SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results and discussion
  5. Experimental procedures
  6. Acknowledgements
  7. References

In the presence of toluene and other structural analogues, the enhancer binding protein XylR activates the σ54 promoter Pu of the TOL (toluene degradation) plasmid pWW0 of Pseudomonas putida. Introduction of amino acid changes Val-219Asp and Ala-220Pro, which enter a proline kink at the interdomain region (B linker) between the A (signal reception) module and the central portion of XylR, originated a protein with unforeseen properties. These included a minor ability to activate Pu in the absence of aromatic effectors, a much higher responsiveness to m-xylene and a significant response to a large collection of aromatic inducers. Such changes could not be attributed to variations in XylR expression levels or to the fortuitous creation of a novel promoter, but to a genuine change in the properties of the activator. Structural predictions suggested that the mutation entirely disrupted an otherwise probable coiled-coil structure. A second directed mutant within the same region consisting of a major replacement of amino acids A220–N221 by the peptide HHHR produced an even more exacerbated phenotype. These data support a model in which the linker B region influences the effector profile by modifying at a distance the operative shape of the effector pocket and fixing the protein in an intermediate step of the activation process.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results and discussion
  5. Experimental procedures
  6. Acknowledgements
  7. References

The toluene-responsive XylR protein is the regulator of the promoter Pu of the TOL plasmid pWW0 of Pseudomonas putida mt-2 (Ramos et al., 1997). Pu drives the expression of the first (upper) TOL for the bioconversion of toluene, m- and p-xylene, m-ethyltoluene and 1,2,4-trimethylbenzene into the corresponding carboxylic acids and is thus the master regulatory asset of the whole catabolic system (Abril et al., 1989; Ramos et al., 1997). XylR belongs to the class of regulators generically known as the NtrC family of prokaryotic enhancer binding proteins, which activate transcription at a distance of promoters dependent on the alternative sigma factor σ54 (Dixon, 1986; Inouye et al., 1988), to which type Pu belongs. The regulators belonging to this family are composed of four domains (Inouye et al., 1988; North et al., 1993; Morett and Segovia, 1993; Fig. 1). The amino-terminal module o signal response domain used to be the most variable among the regulators, as it is the one that directly or indirectly receives the signal leading to the activation of the protein into a transcriptionally competent form (Shingler, 1996). XylR belongs to the subclass of enhancer binding proteins whose A domain interacts directly with the chemical inducer (i.e. toluene, xylenes), followed by the relief of an intramolecular repression caused by the N-terminal module on the central, activating domain of the protein (Pérez-Martín and de Lorenzo, 1995). Such a central domain is the most conserved region among the protein family members, as it is involved in the binding and hydrolysis of ATP that is the basis of the activation of σ54 promoters. To this end, regulators of this type bind to the upstream region (UAS) of cognate promoters through their C-terminal, DNA-binding domain (Inouye et al., 1988; Morett and Segovia, 1993).

image

Figure 1. Functional domains of XylR and its derivatives. The organization of the wild-type XylR is shown with the boundaries between the functional domains and the localization of relevant functions within the protein sequence. A (signal reception and inducer binding); C (binding and hydrolysis of ATP and contacts with the σ54-RNAP); and D (including a helix–turn–helix motif, HTH, for binding to the UAS of the Pu promoter DNA). The linker B region is enlarged at the top to show the presence of at least three repeats of the consensus motif for coiled-coil structures (Lupas, 1996), in which the key amino acid signatures are marked in bold. The changes within the B linker effected in mutants XylR16-1 and XylR3H are indicated. The leading residues of XylRΔA protein used in this work are specified with reference to the position formerly occupied by the nearest unchanged residue (amino acid 226).

Download figure to PowerPoint

One intriguing feature of XylR and other regulators from the same family (Fernández et al., 1995; Pérez-Martín and de Lorenzo, 1995; Shingler, 1996) is how the binding of the aromatic effector gives rise to a protein that activates Pu to the same extent as the protein entirely deleted of the A domain. In other words, how toluene binding results in the alleviation of intramolecular contacts between the A and C domains? In this context, the role of the linker, connecting peptide between the two domains (linker B) has been generally considered as merely that of a hinge to assist the contacts between the other two modules, with no direct contribution to the activation process. This notion was largely based on the early proposal (Wootton and Drummond, 1989) that the abundance of glutamines (Q) through the homologous B regions of a number of enhancer binding proteins could give rise to a flexible and non-structured coiled polypeptide (the so-called Q linker), with little chance of it being more than an interdomain hinge (Bantscheff et al., 1999). The lack of a detectable phenotype in various insertion mutants within such Q linkers of NtrC and NifA (Wootton and Drummond, 1989) substantiated this notion. On the contrary, Fernández et al. (1995) reported that some directed mutations at the B region of XylR originated a partially constitutive protein when expressed in Escherichia coli. The phenotypes of such mutants (which are one of the pillars of the current model of the functioning of XylR; Pérez-Martín and de Lorenzo, 1996a) were interpreted by that time as the result of the offsetting of the interactions between the A and C domains of the protein, which kept the regulator in an inactive state (Pérez-Martín and de Lorenzo, 1996b).

In this work, we have re-examined the contribution of the B region of XylR to the mechanism of activation of the protein by aromatic effectors. To this end, we have monitored in the native host of the system (Pseudomonas putida) the phenotypes endowed by two mutations entered within the linker that disrupt the predicted structure of such a connecting polypeptide. Our results show that the interdomain B sequence is in fact actively engaged in fixing the protein in a form that is competent for transcription activation and, indirectly, in setting the structural tolerance of the protein for aromatic effectors.

Results and discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results and discussion
  5. Experimental procedures
  6. Acknowledgements
  7. References

Assaying the effects of the 16-1 mutation in the Q linkerof XylR in P. putida

As shown in Fig. 1, the directed mutation of XylR that we have named 16-1 enters changes in amino acid positions Val-219 and Ala-220, which become converted in Asp and Pro respectively. These changes set a proline kink in the middle of the B region of the activator that is predicted to break any organized secondary structure at that site (Fernández et al., 1995; Lupas, 1996). When the regulator bearing such changes was expressed in E. coli (Fernández et al., 1995), the multicopy Pu promoter became semi-constitutive but still partially responsive to 3-methyl benzylalcohol. Such an intermediate phenotype is difficult to interpret in view of later observations on XylR and the related regulator DmpR. The A and C domains of these proteins seem to interact strongly if produced as separate polypeptides (Pérez-Martín and de Lorenzo, 1995; Ng et al., 1996; O'Neill et al., 1998), thus disputing the notion that the linker is required to settle the interactions between the domains. In addition, the A domain of XylR or DmpR can repress the ΔA-deleted protein in trans, but such associations (that miss the B connector) are unable to respond to their effectors any more (Pérez-Martín and de Lorenzo, 1995; Ng et al., 1996). These facts suggested a more active role for the B region in the process of XylR activation by aromatic inducers and prompted us to re-examine in more detail the phenotypes endowed by the 16-1 mutation.

As we have noticed repeatedly that the performance of XylR changes significantly depending on whether the bacterial host is E. coli or P. putida, we set out to reconstruct the 16-1 mutation in an assay system that could faithfully reflect the regulation of the Pu promoter in its native gene dose and stoichiometry. To this end, the reference strain was P. putida MAD1 (Fernández et al., 1995) because its chromosome is inserted with a mini-Tn5 transposon that includes the wild-type xylR gene expressed through its natural promoter (Pr) assembled next to a transcriptional Pu–lacZ fusion. This ensures that all regulatory elements controlling Pu activity are placed in one copy per chromosome of the native host. To produce a strain fully isogenic to P. putida MAD1 but encoding the XylR16-1 variant instead of the wild type, we used the method explained in Experimental procedures, which involved a two-step exchange between the insert of a donor plasmid carrying the mutated sequence and the chromosome of P. putida MAD1. The resulting strain, P. putida MAD16-1, equal is in all respects to P. putida MAD1 except for the sequence of the B region of XylR. With these assay materials, all changes detected in the phenotypes of the two strains can be traced unequivocally to the mutation under scrutiny.

Phenotypes endowed by B region mutant 16-1 (Val-219Asp/Ala-220Pro)

Figure 2 shows a comparison of the accumulation of β-galactosidase by P. putida MAD1 (wild-type xylR+) and P. putida MAD16-1 (xylR16-1+) cells exposed to a collection of mostly aromatic chemical species. A number of features become evident that were not apparent in the former E. coli-based reporter system of Fernández et al. (1995). First, although the mutant does cause a certain level of β-galactosidase in the absence of its natural effector m-xylene (lane 1 versus lane 2 in Fig. 2), such a level is moderate in respect to the fully induced peak reached in the presence of the inducer. In fact, it is manifest that the absolute level of β-gal accumulated by P. putida MAD16-1 induced with m-xylene is at least threefold higher than that of the wild-type P. putida MAD1.

image

Figure 2. Broadening of effector profile and increased promoter activity caused by mutation 16-1 (Val-219Asp and Ala-220Pro) within the B region of XylR. Cultures of each of the Pu–lacZ strains P. putida MAD1 (wt xylR+) and P. putida MAD16-1 (xylR16-1+) were grown in LB medium up to an A600 of 1.2, at which point they were exposed to the aromatic inducer indicated in each case. The accumulation of β-galactosidase (Miller units) originated by promoter activity was recorded 5 h later. Sample 1, no inducer; 2, saturating vapours of m-xylene; 3, 2 mM 3-chlorotoluene (3-CT); 4, 2 mM phenol; 5, saturating vapours of benzene diluted 1:5 with dibutyl phthalate; 6, 2 mM aniline; 7, 2 mM 4-fluor benzylalcohol (4-FBA); 8, saturating vapours of 4-chloro benzylalcohol (4-CBA); 9, 2 mM 4-nitrotoluene (4-NT); 10, 2 mM naphthalene; 11, 2 mM 1-methyl naphthalene; 12, saturating vapours of 2-methyl naphthalene; 13, saturating vapours of biphenyl; 14, saturating vapours of n-octane. Figures for β-galactosidase activity represent the average of duplicate samples from a minimum of four independent experiments. Inducers had no significant effect on growth rate under the assay conditions. The suggested protein structures of XylR and XylR16-1 are sketched on top of the corresponding bar diagrams.

Download figure to PowerPoint

A second effect of the mutation, as shown in Fig. 2, is that the XylR16-1 mutant appeared to have a much broader effector range than the non-mutated protein. This was revealed by subjecting the two strains to growth conditions in which bacteria were exposed to well-known XylR inducers (e.g. m-xylene, benzene or 3-chlorotoluene), poor inducers (naphthalene or 4-nitrotoluene) and a suite of non-inducers (phenol, 4-fluor benzylalcohol, 4-chloro benzylalcohol, 1-methyl naphthalene, biphenyl or n-octane). Figure 2 reveals that the mutant kept the responsiveness of the wild type to the standard effectors, but it also displayed a considerable reaction to compounds such as aniline (lane 6), 4-nitrotoluene (lane 9) and naphthalene (lane 10), which only partly induce the activity of the native protein. But, in addition, the experiment shown in Fig. 2 indicated that the mutant protein became responsive to compounds such as phenol (lane 4), 4-fluor benzylalcohol (lane 7) and 1-methyl naphthalene (lane 11), which are plain non-inducers for XylR. In this respect, it should be noticed that the effector profile is expanded in the mutant towards compounds that increasingly diverge from m-xylene in their chemical structure, electronic distribution and molecular size.

Another feature of the experiments summarized in Fig. 2 is that some non-inducers of the native protein, such as 4-chloro benzylalcohol (lane 8), biphenyl (lane 13) and n-octane (lane 14), turn out to abolish the otherwise significant basal level of β-galactosidase caused by the mutant regulator. A final facet of Fig. 2 is that the response of wild-type XylR to the collection of chemicals tested is not identical to that reported by Abril et al. (1989), when the promoter was tested in E. coli, perhaps pointing to the influence of host physiology and solvent tolerance on a suitable response.

The properties determined by the xylR16-1 allele can be rigorously traced to the disturbance of the B region of XylR

Although the data discussed above suggest that the distinct behaviour of P. putida MAD1 compared with P. putida MAD16-1 results from the properties of the mutant protein, they do not prove unequivocally that the phenotypes are linked to the structural mutation in the B domain of the regulator. The stability of reporter product (βgalactosidase) may be sensitive to proteases resulting from the heat shock triggered by exposure of the cells to aromatic or chaotropic compounds (Marqués et al., 1999) and could not therefore reflect the actual output or transcription. The mutant protein could be synthesized at different levels or its A domain could also be proteolytically degraded, thus producing an artifactual activation. Finally, the inhibitory chemicals mentioned above could simply arrest transcription from Pu at a step independent of any XylR-related function. To clarify these issues, we carried out the experiments summarized in Fig. 3. In these, we selected a subset of inducers that represented all possible responses found with the larger collection used in the tests in Fig. 2.

image

Figure 3. Characterization of the 16-1 allele (Val-219Asp and Ala-220Pro) of XylR.

A. Quantitative S1 nuclease protection assay of transcripts raised by XylR and XylR16-1 in the Pu promoter. The RNA extracted from P. putida MAD1 (xylR+) or P. putida MAD16-1 (xylR16-1+) treated with the inducers indicated was hybridized with an excess of a labelled ssDNA probe spanning the Pu promoter region and processed as explained in Experimental procedures. Although the transcription initiation site does correspond to that reported by Inouye et al. (1985), the doublets probably reflect a residual activity of the S1 nuclease at the extremes of the mRNA/DNA hybrid.

B. Expression in vivo of XylR and XylR16-1 under various induction conditions. Approximately 5 µg of protein lysates of P. putida MAD1 (xylR+) or P. putida MAD16-1 (xylR16-1+) cells recovered from cultures amended with the compounds indicated were run in a 10% denaturing polyacrylamide gel and subjected to Western blot analysis with an anti-XylR single-chain monoclonal antibody assembled in an M13 phage carrier. The positive control (C+) is P. putida KT2442 harbouring the multicopy xylR+ plasmid pTK19, whereas the negative control (C–) is the same strain devoid of any exogenous DNA.

C. Effect of positive and negative effectors of XylR16-1 on the constitutive activity of XylRΔA and stability of the LacZ reporter. A culture of P. putida MAD2 (xylRΔA+) was grown in LB to an A600 of 1.2 and exposed for 5 h to the inducers indicated in each case, after which its β-galactosidase levels were registered. The protein structure of XylRΔA is sketched as an insert. 4-FBA, 4-fluor benzylalcohol; 4-CBA, 4-chloro benzylalcohol; 4-NT, 4-nitrotoluene.

Download figure to PowerPoint

First of all, we verified that the data from β-galactosidase accumulation reflected in reality the levels of transcripts produced by Pu under the various induction conditions. Figure 3A shows a quantitative S1 protection assay set up to detect specific Pu transcripts in xylR+ and xylR16-1+P. putida cells collected from seven growth conditions and a variety of inducers. With the method specified in Experimental procedures, the only significant production of Pu-specific mRNA by P. putida MAD1 occurred when the cells were exposed to m-xylene. In contrast, P. putida MAD16-1 originated a detectable level of transcript in the absence of any inducer. Such a level was increased or decreased in a fashion that grossly followed the pattern revealed by the β-galactosidase activities shown in Fig. 2. Interestingly, the lack of β-galactosidase activity detected when cells were exposed to 4-Cl benzylalcohol was correlated with a total absence of transcripts under the same conditions, thus ensuring the significance of the inhibitory effect.

A second set of experiments was made to visualize the level and physical state of both XylR and XylR16-1 in the P. putida cells submitted to the different effectors. This is technically difficult, because the single copy of the regulator/reporter system per chromosome originates just a few molecules of the activator for each cell. Samples therefore failed to give any detectable signal with the polyclonal serum used in previous studies (Fernández et al., 1995). To overcome this problem, we resorted to the use of a high-affinity anti-XylR single-chain (scFv) monoclonal antibody attached as a protein fusion to the apical pIII protein of a full-size M13 phage (Kay et al., 1996). The whole hybrid phage was used to probe the Western blot in Fig. 3B, which was also treated with a second monoclonal antibody against the abundant capside protein of M13, followed by protein A coupled to a chemiluminescent reaction (see Experimental procedures). Such treatment greatly amplified the signals raised by XylR, which could be clearly detected in the blots, as shown in Fig. 3. With only modest differences, it appears that both XylR and its mutant variant 16-1 are expressed to non-limiting levels under all induction conditions regardless of whether the effect on Pu activity is positive (e.g. m-xylene) or negative (e.g. 4-Cl benzylalcohol).

The experiments above strengthened the notion that the unusual performance endowed by the XylR16-1 protein to the Pu–lacZ fusion of P. putida strain was linked to a genuine change in the properties of the activator brought about by the amino acid variations in the B region. However, the data did not elucidate whether the effects of the various inducers occurred before or after the protein had already acquired an active form. To ascertain this issue, we simply assayed the outcome of compounds in P. putida MAD2, as shown in Fig. 3C. This strain is entirely equivalent to P. putida MAD1, but its mini-Tn5 insertion expresses the truncated protein XylRΔA, which is altogether deleted of the A domain and therefore fully constitutive in the absence of any inducer (Fernández et al., 1995). When P. putida MAD2 was grown in the presence of the suite of aromatic effectors shown in Fig. 3C, the accumulation of β-galactosidase was not affected to any significant extent whether the inducers tested had a positive or a negative influence on Pu activity in P. putida MAD1 or P. putida 16-1. This result not only indicated that the effects of the XylR16-1 inducers must occur before full activation of the protein, but also that none of the aromatics tested had any effect on the β-galactosidase enzyme used as reporter. In addition, as 4-chloro benzylalcohol fails to inhibit XylRΔA, the experiment in Fig. 3A also indicated that the inhibition caused by this compound on the low constitutive activity of XylR16-1 requires the presence of the A domain. Some possible explanations for this fact are considered below.

A second mutation A220–N221 × HHHR within the B region of XylR exacerbates the 16-1 phenotype

The results presented above substantiate a veritable change in the properties of the XylR protein when the B sequence was disrupted by the Val-219Asp/Ala-220Pro change. However, as prolines enter a very specific rigid kink in secondary structures (Chou and Fasman, 1978), it is unclear whether the phenotypes can be traced to just this type of alteration or whether other changes may lead to the same functional result as well. To gain some insight into these possibilities, we examined the properties of a second variant of XylR mutated at the same site as XylR16-1 but with the mutation named 3H (Fig. 1). This change consists of the replacement of amino acids A220–N221 of the B region of XylR by the peptide HHHR. This modification enters a major change in the native sequence, involving both substitutions and insertions of amino acid residues. As in the case of the XylR16-1 protein, the mutant sequence was entered into the xylR gene of P. putida MAD1 by homologous recombination so that the resulting strain, P. putida MAD3H, could be faithfully compared with its predecessors. The data shown in Fig. 4 summarize the phenotypes with respect to Pu performance kicked in P. putida by the xylR3H mutant, which (as detected in E. coli;Fernández et al., 1995), produced a certain basal transcription level. As shown in the inserts in Fig. 4, the data from S1 protection assays verified that the measured β-galactosidase contents reflected the actual transcription from Pu. Furthermore, the Western blot experiments showed that the intracellular amounts of the mutant regulator did not vary much with the inducer present in the medium. But the most revealing features of the same data were (i) that the induced mutant protein appeared to be substantially more active than either the wild-type XylR or the previous XylR16-1 mutant (compare Fig. 4 with Fig. 2); and (ii) that the 3H mutation caused the very same broadening of the effector profile observed for XylR16-1 and similar induction/inhibition patterns for the different compounds tested. In other words, the phenotype of the 3H mutant looked like an exacerbation of that already caused by XylR16-1, the difference being more quantitative than qualitative.

image

Figure 4. Phenotypes endowed by mutation 3H at the B domain of XylR. The graphic shows the dramatic broadening of effector profile and increased promoter activity caused by mutation 3H (exchange of A220–N221 for HHHR) within the B domain of XylR. This was revealed by measuring the β-galactosidase levels of cultures of P. putida MAD16-1 (xylR3H+) exposed to the inducers shown under conditions identical to those specified in the legend to Fig. 2. The inserts in the graphic represent the S1 nuclease protection assay of transcripts raised by XylR3H in the Pu promoter (top) and the expression in vivo of the mutant protein under various induction conditions revealed with a Western blot (circumstances and procedures similar to those indicated in the legend to Fig. 3A and B). Inducers: 1, nil; 2, saturating vapours of m-xylene; 3, 2 mM 3-chlorotoluene (3-CT); 4, 2 mM phenol; 5, saturating vapours of benzene diluted 1:5 with dibutyl phthalate; 6, 2 mM aniline; 7, 2 mM 4-fluor benzylalcohol (4-FBA); 8, saturating vapours of 4-chloro benzylalcohol (4-CBA); 9, 2 mM 4-nitrotoluene (4-NT); 10, 2 mM naphthalene; 11, 2 mM 1-methyl naphthalene; 12, saturating vapours of 2-methyl naphthalene; 13, saturating vapours of biphenyl; 14, saturating vapours of n-octane.

Download figure to PowerPoint

Because disrupting the sequence of the B region of XylR through different changes in amino acids originated a similar behaviour in the resulting protein variant, we conclude that the linker is not a mere hinge, as proposed before for this type of protein and other regulators of the family (Wootton and Drummond, 1989; Bantscheff et al., 1999), but it plays a dynamic role in the activating process. In fact, it appears that activation of the protein by m-xylene may require not only the offsetting of the specific contacts between the A and C domains of XylR (Pérez-Martín and de Lorenzo, 1995), but also the disturbance of a protein form stabilized by the structure of the B region. Some further observations to support this notion are described below.

Monitoring the affinity of XylR16-1 and XylR3H for suboptimal inducers in vivo

To detect changes in the apparent affinities of the XylR variants for otherwise suboptimal inducers, we subjected strains P. putida MAD1, P. putida MAD16-1 and P. putida MAD3H to the series of experiments shown in Fig. 5. 3-Cl toluene and naphthalene were selected because they are poor effectors (although to different degrees) of the wild-type XylR protein (Fig. 2, lanes 3 and 10), but they become reliable inducers for both XylR16-1 and XylR3H. In addition, both aromatic compounds are reasonably soluble, so their concentration in the medium can be fixed better than that of other volatile inducers.

image

Figure 5. Dose–response patterns of XylR, XylR16-1 and XylR3H against varying concentrations of 3-chlorotoluene (3-CT) and naphthalene.

A–C. The results of exposing cultures of P. putida MAD1 (wt xylR+), P. putida MAD16-1 (xylR16-1+) and P. putida MAD3H (xylR3H+) to increasing concentrations of the suboptimal inducer 3-CT added to the medium.

D–F. The same experiment performed with increasing concentrations of naphthalene. The effects on Pu promoter activity were monitored by measuring the β-galactosidase levels of the cultures of each of the strains exposed to the inducers for 5 h after reaching an A600 of 1.2. Note the different scales on the y-axes.

Download figure to PowerPoint

To monitor shifts in the operative affinities of the XylR variants for 3-Cl toluene and naphthalene, we used the criteria of Ramos et al. (1990) to estimate in vivo the apparent Ks of effectors for prokaryotic transcriptional regulators. The procedure can be run on the basis that (i) the effector is transported into the cells through non-limiting facilitated diffusion; and (ii) the output of the reporter–lacZ fusion faithfully reflects the activity of the regulator. As these two conditions can be safely assumed in our system, we set out to examine the induction of the Pu–lacZ fusions borne by P. putida through varying effector concentrations. As shown in Fig. 5, each of the strains was subjected to induction experiments with the aromatic compounds ranging from 10−5 mM to 10 mM. The graphs in Fig. 5 show the dose–response curves of each of the P. putida strains to increasing concentrations of the aromatic inducers. The different scales used for the y-axes (β-galactosidase activities) should be noticed for the interpretation of the data discussed below.

Besides the remarkable gain in total activity brought about by the introduction of the 16-1 and the (still more effective) 3H mutations in XylR, comparison of the β-galactosidase levels raised by each inducer revealed some changes in the induction responses of the proteins for either effector. Data were collected over the range 0.0001–2 mM, as the inducers appeared to become too toxic beyond that level. As shown in Fig. 5A–C, the minimum amount of 3-Cl toluene required to induce a consequential response in Pu moved from the 0.1 mM range for the wild-type protein down to 0.01 mM for either XylR16-1 or XylR3H, thus reflecting a probable increase in the affinity of the regulator for its inducer. Such a phenomenon was difficult to quantify because, at higher inducer concentrations, the toxicity of 3-Cl toluene might affect the dose–response curve. However, a conservative comparison of the data in Fig. 5A–C would suggest that such an increase in apparent affinity is within one order of magnitude. A similar enhanced affinity for naphthalene is indicated by the graphs in Fig. 5D–F, although the concentrations at which the phenomenon becomes more manifest were in the 0.1–1 mM range.

Taken together, the experiments shown in Fig. 5 indicate that one of the consequences of the 16-1 and 3H mutations is an increase in the affinity of the protein for otherwise suboptimal effectors. However, we cannot entirely explain the broadening of the effector profile or the phenomenal increase in the intrinsic activity of the protein solely on the basis of such a moderate change in affinity within one order of magnitude. Besides an augmentation in affinity of that kind, the results suggest the concurrent generation of a more permissive inducer binding site and the origination of a protein form that is released from a bottleneck in its activation pathway.

A refined model for the activation of XylR by aromatic effectors

The results presented here oblige us to rework the current model for XylR (and other like proteins; Shingler, 1996) that has been generally accepted for the last 5 years. One key piece of information in this respect is the veritable broadening of the effector profile of the protein caused by mutations that disrupt the sequence of its B region (Figs 2 and 4). This is not easy to explain with the current model (Pérez-Martín and de Lorenzo, 1995; Shingler, 1996), which claims that the aromatic effectors interact exclusively with the A module of the regulator (Fig. 6). We argue in this work that such a relaxed specificity may not simply be attributed to an increase in the affinity of the protein for suboptimal inducers (Fig. 5). On the other hand, the responsiveness to the new compounds is built on partially active proteins, i.e. they are regulators that have partly overcome the intramolecular repression caused by the A domain in the rest of the activator.

image

Figure 6. Model for the outcome of the B region on effector binding. The scheme is a refinement of the current model, which assumes that XylR is a protein organized in three major modules (A, reversible repression; C, activation; D, DNA binding) and a linker B module connecting A and C. The wild-type XylR protein is represented as a dimer upheld (at least in part) by the interactions between the B modules of each monomer (the B domains are predicted with 90% probability to be dimerization elements; see text). The data presented here are consistent with the notion that the contour of the effector pocket may encompass a surface on the central C domain, so that the displacement or loosening of the A and C interactions brought about by distant mutations at the B region could produce a more permissive effector binding site and a more active protein.

Download figure to PowerPoint

Figure 6 shows a plausible explanation for these effects, in which the protein is maintained in an inactive state through the joint action of two sets of intramolecular contacts. One of these is the interaction between the A domain and the central C domain. But, in addition, the B region contributes to the inactivity of the protein in a manner that requires the maintenance of its structure. The high probability that the sequence of the B linker engages a coiled-coil structure and not a simple, flexible coil (Q linker), as proposed by Wootton and Drummond (1989) for other proteins of this type, suggests that the B region is structured and may even determine a dimerization/oligomerization module. The 90% likeness to a coiled-coil, as calculated according to the criteria of Lupas (1996), drops down to nearly zero when predictions are made on the amino acid sequence of the region borne by the mutants 16-1 or 3H. Within this hypothesis, the hindrance of an intermolecular dimerization of the B region with an adjacent protein could be necessary (but not sufficient) for the full activation of the protein. The semi-constitutivity of the mutants could thus be explained because they may have overcome one set of repressive intramolecular contacts. The other set could be determined by the direct interaction of a subdomain of the A module (spanning amino acids 170–210; Pérez-Martín and de Lorenzo, 1996b) with the central C portion of the activator and may be relieved only by effector binding.

The broadening of the effector profile of XylR in the semi-constitutive mutants can easily be understood in the refined model in Fig. 6, which makes compatible present and previous data. Although the size and gross configuration of the aromatic binding site mostly engages protein surface(s) of the A domain, such a pocket may become more permissive by the action at a distance of the mutations in the B region. This case accounts for the inhibition of the semi-constitutivity of the mutant proteins by some compounds such as 4-Cl benzylalcohol (Figs 3 and 4). As the same compound has no effect on the fully constitutive XylRΔA protein (Fig. 3C), it seems possible that the inhibition is the result of locking the interactions between the A and C domains. Along these lines, Salto et al. (1998) observed that the non-inducer 3-nitrotoluene inhibited the semi-constitutive activity of mutants E172K and P85S, indicating that some aromatic compounds may fasten the protein in a non-productive state.

In summary, the novel model for XylR activation that emerges from the experiments shown here is that the effector binding has to overcome at least two steps before the full performance of the protein as a transcriptional factor. The proposed primary effect of m-xylene binding is to disrupt the interdomain interactions. But this is still insufficient until the next stage (e.g. hindrance of the secondary/tertiary structure of the B region) has been overcome as well. Within this scheme, the XylR mutants 16-1 and 3H would simply be locked in the second activation stage, which would still require effector binding for full performance. An alternative could be that the active site of XylR is able to bind a broad series of effectors but, once they are bound, they could induce changes that are processed differentially through the B region. Mutations in the B linker could thus result in a different set of signal transmissions, so that the apparent response to effectors is distinct. This view is, however, more intricate, and we thus favour the simpler option. In any case, each of these models poses a number of genetic and biochemical predictions that deserve further study.

Experimental procedures

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results and discussion
  5. Experimental procedures
  6. Acknowledgements
  7. References

Strains and plasmids

Relevant strains and plasmids used in this work are listed in Table 1. E. coli strains CC118, CC118λpir and HB101 have been published before (Herrero et al., 1990; de Lorenzo and Timmis, 1994). The reference P. putida strain KT2442 has also been described elsewhere (de Lorenzo and Timmis, 1994). P. putida MAD1 bears a hybrid mini-Tn5 transposon, which includes a tellurite resistance (Tel) selection marker, the sequence of the wild-type xylR gene expressed through its native promoter and a transcriptional Pu–lacZ fusion (Fernández et al., 1995). This ensures that all regulatory elements controlling expression from Pu are placed in one copy per chromosome. P. putida MAD2 bears an equivalent insertion that encodes the truncated protein XylRΔA deleted of amino acids 1–223 but expressed under the same translation initiation region and promoter (Pr), which drives expression of the native xylR protein in P. putida MAD1. Other plasmids and strains are referred to briefly in Table 1.

Table 1. Bacterial strains and plasmids.
Strain/plasmidRelevant genotype/phenotypeReference
E. coli
 CC118Δ(ara-leu), araD, ΔlacX74, galE, galK, phoA, thi1, rspE, rpoB, argE (am), recA1 Herrero et al. (1990)
 CC118 λpirSame as CC118 but lysogenized with phage λpir Herrero et al. (1990)
 HB101 rpsL (Smr), recA, thi, pro, leu, hsdRhsdR+ (E. coli K-12/E. coli B hybrid) Sambrook et al. (1989)
P. putida
 KT2442Prototrophic, rifampicin-resistant derivative of reference strain P. putida KT2440 Herrero et al. (1990)
 MAD1KT2442 inserted with mini-Tn5 MAD1 Telr, Pr[RIGHTWARDS ARROW]xylR/Pu[RIGHTWARDS ARROW]lacZ Fernández et al. (1995)
 MAD2KT2442 inserted with mini-Tn5 MAD2 Telr, Pr[RIGHTWARDS ARROW]xylRΔA/Pu[RIGHTWARDS ARROW]lacZ Fernández et al. (1995)
 MAD16-1MAD1 in which the wild-type xylR sequence has been replaced by the xylR161 alleleThis work
 MAD3HMAD1 in which the wild-type xylR sequence has been replaced by the xylR3H alleleThis work
Plasmids
 pCKMutCmr, pSC101 replicon, pCK01 vector inserted with a 1.7 kb NotI fragment bearing the sequence of xylR16-1 allele Fernández et al. (1995)
 pCKHisCmr, pSC101 replicon, pCK01 vector inserted with a 1.7 kb NotI fragment bearing the sequence of xylR3H allele Fernández et al. (1995)
 pKNG101Smr, oriV R6K, oriT RK2+, sacRB+ conditional replication, counterselectable in sucrose, mobilizable Kaniga et al. (1991)
 pKNGMutpKNG101 inserted with the 1.7 kb NotI fragment of pCKMut encoding xylR161This work
 pKNGHispKNG101 inserted with the 1.7 kb NotI fragment of pCKHis encoding xylR3HThis work
 pRK2013Kmr, oriV pBR322, oriT RK2+, tra RK2+ helper for mobilization of oriT RK2+-containing plasmids Herrero et al. (1990)
 pMADApr, pUJ9 vector (de Lorenzo et al., 1990) inserted with a 312 bp EcoRI–BamHI fragment spanning the entire Pu promoter sequence Cases et al. (1996)
 pTK19Kmr, pKT231 vector inserted with a 2.4 kb HpaI segment from the TOL plasmid spanning the wild-type xylR sequence de Lorenzo et al. (1993)

Delivery of xylR alleles into the chromosome of P. putida MAD1

The amino acid changes at the B region of XylR present in variants XylR16-1 and XylR3H (Fig. 1) were targeted to the xylR sequence of the mini-Tn5 MAD1 insertion of P. putida MAD1 by homologous recombination. To this end, we first transferred the 1.7 kb NotI inserts of plasmids pCKMut (encoding xylR16-1) and pCKHis (encoding xylR3H) to plasmid pKNG101 (Kaniga et al., 1991). This vector and the derivatives thus obtained (pKNGMut and pKNGHis) were mobilized towards P. putida MAD1 by triparental matings with donor strains E. coli CC118λpir (pKNGMut) and E. coli CC118λpir (pKNGHis) along with the helper strain E. coli HB101 (pRK2013), which provided the required transfer functions (de Lorenzo and Timmis, 1994). Double cross-over of homologous sequences in the region of interest was promoted by first plating the mating mixtures on selective medium with Sm (60 µg ml−1) and Tel (40 µg ml−1), which selects co-integrates of the inserts borne by pKNGMut and pKNGHis with the mini-Tn5 MAD1. In a second step, Smr Telr exconjugants were grown without selection and streaked out on LB plates with Tel, 5% sucrose (which is toxic for sacB+ cells) and Xgal (5-Br, 4Cl, 3-indolyl-β-d-galactoside). The blue colonies thriving in such a medium were finally screened for loss of resistance to Sm as an indication of co-integrate resolution. The allele exchange between the wild-type xylR sequence and its two variants was verified by amplification of a 1.7 kb portion of the xylR sequence corresponding to amino acids 1–567. The amplified segments were inspected by restriction analysis with BamHI (for xylR16-1) and ClaI (for xylR3H) and finally sequenced to ensure the result of the predicted recombination and the maintenance of the rest of the xylR sequence other than the mutated region. Two fully confirmed strains containing the xylR16-1 and xylR3H alleles were designated P. putida MAD16-1 and P. putida MAD3H respectively (Table 1).

Growth and induction conditions

Unless indicated otherwise, P. putida strains were pregrown overnight at 30°C in LB medium before any procedure. For induction experiments, the cultures were diluted 100-fold in fresh medium and grown with vigorous shaking up to an absorbance at 600 nm (A600) of 1.2. For volatile inducers, the samples were then exposed to saturating vapours of the aromatic compounds under scrutiny in airtight flasks. These were incubated further for 5 h and their β-galactosidase levels measured as explained below. Alternatively, once the cultures had reached an A600 of 1.2, they were added with the desired inducer (predissolved in DMSO) and incubated in the same airtight flasks. The toxicity of benzene was reduced by co-saturating the vapour phase with a 1:5 dilution of the hydrocarbon in dibutyl phthalate (which is not a XylR inducer). Chemicals used for induction experiments were purchased from Aldrich, Fluka or Merck and were always of superior analytical purity (purity grade ≥ 99%).

Monitoring promoter performance

Pu activity was followed in all cases by assaying the accumulation of β-galactosidase in cells carrying one copy per chromosome of a lacZ transcriptional fusion. β-Galactosidase assays were carried out on cells permeabilized with chloroform and SDS as described by Miller (1972) under the conditions specified in each condition. The linearity of the assay within the range of cell densities and the time of reaction with ONPG was verified in all cases. β-Galactosidase activity values given throughout this paper represent the average of at least three independent experiments, each of which was conducted in duplicate samples, with deviations of < 15%.

DNA and protein techniques

Recombinant DNA manipulations and protein analysis were carried out according to published protocols (Sambrook et al., 1989). For detection of the XylR protein and its variants, 5 µg of protein extract of whole P. putida cells was denatured in a sample buffer containing 2% SDS and 5% β-mercaptoethanol and run in 10% polyacrylamide gels. These were subsequently blotted onto an Immobilon-P membrane (Millipore) and probed with an anti-XylR single-chain monoclonal antibody bound to M13 phage (kindly provided by L. A Fernández). The protein bands corresponding to XylR or its derivatives were developed using an anti-M13 monoclonal antibody coupled to horseradish peroxidase (Amersham Pharmacia Biotech), followed by reaction with a chemoluminescent substrate (Amersham).

Quantitative S1 nuclease analysis of Pu-borne transcripts

Total RNA was isolated from bacterial cultures through an adaptation of the standard phenol extraction procedure, as described previously by Monsalve et al. (1995) and Yuste et al. (1998). The source of the single-stranded DNA (ssDNA) probe used in the experiments with the S1 nuclease was plasmid pMAD (Cases et al., 1996; Table 1). This plasmid bears an insert with the sequence spanning positions −209 to +82 relative to the Pu transcription start site. To produce the ssDNA probe used in our assays, pMAD was cleaved with EcoRI, hybridized with the oligonucleotide Lac1 (5′-TCAATCGCTGCTTTTACCTG-3′, labelled in 5′ with 32P) and subjected to 30 cycles of linear amplification with Taq polymerase. As the 5′ end of the labelled primer binds 141 nucleotides downstream of the Pu transcription start site, the ssDNA produced by the amplification spanned positions +141 (5′ end) to −209 (3′ end). For the S1 nuclease protection assays, 25 µg of RNA from each sample was hybridized to an excess (104 c.p.m.) of the labelled ssDNA probe, digested with S1 and processed further as described in detail by Ausubel et al. (1989). Samples were loaded in a DNA sequencing gel with 7 M urea, run at high voltage, dried and autoradiographed on Amersham X-ray film.

Computer-assisted procedures

Structural predictions of the protein segment spanning amino acids 208–237 of the XylR sequence (Inouye et al., 1988) were made with the lupas program (Lupas, 1996), which is available at the http://www.ch.embnet.org/software/COILS_ form.html site. Scanning and quantification of the autoradiographic signals from the S1 nuclease protection experiments (Figs 3A and 4) were carried out with an Agfa ARCUS II system and further processed with the Adobe Photoshop 4.0 program.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results and discussion
  5. Experimental procedures
  6. Acknowledgements
  7. References

The authors are indebted to V. Shingler and E. O'Neill (University of Umeå, Sweden) for sharing results before publication, to I. Canosa for help with the S1 protection assays, and to L. A. Fernández for the kind gift of the anti-XylR monoclonal. I. Cases is also gratefully acknowledged for help with structural predictions and inspiring discussions. This work was supported by contracts BIO4-CT97-2040 and QLRT-99-00041 from the European Commission and by grant BIO98-0808 from the Spanish Comisión Interministerial de Ciencia y Tecnología (CICYT). J.G. was the recipient of a Predoctoral Fellowship from the Eusko Jaurlaritza (Basque government).

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Results and discussion
  5. Experimental procedures
  6. Acknowledgements
  7. References
  • Abril, M.A., Michán, C., Timmis, K.N., Ramos, J.L. (1989) Regulator and enzyme specificities of the TOL plasmid-encoded upper pathway for degradation of aromatic and expansion of the substrate range of the pathway. J Bacteriol 171: 67826790.
  • Ausubel, F.M., Brent, R., Kingston, R.E., Moore, D.D., Seidman, J.G., Smith, J.A., Struhl, K. (eds) (1989) Current Protocols in Molecular Biology. New York: John Wiley and Sons.
  • Bantscheff, M., Weiss, V., Glocker, M.O. (1999) Identification of linker regions and domain borders of the transcription activator protein NtrC from Escherichia coli by limited proteolysis, in-gel digestion and mass spectrometry. Biochemistry 38: 1101211020.
  • Cases, I., De Lorenzo, V., Pérez-Martín, J. (1996) Involvement of σ54 in exponential silencing of the Pseudomonas putida TOL plasmid Pu promoter. Mol Microbiol 19: 717.
  • Chou, P.Y. & Fasman, G.D. (1978) Prediction of the secondary structure of proteins from their amino acid sequence. Adv Enzymol 47: 145148.
  • Dixon, R. (1986) The xylABC promoter from the Pseudomonas putida TOL plasmid is activated by nitrogen regulatory genes in Escherichia coli. Mol Gen Genet 203: 129316.
  • Fernández, S., De Lorenzo, V., Pérez-Martín, J. (1995) Activation of the transcriptional regulator XylR of Pseudomonas putida by release of repression between functional domains. Mol Microbiol 16: 20252213.
  • Herrero, M., De Lorenzo, V., Timmis, K.N. (1990) Transposon vectors containing non-antibiotic resistance selection markers for cloning and stable chromosomal insertion of foreign genes in gram-negative bacteria. J Bacteriol 172: 65576567.
  • Inouye, S., Nakazawa, A., Nakazawa, T. (1985) Determination of the transcription initiation site and identification of the protein product of the regulatory gene xylR for xyl operons on the TOL plasmid. J Bacteriol 163: 863869.
  • Inouye, S., Nakazawa, A., Nakazawa, T. (1988) Nucleotide sequence of the regulatory gene xylR of the TOL plasmid from Pseudomonas putida. Gene 66: 301306.
  • Kaniga, K., Delor, I., Cornelis, G.R. (1991) A wide-host-range suicide vector for improving reverse genetics in Gram-negative bacteria: inactivation of the blaA gene of Yersinia enterocolitica. Gene 109: 137141.
  • Kay, B.K., Winter, J., McCafferty, J. (1996) Phage Display of Peptides and Proteins. A Laboratory Manual. San Diego: Academic Press.
  • De Lorenzo, V. & Timmis, K.N. (1994) Analysis and construction of stable phenotypes in Gram-negative bacteria with Tn5- and Tn10-derived minitransposons. Methods Enzymol 235: 386405.
  • De Lorenzo, V., Herrero, M., Jakubzik, U., Timmis, K.N. (1990) Mini-Tn5 transposon derivatives for insertion mutagenesis, promoter probing, and chromosome insertion of cloned DNA in Gram-negative eubacteria. J Bacteriol 172: 65686572.
  • De Lorenzo, V., Fernández, S., Herrero, M., Jacubzik, U., Timmis, K.N. (1993) Engineering alkyl- and haloaromatic-responsive gene expression with mini-transposons containing regulated promoters of biodegradative pathways of Pseudomonas. Gene 130: 4146.
  • Lupas, A. (1996) Prediction and analysis of coiled-coil structures. Methods Enzymol 266: 513525.
  • Marqués, S., Manzanera, M., Gonzalez-Perez, M.M., Gallegos, M.T., Ramos, J.L. (1999) The XylS-dependent Pm promoter is transcribed in vivo by RNA polymerase with σ32 or σ38 depending on the growth phase. Mol Microbiol 31: 11051113.
  • Miller, J.H. (1972) Experiments in Molecular Genetics. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press.
  • Monsalve, M., Mencía, M., Rojo, F., Salas, M. (1995) Transcriptional regulation in bacteriophage φ29: expression of the viral promoters throughout the infection cycle. Virology 207: 2331.
  • Morett, E. & Segovia, L. (1993) The σ54 bacterial enhancer-binding protein family: mechanism of action and phylogenetic relationship of their functional domains. J Bacteriol 178: 60676074.
  • Ng, L.C., O’Neill, E., Shingler, V. (1996) Genetic evidence for interdomain regulation of the phenol-responsive σ54-dependant activator DmpR. J Biol Chem 271: 1728117286.
  • North, A., Klose, K.E., Stedman, K.M., Kustu, S. (1993) Prokaryotic enhancer-binding proteins reflect eukaryote-like modularity: the puzzle of nitrogen regulatory protein C. J Bacteriol 175: 42674273.
  • O'neill, E., Ng, L.C., Sze, C.C., Shingler, V. (1998) Aromatic ligand binding and intramolecular signalling of the phenol-responsive σ54-dependent regulator DmpR. Mol Microbiol 28: 131141.
  • Pérez-Martín, J. & V. De Lorenzo, (1995) The amino-terminal domain of the prokaryotic enhancer-binding protein XylR is a specific intramolecular repressor. Proc Natl Acad Sci USA 92: 93929396.
  • Pérez-Martín, J. & V. De Lorenzo, (1996a) ATP binding to the σ-54 dependent activator XylR triggers a protein multimerization cycle catalyzed by UAS DNA. Cell 86: 331339.
  • Pérez-Martín, J. & V. Lorenzo, (1996b) Identification of the repressor subdomain within the signal reception module of the prokaryotic enhancer-binding protein XylR of Pseudomonas putida. J Biol Chem 271: 78997902.
  • Ramos, J.L., Michán, C., Rojo, F., Dwyer, D., Timmis, K. (1990) Signal–regulator interactions. Genetic analysis of the effector binding site of XylS, the benzoate-activated positive regulator of Pseudomonas TOL plasmid meta-cleavage pathway operon. J Mol Biol 211: 373382.
  • Ramos, J.L., Marqués, S., Timmis, K.N. (1997) Transcriptional control of the Pseudomonas TOL plasmid catabolic operons is achieved through an interplay of host factors and plasmid encoded regulators. Annu Rev Microbiol 31: 341374.
  • Salto, R., Delgado, A., Michán, C., Marqués, S., Ramos, J.L. (1998) Modulation of the function of the signal receptor domain of XylR, a member of a family of prokaryotic enhancer-like positive regulators. J Bacteriol 180: 600604.
  • Sambrook, J., Fritsch, E.F., Maniatis, T. (1989) Molecular Cloning. A Laboratory Manual, 2nd edn. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press.
  • Shingler, V. (1996) Signal sensing by sigma 54-dependent regulators: derepression as a control mechanism. Mol Microbiol 19: 409416.
  • Wootton, J.C. & Drummond, M.H. (1989) The Q-linker: a class of interdomain sequences found in bacterial multidomain regulatory proteins. Protein Eng 2: 535543.
  • Yuste, L., Canosa, I., Rojo, F. (1998) Carbon-source-dependent expression of the PalkB promoter from the Pseudomonas oleovorans alkane degradation pathway. J Bacteriol 180: 52185226.