Binding of natural and synthetic polyphenols to human dihydrofolate reductase

Abstract

Dihydrofolate reductase (DHFR) is the subject of intensive investigation since it appears to be the primary target enzyme for antifolate drugs. Fluorescence quenching experiments show that the ester bond-containing tea polyphenols (-)-epigallocatechin gallate (EGCG) and (-)-epicatechin gallate (ECG) are potent inhibitors of DHFR with dissociation constants (K(D))of 0.9 and 1.8 microM, respectively, while polyphenols lacking the ester bound gallate moiety [e.g., (-)-epigallocatechin (EGC) and (-)-epicatechin (EC)] did not bind to this enzyme. To avoid stability and bioavailability problems associated with tea catechins we synthesized a methylated derivative of ECG (3-O-(3,4,5-trimethoxybenzoyl)-(-)-epicatechin; TMECG), which effectively binds to DHFR (K(D) = 2.1 microM). In alkaline solution, TMECG generates a stable quinone methide product that strongly binds to the enzyme with a K(D) of 8.2 nM. Quercetin glucuronides also bind to DHFR but its effective binding was highly dependent of the sugar residue, with quercetin-3-xyloside being the stronger inhibitor of the enzyme with a K(D) of 0.6 microM. The finding that natural polyphenols are good inhibitors of human DHFR could explain the epidemiological data on their prophylactic effects for certain forms of cancer and open a possibility for the use of natural and synthetic polyphenols in cancer chemotherapy.

Keywords: antifolates; dihydrofolate reductase; enzyme inhibition; flavonoids; polyphenols; tea catechins.

Figure 1.

Chemical structures of classical and non-classical antifolates compared with natural and synthetic polyphenols. Abbreviations: MTX, methotrexate; TQD, (R)-6-{[methyl-(3,4,5-trimethoxyphenyl)-amino]methyl}-5,6,7,8-tetrahydroquinazoline-2,4-diamine; EC, (-)-epicatechin; EGC, (-)-epigallocatechin; ECG, (-)-epicatechin gallate; EGCG, (-)-epigallocatechin gallate; TMECG, 3-O-(3,4,5-trimethoxybenzoyl)-(-)-epicatechin; Qglc, quercetin-3-β-D-glucoside; Qxyl, quercetin-3-D-xyloside; Qrha, quercetin-3-rhamnoside; Qgal, quercetin-3-D-galactoside.

Figure 2.

Structural comparison of natural (ECG) and synthetic (TMECG) polyphenols with classical (MTX) and non-classical (TQD) antifolates.

Figure 3.

Titration fluorescence experiments for the binding of ECG and EC to human DHFR. In the left panel points are experimental (after correction for enzyme dilution) and the lines are best-fit theoretical curves. The enzyme concentration was 0.1 μM.

Figure 4.

Wall-eyed stereo view of ECG modeled into the folate-binding site of human DHFR. Carbon atoms of the ECG ligand and surrounding protein are colored green and grey respectively. Residue Phe-31, located behind the ECG, is unlabelled. Four different ligands from human and chicken DHFR crystal structures were used to define a binding envelope, shown in cyan; these were placed in a common orientation by superimposing backbone atoms from a common set of protein residues located around the ligands. Ligands from the following PDB structure files were used; 1DR1 (biopterin), 1S3V (TQD), 1S3W, and 1DLR. The figure was prepared using ViewerLite software.

Figure 5.

Titration fluorescence experiments for the binding of TMECG (□) and TMECG-QM (•) to human DHFR. Points are experimental (after correction for enzyme dilution) and the lines are best-fit theoretical curves. The enzyme concentration was 0.1 μM. For comparison the data obtained for ECG (▴) binding to human DHFR is included in the figure.

Figure 6.

Molecular modelling for the binding of TMECG and TMECG-QM to human DHFR. The protein is depicted in ribbon representation and colored by secondary structures (i.e., helix, strand, and loop). The lower panel represents the proposed structure for TMECG-QM and the hydrogen-bonding network of TMECG-QM in human DHFR. Residues hydrogen bonded directly to TMECG-QM are depicted in red and the predicted distances in Å are showed. W represents a water molecule.

Figure 6.

Molecular modelling for the binding of TMECG and TMECG-QM to human DHFR. The protein is depicted in ribbon representation and colored by secondary structures (i.e., helix, strand, and loop). The lower panel represents the proposed structure for TMECG-QM and the hydrogen-bonding network of TMECG-QM in human DHFR. Residues hydrogen bonded directly to TMECG-QM are depicted in red and the predicted distances in Å are showed. W represents a water molecule.

Figure 7.

Titration fluorescence experiments for the binding of quercetin glucuronides to human DHFR. Points are experimental (after correction for enzyme dilution) and the lines are best-fit theoretical curves. The enzyme concentration was 0.1 μM. (•) Qglc (□) Qgal (▴) Qrha (○) Qxyl.