We have proposed a cooperative quinolone-DNA binding model for the inhibition of DNA gyrase. The essential feature of the model is that bound gyrase induces a specific quinolone binding site in the relaxed DNA substrate in the presence of ATP. The binding affinity and specificity are derived from two unique and equally important functional features: the specific conformation of the proposed single-stranded DNA pocket induced by the enzyme and the unique self-association phenomenon (from which the cooperativity is derived) of the drug molecules to fit the binding pocket with a high degree of flexibility. Supporting evidence for and implications of this model are provided.

Norfloxacin is a nalidixic acid analogue and one of the most potent DNA gyrase inhibitors. To study the mechanism of this important class of inhibitors, the binding of [3H]norfloxacin to gyrase and substrate DNA was measured.We found that, contrary to prior belief, norfloxacin does not bind to gyrase but instead binds to DNA. This was demonstrated by both equilibrium dialysis and membrane filtration techniques. Binding to ColEl and pBR322 plasmids showed a primary process that is saturated at a norfloxacin concentration about equal to its supercoiling K; (1.8 x 10-6 M) and is followed by weaker secondary binding. The apparent Kd values are 1 x 10-6 M for both plasmids. The molar binding ratio at this initial saturation point is extremely low: only 4 x 10-4 norfloxacin per nucleotide for both plasmids. The binding of norfloxacin to DNA plasmids is nonintercalative, as shown by the fact that the drug binds preferentially to single-stranded DNA rather than to double-stranded DNA. The binding is reduced at high salt concentration, has a pH optimum between 4.5 and 6.5, and does not require divalent ions. The binding affinities of other nalidixic acid analogues were estimated by an indirect competition method. The calculated apparent Kd values of these analogues correlate well with their K1 values, providing strong evidence that the binding affinity of the drug to DNA determines biological potency.Nalidixic acid was the first member of the quinolone family of antibacterial agents synthesized (1). Subsequent developments have led to increasingly potent derivatives: pipemedic acid, oxolinic acid, and norfloxacin. The minimum inhibitory concentration of norfloxacin of 0.2 ,uM against Escherichia coli is 2 orders of magnitude lower than that of nalidixic acid, and norfloxacin has been shown effective in treatment of urinary tract infections. Its potency makes it a suitable choice for studying the mechanism of action of quinolone antibiotics.The antibacterial activity of this drug family is due to inhibition of DNA synthesis. The drugs inhibit DNA gyrase, a topoisomerase that negatively supercoils DNA in a reaction driven by ATP hydrolysis (for review, see refs. 2 and 3). The enzyme has two subunits, A and B, and it is believed that subunit A is the direct target for the drug. There are three lines of evidence for this conclusion. (i) Quinolone antibacterials are highly specific inhibitors of DNA gyrase. Even other topoisomerases are not inhibited at all or only at orders of magnitude higher concentrations (4-6). (ii) Mutations leading to high-level drug resistance are exclusively in gyrA, the structural gene for subunit A (7). Subunit A purified from the mutant plus wild-type subunit B reconstitutes quinoloneresistant gyrase. Because the minimum inhibitory concentration is raised up to 100-fold in gyrA mutants, no other vital target in E. coli can be important. (iii) Quinolone antibiotics inhibit reactions of DNA gyrase such as supercoiling and relaxation that require DNA breakage and reunion, the active site of w...

Although the functional target of quinolone antibacterials such as nalidixic acid and norfloxacin has been identified as the enzyme DNA gyrase, the direct binding site of the drug is the DNA molecule [Shen, L. L., & Pernet, A. G. (1985) Proc. Natl. Acad. Sci. U.S.A. 82, 307-311]. As described in this paper, binding specificity and cooperativity of quinolones to DNA were further investigated with the use of a variety of DNA species of different structures and different base compositions. Results show that the drug binding specificity is controlled and determined largely by the DNA structure. The drug binds weakly and demonstrates no base preference when DNA strands are paired. The drug binds with much greater affinity when the strands are separated, and consequently, binding preference emerges: it binds better to poly(G) and poly(dG) over their counterparts including poly(dI). The results suggest that the drug binds to unpaired bases via hydrogen bonding and not via ring stacking with DNA bases. The weak binding to relaxed double-stranded DNA and the stronger binding to single-stranded DNA are both nonspecific as they do not demonstrate binding saturation and cooperativity. The specific type of binding, initially demonstrated in our previous publication with the supercoiled DNA and more recently with complex formed between linear DNA and DNA gyrase [Shen, L. L., Kohlbrenner, W. E., Weigl, D., & Baranowski, J. (1989) J. Biol. Chem. (in press)], occurs near the drug's supercoiling inhibition concentration. As shown in this paper, binding saturation curves of this type are highly cooperative (with Hill constant greater than 4).(ABSTRACT TRUNCATED AT 250 WORDS)

A series of novel arylfluoroquinolones has been prepared. These derivatives are characterized by having a fluorine atom at the 6-position, substituted amino groups at the 7-position, and substituted phenyl groups at the 1-position. Structure-activity relationship (SAR) studies indicate that the in vitro antibacterial potency is greatest when the 1-substituent is either p-fluorophenyl or p-hydroxyphenyl and the 7-substituent is either 1-piperazinyl, 4-methyl-1-piperazinyl, or 3-amino-1-pyrrolidinyl. The electronic and spatial properties of the 1-substituent, as well as the steric bulk, play important roles in the antimicrobial potency in this class of antibacterials. As a result of this study, compounds 45 and 41 were found to possess excellent in vitro potency and in vivo efficacy.

The spread of antibiotic resistance in Gram-negative bacteria has sustained a continuing search for new agents with antibacterial activity against this important class of bacterial pathogen. Because the biosynthesis of lipopolysaccharide (LPS) is unique to Gram-negative bacteria and required by them for growth and virulence, attempts have been made to discover or design antibacterial agents acting at this site; however, no such agents have so far been developed. We now present definitive experimental data documenting design of the first member of the class of antibacterial compounds which specifically inhibit LPS synthesis. The target enzyme is 3-deoxy-D-manno-octulosonate cytidylytransferase (CMP-KDO synthetase), a cytoplasmic enzyme which activates 3-deoxy-D-manno-octulosonate (KDO) for incorporation into LPS. A specific inhibitor of CMP-KDO synthetase, alpha-C-(1,5-anhydro-7-amino-2,7-dideoxy-D-manno-heptopyranosyl)-carboxy late was designed using results of our studies of the purified enzyme. LPS synthesis ceased and lipid A precursor accumulated, causing growth stasis and perturbation of outer membrane structure and function, following delivery of the inhibitor to the intracellular target by a peptide carrier. Antibacterial action required an intact oligopeptide permease system and specific intracellular aminopeptidase activity to release inhibitor from the peptide prodrug.