INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

When bacteria are confronted with a mixture of carbon sources, many species preferentially utilize one carbon source until depletion and only then begin to utilize the remaining carbon sources. The genes encoding the carbon catabolic enzymes are often regulated in a similar manner; during growth on the preferred carbon source, other carbon catabolic routes are repressed. Carbon catabolite repression is well described for enteric bacteria and some gram-positive bacteria (for a review, see reference 38). In Escherichia coli, the uptake of its preferred carbon source, glucose, takes place via the phosphotransferase system, which indirectly regulates gene expression by regulating the enzyme adenylate cyclase. The product of this enzyme, cyclic AMP (cAMP), together with its cognate pleiotropic transcriptional regulatory protein, the cAMP receptor protein (CRP), modulates the expression of a multitude of catabolic operons (4).

In Pseudomonas species, the preferred carbon source is usually an organic acid such as lactate or a tricarboxylic acid cycle intermediate like citrate or succinate (29). In the presence of these substrates, the expression of catabolic pathways for terpenes such as camphor (24); for aromatic compounds like benzene (31), styrene (37), aniline (25), protocatechuate (47), phenol (34), and toluene (13, 26, 30); for chloroaromatic compounds (32); and for other nonaromatic compounds and carbohydrates (29) is subject to carbon catabolite repression.

The soil bacterium Pseudomonas oleovorans can utilize n-alkanes as sole carbon and energy sources by sequential oxidation of the terminal methyl group of the substrate into n-alkanols, n-alkanals, and n-alkanoic acids. The latter compounds are fed into the -oxidation pathway for fatty acid degradation. The genes encoding the n-alkane degradation pathway (the alk genes) are clustered in two regions on the catabolic OCT plasmid (6), the alkBFGHJKL operon (16) and the alkST region (14) (see Fig. 1). The genetics and biochemistry of alkane oxidation by P. oleovorans have been reviewed elsewhere (42).

Transcription activation of the alkBFGHJKL operon is mediated by the transcriptional activator AlkS (14) in response to a diverse range of inducers, including substrates like n-alkanes, gratuitous inducers such as dicyclopropylketone (DCPK) (22), and other aliphatic compounds (46) (see Fig. 1).

In the present paper, we show that induction of the P_alkBFGHJKL promoter (P_alk promoter) is repressed at the transcriptional level in P. oleovorans grown on certain organic acids and glucose. In contrast, when the cloned alk system is expressed in E. coli alk⁺ recombinants, there is no carbon catabolite repression of the P_alk promoter. We present data which suggest that carbon catabolite control of the P_alk promoter is a negative control.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Chemicals. The digoxigenin (DIG)-DNA-labeling kit, yeast total RNA, RNase-free DNase I, RNase inhibitor, positively charged nylon membrane, RNA molecular weight marker II, anti-DIG-alkaline phosphatase conjugate (Fab fragments), and disodium 3-(4-methoxyspiro{1,2-dioxetane-3,2'-(5'-chloro)tricyclo-[3.3.1.1^3,7] decan}-4-yl)phenyl phosphate (CSPD) were all from Boehringer Mannheim; X-ray film (Curix) was from Agfa. Amersham supplied ¹⁴C-labeled methylated molecular mass markers and L-[³⁵S]methionine. Casamino Acids were from Difco, DCPK was purchased from Sigma, and the RNeasy kit was supplied by Qiagen. All other chemicals were supplied by Fluka.

Strains and plasmids. P. putida (oleovorans) TF4-1L (ATCC 29347) is a prototroph strain and carries the OCT plasmid (40). P. putida GPo12 is a variant of P. oleovorans cured of the OCT plasmid (7). E. coli HB101 is (gpt-proA)62 leuB6 thi-1 lacY1 hsdS_B20 recA23 rpsL20 ara-14 galK2 xyl-5 mtl-1 supE44 mcrB_B (5). E. coli W3110 [F IN (rrnD-rrnE)1] is a prototrophic K-12 strain (3), and E. coli GEc137 (15) is a fadR derivative of DH1, which is recA1 endA1 supE44 hsdR17 gyrA96 thi-1 relA1 (23). Plasmid pGEc600 carries the alkB gene in pUC19 (36). Plasmid pGEc29 consists of plasmid pLAFR1 and the alkBFGHJKL operon (17), and plasmid pGEc47 consists of the alkBFGHJKL operon and the alkST region in plasmid pLAFR1 (15). The plasmids used in this study are depicted in Fig. 1.

View larger version (13K): [in this window] [in a new window]   FIG. 1.   Genetic organization and induction of the alk genes on the OCT plasmid. (Top) In response to the presence of an inducer (e.g., DCPK or octane), AlkS initiates transcription of the alkBFGHJKL operon. (Bottom) Plasmids used in this study containing part of or all of the alk genes. The bar below pGEc600 denotes the DNA fragment used as the alkB probe. Restriction enzyme recognition sites: A, Asp718; B, BamHI; P, PstI; S, SalI.

Growth and induction conditions. Cells were grown on minimal medium E2 agar plates supplemented with MT microelement solution (27). Liquid cultures were grown in M9 medium (33) supplemented with FeSO₄ to a final concentration of 10 µM (E. coli) or 30 µM (Pseudomonas) in Erlenmeyer flasks, at 30°C (Pseudomonas) or 37°C (E. coli) in a gyratory shaker set at 200 rpm. Various carbon and energy sources were added to the medium at 0.5% (wt/vol) as indicated in the text. Ampicillin and tetracycline were used at concentrations of 100 and 12.5 µg/ml, respectively. For E. coli GEc137, the growth medium was supplemented with 2 µg of thiamine hydrochloride per ml and 10 g of Casamino Acids per liter. For E. coli HB101(pGEc47), the growth medium was supplemented with 92 µg of L-proline per ml, 80 µg of L-leucine per ml, and 2 µg of thiamine hydrochloride per ml.

Nucleic acid techniques. (i) DNA manipulations. Transformation of E. coli, isolation of plasmid and chromosomal DNA, restriction digestions, agarose gel electrophoresis, and Southern transfer of DNA were carried out according to standard protocols (2, 39).

(ii) Conjugative transfer. Mobilization of pGEc47 to P. putida GPo12 was performed according to the triparental mating procedure of Ditta et al. (12). Exconjugants were selected for simultaneous tetracycline resistance and growth on octane (supplied in vapor phase) on minimal medium agar plates.

(iii) DNA probe. A 528-bp BamHI fragment from plasmid pGEc600, which represents an internal fragment of alkB, was isolated from an agarose gel by phenol extraction (2). The fragment was labeled nonradioactively with DIG11-dUTP with the DIG-DNA-labeling kit according to the manufacturer's protocol.

(iv) RNA isolation. Cells were harvested from exponentially growing cultures and spun down at 4°C (8,000 × g, 5 min). Cell pellets were immediately frozen in liquid nitrogen and kept at 70°C until use. Total RNA was isolated from approximately 0.25 mg (dry weight) of cells with the RNeasy kit according to the manufacturer's protocol. The RNA samples were treated with RNase-free DNase I in the presence of RNase inhibitor (2). Purity and RNA concentration of the samples were determined by measuring the A₂₆₀ and A₂₈₀, with the GeneQuant photospectrometer (Pharmacia). RNA samples were stored at 70°C.

(v) RNA-agarose gel electrophoresis, Northern blotting, and detection. Total RNA samples, accompanied by RNA molecular weight markers, were separated on a 1% (wt/vol) agarose gel containing formaldehyde as a denaturant (28). Prior to capillary Northern transfer onto a nylon membrane, the gel was subjected to a partial alkaline hydrolysis to improve the transfer of high-molecular-weight RNA fragments (39). After transfer, the RNA was fixed onto the membrane by UV cross-linking (1 min) on a standard UV transilluminator. Prehybridization and hybridization conditions (the alkB probe was used at ±40 ng/ml) for Northern blots were as specified by Boehringer Mannheim, with 50 µg of yeast total RNA per ml added to the (pre)hybridization solutions to reduce background signals. Bound probes were detected with anti-DIG-alkaline phosphatase conjugate (Fab fragments) and CSPD as chemiluminescent substrate according to Boehringer Mannheim's protocol. Signals were recorded on X-ray film and analyzed in a Molecular Dynamics densitometer.

Pulse-chase labeling, immunoprecipitation, SDS-polyacrylamide gel electrophoresis, fluorography, and quantification of immunoprecipitates. Pulse-chase-labeling with L-[³⁵S]methionine (specific activity, >1,000 Ci/mmol), immunoprecipitation of AlkB from labeled cell lysates followed by sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis, fluorography, and quantification of immunoprecipitates are described in detail elsewhere (41).

Alkane hydroxylase in vivo activity measurement. DCPK-induced P. oleovorans and E. coli alk⁺ recombinant cultures were harvested by centrifugation and resuspended to 2 g (dry weight) of cells per liter in assay buffer (0.1 M KPO₄ [pH 7.0], 10 mM MgSO₄, 0.05% [wt/vol] Triton X-100) containing 1% (wt/vol) of the respective carbon source. Two hundred fifty microliters of the cell suspension was transferred (in duplicate) to glass centrifuge tubes, which could be closed tightly with a screw cap, preventing evaporation of organic solvents. The tubes were placed in a rotary shaker set at 200 rpm at 30°C (P. oleovorans) or 37°C (E. coli alk⁺ recombinants). After 5 to 10 min of preincubation, 5 µl of n-nonene (99.9% pure) (Wiley Organics, Coshocton, Ohio) was added to the tubes. After incubation for another 15 min, the tubes were cooled down in an ice-water bath, the aqueous phase was saturated with NaCl, and the suspension was extracted by vortexing it for 1 min with 250 µl of n-hexane containing 0.01 or 0.1% (vol/vol) 2-octanol as internal standard. The phases were separated by centrifugation and freezing at 80°C. The hexane phase was analyzed for 1,2-epoxynonane by gas chromatography (Fisons Instruments MFC 800) isothermally at 80°C, by split injection onto a WCOT fused-silica column, CP-sil-5-CB, 25 m by 0.32 mm (Chrompack). Compounds were detected with a flame ionization detector. The specific activity of the alkane hydroxylase system was expressed as international units (IU) per gram (dry weight) of cells (micromoles of 1,2-epoxynonane · minute¹ · gram [dry weight] of cells¹). Control experiments showed that the rate of epoxide formation was linear within the first 20 to 30 min after addition of the substrate. The lower detection limit was about 0.1 U/g (dry weight) of cells. The addition of Triton X-100 to the assay buffer was necessary to obtain maximum epoxidation activities, as already noted by de Smet et al. (11). The assay is not compatible with the presence of n-octane in the growth medium, since n-octane is a competitive substrate for the n-nonene epoxidation reaction.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Influence of the carbon source on alkane hydroxylase activity in P. oleovorans. Alkane oxidation activity in various P. aeruginosa (10, 43, 44) and P. putida (21, 22) strains has been reported to be subject to carbon catabolite repression by glucose and other carbon sources. We have pursued this line of investigation and have investigated the alkane hydroxylase activity of cultures of P. oleovorans grown on different carbon sources. Exponentially growing cultures were induced with the gratuitous inducer DCPK, and after 2 h, cells were harvested and incubated with n-nonene. The alkane hydroxylase system converts this substrate into 1,2-epoxynonane, which is readily analyzed by gas chromatography.

AlkB induction kinetics in P. oleovorans. To investigate whether the reduced alkane hydroxylase activity reflected reduced expression of the alk genes, we determined the relative synthesis rate of AlkB in P. oleovorans cultures grown on minimal medium with different carbon sources.

View larger version (45K): [in this window] [in a new window]   FIG. 2.   Synthesis of AlkB by P. oleovorans during growth on glycerol or glucose. The P. oleovorans cultures, growing exponentially on glycerol (A) or glucose (B), were induced with DCPK at t = 0. At the indicated time points (hours), cell samples were pulse-labeled with [35S]methionine, and AlkB was immunoprecipitated from cell lysates containing 100,000 trichloroacetic acid-precipitable counts. After SDS-polyacrylamide gel electrophoresis analysis and treatment with 1 M sodium salicylate as fluorophore, X-ray films were exposed to the gels at 80°C for 142 h. Molecular mass markers (lane Mw) are indicated in kilodaltons.

View larger version (23K): [in this window] [in a new window]   FIG. 3.   Growth and relative AlkB synthesis rates in P. oleovorans cultures growing on different carbon sources. The cultures were induced at t = 0, after which cell samples were pulse-chase-labeled with [35S]methionine at the indicated time points. Growth of noninduced () and induced () cultures is expressed as dry weight of cells. AlkB was immunoprecipitated from cell lysates, and its synthesis rate was quantified relative to total protein synthesis as described in Materials and Methods. AlkB synthesis in noninduced () and induced () cultures was expressed as weight percentage of newly synthesized protein. For induction, 0.05% (vol/vol) DCPK was used for the cultures grown on lactate (A), glucose (B), citrate (C), pyruvate (D), and glycerol (E). The culture in panel F was grown on glycerol and induced with 5% (vol/vol) n-octane. In addition, growth of a P. oleovorans culture with 5% (vol/vol) n-octane as sole carbon source is indicated () in panel F.

1

AlkB induction kinetics in E. coli alk⁺ recombinants. The alkane hydroxylase system was introduced into E. coli on plasmid pGEc47, allowing controlled and functional expression of the alk genes in E. coli (15). The induction kinetics of AlkB in two E. coli alk⁺ recombinants, HB101(pGEc47) and W3110(pGEc47), were investigated as described above for P. oleovorans. As carbon sources for the E. coli alk⁺ recombinants, glycerol (a prototype nonrepressive carbon source) and glucose (a prototype repressive carbon source) were chosen.

View larger version (15K): [in this window] [in a new window]   FIG. 4.   Growth and relative AlkB synthesis levels in E. coli alk+ recombinant cultures growing on different carbon sources. The cultures were induced with 0.05% (vol/vol) DCPK at t = 0, after which cell samples were pulse-chase-labeled with [35S]methionine at the indicated time points and analyzed for newly synthesized AlkB. Growth of noninduced () and induced () cultures is expressed as dry weight of cells, and AlkB synthesis in noninduced () and induced () cultures was expressed as weight percentage of newly synthesized protein. (A) E. coli HB101(pGEc47) grown on glucose; (B) E. coli W3110(pGEc47) grown on glucose; (C) E. coli W3110(pGEc47) grown on glycerol.

Expression from the P_alk promoter. The specificity of the DIG-labeled alkB probe (consisting of an internal alkB sequence [Fig. 1]) was verified on Southern blots of chromosomal and plasmid DNA from various strains (data not shown). No alkB DNA sequences were found in chromosomal DNA samples of P. putida GPo12, thereby confirming the absence of the alk genes from this strain.

View larger version (26K): [in this window] [in a new window]   FIG. 5.   Northern blot analysis of total RNA isolated from Pseudomonas and E. coli recombinant cultures with or without 2-h induction of the alk genes with DCPK. Bound DIG-labeled alkB DNA probes were detected with a chemiluminescent substrate. The signal of the blots in panels A, B, and C was recorded on X-ray film by exposure for 5, 2, and 20 min, respectively. The plus and minus signs above the lanes indicate samples taken from induced and uninduced cultures, respectively. RNA molecular size markers of 5.3, 2.8, 1.9, and 1.6 kb (top to bottom, respectively) are indicated in the right margin of each panel. The alkB signal was detected at about 1.5 kb. (A) Lanes 1 and 2, P. putida GPo12 grown on glycerol; lanes 3 and 4, P. oleovorans grown on succinate; lane 5, P. oleovorans grown on glucose; lane 6, P. oleovorans grown on glycerol. (B) Lanes 1 and 2, E. coli W3110(pGEc47) grown on glucose; lanes 3 and 4, E. coli HB101(pGEc47) grown on glucose; lanes 5 and 6, P. oleovorans grown on glucose; lanes 7 and 8, E. coli W3110(pGEc47) grown on pyruvate; lanes 9 and 10, P. oleovorans grown on pyruvate; lane 11, P. oleovorans grown on succinate. (C) Lanes 1 and 2, E. coli W3110(pGEc47) grown on glycerol; lanes 3 and 4, P. oleovorans grown on glycerol; lanes 5 and 6, E. coli HB101(pGEc47) grown on lactate; lanes 7 and 8, P. oleovorans grown on lactate; lanes 9 and 10, E. coli GEc137(pGEc29) grown on glucose.

Carbon catabolite repression of the P_alk promoter in P. putida GPo12(pGEc47) is reduced. Reintroduction of the alk genes on plasmid pGEc47 restores the ability of P. putida GPo12 to use octane as a carbon source (15). We investigated whether the carbon catabolite repression of alkane hydroxylase activity was controlled identically in P. oleovorans and in P. putida GPo12(pGEc47). Analogously to the experiments performed with P. oleovorans, the alkane hydroxylase activities of induced and noninduced cultures grown on different carbon sources were measured (Table 1, right columns).

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Carbon catabolite repression in pseudomonads appears to be regulated differently from that in Bacillus subtilis or E. coli (9). Whereas in the latter microorganism, the cAMP-CRP complex plays an important role in carbon catabolite repression, no such role for cAMP was found in pseudomonads (29). The only known Pseudomonas CRP homolog, Vfr, was found to be involved primarily in quorum sensing, not in carbon catabolite repression control (1). Thus, the molecular mechanisms of carbon catabolite repression in pseudomonads remain elusive to date.

We have studied the carbon catabolite repression of alkane degradation in P. oleovorans, a P. putida strain carrying the OCT plasmid, in some detail. The induction of the alkane hydroxylase activity was eliminated or reduced significantly during growth on succinate and lactate. Analysis of AlkB synthesis induction indicated that during growth on succinate, lactate, or glucose the expression of the alkBFGHJKL operon was repressed. These results were confirmed by alkB mRNA analysis, showing that carbon catabolite repression of the P_alk promoter in P. oleovorans takes place at the transcriptional level.

Induction with DCPK during growth of P. oleovorans on glycerol, citrate, and pyruvate, but not on lactate or glucose, led to significant expression of AlkB (>0.5% of new protein within 2 h after induction) and to a decrease in growth rate. As we have found for other alk⁺ recombinants (8, 18, 36), the higher the expression of the alkane hydroxylase system and of the alkB gene in particular, the greater the difference in specific growth rates between the noninduced and the induced cultures. As a result, repressive carbon sources are good growth substrates for P. oleovorans even in the presence of alk system inducers, as is indicated by their respective growth rates (µ) in Table 2. The contrary statement is not true: citrate is an excellent growth substrate for P. oleovorans, yet it allows full induction of the P_alk promoter at the expense of a distinct reduction in growth rate.

In order to also study the carbon catabolite repression of the P_alk promoter-AlkS system in enteric bacteria, we transformed two E. coli strains with broad-host-range plasmid pGEc47 containing the alk genes. The induction of alkB mRNA and that of AlkB synthesis were not repressed in the two E. coli alk⁺ recombinants growing on glycerol, lactate, or glucose. Thus, carbon catabolite repression of the P_alk promoter-AlkS system is completely absent in E. coli strains.

To study the carbon catabolite repression behavior of the cloned alk genes in their original genetic background, plasmid pGEc47 was introduced into P. putida GPo12. The resulting alk⁺ recombinant displayed high alkane hydroxylase activities not only when grown on glycerol but also when grown on lactate or succinate (Table 1). Thus, carbon catabolite repression of alkane hydroxylase activity was almost completely abolished in this recombinant strain.

There are several possible explanations for these findings. First, a putative negative control factor might reduce alkS expression in P. oleovorans under carbon catabolite repression conditions, thereby reducing transcription of the alkBFGHJKL operon. Second, a putative negative control factor might bind to AlkS and/or to P_alk, thus compromising the ability of AlkS to initiate transcription from the P_alk promoter.

The OCT plasmid of P. oleovorans has a copy number of 1 to 2 (19). Plasmid pGEc47, an RK2 derivative, has a copy number between 5 and 10 in E. coli and pseudomonads (20). The putative factor exerting negative control over P_alk induction appears to be specific for pseudomonads, since it is completely absent in the E. coli alk⁺ recombinants. The fact that carbon catabolite repression of the P_alk promoter is almost completely abolished when the alk genes are reintroduced into P. putida GPo12 suggests that the control factor might be encoded on the OCT plasmid (outside of the two EcoRI fragments that contain the alkST region and the alkBFGHJKL operon [Fig. 1]). Alternatively, the putative control factor (conferring carbon catabolite repression on the P_alk promoter in P. oleovorans) could be titrated by the extra copies of the P_alk promoter genes present in the P. putida GPo12(pGEc47) recombinants. Additional experimental work will be necessary to clarify these issues.

The finding that the P_alk promoter-AlkS system is not subject to carbon catabolite repression in E. coli allows the use of this system for expression of target genes in recombinant E. coli strains. This strategy can combine several virtues: low-cost inducers (e.g., octane and DCPK), the tightly controlled and efficient gene expression of the P_alk promoter-AlkS system, and the high cell and product yields and low costs associated with the use of glucose as a carbon source.