ABSTRACT
Efficient gene regulation of metabolic pathways implies that the profile of molecules inducing the pathway matches that of the molecules that are metabolized. Gratuitous induction, a well-known phenomenon in catabolic pathways, is the consequence of differences in the substrate and inducer profiles. This phenomenon is particularly evident in pathways for biodegradation of organic contaminants that can be induced by a variety of molecules similar to the real substrates. Analysis of the regulation of tetralin biodegradation genes in mutant strains with mutations that affect each component of the initial dioxygenase enzymatic complex indicated that the response of the regulatory system to potential inducers is altered differently depending on the mutated component. Based on the expression phenotypes of a number of single or double mutants, we propose a model that represents an unprecedented way of communication between a catabolic pathway and its regulatory system to prevent efficient induction by a molecule that is not a real substrate. This communication allows a better fit of the substrate and inducer profiles, thus minimizing gratuitous induction, without a requirement for optimal coevolution to match the specificity of catabolic enzymes and their regulatory systems. Modulation of the regulatory system in this way not only provides a more appropriate response to potential inducers recognized by the regulatory system but also may properly adjust the levels of gene expression to the substrate availability.
The extraordinary versatility of bacteria that allows them to use different organic molecules as carbon and energy sources is indicated by their ability to oxidize a plethora of different xenobiotic compounds recently created by chemical synthesis and discharged into the environment through industrial and urban activities. Bacteria have been able to assemble a multitude of catabolic pathways with the capacity to transform very different molecules, some of which are synthesized and therefore have been released into the environment in just the last few decades. Many of these catabolic pathways, most of which have recognizable common evolutionary origins, have been found in very different microorganisms and characterized (11, 37). This implies that there has not been evolution of just one enzyme but there has been coevolution of a whole set of enzymes, which together allow the bacteria to grow on a particular chemical.
The success of a bacterium in a contaminated niche obviously depends on the ability of its enzymes to use the contaminant in that niche. Its success also depends on an efficient regulatory system that ensures expression of the catabolic genes only under the appropriate environmental conditions (7, 8, 35), which provides an advantage during competition with other residents. Thus, evolution of regulatory proteins and evolution of regulated promoters are key elements in the successful establishment of a catabolic pathway. A large variety of regulatory systems, which control the expression of catabolic genes through proteins belonging to different families and through different mechanisms, have also been characterized, meaning that evolution of catabolic genes and evolution of regulatory systems are not always parallel (for a review, see reference 37). Different regulatory systems seem to be incorporated into newly assembled pathways, since it is not unusual to find similar catabolic pathways with entirely different regulatory systems (5, 7).
Regulatory proteins of biodegradation pathways are quite promiscuous, responding to a number of similar molecules (35). For efficient regulation, it is crucial that the range of inducers to which a regulatory system responds and the range of substrates that the catabolic pathway can transform are the same. Otherwise, expression of the catabolic pathway may result in gratuitous induction by a nonmetabolizable molecule, a phenomenon often found not only in organic contaminant biodegradation pathways but also in catabolic pathways for sugars and other carbon sources (for instance, gratuitous induction of the lac operon by isopropyl-β-d-thiogalactopyranoside [IPTG] [14, 22]). Alternatively, the pathway may not be induced in the presence of a suitable substrate. Expression of some biodegradation genes is not induced in response to the substrate but is induced in response to some intermediate in the catabolism of the substrate (3, 25, 31, 34). This represents a simple way of preventing uncoordinated induction. However, an efficient response implies that the basal levels of expression have to be high enough to allow substantial degradation of the substrate so that a sufficient concentration of the inducer intermediate accumulates. In addition, gratuitous induction is not fully prevented since central intermediates, such as cis,cis-muconate, may be produced from different substrates by peripheral catabolic pathways.
Tetralin (1,2,3,4-tetrahydronaphthalene) is found at low concentrations in different crude oils, and it is also industrially produced and used as an organic solvent. Biodegradation of tetralin has been characterized most extensively in Sphingomonas macrogolitabida strain TFA. Genes required for tetralin biodegradation (Fig. 1) have been sequenced, and the functions of their gene products have been characterized to elucidate the biodegradation pathway (1, 2, 16-18, 27). As observed for other aromatic compounds, degradation of tetralin is initiated by dioxygenation of the aromatic ring, a reaction requiring oxygen and an external electron supply, which is catalyzed by the tetralin dioxygenase enzymatic complex encoded by four thnA genes (27) (Fig. 1).
(A) Arrangement of the two divergent operons containing genes involved in catabolism of tetralin. B, thnB (encoding a dehydrogenase); D, thnD (encoding a hydrolase); E, thnE (encoding a hydratase); F, thnF (encoding an aldolase); A1, thnA1 (encoding a dioxygenase α subunit); A2, thnA2 (encoding a dioxygenase β subunit); C, thnC (encoding an extradiol dioxygenase); A3, thnA3 (encoding a ferredoxin); A4, thnA4 (encoding a ferredoxin reductase); R, thnR (encoding a LysR-type regulator); Y, thnY (encoding a coregulator). (B) Schematic diagram of the initial dioxygenation of tetralin and the electron transport system.
Expression of thn operons is regulated at the transcriptional level. Transcription is induced in the presence of tetralin but is prevented if other preferential carbon sources are available (24), indicating that there is complex regulation that integrates information concerning the nutritional status of the bacteria. The thnR regulatory gene, which is cotranscribed together with other thn structural genes (Fig. 1), encodes a LysR-type activator that is essential for expression of thn genes. Unlike the expression of most of the systems controlled by LysR-type regulators, the expression of thn genes requires an additional factor, encoded by thnY, which presumably controls the function of the transcriptional activator (24).
In this study, expression of thn genes in response to different chemicals was analyzed using a collection of Thn mutants. The results indicate that the regulation of expression of thn genes includes an additional modulation system, not previously described for catabolic pathways, which prevents gratuitous induction by integrating into the regulatory system information concerning the ability of the pathway to metabolize the potential inducer molecules.
MATERIALS AND METHODS
Media and growth conditions. Escherichia coli strains were routinely grown in Luria-Beltrani medium at 37°C. S. macrogolitabida strains (Table 1) were grown at 30°C in MML rich medium (1) or MM medium (9) with β-hydroxybutyrate (βHB) as the carbon source.
Bacterial strains, plasmids, and primers
Induction assays.Strains harboring lacZ fusions integrated into the chromosome were grown at 30°C in mineral medium containing 40 mM βHB as the only carbon and energy source to the exponential phase (optical density at 600 nm, 0.8 to 1.0). Then cells were washed to remove the carbon source and diluted to obtain a final optical density of about 0.1 in MM medium supplemented with 8 mM βHB in the absence or presence of the inducer compound. Cultures were grown at 30°C for 24 h, and β-galactosidase activity was assayed as described by Miller (26). Tetralin, naphthalene, cyclohexane, cis-decalin, trans-decalin, benzene, and toluene were supplied in the gas phase. Soluble compounds were added at the following concentrations: 0.3 mM octanol, 0.075 mM decanol, 2 mM salicylate, 2 mM benzoate, 2 mM toluate, 0.5 mM tretrahydro-2-naphthol, 0.3 mM indole, and 0.5 mM biphenyl.
Construction of strains and plasmids.An in-frame deletion of thnA3 was constructed by overlapping PCR (19) using pIZ612 (18) as the template and two set of primers, primers T3 and ThnA3-1 and primers ThnA3-2 and A282G, to obtain 1.2- and 0.2-kb fragments, respectively. These fragments were used as templates with primers T3 and A282G to obtain a 1.4-kb fragment that included an in-frame deletion of thnA3 encoding a 16-amino-acid protein. Replacement of the wild-type thnA3 gene in pIZ612 by the mutated gene resulted in plasmid pIZ1031. Strain T1031 was obtained by marker exchange of the thnA3::KIXX copy present in strain T661 (18) and the in-frame thnA3 deletion of pIZ1031.
A single amino acid substitution, Asp221Ala, in ThnA1 was generated by PCR using pIZ643 (18) as the template and primers T3 plus and ThnA1-1 introducing a NarI site overlapping the mutation. Replacement of the wild-type gene in plasmid pIZ643 by the PCR product, which contained the thnA1 D221A mutation, resulted in plasmid pIZ1041. Replacement of the thnA1::KIXX insertion in strain T659 (18) by the mutant copy in pIZ1041 by marker exchange resulted in strain T1034. A single amino acid substitution, Asp372Ala, in ThnA1 was generated by overlapping PCR using primers T3 plus and ThnA1-2 to obtain a 180-bp fragment and primers ThnA1-3 and ThnA1-4 to obtain a 500-bp fragment. Both DNA fragments were used as templates in a new amplification reaction, using primers T3 plus and ThnA1-4, that resulted in a 680-bp fragment containing the thnA1 D372A mutation and a NaeI restriction site overlapping this mutation. The approach described above for T1034 was used to clone the thnA1 D372A mutation in pIZ1042 and to introduce it into strain T659 in order to obtain strain T1035.
Strain T1033, a thnA1::KIXX ΔthnA3 double mutant, was obtained by marker exchange replacement of the wild-type thnA1 gene of strain T1031 (ΔthnA3) with the thnA1::KIXX copy of plasmid pIZ659 (18). Strain T1037, a thnY::KIXX ΔthnA3 double mutant, was constructed similarly by marker exchange replacement of the wild-type thnY gene of strain T1031 (ΔthnA3) with the thnY::KIXX copy of plasmid pIZ669 (24).
To monitor the levels of expression of thn genes in all strains, pIZ1002 containing a thnC-lacZ gene fusion (24) was introduced by conjugation and integrated by homologous recombination. Streptomycin- and ampicillin-resistant strains carrying the thnC-lacZ fusion were designated T1031-1002, T1033-1002, T1034-1002, T1035-1002, and T1037-1002.
Plasmid pIZ1033 was obtained by insertion of a 0.57-kb EcoRI DNA fragment containing thnA3 from pIZ633 into the broad-host-range expression vector pIZ1016 (24), resulting in thnA3 gene expression from the tac promoter. For complementation analysis, plasmid pIZ1033 was transferred into recipient strains T1031-1002 and T1033-1002 by triparental mating. Transconjugant colonies were selected on MML agar plates by using resistance to gentamicin, ampicillin, and streptomycin.
The genetic arrangements of all strains constructed by marker exchange were confirmed by Southern blot hybridization.
RESULTS
Inducer profile for thn operon expression.Transcription of the thn operons was induced in the presence of tetralin, the substrate of the biodegradation pathway encoded by the thn genes (24). A more complete analysis of the specificity of induction was carried out by examining expression of thnB-lacZ and thnC-lacZ translational fusions in the presence of molecules that may share some structural or functional features with tetralin. Data in Table 2 obtained by using thnC-lacZ (data obtained by using thnB-lacZ were similar [not shown]) indicate that tetralin, biphenyl, and naphthalene induced similar high levels of expression of thn genes. In addition, molecules containing aromatic or alicyclic rings also induced different levels of thn expression, although they were significantly less efficient inducers (the levels were 19 to 55% of the levels obtained with tetralin). On the other hand, monoaromatic molecules containing hydroxyl or carboxyl groups did not induce the thn genes at all.
Inducer profile for thn operons
All molecules that are able to induce thn gene expression to some extent are hydrophobic; therefore, they intercalate into the membrane, resulting in membrane stress, which is toxic to the cell. However, the hydrophobicity and toxicity of the compounds containing hydroxyl or carboxyl groups are much lower. A comparison of the inducing capabilities of toluene, benzoate, and toluate suggested that membrane stress, to which the cells respond by modifying the cis/trans ratio of their membrane fatty acids (32, 36), might be an inducing factor. However, neither octanol nor decanol, which are not structurally related to tetralin but are known to cause membrane stress and a strong cell response (21, 36), induced thn expression, suggesting that induction of thn genes is not just a stress response but implies that there is molecular recognition of the inducer.
Catabolism of tetralin is not required for thn operon induction.Previous expression analyses were carried out with the wild-type strain and with a number of mutants with mutations that affected different steps of the biodegradation pathway (16, 18, 24, 27). These mutants included a thnA3::KIXX mutant, which lacks the ferredoxin ThnA3, a transporter of electrons to the active center of the dioxygenase enzyme, which catalyzes the initial dioxygenation of tetralin, yielding a cis-dihydrodiol derivative. The expression of a thnC-lacZ fusion in these mutants was not limited at all, indicating that metabolism of tetralin is not necessary to induce the thn genes (24).
However, a subsequent expression analysis using KIXX insertion mutants with mutations in the genes encoding each of the components of the ring-hydroxylating dioxygenase enzymatic complex indicated that the expression phenotypes of these mutants were not homogeneous. As shown in Fig. 2, mutants lacking either the ferredoxin ThnA3 or the ferredoxin reductase ThnA4 induced levels of expression of the thnC-lacZ fusion that were even higher than those in the wild type, while, surprisingly, mutants lacking either the α (ThnA1) or β (ThnA2) subunit of the dioxygenase enzyme exhibited levels of expression that were reduced fivefold. A simple interpretation of these data is that the real inducer is not tetralin itself but the catabolic product of its dioxygenation and that the electrons required for the dioxygenase could be provided, at least to some extent, by an alternative nonspecific electron transport system in the mutants lacking ThnA3 or ThnA4. However, although ThnA4 may be functionally replaced to some extent by unknown ferredoxin reductases (27), the mutant lacking ThnA3 is unable to grow on tetralin as the sole carbon and energy source and has no detectable dioxygenase activity. Additionally, it has been shown by cloning different combinations of thn genes that ThnA3 is essential for detection of tetralin dioxygenase activity in three different hosts, S. macrogolitabida, E. coli, and Pseudomonas putida (27). These data provide strong evidence that the cis-dihydrodiol derivative is not the real inducer. Also, tetralin can induce thn operon expression when strain TFA grows anaerobically using nitrate as an alternative electron acceptor (F. Reyes-Ramírez, unpublished results), which further supports the view that tetralin transformation is not required for induction since molecular dioxygen is required as a substrate in the initial dioxygenation reaction.
Tetralin-induced expression of a thnC-lacZ fusion in KIXX insertion mutants lacking the components of the tetralin dioxygenase enzymatic complex, including thnA1 (T659), thnA2 (T673), thnA3 (T661), and thnA4 (T663). wt, wild type.
Tetralin dioxygenase activity per se, not the reaction product, modulates thn gene induction.The results obtained using the ThnA1 and ThnA2 mutants and the results obtained using the ThnA3 and ThnA4 mutants apparently conflict. However, there may be different explanations for the heterogeneous expression phenotype.
Previous complementation analysis indicated that the regulatory gene thnR belongs to the same transcriptional unit as the thnCA3A4 genes, which are located upstream (18, 24) (Fig. 1). One possibility that would reconcile all the results is that the expression phenotype of the thnA3::KIXX and thnA4::KIXX mutants was not due to the lack of ferredoxin but to aberrant expression of thnR and/or thnY from a promoter transcribing downstream sequences known to be present in the KIXX cassette. An in-frame deletion of thnA3 was constructed and inserted into the chromosome of the thnA3::KIXX mutant by marker exchange. The deletion mutant containing the thnC-lacZ gene fusion was not able to grow using tetralin as the sole carbon and energy source and, as shown in Fig. 3, exhibited the same expression phenotype as the thnA3::KIXX mutant (Fig. 2). Additionally, complementation of ΔthnA3 with plasmid pIZ1033, expressing thnA3 from the tac promoter, restored the levels of expression to the levels of the wild type, which clearly indicated that it is indeed the lack of ferredoxin that causes the expression phenotype of the ThnA3 mutants.
Tetralin-induced expression of a thnC-lacZ fusion in strain TFA, thnA1 D221A (T1034) and thnA1 D372A (T1035) mutant strains, a ΔthnA3 mutant (T1031), and a thnA1::KIXX ΔthnA3 double mutant (T1033) and complementation with plasmid pIZ1033 containing thnA3 expressed from the tac promoter. wt, wild type.
Tetralin is very hydrophobic, and, therefore, most of it is associated with the membranes, which can make binding the molecule a difficult task for a soluble cytosolic regulator. It is obvious that tetralin is bound by tetralin dioxygenase since this enzyme transforms tetralin. A possible explanation for the conflicting results is that the tetrameric dioxygenase (α2β2) somehow facilitated the binding of tetralin by the activator ThnR. Insertion mutations in thnA1 or thnA2 should reduce the levels of thn expression, while insertions in thnA3 or thnA4 should not. In fact, the expression in ThnA3 mutants could be even higher than the expression in the wild type since tetralin bound to the dioxygenase complex could not be transformed, thus making all the tetralin available for ThnR. To test this hypothesis, mutations resulting in two single-amino-acid substitutions, Asp221Ala and Asp372Ala, in thnA1 were independently constructed and inserted into the chromosome by marker exchange of the thnA1::KIXX insertion. As determined by sequence comparison, these substitutions are equivalent to previously constructed Asp205Ala and Asp362Ala substitutions in the protein encoded by nahAa in Pseudomonas sp., which do not significantly alter the crystal structure of a similar naphthalene dioxygenase but result in fully inactive enzymes by apparently preventing electron transfer from the ferredoxin to the catalytic center and, subsequently, to the substrate (29, 30). None of the thnA1 mutants containing these mutations was able to grow using tetralin as the sole source of carbon and energy, which confirms that these residues are important for dioxygenase activity. As shown in Fig. 3, the expression of thnC-lacZ in these mutants was also reduced compared to that in the wild type, suggesting that what is required for normal levels of expression is not the structure of the tetralin dioxygenase complex but, indeed, its activity.
Expression of thn genes in a thnA1::KIXX ΔthnA3 double mutant was also tested. After insertion of the thnC-lacZ gene fusion into the chromosome of the double mutant, expression analysis (Fig. 3) indicated that, interestingly, the expression phenotype of the double mutant was like that of the single ThnA3 mutants. In turn, complementation of the double mutant with thnA3 resulted in levels of expression that were similar to those of the ThnA1 mutants. Together, these data clearly indicate that components of the tetralin dioxygenase enzymatic complex modulate in opposite directions the efficiency of tetralin as an inducer. The epistatic effect of ΔthnA3 on the thnA1::KIXX mutation suggests that the positive effect of the tetralin dioxygenase is fully dispensable in the absence of the ferredoxin ThnA3. More precisely, it suggests that ThnA3 has a negative effect on the inducing capability of tetralin, which may be counteracted, at least partially, by the tetralin dioxygenase itself.
Finally, to determine whether thn gene expression in the absence of ferredoxin required the positive regulator ThnY, which may modulate the function of the activator ThnR, a double thnY::KIXX ΔthnA3 mutant was constructed using the same approach. The lack of thn gene induction in the double mutant (not shown) clearly indicated that ThnY is required for thn expression even when the negative effect of ThnA3 is removed.
Tetralin dioxygenase restricts the thn gene-inducing abilities of some molecules.The effects of mutations in genes coding for different components of the dioxygenase complex were also examined with six other inducer molecules (naphthalene, cis-decalin, trans-decalin, cyclohexane, benzene, and indole) in order to determine whether the effects were similar in all cases. Figure 4 shows the results obtained with two of these molecules, which represented low-efficiency inducers. In general, the phenotypes of the mutants and the complemented strains were similar regardless of the inducer. The ThnA1 mutation resulted in reduced expression, although the levels of reduction were significantly different depending on the inducer (ranging from 1.9-fold for cis-decalin to 6.2-fold for indole; 5-fold for tetralin). On the other hand, thn gene expression was clearly increased in the ΔthnA3 mutant when any of the inducers was used (Fig. 4). In this mutant, the levels of the effect of ΔthnA3 on induction by inefficient inducers (range, 6.7-fold for benzene to 3.8-fold for indole) were always higher than the levels observed when the efficient inducers tetralin and naphthalene were used (2.5-fold and 2-fold, respectively). Consequently, the molecules that were poor inducers in the wild-type strain became better inducers in the ΔthnA3 mutant as compared to tetralin (Fig. 5), which indicates that thnA3 provides a discriminatory capacity to the regulatory system.
Expression of a thnC-lacZ fusion induced by cis-decalin (A) or indole (B) in strain TFA and thnA1::KIXX (T659), ΔthnA3 (T1031), and thnA1::KIXX ΔthnA3 (T1033) mutant strains and complementation with plasmid pIZ1033 containing thnA3 expressed from the tac promoter. wt, wild type.
Inducing efficiencies of different molecules in wild-type strain TFA (solid bars) and ΔthnA3 mutant (open bars) compared to the efficiency of the substrate tetralin. The absolute tetralin-induced levels were 4,970 and 12,480 Miller units for the wild-type and ΔthnA3 strains, respectively.
The basal levels of expression and inducing capacities of molecules considered noninducers in the wild-type strain (Table 2) were also determined in mutant ΔthnA3. None of these molecules induced thn operons to a significant extent in the ΔthnA3 strain, suggesting that recognition of the inducer molecule by the regulator is still required for high levels of thn gene expression even in this mutant (data not shown).
DISCUSSION
The expression studies of the divergent operons encoding the enzymes responsible for the catabolism of tetralin indicate that thn gene expression is inducible by the substrate tetralin and also, to some extent, by a number of structurally related molecules, as observed in many other catabolic pathways. Analysis of the inducer profile clearly suggested that the regulatory system, composed of at least the activator ThnR and the coactivator ThnY, recognizes some structural features of the inducer molecules rather than responding to other more general signals, such as membrane stress caused by hydrophobic compounds. Apparently, as in other systems, the recognition is not very specific since a number different molecules can induce thn operon expression, although they do so inefficiently (35).
A previous analysis using mutants blocked at different steps of the catabolic pathway (24) suggested that tetralin itself, and not an intermediate of the catabolic pathway, induced thn expression. On the other hand, strains with mutations in the genes encoding the subunits of the dioxygenase enzyme, which catalyzes dioxygenation of tetralin, exhibited reduced levels of expression (Fig. 2), which suggested that tetralin had to be transformed into its cis-dihydrodiol derivative to efficiently induce. However, the following findings allowed us to conclude that tetralin itself is recognized as an inducer by the regulatory system: (i) the level of tetralin-induced expression was high in the mutants lacking the ferredoxin ThnA3, which is essential for tetralin dioxygenase activity (27) and, therefore, for the initial tetralin transformation; (ii) the same expression phenotype was observed for the double mutant lacking the ferredoxin and the α subunit of the dioxygenase enzyme (Fig. 2 and 3); and (iii) the thn genes were induced even in the absence of oxygen, one of the substrates of the tetralin dioxygenation reaction (Fig. 1) (not shown).
Still, the strains with mutations that affected components of the dioxygenase enzymatic complex had an altered expression phenotype. Mutants with insertion and point mutations that resulted in an inactive dioxygenase enzyme exhibited reduced levels of expression, which suggested that tetralin dioxygenase activity facilitates high levels of expression. On the other hand, mutations affecting (ThnA4 mutant) or completely blocking (ThnA3 insertion or deletion mutants) the electron transport to the dioxygenase had the opposite effect, suggesting that the electron transport components prevent full expression of the thn genes. The expression phenotype of the thnA1::KIXX ΔthnA3 double mutant provided key information indicating that in the absence of the ferredoxin ThnA3, thn expression is fully induced independent of whether the dioxygenase is active.
Together, these results allowed us to propose a model, shown in Fig. 6, which establishes that the ferredoxin ThnA3, in its reduced form, prevents transcriptional activation by the regulatory system. This model is conceptually simple, since it only ascribes to the reduced form of ferredoxin an additional function as a negative modulator and provides a full explanation for the mutant expression phenotypes. Mutants lacking tetralin dioxygenase or having an inactive enzyme do not transfer the electrons from the ferredoxin to the substrate; therefore, ThnA3 remains in its reduced form, negatively modulating thn gene transcription. On the other hand, mutations that reduce electron transport to the ferredoxin or eliminate the ferredoxin itself prevent negative modulation by reduced ferredoxin, resulting in full expression of the thn genes.
Proposed model for communication between the catabolic pathway and its regulatory system. The upper section represents the reaction catalyzed by the tetralin dioxygenase enzymatic complex, which connects with the regulatory system in the lower section through reduced ferredoxin.
The model is not just an ad hoc explanation of the mutant expression phenotypes. It predicts that high levels of thn expression can occur in the wild-type strain only if there is an available substrate that is efficiently transformed by the dioxygenase, which acts as an electron sink that prevents accumulation of reduced ferredoxin. This view is supported by the results shown in Fig. 4 and 5, which indicate that molecules shown to be relatively inefficient inducers in the wild-type strain are able to induce very high levels of expression, similar to those induced by tetralin, when they are tested with the ΔthnA3 mutant strain.
Of course, the model still assigns the ability to discriminate between different inducer molecules to the regulatory system since thn expression is not constitutive in the ΔthnA3 mutant but higher levels of thn expression are induced by a broader range of molecules. According to this model, the efficiency of a molecule as an inducer of thn genes has two components, whose relative values may vary in different molecules. The first component, which is isolated from the second component in the ΔthnA3 mutant, is the capacity to be recognized as an inducer by the regulatory system; the second component is the capacity to be a substrate of the dioxygenase. This notion was confirmed by a comparison of the effects of ThnA mutations on the levels of expression induced by cis-decalin and by indole, two very different molecules that are similarly inefficient inducers in the wild-type strain (1,703 and 1,701 Miller units for cis-decalin and indole, respectively). Indole is a well-known substrate for dioxygenases and monooxygenases of molecules containing aromatic rings (4, 28), and in fact, it has been used in colorimetric oxygenase enzymatic assays, including tetralin dioxygenase assays (27, 33), since its dioxygenation product is spontaneously transformed into the textile dye indigo. On the other hand, cis-decalin is not expected to be a good substrate for tetralin dioxygenase since this type of enzyme hydroxylates aromatic rings, and cis-decalin is composed of two alicyclic rings (Table 2). As predicted, the ThnA1 mutation had a slight effect (1.9-fold reduction) on induction by cis-decalin, since it is a poor substrate, and a stronger effect (6.2-fold reduction) on induction by indole. On the other hand, the effect of the ΔthnA3 mutation on induction by cis-decalin was greater (6-fold increase) than the effect of the ΔthnA3 mutation on induction by indole (3.9-fold increase), suggesting that cis-decalin is more efficiently recognized as an inducer by the regulatory system.
Also, a comparison of the tetralin-induced levels of expression in the wild type and ThnA3 mutants suggested that even in the presence of an appropriate substrate, the expression potential is not fully developed in the wild type. Possibly, modulation of the regulatory system not only provides a more specific response to potential inducers recognized by the regulatory system but also allows adjustment of the levels of gene expression to the substrate availability.
Many biodegradation pathways of organic contaminants have simple regulatory systems based on a regulatory protein whose activity is directly controlled by a small molecule normally acting as an inducer. Very frequently, the regulator is an activator of the LysR family (37), and in most instances it is the only regulatory factor identified. However, in some catabolic pathways more complex regulatory systems have evolved, which allow the pathway to be expressed in response to other environmental signals not directly sensed by the regulator (6). One such complex regulatory system seems to have evolved to control biodegradation of tetralin in S. macrogolitabida, since it is one of the few examples of systems controlled by LysR-type regulators that require additional regulatory factors (10, 12, 13, 20, 23, 24), it integrates additional signals of the physiological status of the cell, such as catabolite repression (24), and it also modulates the response to different potential inducers according to their suitability as substrates of the pathway (this work).
In summary, the proposed model describes an unprecedented method of communication between the activity of the first enzyme of a catabolic pathway and the regulatory system controlling expression of the genes encoding the catabolic enzymes, which facilitates appropriate gene induction by transducing to the regulatory system information concerning whether a potential inducer can be metabolized by the enzymes of the pathway. In principle, this information may be transduced directly to ThnR, although it is tempting to speculate that modulation may be exerted through ThnY since this molecule is an essential coactivator presumably controlling the function of the ThnR activator.
ACKNOWLEDGMENTS
This work was supported by Spanish Comisión Interministerial de Ciencia y Tecnología grant BIO2005-03094, by the Andalusian Autonomic Government grant Proyecto de Excelencia CVI-131, and by fellowships from the Spanish Ministerio de Educación to O.M.-P. and A.L.-S.
FOOTNOTES
- Received 11 January 2007.
- Accepted 1 March 2007.
- Copyright © 2007 American Society for Microbiology
