The Na+-specific interaction between the LysR-type regulator, NhaR, and the nhaA gene encoding the Na+/H+ antiporter of Escherichia coli
1997; Springer Nature; Volume: 16; Issue: 19 Linguagem: Inglês
10.1093/emboj/16.19.5922
ISSN1460-2075
AutoresO. Carmel, O. Rahav-Manor, Nir Dover, Boaz Shaanan, Etana Padan,
Tópico(s)Plant nutrient uptake and metabolism
ResumoArticle1 October 1997free access The Na+-specific interaction between the LysR-type regulator, NhaR, and the nhaA gene encoding the Na+/H+ antiporter of Escherichia coli O. Carmel O. Carmel Division of Microbial and Molecular Ecology, The Institute of Life Sciences, The Hebrew University of Jerusalem, 91904 Jerusalem, Israel Search for more papers by this author O. Rahav-Manor O. Rahav-Manor Division of Microbial and Molecular Ecology, The Institute of Life Sciences, The Hebrew University of Jerusalem, 91904 Jerusalem, Israel Search for more papers by this author N. Dover N. Dover Division of Microbial and Molecular Ecology, The Institute of Life Sciences, The Hebrew University of Jerusalem, 91904 Jerusalem, Israel Search for more papers by this author B. Shaanan B. Shaanan Department of Biological Chemistry, The Institute of Life Sciences, The Hebrew University of Jerusalem, 91904 Jerusalem, Israel Search for more papers by this author E. Padan Corresponding Author E. Padan Division of Microbial and Molecular Ecology, The Institute of Life Sciences, The Hebrew University of Jerusalem, 91904 Jerusalem, Israel Search for more papers by this author O. Carmel O. Carmel Division of Microbial and Molecular Ecology, The Institute of Life Sciences, The Hebrew University of Jerusalem, 91904 Jerusalem, Israel Search for more papers by this author O. Rahav-Manor O. Rahav-Manor Division of Microbial and Molecular Ecology, The Institute of Life Sciences, The Hebrew University of Jerusalem, 91904 Jerusalem, Israel Search for more papers by this author N. Dover N. Dover Division of Microbial and Molecular Ecology, The Institute of Life Sciences, The Hebrew University of Jerusalem, 91904 Jerusalem, Israel Search for more papers by this author B. Shaanan B. Shaanan Department of Biological Chemistry, The Institute of Life Sciences, The Hebrew University of Jerusalem, 91904 Jerusalem, Israel Search for more papers by this author E. Padan Corresponding Author E. Padan Division of Microbial and Molecular Ecology, The Institute of Life Sciences, The Hebrew University of Jerusalem, 91904 Jerusalem, Israel Search for more papers by this author Author Information O. Carmel1, O. Rahav-Manor1, N. Dover1, B. Shaanan2 and E. Padan 1 1Division of Microbial and Molecular Ecology, The Institute of Life Sciences, The Hebrew University of Jerusalem, 91904 Jerusalem, Israel 2Department of Biological Chemistry, The Institute of Life Sciences, The Hebrew University of Jerusalem, 91904 Jerusalem, Israel *Corresponding author. E-mail: [email protected] The EMBO Journal (1997)16:5922-5929https://doi.org/10.1093/emboj/16.19.5922 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info We used partially purified NhaR and a highly purified His-tagged NhaR derivative to identify the cis-regulatory sequences of nhaA recognized by NhaR and to study the specific effect of Na+ on this interaction. Gel retardation assay with DNase I footprinting analysis showed that NhaR binds a region of nhaA which spans 92 bp and contains three copies of the conserved LysR-binding motif. Na+, up to 100 mM, had no effect on the binding of NhaR to nhaA. The dimethylsulfate methylation protection assay in vivo and in vitro, showed that bases G−92, G−60, G−29 and A−24 form direct contacts with NhaR; in the absence of added Na+ in vivo, these bases were protected but became exposed to methylation in a ΔnhaR strain; accordingly, these bases were protected in vitro by the purified His-tagged NhaR. 100 mM Na+, but not K+, removed the protection of G−60 conferred by His-tagged NhaR in vitro. Exposure of intact cells to 100 mM Na+, but not K+, exposed G−60. The maximal effect of Na+ in vitro was observed at 20 mM and was pH dependent, vanishing below pH 7.5. In contrast to G−60, G−92 was exposed to methylation by the ion only in vivo, suggesting a requirement for another factor existing only in vivo for this interaction. We suggest that NhaR is both sensor and transducer of the Na+ signal and that it regulates nhaA expression by undergoing a conformational change upon Na+ binding which modifies the NhaR–nhaA contact points. Introduction Salt stress is one of the most common growth-arresting factors encountered by bacteria. This stress is multifactorial since it involves stress of osmolarity, ionic strength and desiccation, as well as a specific toxic effect of Na+ on certain essential metabolic reactions, common to all cells (reviewed in Padan and Schuldiner, 1992). Accordingly, all cells have Na+-excreting systems to eliminate toxicity (Padan et al., 1989; Padan and Schuldiner, 1992, 1994, 1996) and an intricate regulatory network responsive to various aspects of the stress of salinity. We have discovered a specific Na+-responsive adaptation in Escherichia coli (Karpel et al., 1991; Rahav-Manor et al., 1992; Carmel et al., 1994) regulating nhaA, the key Na+/H+ antiporter in the tolerance of this bacterium to high Na+ and alkaline pH (in the presence of Na+) (Padan and Schuldiner, 1994, 1996). Northern analysis of nhaA mRNA (Dover et al., 1996) and study of the expression of a nhaA′–′lacZ translational fusion in cells (Karpel et al., 1991; Rahav-Manor et al., 1992) grown at various salt concentrations showed that Na+ and Li+ specifically induce nhaA transcription. Furthermore, a novel regulatory gene nhaR, which is responsible for the Na+-specific induction of nhaA, was identified (Rahav-Manor et al., 1992; Carmel et al., 1994). nhaR is located downstream of nhaA and encodes a protein (NhaR) of 34.2 kDa. NhaR is a positive regulator required, in addition to nhaA, in order to tolerate high Na+ and Li+ concentrations (Rahav-Manor et al., 1992; Carmel et al., 1994). The enhancing effect of plasmidic multicopy nhaR on the Na+-induced expression of nhaA′–′lacZ showed that NhaR works in trans and requires Na+ for its activity. A DNA mobility test showed that a cell-free extract from cells overexpressing NhaR contains a protein which binds to the DNA at the upstream region of nhaA. NhaR is homologous to a large family of positive regulators in prokaryotes, the LysR-OxyR family (Henikoff et al., 1988; Christman et al., 1989; Rahav-Manor et al., 1992). All these proteins share, at their N-terminus, conserved sequences containing a helix–turn–helix motif, implicated in DNA binding. Interestingly, several members of this large group are proteins that are involved in the response of the organism to environmental stress (Christman et al., 1989; Storz et al., 1990; Schell, 1993). We have suggested that NhaR is a component of yet another type of stress response, essential for Li+ and Na+ tolerance, of the LysR family. Recently we have shown that the NhaR-dependent regulation of nhaA is affected by H-NS (Dover et al., 1996), a major DNA-binding protein and a global regulator involved in salt stress in bacteria (Owen-Hughes et al., 1992; Ussery et al., 1994). We have purified the NhaR protein (partially) and its His-tagged derivative (to homogeneity), identified their binding sites to cis-regulatory elements of nhaA and discovered a specific effect of Na+ on the NhaR–nhaA interaction both in vivo and in vitro. Results Construction of His-tagged NhaR and purification of both NhaR and its His-tagged derivative Our previous in vivo experiments showed that NhaR is a positive regulator of nhaA, whose activity is dependent on the concentration of intracellular Na+ (Dover et al., 1996). In the present work, a direct biochemical approach has been undertaken to study the interaction between Na+, NhaR and the nhaA DNA in a molecularly defined system. For the purification of the regulatory protein, we have constructed plasmid pOCRXH. In this plasmid, nhaR is fused in-frame at its 3′ end to a sequence encoding two cleavage sites of the protease factor Xa followed by six histidines. To test whether the chimeric protein (His-tagged NhaR) is active, the plasmid was transformed into RK33Z, a strain bearing a chromosomal nhaA′–′lacZ protein fusion. For a control, we used RK33Z cells transformed with pGM42T, a plasmid harboring wild-type nhaR. As shown previously, these cells showed marked induction of β-galactosidase activity upon addition of Na+ (Rahav-Manor et al., 1992). Similar Na+ induction was obtained with transformants of a plasmid encoding the chimeric His-tagged nhaR. These results show that the His-tagged NhaR is as active as the wild-type protein in promoting in vivo Na+ induction of nhaA. The His-tagged NhaR was overexpressed (compare lane 2 with lane 1 in Figure 1A) and bound readily to the Ni2+ column. Out of the many cytoplasmic proteins (Figure 1A, lane 2) exposed to the resin, many did not bind (Figure 1A, lane 3) or were eluted by the washes at low imidazole concentrations (≤60 mM, Figure 1A, lanes 4 and 5). At 400 mM imidazole, the His-tagged NhaR eluted as a single prominent band (Figure 1A, lanes 6 and 7). As expected from its longer C-terminus, His-tagged NhaR was slightly heavier (36.2 kDa) than the native NhaR (34.2 kDa) (Figure 1A, lane 8). To assess the degree of purification, the fraction eluted from the Ni2+ column was separated by HPLC. A single homogenous band peaking at 72.5 kDa appeared, suggesting that His-tagged NhaR is a dimer. Importantly, the activity of the His-tagged NhaR was the same, whether purified in a single step by the Ni2+ column or in two steps with an additional gel filtration step. With both procedures, no more than 1% of contaminants were observed by silver staining of the proteins, suggesting a very high degree of purification. Figure 1.Overexpression and purification of His-tagged and wild-type NhaR. (A) His-tagged NhaR was overexpressed and separated on a Ni2+-NTA–agarose column as described in Materials and methods. Samples (30 μg of protein) from each fraction applied on or eluted from the column were run on SDS–PAGE to resolve the proteins. Lane 1, non-induced cells; lane 2, cells induced for 2 h; lane 3, void volume; lane 4, binding wash; lane 5, wash with 60 mM imidazole; lanes 6 and 7 elution with 400 mM imidazole; lane 8 shows partially purified native NhaR (20 amino acids shorter than His-tagged NhaR). (B) NhaR was overexpressed and specifically labeled with [35S]methionine, as described in Materials and methods. A mixture of the cell-free extracts was applied to a heparin–Sepharose column and fractions collected for determination of radioactivity (●) and protein concentrations (○). Download figure Download PowerPoint To compare the biochemical properties of His-tagged NhaR with those of the wild-type protein, we also partially purified the wild-type molecule. For this purpose, we used a mixture of cell-free extracts, one containing overexpressed NhaR and the other NhaR specifically labeled with [35S]methionine. The radioactively-labeled protein allowed the NhaR protein to be followed during the purification and allowed it to be optimized by determining the amount of 35S-labeled protein in each fraction. Figure 1B shows that fractions 21–23, highly enriched in the specifically radioactively labeled NhaR, were obtained by chromatography on a heparin–Sepharose column. This conclusion was supported both by silver staining of samples containing equal amounts of radioactive counts eluted in these fractions and by Western analysis using anti-NhaR antibody (Rahav-Manor et al., 1992). These results showed a prominent band at 34 kDa which cross-reacted with the antibody. Fraction 21–23 represented the highest enrichment of NhaR over other contaminating proteins, mainly of higher molecular weights. These fractions were pooled and used in some in vitro experiments as indicated. The other fractions which eluted before or after the peak (19, 20, 24 and 25) also contained a protein(s) of 34 kDa. However, since this protein did not cross-react with the anti-NhaR antibody, we assumed it to be a contaminant which co-purified with NhaR. Deletion mapping of the nhaA DNA region containing the regulatory signals recognized by NhaR Two promoters of nhaA were mapped previously (Karpel et al., 1991 and Figure 2A). To identify the DNA region containing the cis-elements recognized by NhaR, we PCR-amplified various sequences overlapping the nhaA promoter region (Figure 2A). Each fragment was end labeled and tested for binding to the partially purified native NhaR in a DNA gel retardation assay (Figure 2B). As shown previously with a cell-free extract obtained from cells overexpressing native NhaR (Carmel et al., 1994), the partially purified NhaR binds specifically to a DNA fragment containing base pairs −424 to 130 of the upstream sequences of nhaA including the nhaA promoters (Figure 2, fragment A). Figure 2 also shows that whereas the sequences from the 5′ end of this fragment down to bp 121 (fragments B and E) and sequences from the 3′ end up to bp 14 (fragment D) do not bind, DNA fragments overlapping the sequences in between (fragments C, F and G) contain nhaA sequences recognized by NhaR. We have therefore concluded that the NhaR-binding sites are located between bp −120 and 14 (also indicated on the nhaA sequence in Figure 6A). In accordance with this conclusion, sequences between bp −424 and −191 did not bind but those between bp −424 and −78, −190 and 14, and −77 and 130 did (not shown). Figure 2.Deletion mapping of the DNA region containing the cis-regulatory elements of nhaA recognized by NhaR. (A) DNA fragments containing the nhaA sequences marked at their ends by the number of base pairs from the first base of the initiation codon (=1) are shown (DDBJ/EMBL/GenBank accession Nos X17311, S67239 and J03897). (B) Each fragment on its own (odd numbers), or after exposure to partially purified native NhaR (even numbers), was tested in the DNA gel retardation assay. +, retardation; −, no retardation; P1 and P2 are nhaA promoters (Karpel et al., 1991). I, II and III are the conserved LysR motifs shown in Figure 6. Numbers in brackets refer to the transcript start site and otherwise to the first base of the initiation codon GTG. Download figure Download PowerPoint The purified His-tagged NhaR was as active as NhaR in the gel retardation assay (not shown). Hence the purified His-tagged NhaR and the DNA fragments containing the NhaR-binding sites provide the essential tools needed for the study of the NhaR–nhaA molecular interaction. With the gel retardation assay, we have not found an effect of addition of Na+ or K+ (100 mM each) on the binding, either at pH 7 or at pH 8.5. DNase I footprint of NhaR on a linear DNA fragment of nhaA The sequences of nhaA protected by either NhaR (not shown) or His-tagged NhaR (Figure 3) from a limited DNase I digestion were identical. The purified His-tagged NhaR and a linear DNA fragment (from −190 to 52 of the coding sequence, Figure 6A) were used in these experiments. A reaction mixture lacking His-tagged NhaR served as a control (Figure 3A and B). As shown in Figure 3, a very long sequence on each strand of the nhaA promoter region was protected by His-tagged NhaR, extending over 92 bp [from bp −109 to −17 of the bottom strand (Figure 3A) and from −109 to −24 of the top strand (Figure 3B)]. This protected region is not continuous since it is interrupted by sites which became hypersensitive to the enzyme in the presence of NhaR (Figures 3A and B and 6A). Figure 3.DNase I protection footprint of NhaR. A DNA fragment (242 bp) end labeled with 32P at the 3′ (bottom strand) in (A) or at the 5′ (top strand) in (B) were incubated with His-tagged NhaR (250 and 500 ng in lanes b and c, respectively, of A and 500 ng in lane b of B) as indicated and then cleaved with DNase I as described in Materials and methods. The DNase I-protected nhaA regions are marked by the vertical lines adjacent to the sequence. Numbers indicate the position of each base relative to the first base of the initiation codon (Figure 6A). Download figure Download PowerPoint Addition of either Na+ (up to 100 mM) or equimolar K+ to the footprint reaction mixture had no effect on the footprint. Since Na+ contaminants can be as high as 7 mM (Carmel et al., 1994), it was considered that the system was already saturated with Na+ and therefore, further addition of the ion was without effect. To exclude this possibility, the reaction mixture was purified by gel filtration, and the Na+ concentration, as measured by atomic absorption, was reduced to 50 μM. Nevertheless, addition of Na+ or K+ (100 mM each) was still without effect on the footprint (not shown). DMS methylation protection assay in vitro Since the DNase I protection assay is limited in its resolution and DNase I attacks sequences located mainly in the minor groove of the DNA (Sasse-Dwight and Gralla, 1991), we next focused on the major groove with a more sensitive method: probing the NhaR footprint with primer extension following dimethylsulfate (DMS) methylation and subsequent breakage by piperidine of the unprotected methylated sites. DMS modifies mainly guanines and, to a lesser extent, adenines in the major groove of the DNA (Sasse-Dwight and Gralla, 1991). Figure 4A shows that G at −92 is protected specifically by His-tagged NhaR, but addition of either KCl or NaCl (100 mM each) had no effect on the protection pattern. Similarly, the bases, A at −24 and G at −29, were protected by NhaR with no effect of either ion (Figure 4A). Strikingly, the protection of G at −60 by NhaR was affected differently by the ions (Figure 4A); it remained protected in the absence or presence of 100 mM KCl (Figure 4A, lanes b and d) but 100 mM NaCl specifically removed the protection of G−60 by NhaR and exposed it to methylation and subsequent breakage (Figure 4A, compare lane f with lane d). Figure 4.DMS methylation protection by NhaR. In all panels, numbers on the left indicate the position of bases in the promoter region relative to the first base of the initiation codon (see also Figure 6A). (A) In vitro: DNA was incubated with His-tagged NhaR in the presence or absence of added KCl or NaCl as indicated in the figure, subjected to DMS methylation followed by piperidine cleavage and the products were analyzed by primer extension as described in Materials and methods. Arrows, identified bases contacting His-tagged NhaR. (B and C) In vivo: the cells used in (B) were HB101 transformed with pGM42T, a plasmid harboring all upstream sequences of an inactive nhaA and wild-type nhaR. The cells used in (C) were ORC100, a strain deleted of nhaR and transformed with either pKR107 (lanes a and b), a plasmid harboring only the upstream sequences of nhaA without nhaR, or pGM42T, an nhaR-bearing plasmid (lane c). The cells were grown in the presence of the inducer (100 mM Na+) as indicated in the figure, exposed to DMS, plasmid DNA isolated and treated with piperidine and the resulting fragments were analyzed by primer extension as described in Materials and methods. Arrows, identified bases contacting NhaR; the starred arrow points to an unreproducible NhaR-independent modification. Download figure Download PowerPoint We next titrated the Na+ concentration needed to give the specific effect of Na+. Whereas at 7 mM Na+, G−60 was as protected as in 100 mM K+, 20 mM Na+ was sufficient to give the maximal exposure to methylation and subsequent cleavage (not shown), as seen in the presence of 100 mM Na+ (Figure 4A, lanes e and f). These results suggest that the Na+ concentration yielding the maximal effect is ∼20 mM Na+. There was no effect of Na+ on the methylation reaction in the absence of NhaR (Figure 4A). The pH dependence of the Na+ effect on the methylation protection assay is summarized in Figure 5. The bases protected by NhaR which were not affected by Na+, i.e. −92, −29 and −24, were not affected by pH either. In contrast, the Na+-sensitive G−60 was affected drastically by pH; whereas at pH 6.5 it remained protected in the presence of either K+ or Na+ (100 mM each, Figure 5, lanes a–c), at pH 7.5 and pH 7.9 (Figure 5, lanes d–i), and even up to pH 9 (not shown), it was exposed to methylation in the presence of Na+ (100 mM) but not of K+ (100 mM). Figure 5.Effect of pH on the in vitro methylation protection pattern. The DNA was incubated with His-tagged NhaR at the indicated pH obtained by titration of the binding buffer with HCl, otherwise the experimental system was as in Figure 4A. Download figure Download PowerPoint Figure 6.The nhaA sequence bound by NhaR is modified by Na+. (A) The upstream DNA sequences (see Table I for accession No.) containing the cis-regulatory sequences of nhaA are shown. The shortest fragment (bp −120 to 14) binding His-tagged NhaR in the gel retardation assay (Figure 2) is delimited. The shaded sequences show the His-tagged NhaR domain protected from DNase I digestion (Figure 3). G−92 and G−60 specifically affected by Na+ in the DMS methylation assay, in vivo or both in vivo and in vitro respectively (Figure 4), are marked by dark stars. G−24 and A−29 protected by NhaR but not affected by Na+ in the DMS methylation protection assay are indicated by open stars. Open vertical arrows show the DNase I-hypersensitive sites. Three sequential consensus motifs of the lysR family (Schell, 1993) designated I, II and III are shown by interrupted lines above the nhaA sequence (see also B). Numbers in parentheses relate to the indicated promoters P1 and P2 of nhaA. Other numbers relate to the first base (=1) of the initiation codon GTG, in bold, while its upstream neighboring base = −1. (B) The generic consensus sequence of the LysR family according to (Schell, 1993). The consensus sequences recognized by NhaR which appear sequentially three times in the NhaR-binding domain and are designated I, II and III (Figure 6A) are also shown. Download figure Download PowerPoint Identification of the specific effect of Na+ on NhaR–nhaA interaction in vivo The DMS protection assay was conducted in vivo in order to identify the in vivo footprint of NhaR on nhaA. Figure 4B shows that, similarly to the in vitro results, a G at position −60 is less protected when the cells are exposed to 100 mM Na+ as compared with its exposure to 100 mM K+. Strikingly, the G at −92, which did not show any response to Na+ in vitro, is dramatically exposed to methylation when the cells are exposed to Na+ (100 mM, Figure 4B, lane a) and is not affected by an exposure to K+ (100 mM, Figure 4B, lane b). It was critical to show that these specific in vivo effects of Na+ are indeed dependent on NhaR. Support for this contention was obtained by the fact that these in vivo Na+ effects were conspicuous only in cells transformed with a multicopy plasmid bearing nhaR but not in cells having only the single chromosomal copy (not shown). Nevertheless, to prove the dependence of the Na+ effects on NhaR, we constructed a ΔnhaR strain (ORC100) and used it, either transformed or not, with plasmidic nhaR to repeat the methylation protection assay (Figure 4C). In the ΔnhaR strain, all bases at −24, −29, −60 and −92 were similarly exposed to DMS methylation when either Na+ or K+ (100 mM each) were present (Figure 4C, lanes a and b). Indeed transformation with nhaR plasmid restored protection (Figure 4C, lane c) and the specific effects shown in the presence of Na+ in vivo (not shown). Discussion Our previous in vivo studies suggested that as an essential part of Na+ homeostasis in E.coli, the regulation of nhaA expression by NhaR is induced specifically by a change in Na+ concentration rather than by its outcome: a change in ionic strength or osmolarity (Karpel et al., 1991). A similar role has been assigned recently to Na+ in the regulation of expression of the Na+/ATPase of Enterococcus hirae (Murata et al., 1996). In the present study, by molecular dissection of the system in E.coli, we have proven that indeed Na+ itself is the signal for nhaA expression via NhaR, identified the regulatory cis-elements of nhaA which bind NhaR and established both in vivo and in vitro that Na+ changes the footprint of NhaR on nhaA. Different molecular sizes were obtained in the two separation procedures of His-tagged NhaR, 36.2 kDa by SDS–PAGE and 72.5 kDa by gel filtration. The lower molecular weight value obtained under the denaturing conditions (SDS–PAGE) agrees with a monomeric form of His-tagged NhaR which, as expected, is slightly heavier than the native NhaR (34.2 kDa). The molecular weight value obtained under the non-denaturing conditions (HPLC, gel filtration) suggests that His-tagged NhaR exists as a dimer. Many of the LysR-type transcriptional regulators exist and function as dimers (Schell, 1993) although, in several cases, higher multimeric forms are also known (Toledano et al., 1994; Kullik et al., 1995). The multimeric nature of the LysR family members is reflected in the mode of binding to their DNA target promoters; the size of their binding region is unusually long, extending over several tens of base pairs, i.e. several turns of the DNA helix. The NhaR appears to be an extreme case. It protects ∼90 bp against DNase I digestion. Accordingly, the nhaA sequences binding NhaR that are revealed by the gel retardation assay (Figure 2) align with the DNase I-protected sequences (Figures 3 and 6A). Since the LysR regulatory proteins including NhaR each have only one helix–turn–helix motif in their N-terminus, through which binding to DNA is mediated, a single molecule is unlikely to span more than one helix turn. Hence, we suggest that similarly to other members of the LysR family, the His-tagged NhaR binds as a multimer in an as yet unknown NhaR–DNA stoichiometry. A peculiarity of the LysR-type proteins is the paucity of conserved bases involved in DNA binding and the fact that they are dispersed throughout their long binding site. Recently, a detailed consensus motif was defined for the binding of OxyR (Toledano et al., 1994). It shows a 2-fold symmetry, and the spacing of the elements suggests that OxyR contacts four helical turns. This motif also fits the generic LysR family consensus sequence (T-N11-A), which is based on a comparison of binding sites from a variety of species (Goethals et al., 1992; Schell, 1993; and see Figure 6B). Most interestingly, the deletion mapping of the NhaR binding domain on nhaA shows that each of the DNA fragments which bind NhaR contain one or more of these consensus motifs designated I, II and III (Figures 2A and 6A), which are very close to each other but yet separated by spanning sequences. Accordingly, the DNase I-protected sequences of nhaA by NhaR align with these three motifs and show that the spanning sequences separating them contain hypersensitive DNase I sites (Figures 3A and 6A). These spanning sequences separating the consensus motifs further corroborate our suggestion regarding the multimeric nature of bound NhaR. It is remarkable that within the three consecutive consensus motifs, I, II and III, in the binding domain of NhaR, we identified by the DMS methylation protection assay, but not by the DNase I assay, four single bases which form direct contacts with NhaR: G−92 in I, G−60 in II and G−29 and A−24 in III. In the absence of Na+ both in vivo and in vitro, these bases were protected by NhaR or His-tagged NhaR respectively and exposed to methylation in the absence of the regulator (Figure 4A and C). The fact that the DNase I protection assay did not reveal these His-tagged NhaR contacts most probably stems from the difference in the sensitivity and mechanism of these assays. DNase I digests the DNA in unprotected sites which reside mainly in the minor groove of the DNA (Saase-Dwight and Gralla, 1991). DMS methylates mainly the N-7 position of guanine residues in the major groove of the DNA. Hence, we suggest that each contact site is located in different consecutive major grooves separated from each other by two turns of the helix (20 bp, Figure 6A). It is conceivable that additional binding bases exist which cannot be identified by the DMS methylation protection assay. Na+ had no effect on the binding of NhaR to nhaA as measured by the gel retardation assay. This result suggests that whether Na+ is present or not, NhaR is constantly bound to the nhaA DNA. This behavior is characteristic of many members of the LysR family; these regulators remain bound to their target DNA, with no change in affinity even in the absence of the specific inducer. It is only the footprint which is changed upon addition of the inducer (Storz et al., 1990; Schell, 1993; Toledano et al., 1994). Indeed, while Na+ had no effect on the footprint assayed by DNase I protection, the footprint discovered by the DMS methylation protection assay showed an effect of Na+. The binding of the His-tagged NhaR to two guanines was changed dramatically upon addition of Na+; G−60 was exposed specifically to DMS methylation by Na+ (100 mM) since in the absence of the ion or in the presence of K+ (100 mM) it was protected by His-tagged NhaR. The specific Na+ effect on G−60 was found both in vivo and in vitro with both linear and supercoiled plasmidic DNA. On the other hand, G−92 was exposed to methylation by the ion only in vivo. We therefore suggest that Na+ directly affects the interaction of NhaR with G−60 of nhaA but indirectly affects the interaction with G−92. The latter most probably requires either a particular topology of the DNA or another factor existing only in vivo. In this respect, we recently have established a connection between the Na+-specific, NhaR-dependent regulation of nhaA and H-NS, a DNA-binding protein and a global regulator (Dover et al., 1996). Although the mechani
Referência(s)