- Email: [email protected]
Molecular characterization of a recombinant replication protein (Rep) from the Antarctic bacterium Psychrobacter sp. TA144
FEMS Microbiology Letters 198 (2001) 49^55
www.fems-microbiology.org
Molecular characterization of a recombinant replication protein (Rep) from the Antarctic bacterium Psychrobacter sp. TA144 Angela Duilio, M. Luisa Tutino, Vittoria Matafora, Giovanni Sannia, Gennaro Marino * Dipartimento di Chimica Organica e Biochimica, Universita© di Napoli `Federico II', Complesso universitario Monte S. Angelo, via Cinthia, 80100 Naples, Italy Received 6 December 2000; received in revised form 27 February 2001; accepted 27 February 2001
Abstract The Antarctic Gram-negative bacterium Psychrobacter sp. TA144 contains two small cryptic plasmids, called pTAUp and pTADw. pTAUp encodes a replication enzyme (PsyRep) whose activity is responsible for plasmid replication via the rolling circle replication pathway. Several attempts to produce the wild-type biologically active PsyRep in Escherichia coli failed, possibly due to auto-regulation of the protein population. However, the serendipitous occurrence of a frameshift mutation during the preparation of an expression vector resulted in the over-production of a recombinant protein, changed in its last 14 amino acid residues (PsyRep*), that precipitates in insoluble form. The purification of PsyRep* inclusion bodies and the successful refolding of the cold adapted enzyme allowed us to carry out its functional characterization. The mutated protein still displays a double stranded DNA nicking activity, while the change at the C-terminus impairs the enzyme specificity for the pTAUp cognate Ori+ sequence. ß 2001 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved. Keywords : Rep protein ; Psychrophile; Plasmid; Rolling circle; Refolding
1. Introduction Numerous plasmids isolated from either prokaryotes or eukaryotes replicate by the rolling circle replication (RCR) mechanism similar to that described for single stranded DNA phages in Gram-negatives. These replicons, called rolling circle plasmids (RCP), have been extensively studied (for reviews see [1^3]). The key element of the RCR mechanism is the so-called Rep, i.e. the plasmid encoded initiator protein. This protein is a replicon speci¢c initiator enzyme that belongs to the strand transferase superfamily ; it is responsible for both the initiation and termination steps in the RCR mechanism. In this mechanism a Rep protein recognizes a speci¢c DNA sequence (Ori+), and cleaves only one strand of DNA at a speci¢c site (nick site) leaving a free 3P-OH terminus which can be used by the host DNA poly-
* Corresponding author. Tel.: +39 (81) 674312; Fax: +39 (81) 674313; E-mail : [email protected] Abbreviations : MALDI-TOF, matrix assisted laser desorption ionization time of £ight; Ori+, plus origin of replication; PsyRep, Rep protein from Psychrobacter sp. TA144; Psyrep, PsyRep coding gene; RCP, rolling circle plasmid ; RCR, rolling circle replication
merase to extend the DNA sequence. After one round of polymerization, usually the same Rep recognizes a terminator sequence, cleaves the newly synthesized strand and ligates the ends of the displaced DNA ¢lament, thus forming a circular single stranded intermediate that is converted into a double stranded form by host factors, using the so-called minus replication origin as the initiation site [1^4]. The Antarctic Gram-negative bacterium Psychrobacter sp. TA144 contains two small cryptic plasmids, called pTAUp and pTADw, whose molecular characterization was carried out and reported in a previous work [5]. In particular, pTAUp encodes a replication enzyme (PsyRep) whose activity is responsible for plasmid replication via the RCR pathway. Structural similarities at the level of gene organization, protein sequence, and nick site sequence strongly suggested that the psychrophilic protein and the replication enzymes encoded by two Xanthomonas plasmids, pXV64 [6] and pXG31 (EMBL accession number AF069766), and by the I2-2 ¢lamentous bacteriophage [7] belong to a new subfamily of replication enzymes, possibly derived from a common ancestor by divergent evolution [5]. Since no members of this group have been studied in detail so far, we embarked on the over-production of PsyRep in Escherichia coli using many expression
0378-1097 / 01 / $20.00 ß 2001 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved. PII: S 0 3 7 8 - 1 0 9 7 ( 0 1 ) 0 0 1 2 1 - 5
FEMSLE 9881 18-4-01
50
A. Duilio et al. / FEMS Microbiology Letters 198 (2001) 49^55
systems, but all attempts to obtain the wild-type biologically active enzyme failed, possibly due to an auto-regulation of protein concentration within the host cell. However, the serendipitous occurrence of a frameshift mutation during the preparation of an expression vector resulted in the over-production of a protein, changed in its last 14 amino acid residues (PsyRep*), that accumulates in insoluble form. The puri¢cation of PsyRep* inclusion bodies and the successful refolding of the cold adapted enzyme allowed us to carry out its functional characterization. 2. Materials and methods 2.1. Enzymes and reagents Restriction enzymes, T4 DNA ligase, alkaline phosphatase, and Taq DNA polymerase were supplied by Boehringer, Amersham, Promega, or New England Biolabs. Enzyme assay conditions are those suggested by the manufacturers. DNA fragment puri¢cation was carried out using a QIAEX II kit from Qiagen. Acrylamide, agarose, and all other reagents were purchased from Sigma. 2.2. Bacterial strains Psychrobacter sp. strain TA144 [5,8,9] was routinely grown in aerobic conditions at 15³C in LB broth at pH 8.5 [10]. E. coli DH5K [11] was used as the host for gene cloning. E. coli cells were routinely grown in Terri¢c broth [10] containing 100 Wg ml31 of ampicillin, if transformed. E. coli strain RB791 [W3110 lac Iq L8] [12] was used as the bacterial host for recombinant protein production. 2.3. Expression of Psyrep in E. coli The pTAUp plasmid was isolated from the cell of Psychrobacter sp. strain TA144 by the alkaline lysis procedure as described in [10], followed by ultracentrifugation on a CsCl gradient. The 1008 bp long open reading frame (ORF), encoding the cold adapted Rep, was ampli¢ed by PCR using pTAUp as template and as primers the oligonucleotides: Nterminal pQE5P: 5P-GGAATGGATCCATGGTTGATTGGG-3P and C-terminal pQE3P: 5P-CTCATGTTCTAGATCTGCAAGCAAC-3P (boldface type shows the pTAUp speci¢c sequence, underlines show the NcoI and BglII sites, respectively). The primers were designed to insert the NcoI or the BglII restriction sites overlapping the Rep start and stop codons, respectively. PCR reaction was performed in a mixture containing 200 ng of template, 50 pmol of each oligonucleotide primer, 1.8 mM MgCl2 , 50 mM KCl, 20 mM Tris^HCl pH 8.3, 0.1% gelatin, and
200 WM dNTP in a ¢nal volume of 50 Wl. The mixture was incubated at 95³C for 10 min, after which 1.25 U Taq DNA polymerase was added. Twenty cycles of ampli¢cation (consisting of 1 min at 95³C, 1.5 min at 55³C and 1 min 5 s/cycle at 72³C) were carried out and were followed by a cycle in which the extension reaction at 72³C was prolonged for 15 min in order to complete DNA synthesis. The ampli¢ed fragment was digested with NcoI and BglII, then ligated into the corresponding sites of the expression vector pQE60 (Qiagen). The recombinant DNA (pQE-rep) was transferred in E. coli RB791 cells and several independent clones were tested for the IPTG induced production of PsyRep. Only one clone out of that analyzed was able to induce a large production of the recombinant enzyme. The sequencing of the recombinant expression vectors was performed using the T7 Sequenase sequencing kit (Amersham-USB) and suitable synthetic oligonucleotides as primers. It turned out that only the unique producing construct contained a frameshift mutation in the rep gene that resulted in the change of the last 14 amino acid residues and in the expression of a mutant enzyme (hereafter referred to as PsyRep*). The unique clone over-producing the recombinant PsyRep* was grown at 37³C in Terri¢c broth [10] supplemented with 100 Wg ml31 ampicillin. An overnight culture was inoculated in 200 ml of the same medium at OD595 0.2 and incubated at 37³C, under continuous shaking. When the culture reached OD595 2, expression of PsyRep was induced by adding IPTG at a ¢nal concentration of 1 mM. Cells were harvested, after incubation for a further 4 h at 37³C, by centrifugation for 15 min, 2700Ug, 4³C, and deeply frozen at 380³C. 2.4. Puri¢cation and in vitro refolding of PsyRep* Recombinant PsyRep* was puri¢ed from the inclusion bodies of IPTG induced RB791 E. coli cells harboring the pQE-rep* vector. The cell pellet from 200 ml culture was resuspended in 10 ml of 50 mM Tris^HCl, 1 mM EDTA, 1 mM phenylmethylsulfonyl £uoride pH 8.0, and lysozyme was added at 1 mg ml31 . After 30 min of incubation on ice, 5.5 ml of lysis solution (50 mM Tris^HCl, 2% Triton X-100, 63 mM EDTA, pH 8.0) was added and the mixture was stirred for 15 min at 4³C. The cell breaking was enhanced by six cycles of sonication (30 s on, 1 min o¡) on ice (Branson sonicator, model B-15, intensity 4.5). The sample was centrifuged at 2000Ug, 4³C, for 20 min and the pellet was resuspended in 10 ml of 50 mM Tris^HCl, 1 mM EDTA, pH 8, divided into aliquots of 1 ml and centrifuged at 1800Ug, 4³C, for 20 min. This wash was repeated two times and the inclusion bodies were lastly dried by a centrifugation step of 5 min at 9500Ug, 4³C, and stored as dry pellets at 4³C. Each aliquot of PsyRep* inclusion bodies (1.75 mg) was completely dissolved in 25 ml of ice-cold denaturing solution (50 mM Tris^HCl, 6 M guanidinium chloride, 1 mM
FEMSLE 9881 18-4-01
A. Duilio et al. / FEMS Microbiology Letters 198 (2001) 49^55
dithiothreitol (DTT), pH 8.0) and then diluted 10 times in ice-cold refolding bu¡er (10 mM Tris^HCl, 100 mM KCl, 10 mM Mg(OAc)2 , pH 8.2). The protein solution was then concentrated by ultra¢ltration on an Amicon PM-10 membrane (diameter 63 mm) to a ¢nal volume of 20 ml and dialyzed two times against 2 l of ice-cold refolding solution at 4³C. Protein concentration was determined using the BioRad protein assay with bovine serum albumin as the standard. 2.5. In-gel digestion and MALDI mass mapping of recombinant PsyRep* The PsyRep containing band was excised from an SDS^ PAGE gel, washed in acetonitrile and 0.1 M NH4 HCO3 by turns, reduced and S-alkylated with iodoacetamide. The treated band was then incubated with an excess of bovine trypsin or of Staphylococcus aureus V-8 protease (Sigma) overnight at 37³C. The resulting peptide mixture was eluted from the gel and a fraction of 0.3 Wl out of a total of 100 Wl digest solution was used for matrix assisted laser desorption ionization time of £ight (MALDI-TOF) mass spectrometric analysis. MALDI mass spectra were recorded using a Voyager DE MALDI-TOF mass spectrometer equipped with a 337-nm N2 laser and a delay extraction device (PE Biosystems). A mixture of analyte solution, K-cyano-4-hydroxycinnamic acid (10 mg ml31 in CH3 CN/EtOH/0.1% tri£uoroacetic acid, 1:1:1 v/v) and bovine insulin was applied to the sample plate and air dried. Mass calibration was performed using the molecular ions from the bovine insulin at 5734.5 Da and the matrix peak at 379.1 Da as internal standards. Raw data were analyzed using computer software provided by the manufacturer and are reported as average masses. 2.6. Relaxation reactions Reaction mixture (30 Wl) typically contained 10 mM Tris^HCl, 100 mM KCl, 10 mM Mg(OAc)2 pH 8.2, 20% ethylene glycol, 160 ng of supercoiled plasmid substrate (pTAUp/Dw, pMtBL), and di¡erent amounts of PsyRep* were added. After incubation at 4³C for 3.5 h, reactions were stopped by addition of EDTA at a ¢nal concentration of 20 mM, and the samples were loaded on 0.8% agarose gels. Electrophoresis was carried out at room temperature in TAE 1Ubu¡er [10] containing 1 Wg ml31 ethidium bromide for 2 h at 6 V cm31 . 3. Results 3.1. Expression of recombinant Psyrep in E. coli Following the molecular characterization of a plasmid
51
(pTAUp) from Psychrobacter sp. TA144, encoding a replication enzyme (PsyRep) [5], we attempted to over-produce PsyRep in E. coli using many commercially available expression systems. These e¡orts failed since we were unable to the detect the over-expressed product (data not shown). However, during the setting up of one of the tested expression systems (pQE60 vector), we succeeded in obtaining a considerable amount of a recombinant product. The 1008 bp long PsyRep encoding ORF was ampli¢ed by PCR by using as primers two synthetic oligonucleotides designed to insert the NcoI or the BglII restriction sites overlapping the Rep start and stop codons, respectively. In particular, the insertion of the BglII site resulted in the removal of the ORF natural stop codon, allowing the expression of the recombinant PsyRep as a fusion protein with the C-terminal His6 moiety encoded by the vector (see Fig. 1, PsyRepFUS). By using the newly inserted restriction sites, the ampli¢ed coding region was cloned downstream of the IPTG inducible promoter in the pQE60 expression vector, and the resulting plasmid (pQErep) was used to transform the E. coli host strain RB791. Several independent recombinant clones were tested for their ability to over-produce PsyRep but, although di¡erent expression conditions (i.e. growth temperature, cell density, and induction time) were tested (data not shown), only one clone was positive for the production of the recombinant enzyme (see Fig. 2). However, in this positive clone the product was always associated with the insoluble fraction of the extract (Fig. 2, lane 4). DNA sequencing of the ampli¢ed ORFs inserted either into the producing vector or into a nonproducing one revealed that the latter was correctly synthesized during the PCR reaction, while the producing construction contained a deletion of G at position 967 (see Fig. 1). This spontaneous mutation resulted in the frameshift of the translation frame that a¡ected the C-terminal portion of PsyRep from Ala322 to the C-terminal end. Translation of this construct resulted in the production of a mutant protein (named PsyRep*), 332 residues long, in which the Cterminus sequence has changed to Glu-Tyr-Leu-Leu-GlyHis-Thr-Ser-Lys-Asp. Since the mutant enzyme was expressed in E. coli as inclusion bodies, PsyRep* production was carried out routinely in cells grown at 37³C and recovered after 4 h of IPTG induction, which started at OD600 2. 3.2. Puri¢cation and refolding of recombinant PsyRep* The cell pellet collected from 200 ml culture of induced recombinant RB791 E. coli cells was disrupted by sonication and the PsyRep* inclusion bodies were puri¢ed from the homogenate by di¡erential centrifugation. The insoluble fraction was dissolved in ice-cold denaturing bu¡er containing 6 M guanidinium chloride and 1 mM DTT at a ¢nal protein concentration of 0.07 mg ml31 . PsyRep* refolding was obtained by a 10-fold rapid dilution in cold
FEMSLE 9881 18-4-01
52
A. Duilio et al. / FEMS Microbiology Letters 198 (2001) 49^55
Fig. 1. Nucleotide sequence of the ampli¢ed Psyrep gene cloned into the pQE60 vector and its translation. The amino acid sequences underlined represent the PsyRep* peptides, resulting from the protein digestion by either trypsin or V-8 proteases, assigned by MALDI mass spectrometry. The PsyRep active site is highlighted in gray, while the pTAUp nick site is indicated by a ¢lled box. PsyRepFUS is the fusion protein between PsyRep and the His6 tag epitope ; PsyRep* is the frameshifted product derived from the translation of an ORF containing the deletion of nucleotide 967.
refolding bu¡er, followed by a dialysis step for removing the denaturant, resulting in 60% recovery of soluble protein (see Fig. 2, lane 5). The renaturation process was monitored by recording the £uorescence spectra of the protein before and after the removal of the denaturing agent. Both the decrease in £uorescence intensity and the change of the maximum emission wavelength from 354.5 to 345.5 nm suggested that the protein underwent a structural rearrangement (data not shown). 3.3. Structural characterization of recombinant PsyRep* The soluble protein was fractionated by SDS^PAGE and the PsyRep* band subjected to in situ reduction, carboxyamidomethylation, and digestion with either trypsin or V-8 protease. The resulting peptide mixtures were extracted from the gel and directly analyzed by MALDI mass spectrometry without any puri¢cation step. All mass signals were assigned to the corresponding peptides within the PsyRep* sequence on the basis of their mass value and the speci¢city of the proteolytic enzyme used,
Fig. 2. Expression of recombinant PsyRep* in E. coli. RB791 E. coli cells harboring the pQE-rep* vector, were induced with IPTG as described in Section 2. Protein samples were resolved on a SDS^10% polyacrylamide gel and stained with Coomassie brilliant blue. IPTG induction determined the appearance of an extra band in the whole cell lysate (lane 2) with respect to the same sample from the uninduced cells (lane 1). Lane 3 corresponds to the soluble lysate fraction, while lane 4 shows the inclusion bodies fraction. Lane 5, soluble PsyRep* after refolding. The location of PsyRep* is indicated by arrows. M, molecular mass markers.
FEMSLE 9881 18-4-01
A. Duilio et al. / FEMS Microbiology Letters 198 (2001) 49^55
leading to the veri¢cation of about 80% of the protein primary structure, as reported in Fig. 1. The mass signal at m/z 1842.9 con¢rmed the occurrence of the frameshift mutation, since this value could not be assigned to any peptide in the PsyRepFUS sequence (Fig. 1), while it is in perfect agreement with the expected m/z 1843.1 value for the mutant C-terminal peptide 316^332 of PsyRep* (Fig. 1). 3.4. Functional studies Although mutated in the last 14 amino acid residues of its C-terminus, the refolded pTAUp PsyRep* was tested for its nicking activity in vitro. A relaxation assay was set up modifying the conditions reported in [13], as described in Section 2. Di¡erent amounts of refolded PsyRep* were incubated at 4³C with about 150 ng of pTAUp/Dw plasmid preparation, and the reaction products were resolved by agarose gel electrophoresis. As shown in Fig. 3A, PsyRep* exhibited a nicking activity on both TA144 natural plasmids since it induced a change of substrate topology from a supercoiled conformation to an open circular one in a dose dependent manner. Interestingly, pTADw was nicked by the recombinant enzyme as well, and this result could indicate that PsyRep is responsible for the replication of both TA144 plasmids
Fig. 3. Relaxation of pTAUp/pTADw (A) and pMtBL (B) DNA by the refolded PsyRep* protein. A: E¡ect of increasing PsyRep* on the Psychrobacter sp. TA144 plasmids; lane 1, no PsyRep*; lane 2, 240 ng; lane 3, 360 ng; lane 4, 600 ng. B: E¡ect of increasing PsyRep* on the pMtBL plasmid (9); lane 1, no PsyRep*; lane 2, 360 ng; lane 3, 600 ng. M, molecular mass marker; sc, supercoiled form; oc, open circular form.
53
in vivo. Furthermore, the same relaxation assay was carried out on an unrelated substrate, plasmid pMtBL from the psychrophilic bacterium Pseudoalteromonas haloplanktis TAC 125 [9]. As shown in Fig. 3B, pMtBL was nicked by PsyRep* as well, suggesting that the mutated enzyme could have lost its substrate speci¢city. 4. Discussion In this paper we report the expression of a RCR initiator of the pTAUp subfamily and its functional characterization. To obtain the recombinant protein, we attempted to over-produce PsyRep in an IPTG inducible system as a fusion protein with the His6 tag epitope at its C-terminal end. Several independent expression constructs were tested for their ability to induce PsyRep production but accumulation of the recombinant enzyme was achieved only in the cell extract of one of them. Further analysis demonstrated that the psychrophilic product was always associated with the insoluble fraction, although the expression was carried out in di¡erent experimental conditions. Structural characterization of the recombinant enzyme at the level of both its coding gene and its amino acid sequence revealed that PCR synthesis caused a nucleotide deletion (position 967) that resulted in a change of the protein's last 14 amino acid residues (PsyRep*). This serendipitous mutation and the consequent change in the expressed product is responsible for PsyRep* precipitation and aggregation, since E. coli recombinant cells harboring the correct expression construct (pQE-rep) or other commercially available expression vectors are unable to produce the wildtype PsyRep protein either in soluble or in insoluble form. Since the pTAUp Ori+ is actually contained in the PsyRep coding ORF [5], interaction between the expressed product and its cognate target is indeed possible even in E. coli. This in vivo interaction may cause several e¡ects, one of which is the auto-regulation of the PsyRep population by a mechanism possibly similar to that reported for the plasmids pT181 [14] and pUB110 [15]. This attractive hypothesis is currently under investigation. The initiator proteins of RCP have origin speci¢c DNA binding and nicking^closing activities. It is highly likely that these enzymes display a modular arrangement with well-de¢ned regions involved in the above functions [13,16^18]. In particular, Rep proteins belonging to the same family are characterized by conserved nicking domains, containing the tyrosine residue of the active site, and variable regions responsible for the DNA binding activity, which in the initiators of the pT181 family are located at their carboxy-terminal end [19^21]. The same structural organization has also been reported for the Rep proteins belonging to the pTAUp subfamily, in which the putative nicking active site was positioned in a region corresponding to residues 155^172 of the PsyRep amino acid
FEMSLE 9881 18-4-01
54
A. Duilio et al. / FEMS Microbiology Letters 198 (2001) 49^55
sequence (see Fig. 1) [5]. In this context, it seemed interesting to test if the PsyRep* mutation, i.e. the change of the last 14 C-terminal amino acids, alters the enzyme activity. Several protocols for the PsyRep* refolding were tested and the highest yield (about 60%) was achieved by the rapid dilution of the denaturing agent in which the puri¢ed inclusion bodies were dissolved. A relaxation assay was set up to monitor the activity of the soluble enzyme in the presence of the TA144 natural plasmids as substrate (pTAUp and pTADw, Fig. 3A), and taking into account the cold adaptation of the source strain, the experiments were carried out at low temperature (i.e. 4³C) only. The change in substrate topology demonstrates that the refolded PsyRep* exhibits a nicking activity on both TA144 plasmids in a dose dependent manner. Although this result could be suggestive of a common role for PsyRep as initiator in the replication of either pTAUp or pTADw, this hypothesis is not supported by the structural comparison of the two plasmids. In fact, pTADw does not contain any secondary structure identical or at least similar to the pTAUp Ori+ region [5]. Furthermore, the mutant psychrophilic enzyme exerts its nicking activity even on an unrelated supercoiled DNA (pMtBL, Fig. 3B). Taken together, these functional results indicate that the change of its last 14 C-terminal amino acid residues altered or even destroyed the DNA recognition and binding properties of PsyRep*. This hypothesis is in agreement with the observed inability of the mutated enzyme to catalyze the second step of topoisomerase activity, i.e. the closing reaction, which should have generated circular relaxed dsDNA. In fact, a stable protein^DNA interaction has been reported to be a prerequisite for full topoisomerase activity (i.e. nicking^closing) of pT181 RepC in vitro [22]. The above results strongly suggest that the psychrophilic initiator protein also exhibits a modular structural organization with independent functional domains. PsyRep* may be of interest since it uncouples its two functions, i.e. it still retains its nicking activity while it seems to have lost the DNA binding speci¢city. Further work is needed to investigate if the lack of PsyRep* substrate recognition is linked to the speci¢c amino acid sequence altered in the mutant enzyme, or to a not proper folding of the C-terminal domain, impaired by the frameshift mutation. Acknowledgements This work was supported by grants of the European Union (Program COLDZYME, Contract ERB BIO4 CT96 0051 ; Program EUROCOLD, Contract ERB BIO4 CT95 0017 ; Program COLDNET, Contract ERB FMRX CT97 0131), of the Ministero dell' Universita© e della Ricerca Scienti¢ca (Progetti di Rilevante Interesse
Nazionale 1999) and of the Consiglio Nazionale delle Ricerche (CNR Contract 97.01138.PF49, Progetto Finalizzato `Biotecnologie').
References [1] Gruss, A. and Ehrich, S.D. (1989) The family of highly interrelated single-stranded deoxyribonucleic acid plasmids. Microbiol. Rev. 53, 231^241. [2] Khan, A.K. (1997) Rolling-circle replication of bacterial plasmids. Microbiol. Mol. Biol. Rev. 61, 442^455. [3] del Solar, G., Giraldo, R., Ruiz-Echevarr|©a, M.J., Espinosa, M. and D|©az-Orejas, R. (1998) Replication and control of circular bacterial plasmids. Microbiol. Mol. Biol. Rev. 62, 434^464. [4] Novick, R.P. (1998) Contrasting lifestyles of rolling circle phages and plasmids. Trends Biol. Sci. 23, 434^438. [5] Tutino, M.L., Duilio, A., Moretti, M.A., Sannia, G. and Marino, G. (2000) A rolling-circle plasmid from Psychrobacter sp. TA144: evidence for a novel Rep subfamily. Biochem. Biophys. Res. Commun. 274, 488^495. [6] Weng, S.F., Fan, Y.F., Tseng, Y.H. and Lin, J.W. (1997) Sequence analysis of the small cryptic Xanthomonas campestris pv. vescicatoria plasmid pXV64 encoding a Rep protein similar to gene II protein of phage I2-2. Biochem. Biophys. Res. Commun. 231, 121^125. [7] Stassen, A.P.M., Schoenmakers, E.F.P.M., Maoxiao, M., Schoenmakers, J.G.G. and Konings, R.N.H. (1992) Nucleotide sequence of the genome of the ¢lamentous bacteriophage I2-2: module evolution of the ¢lamentous phage genome. J. Mol. Evol. 34, 141^152. [8] Feller, G., Thiry, M., Arpigny, J.L., Mergeay, M. and Gerday, C. (1990) Lipases from psychrotrophic antarctic bacteria. FEMS Microbiol. Lett. 66, 239^244. [9] Tutino, M.L., Duilio, A., Fontanella, B., Moretti, M.A., Sannia, G. and Marino, G. (1999) Plasmids from Antarctic bacteria. In: ColdAdapted Organisms : Ecology, Physiology, Enzymology and Molecular Biology (Margesin, R., and Schinner, F., Eds.), pp. 335^348. Springer Verlag, Berlin. [10] Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual, 2nd edn. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. [11] Hanahan, D. (1983) Studies on transformation of Escherichia coli with plasmids. J. Mol. Biol. 166, 557^580. [12] Andreotti, G., Tutino, M.L., Sannia, G., Marino, G. and Cubellis, M.V. (1994) Indole-3-glycerol-phosphate synthetase from Sulfolobus solfataricus as a model for studying thermostable TIM-barrel enzymes. Biochim. Biophys. Acta 1208, 310^315. [13] Koepsel, R.R., Murray, R.W., Rosenblum, W.D. and Khan, S.A. (1985) The replication initiator protein of plasmid pT181 has sequence-speci¢c endonuclease and topoisomerase-like activities. Proc. Natl. Acad. Sci. USA 82, 6845^6849. [14] Iordanescu, S. (1995) Plasmid pT181 replication is decreased at high levels of RepC per plasmid copy. Mol. Microbiol. 16, 477^484. [15] Muller, A.K., Rojo, F. and Alonso, J.C. (1995) The level of the pUB110 replication initiator protein is autoregulated, which provides an additional control for plasmid copy number. Nucleic Acids Res. 23, 1894^1900. [16] Gros, M.F., te Riele, H. and Ehrlich, S.D. (1987) Rolling circle replication of single-stranded DNA plasmid pC194. EMBO J. 6, 3863^ 3869. [17] Moscoso, M., del Solar, G. and Espinosa, M. (1995) In vitro recognition of the replication origin of pLS1 and of plasmids of the pLS1 family by the RepB initiator protein. J. Bacteriol. 177, 7041^ 7049. [18] Chang, T.L., Kramer, M.G., Ansari, R.A. and Khan, S.A. (2000) Role of individual monomers of a dimeric initiator protein in the
FEMSLE 9881 18-4-01
A. Duilio et al. / FEMS Microbiology Letters 198 (2001) 49^55 initiation and termination of plasmid rolling circle replication. J. Biol. Chem. 275, 13529^13534. [19] Thomas, C.D., Nikiforov, T.T., Connolly, B.A. and Shaw, W.V. (1995) Determination of sequence speci¢city between a plasmid replication initiator protein and the origin of replication. J. Mol. Biol. 254, 381^391. [20] Dempsey, L.A., Birch, P. and Khan, S.A. (1992) Six amino acids determine the sequence-speci¢c DNA binding and replication specificity of the initiator proteins of the pT181 family. J. Biol. Chem. 267, 24538^24543.
55
[21] Wang, P.Z., Projan, S.J., Henriquez, V. and Novick, R.P. (1992) Speci¢city of origin recognition by replication initiator protein in plasmids of the pT181 family is determined by a six amino acid residue element. J. Mol. Biol. 223, 145^158. [22] Dempsey, L.A., Birch, P. and Khan, S.A. (1992) Uncoupling of the DNA topoisomerase and replication activities of an initiator protein. Proc. Natl. Acad. Sci. USA 89, 3083^3087.
FEMSLE 9881 18-4-01
www.fems-microbiology.org
Molecular characterization of a recombinant replication protein (Rep) from the Antarctic bacterium Psychrobacter sp. TA144 Angela Duilio, M. Luisa Tutino, Vittoria Matafora, Giovanni Sannia, Gennaro Marino * Dipartimento di Chimica Organica e Biochimica, Universita© di Napoli `Federico II', Complesso universitario Monte S. Angelo, via Cinthia, 80100 Naples, Italy Received 6 December 2000; received in revised form 27 February 2001; accepted 27 February 2001
Abstract The Antarctic Gram-negative bacterium Psychrobacter sp. TA144 contains two small cryptic plasmids, called pTAUp and pTADw. pTAUp encodes a replication enzyme (PsyRep) whose activity is responsible for plasmid replication via the rolling circle replication pathway. Several attempts to produce the wild-type biologically active PsyRep in Escherichia coli failed, possibly due to auto-regulation of the protein population. However, the serendipitous occurrence of a frameshift mutation during the preparation of an expression vector resulted in the over-production of a recombinant protein, changed in its last 14 amino acid residues (PsyRep*), that precipitates in insoluble form. The purification of PsyRep* inclusion bodies and the successful refolding of the cold adapted enzyme allowed us to carry out its functional characterization. The mutated protein still displays a double stranded DNA nicking activity, while the change at the C-terminus impairs the enzyme specificity for the pTAUp cognate Ori+ sequence. ß 2001 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved. Keywords : Rep protein ; Psychrophile; Plasmid; Rolling circle; Refolding
1. Introduction Numerous plasmids isolated from either prokaryotes or eukaryotes replicate by the rolling circle replication (RCR) mechanism similar to that described for single stranded DNA phages in Gram-negatives. These replicons, called rolling circle plasmids (RCP), have been extensively studied (for reviews see [1^3]). The key element of the RCR mechanism is the so-called Rep, i.e. the plasmid encoded initiator protein. This protein is a replicon speci¢c initiator enzyme that belongs to the strand transferase superfamily ; it is responsible for both the initiation and termination steps in the RCR mechanism. In this mechanism a Rep protein recognizes a speci¢c DNA sequence (Ori+), and cleaves only one strand of DNA at a speci¢c site (nick site) leaving a free 3P-OH terminus which can be used by the host DNA poly-
* Corresponding author. Tel.: +39 (81) 674312; Fax: +39 (81) 674313; E-mail : [email protected] Abbreviations : MALDI-TOF, matrix assisted laser desorption ionization time of £ight; Ori+, plus origin of replication; PsyRep, Rep protein from Psychrobacter sp. TA144; Psyrep, PsyRep coding gene; RCP, rolling circle plasmid ; RCR, rolling circle replication
merase to extend the DNA sequence. After one round of polymerization, usually the same Rep recognizes a terminator sequence, cleaves the newly synthesized strand and ligates the ends of the displaced DNA ¢lament, thus forming a circular single stranded intermediate that is converted into a double stranded form by host factors, using the so-called minus replication origin as the initiation site [1^4]. The Antarctic Gram-negative bacterium Psychrobacter sp. TA144 contains two small cryptic plasmids, called pTAUp and pTADw, whose molecular characterization was carried out and reported in a previous work [5]. In particular, pTAUp encodes a replication enzyme (PsyRep) whose activity is responsible for plasmid replication via the RCR pathway. Structural similarities at the level of gene organization, protein sequence, and nick site sequence strongly suggested that the psychrophilic protein and the replication enzymes encoded by two Xanthomonas plasmids, pXV64 [6] and pXG31 (EMBL accession number AF069766), and by the I2-2 ¢lamentous bacteriophage [7] belong to a new subfamily of replication enzymes, possibly derived from a common ancestor by divergent evolution [5]. Since no members of this group have been studied in detail so far, we embarked on the over-production of PsyRep in Escherichia coli using many expression
0378-1097 / 01 / $20.00 ß 2001 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved. PII: S 0 3 7 8 - 1 0 9 7 ( 0 1 ) 0 0 1 2 1 - 5
FEMSLE 9881 18-4-01
50
A. Duilio et al. / FEMS Microbiology Letters 198 (2001) 49^55
systems, but all attempts to obtain the wild-type biologically active enzyme failed, possibly due to an auto-regulation of protein concentration within the host cell. However, the serendipitous occurrence of a frameshift mutation during the preparation of an expression vector resulted in the over-production of a protein, changed in its last 14 amino acid residues (PsyRep*), that accumulates in insoluble form. The puri¢cation of PsyRep* inclusion bodies and the successful refolding of the cold adapted enzyme allowed us to carry out its functional characterization. 2. Materials and methods 2.1. Enzymes and reagents Restriction enzymes, T4 DNA ligase, alkaline phosphatase, and Taq DNA polymerase were supplied by Boehringer, Amersham, Promega, or New England Biolabs. Enzyme assay conditions are those suggested by the manufacturers. DNA fragment puri¢cation was carried out using a QIAEX II kit from Qiagen. Acrylamide, agarose, and all other reagents were purchased from Sigma. 2.2. Bacterial strains Psychrobacter sp. strain TA144 [5,8,9] was routinely grown in aerobic conditions at 15³C in LB broth at pH 8.5 [10]. E. coli DH5K [11] was used as the host for gene cloning. E. coli cells were routinely grown in Terri¢c broth [10] containing 100 Wg ml31 of ampicillin, if transformed. E. coli strain RB791 [W3110 lac Iq L8] [12] was used as the bacterial host for recombinant protein production. 2.3. Expression of Psyrep in E. coli The pTAUp plasmid was isolated from the cell of Psychrobacter sp. strain TA144 by the alkaline lysis procedure as described in [10], followed by ultracentrifugation on a CsCl gradient. The 1008 bp long open reading frame (ORF), encoding the cold adapted Rep, was ampli¢ed by PCR using pTAUp as template and as primers the oligonucleotides: Nterminal pQE5P: 5P-GGAATGGATCCATGGTTGATTGGG-3P and C-terminal pQE3P: 5P-CTCATGTTCTAGATCTGCAAGCAAC-3P (boldface type shows the pTAUp speci¢c sequence, underlines show the NcoI and BglII sites, respectively). The primers were designed to insert the NcoI or the BglII restriction sites overlapping the Rep start and stop codons, respectively. PCR reaction was performed in a mixture containing 200 ng of template, 50 pmol of each oligonucleotide primer, 1.8 mM MgCl2 , 50 mM KCl, 20 mM Tris^HCl pH 8.3, 0.1% gelatin, and
200 WM dNTP in a ¢nal volume of 50 Wl. The mixture was incubated at 95³C for 10 min, after which 1.25 U Taq DNA polymerase was added. Twenty cycles of ampli¢cation (consisting of 1 min at 95³C, 1.5 min at 55³C and 1 min 5 s/cycle at 72³C) were carried out and were followed by a cycle in which the extension reaction at 72³C was prolonged for 15 min in order to complete DNA synthesis. The ampli¢ed fragment was digested with NcoI and BglII, then ligated into the corresponding sites of the expression vector pQE60 (Qiagen). The recombinant DNA (pQE-rep) was transferred in E. coli RB791 cells and several independent clones were tested for the IPTG induced production of PsyRep. Only one clone out of that analyzed was able to induce a large production of the recombinant enzyme. The sequencing of the recombinant expression vectors was performed using the T7 Sequenase sequencing kit (Amersham-USB) and suitable synthetic oligonucleotides as primers. It turned out that only the unique producing construct contained a frameshift mutation in the rep gene that resulted in the change of the last 14 amino acid residues and in the expression of a mutant enzyme (hereafter referred to as PsyRep*). The unique clone over-producing the recombinant PsyRep* was grown at 37³C in Terri¢c broth [10] supplemented with 100 Wg ml31 ampicillin. An overnight culture was inoculated in 200 ml of the same medium at OD595 0.2 and incubated at 37³C, under continuous shaking. When the culture reached OD595 2, expression of PsyRep was induced by adding IPTG at a ¢nal concentration of 1 mM. Cells were harvested, after incubation for a further 4 h at 37³C, by centrifugation for 15 min, 2700Ug, 4³C, and deeply frozen at 380³C. 2.4. Puri¢cation and in vitro refolding of PsyRep* Recombinant PsyRep* was puri¢ed from the inclusion bodies of IPTG induced RB791 E. coli cells harboring the pQE-rep* vector. The cell pellet from 200 ml culture was resuspended in 10 ml of 50 mM Tris^HCl, 1 mM EDTA, 1 mM phenylmethylsulfonyl £uoride pH 8.0, and lysozyme was added at 1 mg ml31 . After 30 min of incubation on ice, 5.5 ml of lysis solution (50 mM Tris^HCl, 2% Triton X-100, 63 mM EDTA, pH 8.0) was added and the mixture was stirred for 15 min at 4³C. The cell breaking was enhanced by six cycles of sonication (30 s on, 1 min o¡) on ice (Branson sonicator, model B-15, intensity 4.5). The sample was centrifuged at 2000Ug, 4³C, for 20 min and the pellet was resuspended in 10 ml of 50 mM Tris^HCl, 1 mM EDTA, pH 8, divided into aliquots of 1 ml and centrifuged at 1800Ug, 4³C, for 20 min. This wash was repeated two times and the inclusion bodies were lastly dried by a centrifugation step of 5 min at 9500Ug, 4³C, and stored as dry pellets at 4³C. Each aliquot of PsyRep* inclusion bodies (1.75 mg) was completely dissolved in 25 ml of ice-cold denaturing solution (50 mM Tris^HCl, 6 M guanidinium chloride, 1 mM
FEMSLE 9881 18-4-01
A. Duilio et al. / FEMS Microbiology Letters 198 (2001) 49^55
dithiothreitol (DTT), pH 8.0) and then diluted 10 times in ice-cold refolding bu¡er (10 mM Tris^HCl, 100 mM KCl, 10 mM Mg(OAc)2 , pH 8.2). The protein solution was then concentrated by ultra¢ltration on an Amicon PM-10 membrane (diameter 63 mm) to a ¢nal volume of 20 ml and dialyzed two times against 2 l of ice-cold refolding solution at 4³C. Protein concentration was determined using the BioRad protein assay with bovine serum albumin as the standard. 2.5. In-gel digestion and MALDI mass mapping of recombinant PsyRep* The PsyRep containing band was excised from an SDS^ PAGE gel, washed in acetonitrile and 0.1 M NH4 HCO3 by turns, reduced and S-alkylated with iodoacetamide. The treated band was then incubated with an excess of bovine trypsin or of Staphylococcus aureus V-8 protease (Sigma) overnight at 37³C. The resulting peptide mixture was eluted from the gel and a fraction of 0.3 Wl out of a total of 100 Wl digest solution was used for matrix assisted laser desorption ionization time of £ight (MALDI-TOF) mass spectrometric analysis. MALDI mass spectra were recorded using a Voyager DE MALDI-TOF mass spectrometer equipped with a 337-nm N2 laser and a delay extraction device (PE Biosystems). A mixture of analyte solution, K-cyano-4-hydroxycinnamic acid (10 mg ml31 in CH3 CN/EtOH/0.1% tri£uoroacetic acid, 1:1:1 v/v) and bovine insulin was applied to the sample plate and air dried. Mass calibration was performed using the molecular ions from the bovine insulin at 5734.5 Da and the matrix peak at 379.1 Da as internal standards. Raw data were analyzed using computer software provided by the manufacturer and are reported as average masses. 2.6. Relaxation reactions Reaction mixture (30 Wl) typically contained 10 mM Tris^HCl, 100 mM KCl, 10 mM Mg(OAc)2 pH 8.2, 20% ethylene glycol, 160 ng of supercoiled plasmid substrate (pTAUp/Dw, pMtBL), and di¡erent amounts of PsyRep* were added. After incubation at 4³C for 3.5 h, reactions were stopped by addition of EDTA at a ¢nal concentration of 20 mM, and the samples were loaded on 0.8% agarose gels. Electrophoresis was carried out at room temperature in TAE 1Ubu¡er [10] containing 1 Wg ml31 ethidium bromide for 2 h at 6 V cm31 . 3. Results 3.1. Expression of recombinant Psyrep in E. coli Following the molecular characterization of a plasmid
51
(pTAUp) from Psychrobacter sp. TA144, encoding a replication enzyme (PsyRep) [5], we attempted to over-produce PsyRep in E. coli using many commercially available expression systems. These e¡orts failed since we were unable to the detect the over-expressed product (data not shown). However, during the setting up of one of the tested expression systems (pQE60 vector), we succeeded in obtaining a considerable amount of a recombinant product. The 1008 bp long PsyRep encoding ORF was ampli¢ed by PCR by using as primers two synthetic oligonucleotides designed to insert the NcoI or the BglII restriction sites overlapping the Rep start and stop codons, respectively. In particular, the insertion of the BglII site resulted in the removal of the ORF natural stop codon, allowing the expression of the recombinant PsyRep as a fusion protein with the C-terminal His6 moiety encoded by the vector (see Fig. 1, PsyRepFUS). By using the newly inserted restriction sites, the ampli¢ed coding region was cloned downstream of the IPTG inducible promoter in the pQE60 expression vector, and the resulting plasmid (pQErep) was used to transform the E. coli host strain RB791. Several independent recombinant clones were tested for their ability to over-produce PsyRep but, although di¡erent expression conditions (i.e. growth temperature, cell density, and induction time) were tested (data not shown), only one clone was positive for the production of the recombinant enzyme (see Fig. 2). However, in this positive clone the product was always associated with the insoluble fraction of the extract (Fig. 2, lane 4). DNA sequencing of the ampli¢ed ORFs inserted either into the producing vector or into a nonproducing one revealed that the latter was correctly synthesized during the PCR reaction, while the producing construction contained a deletion of G at position 967 (see Fig. 1). This spontaneous mutation resulted in the frameshift of the translation frame that a¡ected the C-terminal portion of PsyRep from Ala322 to the C-terminal end. Translation of this construct resulted in the production of a mutant protein (named PsyRep*), 332 residues long, in which the Cterminus sequence has changed to Glu-Tyr-Leu-Leu-GlyHis-Thr-Ser-Lys-Asp. Since the mutant enzyme was expressed in E. coli as inclusion bodies, PsyRep* production was carried out routinely in cells grown at 37³C and recovered after 4 h of IPTG induction, which started at OD600 2. 3.2. Puri¢cation and refolding of recombinant PsyRep* The cell pellet collected from 200 ml culture of induced recombinant RB791 E. coli cells was disrupted by sonication and the PsyRep* inclusion bodies were puri¢ed from the homogenate by di¡erential centrifugation. The insoluble fraction was dissolved in ice-cold denaturing bu¡er containing 6 M guanidinium chloride and 1 mM DTT at a ¢nal protein concentration of 0.07 mg ml31 . PsyRep* refolding was obtained by a 10-fold rapid dilution in cold
FEMSLE 9881 18-4-01
52
A. Duilio et al. / FEMS Microbiology Letters 198 (2001) 49^55
Fig. 1. Nucleotide sequence of the ampli¢ed Psyrep gene cloned into the pQE60 vector and its translation. The amino acid sequences underlined represent the PsyRep* peptides, resulting from the protein digestion by either trypsin or V-8 proteases, assigned by MALDI mass spectrometry. The PsyRep active site is highlighted in gray, while the pTAUp nick site is indicated by a ¢lled box. PsyRepFUS is the fusion protein between PsyRep and the His6 tag epitope ; PsyRep* is the frameshifted product derived from the translation of an ORF containing the deletion of nucleotide 967.
refolding bu¡er, followed by a dialysis step for removing the denaturant, resulting in 60% recovery of soluble protein (see Fig. 2, lane 5). The renaturation process was monitored by recording the £uorescence spectra of the protein before and after the removal of the denaturing agent. Both the decrease in £uorescence intensity and the change of the maximum emission wavelength from 354.5 to 345.5 nm suggested that the protein underwent a structural rearrangement (data not shown). 3.3. Structural characterization of recombinant PsyRep* The soluble protein was fractionated by SDS^PAGE and the PsyRep* band subjected to in situ reduction, carboxyamidomethylation, and digestion with either trypsin or V-8 protease. The resulting peptide mixtures were extracted from the gel and directly analyzed by MALDI mass spectrometry without any puri¢cation step. All mass signals were assigned to the corresponding peptides within the PsyRep* sequence on the basis of their mass value and the speci¢city of the proteolytic enzyme used,
Fig. 2. Expression of recombinant PsyRep* in E. coli. RB791 E. coli cells harboring the pQE-rep* vector, were induced with IPTG as described in Section 2. Protein samples were resolved on a SDS^10% polyacrylamide gel and stained with Coomassie brilliant blue. IPTG induction determined the appearance of an extra band in the whole cell lysate (lane 2) with respect to the same sample from the uninduced cells (lane 1). Lane 3 corresponds to the soluble lysate fraction, while lane 4 shows the inclusion bodies fraction. Lane 5, soluble PsyRep* after refolding. The location of PsyRep* is indicated by arrows. M, molecular mass markers.
FEMSLE 9881 18-4-01
A. Duilio et al. / FEMS Microbiology Letters 198 (2001) 49^55
leading to the veri¢cation of about 80% of the protein primary structure, as reported in Fig. 1. The mass signal at m/z 1842.9 con¢rmed the occurrence of the frameshift mutation, since this value could not be assigned to any peptide in the PsyRepFUS sequence (Fig. 1), while it is in perfect agreement with the expected m/z 1843.1 value for the mutant C-terminal peptide 316^332 of PsyRep* (Fig. 1). 3.4. Functional studies Although mutated in the last 14 amino acid residues of its C-terminus, the refolded pTAUp PsyRep* was tested for its nicking activity in vitro. A relaxation assay was set up modifying the conditions reported in [13], as described in Section 2. Di¡erent amounts of refolded PsyRep* were incubated at 4³C with about 150 ng of pTAUp/Dw plasmid preparation, and the reaction products were resolved by agarose gel electrophoresis. As shown in Fig. 3A, PsyRep* exhibited a nicking activity on both TA144 natural plasmids since it induced a change of substrate topology from a supercoiled conformation to an open circular one in a dose dependent manner. Interestingly, pTADw was nicked by the recombinant enzyme as well, and this result could indicate that PsyRep is responsible for the replication of both TA144 plasmids
Fig. 3. Relaxation of pTAUp/pTADw (A) and pMtBL (B) DNA by the refolded PsyRep* protein. A: E¡ect of increasing PsyRep* on the Psychrobacter sp. TA144 plasmids; lane 1, no PsyRep*; lane 2, 240 ng; lane 3, 360 ng; lane 4, 600 ng. B: E¡ect of increasing PsyRep* on the pMtBL plasmid (9); lane 1, no PsyRep*; lane 2, 360 ng; lane 3, 600 ng. M, molecular mass marker; sc, supercoiled form; oc, open circular form.
53
in vivo. Furthermore, the same relaxation assay was carried out on an unrelated substrate, plasmid pMtBL from the psychrophilic bacterium Pseudoalteromonas haloplanktis TAC 125 [9]. As shown in Fig. 3B, pMtBL was nicked by PsyRep* as well, suggesting that the mutated enzyme could have lost its substrate speci¢city. 4. Discussion In this paper we report the expression of a RCR initiator of the pTAUp subfamily and its functional characterization. To obtain the recombinant protein, we attempted to over-produce PsyRep in an IPTG inducible system as a fusion protein with the His6 tag epitope at its C-terminal end. Several independent expression constructs were tested for their ability to induce PsyRep production but accumulation of the recombinant enzyme was achieved only in the cell extract of one of them. Further analysis demonstrated that the psychrophilic product was always associated with the insoluble fraction, although the expression was carried out in di¡erent experimental conditions. Structural characterization of the recombinant enzyme at the level of both its coding gene and its amino acid sequence revealed that PCR synthesis caused a nucleotide deletion (position 967) that resulted in a change of the protein's last 14 amino acid residues (PsyRep*). This serendipitous mutation and the consequent change in the expressed product is responsible for PsyRep* precipitation and aggregation, since E. coli recombinant cells harboring the correct expression construct (pQE-rep) or other commercially available expression vectors are unable to produce the wildtype PsyRep protein either in soluble or in insoluble form. Since the pTAUp Ori+ is actually contained in the PsyRep coding ORF [5], interaction between the expressed product and its cognate target is indeed possible even in E. coli. This in vivo interaction may cause several e¡ects, one of which is the auto-regulation of the PsyRep population by a mechanism possibly similar to that reported for the plasmids pT181 [14] and pUB110 [15]. This attractive hypothesis is currently under investigation. The initiator proteins of RCP have origin speci¢c DNA binding and nicking^closing activities. It is highly likely that these enzymes display a modular arrangement with well-de¢ned regions involved in the above functions [13,16^18]. In particular, Rep proteins belonging to the same family are characterized by conserved nicking domains, containing the tyrosine residue of the active site, and variable regions responsible for the DNA binding activity, which in the initiators of the pT181 family are located at their carboxy-terminal end [19^21]. The same structural organization has also been reported for the Rep proteins belonging to the pTAUp subfamily, in which the putative nicking active site was positioned in a region corresponding to residues 155^172 of the PsyRep amino acid
FEMSLE 9881 18-4-01
54
A. Duilio et al. / FEMS Microbiology Letters 198 (2001) 49^55
sequence (see Fig. 1) [5]. In this context, it seemed interesting to test if the PsyRep* mutation, i.e. the change of the last 14 C-terminal amino acids, alters the enzyme activity. Several protocols for the PsyRep* refolding were tested and the highest yield (about 60%) was achieved by the rapid dilution of the denaturing agent in which the puri¢ed inclusion bodies were dissolved. A relaxation assay was set up to monitor the activity of the soluble enzyme in the presence of the TA144 natural plasmids as substrate (pTAUp and pTADw, Fig. 3A), and taking into account the cold adaptation of the source strain, the experiments were carried out at low temperature (i.e. 4³C) only. The change in substrate topology demonstrates that the refolded PsyRep* exhibits a nicking activity on both TA144 plasmids in a dose dependent manner. Although this result could be suggestive of a common role for PsyRep as initiator in the replication of either pTAUp or pTADw, this hypothesis is not supported by the structural comparison of the two plasmids. In fact, pTADw does not contain any secondary structure identical or at least similar to the pTAUp Ori+ region [5]. Furthermore, the mutant psychrophilic enzyme exerts its nicking activity even on an unrelated supercoiled DNA (pMtBL, Fig. 3B). Taken together, these functional results indicate that the change of its last 14 C-terminal amino acid residues altered or even destroyed the DNA recognition and binding properties of PsyRep*. This hypothesis is in agreement with the observed inability of the mutated enzyme to catalyze the second step of topoisomerase activity, i.e. the closing reaction, which should have generated circular relaxed dsDNA. In fact, a stable protein^DNA interaction has been reported to be a prerequisite for full topoisomerase activity (i.e. nicking^closing) of pT181 RepC in vitro [22]. The above results strongly suggest that the psychrophilic initiator protein also exhibits a modular structural organization with independent functional domains. PsyRep* may be of interest since it uncouples its two functions, i.e. it still retains its nicking activity while it seems to have lost the DNA binding speci¢city. Further work is needed to investigate if the lack of PsyRep* substrate recognition is linked to the speci¢c amino acid sequence altered in the mutant enzyme, or to a not proper folding of the C-terminal domain, impaired by the frameshift mutation. Acknowledgements This work was supported by grants of the European Union (Program COLDZYME, Contract ERB BIO4 CT96 0051 ; Program EUROCOLD, Contract ERB BIO4 CT95 0017 ; Program COLDNET, Contract ERB FMRX CT97 0131), of the Ministero dell' Universita© e della Ricerca Scienti¢ca (Progetti di Rilevante Interesse
Nazionale 1999) and of the Consiglio Nazionale delle Ricerche (CNR Contract 97.01138.PF49, Progetto Finalizzato `Biotecnologie').
References [1] Gruss, A. and Ehrich, S.D. (1989) The family of highly interrelated single-stranded deoxyribonucleic acid plasmids. Microbiol. Rev. 53, 231^241. [2] Khan, A.K. (1997) Rolling-circle replication of bacterial plasmids. Microbiol. Mol. Biol. Rev. 61, 442^455. [3] del Solar, G., Giraldo, R., Ruiz-Echevarr|©a, M.J., Espinosa, M. and D|©az-Orejas, R. (1998) Replication and control of circular bacterial plasmids. Microbiol. Mol. Biol. Rev. 62, 434^464. [4] Novick, R.P. (1998) Contrasting lifestyles of rolling circle phages and plasmids. Trends Biol. Sci. 23, 434^438. [5] Tutino, M.L., Duilio, A., Moretti, M.A., Sannia, G. and Marino, G. (2000) A rolling-circle plasmid from Psychrobacter sp. TA144: evidence for a novel Rep subfamily. Biochem. Biophys. Res. Commun. 274, 488^495. [6] Weng, S.F., Fan, Y.F., Tseng, Y.H. and Lin, J.W. (1997) Sequence analysis of the small cryptic Xanthomonas campestris pv. vescicatoria plasmid pXV64 encoding a Rep protein similar to gene II protein of phage I2-2. Biochem. Biophys. Res. Commun. 231, 121^125. [7] Stassen, A.P.M., Schoenmakers, E.F.P.M., Maoxiao, M., Schoenmakers, J.G.G. and Konings, R.N.H. (1992) Nucleotide sequence of the genome of the ¢lamentous bacteriophage I2-2: module evolution of the ¢lamentous phage genome. J. Mol. Evol. 34, 141^152. [8] Feller, G., Thiry, M., Arpigny, J.L., Mergeay, M. and Gerday, C. (1990) Lipases from psychrotrophic antarctic bacteria. FEMS Microbiol. Lett. 66, 239^244. [9] Tutino, M.L., Duilio, A., Fontanella, B., Moretti, M.A., Sannia, G. and Marino, G. (1999) Plasmids from Antarctic bacteria. In: ColdAdapted Organisms : Ecology, Physiology, Enzymology and Molecular Biology (Margesin, R., and Schinner, F., Eds.), pp. 335^348. Springer Verlag, Berlin. [10] Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) Molecular Cloning: A Laboratory Manual, 2nd edn. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. [11] Hanahan, D. (1983) Studies on transformation of Escherichia coli with plasmids. J. Mol. Biol. 166, 557^580. [12] Andreotti, G., Tutino, M.L., Sannia, G., Marino, G. and Cubellis, M.V. (1994) Indole-3-glycerol-phosphate synthetase from Sulfolobus solfataricus as a model for studying thermostable TIM-barrel enzymes. Biochim. Biophys. Acta 1208, 310^315. [13] Koepsel, R.R., Murray, R.W., Rosenblum, W.D. and Khan, S.A. (1985) The replication initiator protein of plasmid pT181 has sequence-speci¢c endonuclease and topoisomerase-like activities. Proc. Natl. Acad. Sci. USA 82, 6845^6849. [14] Iordanescu, S. (1995) Plasmid pT181 replication is decreased at high levels of RepC per plasmid copy. Mol. Microbiol. 16, 477^484. [15] Muller, A.K., Rojo, F. and Alonso, J.C. (1995) The level of the pUB110 replication initiator protein is autoregulated, which provides an additional control for plasmid copy number. Nucleic Acids Res. 23, 1894^1900. [16] Gros, M.F., te Riele, H. and Ehrlich, S.D. (1987) Rolling circle replication of single-stranded DNA plasmid pC194. EMBO J. 6, 3863^ 3869. [17] Moscoso, M., del Solar, G. and Espinosa, M. (1995) In vitro recognition of the replication origin of pLS1 and of plasmids of the pLS1 family by the RepB initiator protein. J. Bacteriol. 177, 7041^ 7049. [18] Chang, T.L., Kramer, M.G., Ansari, R.A. and Khan, S.A. (2000) Role of individual monomers of a dimeric initiator protein in the
FEMSLE 9881 18-4-01
A. Duilio et al. / FEMS Microbiology Letters 198 (2001) 49^55 initiation and termination of plasmid rolling circle replication. J. Biol. Chem. 275, 13529^13534. [19] Thomas, C.D., Nikiforov, T.T., Connolly, B.A. and Shaw, W.V. (1995) Determination of sequence speci¢city between a plasmid replication initiator protein and the origin of replication. J. Mol. Biol. 254, 381^391. [20] Dempsey, L.A., Birch, P. and Khan, S.A. (1992) Six amino acids determine the sequence-speci¢c DNA binding and replication specificity of the initiator proteins of the pT181 family. J. Biol. Chem. 267, 24538^24543.
55
[21] Wang, P.Z., Projan, S.J., Henriquez, V. and Novick, R.P. (1992) Speci¢city of origin recognition by replication initiator protein in plasmids of the pT181 family is determined by a six amino acid residue element. J. Mol. Biol. 223, 145^158. [22] Dempsey, L.A., Birch, P. and Khan, S.A. (1992) Uncoupling of the DNA topoisomerase and replication activities of an initiator protein. Proc. Natl. Acad. Sci. USA 89, 3083^3087.
FEMSLE 9881 18-4-01