The Crc protein is a global regulator that controls the hierarchical assimilation of carbon sources in Pseudomonads by inhibiting expression of several catabolic pathways. Crc does not bind DNA and its mechanism of action has remained elusive. Among other genes, Crc inhibits expression of alkS, the transcriptional activator of the Pseudomonas putida OCT plasmid alkane degradation pathway. AlkS activates expression of its own gene. In the presence of saturating AlkS levels, translational fusions of alkS to the lacZ reporter gene were responsive to Crc, but transcriptional fusions were not. In translational fusions, the first 33 nt of alkS mRNA, which includes up to position +3 relative to the translation start site, were sufficient to confer an efficient response to Crc. In vitro, purified Crc could bind specifically to an alkS mRNA fragment spanning positions +1 to +43, comprising the translation initiation region. We have previously shown that Crc has little effect on the stability of alkS mRNA. We conclude that Crc modulates AlkS levels by binding to the translation initiation region of alkS mRNA, thereby inhibiting translation. Because AlkS is an unstable protein present in limiting amounts, reducing its levels leads to decreased expression of all genes in the pathway.
Free-living bacteria containing large genomes are usually metabolically very versatile, and are endowed with sophisticated regulatory networks that co-ordinate expression of different sets of genes that help coping with environmental changes. These networks are formed by a number of sensors and regulatory proteins that respond to external (environmental) and internal (physiological) signals, integrate them, and derive responses that end up modifying the expression of large sets of genes. This has an important effect on expression of the bacterial catabolic pathways, which determine the metabolic versatility and the ability of the bacteria to colonize different habitats (Cases and de Lorenzo, 2005).
Pseudomonads are ubiquitous Gram-negative bacteria, metabolically very versatile, and adapted to thrive in diverse habitats. Some species are important pathogens, like Pseudomonas aeruginosa for animals and humans (Mahajan-Miklos et al., 2000), or Pseudomonas syringae for plants (Hirano and Upper, 2000). Other species, like Pseudomonas fluorescens (Lugtenberg and Dekkers, 1999) or Pseudomonas putida (Martins dos Santos et al., 2004), are not pathogenic and can even live in the rhizosphere as plant commensals, protecting plants from several diseases. A key aspect of Pseudomonads lifestyle is their ability to use a wide range of compounds as carbon sources. Expression of the genes of the catabolic pathways involved in the assimilation of these compounds is usually tightly regulated. A specific regulator assures that the pathway genes are expressed only in the presence of the compound to be assimilated. Frequently, superimposed to this specific regulation there is a complex control, mediated by global regulators, which assures that the pathways are induced only under appropriate physiological or environmental conditions. This global control can, for example, co-ordinate the induction of several pathways to assure a hierarchical assimilation of individual carbon sources when cells are faced to a mixture of potentially assimilable compounds, a process termed catabolite repression. The global regulation networks involved in catabolite repression are not well characterized, in part due to their complexity and also because they differ significantly from those characterized in other model bacteria (reviewed in Shingler, 2003; Rojo and Dinamarca, 2004; Cases and de Lorenzo, 2005).
In Pseudomonads, the Crc protein (catabolite repression control) is an important and unique global regulator. Crc is a master regulator of carbon metabolism. It is involved in the catabolite repression generated by succinate, or by amino acids, on the expression of genes involved in the metabolism of several compounds. Crc affects the expression of genes involved in the assimilation of sugars and nitrogenated compounds, both in P. aeruginosa (Wolff et al., 1991; Collier et al., 1996; MacGregor et al., 1996) and in P. putida (Hester et al., 2000a,b). Crc also modulates a number of pathways for the assimilation of linear and aromatic hydrocarbons (Yuste and Rojo, 2001; Aranda-Olmedo et al., 2005), as well as the expression of the homogentisate, catechol and protocatechuate pathways (Morales et al., 2004). In addition, at least in P. aeruginosa, the lack of Crc affects biofilm development (O'Toole et al., 2000). Crc levels are regulated and are higher under conditions imposing a strong catabolite repression over the catabolic pathways it regulates (Ruiz-Manzano et al., 2005). It is at present unknown how Crc responds to metabolic signals, and what these signals could be.
Crc ultimately affects the expression of target genes, although its molecular mechanism is unclear. Purified Crc does not bind DNA, suggesting that it is not a classical DNA binding repressor (Collier et al., 1996; MacGregor et al., 1996; Hester et al., 2000b). Its precise target and mode of action is unknown, although it has been proposed to act post-transcriptionally (Hester et al., 2000b; Yuste and Rojo, 2001). Crc has little influence on the stability of the mRNAs it regulates (Yuste and Rojo, 2001). This suggests that Crc may inhibit translation, but direct evidence supporting this hypothesis has not been reported.
In this work we have analysed how Crc modulates gene expression using as model the alkane degradation pathway encoded in the P. putida OCT plasmid (van Beilen et al., 2001). The genes of this pathway are grouped in two clusters, alkBFGHJKL and alkST (see Fig. 1). The alkBFGHJKL operon is transcribed from a promoter, named PalkB, whose expression requires the transcriptional activator AlkS and the presence of alkanes (Panke et al., 1999). In the absence of alkanes, the alkST genes are expressed at low levels from promoter PalkS1 (Canosa et al., 1999; 2000). When alkanes are present, the AlkS regulator represses PalkS1 and activates promoter PalkS2, located 38 nt downstream from PalkS1 and which provides high expression of the alkST genes (Canosa et al., 2000). Therefore, the pathway is controlled by a positive feedback mechanism governed by AlkS. In addition, activation of promoters PalkB and PalkS2 by AlkS is negatively modulated by a dominant global control when cells grow in a complete medium (Yuste et al., 1998; Staijen et al., 1999; Canosa et al., 2000). This global control depends on the additive effects of Crc (Yuste and Rojo, 2001) and of the cytochrome o ubiquinol oxidase (Cyo), a component of the electron transport chain (Dinamarca et al., 2002; 2003). Repression is particularly strong during exponential growth, but rapidly disappears when cells enter into stationary phase (Yuste et al., 1998). At least in a complete medium, the inhibition process generates a strong decrease in the levels of the AlkS transcriptional activator, an unstable protein present in the cell in limiting amounts even under inducing conditions (Yuste and Rojo, 2001). By keeping AlkS levels below those required for maximal induction of the pathway, expression of the alkST and alkBFGHJKL operons can be down-modulated in a simple and co-ordinated way (see Fig. 1). Because it activates expression of its own gene, modulation of AlkS levels could be achieved by limiting either alkS transcription from promoter PalkS2, or translation of the alkS mRNA. Both alternatives would lead to the same final result. In addition, although Crc decreases the activity of promoter PalkB (Yuste and Rojo, 2001), it has not been demonstrated whether it does so directly (acting on transcription initiation from PalkB) or indirectly (limiting the levels of AlkS below those required for full induction of PalkB). It this work we demonstrate that Crc acts by inhibiting alkS translation and that the effect on PalkB is indirect. We have localized the target for Crc and propose a model explaining how this protein may work.
Crc inhibits translation of the alkS mRNA
Transcriptional and translational fusions of alkS to the lacZ reporter gene were used to discern whether Crc affects AlkS protein levels by inhibiting either transcription or translation of alkS. The transcriptional fusion has been described before (Canosa et al., 2000), and includes the AlkS binding site upstream of PalkS2, and downstream sequences just up to the Shine-Dalgarno (SD) sequence, that was substituted by the lacZ SD. Translational fusions were constructed that include the alkS SD and downstream sequences up to the 1st, 5th or 14th codon of alkS, followed by a lacZ reporter gene devoid of its own SD (see Fig. 2). The fusions were independently introduced into the chromosome of P. putida PPSH1, which contains the alkS gene expressed from the heterologous and strong Ptrc promoter. This assures that AlkS levels do not fluctuate depending on growth conditions, and are sufficient to saturate the AlkS binding site of the PalkS2 promoter (Yuste and Rojo, 2001). Therefore, in this genetic background, and in the presence of an AlkS effector such as the non-metabolizable analogue dicyclopropylketone (DCPK), the levels of β-galactosidase generated from the different fusions reflects the strength of the global control acting on PalkS2, or on the mRNA arising from this promoter. It should be noted that, in the absence of DCPK, β-galactosidase activity was negligible in all the transcriptional and translational fusions analysed, showing that lacZ expression from promoter PalkS2 requires activation by the AlkS regulator.
During exponential growth in Luria–Bertani (LB) medium, expression of lacZ from promoter PalkS2 was efficient in the strain containing the transcriptional PalkS2-lacZ fusion (strain PSH2; see Fig. 3A), showing no signs of inhibition by global control. In contrast, lacZ expression was severely inhibited in the strain harbouring the translational alkS'-'lacZ fusion containing the first 14th codons of alkS (strain PSH50; Fig. 3A), irrespective of the presence of the inducer in the medium. Expression increased moderately when cells entered into the stationary phase of growth, at turbidity values of about 2 (Fig. 3B and C). However, β-galactosidase activity remained low and far from the levels observed for the strain containing the transcriptional fusion. Similar strains containing smaller translational fusions including only the first five alkS codons (strain PSH30; not shown), or just the first alkS codon (strain PSH14, see Fig. 3A), had the same behaviour. Therefore, the sequences required to observe this inhibitory effect in the translational fusions can be trimmed down to a region spanning positions +1 to +33 relative to the PalkS2 transcription start site, which corresponds to positions −30 to +3 relative to the alkS AUG translation initiation codon. Inactivation of the crc gene of strain PSH50 by gene replacement with a crc::tet allele (strain PSH50C) led to a clear and significant increase in β-galactosidase activity during exponential growth (strain PSH50C; see Fig. 3B). The same behaviour was observed upon inactivation of the crc gene of strain PSH14 (strain PSH14C, see Fig. 3C). These results indicate that the Crc protein inhibits gene expression in the alkS'-‘lacZ translational fusions analysed. Therefore, the Crc global regulator modulates PalkS2 expression during exponential growth in LB medium at a post-transcriptional level, most likely by inhibiting translation of alkS.
The effect of Crc on expression of the alkBFGHJKL operon is indirect
Crc decreases the mRNA levels arising from promoter PalkB by about sevenfold (Yuste and Rojo, 2001). Transcription from PalkB requires the AlkS transcriptional activator and the presence of alkanes (Panke et al., 1999). AlkS levels are limiting in the cell even when alkanes are present and expression of alkS from PalkS2 is fully induced (Yuste and Rojo, 2001). If Crc decreases AlkS levels by inhibiting alkS translation, as described above, the effect of Crc on PalkB could be indirect, due to its ability to reduce AlkS levels. If this prediction is correct, in the presence of saturating AlkS levels a translational fusion of alkB to the lacZ reporter gene should be insensitive to Crc. To test this idea, a translational fusion was constructed that included PalkB and downstream sequences up to the 18th codon of alkB, followed by the lacZ reporter gene devoid of its own SD (see Fig. 2). The fusion was introduced into P. putida PPSH1, in which the alkS gene is expressed constitutively at high levels from the heterologous Ptrc promoter. In the presence of the effector DCPK, the strain obtained, named PBSH11, expressed lacZ efficiently throughout the exponential phase (Fig. 3D). In the absence of DCPK, β-galactosidase activity remained at background levels. Inactivation of the crc gene in this strain did not result in higher β-galactosidase activities (strain PBSH11C, see Fig. 3D). We conclude that Crc does not modulate alkB translation, and that the effect of Crc on the activity of promoter PalkB is an indirect result of its ability to reduce AlkS levels below those required for full induction of the PalkB promoter.
Crc binds specifically to the alkS mRNA in vitro
To investigate whether Crc inhibits alkS translation by binding to alkS mRNA, a Crc derivative containing a 6xHis-tag at the carboxyl-end was constructed. The gene coding for this protein was introduced into the Crc-deficient strain PBS4C1. The Crc(6xHis) protein complemented the absence of wild-type Crc (not shown), indicating that the His-tag does not impair protein function. The protein was overproduced in Escherichia coli, purified and used to analyse its ability to interact with the alkS mRNA in vitro. The RNA fragment used as substrate was obtained by in vitro transcription and included alkS mRNA sequences spanning positions +1 to +99 relative to PalkS2 transcription start site. Therefore, it included the alkS untranslated region and alkS sequences up to codon 23. The RNA, radioactively labelled, was purified from a denaturing polyacrylamide gel, briefly heated to 90°C and allowed to renature at room temperature. As shown in Fig. 4A, the purified Crc(6xHis) protein generated a clear retardation band in the presence of a large excess of unlabelled competitor tRNA. Under the conditions tested, about 125 ng (200 nM) of Crc was enough to retard 50% of the labelled RNA fragment used (Fig. 4A, lane 4). The protein–RNA complex was not observed if a 50-fold excess of unlabelled alkS mRNA fragment was included in the binding reaction, which acted as a specific competitor (Fig. 4B, lanes 2 and 3). However, addition of an unrelated unlabelled mRNA fragment of the same size as that of the labelled probe, but with a different sequence, did not impair binding of Crc(6xHis) to the labelled RNA fragment (Fig. 4B, lanes 4 and 5). Under the same conditions, Crc was unable to bind to an unrelated labelled RNA fragment, 68 nt in length and with a G + C content of 56% (not shown). Substitution of Crc(6xHis) by purified SmeT or RoxR, two bacterial DNA binding proteins, did not generate a retardation band (not shown). We conclude therefore that purified Crc(6xHis) can specifically bind to alkS mRNA under stringent conditions.
Delimitation of the Crc binding region
To identify the minimum region required for Crc binding, the 3′-end of the alkS mRNA fragment used in the gel-retardation assays was sequentially trimmed. RNA fragments containing up to position +81, +59, +43 or +33 were labelled and the ability of purified Crc(6xHis) protein to recognize them was analysed in band-shift assays. As shown in Fig. 5, deletion of the 3′-end of the alkS mRNA up to positions 81, 59 or 43 did not impair binding of Crc(6xHis) to it. However, further reduction of the 3′-end of the RNA up to position +33 impaired Crc(6xHis) binding (Fig. 5D). A very weak retardation band with a smear below it could be envisioned, suggesting dissociation of the complex during the electrophoresis.
Although the Crc global regulator is known to act as a negative modulator of many genes depending on the carbon source being used, its mechanism of action had not been defined. Crc has sequence similarity to members of the Endonuclease-Exonuclease-Phosphatase family of proteins, which includes Mg2+-dependent endonucleases and a large number of phosphatases involved in intracellular signalling (Dlakic, 2000). Highest similarity scores correspond to Exonuclease III from several bacterial species (67%, 54% and 43% similarity, respectively, to that of Coxiella burnetii, Bacillus subtilis and E. coli). The similarity to endonucleases suggests that Crc may act on, or bind to, nucleic acids. Purified Crc does not show exo- or endonuclase activity over DNA (Collier et al., 1996; MacGregor et al., 1996). Moreover, mutations at residues that are known to be critical for catalysis in Exonuclease III and Mg2+-dependent endonucleases, and that are conserved in Crc, do not decrease Crc activity in vivo (Ruiz-Manzano et al., 2005). Attempts to detect binding of Crc to DNA proved unsuccessful (MacGregor et al., 1996; A. Ruiz-Manzano and F. Rojo, unpubl. obs.), suggesting that it is not a DNA binding protein. In this work we have investigated whether Crc inhibits transcription or translation of the alkS gene. We show that, under conditions in which the levels of the AlkS activator are not limiting, translational fusions of the alkS promoter to the lacZ reporter gene are responsive to Crc, but transcriptional fusions are not. This clearly indicates that Crc regulates gene expression post-transcriptionally. This conclusion is consistent with previous results suggesting that Crc modulates expression of the P. putida BkdR regulator post-transcriptionally, although the precise step affected, and the mechanism involved, had not been investigated (Hester et al., 2000b). We have previously shown that Crc decreases alkS mRNA levels six- to sevenfold in cells growing exponentially in a rich medium (Yuste and Rojo, 2001). Although it may appear contradictory with a post-transcriptional role for Crc, it should be recalled that AlkS is an unstable protein, normally present in limiting amounts, which activates expression of its own gene. Therefore, decreasing alkS mRNA translation reduces the amount of AlkS available to activate transcription from promoter PalkS2, which finally decreases alkS mRNA levels. This is why the transcriptional and translational fusions of alkS to lacZ were introduced and analysed in a strain expressing AlkS constitutively from an heterologous promoter, which provides constant and saturating AlkS concentrations (Yuste and Rojo, 2001).
If Crc acts post-transcriptionally, it could do so by affecting translation or degradation of the mRNA. Several evidences discard a direct role in RNA degradation. We have previously shown that Crc has little influence on the stability of alkS mRNA (Yuste and Rojo, 2001). Furthermore, incubation of purified Crc with alkS mRNA did not promote RNA degradation, even after prolonged incubation at 37°C. Rather, Crc could bind to the RNA generating a protein–RNA complex that could be detected in band-shift assays. Binding was competed by an excess of alkS mRNA, but not by a similar mRNA of unrelated sequence. Other RNA sequences unrelated to alkS did not bind to Crc, showing that binding is specific. Trimming of the 3′-end of alkS mRNA up to position 43 did not impede Crc binding. However, an alkS RNA fragment spanning positions 1–33 was unable to bind Crc, or did so very poorly. It should be noted that this 33 nt RNA fragment did respond to Crc in vivo when fused to the lacZ reporter gene. This indicates that nucleotides between position 33 and 43 are important for Crc binding to the RNA, but their absence can be compensated by lacZ sequences. The lacZ sequences following nt 33 of alkS in the translational fusion showed no obvious similarity to the corresponding alkS sequence. This suggests that at least most of the determinants required for Crc binding are included within the first 33 nt of alkS RNA, but efficient Crc binding requires an RNA substrate of certain length, containing about 10 extra nucleotides. Altogether, the results presented allow to conclude that Crc can bind specifically to the translation initiation region of alkS mRNA. This supports a model in which Crc would modulate gene expression by binding to the mRNA in a way that hinders the access of ribosomes to the ribosome binding site, therefore impeding translation initiation (see Fig. 6).
Crc activity depends on the carbon source being used and on the growth phase (Yuste and Rojo, 2001). However, it is unclear how the action of Crc is modulated. Crc levels vary depending on the carbon source being used, and decrease about fivefold when cells reach to stationary phase in LB medium (Ruiz-Manzano et al., 2005). The activity of some RNA binding proteins is regulated by other factors. A well-studied example is that of the CsrA/RsmA family of proteins, which are important global regulators that inhibit translation of target genes by binding to their mRNAs (Romeo, 1998; Pessi et al., 2001; Baker et al., 2002; Dubey et al., 2003; Heurlier et al., 2004). The action of these proteins is modulated by a number of small RNAs that sequester the CsrA/RsmA protein (Weilbacher et al., 2003; Heurlier et al., 2004; Kay et al., 2005). Although the affinity of Crc for RNA could allow a small RNA to modulate its activity, there is at present no evidence for such a mechanism.
There are several examples of regulatory proteins that inhibit translation by impeding the access of the ribosome to the translation initiation region (reviewed in Schlax and Worhunsky, 2003; Kozak, 2005). However, this way of controlling gene expression to assure a hierarchical assimilation of carbon sources is very different from that observed in well-studied model systems such as E. coli or B. subtilis (reviewed in Saier, 1998; Stülke and Hillen, 1999). In these two cases, catabolite repression is based either on a direct inhibition of the transporters of the non-preferred substrates to the cell cytoplasm by protein-protein contacts, or on a complex process that inhibits transcription initiation (but not translation) of many genes. The molecular details and proteins involved are different in the two bacterial species. In Pseudomonads, the regulatory networks controlling nutritional choices are poorly characterized. The results described in this report further support that the molecular mechanisms involved are substantially different from those found in other model organisms (reviewed in Shingler, 2003; Rojo and Dinamarca, 2004; Cases and de Lorenzo, 2005; this work). Given the importance of Pseudomonads in medicine, biotechnology and in the environment (Ramos et al., 1997; Stover et al., 2000; Timmis, 2002; Wackett, 2003), detailed characterization of the regulatory mechanism controlling regulation of transcriptional programs in response to a changing environment deserves further investigation.
Bacterial strains and growth conditions
Pseudomonas putida strains were grown at 30°C in LB medium or in M9 minimal salts medium (Sambrook and Russell, 2001), the latter supplemented with trace elements (Bauchop and Eldsen, 1960) and 30 mM citrate or succinate as the carbon source. Expression of the PalkB or PalkS2 promoters was induced by addition of 0.05% (v/v) DCPK, a non-metabolizable inducer that mimics the effect of alkanes (Grund et al., 1975). E. coli strains were grown in LB medium at 37°C. Antibiotics were added when appropriate at the following concentrations: ampicillin 100 μg ml−1, tetracycline 8 μg ml−1, kanamycin, streptomycin or gentamicin, 50 μg ml−1. Cell growth was followed measuring turbidity at 600 nm.
The bacterial strains and plasmids used are described in Table 1. pUT-mini-Tn5 suicide-donor plasmids containing transcriptional fusions of promoters PalkS2 (−306 to +91 relative to the transcription start point, named pSS1) or PalkB (−520 to +71 relative to the transcriptional start point; named pPBK2) to lacZ have been described before (Yuste et al., 1998; Canosa et al., 2000). The transcriptional fusions contained in plasmids pSS1 and pPBK2 were delivered to the chromosome of P. putida strain PPSH1 by means of triparental matings with plasmid pRK600 as donor of transfer functions. Strain PPSH1 is a derivative of KT2442 that contains a Ptrc-alkS transcriptional fusion in the chromosome. This strain, which expresses alkS constitutively from the Ptrc promoter, was obtained using the Ptrc-alkS suicide-donor plasmid pHLS1 (Yuste and Rojo, 2001). Translational fusions of alkS to lacZ were obtained by cloning DNA fragments including positions −180 to +33 (includes the alkS binding site required for PalkS2 activation, the 5′-alkS mRNA untranslated region and the first alkS codon), −180 to +45 (up to the first five alkS codons), or −180 to +70 (up to the first 14 alkS codons), relative to PalkS2 transcriptional start site, into pUJ9, a plasmid designed to generate translational fusions to lacZ. The plasmids obtained were named pFTS0, pFTS10 and pFTS1 respectively. DNA fragments containing the corresponding alkS’-'lacZ translational fusions were excised from these plasmids with NotI and cloned into the NotI site of pUT-mini-Tn5Km, obtaining plasmids pFTS3, pFTS11 and pFTS2 respectively. These suicide-donor plasmids were used to deliver each of the alkS'-'lacZ translational fusions to the chromosome of P. putida PPSH1, obtaining strains PSH50, PSH30 and PSH14 (see Table 1). The crc gene of the indicated strains was inactivated by allelic exchange with a crc::tet or a crc::aacC1 allele as described (Yuste and Rojo, 2001), using plasmids pCRC10 or pCRC10Gm respectively. Plasmid pCRC10 has been previously described (Yuste and Rojo, 2001). To obtain pCRC10Gm, the aacC1 gene (Gmr determinant) was excised from plasmid pTnMod-OGm as a 0.9 kb SstI segment and, after blunting the ends with T4 DNA polymerase, inserted into the NruI site of the crc gene contained in pCRC5, obtaining plasmid pCRC5Gm. The crc::accC1 allele was excised from pCRC5Gm as a BamHI DNA segment and inserted into the BamHI site of pKNG101, generating pCRC10Gm. The alkB'-'lacZ translational fusion was obtained by cloning a DNA fragment spanning positions −131 to +155 relative to the PalkB transcription start site (includes the AlkS binding site required for PalkB activation, the 5′-untranslated region of alkB mRNA, and the first 18 alkB codons) into plasmid pUJ9, obtaining plasmid pFTB1. The alkB'-'lacZ translational fusion was excised from pFTB2 as a NotI fragment and cloned at the NotI site of pUT-Mini-Tn5Km, obtaining pFTB2. This suicide-donor plasmid was used to deliver the translational fusion into the chromosome of P. putida PPSH1, obtaining strain PBSH11. The crc gene of PBSH11 was inactivated by allelic exchange with plasmid pCRC10Gm, as described above, obtaining strain PBSH11C. The DNA sequence of all constructions was determined to verify the modifications introduced and to assure the absence of spurious mutations.
A derivative of the crc gene containing a 6xHis tag at the C-terminus was obtained by polymerase chain reaction (PCR) amplification using pCRC5 as template and oligonucleotides 5′-ggttttcccagtcacgacgt-3′ (hybridizes on vector sequences) and 5′-gtggtggtggtggtggtgaatggcctttttgatggtcag-3′. The amplified DNA fragment, which contained the crc(6xHis) gene expressed from the native Pcrc promoter, was cloned into pGEM-T Easy, generating pGCRCH. A NotI DNA fragment containing crc(6xHis) was excised from this plasmid and cloned at the NotI site of pUT-Mini-Tn5Sm, obtaining pSM-CH. This suicide-donor plasmid was used to deliver the crc(6xHis) into the chromosome of P. putida PBS4C1, which contains an inactivated crc::tet allele. The strain obtained was named PBS4C1-CH. Protein Crc(6xHis) was purified using the expression plasmid pCRCH, which derives from pET-22b(+). To construct this plasmid, the crc gene was PCR-amplified from pCRC5 using oligonucleotides 5′-ggaattccatatgatgcggatcatcagtgtg-3′ and 5′-cccaagcttgatggtcagcgtcca-3′, which introduce NdeI and HindIII restriction sites that were used to clone the DNA fragment into the corresponding sites of pET-22b(+).
An overnight culture was diluted to a final turbidity (A600) of 0.04 in fresh medium. When turbidity reached to 0.08, the non-metabolizable inducer DCPK (0.05% [v/v]) was added, where indicated, to induce expression of the alkane degradation pathway promoters. Cultures were grown at 30°C. At different time points, aliquots were taken and β-galactosidase activity was measured as described by Miller (Miller, 1972), using o-nitrophenyl-β-d-galactoside as substrate. At least three independent assays were performed.
To purify Crc(6xHis), E. coli BL21(DE3)-pLysS containing plasmid pCRCH was grown in LB medium at 37°C to a turbidity (A600) of 0.5. Expression of crc(6xHis) from the T7 RNA polymerase-dependent promoter of the plasmid was induced by addition of 1 mM IPTG (isopropyl-β-d-thiogalactopyranoside). After 4 h, cells were harvested by centrifugation, resuspended in 20 mM Tris-HCl, pH 7.5, 30 mM imidazole, 300 mM NaCl and disrupted by sonication. After eliminating cell debris by centrifugation, the supernatant was loaded into a Ni–NTA column. The column was washed with the same buffer and eluted with increasing concentrations imidazole. Crc(6xHis) protein was eluted at 230 mM imidazole. The protein obtained was > 95% pure as judged by SDS-polyacrylamide gel electrophoresis.
RNA band-shift assays
Reactions contained, in 20 μl, 10 mM Hepes-KOH, pH 7.9, 35 mM KCl, 2 mM MgCl2, 0.1 nM radioactively labelled RNA, 1 μg yeast tRNA and, were indicated, purified Crc(6xHis). Where specified, reactions contained as well 5 nM of the specified unlabelled mRNA, or the purified DNA binding proteins SmeT or RoxR (from Stenotrophomonas maltophilia and P. putida, respectively, kindly provided by A. Henández and A. Ugidos). The non-specific competitor RNA, 235 nt in length, had a G + C content of 60.4%. After 1 h incubation at room temperature, 4 μl of loading buffer (60% glycerol, 0.025% xylene cyanol) were added and samples were loaded onto a non-denaturing 4% polyacrylamide gel containing TBM buffer (45 mM Tris-HCl, pH 8.3, 43 mM boric acid, 2 mM MgCl2, 5% glycerol). Electrophoresis was performed at 4°C using TBM as running buffer.
The radioactively labelled RNA fragments used as substrate including alkS mRNA sequences up to positions 43, 59, 81 or 99 were obtained by in vitro transcription using plasmids pΔalkS-43, pΔalkS-59, pΔalkS-81 or pΔalkS-SN as template respectively. Transcription was performed for 1 h at 37°C; reactions contained 40 mM Tris-HCl (pH 7.9), 6 mM MgCl2, 2 mM spermidine, 10 mM NaCl, 1 μg of DNA template (linearized with PstI or EcoRI; see below), 10 mM DTT, 0.5 mM ATP, CTP and GTP and [α32P]-UTP (3000 Ci mmol−1) and 20 U of T7 or SP6 RNA polymerase. The reaction was loaded into a denaturing 6% polyacrylamide gel and the transcript corresponding to the alkS 5′-end was excised from the gel and purified.
To generate the plasmids pΔalkS-43, pΔalkS-59 and pΔalkS-81, DNA fragments of appropriate lengths were PCR-amplified with oligonucleotides providing targets for ApaI (5′-end of the alkS mRNA) and PstI (3′-end). The DNA fragments obtained were digested with these endonucleases and cloned between the corresponding sites of pGEM-T Easy. The plasmids generated allowed to obtain the desired RNA transcripts from the T7 promoter of the vector after digesting them with PstI. To obtain plasmid pΔalkS-SN, a 99 bp DNA fragment spanning positions +1 to +99 of the alkS mRNA (relative to PalkS2 transcription start site) was PCR-amplified and introduced into pGEM-T Easy, obtaining pΔalkS. Vector sequences between the SP6 promoter and the start of the alkS mRNA were trimmed down by digesting pΔalkS with SpeI and NsiI, blunting the DNA ends with T4 DNA polymerase, and self-ligation. The resulting plasmid, named pΔalkS-SN, allowed obtaining the desired RNA transcript from the SP6 promoter of the vector after digestion with EcoRI. The DNA sequence of these constructions was verified in all cases. The RNA fragment containing alkS mRNA sequences up to position 33 was chemically synthesized by Invitrogen.
We are grateful to E. Martínez-Salas for advice and helpful discussions, and to A. Ugidos and A. Hernández for providing purified RoxR and SmeT proteins respectively. This work was supported by Grants BMC2003-00063 and BFU2006-00767/BMC from the Spanish Ministry of Education and Science. A.R.-M. was recipient of a predoctoral fellowship from the Government of the Basque Country.