conformational isomers of DHFR and TS inhibitors, delineate their differing degrees of flexibility, determine the influence of specific substituents upon conformation, and define the nature of the active site interactions in these inhibitors. Although methotrexate is the most widely used antifolate antineoplastic agent, its limitations include poor cellular uptake, failure to cross the blood-brain barrier and frequent occurrence of cell resistance. In contrast, lipophilic antifolates readily diffuse through cell membranes and cross the blood-brain barrier. However, their insolubility in physiological media and their central nervous system toxicity are major problems in their use. The recent discovery that polyglutamation of methotrexate enhances its cytotoxic action on a number of folate-dependent enzymes, in particular TS, thereby altering folate pools, suggests that polyglutamation plays an important role in enzyme selectivity and specificity of action. These factors further indicate that a detailed knowledge of these antifolates at the molecular level is required to gain insight into their mechanisms of action and to permit a more rational basis for anticancer drug design.
