Peculiarities of the Regulation of Gene Expression in the Ecl18kI Restriction–Modification System

Transcription regulation in bacterial restriction–modification (R–M) systems is an important process, which provides coordinated expression levels of tandem enzymes, DNA methyltransferase (MTase) and restriction endonuclease (RE) protecting cells against penetration of alien DNA. The present study focuses on (cytosine-5)-DNA methyltransferase Ecl18kI (M.Ecl18kI), which is almost identical to DNA methyltransferase SsoII (M.SsoII) in terms of its structure and properties. Each of these enzymes inhibits expression of the intrinsic gene and activates expression of the corresponding RE gene via binding to the regulatory site in the promoter region of these genes. In the present work, complex formation of M.Ecl18kI and RNA polymerase from Escherichia сoli with the promoter regions of the MTase and RE genes is studied. The mechanism of regulation of gene expression in the Ecl18kI R–M system is thoroughly investigated. M.Ecl18kI and RNA polymerase are shown to compete for binding to the promoter region. However, no direct contacts between M.Ecl18kI and RNA polymerase are detected. The properties of M.Ecl18kI and M.SsoII mutants are studied. Amino acid substitutions in the N-terminal region of M.Ecl18kI, which performs the regulatory function, are shown to influence not only M.Ecl18kI capability to interact with the regulatory site and to act as a transcription factor, but also its ability to bind and methylate the substrate DNA. The loss of methylation activity does not prevent MTase from performing its regulatory function and even increases its affinity to the regulatory site. However, the presence of the domain responsible for methylation in the M.Ecl18kI molecule is necessary for M.Ecl18kI to perform its regulatory function.


INTRODUCTION
restriction-modification (r-M) systems are abundant in bacterial cells; they contain genes that encode restriction endonucleases (re) and DnA methyltransferases (Mtases). re hydrolyzes a certain sequence in a double-stranded DnA (dsDnA), while Mtase meth-ylates the same sequence at a strictly determined position, thus preventing its cleavage by re. the r-M functions as a primitive immune system that protects a host bacterium from penetration by alien DnA: re hydrolyses the intruding DnA that is not methylated by the corresponding Mtase [1]. the activity levels of the re and the Mtase in the cell are to be strictly coordinated. An extremely low level of the Mtase gene expression as compared with the re gene may cause cell death via hydrolysis of cellular DnA, whereas its excessively high level cannot protect the cell against penetration by an alien DnA.
Although there is no doubt that gene expression in the r-M systems is regulated, the mechanisms underlying this process are poorly studied. It has been demonstrated by recent research that coordinated gene expression in r-M systems is presumably determined by regulation at the transcriptional level. three major types of regulation can be distinguished: via c-proteins, via methylation of the promoter region of the r-M system by the Mtase, and via the interaction between the Mtase and the regulatory sites in DnA, which differ from the methylation site [2]. this study focuses on the latter type of regulation, which is typical of (cytosine-5)-DnA Mtases (enzymes that methylate the cytosine residue at position 5) belonging to the type II r-M systems. Over 300 (cytosine-5)-DnA Mtases have been recently characterized; however, the existence of the regulatory function has been experimentally confirmed only for six of them (M.MspI, M.ecorII, M.ScrFIA, M1.LlaJI, M.SsoII, and M.ecl18kI) [2].
the type II r-M system SsoII has been most thoroughly studied. the genes of this system are located in natural plasmid P4 (4250 bp) from the Shigella sonnei 47 strain; they are divergently oriented; the intergenic region consists of 109 bp [3]. the other four SsoII-like r-M systems isolated from various bacterial strains have been described; their Mtases are either identical to M.SsoII in terms of the amino acid sequence (M.Kpn2kI from Klebsiella pneumoniae 2k) or differ insignificantly. thus, Mtases ecl18kI from Enterobacter cloacae 18k and StyD4I from Salmonella typhi D4 carry Met instead of Ile at position 56, while Mtase SenPI from Salmonella enteritidis P1 contains Ile56 and, in addition, Gly instead of Glu at position 11 [4][5][6][7]. the nucleotide sequences of the corresponding genes share 99-100% identity; those of the intergenic regions are absolutely identical. Hence, the data on the functioning of the enzymes from one of these systems can be extrapolated to the other systems as well.
All SsoII-like r-M systems recognize sequence 5'-ccnGG-3'/3'-GGncc-5' (n = A, G, c or t) in dsDnA and methylate the inner c residue in this sequence in the presence of the cofactor S-adenosyl-Lmethionine (AdoMet) forming 5-methyl-2'-deoxycytidine [4]. the promoter elements of the genes encoding the re and Mtase of the SsoII-like r-M systems have been determined using the ecl18kI system as an example; the in vitro transcriptional regulation of these genes by M.ecl18kI has been also shown. In or-der to regulate transcription, M.ecl18kI binds to the so-called regulatory site, the 15-mer inverted repeat 5'-GGAcAAAttGtcct-3'/3'-cctGttTAAcAG-GA-5', which is localized inside the promoter region of the genes of the ecl18kI r-M system [9]. the nucleotides that participate in the formation of specific DnA-protein contacts with the Mtase are located inside the regulatory site ( Fig. 1) [10,11]. All SsoII-like Mtases are two-domain proteins whose n-terminal region (residues 1-71) provides transcriptional regulation, while the region of 72-379 residues is responsible for DnA methylation. the n-terminal region of M.SsoII has been shown to have a strongly pronounced secondary structure [12] in which the "helix-turnhelix" (HtH) motif is predicted with a high probability. two M.SsoII molecules (which are monomeric in the apo-form) interact with the regulatory site [12]. the data [13] regarding the putative contacts in the complex between the M.SsoII n-terminal region and the regulatory site are summarized in Fig. 1.
In order to refine the mechanism of gene transcription regulation in the SsoII-like r-M systems, the efficiency of complex formation of M.ecl18kI and E. coli rnA polymerase (rnAP) with DnA fragments containing the regulatory elements of the genes of the ecl18kI r-M system is assessed in this study. All known SsoII-like r-M systems have been isolated from various enterobacterial strains (E. coli belonging to them as well); thus, the use of E. coli rnAP is reasonable. the role of residues Lys21, Lys31, Lys46, and Lys53 in the М.ecl18kI n-terminal region for the binding of this protein to the regulatory site, as well as their effect on the Mtase ability to act as a transcription factor and on the interaction between the enzyme and the methylation site are being studied for the first time.

Protein purification
Mtase ecl18kI and its mutant forms were purified by affinity chromatography on ni-ntA agarose [4]. E. coli rnA polymerase was sequentially purified by ni-ntA-agarose and heparin-sepharose affinity chromatography, followed by DeAe cellulose ion-exchange chromatography [14].
Equilibrium binding of the proteins to the DNA ligands the 5'-ends of the oligonucleotides were radioactively labeled using t4 polynucleotide kinase (10 units, Fermentas, Lithuania) and [γ-32 Р]АТP. the complex formation between the М.ecl18kI and DnA fragments I-II, as well as between the rnAP and DnA fragments I-III, was conducted in 10 µl of the binding buffer (50 mM tris-Hcl (pH 7.6), 150 mM nacl, 5 mM β-mercaptoethanol) in the presence of heparin (equimolar amount to the protein) for 40 min at 37°С. In the case of M.ecl18kI, the reaction mixture contained 1 mM AdoMet. the DnA-protein complex and the unbound DnA duplex were separated by gel electrophoresis in 1% agarose gel. After the electrophoresis, the agarose gels were dried on a supporting plate at 90 о С in a hot air flow. the dissociation constants (K d ) of the DnA-protein complexes were determined by the Scatchard technique [15]. the concentrations of М.ecl18kI and rnAP were 60 and 30 nM, respectively. the concentrations of the DnA duplex II were varied within a range from 5 to 120 nM. the complex formation of the mutants М.ecl18kI(K46A), М.ecl18kI(K53A), and М.ecl18kI(K21A) with the DnA fragments IV and V was conducted in 20 µl of the binding buffer (50 mM tris-Hcl (pH 7.6), 150 mM nacl, 5 mM Dtt, 50 ng/µl poly(dI·dc)) for 20 min at 37°С. the concentrations of the DnA duplexes IV and V were varied within a range from 20 to 100 nM. the concentrations of М.ecl18kI(K46A), М.ecl18kI(K53A), and М.ecl18kI(K21A) were equal to 560, 400, and 400 nM, respectively, when binding to the DnA fragment IV and were equal to 200, 1600, and 5600 nM, respectively, when binding to the DnA fragment V.
Determination of the initial rate of the substrate DNA methylation the initial rate of the substrate DnA methylation by Mtases ecl18kI, SsoII, and their mutant forms was determined as previously described [9], on the basis of the degree of the duplex V "protection" against hydrolysis by re ecl18kI (r.ecl18kI). For this purpose, 350 nM of the radiolabeled DnA duplex V was incubated with Mtase in the binding buffer containing 1 mM AdoMet for 0.5-60 min at 37°c. the reaction mixture was then kept at 65°c for 10 min to inactivate the enzyme, and cooled to 25°c. next, Mgcl 2 (up to 10 mM) and r.ecl18kI (up to 240 nM) were added and the reaction mixture was incubated at 37°c for 1 h. the initial active concentrations of the Mtases were identical (14 nM). the degree of hydrolysis of the unmethylated DnA duplex V by r.ecl18kI was taken as 100%. the degree of methylation of the DnA duplex V by the Mtases was calculated with respect to this value, and the kinetic curves were plotted. the initial methylation rate (v 0 ) of the DnA duplex V by Mtase was calculated as an angular coefficient (slope ratio) of the initial linear region on the kinetic curve.
Characterization of the regulatory activity of the methyltransferases the regulatory activity of the mutant forms of M.ecl18kI and M.SsoII was assessed via in vitro transcription from the DnA fragment I in the presence of the corresponding proteins. the wild-type M.ecl18kI or M.SsoII were used in the control experiments. the reaction mixtures were analyzed by 5% polyacrylamide gel electrophoresis (PAGe; the gel contained 7 M urea) at a field intensity of 5 V/cm in tBe buffer. Only the resulting rnA transcripts contained the radiolabel. In the presence of the SsoII-like Mtases capable of acting as regulatory proteins, the following changes were observed: an increase in the radioactivity of the region corresponding to the rnA transcript from the re gene promoter and a decrease in the radioactivity of the region corresponding to the rnA transcript from the Mtase gene promoter. the fraction (%) of the re gene transcript in the total radioactivity of the resulting transcripts (taken as 100%) at various Mtase concentrations was determined. Identical active concentrations of the Mtases were used to ensure a correct comparison of the yields of the transcription products in the reaction. they were obtained from the Scatchard plots used to determine the K d values for the complexes between the proteins and the duplex IV containing the regulatory site [15]. the fraction of the transcript from the re gene promoter was plotted as a function of the Mtase active concentration. the relative yield of this transcript per unit of the Mtase active concentration was then determined. For this purpose, the ratio between the angular coefficients (slope ratios) of the initial linear region on the curves of the mutant Mtase and the wild-type M.ecl18kI (or M.SsoII) was calculated.

RESULTS AND DISCUSSION
Complex formation of RNA polymerase and M.Ecl18kI with the DNA fragments containing the intergenic region of the Ecl18kI R-M system Figure 2 shows the genetic arrangement of the ecl18kI r-M system (based on the data [8,11]) by the example of the 247-bp DnA fragment I. the Mtase gene pro- moter is localized directly before the regulatory site and partially overlaps the region which is protected by M.ecl18kI from DnAse I cleavage. We had assumed that the mechanism of negative regulation of the Mtase gene expression may consist in physical blocking of the rnAP access to the Mtase gene promoter as М.ecl18kI binds to the regulatory site. to verify this hypothesis, complex formation of both proteins with the 116-bp DnA fragment II containing the intergenic sequence of the ecl18kI r-M system (the regulatory site, the transcription initiation point, and the promoter elements of the Mtase gene ecl18kIM) but lacking the promoter elements of the re gene ecl18kIR was studied (Fig. 3A). After rnAP was added to the Mtase-DnA mixture, no other complexes but Mtase-DnA and rnAP-DnA emerged in the reaction mixture. this fact eliminates the possibility of direct contact between М.ecl18kI and rnAP. Moreover, the 5-fold excess of М.ecl18kI (with respect to rnAP) resulted in virtually complete disappearance of the rnAP-DnA complex. therefore, Mtase binding to the regulatory site does impede the interaction between rnAP and the promoter region of the SsoII r-M system genes (Fig. 3B).
the efficiency of rnAP and M.ecl18kI binding to the Mtase promoter and to the regulatory site was assessed by determining the K d values of the DnAprotein complexes. K d = 12 ± 1 nM for the M.ecl18kI complex with the DnA fragment II, while K d = 25 ± 1 nM for the rnAP complex with the same fragment. thus, the control over the Mtase expression level can be attributed to the competition between rnAP and М.ecl18kI for the binding site. the insignificant (2fold) difference in the Mtase and rnAP affinity to this DnA region allows preventing premature inhibition of M.ecl18kI synthesis, i.e. controlling the expression level of the Mtase gene more accurately. thus, the level of M.ecl18kI synthesis does not fall below the minimal value that ensures maintenance of the specific methylation of cellular DnA.
Since the Mtase is localized near the transcription initiation point of the re gene (Fig. 2), it seems quite possible that M.ecl18kI has a negative effect on the ecl-18kIR gene transcription. However, an opposite effect is observed. We had assumed that rnAP and M.ecl18kI could be bound simultaneously to the same DnA fragment, rnAP interacting with the re gene promoter, while the Mtase interacts with its regulatory site. this assumption is verified experimentally (Fig. 3C,D): sequential addition of rnAP and M.ecl18kI to the 247bp DnA fragment I leads to a ternary complex formation (supposedly rnAP-М.ecl18kI-DnA) which has a lower electrophoretic mobility as compared with the rnAP-DnA and М.ecl18kI-DnA complexes. Since two M.SsoII molecules bind to a single regulatory site [12], it is highly probable that each complex (rnAP-М. ecl18kI-DnA and М.ecl18kI-DnA) contains two М.ecl18kI molecules.
complex formation between rnAP and the two different promoters shows that the degree of rnAP binding to the DnA fragment III (Fig. 3E,F), which contains the transcription initiation point and the promoter regions of the ecl18kIR gene only, is 4-fold lower than the degree of rnAP binding to the DnA fragment II, which contains the transcription initiation point and promoter regions of the ecl18kIM gene only. thus, the ecl18kIM gene promoter is stronger than the ecl18kIR gene promoter, and transcription primarily occurs from the Mtase gene promoter in the absence of М.ecl18kI. this phenomenon can also be stipulated by the "sitting duck" mechanism of transcriptional interference [16] when the rates of the open rnAP complex transition into the elongation form for two closely spaced promoters differ considerably and the activity of the weaker promoter is suppressed due to the intensive transcription of the stronger one.

Analysis of the ability of the М.Ecl18kI N-terminal region to regulate gene transcription in the restriction-modification system in vitro
the experiments with deletion mutants have demonstrated that the M.SsoII ability to act as a transcription factor can be attributed to the n-terminal region of this protein, which consists of 71 residues [3]. the amino acid sequences of the n-terminal regions of M.ecl18kI and M.SsoII significantly resemble c-proteins. When comparing the M.SsoII regulatory site with the idealized sequence of c-boxes (5'-GAct...AGtc-3') [17], 6 out of 8 nucleotides coincide. considering the significant variability among the sequences of the c-boxes, the regulatory site recognized by M.ecl18kI can also be classified as a c-box. the deletion mutant Δ(72-379) M.ecl18kI, which is the n-terminal region of M.ecl18kI, retains its strongly pronounced secondary structure and is capable of specific binding to the DnA containing the regulatory site; however, the efficiency of such binding is an order of magnitude lower than that of the full-length protein [12]. the effect of ∆(72-379)M.ecl18kI on the in vitro transcription of the ecl18kIR and ecl18kIM genes has been studied. the full-length M.ecl18kI was used in the control experiment (Fig. 4). transcription from the 247-bp DnA fragment I resulted in two products corresponding to the transcripts from the re gene promoter (~190 nucleotides) and from the Mtase gene promoter (~110 nucleotides). When the reaction mixture was titrated with increasing amounts of М.ecl18kI, the fraction of the Mtase gene transcript decreased considerably,  Fig. 3. Complex formation of М.Ecl18kI and RNA polymerase with the DNA fragments that contain different elements of the intergenic region of the Ecl18kI restriction-modification system. A, C, E -schematic representations of the DNAprotein complexes formation. The directions of the MTase and the RE genes are shown with yellow and green arrows, respectively. P R , P M -transcription initiation points of the RE and MTase genes, respectively (also marked with thin arrows). The promoter elements are shown in blue, the regulatory site is shown in red. B, D -complex formation of RNA polymerase (30 nM) with the DNA fragments II or I, respectively (15 nM) in the presence or absence of М.Ecl18kI excess (150 nM) under specific binding conditions (with 300 nM heparin). F -complex formation between RNA polymerase (190 nM) and the DNA fragments III or II (30 nM). Radioautographs of 1% agarose gels while that of the re gene transcript increased (Figs. 4,5). Meanwhile, the addition of ∆(72-379)M.ecl18kI to the reaction mixture caused no changes in the ratio between the yields of the two transcripts; i.e., this deletion mutant could not function as a transcription factor (Fig. 5). this fact is probably due to the low affinity of ∆(72-379)M.ecl18kI to the DnA carrying the regulatory site [12]: such a protein cannot efficiently compete with rnAP for binding to the promoter region. It is also possible that the deletion mutant covers a considerably smaller DnA fragment as compared with the full-length M.ecl18kI and therefore is not a steric impediment for rnAP. thus, the region responsible for methylation is necessary to maintain the regulatory function of M.ecl18kI. this result agrees with the recently proposed structural model of the complex between the SsoII-like Mtases and the regulatory site within the intergenic region of the r-M system: the n-terminal regions of both Mtase molecules specifically interact with the regulatory site, while the regions responsible for methylation are nonspecifically bound to the DnA flanking the regulatory site [18].

Model of gene transcription regulation in the Ecl18kI restriction-modification system
After the r-M system penetrates into a cell, the Mtase is actively synthesized from the stronger promoter, which is required to protect cellular DnA against the hydrolysis by re. A certain amount of Mtase, which can efficiently protect the cell against bacteriophage infection, is produced with time. then, two Mtase molecules bind to the regulatory site and block the rnAP access to the promoter of the Mtase gene (Fig. 6). no complex formation between Mtase and rnAP occurs in this case; i.e., the mechanism of transcription suppression of the Mtase gene is based exclusively on the competition between the Mtase and rnAP for binding to the intergenic region of the ecl18kI r-M system. the close K d values attest to the fact that even small changes in the Mtase concentration are expected to affect the efficiency of the Mtase gene transcription.
the interaction between the M.SsoII regions responsible for methylation with the DnA flanking the regulatory site described in [18] seems to confer additional strength to the DnA-protein complex. this circumstance allows SsoII-like Mtase to successfully compete with rnAP for binding to the promoter region, resulting in the suppression of the Mtase gene transcription and stabilization of the Mtase concentration in the cell. It can be assumed that binding of the enzyme region responsible for methylation to the DnA flanking the regulatory site is a compensatory mechanism which is required to make the effect on transcription of a Mtase dimer bound to the regulatory site as efficient as that of two c-protein dimers bound to two palindromic sites in DnA. the fact that a deletion mutant, which is the n-terminal region of M.ecl18kI, does not have this "additional" interaction explains the low stability of its complex with the DnA and its inability to control transcription in the ecl18kI r-M system. Binding between M.ecl18kI and the regulatory site results in indirect activation of the re gene promoter by preventing rnAP from binding to the Mtase gene promoter. During transcription from the re gene promoter, rnAP runs against the Mtase region, which is responsible for methylation and nonspecifically interacts with the DnA region flanking the regulatory site [18]. these nonspecific DnA-protein contacts can be relatively easily destroyed by rnAP, which melts DnA in the elongation complex. It is possible that both Mtase subunits are pushed away from the DnA, which can be caused by the reduced affinity of the enzyme to the DnA melted during the elongation process.

Effect of single amino acid substitutions on the regulatory activity of the SsoIIlike methyltransferases
The mutant form of M.SsoII containing Cys142 substitution in the region responsible for methylation. cys142 in the М.ecl18kI (M.SsoII) molecule plays the key role in catalyzing the methyl group transfer from the reaction cofactor AdoMet to the substrate DnA [19]. replacement of cys142 by Ala results in loss of M.SsoII enzymatic activity. the efficiency of the mutant protein binding to the methylation site decreases; however, the mutant has a considerably higher affinity to the regulatory site (Table) [9]. the M.SsoII(c142A) ability to regulate in vitro transcription of the genes in the ecl18kI r-M system was tested in this study.
the yields of the transcripts of the ecl18kIR gene in the presence of M.SsoII, M.ecl18kI, or the mutant protein M.SsoII(c142A) are almost identical (Fig. 7, Table). Hence, loss of the methylation function does not affect Mtase's ability to function as a transcription factor.
The mutant forms of M.Ecl18kI containing substitutions in the region responsible for the regulatory function. Based on the model of the complex between the M.SsoII n-terminal region and the regulatory site [13], a hypothesis has been proposed that residues Lys21, Lys31, Arg35, Arg38, Arg39, and Arg42 interact with DnA ( Fig. 1). We studied the regulatory properties of the М.ecl18kI mutants, where one of the abovementioned residues was replaced by Ala (Table). the М.ecl18kI mutants with one of the residues (Arg15, Lys46, or Lys53) replaced by Ala were used as a control. the regulatory activity of all the M.ecl18kI mutant forms was tested by conducting in vitro transcription in the presence of these proteins. Wild-type M.ecl18kI was used in the control experiment.
It is shown for the first time that the amino acid substitutions in the n-terminal region affect the Mtase ability to regulate transcription in the r-M system (Table). An interesting and unexpected result of the study is the dynamics of the yield changes of the transcripts from the re gene promoter, which differed among different M.ecl18kI mutants at the same active concentrations. the mutants exhibiting high affinity to the regulatory site had been expected to regulate transcription more efficiently, whereas the regulation   [9]. 2 The complex formation was studied using the 31-bp DNA duplex IV containing the regulatory site: 5'-ttGGttttAGGACAATTTGTCCTGttttGat-3' 3'-aaCCaaaaTCCTGTTAAACAGGACaaaaCGa-5' (DNA duplex IV). 3 The complex formation and methylation activity were studied using the 30-bp DNA duplex V containing the methylation site: 5'-GatGCtGCCaaCCTGGCtCtaGGttCataC-3' 3'-CtaCGaCGGttGGACCGaGatCGaaGtatG-5' (DNA duplex V).
of transcription by the Mtases exhibiting lower affin-the ecl18kI restriction-modification system. the inhibition of the Mtase gene transcription is caused by competition between rnAP and the modification enzyme for the binding site near the Mtase gene promoter. transcription of the restriction endonuclease ecl18kI gene is activated due to the attenuation of transcriptional interference resulting from the modification enzyme binding to the regulatory site. It is demonstrated for the first time that the presence of the Mtase region responsible for methylation is required for this enzyme to function as a transcription factor. the point mutation turning off the Mtase catalytic function increases the mutant affinity to the regulatory sequence and does not affect its ability to act as a transcription factor. On the other hand, the mutants M.ecl18kI(K46A) and M.ecl18kI(K53A), which efficiently regulate transcrip-tion in the ecl18kI r-M system, do not modify the substrate DnA because of the extremely low affinity to the methylation site. the replacement of Arg35 or Arg38 in Mtase ecl18kI by Ala not only impairs protein binding to the regulatory site, but also impedes its performing of the regulatory function; however, the efficiency of DnA methylation is considerably enhanced in this case. evidently, there is a relationship between the functioning of the two DnA recognition centers in the SsoII-like Mtases.