Chemical reclosure of opened imidazole ring of guanine

Imidazole ring opened adenine and guanine residues similar to those generated by gamma-irradiation of nucleosides of DNA, were chemically synthesised. Reaction conditions that promote the chemical reclosure of opened imidazole rings of guanine have been identified. The optimal conditions for the reclosure of such rings was found to be 0.2 M HCl at 37 degrees C. These conditions did not promote a reclosure of opened imidazole rings of adenine. The reclosure of opened imidazole rings of guanine was found to follow first order kinetics. The very low pH for this chemical ring reclosure precludes the likelihood that this reaction occurs intracellularly.     

Coiled Rings of DNA Released from Cells Infected with Bacteriophages T7 or T4 or from Uninfected Escherichia coli

The replicating intracellular DNA of phage T7 was labeled at high specific activity with tritiated thymidine. The DNA of uninfected Escherichia coli was similarly labeled. Portions of cells which contained replicating phage T7 or E. coli DNA were lysed by a lysozyme, freeze-thaw, sodium lauryl sulfate procedure, and the DNA was spread on Millipore membranes for visualization by autoradiography. The DNA of phage T7 appeared to be highly concatenated reaching lengths of up to 721 mum. Much of the DNA of phage T7 and E. coli was retained in compact globular structures. In addition, orderly coiled rings of varying diameter up to about 43 mum were regularly observed. Similar coiled ring structures were also observed in autoradiographs of replicating phage T4 DNA which had been prepared in previous experiments. Worcel and Burgi (27) have presented evidence that E. coli chromosomes, when gently extracted from cells, are in a multilooped and superhelically twisted configuration. The coiled rings which we have observed may correspond to the relaxed, multilooped configurations which they find when the superhelical twists have been relieved by one or more nicks in each loop.     

Biological properties of imidazole ring-opened N7-methylguanine in M13mp18 phage DNA.

Guanine residues methylated at the N-7 position (7-MeGua) are susceptible to cleavage of the imidazole ring yielding 2,6-diamino-4-hydroxy-5N-methyl-formamidopyrimidine (Fapy-7-MeGua). The presence of Fapy-7-MeGua in DNA template causes stops in DNA synthesis in vitro by E. coli DNA polymerase I. The biological consequences of Fapy-7-MeGua lesions for survival and mutagenesis were investigated using single-stranded M13mp18 phage DNA. Fapy-7-MeGua lesions were generated in vitro in phage DNA by dimethylsulfate (DMS) methylation and subsequent ring opening of 7-MeGua by treatment with NaOH (DMS-base). The presence of Fapy-7-MeGua residues in M13 phage DNA correlated with a significant decrease in transfection efficiency and an increase in mutation frequency in the lacZ gene, when transfected into SOS-induced JM105 E.coli cells. Sequencing analysis revealed unexpectedly, that mutation rate at guanine sites was only slightly increased, suggesting that Fapy-7-MeGua was not responsible for the overall increase in the mutagenic frequency of DMS-base treated DNA. In contrast, mutation frequency at adenine sites yielding A----G transitions was the most frequent event, 60-fold increased over DMS induced mutations. These results show that treatment with alkali of methylated single-stranded DNA generates a mutagenic adenine derivative, which mispairs with cytosine in SOS induced bacteria. The results also imply that the Fapy-7-MeGua in E. coli cells is primarily a lethal lesion.     

Excision of 5'-terminal deoxyribose phosphate from damaged DNA is catalyzed by the Fpg protein of Escherichia coli.

Homogeneous Fpg protein of Escherichia coli has DNA glycosylase activity which excises some purine bases with damaged imidazole rings, and an activity excising deoxyribose (dR) from DNA at abasic (AP) sites leaving a gap bordered by 5'- and 3'-phosphoryl groups. In addition to these two reported activities, we show that the Fpg protein also catalyzes the excision of 5'-terminal deoxyribose phosphate (dRp) from DNA, which is the principal product formed by the incision of AP endonucleases at abasic sites. Moreover, the rate of the Fpg protein catalysis for the 2,6-diamino-4-hydroxy-5-formamidopyrimidine-DNA glycosylase activity is slower than the activities excising dR from abasic sites and dRp from abasic sites preincised by endonucleases. The product released by the Fpg protein in the excision of 5'-terminal dRp from an abasic site preincised by an AP endonuclease is a single base-free unsaturated dRp, suggesting that the excision results from beta-elimination. The release of 5'-terminal dRp by crude extracts of E. coli from wild type and fpg-mutant strains shows that the Fpg protein is one of the major EDTA-resistant activities catalyzing this reaction.     

Escherichia coli single-stranded DNA binding protein stimulates the DNA deoxyribophosphodiesterase activity of exonuclease I.

The E. coli single-stranded binding protein (SSB) has been demonstrated in vitro to be involved in a number of replicative, DNA renaturation, and protective functions. It was shown previously that SSB can interact with exonuclease I to stimulate the hydrolysis of single-stranded DNA. We demonstrate here that E. coli SSB can also enhance the DNA deoxyribophosphodiesterase (dRpase) activity of exonuclease I by stimulating the release of 2-deoxyribose-5-phosphate from a DNA substrate containing AP endonuclease-incised AP sites, and the release of 4-hydroxy-2-pentenal-5-phosphate from a DNA substrate containing AP lyase-incised AP sites. E. coli SSB and exonuclease I form a protein complex as demonstrated by Superose 12 gel filtration chromatography. These results suggest that SSB may have an important role in the DNA base excision repair pathway.     

Release of normal bases from intact DNA by a native DNA repair enzyme.

Base excision repair is initiated by DNA glycosylases removing inappropriate bases from DNA. One group of these enzymes, comprising 3-methyladenine DNA glycosylase II (AlkA) from Escherichia coli and related enzymes from other organisms, has been found to have an unusual broad specificity towards quite different base structures. We tested whether such enzymes might also be capable of removing normal base residues from DNA. The native enzymes from E.coli, Saccharomyces cerevisiae and human cells promoted release of intact guanines with significant frequencies, and further analysis of AlkA showed that all the normal bases can be removed. Transformation of E. coli with plasmids expressing different levels of AlkA produced an increased spontaneous mutation frequency correlated with the expression levels, indicating that excision of normal bases occurs at biologically significant rates. We propose that the broad specificity 3-methyladenine DNA glycosylases represent a general type of repair enzyme 'pulling' bases in DNA largely at random, without much preference for a specific structure. The specificity for release of damaged bases occurs because base structure alterations cause instability of the base-sugar bonds. Damaged bases are therefore released more readily than normal bases once the bond activation energy is reduced further by the enzyme. Qualitatively, the model correlates quite well with the relative rate of excision observed for most, if not all, of the substrates described for AlkA and analogues.     

Real-time imaging of DNA ejection from single phage particles

Infection by tailed dsDNA phages is initiated by release of the viral DNA from the capsid and its polarized injection into the host. The driving force for the genome transport remains poorly defined. Among many hypothesis [1], it has been proposed that the internal pressure built up during packaging of the DNA in the capsid is responsible for its injection [2-4]. Whether the energy stored during packaging is sufficient to cause full DNA ejection or only to initiate the process was tested on phage T5 whose DNA (121,400 bp) can be released in vitro by mere interaction of the phage with its E. coli membrane receptor FhuA [5-7]. We present a fluorescence microscopy study investigating in real time the dynamics of DNA ejection from single T5 phages adsorbed onto a microfluidic cell. The ejected DNA was fluorescently stained, and its length was measured at different stages of the ejection after being stretched in a hydrodynamic flow. We conclude that DNA release is not an all-or-none process but occurs in a stepwise fashion and at a rate reaching 75,000 bp/sec. The relevance of this stepwise ejection to the in vivo DNA transfer is discussed.     

Cloning and expression in Escherichia coli of a human cDNA encoding the DNA repair protein N-methylpurine-DNA glycosylase

A 871-base pair cDNA encoding the human N-methylpurine-DNA glycosylase (MPG) was cloned from a HeLa S3 cDNA expression library in a pUC vector by phenotypic screening of MPG-negative (tag- alkA-) Escherichia coli cells exposed to methylmethane sulfonate. The active MPG is expressed as a 31-kDa fusion protein. The human cDNA-encoded MPG releases 3-methyladenine, 7-methylguanine, and 3-methylguanine from DNA and thus has a substrate range similar to that of the indigenous enzyme and the E. coli AlkA protein. The cDNA hybridizes with distinct restriction fragments of mammalian DNAs but not with E. coli or yeast DNA. A search in the GenBank data bank failed to show any other cloned DNA with a similar sequence. Although the human protein has 62% sequence homology with the corresponding rat enzyme, only a few amino acid residues are conserved between the human protein and the E. coli and yeast MPGs. However, a conserved glutamine residue in all MPGs that release 3-alkyladenine and an arginine residue in eukaryotic MPGs and E. coli AlkA that act additionally on N-alkylguanines suggest that these residues are involved in recognition of adenine and guanine adducts in DNA, respectively. Although the 1.1-kilobase mRNAs of MPG from human and rodents are similar in size, liver and cultured cells of rat have much lower levels of MPG mRNA than do human and mouse cells. A hamster cell line variant isolated as being resistant to methylmethane sulfonate does not have a higher level of MPG mRNA than the parent cell line.     

Reduction of the toxicity and mutagenicity of aziridine in mammalian cells harboring the Escherichia coli fpg gene.

Aziridine (ethyleneimine) reacts with DNA in vitro, mainly at the N7 position of guanine and N3 of adenine, then imidazole ring opening of the modified guanine results in formation of formamidopyrimidine (FaPy) residues. The Escherichia coli fpg gene encodes a DNA glycosylase that removes FaPy residues from DNA. To determine whether aziridine produces FaPy lesions in mammalian cells we have expressed the E.coli fpg gene in CHO cells. The transfected cells, expressing high levels of the bacterial protein, are more resistant to the toxic and mutagenic effects of aziridine than the control population. Less DNA damage was measured by quantitative PCR analysis in transfected than in control cells treated with equimolar concentrations of aziridine. The results suggest that aziridine produces in vivo FaPy residues that could account for the deleterious effects of this compound.     

Enzymatic excision from gamma-irradiated polydeoxyribonucleotides of adenine residues whose imidazole rings have been ruptured

The main forms of base damage in polydeoxyadenylic acid gamma-irradiated under hypoxic conditions are due to saturation and fragmentation of the adenine imidazole ring. An irradiated polymer was annealed with an equimolar amount of poly (dT) to generate a double-stranded polydeoxyribonucleotide containing scattered damaged base residues. On incubation of the latter with partially purified cell extracts of E.coli, imidazole ring-opened adenine, i.e. 4,6-diamino-5-formamidopyrimidine, was released in free form by a DNA glycosylase activity. The enzyme has been purified 4,500-fold, has Mr = 29,000, and appears to be identical with the previously described DNA repair enzyme formamidopyrimidine-DNA glycosylase.     

Structural polymorphism of the RecA protein from the thermophilic bacterium Thermus aquaticus.

The Escherichia coli RecA protein has served as a model for understanding protein-catalyzed homologous recombination, both in vitro and in vivo. Although RecA proteins have now been sequenced from over 60 different bacteria, almost all of our structural knowledge about RecA has come from studies of the E. coli protein. We have used electron microscopy and image analysis to examine three different structures formed by the RecA protein from the thermophilic bacterium Thermus aquaticus. This protein has previously been shown to catalyze an in vitro strand exchange reaction at an optimal temperature of about 60 degrees C. We show that the active filament formed by the T. aquaticus RecA on DNA in the presence of a nucleotide cofactor is extremely similar to the filament formed by the E. coli protein, including the extension of DNA to a 5.1-A rise per base pair within this filament. This parameter appears highly conserved through evolution, as it has been observed for the eukaryotic RecA analogs as well. We have also characterized bundles of filaments formed by the T. aquaticus RecA in the absence of both DNA and nucleotide cofactor, as well as hexameric rings of the protein formed under all conditions examined. The bundles display a very large plasticity of mass within the RecA filament, as well as showing a polymorphism in filament-filament contacts that may be important to understanding mutations that affect surface residues on the RecA filament.     

Substrate discrimination by formamidopyrimidine-DNA glycosylase: distinguishing interactions within the active site

Reactive oxygen species are byproducts of normal aerobic respiration and ionizing radiation, and they readily react with DNA to form a number of base lesions, including the mutagenic 8-oxo-7,8-dihydroguanine (8-oxoG), 2,6-diamino-4-hydroxy-5-formamidopyrimidine (FapyG), 4,6-diamino-5-formamidopyrimidine (FapyA), and 8-oxo-7,8-dihydroadenine (8-oxoA). Such oxidative lesions are removed by the base excision repair pathway, which is initiated by DNA glycosylases such as the formamidopyrimidine-DNA glycosylase (Fpg) in Escherichia coli. The 8-oxoG, FapyG, and FapyA lesions are bound and excised by Fpg, while structurally similar 8-oxoA is excised by Fpg very poorly. We carried out molecular modeling and molecular dynamics simulations to interpret substrate discrimination within the active site of E. coli Fpg. Lys-217 and Met-73 were identified as residues playing important roles in the recognition of the oxidized imidazole ring in the substrate bases, and the Watson-Crick edge of the damaged base plays a role in optimally positioning the base within the active site. The recognition and excision of FapyA likely result from the opened imidazole ring, while 8-oxoA's lack of flexibility and closed imidazole ring may contribute to Fpg's inability to excise this base. Different interactions between each base and the enzyme specificity pocket account for differential treatment of the various lesions by this enzyme, and thus elucidate the structure-function relationship involved in an initial step of base excision repair.     

Imidazole ring-opened DNA purines and their biological significance

Fragmentation of purine imidazole ring and production of formamidopyrimidines in deoxynucleosides (Fapy lesions) occurs upon DNA oxidation as well as upon spontaneous or alkali-triggered rearrangement of certain alkylated bases. Many chemotherapeutic agents such as cyclophosphamide or thiotepa produce such lesions in DNA. Unsubstituted FapyA and FapyG, formed upon DNA oxidation cause moderate inhibition of DNA synthesis, which is DNA polymerase and sequence dependent. Fapy-7MeG, a methylated counterpart of FapyG-, a efficiently inhibits DNA replication in vitro and in E.coli, however its mutagenic potency is low. This is probably due to preferential incorporation of cytosine opposite Fapy-7MeG and preferential extension of Fapy-7MeG:C pair. In contrast, FapyA and Fapy-7MeA possess miscoding potential. Both lesions in SOS induced E.coli preferentially mispair with cytosine giving rise to A-->G transitions. Fapy lesions substituted with longer chain alkyl groups also show simult aneous lethal and mutagenic properties. Fapy lesions are actively eliminated from DNA by repair glycosylases specific for oxidized purines and pyrimidines both in bacteria and eukaryotic cells. Bacterial enzymes include E.coli formamidopyrimidine-DNA-glycosylase (Fpg protein), endonuclease III (Nth protein) and endonuclease VIII (Nei protein).     

Enzymatic release of 5'-terminal deoxyribose phosphate residues from damaged DNA in human cells.

Activities that catalyze or promote the release of 5'-terminal deoxyribose phosphate residues from DNA abasic sites previously incised by an AP endonuclease have been identified in soluble extracts of several human cell lines and calf thymus. Such excision of base-free sugar phosphate residues from apurinic/apyrimidinic sites is expected to be obligatory prior to repair by gap filling and ligation. The most efficient excision function is due to a DNA deoxyribophosphodiesterase similar to the protein found in Escherichia coli. The human enzyme has been partially purified and freed from detectable exonuclease activity. This DNA deoxyribophosphodiesterase is a Mg(2+)-requiring hydrolytic enzyme with an apparent molecular mass of approximately 47 kDa and is located in the cell nucleus. By comparison, the major nuclear 5'----3' exonuclease, DNase IV, is unable to catalyze the release of 5'-terminal deoxyribose phosphate residues as free sugar phosphates but can liberate them at a slow rate as part of small oligonucleotides. Nonenzymatic removal of 5'-terminal deoxyribose phosphate from DNA by beta-elimination promoted by polyamines and basic proteins is a very slow mechanism of release compared to enzymatic hydrolysis. We conclude that a DNA deoxyribophosphodiesterase acts at an intermediate stage between an AP endonuclease and a DNA polymerase during DNA repair at apurinic/apyrimidinc sites in mammalian cells, but several alternative routes also exist for the excision of deoxyribose phosphate residues.     

Complex Formation of E. coli RNA Polymerase with Bacteriophage T2 DNA: Long-Range Effects in DNA

Conformation behavior of phase T2 DNA in the process of its interaction with it E. coli RNA polymerase was studied using spin labeling technique. T2 DNA was modified by the spin-labeled imidazole at OH-groups of glucosylated cytidine residues. It was shown that the binding of RNA polymerase under the conditions favoring the formation of open promoter complexes induces specific conformational changes in the spin-labeled DNA. The observed conformational changes encompass not only the promoter regions of DNA which are involved in direct contacts with RNA polymerase molecules but extend over remote DNA sites (long-range effect). In relation to this effect, current theoretical models of DNA dynamics are discussed.     

Methylpurine DNA glycosylase of the hyperthermophilic archaeon Archaeoglobus fulgidus

Base excision repair of DNA alkylation damage is initiated by a methylpurine DNA glycosylase (MPG) function. Such enzymes have previously been characterized from bacteria and eukarya, but not from archaea. We identified activity for the release of methylated bases from DNA in cell-free extracts of Archaeoglobus fulgidus, an archaeon growing optimally at 83 degrees C. An open reading frame homologous to the alkA gene of Escherichia coli was overexpressed and identified as a gene encoding an MPG enzyme (M(r) = 34 251), hereafter designated afalkA. The purified AfalkA protein differs from E. coli AlkA by excising alkylated bases only, from DNA, in the following order of efficiency: 3-methyladenine (m(3)A) > 3-methylguanine approximately 7-methyladenine > 7-methylguanine. Although the rate of enzymatic release of m(3)A is highest in the temperature range of 65-75 degrees C, it is only reduced by 50% at 45 degrees C, a temperature that does not support growth of A. fulgidus. At temperatures above 75 degrees C, nonenzymatic release of methylpurines predominates. The results suggest that the biological function of AfalkA is to excise m(3)A from DNA at suboptimal and maybe even mesophilic temperatures. This hypothesis is further supported by the observation that the afalkA gene function suppresses the alkylation sensitivity of the E. coli tag alkA double mutant. The amino acid sequence similarity and evolutionary relationship of AfalkA with other MPG enzymes from the three domains of life are described and discussed.     

Release of 5'-terminal deoxyribose-phosphate residues from incised abasic sites in DNA by the Escherichia coli RecJ protein.

Excision of deoxyribose-phosphate residues from enzymatically incised abasic sites in double-stranded DNA is required prior to gap-filling and ligation during DNA base excision-repair, and a candidate deoxyribophosphodiesterase (dRpase) activity has been identified in E. coli. This activity is shown here to be a function of the E. coli RecJ protein, previously described as a 5'-->3' single-strand specific DNA exonuclease involved in a recombination pathway and in mismatch repair. Highly purified preparations of dRpase contained 5'-->3' exonuclease activity for single-stranded DNA, and homogeneous RecJ protein purified from an overproducer strain had both 5'-->3' exonuclease and dRpase activity. Moreover, E. coli recJ strains were deficient in dRpase activity. The hydrolytic dRpase function of the RecJ protein requires Mg2+; in contrast, the activity of E. coli Fpg protein, that promotes the liberation of 5'-->3'Rp residues from DNA by beta-elimination, is suppressed by Mg2+. Several other E. coli nucleases, including exonucleases I, III, V, and VII, endonucleases I, III and IV and the 5'-->3' exonuclease function of DNA polymerase I, are unable to act as a dRpase. Nevertheless, E. coli fpg recJ double mutants retain capacity to repair abasic sites in DNA, indicating the presence of a back-up excision function.     

Cloning, sequencing, and expression of the DNA gyrase genes from Staphylococcus aureus.

We have isolated and cloned the gyrA and gyrB genes from Staphylococcus aureus. These adjacent genes encode the subunits of DNA gyrase. The nucleotide sequence of a 5.9-kb region which includes part of an upstream recF gene, the whole of gyrB and gyrA, and about 1 kb of unknown downstream sequence has been determined. The gyrB and gyrA gene sequences predict proteins of 886 and 644 amino acid residues, respectively, which have significant homologies with the gyrase subunits of Escherichia coli and Bacillus subtilis. Residues thought to be important to the structure and function of the subunits are conserved. These genes have been expressed separately by using a T7 promoter vector. N-terminal sequencing of the cloned gene products suggests that the mature GyrB subunit exists mainly with its initial five residues removed. Protein sequencing also supports the interpretation of our DNA sequencing data, which are inconsistent in several placed with the recently published sequence of the same genes (E. E. C. Margerrison, R. Hopewell, and L. M. Fisher, J. Bacteriol. 174:1596-1603, 1992).