The 3C-like protease (3CL(pro)) of severe acute respiratory syndrome-associated coronavirus (SARS-CoV) is one of the most promising targets for discovery of drugs against SARS, because of its critical role in the viral life cycle. In this study, a natural compound called quercetin-3-beta-galactoside was identified as an inhibitor of the protease by molecular docking, SPR/FRET-based bioassays, and mutagenesis studies. Both molecular modeling and Q189A mutation revealed that Gln189 plays a key role in the binding. Furthermore, experimental evidence showed that the secondary structure and enzymatic activity of SARS-CoV 3CL(pro) were not affected by the Q189A mutation. With the help of molecular modeling, eight new derivatives of the natural product were designed and synthesized. Bioassay results reveal salient features of the structure-activity relationship of the new compounds: (1) removal of the 7-hydroxy group of the quercetin moiety decreases the bioactivity of the derivatives; (2) acetoxylation of the sugar moiety abolishes inhibitor action; (3) introduction of a large sugar substituent on 7-hydroxy of quercetin can be tolerated; (4) replacement of the galactose moiety with other sugars does not affect inhibitor potency. This study not only reveals a new class of compounds as potential drug leads against the SARS virus, but also provides a solid understanding of the mechanism of inhibition against the target enzyme.

Density functional theory (DFT) calculations were performed at the B3LYP/6-311++G(d,p) level to systematically explore the geometrical multiplicity and binding strength for the complexes formed by alkaline and alkaline earth metal cations, viz. Li + , Na + , K + , Be 2+ , Mg 2+ , and Ca 2+ (M n+ , hereinafter), with nucleobases, namely, adenine, cytosine, guanine, thymine, and uracil. Morokuma decomposition and orbital analysis were used to analyze the binding components. A total of 150 initial structures were designed and optimized, of which 93 optimized structures were found, which could be divided into two different types: cation-π complex and cation-heteroatom complex. In the former, a M n+ is located above the nucleobase ring, while in the latter a M n+ directly interacts in flank with the heteroatom(s) of a nucleobase. The strongest binding of -319.2 kcal/mol was found in the Be 2+ -guanine complex. Furthermore, the planar ring structures of the nucleobases in some cation-π complexes were deformed, destroying more or less the aromaticity of the corresponding nucleobases. In the cation-heteroatom complex, bidentate binding is generally stronger than unidentate binding, and of which the bidentate binding with five-membered ring structure has the strongest interaction. Moreover, the calculated Mulliken charges showed that the transferred charge is linearly proportional to the binding strength. Molecular orbital coefficient analysis indicated a significant orbital interaction in cation-π complex, but not in cation-heteroatom interaction. In addition, Morokuma decomposition revealed that electrostatic interaction is more important for cation-heteroatom binding. The majority of the calculated ∆H values are in good agreement with the experimental results. In those cases with significant differences, the experimental results are proximate to an average of the ∆H values of two isomers formed by the same nucleobase and cation.

DFT/B3LYP calculations were carried out on complexes formed by NH 4 + with aromatics, viz. benzene, phenol, pyrrole, imidazole, pyridine, indole, furane, and thiophene, to characterize the forces involved in such interactions and to gain further insight into the nature and diversity of cation-aromatic interactions. Such calculations may provide valuable information for understanding molecular recognition in biological systems and for force-field development. B3LYP/6-31G** optimization on 35 initial structures resulted in 11 different finally optimized geometries, which could be divided into three types: NH 4 + -π complexes, protonated heterocyclic-NH 3 hydrogen bond complexes, and heterocyclic-NH 4 + hydrogen bond complexes. For NH 4 + -π complexes, NH 4 + always tilts toward the carbon-carbon bond rather than toward the heteroatom or the carbonheteroatom bond. The calculated CHelpG charges suggest that the charge distribution of a free heterocyclic may be used to predict the geometry of its complex. Charge population and electrostatic interaction estimations show that the NH 4 + -π interaction has the largest nonelectrostatic interaction fraction (∼47%) of the total binding energy, while the NH 4 + -aromatic hydrogen bond interaction has the largest electrostatic fraction (∼90%). A good correlation between binding energy and electrostatic interaction in the NH 4 + -π complexes is found, which shows that nonelectrostatic interaction is important for cation-π binding. The results calculated with basis sets from 6-31G to 6-311++G(2df, 2dp) show that ∆E corr and ∆H corr do not require a basis-set superposition error (BSSE) correction, in view of experimental error, if a larger basis set is used in the calculation. The calculated ∆H corr values for the NH 4 + -C 6 H 6 complex with different basis sets suggest that the experimental ∆H may be overestimated.