SUMMARYWe compared the anti-poliovirus activities of three flavones, quercetin, luteolin and 3-methylquercetin, which differ only at ring position 3. 3-Methylquercetin was the most potent compound. Quercetin exhibited antiviral activity only when protected against oxidative degradation by ascorbate. The antiviral activity of luteolin was comparable to that of ascorbate-stabilized quercetin.Several 3-methoxy derivatives of quercetin (3,3',4',5,7-pentahydroxyflavone) extracted from various plants are known to exhibit antiviral activity. They are not virucidal, but interfere with the replication of several viruses, including poliovirus (Ishitsuka et al., 1982;Van Hoof et al., 1984;Castrillo et al., 1986;Vrijsen et al., 1987). An anti-poliovirus activity of quercetin itself, i.e. a direct virucidal effect as well as interference with plaque formation, has been shown recently (Kaul et al., 1985). However, these results are at variance with previous reports according to which quercetin was virucidal only to enveloped, but not to unenveloped viruses (B61adi et al., 1977;Veckenstedt & Pusztai, 1981).The discrepancies about the antiviral properties of quercetin may originate in the oxidative degradation of this flavone in aqueous solution (Nishinaga et aL, 1979;Rajananda & Brown, 1981). In this communication, the anti-poliovirus activity of quercetin was determined in the presence and absenceof the antioxidant ascorbate. Comparative experiments were also carried out with 3-methylquercetin and luteolin, which differ from quercetin only at ring position 3 (Fig. 1).The antiviral effect of quercetin, luteolin and 3-methylquercetin was tested by the plaque reduction assay (Fig. 2). Both luteolin (not previously reported to possess antiviral activity) and 3-methylquercetin were active; 40 ktM-luteolin or 10 ktu-3-methylquercetin reduced the plaque count by half. Quercetin was inactive, even at the highest concentration tested (100 p,M). However, in the presence of ascorbate it did cause plaque reduction and this enhancement of its activity reached saturation at an ascorbate concentration of 100 ~tM. D-and L-ascorbate were equivalent, whereas dehydroascorbate was inactive (Fig. 3). Ascorbate alone failed to show any antiviral activity.With maximum enhancement by ascorbate, 30 ~tM-quercetin caused 50 % plaque reduction; this figure is close to that previously noted for luteolin (Fig. 2b). The antiviral effect of luteolin OH HO~R~OH OH o
0000-8341 © 1988 SGM