INTRODUCTION

Top Abstract Introduction Materials and Methods Results and Discussion References

The Pu promoter of Pseudomonas putida drives transcription of the upper operon for degradation of toluene and p- and m-xylenes of plasmid pWW0 (3). This promoter depends on the alternative sigma factor ⁵⁴ and requires the activator protein XylR. In response to the presence of pathway inducers, XylR, in coordination with the integration host factor (IHF) and a ⁵⁴-containing RNA polymerase, activates Pu (29). In vivo, this promoter is also affected by at least two dissimilar regulatory circuits that couple Pu activity to the general physiology of the cell. One of these circuits involves so-called exponential silencing, which avoids Pu activation during rapid growth in rich media despite the presence of inducer molecules (6, 9, 14, 21). The other circuit involves modulation of Pu by other available carbon sources. In particular, glucose or gluconate can inhibit Pu, reducing its activity to one-third the normal activity (8, 13). While exponential silencing is believed to act by modulating ⁵⁴ activity (6), carbon-mediated inhibition affects Pu through the product of the ptsN gene; P. putida mutants lacking ptsN are not sensitive to glucose inhibition (8), although exponential silencing is not affected (5).

ptsN maps in a gene cluster that also includes rpoN, the gene that encodes ⁵⁴, and three other open reading frames: ORF102, ORF284, and the ptsO gene (8). The whole cluster is well conserved in gram-negative bacteria. Mutations in some of these reading frames have been characterized in different species (15, 16, 22, 23, 27). In P. putida, no insertion mutation in any of the three genes increases the activity of Pu in the absence of glucose, unlike disruption of ptsN. Neither do such mutations suppress exponential silencing (8). While the sequences of ORF102 and ORF284 provide no clues about the possible functions of these open reading frames, the ptsN- and ptsO-encoded proteins (IIA^Ntr and NPr, respectively) show similarity to phosphotransferases belonging to the phosphoenolpyruvate:sugar phosphotransferase system (PTS) family, particularly members of the IIA and HPr subfamilies (8, 18). IIA and HPr proteins participate in a phosphorylation cascade that also involves the enzyme I (EI) protein. EI uses phosphoenolpyruvate as a phospho donor and phosphorylates HPr, which is then able to transfer the phosphate moiety to IIA domains (26). In gram-positive and gram-negative bacteria, the main role of the PTS family of phosphotransferases thus involves transport of sugars across cell membranes simultaneously with activation of the sugars by phosphorylation. This is done in such a way that the presence or absence of sugars affects the phosphorylation of these PTS family proteins (26).

These changes in phosphorylation have regulatory effects on transcription, chemotaxis, and general metabolism (26, 31). While a role for the ptsN gene product in glucose transport has been discredited by previous data (8), phosphorylation of this protein seems to be a key event in Pu regulation. A modification of the IIA^Ntr sequence that prevents phosphorylation renders Pu unresponsive to glucose, while a similar mutation believed to mimic the phosphorylation state of IIA^Ntr gives rise to a superrepressed phenotype (i.e., low levels of Pu activity either in the presence or in the absence of glucose) (8). In the present investigation we examined the role of ptsO in Pu regulation and its connections with ptsN. We obtained genetic evidence showing that ptsO and ptsN operate together in Pu regulation and that phosphorylation of NPr is necessary for the normal response of Pu to glucose. The relationships between this phenomenon and other P. putida PTS proteins are discussed below in the context of Pu regulation.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results and Discussion References

Strains, plasmids, media, and general methods To examine Pu activity, P. putida MAD2 (12) or derivatives of this strain were used. MAD2 is a derivative of P. putida KT2442 with a chromosomal Pu::lacZ fusion and the xylR allele named xylRA in a tellurite resistance (Tel^r) minitransposon vector (33). The loss of the N-terminal A domain of XylR renders the protein constitutively active in the absence of aromatic inducers (12). The ptsN::Km and ptsO::Km strains used are described in detail elsewhere (8). Essentially, these strains are MAD2 derivatives in which the ptsN or ptsO gene has been disrupted by gene replacement with a promoterless Km^r cassette.

Construction of insertion double mutants To construct the ptsN ptsO double-mutant strain, it was necessary to first disrupt the ptsN gene with a compatible marker and then recombine the resulting DNA fragments in the ptsO::Km strain. To do this, a xylE gene with neither transcription start nor termination signals was first excised as an XmaI fragment from the pXyLE10 plasmid (35) and introduced into a pBluescript (1) derivative, pBSNTR (8), containing a P. putida chromosome fragment with the ORF102 and ptsN genes and parts of the rpoN and ORF284 genes. This plasmid contained a unique XmaI site in the ptsN gene. The resulting plasmid, pBSNTR154X, was then digested with enzymes ApaI and XbaI to liberate the chromosome region with the ptsN gene disrupted. The resulting fragment was introduced into the pKNG101 plasmid (17), previously linearized with the same enzymes, resulting in pKNG154X. This plasmid was introduced by triparental mating into the P. putida ptsO::Km strain. Integration of the plasmid into the chromosome was determined by selection from the mating mixture by plating in the presence of streptomycin.

Site-directed mutagenesis of ptsO and expression of the mutant protein. The method developed by Kunkel et al. (19) was used to generate the H15A variant of ptsO. To do this, the SalI fragment of pARG5.1 (see above) was cloned in vector pGC1 (25). In vitro extension of the single-stranded, uracil-containing DNA from the resulting plasmid (pGC90) was primed independently with mutagenic oligonucleotide PTSOH15A (5' GGCTGCCGCCCGGGCCGCTAGCCCCAGCTTGTT 3' [the changed nucleotides are in boldface type, and the triplet corresponding to the new amino acid residue is underlined]). The changes introduced a new NheI restriction site to facilitate screening of mutants. The resulting plasmid, pGC90-H15A, was digested with EcoRI and HindIII, and the released 1.1-kb fragment was cloned in the pVLT35 expression vector (10), which resulted in plasmid pVLT290-H15A. An equivalent plasmid, pVLT290, containing the wild-type ptsO allele, was also constructed and used as a control. Each of the three resulting plasmids was transferred to the P. putida MAD2 ptsO::Km strain by triparental mating as described above.

Construction of ptsP and multiphosporyl transfer protein (MTP) insertion mutants. The complete genome sequence of P. putida KT2440, available at The Institute for Genome Research (TIGR) website (http://www.tigr.org/), was studied to determine the presence of proteins with an EI domain that might be involved in NPr phosphorylation. To do this, the E. coli EI^Ntr protein (SwissProt accession code PT1P  ECOLI) was used as an electronic probe with the TBLASTN program at the TIGR website (http://www.tigr.org/cgi-bin/BlastSeach/blast.cgi?). The two resulting sequences were used to design oligonucleotide primers that amplified partial gene sequences from the P. putida KT2442 chromosome, a fragment long enough to be useful for homologous recombination.

Protein techniques. Western blot assays to detect the ptsN product were performed as described by Cases et al. (6). Equal amounts of whole Pseudomonas cells (typically 10⁸ cells) were lysed in a sample buffer with 2% sodium dodecyl sulfate and 5% -mercapthoethanol and electrophoresed in denaturing 12% polyacrylamide gels. These gels were subsequently blotted and probed with a 1:1,000 dilution of a preadsorbed rabbit serum raised against a purified MalE-IIA^Ntr fusion protein expressed in E. coli for detection of IIA^Ntr and with a mouse antiserum raised against NPr overexpressed in E. coli and purified from inclusion bodies. The blots were finally developed with protein A coupled to horseradish peroxidase by using an ECL+ kit (Amersham) as recommended by the manufacturer.

	RESULTS AND DISCUSSION

Top Abstract Introduction Materials and Methods Results and Discussion References

ptsO mutant strain showed low Pu activity. In a previous study (8), it was observed that one mutation in the ptsN gene of the rpoN cluster led to a complete lack of sensitivity of the Pu promoter to the repressive action of glucose. In that study, it was impossible to assign any phenotype to the disruption of ptsO. This was unexpected, since based on sequence similarity to the HPr and IIA domains (8, 18), the ptsN and ptsO genes had the potential to interact via a phospho transfer event. However, when expression of Pu was assayed in the absence of glucose, an unanticipated phenotype was detected (Fig. 1). Pu activity was strongly reduced compared to the activity in the wild-type strain; it was reduced to levels similar to those observed in the presence of the sugar. The effect of the mutation could be reversed by transforming the mutant strain with a plasmid expressing a wild-type copy of the gene. This indicates that the observed phenotype was indeed due to the absence of the NPr protein. By using polyclonal antibodies against purified NPr, this absence was confirmed in the mutant strain, as was effective expression of the phenotype from the plasmid-borne copy. This phenotype resembled that exhibited by a strain expressing a ptsN allele in which the phosphorylatable histidine residue at position 68 was replaced by aspartic acid, thus mimicking the phosphorylated state of the IIA^Ntr protein (8). Although this is not the only possible interpretation of the ptsO::Km phenotype, it may be that the loss of NPr could lead to accumulation of phosphorylated IIA^Ntr and to subsequent inhibition of Pu.

View larger version (35K): [in this window] [in a new window]   FIG. 1.   Effect of ptsO::Km mutation on Pu activity. P. putida MAD2-derived strains lacking ptsO were grown in M9-CAA medium with or without glucose amended with 1 mM IPTG (isopropyl--D-thiogalactopyranoside). -Galactosidase (-Gal) activity was provided by the Pu::lacZ fusion and was determined at the beginning of the stationary phase (A). Note how the low levels of the mutant strains could be changed to wild-type levels by the presence of a plasmid carrying a copy of ptsO but not by the presence of a plasmid without the insert. The levels of expression of the ptsO gene in each of the strains under each type of growth condition were monitored by using blots probed with an anti-NPr serum (B). Equal amounts of total cell protein were loaded in the lanes, so the intensities of the bands in the blots represented the relative intracellular concentrations of the ptsO products.

Double ptsO ptsN mutant had a ptsN-like phenotype. To further investigate the functional relationship between ptsO and ptsN and the role of these genes in Pu regulation, a P. putida MAD2 derivative lacking both genes was constructed. It was hypothesized that if the effect of ptsO disruption was not related to ptsN, then the levels of Pu activity should remain low despite the loss of this gene. The double mutant was assayed for Pu activity both in the presence and in the absence of glucose, and its performance was compared to the performance of the single-mutant strains (Fig. 2). The phenotype of the ptsN::xylE ptsO::Km strain was almost identical to that of the ptsN::Km mutant. Furthermore, in clear contrast to the ptsO mutant strain, the double mutant showed high levels of Pu activity that were not affected by the sugar. Introduction of a wild-type copy of ptsN into a plasmid restored the low levels of Pu activity previously observed in the ptsO::Km strain. When NPr and IIA^Ntr protein levels were monitored by immunoblotting for all of the strains tested, a significant polar effect of the ptsN::Km mutation on expression of ptsO was detected, although a small amount of NPr protein could still be found. We believe that this amount is sufficient for Pu regulation, since the ptsN mutation can be complemented by a plasmid carrying only the ptsN gene (8). The extracts obtained from the strains containing the ptsO::Km mutation showed no sign of NPr protein. We believe that these results provide strong evidence that there is a functional interdependence between ptsO and ptsN and that both genes are required for correct coupling of Pu activity to the availability of additional carbon sources.

View larger version (37K): [in this window] [in a new window]   FIG. 2.   Effect of ptsO::Km ptsN::xylE double mutation on Pu activity. P. putida MAD2-derived strains lacking ptsO or ptsN or both were grown on M9-CAA medium with or without glucose. -Galactosidase (-Gal) activity was provided by the Pu::lacZ fusion and was determined at the beginning of the stationary phase (A). The double-mutant levels were indistinguishable from the ptsN single-mutant levels. Note, however, how the low levels of the ptsO mutant strains were restored by the presence of a plasmid carrying a copy of ptsN (indicated in parentheses). The levels of expression of the ptsO and ptsN genes in each of the strains were monitored by using blots probed with an anti-NPr serum or an anti-IIANtr serum (B). The intensities of the bands in the blots represented the relative intracellular concentrations of the ptsO and ptsN products. Note that the ptsN::Km insertion had a notable polar effect on ptsO expression.

Conserved His residue at position 15 is required for ptsO activity. As both NPr and IIA^Ntr belong to the PTS family of phosphotransferases, we speculated that phosphorylation was the most likely type of interaction. Based on the homology of NPr to HPr, it can be assumed that the conserved His residue at position 15 is the residue that is the acceptor and donor of the phosphate moiety (26) (Fig. 3A). In order to determine the influence of phosphorylation on the action of the ptsO product in Pu regulation, an allele of ptsO was generated in which the His residue at position 15 was replaced by alanine. The new version of NPr could not be phosphorylated. This allele was cloned in a plasmid under the control of a Ptrc promoter, and the plasmid was transformed into the ptsO::Km strain. As shown in Fig. 3B, this plasmid was not able to restore the wild-type Pu activity phenotype. To eliminate the possibility that ptsOH15A was not properly expressed from the plasmid under the conditions of the experiment, the presence of the NPr protein in all strains was determined by immunoblotting. The results strongly support the hypothesis that phosphorylation of NPr is required for a functional interaction with IIA^Ntr. Although other possibilities (e.g., misfolding of the protein) cannot be ruled out, the inability of NPr H15A to become phosphorylated is the most likely cause of the phenotype.

View larger version (46K): [in this window] [in a new window]   FIG. 3.   Effect of the H15A substitution on NPr functionality. (A) Alignment of the region consisting of three NPr proteins containing the phosphorylation motif of the HPr family of PTS proteins. Two HPr proteins were included as a reference. The conserved His residue at position 15 (indicated by boldface type) was changed to Ala. (B) Phenotype endowed by the ptsOH15A allele. Plasmid pVLT290H15A, which expresses the ptsOH15A allele from an IPTG-inducible Ptrc promoter, and the insertless vector pVLT35 were transferred to the ptsO::Km P. putida MAD2 derivative (indicated in parentheses below the relevant genotype). The resulting exconjugants were grown on M9-CAA medium with or without glucose amended with IPTG. Note the inability of the mutant H15A allele to restore Pu regulation. The levels of expression of the ptsO gene in each of the strains under the different growth conditions were monitored by using blots probed with an anti-NPr serum (C). Equal amounts of total cell protein were loaded in the lanes, so the intensities of the bands in the blots represented the relative intracellular concentrations of the ptsO products. -Gal, -galactosidase.

Genome-wide search for EI-like enzymes. The results described above suggested that at least one other protein responsible for NPr phosphorylation participates in the signal transduction pathway that leads to Pu inhibition by glucose. HPr-like proteins are normally phosphorylated by members of the EI family of PTS proteins (26). It has been reported that the E. coli NPr protein can be phosphorylated in vitro by the EI^Ntr protein, the product of the ptsP gene (28), and (less efficiently) by the EI protein (27). Since these proteins have not been found in P. putida, we wondered whether they might be present in the P. putida genome and whether they are involved in Pu regulation. To address these questions, the complete (albeit not ordered) genome sequence of P. putida KT2440 available at TIGR website (http://www.tigr.org; updated October 2000) was obtained and scanned for the presence of proteins with an EI domain that might be involved in NPr phosphorylation. The E. coli EI^Ntr protein was used as an electronic probe with the TBLASTN program at TIGR website (http://www.tigr.org/cgi-bin/BlastSearch/blast.cgi?). Two chromosomal DNA fragments with a likelihood of encoding EI-like enzymes were found (Fig. 4A). One of these fragments, designated ptsP, codes for a 759-amino-acid protein with 43% identity and 63% similarity to the EI^Ntr E. coli homologue. Like the E. coli protein (30), this fragment has two domains: a GAF domain, which also occurs in eukaryotic phosphodiesterases and in the NifA family of bacterial transcriptional regulators (2), and the EI-like domain. The other fragment, designated mtp, codes for a 990-amino-acid multidomain protein that comprises a IIA domain, an HPr-like domain, and the EI-like domain. The closest relative of this fragment in the data banks is MTP from Rhodobacter capsulatus, with which it exhibits 38% identity and 50% similarity. The function of the R. capsulatus homolog is transport of fructose across the inner cell membrane (37). It is likely that the P. putida protein has the same function. In fact, it has been shown that in contrast to glucose, gluconate, and other sugars, fructose is transported in P. putida by a PTS-dependent mechanism (20, 34).

View larger version (29K): [in this window] [in a new window]   FIG. 4.   Effect of ptsP, mtp, and ptsP mtp mutations on Pu activity. (A) Domain structure of P. putida EINtr, encoded by ptsP, and MTP, as assigned by the InterPro server (http://www.ebi.ac.uk/interpro/scan.html). The two characteristic motifs present in EI-like enzymes, as described by PROSITE (http://www.expasy.ch/tools/scnpsite.html), are also shown. (B) P. putida MAD2-derived strains lacking ptsP or mtp or both were grown on M9-CAA medium with or without glucose. -Galacatosidase (-Gal) activity was provided by the Pu::lacZ fusion and was determined at the beginning of the stationary phase.

Mutations in ptsP or mtp or both do not affect Pu activity. The ptsP and mtp genes were disrupted as described in Materials and Methods, and the Pu activity phenotypes of the mutant strains were determined (Fig. 4B). None of the mutations had any influence on the levels of Pu activity or on the response of Pu activity to the presence of glucose. In order to eliminate the possibility that there is any functional redundancy in these genes, a double mutant that lacked both of the genes was also constructed, and no effects on Pu were observed. These results rule out the possibility that these two genes play any role in Pu regulation. The possibility that there are other genes belonging to the EI family which were not detected in these screening experiments is highly unlikely, but as the genome is not yet completely determined, other EI-like enzymes may be identified when sequencing is finished.

Role of ptsO and ptsN in Pu regulation. Based on previous results and results presented in this paper, a refined model that accounts for carbon regulation of Pu which involves both ptsO and ptsN can be proposed (Fig. 5). Based on the phenotypes observed for the strain lacking ptsO and the ptsN::xylE ptsO::Km strain, it can be deduced that NPr probably modulates IIA^Ntr activity. It has been proposed that the phosphorylated form of IIA^Ntr is responsible for Pu inhibition (8). NPr could then promote dephosphorylation of IIA^Ntr. This would explain why the His residue at position 15 of NPr is definitely required: it would act as the phospho acceptor group for IIA^Ntr dephosphorylation. It is tempting to speculate that IIA^Ntr and NPr share an independent pool of phosphate which flows from one protein to the other in response to glucose availability by an unknown mechanism. In the presence of glucose, IIA^Ntr~P would accumulate as a consequence of phosphate transfer from NPr~P. Once glucose is depleted, phosphate would be transferred back to NPr. The unphosphorylated IIA^Ntr concentration would then increase, and inhibition of Pu would disappear. This explanation is consistent with the apparent absence of any glucose-related phenotype in the mutants lacking EI-like enzymes, since these enzymes are not required for performance of the system proposed. However, it is possible that ptsN, ptsO, and ptsP are all involved in a regulatory cascade, which although not operating on Pu, regulates other genes. It is interesting that ptsN and the ptsP Rhizobium etli homologue, ptsA, have been reported to participate in regulation of the ⁵⁴-dependent promoter nifH in this bacterium (23). In this regard, it should be noted that as determined by two-dimensional gel electrophoresis, disruption of ptsN leads to changes in the levels of more than 8% of P. putida proteins (7). This hypothesis requires the existence of at least one additional regulatory component that has not been identified yet. In the ptsO mutant strain, Pu is inhibited even in the absence of glucose. This could be interpreted as an effect that is due to accumulation of IIA^Ntr~P. This observation cannot be explained without hypothesizing that there is a source of high-energy phosphate able to phosphorylate IIA^Ntr other than ptsO. One plausible candidate is MTP, which, in addition to an EI domain, harbors an HPr domain with the potential to phosphorylate IIA (Fig. 4). This is in contrast to the EI^Ntr protein, which has only one EI domain, which makes a possible role in IIA^Ntr phosphorylation unlikely. Although other models might be proposed, in the absence of more experimental data the simpler interpretation should be favored (Fig. 5). In any event, the results reported here reveal the existence of an intricate regulatory mechanism involving ptsO, ptsN, and another factor(s) that is not known. Experiments to obtain more information concerning the system are under way.

View larger version (7K): [in this window] [in a new window]   FIG. 5.   Plausible model for the role of IIANtr and NPr in Pu regulation. The model shows the reversible flow of the phosphoryl group between IIANtr and NPr. In the presence of glucose, phosphorylation of IIANtr would be favored. Accumulation of IIANtr~P would in turn lead to Pu inhibition. In the absence of glucose, the equilibrium would be reestablished, and Pu inhibition would disappear. For clarity, the alternative IIANtr phospho donor that would be required in the absence of ptsO is not shown. See the text for further discussion of the model.