TEXT

Top Abstract Text References

Active export of tetracycline (TC) is the most frequently found resistance mechanism among gram-negative bacteria (19). The expression of resistance underlies stringent regulation depending on the presence or absence of TC. Seven closely related resistance determinants have been isolated (10). We examine Tet repressors (TetR) encoded by transposon Tn10 (class B) or plasmid RA1 (class D), which have 63% identical amino acids. Genetic and biochemical results reported for TetR(B) are consistent with the crystal structures of TetR(D)-TC complexes (11, 12). TetR is an all--helical protein. Its structure is shown in Fig. 1. Helices 1 to 3 form the reading head and contain a DNA binding helix-turn-helix motif (HTH) (2, 27). It is connected via helix 4 to the protein core, which is formed by helices 5 to 10 and which contains the TC binding pocket and the dimer interface. Detailed models of operator recognition (9) and TC binding (6, 11, 12, 14, 16) are available. Based on the crystal structure, a model for induction upon TC binding has been presented. It involves a seesaw-like movement of helix 4 on 6, thereby adjusting the distance and angle between the two reading heads (11, 12). An analysis of induction-deficient TetR mutants (TetR^s) (4, 8, 16, 23, 28) led Müller et al. (16) to expand this model by proposing that the distance between the two reading heads is adjusted by the 4/6 movement, while the angle between them is reduced by a shear movement of 8 and 10 from both subunits. The two helices from both subunits form a four-helix bundle, the main dimerization motif of TetR (11, 12).

View larger version (76K): [in this window] [in a new window]   View larger version (64K): [in this window] [in a new window]   FIG. 1.   Position and phenotype of TetRs suppressor mutants. (A, C, E, G, and I) Crystal structure of the TetR(D)-TC complex. The two monomers are shown in white and yellow; TC is in blue. The position of the TetRs mutation is shown in red for both monomers, whereas the suppressors are only shown in the yellow monomer. Suppressors identified based on reduced operator binding affinity are pink; those identified based on mechanistic effects are green. Suppressors which could not be classified this way are orange. N and C indicate the N and C termini of the protein, respectively. The numbers give the positions of the respective amino acids and are marked with an additional prime in the white monomer. (B, D, F, H, and J) DNA binding (black bars) and inducibility by TC (hatched bars) are shown for TetR (wt) and the TetRs mutant with the indicated mutation in the left part of each panel and for the TetRs mutant with the indicated mutation in combination with the respective suppressor mutation in the right part of each panel. Standard deviations of repeated experiments are shown as vertical error bars.

Insight into movements within a protein can be gained from intragenic suppression, in which a functional deficiency caused by one mutation is compensated for by a second mutation. Examples of this approach include the identification of interacting regions within Escherichia coli RNA polymerase (26) or yeast mitochondrial cytochrome b (7). Poteete et al. (18) presented a very detailed structural analysis of second-site suppressors restoring the activity of a T4 lysozyme mutant. In these three cases, the intragenic suppressors are located near the sites of the respective primary mutations.

The large number of mutant residues in the dimer interface of induction-deficient TetR (8) suggested that this region is crucial for induction by TC (16). We reasoned that second-site suppressors of TetR^s mutants might help identify contacts between the two monomers that are important for induction. Eight TetR^s mutant residues were selected, and the positions of five of them in the TetR(D)-[Mg-TC]⁺ crystal structure are shown in Fig. 1A (LS117), 1C (RQ49), 1E (DN53), 1G (EG150), and 1I (164-166). These residues are located at or close to the dimer interface, being not further than 4 Å away from amino acid residues in the second TetR monomer. They were selected from the three structural classes defined by Müller et al. (16) for mutants with a TetR^s phenotype. Position L-117 directly contacts TC via a hydrophobic interaction (12) and, thus, LS117 belongs to the TC proximity class. RQ49, DY53, and EG150 belong to the domain connection class, since they are located at the interface between the reading head and the protein core. YC110, LF146, HR151, and 164-166 are less than 4 Å distant from residues in the other monomer of TetR and are grouped in the dimer interface class. It has been shown for all mutations selected that they exert their effects primarily by interfering with the conformational change which occurs upon TC binding and not by affecting TC binding itself (5, 16).

A stock of second-site mutants was prepared by fusing a pool of chemically mutagenized tetR fragments (8) to the respective tetR alleles encoding induction-deficient TetR mutants by using appropriate restriction sites (XbaI/BstXI for the N-terminal part (codons for residues 1 to 33), BstXI/MluI, for the central part (codons for residues 33 to 130), and MluI/SphI for the C-terminal fragment of TetR. The mutagenized fragments were chosen to encode residues in the other monomer of the TetR dimer which are located close to the respective primary mutations in the crystal structure. The tetR 164-166 allele was randomly mutagenized by error prone PCR (30).

The selection system consists of the tet50 prophage (23) containing a single-copy tetP_AO-lacZ transcriptional fusion. TetR is constitutively expressed in trans by plasmid pWH520 (3) or pWH1919 (Table 1) (8) and represses -galactosidase (-Gal) expression by binding to tetO in the absence of TC. Thus, growth on minimal medium with lactose as the sole carbon source is not possible without TC except for strains expressing TetR mutants deficient in operator binding. TetR^s mutants are not able to grow on this medium even in the presence of TC. We selected for candidates from the second-site mutant stocks with the ability to grow on lactose minimal medium containing 0.4 or 0.5 µg of TC per ml or screened for dark-colored colonies on MacConkey lactose agar in the presence of 0.5 µg of TC per ml. Nonfunctional repressor mutants were identified by growth on selective media without TC and eliminated from further analysis.

Apart from allosteric suppression, TetR^s suppression may also result from reduced intracellular protein amounts, reduced operator affinity, or increased TC binding. The intracellular amounts of several TetR^s suppressors were checked by Western blotting (data not shown). No mutants with reduced intracellular TetR levels were found. Reduced operator binding affinity was tested by introducing mutations conferring reduced DNA binding activity into tetR 164-166 (5). The -Gal expression levels in the presence of TC of strains carrying the resulting double mutants were plotted in Fig. 2 versus their -Gal expression levels in the absence of TC. The semilogarithmic graph is nearly linear between the lower resolution limit of expression without TC and the upper limit corresponding to wild-type (wt) inducibility. This demonstrates that decreased tetO binding leads to increased apparent inducibility. Suppressor mutants whose expression levels lie on the reference line most likely act by reducing tetO binding. Induction of -Gal expression to levels above the line would indicate suppression that is not solely due to decreased tetO binding. Induction to a level below the line would suggest a stronger TetR^s phenotype. Suppressors with reduced or ambiguous tetO binding will not be discussed in detail. Increased TC binding needs to be determined in vitro and is not experimentally addressed in this study. Judging by the crystal structures, we do not consider increased TC binding of the suppressor mutant residues likely, since they do not contact TC.

View larger version (18K): [in this window] [in a new window]   FIG. 2.   Semilogarithmic plot of -Gal activities in the presence of inducer versus -Gal activities in the absence of TC. -Gal expression was determined as described by Miller (15) with cells grown at 37°C in Luria-Bertani medium supplemented with the appropriate antibiotics. For measuring inducibility, TC was added at a concentration of 0.2 µg/ml. Three to four independent cultures were assayed for each strain, and measurements were carried out at least twice. The open circles correspond to the TetRs mutant TetR 164-166 (0.8% ± 0.0% -Gal activity under repressed conditions and 0.9% ± 0.0% activity under induced conditions) and the TetR 164-166 double mutants with substitutions within the HTH. The mutations and the corresponding activities under repressed and induced conditions, respectively, are as follows: TG40, 1.6% ± 0.1% and 31% ± 1.6%; TQ27, 1.9% ± 0.1% and 42% ± 3.0%; VS36, 2.0% ± 0.1% and 31% ± 2.2%; LT41, 9.6% ± 0.5% and 68% ± 2.7%; QG38, 10% ± 0.9% and 74% ± 3.0%; VW36, 15% ± 0.7% and 85% ± 2.5%; WG43, 17% ± 2.1% and 87% ± 4.4%; KT48, 20% ± 1.1% and 82% ± 2.9%. The filled diamonds represent the suppressors isolated or constructed for the mutant TetR 164-166. The reference line was fitted by linear regression. Mutants above or on this line are labelled.

TC proximity class. TetR LS117 was suppressed by mutations SI135 and LS146 (Fig. 1A). (The nomenclature for suppressors is as follows. The primary mutation is given first and is followed by the suppressor mutation(s), with amino acids abbreviated by standard one-letter coding. Individual mutation designations are as in the following example: LS117 denotes a leucine-to-serine exchange at position 117 in the TetR primary structure.) In Fig. 1B, TetR LS117;LS146 has reduced repression efficiency compared to the wt, while TetR LS117;SI135 shows almost wt repression, indicating allosteric suppression. Both primary and suppressing mutations approach helix 9 from the other monomer. This helix is important for induction. Its proposed role is to act as a bolt locking the four-helix bundle in its conformation in the induced structure (16).

Domain connection class. Eight mutations were isolated as intragenic suppressors for TetR RQ49 (Fig. 1C and D). Two of them (FI186 and LF191) confer significantly reduced tetO affinity. The other six suppressor mutations cause no (ED147 and FL188) or only a less-than-twofold (SI135, LI142, LS146, and FS188) reduction of operator affinity. Five mutants exhibit a high level of inducibility (70% of wt -Gal activity), indicating allosteric suppression of the TetR^s phenotype, while TetR RQ49;ED147 is not as inducible.

Domain interface class. Two second-site suppressor mutations were found for TetR YC110 (ED147 and FI186), one for TetR LF146 (TA40), and five for TetR HR151 (VA36, TA40, RW49, WR75, and RT87). None of them conferred a high level of inducibility without significant increase of -Gal expression in the absence of TC, indicating that these suppressor mutations act by reducing the tetO affinity (data not shown).

Conclusions. Except for one (TetR EG150;DY95), all allosteric suppressor mutations affect residues which are not in direct contact with the primary mutation in the crystal structure of the TetR(D)-[Mg-TC]⁺ complex (11, 12). This hints that global structural changes involving both mutations occur between the operator binding and induced forms of TetR. Interference of a TetR^s mutant with these conformational changes could then be suppressed by mutations at distant sites. The numerous suppressor residues at the dimer interface lend strong support to the proposal by Müller et al. (16) that a conformational change in the four-helix bundle contributes to the induction of TetR by TC.