Different Modes of Binding of Mono- and Biaromatic Effectors to the Transcriptional Regulator TTGV
2007; Elsevier BV; Volume: 282; Issue: 22 Linguagem: Inglês
10.1074/jbc.m610032200
ISSN1083-351X
AutoresMaría‐Eugenia Guazzaroni, Marı́a-Trinidad Gallegos, Juan L. Ramos, Tino Krell,
Tópico(s)Receptor Mechanisms and Signaling
ResumoMembers of the IclR family of regulators exhibit a highly conserved effector recognition domain and interact with a limited number of effectors. In contrast with most IclR family members, TtgV, the transcriptional repressor of the TtgGHI efflux pump, exhibits multidrug recognition properties. A three-dimensional model of the effector domain of TtgV was generated based on the available three-dimensional structure of several IclR members, and a series of point mutants was created. Using isothermal titration calorimetry, we determined the binding parameters of the most efficient effectors for TtgV and its mutant variants. All mutants bound biaromatic compounds with higher affinity than the wild-type protein, whereas monoaromatic compounds were bound with lower affinity. This tendency was particularly pronounced for mutants F134A and H200A. TtgVF134A bound 4-nitrotoluene with an affinity 13-fold lower than that of TtgV (17.4 ± 0.6 μm). This mutant bound 1-naphthol with an affinity of 5.7 μm, which is seven times as great as that of TtgV (40 μm). The TtgVV223A mutant bound to DNA with the same affinity as the wild-type TtgV protein, but it remained bound to the target operator in the presence of effectors, suggesting that Val-223 could be part of an intra-TtgV signal recognition pathway. Thermodynamic analyses of the binding of effectors to TtgV and to its mutants in complex with their target DNA revealed that the binding of biaromatic compounds resulted in a more efficient release of the repressor protein than the binding of monoaromatics. The physiological significance of these findings is discussed. Members of the IclR family of regulators exhibit a highly conserved effector recognition domain and interact with a limited number of effectors. In contrast with most IclR family members, TtgV, the transcriptional repressor of the TtgGHI efflux pump, exhibits multidrug recognition properties. A three-dimensional model of the effector domain of TtgV was generated based on the available three-dimensional structure of several IclR members, and a series of point mutants was created. Using isothermal titration calorimetry, we determined the binding parameters of the most efficient effectors for TtgV and its mutant variants. All mutants bound biaromatic compounds with higher affinity than the wild-type protein, whereas monoaromatic compounds were bound with lower affinity. This tendency was particularly pronounced for mutants F134A and H200A. TtgVF134A bound 4-nitrotoluene with an affinity 13-fold lower than that of TtgV (17.4 ± 0.6 μm). This mutant bound 1-naphthol with an affinity of 5.7 μm, which is seven times as great as that of TtgV (40 μm). The TtgVV223A mutant bound to DNA with the same affinity as the wild-type TtgV protein, but it remained bound to the target operator in the presence of effectors, suggesting that Val-223 could be part of an intra-TtgV signal recognition pathway. Thermodynamic analyses of the binding of effectors to TtgV and to its mutants in complex with their target DNA revealed that the binding of biaromatic compounds resulted in a more efficient release of the repressor protein than the binding of monoaromatics. The physiological significance of these findings is discussed. The DOT-T1E strain of Pseudomonas putida has the extraordinary capacity to withstand, and even grow in, the presence of high concentrations of organic solvents such as an aqueous solution saturated with toluene, a highly toxic compound (1.Ramos J.L. Duque E. Huertas M.J. Haïdour A. J. Bacteriol. 1995; 177: 3911-3916Crossref PubMed Google Scholar). The main mechanism underlying this resistance lies in the action of three RND (resistance-nodulation-cell Division) efflux pumps, termed TtgABC, TtgDEF, and TtgGHI (pump encoded by toluene tolerance genes ttgGHI) (2.Rojas A. Duque E. Mosqueda G. Golden G. Hurtado A. Ramos J.L. Segura A. J. Bacteriol. 2001; 183: 3967-3973Crossref PubMed Scopus (210) Google Scholar), which extrude organic solvents and other toxic compounds from the cells. These three efflux pumps show a high degree of similarity to the AcrAB-TolC multidrug efflux pump, which is the best characterized member of this family (3.Murakami S. Nakashima R. Yamashita E. Yamaguchi A. Nature. 2002; 419: 587-593Crossref PubMed Scopus (766) Google Scholar, 4.Zgurskaya H.I. Nikaido H. J. Bacteriol. 2000; 182: 4264-4267Crossref PubMed Scopus (141) Google Scholar, 5.Yu E.W. McDermont G. Zgurskaya H.I. Nikaido H. Koshland Jr., D.E. Science. 2003; 300: 976-980Crossref PubMed Scopus (338) Google Scholar, 6.Koronakis V. Sharff A. Koronakis E. Luisi B. Hugues C. Nature. 2000; 405: 914-919Crossref PubMed Scopus (870) Google Scholar, 7.Su C.-C. Li M. Gu R. Takatsuka Y. McDermont G. Nikaido H. Yu E.W. J. Bacteriol. 2006; 188: 7290-7296Crossref PubMed Scopus (101) Google Scholar, 29.Yu E.W. Aires J.R. McDermont G. Nikaido H. J. Bacteriol. 2005; 187: 6804-6815Crossref PubMed Scopus (167) Google Scholar). Expression of the P. putida efflux pumps is controlled by transcriptional repressors; TtgR controls the expression of the ttgABC operon (8.Duque E. Segura A. Mosqueda G. Ramos J.L. Mol. Microbiol. 2001; 39: 1100-1106Crossref PubMed Scopus (92) Google Scholar, 9.Tera´n W. Krell T. Ramos J.L. Gallegos M.T. J. Biol. Chem. 2006; 281: 7102-7109Abstract Full Text Full Text PDF PubMed Scopus (64) Google Scholar), whereas TtgV (10.Rojas A. Segura A. Guazzaroni M.E. Tera´n W. Hurtado A. Gallegos M.-T. Ramos J.L. J. Bacteriol. 2003; 185: 4755-4763Crossref PubMed Scopus (65) Google Scholar, 11.Guazzaroni M.-E. Tera´n W. Zhang X. Gallegos M.-T. Ramos J.L. J. Bacteriol. 2004; 186: 2921-2927Crossref PubMed Scopus (35) Google Scholar, 12.Guazzaroni M.-E. Krell T. Felipe A. Ruiz R. Meng C. Zhang X. Gallegos M.-T. Ramos J.L. J. Biol. Chem. 2005; 280: 20887-20893Abstract Full Text Full Text PDF PubMed Scopus (52) Google Scholar) is the main regulator controlling the expression of the ttgDEF and ttgGHI operons. TtgV, a member of the IclR family of regulators (13.Molina-Henares A.-J. Krell T. Guazzaroni M.-E. Segura A. Ramos J.L. FEMS Microbiol. Rev. 2006; 30: 157-186Crossref PubMed Scopus (153) Google Scholar, 14.Krell T. Molina-Henares A.J. Ramos J.L. Protein Sci. 2006; 15: 1207-1213Crossref PubMed Scopus (35) Google Scholar), exhibits multidrug binding properties (12.Guazzaroni M.-E. Krell T. Felipe A. Ruiz R. Meng C. Zhang X. Gallegos M.-T. Ramos J.L. J. Biol. Chem. 2005; 280: 20887-20893Abstract Full Text Full Text PDF PubMed Scopus (52) Google Scholar) in contrast to other members of this family, which are generally characterized by their high specificity for effector molecules (15.Kok R.D. D'Argenio D.A. Ornston L.N. J. Bacteriol. 1998; 180: 5058-5069Crossref PubMed Google Scholar, 16.Rintoul M.R. Cusa E. Baldoma L. Badia J. Reitzer L. Aguilar J. J. Mol. Biol. 2002; 324: 599-610Crossref PubMed Scopus (36) Google Scholar, 17.Arias-Barrau E. Olivera E.R. Luengo J.M. Ferna´ndez C. Gala´n B. Garci´a J.L. Di´az E. Miñambres B. J. Bacteriol. 2004; 186: 5062-5067Crossref PubMed Scopus (184) Google Scholar). TtgV is a repressor that operates according to effector-mediated derepression. In the absence of effector, the protein is bound to the promotor DNA region repressing transcription. Effector binding to the TtgV-DNA complex is thought to produce an intramolecular signal that is transmitted to the DNA-binding domain, giving rise to protein dissociation from the operator. The RNA polymerase then accesses the promoter and transcribes the corresponding genes (11.Guazzaroni M.-E. Tera´n W. Zhang X. Gallegos M.-T. Ramos J.L. J. Bacteriol. 2004; 186: 2921-2927Crossref PubMed Scopus (35) Google Scholar). The most efficient effectors in vivo are two-ring aromatic compounds such as 1-naphthol (1NL) 2The abbreviations used are: 1NL, 1-naphthol; BN, benzonitrile; EMSA, electrophoretic mobility shift assay; IND, indole; ITC, isothermal titration calorimetry; MIC, minimal inhibitory concentration; 4NT, 4-nitrotoluene; PDB, Protein Data Bank. and indole (IND) and one-ring compounds such as 4-nitrotoluene (4NT) and benzonitrile (BN) (12.Guazzaroni M.-E. Krell T. Felipe A. Ruiz R. Meng C. Zhang X. Gallegos M.-T. Ramos J.L. J. Biol. Chem. 2005; 280: 20887-20893Abstract Full Text Full Text PDF PubMed Scopus (52) Google Scholar). In the framework of structural genomic studies, two research groups reported the three-dimensional structure of two members of the IclR family members, that of the IclR-TM protein isolated from Thermotoga maritima (PDB: 1MKM (18.Zhang R.G. Kim Y. Skarina T. Beasley S. Laskowski R. Arrowsmith C. Edwards A. Joachimiak A. Savchenko A. J. Biol. Chem. 2002; 277: 19183-19190Abstract Full Text Full Text PDF PubMed Scopus (59) Google Scholar)) and that of a regulator of unknown function purified from Rhodococcus sp. RHA1 (PDB: 2G7U). Both proteins consist of two well separated domains. The N-terminal DNA-binding domain is linked by a long helix to the conserved effector-binding domain. Furthermore, the coordinates of several individual effector-binding domains of IclR family members have been released on the PDB data base. However, none of the structures available forms a complex with a physiologically relevant ligand. A certain body of information is available on the interaction of multidrug-binding transcriptional regulators with effectors (19.Va´zquez-Laslop N. Markham P.N. Neyfakh A.A. Biochemistry. 1999; 38: 16925-16931Crossref PubMed Scopus (45) Google Scholar, 20.Schumacher M.A. Miller M.C. Brennan R.G. EMBO J. 2004; 23: 2923-2930Crossref PubMed Scopus (106) Google Scholar). In several cases, affinities do not correlate with the potential of the effectors to release protein in vitro or with their capacity to modulate gene expression in vivo (21.Grkovic S. Brown M.H. Roberts N.J. Paulsen I.T. Skurray R.A. J. Biol. Chem. 1998; 273: 18665-18673Abstract Full Text Full Text PDF PubMed Scopus (156) Google Scholar, 22.DiRusso C.C. Tsvetnitsky V. Højrup P. Knudsen J. J. Biol. Chem. 1998; 273: 33652-33659Abstract Full Text Full Text PDF PubMed Scopus (50) Google Scholar). This raises questions concerning the efficiency of effector-induced intramolecular signal transmission, an issue that has rarely been addressed. From structural studies, it appears that the effector structure determines with which set of amino acids the effector interacts in the binding pocket (23.Schumacher M.A. Millar M.C. Grkovic S. Brown M. Skurray R.A. Brennan R.G. Science. 2001; 294: 2158-2163Crossref PubMed Scopus (328) Google Scholar), which in turn determines the efficiency of signal transduction (21.Grkovic S. Brown M.H. Roberts N.J. Paulsen I.T. Skurray R.A. J. Biol. Chem. 1998; 273: 18665-18673Abstract Full Text Full Text PDF PubMed Scopus (156) Google Scholar). It is unknown whether this efficiency can also be modulated by mutations in the regulator binding pocket, which would be of relevance for understanding the evolution of multidrug-binding proteins. Based on the three-dimensional homology model of TtgV, we have identified the potential effector-binding pocket of TtgV. Six site-directed alanine replacement mutants of amino acids located in this pocket were generated and characterized in this study. Isothermal titration calorimetry (ITC) was used to determine the thermodynamic parameters for the binding of four different effectors to the wild type and to all TtgV mutants. Further experiments were aimed at evaluating the impact of the mutations on DNA binding and at characterizing the efficiency of the effector in triggering the release of the bound repressor. Site-directed Mutagenesis−TtgV mutants, in which amino acid residues at positions 118, 134, 140, 200, 204, and 223 were replaced by alanine and valine (positions 134 and 200), were generated by overlapping PCR mutagenesis (24.Ho S.N. Hunt H.D. Horton M. Pullen J.K. Pease L.R. Gene (Amst.). 1989; 77: 51-59Crossref PubMed Scopus (6851) Google Scholar) using plasmid pANA126 (10.Rojas A. Segura A. Guazzaroni M.E. Tera´n W. Hurtado A. Gallegos M.-T. Ramos J.L. J. Bacteriol. 2003; 185: 4755-4763Crossref PubMed Scopus (65) Google Scholar) as a source of the ttgV wild-type allele. For each mutant, three PCRs were carried out. The initial two PCRs involved amplifications using the upstream primer (5′-CGCTCCACCGTTCAGAGAAT-3′, corresponding to nucleotides 139–158 of ttgV coding sequence) and a mismatch primer covering the segment to be mutated as well as a PCR amplification using the downstream primer (5′-CTTGTCGACGGAGCTCGAAT-3′, nucleotides 791–810 of the ttgV coding sequence) and an oligonucleotide complementary to the mismatch primer. The following mismatch primers were used, and the mismatch codon is underlined: I118A, 5-AGACAAAGCGTACGTGCTT-3; F134A, 5′-GGTAGTGGCGCCGATTGGTA-3′; V140F, 5′-GTATTAACTTCCCCGCGCA-3′; V140A, 5′-GTATTAACGCGCCCGCCGCA-3′; H200A, 5′-TGGACGAGGCGATTGATGGC-3′; V204A, 5′-ATTGATGGCGCGTGCTCATT-3′; V223A: 5′-CTCGCGATCGCGATGCCGAG-3′. The resulting overlapping PCR products were annealed, supplemented with upstream and downstream primers, and submitted to the third PCR. For the I118A, F134A, and V160A mutations, the final PCR product was cut with BbvCI and BlnI, which produced a 223-bp fragment that was cloned into pANA126 linearized with the same enzymes. For the H200A, H200V, V204A, and V223A mutations, the final PCR product was digested with BlnI and PstI, and the resulting 200-bp fragment was equally cloned into pANA126. Cell Culture and Protein Expression−Escherichia coli B834 (DE3) was transformed with pANA126 bearing the wild-type ttgV allele (10.Rojas A. Segura A. Guazzaroni M.E. Tera´n W. Hurtado A. Gallegos M.-T. Ramos J.L. J. Bacteriol. 2003; 185: 4755-4763Crossref PubMed Scopus (65) Google Scholar) and a series of pANA126 derivatives that encode the six different TtgV mutants. Cells were grown in 2-liter conical flasks with 500 ml of LB supplemented with 25 μg/ml kanamycin. Cultures were incubated at 30 °C with shaking and induced with 0.1 mm isopropyl β-d-thiogalactopyranoside when the culture reached a turbidity at 660 nm (OD660) of 0.7. Cultures were then transferred at 18 °C, and after growth for 3 h, cells were harvested by centrifugation (10 min at 4000 g) and stored at –80 °C. Protein Purification−Cells from a 1-liter culture were suspended in 50 ml of buffer A (25 mm NaH2PO4/Na2HPO4, 0.5 m NaCl, 10 mm imidazole, 5% (v/v) glycerol, 0.1 mm dithiothreitol, pH 7.5) containing 10 units/ml Benzonase (Novagen, Madrid, Spain) and one tablet of Roche Applied Science Complete™ EDTA-free protease inhibitor mixture. Cells were broken by two passages through a French press at 1000 p.s.i., and the resulting suspension was centrifuged at 19,000 × g for 45 min. The supernatant was filtered and loaded onto a 5-ml Hi-Trap chelating column (GE Healthcare, St. Gibes, UK). His6-TtgV and its mutant variants were eluted with a 45–500 mm gradient of imidazole in buffer A. Protein was dialyzed against 20 mm Tris-HCl, 8 mm magnesium acetate, 300 mm NaCl, pH 7.2. For storage at –80 °C, the samples were mixed with 10% (v/v) glycerol. The purity of the protein was between 90 and 95%, as judged from SDS-PAGE gels. Protein samples were aliquoted prior to freezing. All experiments were carried out with a single batch of each protein. Protein aliquots were thawed for immediate use, and excess protein was discarded. ITC−Measurements were made with a VP-Microcalorimeter (MicroCal, Northampton, MA) at 25 °C. The protein was thoroughly dialyzed against 20 mm Tris-HCl, 8 mm magnesium acetate, 100 mm NaCl, 10% (v/v) glycerol, and 1 mm dithiothreitol, pH 7.2. The buffer used in our initial analysis (12.Guazzaroni M.-E. Krell T. Felipe A. Ruiz R. Meng C. Zhang X. Gallegos M.-T. Ramos J.L. J. Biol. Chem. 2005; 280: 20887-20893Abstract Full Text Full Text PDF PubMed Scopus (52) Google Scholar) and in the present study differed in that the buffer used for the assays reported here included 10% (v/v) glycerol, which had a stabilizing effect on the protein, as well as a higher ionic strength and lower pH (7.2 rather than 8.0), which corresponds more closely to physiological conditions. Protein concentration was determined with the Bradford assay. Stock solutions of 1NL, BN, 4NT, and IND at a concentration of 500 mm were prepared in dimethyl sulfoxide, and the solutions were diluted with dialysis buffer to final concentrations of 0.5 to 1 mm. The corresponding amount of dimethyl sulfoxide was added to the protein sample. DNA duplex samples were prepared as described by Guazzaroni et al. (12.Guazzaroni M.-E. Krell T. Felipe A. Ruiz R. Meng C. Zhang X. Gallegos M.-T. Ramos J.L. J. Biol. Chem. 2005; 280: 20887-20893Abstract Full Text Full Text PDF PubMed Scopus (52) Google Scholar). Each titration involved a single 1.6-μl injection and a series of 4.8-μl injections of effectors into a 34–40 μm protein solution. The mean enthalpies measured from injection of the ligands into the buffer were subtracted from raw titration data prior to data analysis with a model for the binding of a ligand to identical independent sites of a macromolecule (MicroCal). Data analysis with this model produced satisfactory statistical data. Circular Dichroism−The CD spectra of each protein were recorded on a Jasco 715 spectropolarimeter (Great Dunmow, UK). Spectra in the far UV region (195–260 nm) were recorded in cylindrical quartz cells (0.02-cm path length) at a protein concentration of 0.6 mg/ml. All protein solutions were dialyzed against the buffer used for ITC. EMSA−Experiments were carried out as described previously (12.Guazzaroni M.-E. Krell T. Felipe A. Ruiz R. Meng C. Zhang X. Gallegos M.-T. Ramos J.L. J. Biol. Chem. 2005; 280: 20887-20893Abstract Full Text Full Text PDF PubMed Scopus (52) Google Scholar). One nm labeled DNA probe (∼1.5 × 104 cpm) was incubated with increasing concentration of TtgV (10–1500 nm) for 10 min at 30 °C in 10 μl of 20 mm Tris-HCl, 8 mm magnesium acetate, 100 mm NaCl, 10% (v/v) glycerol, and 1 mm dithiothreitol, pH 7.2, containing 20 μg/ml poly(dI-dC) and 200 μg/ml bovine serum albumin. Determination of MICs−P. putida DOT-T1E and isogenic mutants lacking one or several Ttg efflux pumps (2.Rojas A. Duque E. Mosqueda G. Golden G. Hurtado A. Ramos J.L. Segura A. J. Bacteriol. 2001; 183: 3967-3973Crossref PubMed Scopus (210) Google Scholar, 25.Mosqueda G. Ramos J.L. J. Bacteriol. 2000; 182: 937-943Crossref PubMed Scopus (100) Google Scholar) were grown overnight in LB medium with the appropriate antibiotics. Cultures were diluted 100-fold in LB supplemented with solvents at different concentrations and incubated for 20 h at 30 °C. The MIC value corresponds to the lowest concentration that reduced growth by more than 90%. Results are the average of at least five independent assays. Homology Modeling of TtgV and Identification of Amino Acids in the Effector-binding Site−We have previously shown that the IclR family of regulators comprises more than 500 members and that a distinct signature profile can be derived from the alignment of the whole set of proteins. In contrast with other families of regulators, the region that best defines the IclR family is not the DNA-binding domain but the effector-binding region (13.Molina-Henares A.-J. Krell T. Guazzaroni M.-E. Segura A. Ramos J.L. FEMS Microbiol. Rev. 2006; 30: 157-186Crossref PubMed Scopus (153) Google Scholar, 25.Mosqueda G. Ramos J.L. J. Bacteriol. 2000; 182: 937-943Crossref PubMed Scopus (100) Google Scholar). The predicted secondary structure of TtgV can be closely aligned to the secondary structure elements found in the three-dimensional structure of the full-length IclR-TM protein of T. maritima (18.Zhang R.G. Kim Y. Skarina T. Beasley S. Laskowski R. Arrowsmith C. Edwards A. Joachimiak A. Savchenko A. J. Biol. Chem. 2002; 277: 19183-19190Abstract Full Text Full Text PDF PubMed Scopus (59) Google Scholar). The TtgV sequence corresponding to the effector-binding domain was subsequently subjected to homology modeling with Geno3D (26.Combet C. Jambon M. Deléage G. Geourjon C. Bioinformatics. 2002; 18: 213-214Crossref PubMed Scopus (371) Google Scholar) software. The model was based on the four following templates: T. maritima IclR-TM (PDB: 1MKM (18.Zhang R.G. Kim Y. Skarina T. Beasley S. Laskowski R. Arrowsmith C. Edwards A. Joachimiak A. Savchenko A. J. Biol. Chem. 2002; 277: 19183-19190Abstract Full Text Full Text PDF PubMed Scopus (59) Google Scholar)), the effector-binding domains of E. coli IclR (PDB: 1TD5), the glyoxylate regulatory protein (PDB: 1TF1), and KdgR (PDB: 1YSP). These templates share 21–29% sequence identity with TtgV. The model obtained was submitted to What_check (27.Hooft R.W.W. Vriend G. Sander C. Abola E.E. Nature. 1996; 381: 272Crossref PubMed Scopus (1818) Google Scholar) and was found to have an acceptable geometry. A Ramachandran plot showed that over 97% of the residues were in allowed regions. With the DALI algorithm, it was possible to superimpose the TtgV effector-binding domain model onto its templates with Cα root mean square deviation values between 1.8 and 2.4 Å. A surface representation of this model (Fig. 1A) shows a hydrophobic cavity with a volume of 1200 Å 3W. Tera´n, A. Felipe, M.-E. Guazzaroni, T. Krell, R. Ruiz, J. L. Ramos, and M.-T. Gallegos, submitted for publication. as determined by PASS software (28.Brady Jr, G.P. Stouten P.F.W. J. Comput. Aided Mol. Des. 2000; 14: 383-401Crossref PubMed Scopus (462) Google Scholar). This model proposes that the ligand-binding pocket of TtgV is formed by a long loop connecting S2 with H3 and the β-sheet (Fig. 1B). A number of residues within this cavity were selected for mutagenesis according to two criteria: 1) their location and projection in the binding pocket of the model and 2) their conservation in TtgT, a transcriptional regulator sharing 56% sequence identity with TtgV, which has a very similar effector profile.3 Based on these criteria, 6 residues, Ile-118, Phe-134, Val-140, His-200, Val-204, and Val-233, were chosen (Fig. 1B), and alanine replacement mutants were generated in each position. In the model, 2 amino acids were located on β-strands: Ile-118 on S2 and Val-223 on S4. Residues Phe-134 and Val-140 were located on the loop connecting S2 with H3. The side chain of Phe-134 lies in the lower side of the effector-binding pocket and appears to play a central role in effector binding. Amino acids His-200 and Val-204 were mutated in the upper part of the pocket. Mutations in the predicted effector-binding pocket seemed not to alter the secondary protein structure of the mutant proteins, as deduced from analysis of mutant and wild-type TtgV proteins by far UV circular dichroism spectroscopy. The spectra of all proteins could be closely superimposed (Fig. 2), indicating that the amino acid replacements did not significantly affect the protein secondary structure. Most Mutants Show Increased Affinities for Biaromatic Effectors but Bind Monoaromatic Compounds with Lower Affinity−We recently showed that TtgV effectors are primarily monoaromatic and biaromatic compounds (12.Guazzaroni M.-E. Krell T. Felipe A. Ruiz R. Meng C. Zhang X. Gallegos M.-T. Ramos J.L. J. Biol. Chem. 2005; 280: 20887-20893Abstract Full Text Full Text PDF PubMed Scopus (52) Google Scholar). Two representatives of each class, which are efficient effectors in vivo, the monoaromatic compounds 4NT and BN and the biaromatic effectors 1NL and IND, were chosen to study their binding parameters to native and mutant proteins using ITC. With this technique, all thermodynamic binding parameters can be determined in a single experiment. Fig. 3A shows the titration of TtgV and the F134A mutant with 4NT. The derived binding parameters are given in Table 1. Binding to the TtgV protein was enthalpy-driven (ΔH = –9.7 ± 0.2 kcal/mol) and characterized by a KD of 17.4 ± 0.6 μm. When the same experiment was repeated with the F134A mutant, the heat changes were much smaller, and affinity decreased by a factor of almost 13-fold (Fig. 3A and Table 1). An even more dramatic decrease in affinity for 4NT (27-fold) was obtained for the H200A mutant. For both mutants, enthalpy changes were reduced to –3.0 and –3.5 kcal/mol, respectively (Table 1), which is consistent with fewer molecular interactions between the effector and these mutants. Affinity of the TtgVV140A, TtgVV204A, and TtgVV223A mutants was reduced by one-half to two-thirds of that of the native TtgV protein (Table 1). We also found that the single mutant TtgV I118A showed 2-fold higher affinity for 4NT than the wild-type protein.TABLE 1Thermodynamic parameters derived from the microcalorimetric titration of TtgV with one- and two-ring aromatic effectorsProteinEffectorKDKDwt/KDΔHEffectorKDKDwt/KDΔHμmkcal/molμmkcal/molWTaWild-type TtgV protein.4NT17.4 ± 0.61.0-9.7 ± 0.2BN50 ± 31.0-5.7 ± 0.1F134A4NT221 ± 80.08-3.0 ± 0.1BN330 ± 100.15-3.5 ± 0.1F134V4NTNo bindingBNNo bindingH200A4NT470 ± 200.04-3.5 ± 0.1BNNo bindingH200V4NT83 ± 100.2-13 ± 2BNNo bindingI118A4NT8.9 ± 0.42.0-8.6 ± 0.2BN46 ± 21.1-14 ± 1V140A4NT30 ± 10.6-6.6 ± 0.2BN85 ± 30.6-6.3 ± 0.5V140F4NT24 ± 10.7-5.9 ± 0.1BN77 ± 70.6-4.8 ± 0.8V204A4NT41 ± 20.4-14.3 ± 0.1BN71 ± 50.7-13 ± 4V223A4NT59 ± 30.3-6.4 ± 0.4BN58 ± 40.9-10 ± 2WT1NL40 ± 31.0-21 ± 3IND78 ± 71.0-6 ± 2F134A1NL5.7 ± 0.27.0-22.5 ± 0.6IND32 ± 22.4-12 ± 1F134V1NL8 ± 16.7-47 ± 2IND44 ± 61.8-11 ± 1H200A1NL2.9 ± 0.413.8-16 ± 3IND39 ± 32.0-2.9 ± 0.6H200V1NL5.1 ± 0.87.8-26.5 ± 0.1IND49 ± 52.0-13 ± 9I118A1NL24 ± 21.7-16 ± 1IND46 ± 21.7-16 ± 2V140A1NL6.5 ± 0.56.2-22 ± 1IND21 ± 23.7-14 ± 2V140F1NL2.5 ± 0.116-20.5 ± 0.5IND19 ± 24.1-15.7 ± 0.6V204A1NL23.8 ± 0.81.7-23 ± 1IND37 ± 22.1-10 ± 1V223A1NL6.2 ± 0.46.5-16.8 ± 0.6IND35 ± 22.2-9 ± 1a Wild-type TtgV protein. Open table in a new tab The tendency in affinity observed for 4NT was completely opposite when the same experiments were repeated with the biaromatic effector 1NL; all mutants showed a higher affinity than the wild-type protein (Table 1). This is illustrated in Fig. 3B for the wild type and the F134A mutant, which bound 1NL with KD values of 40.1 ± 3 and 5.7 ± 0.2 μm, respectively. All six mutants exhibited affinities for 1NL that were between 1.7- and 13.8-fold higher. Most interestingly, the gain in affinity for 1NL was most pronounced for the mutants that had the lowest affinity for the monoaromatic 4NT, i.e. the F134A and H200A mutants. To establish whether the differential binding behavior of these two effectors is a general phenomenon for mono- and biaromatic effectors, ITC experiments were conducted with two other representative effectors, namely BN (monoaromatic) and IND (biaromatic). In a manner exactly analogous to 4NT, all mutants except I118A had decreased affinity for the monoaromatic compound BN. This reduction in affinity was again particularly pronounced for F134A and H200A (Table 1). In analogy to 4NT, the enthalpic contribution to the binding of BN to F134A was again reduced, and no detectable affinity was observed for the H200A protein. As in the experiments with 1NL, all mutants were found to have increased affinities with respect to IND (Table 1). We hypothesized that the increase in affinity for bicyclic compounds is founded on the increase in the volume of the binding pocket caused by the replacement of bulky amino acids. To further study this unexpected finding, we generated mutants in which Phe-134 and His-200 were replaced by valine. The microcalorimetric titrations also showed that both mutants bound monocyclic compounds with decreased affinity when compared with the wild-type protein (Table 1). Although F134A and H200V bound bicyclic compounds with higher affinity than the parental wild-type protein, affinity was lower than that of the corresponding alanine replacement mutants. Therefore, these results support the above hypothesis. We have also generated the V140F change. In this case, we found that the decrease in the size of the pocket resulted in small changes in affinity for the 4NT, BN, and IND compounds when compared with the V140A mutant. In contrast, 1NL bound with a higher affinity to V140F than to V140A. This indicates that apart from the volume increase, there are other factors, such as the establishment of additional van der Waals interactions between Val-140 and 1-NL, that can influence effector recognition (3.Murakami S. Nakashima R. Yamashita E. Yamaguchi A. Nature. 2002; 419: 587-593Crossref PubMed Scopus (766) Google Scholar, 5.Yu E.W. McDermont G. Zgurskaya H.I. Nikaido H. Koshland Jr., D.E. Science. 2003; 300: 976-980Crossref PubMed Scopus (338) Google Scholar, 20.Schumacher M.A. Miller M.C. Brennan R.G. EMBO J. 2004; 23: 2923-2930Crossref PubMed Scopus (106) Google Scholar, 21.Grkovic S. Brown M.H. Roberts N.J. Paulsen I.T. Skurray R.A. J. Biol. Chem. 1998; 273: 18665-18673Abstract Full Text Full Text PDF PubMed Scopus (156) Google Scholar). We also determined the affinity of the wild-type TtgV and its mutant variants when complexed with DNA for mono- and biaromatic effectors. We found, in agreement with previous findings (12.Guazzaroni M.-E. Krell T. Felipe A. Ruiz R. Meng C. Zhang X. Gallegos M.-T. Ramos J.L. J. Biol. Chem. 2005; 280: 20887-20893Abstract Full Text Full Text PDF PubMed Scopus (52) Google Scholar), that affinity of TtgV and TtgV mutants for their effectors when bound to DNA was not altered in a significant manner. To illustrate this, we performed experiments involving the titration of free and DNA-bound TtgVV223A with 1NL (Fig. 4). 1NL bound to free and DNA-bound TtgVV223A with similar affinities, as evidenced by their respective dissociation constants of 6.2 ± 0.4 and 4.8 ± 0.1 μm. Effect of Mutations on Affinity for Operator DNA−Subsequent experiments were aimed at elucidating whether the mutations in the effector-binding site altered the DNA binding characteristics of the mutants. EMSAs using 1 nm 210-bp fragment correspon
Referência(s)