Mutational Analysis of TraR
2003; Elsevier BV; Volume: 278; Issue: 15 Linguagem: Inglês
10.1074/jbc.m210035200
ISSN1083-351X
AutoresZhao‐Qing Luo, Audra J. Smyth, Ping Gao, Yinping Qin, Stephen K. Farrand,
Tópico(s)Legume Nitrogen Fixing Symbiosis
ResumoTraR, the quorum-sensing activator of theAgrobacterium tumefaciens Ti plasmid conjugation system, induces gene expression in response to its quormone,N-(3-oxooctanoyl)-l-homoserine lactone. Ligand binding results in dimerization of TraR and is required for its activity. Analysis of N- and C-terminal deletion mutants of TraR localized the quormone-binding domain to a region between residues 39 and 140 and the primary dimerization domain to a region between residues 119 and 156. The dominant-negative properties of these mutants predicted a second dimerization domain at the C terminus of the protein. Analysis of fusions of N-terminal fragments of TraR to λcI′ confirmed the dimerization activity of these two domains. Fifteen single amino acid substitution mutants of TraR defective in dimerization were isolated. According to the analysis of these mutants, Asp-70 and Gly-113 are essential for quormone binding, whereas Ala-38 and Ala-105 are important, but not essential. Additional residues located within the N-terminal half of TraR, including three located in α-helix 9, contribute to dimerization, but are not required for ligand binding. These results and the recently reported crystal structure of TraR are consistent with and complement each other and together define some of the structural and functional relationships of this quorum-sensing activator. TraR, the quorum-sensing activator of theAgrobacterium tumefaciens Ti plasmid conjugation system, induces gene expression in response to its quormone,N-(3-oxooctanoyl)-l-homoserine lactone. Ligand binding results in dimerization of TraR and is required for its activity. Analysis of N- and C-terminal deletion mutants of TraR localized the quormone-binding domain to a region between residues 39 and 140 and the primary dimerization domain to a region between residues 119 and 156. The dominant-negative properties of these mutants predicted a second dimerization domain at the C terminus of the protein. Analysis of fusions of N-terminal fragments of TraR to λcI′ confirmed the dimerization activity of these two domains. Fifteen single amino acid substitution mutants of TraR defective in dimerization were isolated. According to the analysis of these mutants, Asp-70 and Gly-113 are essential for quormone binding, whereas Ala-38 and Ala-105 are important, but not essential. Additional residues located within the N-terminal half of TraR, including three located in α-helix 9, contribute to dimerization, but are not required for ligand binding. These results and the recently reported crystal structure of TraR are consistent with and complement each other and together define some of the structural and functional relationships of this quorum-sensing activator. N-acylated homoserine lactone N-(3-oxohexanoyl)-l-homoserine lactone N-(3-oxooctanoyl)-l-homoserine lactone N-octanoylhomoserine lactone 5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside isopropyl-β-d-thiogalactopyranoside Many bacteria use a cell-cell signaling system to control expression of specialized gene sets in response to their population size. Called quorum sensing, the bacteria measure their cell numbers by sensing the accumulation within the environment of a small diffusible signal molecule, the quormone, which they themselves produce (reviewed in Ref. 1Whitehead N.A. Barnard A.M.L. Slater H. Simpson N.J.L. Salmond G.P.C. FEMS Microbiol. Rev. 2001; 25: 365-404Crossref PubMed Google Scholar). Gram-negative bacteria often use N-acylated homoserine lactones (acyl-HSL)1 as their quormone, with signal specificity being determined by the length and degree of saturation of the acyl side chain as well as the nature of the chemical group at C-3. For example, Vibrio fischeri, which regulates the expression of the luxoperon by quorum sensing (2Sitnikov D.M. Schineller J.B. Baldwin T.O. Mol. Microbiol. 1995; 17: 801-812Crossref PubMed Scopus (131) Google Scholar), produces and responds toN-(3-oxohexanoyl)-l-homoserine lactone (3-oxo-C6-HSL) (3Eberhard A. Burlingame A.L. Eberhard C. Kenyon G.L. Nealson K.H. Oppenheimer N.J. Biochemistry. 1981; 20: 2444-2449Crossref PubMed Scopus (668) Google Scholar), whereas Rhizobium leguminosarum regulates expression of a network of genes (4Lithgow J.K. Wilkinson A. Hardman A. Rodelas B. Wisniewski-Dye F. Williams P. Downie J.A. Mol. Microbiol. 2000; 37: 81-97Crossref PubMed Scopus (170) Google Scholar) in a quorum-dependent manner usingN-(3-hydroxy-7-cis-tetradecanoyl)-l-homoserine lactone as the signal (5Gray K.M. Pearson J.P. Downie J.A. Boboye B.E.A. Greenberg E.P. J. Bacteriol. 1996; 178: 372-376Crossref PubMed Google Scholar, 6Schripsema J. de Rudder K.E.E. van Vliet T.B. Lankhorst P.P. de Vroom E. Kijne J.W. van Brussel A.A.N. J. Bacteriol. 1996; 178: 366-371Crossref PubMed Google Scholar).In the acyl-HSL-mediated systems examined to date, expression of the target genes is controlled by a dedicated transcription factor, usually an activator (1Whitehead N.A. Barnard A.M.L. Slater H. Simpson N.J.L. Salmond G.P.C. FEMS Microbiol. Rev. 2001; 25: 365-404Crossref PubMed Google Scholar). These proteins, of which LuxR is the prototype, require the quormone to activate transcription. Genetic evidence from studies of LuxR suggests that the activator binds the acyl-HSL ligand and, in doing so, converts from an inactive to an active form (7Fuqua W.C. Winans S.C. Greenberg E.P. J. Bacteriol. 1994; 176: 269-275Crossref PubMed Scopus (2108) Google Scholar). LuxR (and by inference, all other members of the family) is believed to be composed of two functional regions, a C-terminal DNA recognition and interaction domain and an N-terminal quormone-binding domain (7Fuqua W.C. Winans S.C. Greenberg E.P. J. Bacteriol. 1994; 176: 269-275Crossref PubMed Scopus (2108) Google Scholar).Recent studies of TraR, the quorum-sensing transcriptional activator responsible for controlling conjugal transfer of theAgrobacterium tumefaciens Ti plasmids, are consistent with this model. TraR activates expression of the three operons of the Ti plasmid tra regulon from promoters that contain an 18-bp inverted repeat called the tra box (8Farrand S.K. Spaink H.P. Kondorosi A. Hooykaas P.J.J. The Rhizobiaceae: Molecular Biology of Plant-associated Bacteria. Kluwer Academic Publishers Group, Dordrecht, The Netherlands1998: 199-233Crossref Google Scholar). The acyl-HSL quormone N-(3-oxooctanoyl)-l-homoserine lactone (3-oxo-C8-HSL) (9Zhang L. Murphy P.J. Kerr A. Tate M.E. Nature. 1993; 362: 446-448Crossref PubMed Scopus (355) Google Scholar) is required for TraR to bind thetra box (10Luo Z.-Q. Farrand S.K. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 9009-9014Crossref PubMed Scopus (85) Google Scholar). Moreover, purified TraR binds the ligand tightly (11Qin Y Luo Z.-Q. Smyth A.J. Gao P. Beck von Bodman S. Farrand S.K. EMBO J. 2000; 19: 5212-5221Crossref PubMed Scopus (128) Google Scholar, 12Zhu J. Winans S.C. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 4832-4837Crossref PubMed Scopus (220) Google Scholar), with such binding resulting in dimerization of the protein (11Qin Y Luo Z.-Q. Smyth A.J. Gao P. Beck von Bodman S. Farrand S.K. EMBO J. 2000; 19: 5212-5221Crossref PubMed Scopus (128) Google Scholar). No higher order multimers are formed at detectable levels (11Qin Y Luo Z.-Q. Smyth A.J. Gao P. Beck von Bodman S. Farrand S.K. EMBO J. 2000; 19: 5212-5221Crossref PubMed Scopus (128) Google Scholar), indicating that TraR is active only in the dimer form. These observations suggest a model in which, in the absence of the quormone, TraR fails to dimerize and cannot bind the tra box. In the presence of the signal, TraR binds the quormone, thereby forming stable homodimers. The dimer binds the tra boxes and, in interaction with RNA polymerase, activates transcription.TraR is the only activator of the LuxR family that can be purified in its full-sized active form, making it an ideal candidate for molecular and biochemical analysis (12Zhu J. Winans S.C. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 4832-4837Crossref PubMed Scopus (220) Google Scholar). Moreover, genetic assays and biochemical techniques are available to examine the DNA-binding, dimerization, and transcriptional activation properties of this regulator (10Luo Z.-Q. Farrand S.K. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 9009-9014Crossref PubMed Scopus (85) Google Scholar, 11Qin Y Luo Z.-Q. Smyth A.J. Gao P. Beck von Bodman S. Farrand S.K. EMBO J. 2000; 19: 5212-5221Crossref PubMed Scopus (128) Google Scholar, 12Zhu J. Winans S.C. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 4832-4837Crossref PubMed Scopus (220) Google Scholar). Mutations in the C-terminal domain of TraR abolish DNA binding (10Luo Z.-Q. Farrand S.K. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 9009-9014Crossref PubMed Scopus (85) Google Scholar), consistent with the presence of a helix-turn-helix motif located within this region of the protein (13Piper K.R. Beck von Bodman S. Farrand S.K. Nature. 1993; 362: 448-450Crossref PubMed Scopus (381) Google Scholar, 14Fuqua W.C. Winans S.C. J. Bacteriol. 1994; 176: 2796-2806Crossref PubMed Scopus (390) Google Scholar). In addition, certain residues located at the far N terminus and in the middle of the protein are important for transcriptional activation, but not for DNA binding (10Luo Z.-Q. Farrand S.K. Proc. Natl. Acad. Sci. U. S. A. 1999; 96: 9009-9014Crossref PubMed Scopus (85) Google Scholar). Analysis of a series of stable N- and C-terminal deletion mutants identified a region of TraR located between residues 49 and 156 that is important for dimerization (11Qin Y Luo Z.-Q. Smyth A.J. Gao P. Beck von Bodman S. Farrand S.K. EMBO J. 2000; 19: 5212-5221Crossref PubMed Scopus (128) Google Scholar). These results suggest that, like LuxR, TraR is composed of an N-terminal dimerization domain and a C-terminal DNA-binding domain, a conclusion that is supported by the recent reports of the x-ray crystal structure of TraR bound to its DNA recognition element (15Zhang R.-G. Pappas T. Brace J.L. Miller P.C. Oulmassov T. Molyneaux J.M. Anderson J.C. Bashkin J.K. Winans S.C. Joachimiak A. Nature. 2002; 417: 971-974Crossref PubMed Scopus (358) Google Scholar, 16Vannini A. Volpari C. Gargioli C. Muraglia E. Cortese R. De Francesco R. Neddermann P. Di Marco S. EMBO J. 2002; 21: 4393-4401Crossref PubMed Scopus (284) Google Scholar). However, although the x-ray structure predicts regions and residues involved in ligand binding as well as in dimerization, the importance and contributions of these residues in the two processes have not been assessed by biochemical or genetic analyses.In this study, we combine a genetic analysis with biochemical tests to identify regions of TraR required for ligand binding and dimerization. We also report the isolation of a series of substitution mutants of TraR defective in these properties and have used these mutants to identify residues of the activator that are critical for binding the quormone and for forming dimers. Our results indicate that certain conserved residues are essential for binding acyl-HSL; but, although ligand binding is required for the process, no single residue is essential for dimerization per se. In addition, although two residues were identified as being essential for ligand binding, substitutions at several positions that hinder dimerization alter, but do not abolish, the ligand-binding properties of the protein.DISCUSSIONWe reported previously that purified active TraR is a homodimer and that formation of stable dimers depends upon binding the acyl-HSL quormone (11Qin Y Luo Z.-Q. Smyth A.J. Gao P. Beck von Bodman S. Farrand S.K. EMBO J. 2000; 19: 5212-5221Crossref PubMed Scopus (128) Google Scholar). Recently reported crystallographic analyses confirm this prediction and identify sets of residues, all in the N-terminal half of TraR, that contribute to ligand binding and to dimerization (15Zhang R.-G. Pappas T. Brace J.L. Miller P.C. Oulmassov T. Molyneaux J.M. Anderson J.C. Bashkin J.K. Winans S.C. Joachimiak A. Nature. 2002; 417: 971-974Crossref PubMed Scopus (358) Google Scholar, 16Vannini A. Volpari C. Gargioli C. Muraglia E. Cortese R. De Francesco R. Neddermann P. Di Marco S. EMBO J. 2002; 21: 4393-4401Crossref PubMed Scopus (284) Google Scholar). Our studies of dimerization-defective mutants of TraR generally support these assignments. However, we show that substitutions at residues predicted to be involved in quormone binding exert different effects, ranging from abolishing binding outright to modifying the nature of the binding, resulting in altered responses to cognate and non-cognate signals.Of the 15 substitution mutants we isolated, alterations at only two residues, Asp-70 and Gly-113, completely abolished quormone binding. Not surprisingly, both mutants failed to dimerize and lacked detectable biological activity. The two amino acids are among five residues that are conserved in all acyl-HSL-dependent members of the LuxR family (1Whitehead N.A. Barnard A.M.L. Slater H. Simpson N.J.L. Salmond G.P.C. FEMS Microbiol. Rev. 2001; 25: 365-404Crossref PubMed Google Scholar). Moreover, the corresponding residues of LuxR (Asp-79) and LasR (Asp-73) also are important for transcriptional activation (33Shadel G.S. Young R. Baldwin T.O. J. Bacteriol. 1990; 172: 3980-3987Crossref PubMed Google Scholar, 34Kiratisin P. Tucker K.D. Passador L. J. Bacteriol. 2002; 184: 4912-4919Crossref PubMed Scopus (142) Google Scholar) and, in the case of LasR, for dimerization (34Kiratisin P. Tucker K.D. Passador L. J. Bacteriol. 2002; 184: 4912-4919Crossref PubMed Scopus (142) Google Scholar). That a mutation at Asp-70 affects binding is predictable; in the crystal structure, the side chain carboxyl group of this residue forms a hydrogen bond with the imino group of the quormone (Fig.6A) (15Zhang R.-G. Pappas T. Brace J.L. Miller P.C. Oulmassov T. Molyneaux J.M. Anderson J.C. Bashkin J.K. Winans S.C. Joachimiak A. Nature. 2002; 417: 971-974Crossref PubMed Scopus (358) Google Scholar, 16Vannini A. Volpari C. Gargioli C. Muraglia E. Cortese R. De Francesco R. Neddermann P. Di Marco S. EMBO J. 2002; 21: 4393-4401Crossref PubMed Scopus (284) Google Scholar). Our analyses show that Asp-70 and its contributed hydrogen bond are essential for binding signal.The two replacements at Asp-70, D70N and D70E, provide an informative contrast. Whereas C-terminally repaired D70N was completely inactive, the D70E mutant was very weakly active in cells grown with high levels of quormone (Table II). Asparagine differs from the native Asp predominantly in charge, whereas the glutamate retains the acidic function, but the side chain is larger by one methylene group. We conclude that, although the bulkiness of the side group at position 70 is important, this residue must contain an acidic function, a prediction that is consistent with the structural studies (15Zhang R.-G. Pappas T. Brace J.L. Miller P.C. Oulmassov T. Molyneaux J.M. Anderson J.C. Bashkin J.K. Winans S.C. Joachimiak A. Nature. 2002; 417: 971-974Crossref PubMed Scopus (358) Google Scholar, 16Vannini A. Volpari C. Gargioli C. Muraglia E. Cortese R. De Francesco R. Neddermann P. Di Marco S. EMBO J. 2002; 21: 4393-4401Crossref PubMed Scopus (284) Google Scholar).The G113S mutation exerted a profound effect on TraR activity; and clearly, the glycine at this position is essential for quormone binding and subsequent dimerization. The corresponding residues in LuxR (Gly-121) and LasR (Gly-113) also are required for biological activity (2Sitnikov D.M. Schineller J.B. Baldwin T.O. Mol. Microbiol. 1995; 17: 801-812Crossref PubMed Scopus (131) Google Scholar, 33Shadel G.S. Young R. Baldwin T.O. J. Bacteriol. 1990; 172: 3980-3987Crossref PubMed Google Scholar, 34Kiratisin P. Tucker K.D. Passador L. J. Bacteriol. 2002; 184: 4912-4919Crossref PubMed Scopus (142) Google Scholar) and, in LasR, for dimerization (34Kiratisin P. Tucker K.D. Passador L. J. Bacteriol. 2002; 184: 4912-4919Crossref PubMed Scopus (142) Google Scholar). The structural analysis did not predict a direct interaction between Gly-113, located in β-sheet 4, and the acyl-HSL (Figs. 3 and 6A) (15Zhang R.-G. Pappas T. Brace J.L. Miller P.C. Oulmassov T. Molyneaux J.M. Anderson J.C. Bashkin J.K. Winans S.C. Joachimiak A. Nature. 2002; 417: 971-974Crossref PubMed Scopus (358) Google Scholar, 16Vannini A. Volpari C. Gargioli C. Muraglia E. Cortese R. De Francesco R. Neddermann P. Di Marco S. EMBO J. 2002; 21: 4393-4401Crossref PubMed Scopus (284) Google Scholar), but the glycine may position Trp-85, located in β-sheet 3, which could interact with the quormone (Fig.7A) (15Zhang R.-G. Pappas T. Brace J.L. Miller P.C. Oulmassov T. Molyneaux J.M. Anderson J.C. Bashkin J.K. Winans S.C. Joachimiak A. Nature. 2002; 417: 971-974Crossref PubMed Scopus (358) Google Scholar). Zhang et al. (15Zhang R.-G. Pappas T. Brace J.L. Miller P.C. Oulmassov T. Molyneaux J.M. Anderson J.C. Bashkin J.K. Winans S.C. Joachimiak A. Nature. 2002; 417: 971-974Crossref PubMed Scopus (358) Google Scholar) concluded that Trp-85 interacts with the acyl side chain, whereas Vannini et al. (16Vannini A. Volpari C. Gargioli C. Muraglia E. Cortese R. De Francesco R. Neddermann P. Di Marco S. EMBO J. 2002; 21: 4393-4401Crossref PubMed Scopus (284) Google Scholar) predicted that this residue interacts with the lactone moiety. Our examination of the crystal structure is consistent with the latter interpretation (Fig.7A). We suggest that the bulkier side group of the serine substitution in the G113S mutant perturbs the correct positioning of Trp-85 (Fig. 7B), thereby inhibiting quormone binding.Figure 7Specific interactions between residues of TraR and between residues of TraR and the acyl-HSL ligand. A, possible interactions of Trp-85 with the homoserine lactone moiety of 3-oxo-C8-HSL and with Gly-113.B, the serine substitution in the G113S mutant with its altered side group in white. C, possible main chain hydrogen bonding between Thr-190 and the amide of Glu-193 located in α-helix 11, the scaffold helix of the helix-turn-helix motif.D, the isoleucine substitution at Thr-190. Groups with oxygen atoms are shown in red, and those with nitrogen atoms are shown in blue. Distances are in Angstroms.View Large Image Figure ViewerDownload (PPT)A mutation at Ala-38 also affected signal retention. However, although the mutant was defective, unlike the Asp-70 and G-113 mutants, activity was restored by addition of excess quormone (Fig. 5). Ala-38 is located in the ligand-binding site (Fig. 6A) and, along with Thr-129, is predicted to contribute to a water bridge with the 3-oxo group of the quormone (15Zhang R.-G. Pappas T. Brace J.L. Miller P.C. Oulmassov T. Molyneaux J.M. Anderson J.C. Bashkin J.K. Winans S.C. Joachimiak A. Nature. 2002; 417: 971-974Crossref PubMed Scopus (358) Google Scholar). The A38V mutant still bound a small amount of signal, suggesting that the valine substitution can participate, although poorly, in the coordination of the water molecule. Alternatively, the decreased binding may reflect the contribution of Thr-129 alone. Consistent with its limited signal binding, the A38V mutant formed heterodimers, albeit at low levels, with His-tagged TraR (Fig. 4B). Moreover, although functional, the C-terminally repaired mutant activated expression of the reporter to levels only half those of the wild-type activator (Fig. 5). Furthermore, the mutant was as sensitive as wild-type TraR to the cognate acyl-HSL, but was considerably less sensitive to the non-cognate signals (Fig. 5). We suggest that the bulkier valine substitution interferes with the water bridge, resulting in a decreased affinity for the quormone. However, this defect can be partially overcome by addition of excess ligand. Given the disproportionately decreased sensitivity to 3-oxo-C6-HSL, we speculate that, in the absence of a water bridge, the van der Waals interactions with the acyl side chain take on added importance for signal binding. Ala-38 is not highly conserved among sequenced homologs of LuxR (1Whitehead N.A. Barnard A.M.L. Slater H. Simpson N.J.L. Salmond G.P.C. FEMS Microbiol. Rev. 2001; 25: 365-404Crossref PubMed Google Scholar). Among those members that recognize acyl-HSLs with oxo or hydroxy substitutions at C-3 in which a water bridge might be important, the analogous positions can be occupied by alanine, serine, and the considerably bulkier residues threonine and leucine, but none contain valine (1Whitehead N.A. Barnard A.M.L. Slater H. Simpson N.J.L. Salmond G.P.C. FEMS Microbiol. Rev. 2001; 25: 365-404Crossref PubMed Google Scholar).Like the A38V mutant, cells expressing the A105V mutant retained only small amounts of quormone (Fig. 2C). Moreover, the two mutants formed very weak heterodimers with His6-TraR (Fig.4) and also exhibited reduced sensitivity to the non-cognate signals (Fig. 5). However, whereas A38V activated the reporter to only about half its maximal level, the A105V mutant induced the same reporter to wild-type levels. Ala-105 contributes to quormone binding through a van der Waals interaction with the homoserine lactone moiety (Fig.6A) (16Vannini A. Volpari C. Gargioli C. Muraglia E. Cortese R. De Francesco R. Neddermann P. Di Marco S. EMBO J. 2002; 21: 4393-4401Crossref PubMed Scopus (284) Google Scholar). Moreover, the residue is conserved in most members of the LuxR family (1Whitehead N.A. Barnard A.M.L. Slater H. Simpson N.J.L. Salmond G.P.C. FEMS Microbiol. Rev. 2001; 25: 365-404Crossref PubMed Google Scholar), and an A105V mutation in LasR has similar effects on the activity of the Pseudomonas aeruginosaactivator (34Kiratisin P. Tucker K.D. Passador L. J. Bacteriol. 2002; 184: 4912-4919Crossref PubMed Scopus (142) Google Scholar). We suggest that the bulkier valine substitution at this position affects the capacity of TraR to initially bind the quormone. However, once bound, the protein forms stable dimers, and the ligand is retained strongly. This conclusion is consistent with the wild-type levels of activity exhibited by the mutant when exposed to higher levels of quormone (Fig. 5). Moreover, when tested at high levels of signal, cells expressing the A105V mutant retained as much signal as wild-type TraR (data not shown).Strains expressing the T51I mutant retained levels of ligand indistinguishable from those expressing wild-type TraR. Thr-51 is predicted to form a van der Waals contact with the distal region of the acyl chain of the quormone (Fig. 6A) (15Zhang R.-G. Pappas T. Brace J.L. Miller P.C. Oulmassov T. Molyneaux J.M. Anderson J.C. Bashkin J.K. Winans S.C. Joachimiak A. Nature. 2002; 417: 971-974Crossref PubMed Scopus (358) Google Scholar). That strains expressing the T51I mutant retained ligand indicates that the isoleucine substitution has little effect on signal binding. However, the mutant lost most of the dominant negativity of the parent (TableII) and did not form strong heterodimers with His6-TraR. Moreover, the C-terminally repaired T51I mutant was activable by excess quormone, but to levels only one-fourth that of wild-type TraR (Fig.5). These observations suggest that the isoleucine substitution, although having little effect on signal binding, interferes strongly with the formation or stability of the dimers. It is conceivable that replacing the hydrophilic threonine residue with the bulkier hydrophobic isoleucine introduces a steric perturbation that inhibits dimerization.Of the remaining nine dimerization-defective mutants, three, A149V, L155F, and S160F, map to α-helix 9 (Ref. 15Zhang R.-G. Pappas T. Brace J.L. Miller P.C. Oulmassov T. Molyneaux J.M. Anderson J.C. Bashkin J.K. Winans S.C. Joachimiak A. Nature. 2002; 417: 971-974Crossref PubMed Scopus (358) Google Scholar; α-helix 6 in Ref. 16Vannini A. Volpari C. Gargioli C. Muraglia E. Cortese R. De Francesco R. Neddermann P. Di Marco S. EMBO J. 2002; 21: 4393-4401Crossref PubMed Scopus (284) Google Scholar), the motif predicted to constitute the primary dimer interface of TraR (Figs. 3 and 6B) (15Zhang R.-G. Pappas T. Brace J.L. Miller P.C. Oulmassov T. Molyneaux J.M. Anderson J.C. Bashkin J.K. Winans S.C. Joachimiak A. Nature. 2002; 417: 971-974Crossref PubMed Scopus (358) Google Scholar, 16Vannini A. Volpari C. Gargioli C. Muraglia E. Cortese R. De Francesco R. Neddermann P. Di Marco S. EMBO J. 2002; 21: 4393-4401Crossref PubMed Scopus (284) Google Scholar). All three mutants bound quormone as well as wild-type TraR (Fig. 2) (data not shown). Moreover, all three retained some dominant negativity and, in their C-terminally repaired forms, activated the traG reporter, albeit to low levels, suggesting that they still form dimers (Table II). These properties are consistent with the analysis of the C-terminal deletion mutants; TraRΔC115, from which all of α-helix 9 was removed, retained some dominant negativity (Fig. 1), suggesting that it can form weak heterodimers with wild-type TraR. The response of the L155F mutant to cognate and non-cognate quormones emphasizes the importance of the α-helix 9 region; the mutant activated the reporter, but only to very low levels even at saturating amounts of signal (Fig. 5). Moreover, the mutant responded to the non-cognate quormones in a manner similar to that of the cognate acyl-HSL, consistent with our conclusion that the phenylalanine substitution alters dimerization of the protein, but not its signal-binding properties.Four mutants, L25F, H54Y, E88K, and S112F, map to the N-terminal region of TraR (Fig. 3), but are not predicted to contribute to dimerization or signal binding (15Zhang R.-G. Pappas T. Brace J.L. Miller P.C. Oulmassov T. Molyneaux J.M. Anderson J.C. Bashkin J.K. Winans S.C. Joachimiak A. Nature. 2002; 417: 971-974Crossref PubMed Scopus (358) Google Scholar, 16Vannini A. Volpari C. Gargioli C. Muraglia E. Cortese R. De Francesco R. Neddermann P. Di Marco S. EMBO J. 2002; 21: 4393-4401Crossref PubMed Scopus (284) Google Scholar). All four mutant proteins bound quormone as well as wild-type TraR (Fig. 2) (data not shown), and all retained some biological activity (Table II). However, the mutants did not interact with His6-TraR (Fig. 4B) (data not shown). Leu-25 and Glu-88 lie within α-helices 2 and 6, respectively, and Ser-112 is located within β-sheet 4 (Fig. 6D) (15Zhang R.-G. Pappas T. Brace J.L. Miller P.C. Oulmassov T. Molyneaux J.M. Anderson J.C. Bashkin J.K. Winans S.C. Joachimiak A. Nature. 2002; 417: 971-974Crossref PubMed Scopus (358) Google Scholar). His-54 is located within a group of residues involved in quormone binding (15Zhang R.-G. Pappas T. Brace J.L. Miller P.C. Oulmassov T. Molyneaux J.M. Anderson J.C. Bashkin J.K. Winans S.C. Joachimiak A. Nature. 2002; 417: 971-974Crossref PubMed Scopus (358) Google Scholar); but, in the three-dimensional structure, this amino acid is distant from the binding site (Fig. 6D). All four residues cluster in the same general region (Fig. 6D), and it is conceivable that steric effects induced by substitutions at these sites adversely alter the dimer properties of the protein.The remaining two mutations are located C-terminal to the linker joining the N- and C-terminal domains (Fig. 3). E211K is altered in a residue located in the recognition helix of the DNA-binding domain (Fig. 6C). Zhang et al. (15Zhang R.-G. Pappas T. Brace J.L. Miller P.C. Oulmassov T. Molyneaux J.M. Anderson J.C. Bashkin J.K. Winans S.C. Joachimiak A. Nature. 2002; 417: 971-974Crossref PubMed Scopus (358) Google Scholar) predicted a van der Waals interaction between Glu-211 and T-8. In agreement with this assignment, the mutation substantially inhibited the dominant-negative properties of TraRΔC1 (Table II). This defect is consistent with the observation that E211K formed weak heterodimers with His6-TraR (data not shown). Moreover, the C-terminally repaired mutant showed only weak activator activity (Table II). In contrast to these observations, an alanine substitution at Asn-220, the corresponding residue of LuxR, has little to no effect on transcriptional activation by this protein (35Egland K.R. Greenberg E.P. J. Bacteriol. 2001; 183: 382-386Crossref PubMed Scopus (65) Google Scholar, 36Trott A.E. Stevens A.M. J. Bacteriol. 2001; 183: 387-392Crossref PubMed Scopus (23) Google Scholar).The T190I mutant is unique among our isolates. Strains expressing this mutant retained normal levels of quormone, and the protein only weakly dimerized with His6-TraR. However, like the ligand binding-defective mutants, in its C-terminally reconstituted form, T190I failed to detectably activate the lacZ reporter even at high quormone concentrations (Fig. 5). Although Thr-190 is not predicted to participate in signal binding or dimerization (15Zhang R.-G. Pappas T. Brace J.L. Miller P.C. Oulmassov T. Molyneaux J.M. Anderson J.C. Bashkin J.K. Winans S.C. Joachimiak A. Nature. 2002; 417: 971-974Crossref PubMed Scopus (358) Google Scholar, 16Vannini A. Volpari C. Gargioli C. Muraglia E. Cortese R. De Francesco R. Neddermann P. Di Marco S. EMBO J. 2002; 21: 4393-4401Crossref PubMed Scopus (284) Google Scholar), Vannini et al. (16Vannini A. Volpari C. Gargioli C. Muraglia E. Cortese R. De Francesco R. Neddermann P. Di Marco S. EMBO J. 2002; 21: 4393-4401Crossref PubMed Scopus (284) Google Scholar) proposed that this residue makes hydrophobic contact with the phosphate group of cytosine 11 of thetra box, a conclusion consistent with its location near the helix-turn-helix domain (Fig. 6C). However, if this interaction is correct, it is difficult to understand how a substitution at this residue results in such a drastic loss of activity. Our analysis of the crystal structure predicts that the side group hydroxyl of Thr-190 forms a main chain hydrogen bond with the amide of Glu-193, which is located in α-helix 11, the scaffold component of the helix-turn-helix motif (Fig. 7C). An isoleucine sub
Referência(s)