INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

The TOL plasmid pWW0 of Pseudomonas putida encodes the enzymes for the metabolism of toluene and related hydrocarbons (36). The main regulator in the transcriptional control of the catabolic pathways is the XylR protein (reviewed in reference 31). The xylR gene is expressed constitutively from two tandem promoters (16, 20), but the regulator only induces expression from the target promoters in response to a wide variety of effectors (1). Activated XylR protein stimulates transcription from the Pu promoter for the upper operon, which encodes the enzymes for the oxidation of the aromatic hydrocarbons to the corresponding benzoates, and from the xylS Ps1 promoter, the regulator of the meta-cleavage pathway responsible for the further metabolism of aromatic carboxylic acid derivatives (12, 16, 21, 32). The Pu and Ps1 promoters belong to the ⁵⁴ class.

The XylR protein belongs to the NtrC family of transcriptional regulators (17, 25). These regulators exhibit four domains, three of which are highly conserved among members of the family (Fig. 1). The nonhomologous N-terminal domains have been implicated in signal reception, either via a sensory protein, as in the NtrB-NtrC pair (18, 24), or via interaction with a chemical signal, as in FhlR, DmpR, and XylR (31, 34). This has been genetically confirmed for the last two regulators, in which mutations in the N-terminal region altered effector recognition (6, 26). The central domains of these regulators contain an ATP binding domain, and ATPase activity is a sine qua non for transcriptional activity (34). At the C-terminal domain these regulators exhibit a typical helix-turn-helix DNA binding domain.

View larger version (12K): [in this window] [in a new window]   FIG. 1.   Domains of the XylR regulator and locations of point mutations. The organization of the XylR domains is according to Inouye et al. (17). The mutations located at the N-terminal end of this regulator are shown.

A surprising feature of the XylR regulator is that it recognizes a wide variety of effectors, namely, an aromatic benzene ring with certain mono-, di-, and trisubstituents. Toluene and benzyl alcohol are examples of monosubstituted benzene rings recognized by XylR. o-, m-, and p-xylene, o- and p-nitrotoluene, and m-methylbenzyl alcohol are examples of disubstituted aromatic rings recognized by XylR. 2,3-Dichlorotoluene is an example of a trisubstituted ring recognized by this regulator (1, 7).

Removal of the first 210 amino acids of the N-terminal domain of XylR results in a mutant regulator that stimulates transcription in the absence of effectors (10). This suggested that the N-terminal domain acts as an intramolecular inhibitor of the transcriptional activity of XylR (10). Because XylR binds persistently to target DNA sequences (2, 8) regardless of the presence of an effector, Pérez-Martín and de Lorenzo suggested that the N-terminal domain prevents either ATP binding or ATPase activity of this regulator (27-30).

We report that in vivo XylR-dependent transcription stimulation from the Pu promoter shows an effector dose-dependent response that gives rise to different induction kinetics with different effectors. The nature of the substituent on the aromatic ring of the effector is critical to bring XylR to a productive or an unproductive transcriptional state. These observations are being substantiated by mutant regulators with altered effector specificity (XylRE172K) or blocked in transcriptionally active (XylRD135N and XylRD135Q) or inactive (XylRD135E) forms.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Bacterial strains, plasmids, and growth conditions. The bacterial strain used in this work was Escherichia coli ET8000 [rbs lacZ::IS1 gyrA hutC(Con)] (22). Plasmids used were pTS174 (Cm^r xylR, p15 replicon) (15); pAD1 (Cm^r, carrying an xylR mutant allele encoding XylRE172K, p15 replicon) (6); pAD6, pAD49, pRS1, and pRS2, which were similar to pAD1 except that the mutant xylR allele encoded, respectively, XylRP85S, XylRD135N, XylRD135E, and XylRD135Q (reference 7 and this study); pRD579 (Ap^r Pu::lacZ, pR1 replicon) (9); and pERD401 (Ap^r Pu::lacZ, pBR replicon) (2). Plasmid pTS174HincII was constructed in this study by deleting a HincII restriction fragment of plasmid pTS174. This removed a BglI site from pTS174 and resulted in a plasmid containing unique PstI and BglI sites within the xylR gene, facilitating the cloning of mutated PstI-BglI fragments of xylR. Bacteria were grown at 30°C in Luria-Bertani broth supplemented, when required, with 100 µg of ampicillin per ml and 30 µg of chloramphenicol per ml.

-Galactosidase assays. E. coli ET8000(pRD579 Pu::'lacZ) cells also bearing a plasmid encoding wild-type XylR or a mutant XylR protein were grown for 5 h with vigorous shaking in the absence or presence of effectors. -Galactosidase activity was measured in permeabilized cells as described before (6), and activity was expressed in Miller units (23).

DNA techniques. DNA manipulations were done according to standard procedures (33). The 5' mRNA start of the upper operon transcript was determined by primer extension analysis (20). The oligonucleotide 5'-GATGTGCTGCAAGGCGATTAAGTTG-3' was 5' end labeled with [-³²P]ATP and annealed to 20 µg of total RNA prepared from E. coli ET8000(pERD401) also bearing pTS174.

Site-directed xylR mutagenesis. XylR mutants were constructed by a PCR technique. The plasmid pTS174 was used as a template for the PCRs with a pair of internal oligonucleotides that carried the desired mutation and two external oligonucleotides. The internal oligonucleotide 5'-CCTTTGAAGTGGAGATCTGCC-3' and its complementary oligonucleotide were used to replace Asp135 in XylR with Glu (the underlined nucleotide indicates the point mutation introduced to generate the desired substitution at the protein level). The oligonucleotide 5'-CCTTTGAAGTGCAGATCTGCC-3' and its complementary oligonucleotide were used to replace Asp135 with Gln. As external oligonucleotides we used 5'-AGGCCTTGACGTTGCAAGGT-3' and 5'-CCCACCCCAGTCTCACCC-3'. Both PCR products were purified and mixed, and the final PCR was carried out with the external oligonucleotides. The PCR products were digested with BglI and PstI and used to replace the wild-type sequence in plasmid pTS174HincII to yield pRS1 (XylRD135E) and pRS2 (XylRD135Q). The single mutation in the mutant xylR alleles was confirmed by dideoxy sequencing.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Dose-dependent stimulation of transcription by XylR and XylRE172K. Activation of transcription in systems triggered by a chemical stimulus occurs above a certain threshold concentration of the effector molecule (4). The dose dependence of the activation of transcription from Pu by XylR was assayed in vivo by measuring the -galactosidase activity resulting from the expression of lacZ in a strain bearing a Pu::'lacZ fusion. XylR-dependent activation of transcription from Pu with toluene followed a hyperbolic curve, with a maximum transcriptional activity of 2,550 ± 100 Miller units and an apparent K (K_app) of 0.95 ± 0.08 mM (Fig. 2A). Similar results were obtained when m-aminotoluene was used as an effector for XylR (K_app, 0.85 ± 0.15 mM; V_max, 4,400 ± 450 Miller units), whereas m-nitrotoluene was not an effector for XylR (Table 1). However, when activation was assayed with o-nitrotoluene, transcription activation from Pu was observed. The o-nitrotoluene dose-dependent curve was sigmoidal in the range of 20 µM to 1 mM (Fig. 2B). The curve profiles yielded a K_app of 0.55 ± 0.04 mM and a Hill coefficient of 2.65 ± 0.45. This suggests the existence of cooperative effects in the transcriptional process with o-nitrotoluene as an effector. A remarkable effect was the inhibition of transcription of wild-type XylR by concentrations of o-nitrotoluene above 1.5 mM (Fig. 2B). This inhibition suggests that above certain concentrations, o-nitrotoluene may have more than one binding mode in the binding pocket of XylR, with nonproductive binding resembling the binding of m-nitrotoluene, which acts as an inhibitor of transcription (7). XylRE172K is a mutant with altered effector specificity that was isolated as being able to recognize m-nitrotoluene as an effector (7). The E172K mutation in XylR did not alter the profile of activation of Pu in response to toluene; it produced a hyperbolic curve similar to that of the wild-type regulator (Fig. 2A). However, the mutation increased the K_app to 1.42 ± 0.25 mM and decreased the maximum activation to 1,150 ± 75 Miller units. Activation of transcription from Pu by XylRE172K in the presence of o-nitrotoluene followed a sigmoidal curve with an apparent K_app of 1.12 ± 0.12 mM and a Hill coefficient of 3.62 ± 0.25 (Fig. 2B). This suggests that the E172K mutation in the signal recognition domain of XylR resulted in an increase in cooperativity in XylR-dependent stimulation of transcription from Pu with o-nitrotoluene.

View larger version (23K): [in this window] [in a new window]   FIG. 2.   Dose dependence of the activation of transcription from Pu by XylR and XylRE172K in response to toluene and nitrotoluenes. E. coli ET8000(pRD579 Pu::'lacZ) also bearing pTS174 (encoding wild-type XylR) or pAD1 (encoding XylRE172K) was grown for 5 h with vigorous shaking in the absence or presence of increasing concentrations of toluene or nitrotoluenes. The -galactosidase activity was assayed in permeabilized cells as described in Materials and Methods. The data are the averages of at least five independent determinations. The results were fitted to hyperbolic or sigmoidal equations, and the values of Kapp and Hill's coefficient were deduced from these equations. (A) Dose dependence of the activation of transcription from Pu by toluene. Open symbols, wild-type XylR; closed symbols, XylRE172K. (B) Dose dependence of the activation of transcription from Pu by nitrotoluenes. Triangles, wild-type XylR; squares, XylRE172K; open symbols, o-nitrotoluene; closed symbols, m-nitrotoluene.

Further evidence for m-nitrotoluene being an inhibitor of transcription stimulation by wild-type XylR, XylRP85S, and XylRE172K. As mentioned above, m-nitrotoluene is not an effector of XylR (6, 7) (Table 1). Furthermore, we noticed that basal transcription levels from Pu in the presence of this nitroarene were reduced (Table 1). The inhibitory effect of m-nitrotoluene was also confirmed by the fact that this nitroarene decreased o-nitrotoluene-dependent activation of transcription from Pu mediated by the wild-type XylR regulator: for a given o-nitrotoluene concentration, the higher the concentration of m-nitrotoluene, the lower the induction from Pu (Fig. 3A).

View larger version (26K): [in this window] [in a new window]   FIG. 3.   Inhibition of XylR mediated by m-nitrotoluene. (A) Inhibition of expression from Pu mediated by XylR with o-nitrotoluene and m-nitrotoluene. E. coli ET8000(pRD579, pTS174) was grown for 5 h with vigorous shaking in the presence of 0.5 mM o-nitrotoluene (o-NT) and increasing concentrations of m-nitrotoluene (m-NT) (0 to 2.5 mM). After incubation, -galactosidase activity was assayed in permeabilized cells as described in Materials and Methods. The data are the averages and standard deviations of at least five independent determinations. (B) mRNA synthesis from Pu mediated by XylRP85S in the absence of an aromatic (lane 1) and in the presence of m-methylbenzyl alcohol (lane 2) or m-nitrotoluene (lane 3). E. coli ET8000(pERD401, pAD6) was grown overnight on Luria-Bertani medium supplemented with appropriate antibiotics. Bacterial cells were diluted 1/100 in the same medium, and after 1 h of cell growth, the samples were supplemented with m-methylbenzyl alcohol or m-nitrotoluene to a final concentration of 1 mM. After 30 min, samples were withdrawn for mRNA analysis. The extended products (arrow) were 134 nucleotides long.

Importance of the charge and size of the side chain of residue 135 in the constitutive activation of transcription by XylR mutants. XylRD135N is a mutant regulator that is able to stimulate transcription from Pu and Ps1 in the absence of effectors. To further study the role of position 135 in the stimulation of transcription from Pu, the Asp residue at position 135 in the wild-type protein was replaced with Glu or Gln to test the importance of the charge and size of the side chain at this position.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

The XylR protein induces expression of the TOL plasmid upper pathway by stimulating transcription from the Pu promoter (11, 16, 32) and is involved through a cascade regulatory system in activation of the meta pathway for the metabolism of alkyl benzoates (31). XylR binds persistently to its target sequences in the Pu and Ps1 promoters (2, 8); conversion from the inactive to the active form of XylR and induction of transcription are regulated by the presence of aromatic compounds in the medium (1). In vitro studies with the highly homologous DmpR regulator have shown that effector binding to DmpR activates an ATPase activity which lies in the central domain (36). Large deletions in the amino-terminal domain of XylR or DmpR resulted in high rates of ATP hydrolysis by the ATPase activity of these regulators, and consequently they stimulated transcription from the corresponding promoters in the absence of effectors (10, 27-29, 35). This mechanism has also been described for other members of the family that are activated by covalent modification (3, 5, 13, 19).

On the basis of these findings, a model for the activation of XylR and the stimulation of transcription by this regulator is emerging: the N-terminal domain of XylR, in the absence of effectors, prevents either ATP binding or the ATPase activity of this regulator, and hence the regulator is maintained in an inactive form from a transcriptional point of view. Upon binding to effector molecules, the N-terminal domain undergoes conformational changes that allow binding of ATP; this in turn induces further conformational changes that lead to XylR multimerization and increase the ATPase activity of the regulator, which in turn stimulates transcription (27-29).

The first steps in this modelinteraction between effectors and the regulator and how the conformational change induced upon effector binding is transmitted to the other domains of XylRare poorly understood. An interesting aspect of XylR activation by effectors is the dose dependence and the type of kinetic induction. Transcription activation from Pu with toluene-activated and m-aminotoluene-activated XylR followed hyperbolic curves, whereas the response to increasing concentrations of o-nitrotoluene followed a sigmoidal curve. Therefore, the binding of different effector molecules may produce distinct conformational changes in the N-terminal domain of XylR that modulate transcription activation. This may result in cooperative interactions of XylR monomers to achieve the active multimeric XylR state (29); alternatively, it may reflect cooperativity for ATP hydrolysis or cooperativity in the interaction with the transcriptional machinery located in the downstream promoter. XylR is not able to activate transcription with m-nitrotoluene (7); moreover, this compound inhibits basal activity of the wild-type protein and significantly decreases the high basal mRNA levels mediated by the constitutive mutant XylRP85S. These facts suggest that m-nitrotoluene interacts with XylR but that the conformational changes in the protein upon binding to the aromatic are counterproductive.

The Glu172Lys change in the N-terminal domain of XylR reversed this response: mutant XylRE172K induced transcription in the presence of m-nitrotoluene but showed low activity with m-aminotoluene. These results suggest that residue 172 of XylR is probably involved in interactions with effectors, although a more detailed study of substituents on the ring is still needed to define the specificity for interactions with substituents on the aromatic ring. In any case, our results support a model in which interaction of effector molecules with the N-terminal domain of XylR can lead to two different conformational states. One of them is productive; the inhibition that the N-terminal domain exerts over the rest of the regulatory protein is abolished, allowing stimulation of transcription. In the other, an unproductive state, the N-terminal domain is probably locked into a conformation that keeps the N-terminal module in an inhibitory state.

The Asp135 residue in XylR may also be involved in effector recognition and/or conformational changes upon effector binding, as suggested by the finding that a previously isolated XylRD135N mutant induced a high level of transcription in the absence of effectors. Using site-directed mutagenesis, we generated mutants XylRD135E and XylRD135Q. The conservative mutation Asp135Glu produced a mutant regulator that was unable to recognize effectors. This may be because the regulator was locked into a conformation that prevented either interactions with effectors or transmission of the conformational change produced upon binding of the effector to the other domains of the protein. Substitution of Gln for Asp135 yielded a regulator with high basal activity in the absence of effectors. This activity was still enhanced in the presence of effector molecules that activated the wild-type regulator. These results emphasize the role that the removal of the negative charge of Asp135 plays in locking the XylR protein into a conformation that activates transcription.

In summary, the role of the N-terminal domain in the modulation of XylR activity is demonstrated by (i) the existence of activator and inhibitor molecules of XylR, as well as the change in the character of these molecules caused by point mutations at position 172 of the regulator, and (ii) the point mutations at position 135 that lock the regulator into an active conformation that mimics that produced by effector binding (XylRD135N and XylRD135Q) or by a mutation (XylRD135E) that blocks the protein in a conformation unable to stimulate transcription.