In mammalian cells, DNA replication occurs at discrete nuclear sites termed replication factories. Here we demonstrate that DNA ligase I and the large subunit of replication factor C (RF-C p140) have a homologous sequence of~20 amino acids at their N-termini that functions as a replication factory targeting sequence (RFTS). This motif consists of two boxes: box 1 contains the sequence IxxFF whereas box 2 is rich in positively charged residues. N-terminal fragments of DNA ligase I and the RF-C large subunit that contain the RFTS both interact with proliferating cell nuclear antigen (PCNA) in vitro. Moreover, the RFTS of DNA ligase I and of the RF-C large subunit is necessary and sufficient for the interaction with PCNA. Both subnuclear targeting and PCNA binding by the DNA ligase I RFTS are abolished by replacement of the adjacent phenylalanine residues within box 1. Since sequences similar to the RFTS/PCNA-binding motif have been identified in other DNA replication enzymes and in p21 CIP1/WAF1 , we propose that, in addition to functioning as a DNA polymerase processivity factor, PCNA plays a central role in the recruitment and stable association of DNA replication proteins at replication factories.

The binding of plasmid DNA to norfloxacin, a quinolone antibacterial agent, was investigated by fluorescence, electrophoretic DNA unwinding, and affinity chromatography techniques. The amount of quinolone bound to DNA was modulated by the concentration of Mg2+. No interaction was evident in the absence of Mg2+ or in the presence of an excess of Mg2+, whereas maximum binding was observed at a Mg2+ concentration of 1-2 mM. The experimental data can be fitted to the formation of three types of Mg adducts: a binary adduct with norfloxacin and Mg2+, a binary adduct with DNA and Mg2+, and a ternary adduct with quinolone, plasmid, and Mg2+. We propose a model for the ternary complex, in which Mg acts as a bridge between the phosphate groups of the nucleic acid and the carbonyl and carboxyl moieties of norfloxacin. Additional stabilization may arise from stacking interactions between the condensed rings of the drug and DNA bases (especially guanine and adenine), which may account for the preference exhibited by quinolones for single-stranded and purine-rich regions ofnucleic acids. Other possible biochemical pathways of drug action are suggested by the observation that norfloxacin binds Mg2+ under conditions that are close to physiological.Conflicting literature reports have been accumulating on the role played by DNA in the mechanism of action of quinolone compounds. Although a large amount of biological data has indicated that DNA gyrase was the target for quinolone compounds (1-4), recent reports dismissed DNA gyrase as the target and pointed to DNA as the direct binding species (5). In fact, a cooperative interaction was proposed to occur between quinolones and supercoiled DNA. Subsequent publications by the same laboratory have modified this view extensively (6-8). In particular, Shen et al. (7) have proposed that in the presence of ATP bound gyrase induces a specific quinolone binding site in the relaxed DNA substrate. Gelelectrophoresis experiments by Tornaletti and Pedrini (9) showed that norfloxacin (Nor) is able to unwind the DNA double helix in the presence of Mg2+. On the other hand 19F NMR measurements failed to show any direct DNAquinolone interaction (10). We were also unable to detect binding using fluorescence spectroscopy techniques (11).Even if reconsidered in terms of affinity, the interaction with DNA is still of great concern because of the possible long-term genotoxicity of quinolone compounds, which are increasingly adopted as first-choice antibiotics for the treatment of many infections, and because it addresses the real mechanism of action of this class of molecules. To shed some light on this cumbersome problem, we have focused our attention on the role of Mg2+ in the binding of the model quinolone drug Nor to plasmid DNA. Our approach includes fluorescence and affinity chromatography measurements and electrophoretic DNA-unwinding assays. MATERIALS AND METHODSChemicals. Nor and [14C]Nor (specific activity, 46.5 ,uCi/ mg; 1 Ci = 37 GBq) were a kind gift of Merck Sharp & Dohme. Magnesium...

ATP-dependent DNA ligases are essential enzymes in both DNA replication and DNA repair processes. Here we report a functional characterization of the T4 DNA ligase. One N-terminal and two C-terminal deletion mutants were expressed in Escherichia coli as histidine- tagged proteins. An additional mutant bore a substitution of Lys159 in the active site that abolished ATP binding. All the proteins were tested in biochemical assays for ATP-dependent self-adenylation, DNA binding, nick joining, blunt-end ligation and AMP- dependent DNA relaxation. From this analysis we conclude that binding to DNA is mediated by sequences at both protein ends and plays a key role in the reaction. The enzyme establishes two different complexes with DNA: (i) a transient complex (T.complex) involving the adenylated enzyme; (ii) a stable complex (S.complex) requiring the deadenylated T4 DNA ligase. The formation of an S. complex seems to be relevant during both blunt-end ligation and DNA relaxation. Moreover the inactive His-K159L substitution mutant, although unable to self-adenylate, still possesses AMP-dependent DNA nicking activity.

DNA replication in mammalian cells occurs in discrete nuclear foci called ‘replication factories’. Here we show that DNA ligase I, the main DNA ligase activity in proliferating cells, associates with the factories during S phase but displays a diffuse nucleoplasmic distribution in non‐S phase nuclei. Immunolocalization analysis of both chloramphenicol acetyltransferase (CAT)‐DNA ligase I fusion proteins and epitope tagged DNA ligase I mutants allowed the identification of a 13 amino acid functional nuclear localization signal (NLS) located in the N‐terminal regulatory domain of the protein. Furthermore, the NLS is immediately preceded by a 115 amino acid region required for the association of the enzyme with the replication factories. We propose that in vivo the activity of DNA ligase I could be modulated through the control of its sub‐nuclear compartmentalization.