The structure of the glycosomal glyceraldehyde-3-phosphate dehydrogenase (gGAPDH) from Trypanosoma cruzi complexed with chalepin, a natural product from Pilocarpus spicatus, has been determined by X-ray crystallography to 1.95 A î resolution. The structure is in the apo form without cofactors in the subunits of the tetrameric gGAPDH in the asymmetric unit. Unequivocal density corresponding to the inhibitor was clearly identified in one monomer. The final refined model of the complex shows extensive conformational changes when compared with the native structure. The mode of binding of chalepin to gGAPDH and its implications for inhibitor design are discussed. ß

Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) catalyzes the reversible oxidative phosphorylation of d-glyceraldehyde 3-phosphate (GAP) into d-glycerate 1,3-bisphosphate (1,3-diPG) in the presence of NAD(+) and inorganic phosphate (P(i)). Within the active site, two anion-binding sites were ascribed to the binding of the C3 phosphate of GAP (P(s)) and to the binding of the attacking phosphate ion (P(i)). The role played by these two sites in the catalytic mechanism in connection with the functional role of coenzyme exchange (NADH-NAD(+) shuttle) has been investigated by several studies leading to the C3 phosphate flipping model proposed by Skarzynski et al. [Skarzynski, T., Moody, P. C., and Wonacott, A. J. (1987) J. Mol. Biol. 193, 171-187]. This model has not yet received direct confirmation. To gain further insight into the role of both sites, we synthesized irreversible inhibitors which form with the essential cysteine residue a thioacyl enzyme analogue of the catalytic intermediate. Here we report the refined glycosomal Trypanosoma cruzi GAPDH in complex with a covalently bound GAP analogue at an improved resolution of 2.0-2.5 A. For this holo-thioacyl enzyme complex, a flip-flop movement is clearly characterized, the change from the P(i) to the P(s) binding site being correlated with the coenzyme exchange step: the weaker interaction of the intermediate when bound at the P(s) site with the cofactor allows its release and also the binding of the inorganic phosphate for the next catalytic step. This result gives strong experimental support for the generally accepted flip-flop model of the catalytic mechanism in GAPDH.

Crude extracts of Rutales species were tested in vitro against the trypomastigote form of Trypanosoma cruzi at 4 mg/mL, and 20% of them showed significant activity (80%). Their inhibitory activity against the glycolytic enzyme GAPDH from T. cruzi has also been evaluated at the concentrations of 100 and 200 mg/mL. Additionally, the inhibitory activity of 13 purified coumarins were also assayed against T. cruzi-GAPDH. Chalepin was the most active substance with IC50 = 64 mM. The 3D structure of the complex chalepin-enzyme was elucidated by X-ray crystallography, revealing the architecture of the interactions between the inhibitor and the enzyme active site.

Drug design strategies based on Comparative Molecular Field Analysis (CoMFA) have been used to predict the activity of new compounds. The major advantage of this approach is that it permits the analysis of a large number of quantitative descriptors and uses chemometric methods such as partial least squares (PLS) to correlate changes in bioactivity with changes in chemical structure. Because it is often difficult to rationalize all variables affecting the binding affinity of compounds using CoMFA solely, the program GRID was used to describe ligands in terms of their molecular interaction fields, MIFs. The program VolSurf that is able to compress the relevant information present in 3D maps into a few descriptors can treat these GRID fields. The binding affinities of a new set of compounds consisting of 13 coumarins, for one of which the three-dimensional ligand-enzyme bound structure is known, were studied. A final model based on the mentioned programs was independently validated by synthesizing and testing new coumarin derivatives. By relying on our knowledge of the real physical data (i.e., combining crystallographic and binding affinity results), it is also shown that ligand-based design agrees with structure-based design. The compound with the highest binding affinity was the coumarin chalepin, isolated from Rutaceae species, with an IC50 value of 55.5 microM towards the enzyme glyceraldehyde-3-phosphate dehydrogenase (gGAPDH) from glycosomes of the parasite Trypanosoma cruzi, the causative agent of Chagas' disease. The proposed models from GRID MIFs have revealed the importance of lipophilic interactions in modulating the inhibition, but without excluding the dependence on stereo-electronic properties as found from CoMFA fields.