In vitro assembly of relaxosomes at the transfer origin of plasmid RP 4 ( bacterial conjugation / initiation of transfer DNA replication / protein-DNA interaction / electron microscopy )

W. Pansegrau,D. Balzer,V. Kruft,R. Lurz,E. Lanka

Abstract:During tiatin of conjugative transfer of DNA containing the transfer origin (onTiT) of the promiscuous plasmid RP4, the proteins TraI, TraJ, and Trai interact and assemble a specialized nucleoprotein complex (the relaxosome) at onT. The structure can be visualized on electron micrographs. Siteand strand-specific nicking at the transfer origin in vitro is dependent on the proteins TraIl and TraJ and on Mg2+ ions. Substrate specificity is directed exclusively towards the cognate transfer origin: the RP4-specifled TraJ protein cannot recognize the closely related onT ofplasmid R751. After nicking, TraI protein remains attached to the 5'-terminal 2'-deoxycytidyl residue at the nic site [Pansegrau, W., Ziegelin, G. & Lanka, E. (1990) J. Biol. Chem. 265, 10637-10644]. Nicking and relaxosome formation require supercoiled DNA. Thus, a complicated structure involving multiple plasmidspecified proteins and a defined region ofDNA must be formed at the transfer origin to prepare the plasmid for generating the single strand to be transferred. Bacterial conjugation mediates horizontal gene transfer between a wide variety of organisms (1, 2). The general model for initiation of conjugative DNA synthesis proposes that cleavage at the nic site within the transfer origin (oriT) allows creation ofa single strand by subsequent strand displacement through rolling-circle-type replication. A prerequisite for the initial nicking reaction is the formation of a specialized nucleoprotein structure at oriT, the relaxosome. For the promiscuous IncPa plasmid RP4, it has been shown that genes specifying proteins which exert high-precision interactions with oriT sequences directly flank the transfer origin (3). The traJ gene encodes an oriT-recognizing protein essential for initiation of transfer DNA replication (3, 4). Genetic and biochemical data indicate that the product of the traf gene is an additional component ofthe relaxosome (3, 5). However, definite proof that the gene products specified by the relaxase operon (4) are sufficient for relaxosome formation can be obtained only by in vitro assembly from purified components. To address this issue, we attempted overproduction of Tral protein by means of molecular cloning in expression vectors and its purification. We report here that purified TraJ and Tral proteins mimic the specific nicking reaction on superhelical oriT DNA in vitro. However, we demonstrate by gel electrophoresis that stable complex formation with DNA needs a third protein component, the product of the traH gene. Electron microscopy of protein-DNA complexes directly locates the site of relaxosome assembly at the transfer origin. We presume that relaxosome formation follows a sequential order in that Tral protein, which bears the specific nuclease activity, recognizes a preformed TraJ-oriT complex, and that this structure subsequently is stabilized by specific protein-protein interactions of TraH with both TraJ and Tral. MATERIALS AND METHODS Bacterial Strains, Plasmids, and Media. Escherichia coli strain SCS1 (6) was used as host for plasmids. Recombinant plasmids pJF166uA2 (3), pMS2260 (7), pGZ192OA114 (8), pJF145n (3), and pWP392n (7) have been described. Expression systems used were the 17410/T7 gene 10 S/D vector pT7-7 obtained from Stanley Tabor (Harvard Medical School) and the Ptac/lacIQ vector pJF119HE (9). Cells were grown under the conditions described (5). When appropriate, antibiotics were used at the following concentrations: chloramphenicol, 10 /Lg/ml; ampicillin (sodium salt), 100 .g/ml. DNA Techniques. Standard molecular cloning techniques were performed as described (10). Proteins. TraJ protein was purified as described (4). A purification procedure for TraH protein (3) will be published elsewhere. Antiserum to the TraI protein was raised in rabbits by using a TraL/Tral hybrid polypeptide as immunogen (5). Immunoblot Assay. Proteins were electrophoresed in SDS/ 15% (wt/vol) polyacrylamide gels and transferred electrophoretically to a nitrocellulose membrane (12 V/cm, 90 min; ref. 11), followed by reaction with anti-TraL/Tral serum (1:500, 2 hr) and dichlorotriazinylaminofluorescein-conjugated goat anti-rabbit IgGs (Jackson ImmunoResearch; 1:50, 45 min). In Vitro Reconstitution of Relaxosomes. Under standard conditions, mixtures of Tra proteins and supercoiled plasmid DNA (0.7 ;kg) were incubated in a total volume of 30 p.l (20 mM Tris-HCl, pH 7.6/50 mM NaCI/10 mM MgCl2/0.1 mM EDTA with bovine serum albumin at 10 ,ug/ml) for 45 min at 370C. TraH protein was added after the samples were preincubated for 10 min at 370C. When desired, relaxed intermediates were captured by addition of SDS and proteinase K to concentrations of 1% (wt/vol) and 100 tkg/ml, respectively. After incubation at 370C for 5 min, samples were analyzed electrophoretically. The amount of specifically relaxed plasmid DNA was estimated by laser densitometry of photographs obtained from ethidium bromide-stained agarose gels. Since open circular (OC) and covalently closed circular (CCC) forms of plasmid DNA give different specific fluorescence intensities, the yield of specifically relaxed plasmid DNA was calculated from the decrease of the fluorescence signal of the CCC form relative to that of the untreated substrate DNA, ensuring that exposure signals on the film were proportional to the amount of DNA.