No abstract

DNA helicase I, the traI gene product of the Escherichia coli F factor, was shown to be associated with endonuclease activity specific for the transfer origin of the F plasmid, oriT. In the presence of Mg2+, the purified enzyme forms a complex, stable in the presence of sodium dodecylsulfate (SDS) with a negatively superhelical chimeric plasmid containing oriT. The enzyme nicks and, after this, apparently binds to the 5′ nick terminus when this complex is heated in the presence of SDS and/or EDTA or treated with proteinase K. Dideoxy sequencing locates the nick site in the F DNA strand transferred during bacterial conjugation after nucleotide 138 clockwise of the mid‐point of the BglII site at 66.7 kb of the F genetic map. A sequencing stop after nucleotide 137 of this strand (where oriT‐nicking seems to occur in vivo) is possibly an artefact caused by helicase I protein attached to the 5′ terminal nucleotide. Deletion in the amino‐terminal part of the traI polypeptide abolishes the oriT‐nicking activity while leaving the strand‐separating activity intact. These results confirm the prediction from genetic studies that helicase I is bifunctional with site‐specific endonuclease and strand‐separating activities.

The DNA-stimulated ATPase characterized in the accompanying paper is shown to be a DNA unwinding enzyme. Substrates employed were DNA . RNA hybrid duplexes and D N A . DNA partial duplexes prepared by polymerization on fd phage single-stranded DNA template. The enzyme was found to denature these duplexes in an ATP-dependent reaction, without detectably degrading. EDTA, an inhibitor of the Mg2 +-requiring ATPase, was found to prevent denaturation suggesting that dephosphorylation of the ATP and not only its presence is required. These results together with those from enzyme-DNA binding studies lead to ideas regarding the mode of enzymic action. It is proposed that the enzyme binds, in an initial step, to a single-stranded part of the DNA substrate molecule and that from here, energetically supported by ATP dephosphorylation, it invades double-stranded parts separating base-paired strands by processive, zipper-like action. It is further proposed that chain separation results from the combined action of several enzyme molecules and that a tendency of the enzyme to aggregate with itself reflects a tendency of the molecules to cooperate. Various functions are conceivable for the enzyme.In the preceding paper [l] we have described a novel DNA-stimulated ATPase purified from Escherichiu coli. The enzyme, an MgZ +-requiring y-phosphohydrolase, has a peptide molecular weight of approximately 180000 and exists, in high salt, as a monomer of probably highly anisometric shape. In low salt, where the ATPase activity is highest, the enzyme forms aggregates. The enzyme binds, as the purification procedures shows, to single-stranded DNA which also best satisfies its cofactor requirement and does not detectably cleave DNA.The present publication shows that the ATPase activity is used for denaturing base-paired DNA. Abbreviation. EGTA, ethylene glycol-bis(2-aminoethylether) N,N'-tetradcetate. MATERIALS AND METHODS BuffersEnzymes. DNA-stimulated ATPdse (EC 3.6.1.3); DNA polymerase I (EC 2.7.7.7); RNA polymerase (EC 2.7.7.6); S1 nuclease (EC 3.1.4.21).cerol, 100 pg bovine serum albumin/ml. Buffer 11 is 20 mM Tris/maleate buffer (pH 7.2), 2.5 mM MgClZ, 15 mM 2-mercaptoethanol, 10 % glycerol. EnzymesDNA unwinding ATPase, fraction 5, was prepared as described in the accompanying paper. E. coli DNA polymerase I and RNA polymerase were gifts from Dr K . Geider. S1 nuclease from Aspergillus oryzae was purified through the DEAE-cellulose step PI. Nucfeic AcidsDNA preparations were those used in the accompanying paper. 3H-labelled @X DNA was a gift of Dr U. Hess.DNA . RNA hybrid duplexes were prepared according to Wickner et ul. [3] by transcribing on 16 nmol (as nucleotide) fd DNA with 20 pg RNA polymerase in the presence of three non-radioactive ribonucleoside triphosphates (250 pM) and [5-3H]-CTP (New England) (1.1 pM; 23.2 Ci/mmol) in 1.6 ml final volume. After 20 min at 30 "C the reaction was interrupted with 5 mM EDTA and 1 % sodium dodecylsulphate. The mixture (4.6 x lo6 acid-precipitable 3H counts/min) was then extracted with phen...

Evidence from various sources in the literature suggests that, in connection with DNA, ATP dephosphorylation can be used to provide energy for mechanical effects. Starting from this concept we have studied a novel DNA-dependent ATPase purified to 90% homogeneity from Escherichiu coli. The enzyme has a peptide weight near 180000 and, in high salt, is a monomeric, probably highly anisometric molecule. In salt-free buffer, where the ATPase activity is highest, the enzyme forms aggregates. ATP is the preferred substrate (K, 0.27 mM) and dephosphorylated at the y-position at a maximal rate near lo4 molecules per enzyme monomer per min at 35 "C. A requirement for divalent cation is best satisfied by Mg2+ or Ca2+ and the requirement for DNA best by the single-stranded, circular DNA of phages @XI 74 (K, 62 nM nucleotide) and fd indicating that the enzyme recognizes internal DNA regions. When saturated with E. coli DNA unwinding protein, @X DNA is not accepted but, once in contact with the DNA, the enzyme is little inhibited by unwinding protein. Apparently the unwinding protein interferes preferentially with the recognition of DNA.The enzyme does not detectably cleave DNA, and for this and genetic reasons is not identical with the recBC ATPase or the K12 restriction ATPase of the extracted cells. The enzyme is probably not identical either with the dnaB-product-associated ATPase or the ATPase activity found in DNA polymerase I11 holoenzyme under appropriate conditions, and it is certainly not identical with a DNA-dependent ATPase of molecular weight 69000 from E. coli which has recently been purified. Attempts to ascribe the enzyme to other genes, including r e d , lex and rep, have failed.Various enzymes which hydrolyse adenosine 5'-triphosphate to the nucleoside diphosphate and orthophosphate utilize DNA as cofactor. Some of these DNA-stimulated ATPases behave as ATP-stimulated deoxyribonucleases, as the recBC nuclease of Escherichia coli [I, 21, a similar enzyme from Hemophilus influenzae [3] and the strain-specific restriction nucleases of E. coliB and K12 [4]. Other DNA-stimulated ATPases without nucleolytic activity have been detected in close association with purified replication proteins, the dnaB product of E. coli [S] and a complex between the proteins of genes 44 and 62 of phage T4 which is part of the viral replication machinery

In summary, we postulate that DNA unwinding and ATP dephosphorylation are coupled in different ways, depending on whether the fibrous ATPase or one of the globular ATPases provides the catalytic agent. Unanswered is the question of whether there is stoichiometry of ATP utilization during the unwinding of a duplex, and unsolved is the role of the individual enzyme in the cell.