Tài liệu Báo cáo khoa học: Tet repressor mutants with altered effector binding and allostery docx

Tet repressor mutants with altered effector binding and allostery Eva-Maria Henßler, Ralph Bertram, Stefanie Wisshak and Wolfgang Hillen Lehrstuhl fu ¨ r Mikrobiologie, Institut fu ¨ r Biologie, Friedrich-Alexander Universita ¨ t Erlangen-Nu ¨ rnberg, Erlangen, Germany Tetracycline (tc) resistance in Gram-negative bacteria is often regulated by the Tet Repressor (TetR), a tc- responsive allosterical DNA-binding protein. Due to three very advantageous properties, namely the highly specific binding of TetR to tet operator (tetO), the sen- sitive induction by small amounts of tc, and the ability of this drug to penetrate into most cells, the TetR based regulation systems are widely used for condi- tional gene expression [1]. Many biochemical studies and crystal structures of TetR in all complexed forms [2–4] have led to a detailed understanding of the regu- latory mechanism. TetR is an all a-helical, dimeric protein in which tetO recognition is accomplished by a helix-turn-helix motif consisting of helices a2 and a3 at the N-termi- nus. The core domain (a5toa10) contains the tc bind- ing pocket and the dimerization motif. Both domains are connected by helix a4, and their interface is formed by residues of helices a1, a4, and a6 (Fig. 1A). As the [tc-Mg] + binding site is 33 A ˚ away from the tetO binding site, the structural changes associated with induction of TetR must be transfered through the pro- tein. They are initiated at the residues 100–103 which are part of helix a6 in the DNA-binding conformation and assume a type II b-turn to contact [tc-Mg] + in the induced state. The transmission of structural changes to the DNA binding domains occurs via helices a4 and moves them by about 5° in a pendulum-like motion so that the recognition helices no longer fit into successive major grooves of DNA (Fig. 1B). Extensive mutagenesis employing powerful selection and screening systems have led to many TetR variants with new activities. Among them was a TetR variant with changed inducer specificity [5]. Instead of tc or the more powerful inducer anhydrotetracycline (atc) the TetR H64K S135L S138I triple mutant (TetR i2 ) recogni- zes 4-de-dimethylamino-anhydrotetracycline (4-ddma-atc, see Fig. 1C for chemical structures) [5,6], an analog lacking the dimethylamino grouping at position 4 and showing no antibiotic activity. In another effort TetR Keywords allostery; effector specificity; reverse TetR; tetracycline derivatives; Tet repressor Correspondence W. Hillen, Lehrstuhl fu ¨ r Mikrobiologie, Institut fu ¨ r Biologie, Friedrich-Alexander Universita ¨ t Erlangen-Nu ¨ rnberg, Staudtstraße 5, 91058 Erlangen, Germany Fax: +49 9131 ⁄ 85 28082 Tel: +49 9131 ⁄ 85 28081 E-mail: whillen@biologie.uni-erlangen.de (Received 12 May 2005, revised 12 July 2005, accepted 15 July 2005) doi:10.1111/j.1742-4658.2005.04868.x To learn about the correlation between allostery and ligand binding of the Tet repressor (TetR) we analyzed the effect of mutations in the DNA read- ing head–core interface on the effector specific TetR i2 variant. The same mutations in these subdomains can lead to completely different activities, e.g. the V99G exchange in the wild-type leads to corepression by 4-ddma- atc without altering DNA binding. However, in TetR i2 it leads to 4-ddma- atc dependent repression in combination with reduced DNA binding in the absence of effector. The thermodynamic analysis of effector binding revealed decreased affinities and positive cooperativity. Thus, mutations in this interface can influence DNA binding as well as effector binding, albeit both ligand binding sites are not in direct contact to these altered residues. This finding represents a novel communication mode of TetR. Thus, allo- stery may not only operate by the structural change proposed on the basis of the crystal structures. Abbreviations Atc, anhydrotetracycline; 4-ddma-atc, 4-de-dimethylamino-anhydrotetracycline; tc, tetracycline; TetR, tetracycline repressor; b-Gal, b-Galactosidase. FEBS Journal 272 (2005) 4487–4496 ª 2005 FEBS 4487 was converted to reverse TetR (revTetR) by single or multiple mutations affecting the allostery of the protein. RevTetR variants bind tetO only in the presence of tc, thus turning the inducer into a corepressor [7,8]. The underlying mutations occur in residues located in the interface of the core and DNA binding domains which do not move upon induction (Fig. 1A shows their loca- tion in the TetR structure). It has therefore been assumed that they may cause a repositioning of the DNA reading head with respect to the core of TetR, so that the same allosterical change as in wild-type could result in the opposite activity [8]. The combination of mutations resulting in 4-ddma- atc specificity with those yielding revTetR in the same polypeptide did not yield efficient mutants with com- bined activities [9]. Thus, these properties of TetR must be interrelated. We use here the TetR i2 variant exhibiting 4-ddma-atc specificity to combine it with degenerations of residues giving rise to revTetR mutants and screen for the combined phenotype. The results lead to insights about the allostery of TetR. Results Randomization of residues in helices a1, a4, or a6 in TetR i2 As the most efficient and mechanistically most interest- ing mutations leading to revTetR occurred in helices a1, a4, and a6 [8], we decided to combine randomiza- tions in these helices with TetR i2 (containing the muta- tions H64K, S135L and S138I). We screened the resulting candidates for TetR variants with 4-ddma-atc specific corepression in Escherichia coli WH207 ⁄ ktet50 [10]. The specificity is scored against atc, the most efficient effector of TetR known so far. The DNA fragments containing the randomized codons 14–25 (C-terminal part of helix a1 and the following loop) and 93–102 (helix a6 and the b-turns N-terminal and C-terminal of a6) as described previously were intro- duced into pWH1925-tetR i2 . The randomized codons 50–63 in helix a4 generated by PCR mutagenesis using a ‘doped’ oligonucleotide also encoding the H64K mutation were as well inserted in pWH1925-tetR i2 . The three mutant pools were screened for repression in the presence of 0.4 lm 4-ddma-atc and rescreened for induction with 0.4 lm atc and without inducer on MacConkey agar plates. We screened 17 600 colonies with mutations in helix a1, 29 840 with mutations in helix a4 and 3900 out of the helix a6 pool and obtained a total of 15 candidates with the desired properties. These were confirmed by in vivo repression and induction determined in broth cultures of E. coli B A C Fig. 1. Structural depiction of the allostery in TetR. (A) Crystal struc- ture of the TetR-[tc-Mg] + 2 complex. One monomer is shown as dark blue ribbon, the second monomer in light blue. Tc is drawn as a yellow stick model. The mutated parts of helices a1, a4anda6 (see arrows) forming the interface between the DNA-binding head and the protein core are highlighted in red in one subunit. (B) Over- lay of the induced (dark blue) and tetO bound (grey) partial struc- tures of one TetR subunit. Tetracycline is depicted as a yellow stick model. Leu17 (in helix a1), Val99 and Thr103 (in helix a6) and Leu52, Leu56 and His64 in helix a4 are indicated in the induced structure by red side chains and S135 in helix a8 by the green side chain. All amino acids are designated in the three letter code. The C-terminus of helix a4 is connected to the [tc-Mg] + binding pocket by interaction of His64 with tc. Leu52 and Leu56 form the hydro- phobic region contacting Val99. Thr103 is located in the C-terminal helical turn of a6 which is transformed into a type II b-turn upon induction [20]. (C) Chemical structures of anhydrotetracycline (atc) and 4-de-dimethylamino-anhydrotetracycline (4-ddma-atc). Altered TetR effector binding and allostery E M. Henßler et al. 4488 FEBS Journal 272 (2005) 4487–4496 ª 2005 FEBS WH207 ⁄ ktet50 containing a chromosomal tetA-lacZ fusion as summarized in Table 1. Three candidates with mutations in helix a1 exhibit 4-ddma-atc dependent repression, but show different specificities as scored by the lack of atc-mediated repres- sion. TetR i2 -E15A L17V shows repression to 12.5% with 4-ddma-atc and to 58% with atc, thus showing only fivefold improved repression with 4-ddma-atc over atc. A similar result is found for TetR i2 -L17V V20F, while the single exchange mutant TetR i2 -N18Y shows no distinction between atc and 4-ddma-atc. Randomization of helix a4 also led to three candi- dates. The amino-acid exchanges I59L L60M yield a TetR i2 mutant with reduced inducibility. TetR i2 -I59V H63Q exhibits a reverse phenotype but only little specificity with fourfold increased corepression with 4-ddma-atc over atc. The I59S exchange leads to repression of about 20% with 4-ddma-atc and to about 50% with atc. While the efficiency is not great, this mutation nevertheless proves that exchange of a single amino acid can be sufficient to reverse the induc- tion properties of TetR i2 . The pool with the randomizations in helix a6 yielded nine candidates. They contain single, double and triple mutations conferring 4-ddma-atc specific reverse phe- notypes, and all of them were confirmed in liquid cul- ture. The single amino-acid exchanges D95H, G96V or K98N yield revTetR i2 mutants. The G96V exchange leads to an about 20-fold increased repression with 4-ddma-atc compared to atc. It is noteworthy that this mutation does not change the property of wild-type TetR [8]. Therefore, we asked if single residue muta- tions leading to revTetR generally behave different when introduced into TetR i2 . As single exchange rev- TetR variants were found for the residues V99 and L17 [8] and V99G did reverse the TetR i2 activity [9] we analyzed the role of substitutions at these positions for both effectors in the wild-type and TetR i2 sequence backgrounds. TetR i2 variants with mutations at valine 99 We revisited the 19 possible exchanges at positions 99 in TetR [8] and introduced them into TetR i2 . The in vivo repression and induction for these 20 TetR vari- ants is shown in Fig. 2A,B. Fig. 2A shows the activit- ies of V99 exchanges in TetR for the effectors atc and 4-ddma-atc and without effector. 11 out of 19 substitu- tions at V99 do not lead to large changes of the phe- notype. However, 11 out of the 19 exchanges show slightly enhanced repression with 4-ddma-atc com- pared to without thus making it a corepressor for these variants. Interestingly, the mutations V99I, V99M, V99P, V99G and V99F turn 4-ddma-atc into a core- pressor while atc is still an inducer. This is remarkable because a residue at position 99 is not in contact with the effector [2]. Six out of these 11 exchanges exhibit a reverse phenotype with 4-ddma-atc in the TetR i2 back- ground. Thus, 4-ddma-atc still acts as corepressor but repression in the absence of effector is lost, despite of the fact that a residue at position 99 does not contact DNA, either [3]. Thirteen out of the 19 mutations in TetR i2 cause almost complete loss of repression with- out effector, indicating that this mutant is generally more sensitive for additional mutations. While substi- tutions of V99 with the charged amino acids R, K or E in wild-type TetR show pronounced reverse pheno- types, all charged amino acids at this position in TetR i2 lead to only very weak reverse phenotypes with 4-ddma-atc (Fig. 2B). Despite of their chemical similarity, serine and threonine cause contrary effects when replacing V99: TetR i2 -V99S shows fivefold better repression with 4-ddma-atc, while V99T enhances repression with atc sevenfold. Residues with aromatic side chains at posi- tion 99 exhibit regulatory effects according to their size: the V99W exchange shows a 1.5-fold, V99Y a threefold, and V99F a 4.5-fold preference for 4-ddma- atc over atc as corepressor. There is no relationship between phenotype and size at this position in the TetR background. The V99G exchange in TetR only marginally influences induction with atc but leads to Table 1. In vivo repression and induction of TetR i2 variants. The expression of 100% b-galactosidase corresponds to 6300 ± 1050 units. TetR variant b-Gal activity (%) Induction with 4-ddma-atc (0.4 l M) atc (0.4 lM) TetR i2 1.6 ± 0.1 57 ± 5 3.1 ± 0.5 TetR i2 -N18Y 60 ± 1.3 20 ± 1 14 ± 1.2 TetR i2 -L17V V20F 80 ± 6 11 ± 0.5 34 ± 0.7 TetR i2 -E15A L17V 85 ± 2 12.5 ± 0.4 58 ± 8 TetR i2 -I59S 87 ± 1 19 ± 0.6 46 ± 1 TetR i2 -I59V H63Q 84 ± 3 5 ± 4 20 ± 1 TetR i2 -I59L L60M 8 ± 0.8 17 ± 0.9 3 ± 0.1 TetR i2 -G96V 67 ± 0.9 2.4 ± 0.1 51 ± 4 TetR i2 -D95H 58 ± 3.4 2 ± 0.1 4 ± 0.4 TetR i2 -K98N 63 ± 1 6 ± 0.4 18 ± 2 TetR i2 -R94C D95C 79 ± 8 2 ± 0.1 51 ± 5 TetR i2 -D95A N81S 81 ± 2.7 2 ± 0.2 54 ± 4 TetR i2 -Y93C V99M 80 ± 6 4 ± 0.2 8 ± 0.6 TetR i2 -R94H K98I V99R 45 ± 3 1 ± 0.1 17 ± 0.5 TetR i2 -R94S K98I H100Q 58 ± 3 4 ± 0.1 53 ± 3 TetR i2 -R94H D95N G96R 60 ± 1.4 2 ± 0.1 4 ± 0.4 E M. Henßler et al. Altered TetR effector binding and allostery FEBS Journal 272 (2005) 4487–4496 ª 2005 FEBS 4489 improved repression with 4-ddma-atc. The same exchange leads to one of the best 4-ddma-atc specific reverse phenotypes in TetR i2 -V99G. Another very spe- cific reverse phenotype is found for TetR i2 -V99N with a 34-fold better repression in the presence of 4-ddma- atc compared to atc. Again, this exchange has only a slight effect on induction with atc when introduced in the TetR sequence background. Western blot analyses of the V99A, V99G, V99S, V99T and V99N exchanges in wild-type and TetR i2 revealed only small differences of the intracellular pro- tein amounts (Fig. 2C), which may correlate to the alterations seen in the repressed expression levels of the respective mutants. The observed specificity chan- ges are clearly not influenced by protein amounts. TetR i2 variants with mutations at leucine 17 The effects of all possible exchanges of leucine at position 17 in the wild-type or TetR i2 sequence back- grounds are shown in Fig. 3A,B. Most of the larger and charged amino acids at this position lead to repression deficient proteins in both sequence back- grounds. TetR is less proned for activity loss due to mutation at this position as 10 of 19 substitutions lead to altered phenotypes, while 17 out of 19 substi- tutions cause more or less severe loss of repression without effector in TetR i2 . We obtained eight variants showing improved repression with 4-ddma-atc com- pared to without, among them four exchanges where atc is an inducer and 4-ddma-atc a corepressor. Five out of these mutations cause 4-ddma-atc sensitive reverse phenotypes. The TetR i2 -L17M or -L17I exchanges show increased repression with atc, but not with 4-ddma-atc. Exchanges leading to 4-ddma-atc dependent repression in TetR i2 include the aromatic amino acids W and Y, the hydroxyl containing resi- dues S and T, and C. TetR i2 -L17A is the best rev- TetR i2 showing repression to 3% with 4-ddma-atc and only to 59% with atc. In contrast, TetR-L17A is noninducible with atc or 4-ddma-atc, while TetR- L17G shows the best atc dependent reverse pheno- type. TetR i2 -L17G, on the other hand, is inactive. Taken together, it is surprising that single residue exchanges cause quite different effects in these two sequence backgrounds. Moreover, contrary activities are caused by very small differences in side chains. The determination of the intracellular protein amounts (Fig. 3C) excludes contributions to these spe- cificity changes. Specificity determining residues in TetR i2 -V99G TetR i2 -V99G is one of the best revTetR i2 variants with 4-ddma-atc specific repression, almost completely lack- ing repression with atc [9]. As the V99G exchange in the wild-type sequence background displays no reverse phenotype and slightly increased repression in the presence of 4-ddma-atc, we decided to determine the contribution of each amino-acid exchange in TetR i2 - V99G to the combined activity. We constructed and B C A Fig. 2. Regulatory properties of TetR variants with mutations at position 99. b-Galactosidase activities of E. coli WH207 ⁄ ktet50 transformed with plasmids bearing either no tetR or different tetR variants are shown. They were determined in the presence of 0.4 l M atc (white columns), 0.4 lM 4-ddma-atc (grey) or in the absence of effector (black). The b-Gal activity in the absence of tetR was set to 100% and corresponds to 6300 ± 1050 units. (A) Regulatory properties for the wild-type TetR sequence back- ground with all mutations at Val99 and (B) for the TetR i2 sequence background with all possible mutations at Val99. (C) Steady-state levels of selected TetR variants with mutations of V99. The first lane (TetR) contains 50 ng of purified wild-type TetR and the other lanes 50 lg of a soluble protein extract from E. coli WH207 ⁄ ktet50. Altered TetR effector binding and allostery E M. Henßler et al. 4490 FEBS Journal 272 (2005) 4487–4496 ª 2005 FEBS analyzed all possible double and triple mutants that include the mutation V99G. The results obtained in our E. coli indicator strain are shown in Fig. 4. As TetR-H64K V99G S138I has nearly the same proper- ties than the protein with all four exchanges and TetR-V99G S135L is not reverse and atc specific, the S135L mutation does not contribute to the TetR i2 - V99G activity profile. The mutants TetR-V99G S138I, TetR-H64K V99G S135L, and TetR-H64K V99G clearly have reverse activities, albeit with different effi- ciencies, but do not show the effector specificity. TetR-V99G S135L S138I is inactive. These results demonstrate that both the H64K and S138I mutations are necessary in combination with V99G to produce revTetR variants with 4-ddma-atc specificity. It is also remakable that the S138I mutation does not always lead to 4-ddma-atc specificity as seen in TetR-V99G S138I, however, this mutant is only slightly reverse with both effectors. Thermodynamic analysis of atc and 4-ddma-atc binding to TetR i2 -L17A and TetR i2 -V99N For overexpression of the proteins the respective genes were introduced into pWH610 and the resulting plas- mids were transformed in E. coli RB791 [11]. TetR i2 - L17A was purified to homogeneity employing the protocol described for wild-type TetR [11]. The purifi- cation protocol for TetR i2 -V99N had to be modified as described in experimental procedures and resulted in a protein with 50% activity. To quantify atc and 4-ddma-atc binding to the TetR variants we titrated 0.005 lm, 0.01 lm or 0.1 lm atc or 4-ddma-atc with each TetR protein in fluorescence buffer containing 20 mm MgCl 2 . Under these conditions, binding of the Fig. 4. Contribution of mutations H64K, S135L and S138I to the activity of TetR i2 -V99G. b-Gal activities were measured in E. coli WH207 ⁄ ktet50 transformed with a plasmid bearing either no tetR or different tetR variants. b-Gal activities are shown in the presence of 0.4 l M atc (white columns) or 4-ddma-atc (grey) or without effec- tor (black). The combination of mutations is indicated at the bottom of the figure. b-Gal activity in the absence of tetR was set to 100% and corresponds to 6300 ± 1050 units. A B C Fig. 3. Regulatory properties of TetR variants with mutations at position 17. b-Gal activities of E. coli WH207 ⁄ ktet50 transformed with plasmids bearing either no tetR or different tetR variants are shown. They were determined in the presence of 0.4 l M atc (white columns), 0.4 l M 4-ddma-atc (grey) or without effector (black). b-Gal activity of 100% corresponds to 6300 ± 1050 units. (A) Results are shown for all possible residues at position 17 in the wild-type TetR sequence background and (B) for the TetR i2 sequence background. (C) Steady-state levels of selected TetR vari- ants with mutations of L17. The first lane (TetR) contains 50 ng of purified wild-type TetR and the other lanes 50 lg of a soluble pro- tein extract from E. coli WH207 ⁄ ktet50. E M. Henßler et al. Altered TetR effector binding and allostery FEBS Journal 272 (2005) 4487–4496 ª 2005 FEBS 4491 [atc-Mg] + or [4-ddma-atc-Mg] + complex to TetR can be directly monitored. Atc or 4-ddma-atc fluorescence were employed to observe complex formation. The fits of the data for all TetR variants indicated positive cooperativity. As described previously for tetracycline and atc [12,13], we observed only weak cooperativity for binding of 4-ddma-atc to the wild-type TetR and of atc to TetR i2 . In contrast, 4-ddma-atc binding to the revTetR i2 variants showed large cooperativity. Scatchard analysis confirmed positive cooperativity for TetR i2 -L17A and -V99N (Fig. 5). The resulting equi- librium binding constants are summarized in Table 2. As we could not saturate atc binding to TetR i2 -L17A and -V99N at 5 lm of atc, we assume constants below 2 · 10 5 m )1 . The equilibrium binding constant of atc to TetR i2 was determined previously by Mg 2+ -depend- ent titrations [5] and yielded a K A of 1.7 · 10 7 m )1 . The direct titration used here uncovers weak coopera- tivity (a sixfold higher affinity for binding to the sec- ond atc) but the constants are in the same range. Binding of the first and the second 4-ddma-atc to wild-type TetR are roughly 10-fold higher than deter- mined previously [5]. Both revTetR i2 variants exhibit higher affinities for 4-ddma-atc compared to atc but the affinities are lower than the respective ones to TetR i2 . Fig. 5. Binding curves and Scatchard plots of binding of TetR to [4-ddma-atc-Mg] + . Fluorescence titrations were carried out at 0.1 l M, 0.01 lM and 0.005 lM [Atc-Mg] + . m is the average number of 4-ddma-atc mole- cules bound to one TetR monomer. The circles show the data, and the lines indicate the fit according to the binding function. The nonlinear curve progression shows the pres- ence of positive cooperativity for the two 4-ddma-atc binding sites. (A) Fluorescence titration, Langmuir fit and Scatchard plot of the titration of 4-ddma-atc with TetR i2 -L17A. (B) Fluorescence titration, Langmuir fit and Scatchard plot of the titration of 4-ddma-atc with TetR i2 -V99N. Table 2. Atc and 4-ddma-atc binding constants of TetR variants. All constants have been determined by direct titration of 0.1 lM, 0.01 lM or 0.005 l M [atc-Mg] + or [4-ddma-atc-Mg] + with TetR and are compared to binding constants obtained previously by titration at limiting MgCl 2 concentrations [5]. TetR M corresponds to one monomer which can bind one [tc-Mg] + .TetR D represents the dimer that binds [tc-Mg] + which can then bind the second molecule. Equilibrium binding constants a ,(·10 7 M )1 ) [4-ddma-atc-Mg] + [atc-Mg] + TetR TetR i2 TetR i2 -V99N TetR i2 -L17A TetR TetR i2 TetR i2 -V99N TetR i2 -L17A TetR M + [tc-Mg] + ?TetR M [tc-Mg] + 0.3 b 132 b 119600 b 1.7 b < 0.02 c < 0.02 c TetR D + [tc-Mg] + ? TetR D [tc-Mg] + 1.5 – d 0.1 0.4 – d 1.2 < 0.02 c < 0.02 c TetR D [tc-Mg] + + [tc-Mg] + ? TetR D [tc-Mg] + 2 6.3 – d 22 31 – d 3.7 < 0.02 c < 0.02 c a The standard deviations typically range from 10% to 40%. b See [5]. c The affinity was too low to be quantified. d The affinity was too high for quantification by direct titrations. Altered TetR effector binding and allostery E M. Henßler et al. 4492 FEBS Journal 272 (2005) 4487–4496 ª 2005 FEBS TetO binding in the presence of both effectors was qualitatively analyzed by EMSA for the revTetR i2 vari- ants. We used an at least 55-fold excess of 4-ddma-atc or atc over TetR to ensure complete complex forma- tion. The results are shown in Fig. 6. Both proteins exhibit residual binding to tetO without effector and with atc. The strongest tetO binding is observed for TetR i2 -V99N with 4-ddma-atc as corepressor, while TetR i2 -L17A does not show a clear specificity for one of the effectors in this experiment. Discussion The two activities of TetR, DNA recognition and effector binding, albeit located in different parts of the protein, are connected by allostery and can be either mutually exclusive (wild-type) or additive (revTetR) [8]. Changes of DNA recognition specificities were accomplished by mutations in the DNA reading head [14,15] while alterations of effector specificity require mutations near the respective binding pocket [5,6], and changes of allostery can be accomplished by exchan- ging residues in the contacting area between the DNA reading head and the core of TetR (Fig. 1) [7,8]. In addition to these structurally obvious location-function relationships, the data presented here establish that alterations of residues in that interface built by helices a1, a4 and a6 (Fig. 1A) also affect both substrate recognition properties, although they are located far away from either binding site. The same mutations in the wild-type TetR or TetR i2 sequence backgrounds can lead to completely different results, e.g. V99G in the wild-type is induced by atc while the repression is increased in the presence of 4-ddma-atc, but DNA binding without effector is not altered. In the TetR i2 mutant, however, the same exchange leads to loss of DNA binding in the absence of effector and 4-ddma-atc dependent corepression. Moreover, analysis of the contributions of each exchange in TetR i2 -V99G to effector specificity revealed a similar role for V99G and S135L. It was shown for S135L previously that it confers relaxed effector specificity to TetR [5,6]. S135 belongs to the secondary shell of the effector binding pocket which does not directly contact tetracycline in the crystal structure [2] but is located next to tc contacting resi- dues. V99 is not in contact with S135 (Fig. 1B). Thus, the effect of V99G on effector binding must be trans- ferred to the effector binding pocket. V99A also leads to similar properties as it has no effect on the wild-type but shows corepression with atc and 4-ddma-atc in TetR i2 . V99S has unaltered properties in the wild-type TetR, but leads to loss of DNA binding and 4-ddma- atc dependent corepression in TetR i2 . V99T, on the other hand, shows partial DNA binding, corepression by atc and induction by 4-ddma-atc when introduced in TetR i2 . Thus, the addition of a single methyl group- ing has remarkable differential effects on the activities of these two TetR variants. A similar result is also observed for exchanges at position 17. L17C, F, I and V alter the allostery of TetR i2 while the wild-type activ- ity is not affected. Exchanges of L17 for A, T, W or Y influence allostery and effector recognition. The thermodynamic analysis of TetR i2 -L17A and -V99N revealed reduced binding constants for 4-ddma- atc and atc. Thus, the in vivo effects reflect large affin- ity changes reinforcing clear structural influence of these mutations on the effector binding pocket. More- over, the wild-type TetR has no apparent cooperativity for effector binding [12] yet we observed positive coop- erativity for the TetR mutants. Cooperativity has been described for IPTG binding to the Lac repressor-oper- ator DNA complex [16] and for tetracycline binding to the TetR-tetO complex [17] but binding to both free proteins is not cooperative [12,18]. Thus, it seems that not only the affinity but also the nature of effector recognition may be altered by the mutations studied here. The structural details underlying these long range effects are not clear at present. It has been proposed that the reduced effector binding affinities for revTetR-G96E Fig. 6. EMSA of tet operator with TetR i2 -V99N and TetR i2 -L17A. The EMSA was performed without effector and in the presence of 0.1 m M atc or 4-ddma-atc. Hybridized oligonucleotides (0.3 lM) car- rying tetO or a nonpalindromic sequence (usp. DNA) were incuba- ted for 15 min with 0.3 l M,0.9lM or 1.8 lM of the respective TetR variant, electrophoresed on an 8% polyacylamide gel and stained with ethidium bromide. The contents of the mixtures ana- lyzed are indicated below the respective slots. E M. Henßler et al. Altered TetR effector binding and allostery FEBS Journal 272 (2005) 4487–4496 ª 2005 FEBS 4493 L205S may be due to changes in the positioning of the effector contacting residue H100 [7]. This assumption could apply to the properties of TetR i2 -V99N, but the exchanged residue in TetR i2 -L17A is not in proximity to H100 or to any other residue contacting the effector [2]. In conclusion, we propose that revTetR mutations do not only lead to the previously proposed reposition- ing of the DNA reading heads with respect to the core domain [7,8] but also to altered effector binding via structural changes in the effector binding pocket. It was proposed on the basis of the TetR crystal struc- tures [2–4,19] that the structural changes upon effector binding are transmitted through the protein to the DNA-binding head via the interface region. As a result the distance between the two recognition helices is increased, leading to loss of DNA binding. Helices a1, a4 and a6 forming this interface are involved in signal transduction, but there is no structural hint for an influence on effector binding. As we clearly demon- strate specificity effects of residues in this region on effector binding [20] this must be an as of yet unrecog- nized contribution of effector binding site flexibility to TetR allostery. Experimental procedures Materials and general methods Atc was from Acros (Geel, Belgium) and 4-ddma-atc was synthesized by Susanne Lochner and Peter Gmeiner (Phar- mazeutische Chemie, FAU Erlangen-Nu ¨ rnberg). All other chemicals were from Merck (Darmstadt, Germany), Roth (Karlsruhe, Germany) or Sigma (Munich, Germany). Enzymes for DNA restriction and modification were from New England Biolabs (Frankfurt ⁄ Main, Germany), Roche (Mannheim, Germany), Stratagene (Heidelberg, Germany) or Pharmacia (Freiburg, Germany). Oligonucleotides were purchased from MWG Biotech (Ebersberg, Germany). Isolation and manipulation of DNA was performed as described previously [21]. Construction of the TetR mutant pools Escherichia coli DH5a was used for cloning. The DNA con- taining randomized codons for helices a1 and a6 from pWH1925 [8] were introduced in pWH1925-tetR i2 (enco- ding the mutations H64K S135L S138I) via XbaI ⁄ ApaI and ApaI ⁄ FspI, respectively. Randomization of codons 50–63 in helix a4 was performed by PCR mutagenesis with the primers a4deg_H64K (5¢-aataagcgggcccta ctggatgcgctggcggt ggagatcttggcgcgtcataaggattat-3¢; the underlined positions contain 89% wild-type and 11% of the three non-wt bases resulting in a predominant frequency of three to four muta- tions) and 1925gh (5¢-gcaaaccgcctctcgccgc-3¢) using tetR i2 as template. The resulting fragment was introduced in pWH1925 via ApaI ⁄ NcoI for constitutive expression. All other TetR variants were constructed using single restric- tion enzyme sites in pWH1925. E. coli screening system E. coli WH207 ⁄ ktet50 [10,22] was transformed with the mutant pools. It contains a chromosomal tetA-lacZ fusion under tetR control. The cells were plated on MacConkey Agar Base (Becton Dickinson, San Jose, CA, USA) con- taining 14 gÆL )1 lactose, 0.0042% (w ⁄ v) neutral red and 0.0014% (w ⁄ v) crystal violet. The colonies were screened for their ability to repress b-galactosidase in the presence of 0.4 lm 4-ddma-atc and to express b-galactosidase on plates containing 0.4 lm atc. b-Galactosidase assays Repression and induction with different tc analogs was determined in E. coli WH207 ⁄ ktet50. Cells were grown in LB supplemented with 0.4 lm of atc or 4-ddma-atc at 37 °C. b-Galactosidase activities were determined as des- cribed [23]. Three independent cultures were assayed for each mutant and measurements were repeated at least twice. The expression of b-galactosidase in the absence of TetR was set to 100% and corresponds to 6300 ± 1050 units. Protein purification E. coli RB791 was transformed with pWH610 containing the respective mutations. Purification of TetR i2 -L17A to homogeneity was performed as described [11]. For purifica- tion of TetR i2 -V99N the E. coli cells were resuspended in 50 mm Na-phosphate pH 6.8, 50 mm NaCl, 25% (w ⁄ v) sucrose, 1 mm EDTA and 10 mm dithiothreitol. Cell dis- ruption was achieved by sonification following addition of 5 mg lysozyme, 0.25 mg DNaseI, 2 mm MgCl 2 ,1%(v⁄ v) Triton X-100 and 1% (w ⁄ v) Na-deoxycholate and incuba- tion for 30 min at room temperature. The suspension was frozen in liquid N 2 after adjustment to 6 mm EDTA and thawed at 37 °C. The protein was purified from the super- natant by cation exchange and size exclusion chromatogra- phy as described previously [11]. The protein concentrations were determined by UV spectroscopy and their activity was assessed by saturating titration with 4-ddma-atc observing the change of fluores- cence. Fluorescence measurements The fluorescence measurements were performed in a Spex Fluorolog 3 with two double monochromators. To observe Altered TetR effector binding and allostery E M. Henßler et al. 4494 FEBS Journal 272 (2005) 4487–4496 ª 2005 FEBS 4-ddma-atc fluorescence we excited at 420 nm and observed emission at 540 nm. Excitation of atc fluorescence was per- formed at 454 nm and emission was observed at 545 nm. The equilibrium binding constants were obtained from fluorescence titrations under equilibrium conditions. The titrations were carried out in buffer containing 100 mm Tris ⁄ HCl, pH 8.0, 100 mm NaCl and 20 mm MgCl 2 . 0.1 lm, 0.01 lm or 0.005 lm atc or 4-ddma-atc were titra- ted with TetR concentrations from 2 · 10 )10 m to 1 · 10 )5 m. All measurements were repeated at least twice. The binding constants were calculated by fitting of a hyper- bolic binding function and including cooperative binding. Electrophoretic mobility shift assay The synthetic tetO1 containing fragment 5¢-gggtgtgcc gacactctatcattgatagagttattatac-3¢ and tetO2 containing the complementary sequence were used for EMSA. The TetR recognition site is depicted in bold style. For hybridization, equal molar amounts of each oligonucleotide were mixed in water, heated at 94 °C for 2 min and allowed to cool down to room temperature within 2 hÆ6 pmol of the DNA was incubated with atc, 4-ddma-atc or without effector and the indicated amounts of protein. An oligonucleotide contain- ing no palindromic sequence was used as a negative control (5¢-ctaataaaattaatcatttatggcataggcaacaag-3¢). All samples were incubated in complex buffer containing 0.02 m Tris ⁄ HCl (pH 8.0) and 5 mm MgCl 2 . Atc and 4-ddma-atc were added to a final concentration of 0.1 mm. After incu- bation for 15min at room temperature, the DNA was elec- trophoresed on an 8% polyacylamide gel at 100 V in TBM buffer containing 89 mm Tris, 89 mm boric acid and 1 m m MgCl 2 . The DNA was detected by ethidium bromide staining. Acknowledgements We thank Susanne Lochner and Prof. Peter Gmeiner for kindly providing 4-ddma-atc and Dr. Oliver Scholz for fruitful discussions. This work was supported by the Deutsche Fors- chungsgemeinschaft through SFB 473 and the Fonds der Chemischen Industrie. References 1 Berens C & Hillen W (2003) Gene regulation by tetra- cyclines. Constraints of resistance regulation in bacteria shape TetR for application in eukaryotes. Eur J Bio- chem 270, 3109–3121. 2 Kisker C, Hinrichs W, Tovar K, Hillen W & Saenger WB (1995) The complex formed between Tet repressor and tetracycline-Mg2+ reveals mechanism of antibiotic resistance. J Mol Biol 247, 260–280. 3 Orth P, Alings C, Schnappinger D, Saenger W & Hin- richs W (1998) Crystallization and preliminary X-ray analysis of the Tet-repressor ⁄ operator complex. Acta Crystallogr D54, 99–100. 4 Orth P, Schnappinger D, Sum PE, Ellestad GA, Hillen W, Saenger W & Hinrichs W (1999) Crystal structure of the Tet repressor in complex with a novel tetracycline, 9-(N,N-dimethylglycylamido)-6-demethyl- 6-deoxy-tetracycline. J Mol Biol 285, 455–461. 5 Henssler EM, Scholz O, Lochner S, Gmeiner P & Hillen W (2004) Structure-based design of tet repressor to opti- mize a new inducer specificity. Biochemistry 43, 9512– 9518. 6 Scholz O, Ko ¨ stner M, Reich M, Gastiger S & Hillen WB (2003) Teaching TetR to recognize a new inducer. J Mol Biol 329, 217–227. 7 Kamionka A, Bogdanska-Urbaniak J, Scholz O & Hillen W (2004) Two mutations in the tetracycline repressor change the inducer anhydrotetracycline to a corepressor. Nucleic Acids Res 32, 842–847. 8 Scholz O, Henssler EM, Bail J, Schubert P, Bogdanska- Urbaniak J, Sopp S, Reich M, Wisshak S, Ko ¨ stner M, Bertram R & Hillen W (2004) Activity reversal of Tet repressor caused by single amino acid exchanges. Mol Microbiol 53, 777–789. 9 Bertram R, Kraft C, Wisshak S, Mu ¨ ller J, Scholz O & Hillen W (2005) Phenotypes of combined Tet repressor mutants for effector and operator recogni- tion and allostery. J Mol Microbiol Biotechnol 8, 104–110. 10 Wissmann A, Wray LV Jr, Somaggio U, Baumeister R, Geissendo ¨ rfer M & Hillen W (1991) Selection for Tn10 Tet repressor binding to tet operator in Escherichia coli: isolation of temperature-sensitive mutants and combina- torial mutagenesis in the DNA binding motif. Genetics 128, 225–232. 11 Ettner N, Muller G, Berens C, Backes H, Schnappinger D, Schreppel T, Pfleiderer K & Hillen WB (1996) Fast large-scale purification of tetracycline repressor variants from overproducing Escherichia coli strains. J Chroma- togr A 742, 95–105. 12 Takahashi M, Altschmied L & Hillen W (1986) Kinetic and equilibrium characterization of the Tet repressor- tetracycline complex by fluorescence measurements. Evidence for divalent metal ion requirement and energy transfer. J Mol Biol 187, 341–348. 13 Scholz O, Kintrup M, Reich M & Hillen WB (2001) Mechanism of Tet repressor induction by tetracyclines: length compensates for sequence in the alpha8-alpha9 loop. J Mol Biol 310, 979–986. 14 Helbl V & Hillen W (1998) Stepwise selection of TetR variants recognizing tet operator 4C with high affinity and specificity. J Mol Biol 276, 313–318. 15 Helbl V, Tiebel B & Hillen W (1998) Stepwise selec- tion of TetR variants recognizing tet operator 6C E M. Henßler et al. Altered TetR effector binding and allostery FEBS Journal 272 (2005) 4487–4496 ª 2005 FEBS 4495 with high affinity and specificity. J Mol Biol 276, 319–324. 16 O’Gorman RB, Rosenberg JM, Kallai OB, Dickerson RE, Itakura K, Riggs AD & Matthews KS (1980) Equilibrium binding of inducer to lac repressor. operator DNA complex. J Biol Chem 255, 10107– 10114. 17 Lederer T, Takahashi M & Hillen W (1995) Thermo- dynamic analysis of tetracycline-mediated induction of Tet repressor by a quantitative methylation protection assay. Anal Biochem 232, 190–196. 18 Friedman BE, Olson JS & Matthews KS (1977) Inter- action of lac repressor with inducer, kinetic and equili- brium measurements. J Mol Biol 111, 27–39. 19 Hinrichs W, Kisker C, Duvel M, Mu ¨ ller A, Tovar K, Hillen W & Saenger W (1994) Structure of the Tet repressor-tetracycline complex and regulation of anti- biotic resistance. Science 264, 418–420. 20 Orth P, Cordes F, Schnappinger D, Hillen W, Saenger W & Hinrichs W (1998) Conformational changes of the Tet repressor induced by tetracycline trapping. J Mol Biol 279, 439–447. 21 Sambrook J (2001) Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, USA. 22 Smith LD & Bertrand KP (1988) Mutations in the Tn10 tet repressor that interfere with induction. Location of the tetracycline-binding domain. J Mol Biol 203, 949– 959. 23 Miller JH (1972) Experiments in molecular genetics. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, USA. Altered TetR effector binding and allostery E M. Henßler et al. 4496 FEBS Journal 272 (2005) 4487–4496 ª 2005 FEBS . residual binding to tetO without effector and with atc. The strongest tetO binding is observed for TetR i2 -V99N with 4-ddma-atc as corepressor, while TetR i2 -L17A. constants a ,(·10 7 M )1 ) [4-ddma-atc-Mg] + [atc-Mg] + TetR TetR i2 TetR i2 -V99N TetR i2 -L17A TetR TetR i2 TetR i2 -V99N TetR i2 -L17A TetR M + [tc-Mg] + ?TetR M [tc-Mg] + 0.3 b 132 b 119600 b 1.7 b <