The multidrug efflux regulator TtgV recognizes a wide range of structurally different effectors in solution and complexed with target DNA: evidence from isothermal titration calorimetry.

TtgV modulates the expression of the ttgGHI operon, which encodes an efflux pump that extrudes a wide variety of chemicals including mono- and binuclear aromatic hydrocarbons, aliphatic alcohols, and antibiotics of dissimilar chemical structure. Using a 'lacZ fusion to the ttgG promoter, we show that the most efficient in vivo inducers were 1-naphthol, 2,3-dihydroxynaphthalene, 4-nitrotoluene, benzonitrile, and indole. The thermodynamic parameters for the binding of different effector molecules to purified TtgV were determined by isothermal titration calorimetry. For the majority of effectors, the interaction was enthalpy-driven and counterbalance by unfavorable entropy changes. The TtgV-effector dissociation constants were found to vary between 2 and 890 mum. There was a relationship between TtgV affinity for the different effectors and their potential to induce gene expression in vivo, indicating that the effector binding constant is a major determinant for efficient efflux pump gene expression. Equilibrium dialysis and isothermal titration calorimetry studies indicated that a TtgV dimer binds one effector molecule. No evidence for the simultaneous binding of multiple effectors to TtgV was obtained. The binding of TtgV to a 63-bp DNA fragment containing its cognate operator was tight and entropy-driven (K(D) = 2.4 +/- 0.35 nm, DeltaH = 5.5 +/- 0.04 kcal/mol). The TtgV-DNA complex was shown to bind 1-napthol with an affinity comparable with the free soluble TtgV protein, K(D) = 4.8 +/- 0.19 and 3.0 +/- 0.15 mum, respectively. The biological relevance of this finding is discussed.

Pseudomonas putida DOT-T1E is a paradigm of solventtolerant microorganisms because it can grow in the presence of high concentrations of extremely toxic and harmful compounds such as aromatic hydrocarbons (1,2). These compounds preferentially partition in the cell membrane, disorganizing it and leading to cell death (3). Efflux pumps have been shown to play a critical role in the removal of toxic compounds such as anti-biotics, biocides, dyes, detergents, fatty acids, and organic solvents from the cell membranes (2, 4 -13). In P. putida DOT-T1E, the cooperative action of up to three efflux pumps, TtgABC, TtgDEF, and TtgGHI, is needed to achieve maximal tolerance against toluene, one of the most toxic aromatic hydrocarbons. TtgGHI appears to be the most important extrusion element since, in contrast to the other two efflux pumps, a knock-out mutant in which this efflux pump is not functional was not able to withstand a sudden 0.3% (v/v) toluene shock regardless of the growth conditions (14). TtgGHI, like other multidrug-resistant pumps, possesses a broad substrate specificity reflected in its capacity to extrude not only aromatic hydrocarbons such as toluene, xylenes, or styrene but also aliphatic alcohols such as octanol, nonanol, and decanol, as well as antibiotics of different chemical structure, e.g. ampicillin, tetracycline, and nalidixic acid (14,15).
The expression of the ttgGHI operon is regulated by the TtgV protein (16). The ttgV gene is transcribed divergently from the ttgGHI operon, and the corresponding promoters, called P ttgV and P ttgG , overlap each other. The two start codons are separated by only 210 bp, 40 bp of which constitute the TtgV operator so that TtgV covers the Ϫ10 region of ttgG promoter and the Ϫ35 region of ttgV promoter (16,17). Basal expression from the ttgG and ttgV promoters occurs, but expression has been shown to increase in response to the presence of some, but not all, of the pump substrates in the culture medium. Direct evidence of in vitro TtgV binding to drugs has only been obtained with 1-hexanol; Guazzaroni et al. (17) showed in EMSA 1 that this aliphatic alcohol released TtgV from its target operator. The present study was undertaken to determine the effector profile of TtgV, elucidate the TtgV-effector stoichiometry, and determine the thermodynamic parameters for the binding of the most potent effectors. Furthermore, the binding of 1-naphthol, one of the most potent effectors, by free TtgV and the protein complexed to its operator DNA has been compared using ITC.

EXPERIMENTAL PROCEDURES
Bacterial Strains, Plasmids, and Culture Medium-The bacterial strains and plasmids used in this study are shown in Table I. Bacterial strains were grown in LB medium at 30°C as described before (14) or in 2ϫYT for the production of the TtgV protein (18). Liquid cultures were shaken on an orbital platform operating at 200 rpm. When re-quired, the following antibiotics were added to the cultures: Km, 50 g/ml; rifampicin, 20 g/ml; and tetracycline, 20 g/ml.
␤-Galactosidase Assays-Plasmid pANA96 carries a transcriptional fusion of the P ttgG promoter region to the ЈlacZ gene in the low copy pMP220 promoter probe vector. P. putida DOT-T1E (pANA96) was grown overnight on LB medium with tetracycline. Cultures were diluted to an initial OD 660 of 0.05 in the same medium supplemented or not with the chemicals under study at 1 mM. These compounds were dissolved in Me 2 SO when needed (note that the latter did not interfere with the induction assays performed in this study). When cultures reached an OD 660 of 0.9 -1.0, ␤-galactosidase activity was determined in triplicate in permeabilized cells (19).
TtgV Expression and Purification-Plasmid pTGF2 was constructed by cloning a 784-bp NdeI-BamHI fragment bearing the ttgV open reading frame in the Km r pET29a(ϩ) plasmid (Novagen) digested with the same enzymes to allow the expression of the native TtgV protein.
Plasmid pTGF2 was transformed in Escherichia coli B834 (DE3) cells. The cells were grown in two-liter conical flasks containing 500 ml of 2ϫYT culture medium with 50 g/ml Km, incubated at 30°C with shaking, and induced with 1 mM isopropyl ␤-D-thiogalactopyranoside when the turbidity of the culture was around 0.7. Then cultures were grown at 22°C for 3 h, and cells were harvested by centrifugation (10 min at 4000 ϫ g). The cell pellet was resuspended in 0.2 M sodium acetate, 50 mM NaCl, 0.1 mM EDTA, 2 mM dithiothreitol, and 1 mM phenylmethylsulfonyl fluoride and lysed by sonication. After centrifugation at 13,000 ϫ g for 40 min, the supernatant was loaded onto an S-cation column (16/10, Amersham Biosciences) and eluted with a sodium chloride gradient. The fraction containing TtgV was then dialyzed against buffer containing 10 mM Tris-HCl, pH 8.0, 5% (v/v) glycerol, 100 mM NaCl, 0.1 mM EDTA, and 2 mM dithiothreitol, concentrated to 2 ml and loaded onto a Superdex 200 16/20 column (Amersham Biosciences) for gel filtration. The proteins were then concentrated to 4 mg/ml as determined by the Bio-Rad protein assay kit.
Electrophoresis Mobility Shift Assay-A 228-bp DNA fragment containing the wild-type P ttgG promoter was amplified by PCR from pGG1 using appropriate primers, isolated from agarose gels, and end-labeled with 32 P as described before (18,20). About 1 nM labeled DNA (ϳ1.5 ϫ 10 4 cpm) was incubated with the indicated amounts of purified TtgV for 10 min at 30°C in 10 l of TAPS binding buffer (50 mM Tris-acetate, pH 8.0; 100 mM potassium acetate; 8 mM magnesium acetate; 27 mM ammonium acetate; 3.5% (w/v) polyethylene glycol, and 1 mM dithiothreitol) containing 20 g/ml poly(dI-dC) and 200 g/ml bovine serum albumin. Electrophoresis in nondenaturing polyacrylamide gels and analyses were as described before (17).
Single-round in Vitro Transcription Assays with Supercoiled Plasmid DNA-Reactions (20 l) were performed in STA buffer (25 mM Tris-acetate, pH 8.0, 8 mM magnesium acetate, 3.5% (w/v) polyethylene glycol, 10 mM KCl) containing 100 nM 70 -holoenzyme (Epicenter), 20 units of RNAsin (Promega), 0.1 mM GTP, and 10 nM supercoiled pTE103-P ttgG DNA template (10). The reactions were incubated for 20 min at 30°C before the addition of the following elongation mixture: 0.1 mM each of ATP, CTP, and UTP; 0.3 Ci of [␣-32 P]UTP (20 Ci/l); and 100 g/ml heparin. After incubation for a further 10 min at 30°C, the reactions were stopped by chilling to 4°C, and the product was precipitated with 0.25 volumes of 10 M ammonium acetate and 2.5 volumes of ethanol. The pellets were washed with 80% (v/v) ethanol. Dried pellets were resuspended in 8 l of water and 4 l of formamide sequencing dye. Samples were submitted to electrophoresis using a 6.5% (w/v) polyacrylamide denaturing sequencing gel. The results were analyzed using Personal FX equipment software (Bio-Rad).
Isothermal Titration Calorimetry-Measurements were performed on a VP-Microcalorimeter (MicroCal, Northampton, MA) at 30°C. The protein was thoroughly dialyzed against 25 mM Tris acetate, pH 8.0, 8 mM magnesium acetate, 10 mM KCl, and 1 mM dithiothreitol. The protein concentration was determined using the Bradford assay. Stock solutions of 1-naphthol, 2,3-dihydroxynaphthalene, indole, and 4-nitrotoluene at a concentration of 500 mM were prepared in Me 2 SO and subsequently diluted with dialysis buffer to a final concentration of 0.3 mM (1-naphthol and 2,3-dihydroxynaphthalene), 1.5 mM (indole), and 1 mM (4-nitrotoluene). The appropriate amount of Me 2 SO (0.1%) was added to the protein sample in each assay. Solutions of benzonitrile (4 mM) and hexanol and toluene (5 mM) were directly prepared in dialysis buffer. All chemicals were manipulated in glass vessels, and effector samples were neither degassed nor filtered, to avoid evaporation or nonspecific binding. Each titration involved a single 2-l injection and a series of 4-l injections of effector molecules into the protein solution. For DNA binding studies, oligonucleotides corresponding to both strands of the TtgV operator were synthesized (5Ј-GGAATTCTCAAGAGT-ATCACATAATGCTACACTCTACCGCATTACGATTCAGCAACTGCA-GAA-3Ј and its corresponding complementary oligonucleotide). Annealing was carried out by mixing equimolar amounts (at a concentration of 60 M) of each complementary oligonucleotide in 0.5 mM Tris-HCl, pH 8.0, 0.5 mM MgCl 2 . The mixture was incubated 95°C for 5 min and then chilled on ice and dialyzed in the buffer used for ITC studies. The mean enthalpies measured from injection of the ligand in the buffer were subtracted from raw titration data before data analysis with ORIGIN software (MicroCal). Titration curves were fitted by a nonlinear least squares method to a function for the binding of a ligand to a macromolecule (21). From the curve thus fitted, the parameters ⌬H (reaction enthalpy), K A (binding constant, K A ϭ 1/K D ), and n (reaction stoichiometry) were determined. From the values of K A and ⌬H, the change in free energy (⌬G) and in entropy (⌬S) were calculated with the equation: ⌬G ϭ ϪRT lnK A ϭ ⌬ Ϫ T⌬S, where R is the universal molar gas constant and T is the absolute temperature.
Measurements of TtgV/Ligand Binding Ratio Using Equilibrium Dialysis Assays-Four samples of TtgV with different protein concentrations were dialyzed against protein buffer containing effectors using Slidea-Lyzer (Pierce) equipment for 5 days at 4°C with stirring to ensure equilibrium. The proteins inside the cassette were then denatured by incubating at 100°C for 5 min to release the bound effectors into the buffer. The denatured proteins were then centrifuged for 2 min at 13,000 ϫ g. Ultraviolet light absorption at the appropriate wavelength was then measured in the supernatants that contained the effectors, and the effector concentrations were determined using the corresponding ʦ extinction coefficients. The concentration of protein-bound effectors was obtained after correction for the effector concentration in the buffer. Protein concentrations were determined with the Bradford assay (Pierce

RESULTS
In Vivo Effector Profile of TtgV-As an initial approach to the identification of the effectors recognized by TtgV, we used a P ttgG ::ЈlacZ fusion (pANA96) to measure ␤-galactosidase activity in P. putida DOT-T1E cells grown in the absence or in the presence of 1 mM compounds (Fig. 1). The basal level of expression from the ttgG promoter was 438 Ϯ 34 Miller units, and expression increased up to 5-fold in response to 1 mM tested Tc r , ttgG promoter cloned in pMP220 16 pET29a(ϩ) K m r , protein expression vector Novagen pGG1 Ap r , pUC18 bearing an 8-kb BamHI fragment with ttgGHI and ttgVW 14 pMP220 Tc r , promoterless lacZ expression vector 37 pTE103-P ttgG Ap r , promoter of ttgG cloned upstream of the T7 terminator in pTE103 17 pTGF2 Km r , pET29a(ϩ) derivative vector used to produce TtgV This work chemicals. The effectors that yielded the highest induction levels (4 -5-fold increase) were two-ring aromatic compounds such as 1-naphthol, 2,3-dihydroxynaphthalene, and indole and one-ring aromatic compounds such as benzonitrile and 4-nitrotoluene ( Fig. 1). Other compounds such as alkylphenols, halogenated aromatic rings, and aliphatic and aromatic alcohols also behaved as effectors and increased expression from the P ttgG promoter by at least 2-fold.
In Vitro TtgV-Effector Interactions-Equilibrium dialysis experiments were carried out to shed light into the effector-TtgV stoichiometry. The results for the binding of 1-naphthol and 2,3-dihydroxynaphthalene to TtgV, as shown in Tables II and III, indicated that one effector molecule binds to the TtgV dimer. The apparent affinity for these molecules was in the low micromolar range. The thermodynamic parameters for the interaction of these two effectors as well as those of other effectors (biaromatic, monoaromatic compounds, and aliphatic alcohols) were subsequently determined by ITC at 30°C. These effectors were chosen to cover the spectrum of chemically different compounds that behaved as inducers in vivo.
The titration of TtgV with different effectors (1-naphthol, 2,3-dihydroxynaphthalene, benzonitrile, indole, 4-nitrotoluene, toluene, and 1-hexanol) is characterized by exothermal heat changes, giving rise to hyperbolic binding curves. This was exemplified by the titration with 1-naphthol shown in Fig. 2 (left-hand panel). ITC data were analyzed using n fixed at 0.5 (one effector molecule/dimer) determined by equilibrium dialysis experiments (see above), and satisfactory curve fits were obtained. The corresponding thermodynamic parameters are shown in Table IV. For all tested effectors, with the exception of 2,3-dihydroxynaphthalene, the binding was driven by favorable enthalpy changes and counterbalanced by unfavorable entropy changes (Table IV). In contrast, the thermodynamic mode of the binding of 2,3-dihydroxynaphthalene was different from that of the other effectors. Binding gave rise to only very small exothermic heat changes and was thus driven by entropy changes. It should be noted that dissociation constants for the binding of naphthol and 2,3-dihydroxynaphthalene determined by ITC and equilibrium dialysis were very close (3.0 Ϯ 0.15 and 2.2 Ϯ 0.8 M for 1-naphthol and 2.3 Ϯ 0.42 and 1.2 Ϯ 0.6 M for 2,3-dihydroxynaphthalene, respectively). These two bi-aromatic compounds are clearly bound by TtgV with the highest affinity, which is also reflected in its superior in vivo efficiency (Fig. 1).
The dissociation constants for the different effectors span the micromolar range (Table IV). Although benzonitrile and indole were shown to be efficient effectors in vivo, the affinity of TtgV for these two chemicals was lower (in the range of 50 -70 M) than for 1-naphthol. Substantially lower affinities were determined for toluene (K D ϭ 118 M) and 1-hexanol (K D ϭ 892 M), which may account for the relatively modest activity of these effectors in vivo. The nitro substitution of toluene at position 4 resulted in a substantial increase in ⌬H (Ϫ6.3 to Ϫ14.9 kcal/ mol) and an 8-fold increase in affinity (Table IV), which is in agreement with 4-nitrotoluene being a more efficient effector than toluene in vivo. This is compatible with a potential direct recognition of the nitro group at position 4 by TtgV since a nitro group at position 2 or 3 resulted in a less efficient effector.
Different effectors have been shown to bind to different sites in the large binding pocket of the QacR protein (22)(23)(24)(25). This raises the question whether the effector binding pocket of TtgV can accommodate different molecules at a time. We carried out a series of sequential ITC experiments that involved the initial saturation of TtgV with a first effector followed by the titration with a second effector. In a first series of experiments, TtgV (6.4 M) was saturated by the addition of aliquots of 300 M 1-naphthol. This complex was titrated with 300 M 2,3-dihydroxynaphthalene. In a second series of experiments, TtgV (6.7 M) was saturated with 2,3-dihydroxynaphthalene and subsequently titrated with 1 mM benzonitrile. In both cases, heat changes were very small and corresponded to the competition of two effectors to a single site and not to the simultaneous binding of both effectors to the protein with a physiological relevant affinity (data not shown). We were thus unable to provide evidence for the simultaneous binding of multiple effectors to TtgV.
ITC Binding Studies of 1-nNaphthol to the TtgV-DNA Complex-As stated above, TtgV exerts its biological function by an up-regulation of gene expression as a result of the effector-   Table IV. . Experiments were designed so that the protein concentration after saturation with DNA corresponded exactly to the protein concentration used for the titration of unliganded protein with 1-naphthol. After saturation, the TtgV-DNA complex was titrated with 1-naphthol in a similar fashion as the titration of the unliganded protein (Fig. 2, right-hand panel). The resulting heat changes corresponded to the binding of the effector to the protein and to the dissociation of the protein from DNA. Peaks were narrow, indicating that operator/TtgV dissociation occurred immediately upon effector binding by TtgV. After titra-tion, the sample was subjected to EMSA, which demonstrated that protein has dissociated quantitatively from its target operator DNA (data not shown). In a control assay, free DNA at the same concentration as in the titration of the DNA-protein complex was titrated with 1-naphthol. Resulting peaks were small and uniform, indicative of that dilution. ITC data for the titration of the DNA-TtgV complex with 1-naphthol were analyzed assuming that one effector molecule was bound per dimer, and derived thermodynamic parameters are given in  release TtgV from its target site. To test this hypothesis, we used EMSA assays with TtgV at 50 nM, a concentration that shifted about 50% of the DNA (Fig. 3), and the most efficient inducers of P ttgG in vivo were tested at different concentrations. EMSA studies revealed that there was a direct correlation between the concentration of the tested compound and the amount of TtgV that was released from its target DNA (see Fig.  3 for 1-naphthol, 2,3-dihydroxynaphthalene, benzonitrile, naphthalene, indole, 3-ethylphenol, and 4-nitrotoluene). In fact, with 0.5 mM 1-naphthol, 2,3-dihydroxynaphthalene, 3-ethylphenol, and 4-nitrotoluene, 90% of TtgV was released from its operator DNA (Fig. 3B). Indole and benzonitrile seemed to be similarly recognized by TtgV (around 70 -80% of the DNA was freed from TtgV at a concentration of 1 mM compound (Fig.  3B)). These results are, in general, in agreement with the in vivo measurements of ␤-galactosidase activity since the compounds that yielded the highest induction level of the P ttgG promoter in vivo also released more TtgV from its operator site in the in vitro experiments. Furthermore, compounds such as p-isopropylbenzoate, tetracycline, and chloramphenicol, which did not induce in vivo (not shown), failed to release TtgV from its target operator in vitro.

FIG. 2. Isothermal titration calorimetry data for the binding of 1-naphthol to TtgV (left-hand panel) and to TtgV saturated with a 63-bp DNA fragment containing the operator (right-hand panel). Left panel, heat changes (upper panel) and integrated peak areas (lower
From a mechanistic point of view, TtgV was suggested to prevent RNA polymerase from accessing the promoter region (17). This suggestion was based on the observation that transcription inhibition was more effective when TtgV was added before the formation of the open complex by RNA polymerase. It was further reasoned that upon ligand binding, TtgV was released from its operator site and ttgGHI transcription occurred. To test whether the chemicals capable of releasing TtgV from its operator allowed expression from the P ttgG promoter, we performed in vitro transcription assays in which TtgV was added before RNA polymerase in the absence and in the presence of some of the most efficient effectors (Fig. 4). As expected, expression from the ttgG promoter was inhibited (lane 2), whereas the presence of 100 M benzonitrile, 2,3-dihydroxynaphthalene, or 1-naphthol enabled transcription from P ttgG at a level similar to that observed in the absence of TtgV. These results support a mechanism by which TtgV represses ttgG expression by blocking the RNA polymerase binding site. Nevertheless, in the presence of ligand molecules of different structures and substituents, TtgV dissociates from its operator site, allowing RNA polymerase to start transcription from P ttgG . DISCUSSION EMSA, equilibrium dialysis, and ITC experiments (Figs. 2  and 3 and Tables II-IV) suggested that TtgV is able to bind a large number of structurally different compounds and that this interaction leads to the induction of ttgGHI expression. This is the first case in which a regulator belonging to the IclR family is shown to interact directly with different chemicals. Our findings contrast with the relatively narrow effector specificity found for other regulators of the IclR family that recognize a single aromatic compound. For example, PobR and PcaU of Acinetobacter calcoaceticus interact only with 4-hydroxybenzoate and 3,4-dihydroxybenzoate, respectively (26 -28). However, the ability of TtgV to recognize various effectors of dissimilar structure seems to be a particular feature of the regulators that control the expression of multidrug efflux pumps. This is consistent with a mechanism based on the direct recognition of structurally dissimilar compounds rather than the involvement of a secondary messenger, which has also been shown for a limited number of other multidrug pump regulators such as QacR (29) and BmrR (30,31).
Our data indicate that one effector molecule binds to a TtgV dimer. This stoichiometry has also been seen for the QacR regulator (25) but is different from TetR, where two molecules of tetracycline bind to the protein dimer (32,33). QacR accommodates structurally diverse ligands in different parts of the large binding pocket (25). However, signal transduction for all ligands is mediated by an identical induction mechanism, and QacR-drug complexes undergo the same transition (25,33). Our results support that TtgV, a member of the IclR family of repressors, may function similarly to QacR since there is a relation between the in vitro affinity of TtgV for its effectors and their in vivo efficiency. However, details of TtgV-effector interactions await the resolution of the three-dimensional structure of TtgV bound to these chemicals.
The affinity of TtgV for its effectors spans the micromolar range, and the binding constants are similar to those of other repressor proteins such as TrpR (34) or QacR (35), the latter having a K D around 1 M for rhodamine 6G (25). Although there is no strict correlation (R 2 ϭ 0.76) for the linear fit of the plot of lnK D against ␤-galactosidase activity, a clear relation was observed between the affinity of TtgV for different effectors and their potential to induce gene expression (measured by ␤-galactosidase assays). Effectors with the highest affinity, such as 1-naphthol and 2,3-dihydroxynaphthalene, were shown to be more efficient in vivo. Effectors with slightly lower in vivo induction activities, such as indole, 4-nitrotoluene, and benzonitrile, were also found to bind less strongly to the protein in vitro, whereas effectors characterized by moderate in vivo activity (toluene and hexanol) showed significantly reduced affinity. The lack of a strict correlation between in vitro-determined K D values and ␤-galactosidase levels induced in vivo is mainly due to differential cell extrusion of the tested compounds, which is mediated by the TtgGHI pump and other multidrug extrusion elements rather than to the existence of secondary layers of regulators influenced by these chemicals. 2 Furthermore, it should be taken into account that the positive correlation between the affinity of TtgV for an effector molecule and the release of the protein from the promoter is not always found in transcriptional regulators. This is exemplified by FadR, for which the effector palmitoyl-CoA is around 50-fold more efficient than myristoyl-CoA in inhibiting FadR from FIG. 4. TtgV repression of the transcription of P ttgG is alleviated by different ligands. Single-round transcription assays were carried out as described under "Experimental Procedures." The assays were performed at 30°C for 20 min in the absence of TtgV (lane 1) or in the presence of 0.5 M TtgV added prior to 100 nM RNA polymerase (lanes 2-5). Different effectors were added at 100 M: benzonitrile, 2,3dihydroxynaphthalene, and 1-naphthol. The mRNAs synthesized from P ttgG and P sp (a plasmid promoter used as a control) are indicated by arrowheads.
DNA binding. However, ITC assays showed that palmitoyl-CoA binds around one-sixth as strongly to FadR as does myristoyl-CoA (36). The binding affinity of the effector to TtgV can thus be considered the major determinant of gene expression.
In general, the binding modes of structurally similar ligands to proteins with narrow substrate specificity are comparable. This was not the case for binding of TtgV to 1-naphthol and 2,3-dihydroxynaphthalene (Table IV), which were enthalpyand entropy-driven, respectively. It remains to be explored whether the different binding modes observed here for structurally similar effectors is a general feature of multidrug recognition. TtgV is a dimer in solution and binds to its corresponding target, DNA covering four potential direct repeats within the ttgG promoter (17). In the absence of effectors, TtgV binds tightly to its target promoter (K D 2.4 Ϯ 0.35 nM).
The effector 1-naphthol was shown to bind with high affinity to the protein-DNA complex (K D ϭ 4.8 M), an affinity similar to that determined for the binding to free TtgV (K D ϭ 3.0 M). Furthermore, TtgV bound to 1-naphthol (or other effectors) does not interact with its target DNA promoter. This set of observations is of physiological importance and can be critical for survival since the K D of 1-naphthol for the TtgV-DNA complex is lower than the toxicity threshold of this compound (around 1 mM). This implies that 1-naphthol triggers a TtgV response at a concentration at which its toxicity is low. The up-regulation of the expression of the efflux pump and consequently the extrusion of toxic substances probably lead to a dynamic equilibrium between uptake and expulsion, which is characterized by an intracellular concentration of the toxic compound below its toxicity threshold. This model could be a general feature for the regulation of bacterial multidrug transporters.
The finding that the binding of the TtgV-DNA complex to 1-naphthol was less exothermal than the binding to free protein was unexpected (Fig. 2, right-hand panel; Table IV). As stated above, the heat signal from the titration of the TtgV-DNA complex with 1-naphthol should be considered as the sum of heats originating from effector binding and protein dissociation from DNA. We have demonstrated that the binding of the effector to TtgV is characterized by a favorable enthalpy change of Ϫ8.6 Ϯ 0.16 kcal/mol, whereas the binding of protein to DNA is endothermic (⌬H ϭ 5.5 Ϯ 0.04 kcal/mol). This implies that protein dissociation from DNA should give rise to an exothermic signal. During the binding of 1-naphthol to the TtgV-DNA complex, the exothermic heat of TtgV-operator dissociation was expected to combine with the exothermic heat generated by the effector binding, giving rise to a stronger exothermic signal than in the binding of TtgV to the effector. However, this was not the case, and this heat change for the binding of 1-naphthol to the TtgV-DNA complex was less exothermic than that the one observed in the binding to TtgV (Fig.  2, Table IV). The above hypothetical combination effect of heat is only valid for fully reversible systems, in which the nature of the binding reactions corresponds to the inverse of the dissociation reactions. In general, very little is known about the mechanisms of effector-induced protein dissociation. Our data indicate either that the system studied here is not entirely reversible or that the molecular mechanisms of protein-DNA association and dissociation differ. Further studies are necessary to elucidate the conformational changes caused by effector binding to free and TtgV-bound DNA.