INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Autotrophy denotes the ability of an organism to gain the majority of its cell carbon by the assimilation of CO₂. The Calvin-Benson-Bassham carbon reduction cycle (2) is the quantitatively predominating route of CO₂ fixation among autotrophs (28). Ralstonia eutropha (Alcaligenes eutrophus) is an aerobic, facultatively chemoautotrophic -proteobacterium, which fixes CO₂ via the Calvin cycle, by using either hydrogen or formate as an energy source for litho- or organoautotrophic growth, respectively (5). Most genes of Calvin cycle enzymes in R. eutropha, including those of ribulose-1,5-bisphosphate carboxylase oxygenase (RubisCO)the actual CO₂-fixing enzyme of the cycleare encoded within a chromosomal operon (cbb_c operon) of about 15 kbp. Strain H16 harbors a second, highly homologous, and identically organized copy of the operon (cbb_p operon) located on the megaplasmid pHG1 (20). The cbb_p operon lacks, however, the 3'-terminal cbbB_c gene present in the cbb_c operon (4), whereas the large subunit gene cbbL of RubisCO forms the 5' ends of both operons. Although the organizations and sizes of cbb operons vary considerably among autotrophic bacteria, they also show some common features (11, 20, 33).

The cbb_c operon of R. eutropha H16 is separated by 167 bp from the divergently oriented regulatory gene cbbR, whose product is required for transcription activation of the duplicate cbb operons. CbbR belongs to the LysR family of transcriptional regulator proteins (32). A deficient cbbR gene (cbbR') is situated exactly in the position corresponding to the cbb_p operon on pHG1 (37). The intergenic segments between cbbR or cbbR' and cbbL_c or cbbL_p constitute the control regions of the operons, as they contain the operators and promoters of the system (19, 21). CbbR has been shown to bind preferentially between positions 29 and 74 relative to the transcriptional start point of the cbb_c operon (19). Consistent with metabolic economy, the heterotrophic growth of the organism on most organic substrates completely represses the transcription of the cbb operons, to avoid wasting of energy that would be caused by CO₂ fixation under these conditions. Partial derepression occurs during growth on a few substrates, such as fructose and gluconate, but only autotrophic growth results in the full derepression or induction of the operons (6, 10, 21). The transcriptional control is postulated to involve the transduction of a still unknown metabolic signal sensed by CbbR, which modulates the activity of the cbb_c and cbb_p operon promoter P_L (20). No significant subpromoter activity was detected within the operons (31). Structural motifs tentatively assigned to P_L resemble those of ⁷⁰-dependent promoters of Escherichia coli (12), and this applies also, although in a less pronounced way, to the presumed cbbR and cbbR' promoter P_R. The divergent promoters P_L and P_R are arranged in a back-to-back configuration, according to Beck and Warren (3), with overlapping 35 boxes (21).

As a prerequisite to understanding the regulatory mechanism acting upon the cbb operons of R. eutropha H16, critical structural properties of P_L must be known. Therefore, a mutational analysis of chromosomal P_L was performed that aimed at defining sequence elements important for the promoter activity. The activities of altered P_L promoters were determined not only in the homologous host R. eutropha but also in E. coli and in newly constructed cbbR deletion mutants of R. eutropha. The results verified the location and function of the promoter and provided a first hint that, in addition to CbbR, another regulator(s) might participate in the transcriptional control of the cbb operons in R. eutropha.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Bacterial strains, plasmids, and culture conditions. The bacterial strains and plasmids used in this study are shown in Table 1. Strains of R. eutropha were grown in a nutrient broth or mineral medium (MM) at 30°C as described previously (36). MM was routinely supplemented with 0.2% (wt/vol) organic substrate for either heterotrophic or organoautotrophic growth on formate. Mixotrophic cultures were initially grown on 0.1% (wt/vol) fructose until reaching an optical density at 436 nm of 1 to 2, before 0.2% (wt/vol) formate was added and the incubation was continued for another 4 h. Lithoautotrophic cultures were gassed with a mixture of H₂, CO₂, and O₂ (8:1:1 [vol/vol/vol]). E. coli was propagated in Luria-Bertani medium at 37 or 30°C. If required, the media contained ampicillin (50 µg/ml), kanamycin (50 µg/ml for E. coli, 450 µg/ml in MM for R. eutropha, or 120 µg/ml in nutrient broth), or tetracycline (15 µg/ml for E. coli or 20 µg/ml for R. eutropha) as selective antibiotics.

Manipulation and sequencing of DNA. Standard procedures (1, 29) were employed to isolate genomic and plasmid DNA from bacteria, to transform plasmid DNA into E. coli, and for general DNA handling. Restriction endonucleases and DNA-modifying enzymes were used under the reaction conditions recommended by the manufacturers. DNA probes to be applied in Southern hybridizations were nonradioactively labelled with peroxidase by means of a special reaction system (ECL direct labelling and detection kit; Amersham, Brunswick, Germany). For the hybridizations, DNA fragments were separated by agarose gel electrophoresis and transferred onto a nylon membrane (Biodyne B; Pall, Dreieich, Germany) by vacuum blotting. DNA sequences were determined by the dideoxy chain termination method (30) and PCR cycle sequencing (SequiTherm cycle sequencing kit; Biozym, Hessisch Oldendorf, Germany) with ³⁵S- or fluorescence-labelled oligonucleotide primers. Plasmids were conjugally transferred from E. coli S17-1 to strains of R. eutropha by biparental mating (35).

Generation of mutations in P_L and construction of transcriptional fusions. Oligonucleotide-directed mutagenesis carried out in a two-stage PCR (9) was employed to introduce mutations into the cbb_c operon promoter P_L. The specifically designed oligonucleotides M1 through M14 (24-mers; not shown) contained the desired mutations close to their centers. Together with a reverse primer (24-mer) they served individually to perform the first amplification (30 cycles at 95°C for 30 s, 59°C for 30 s, and 72°C for 60 s) of a segment (121 to 147 bp) of the 224-bp BglII-HinfI DNA fragment inserted in pBH2241. The reaction mixtures (100 µl) included 200 ng of pBH2241 as a template, a 2 µM concentration of mutagenesis primer, a 2 µM concentration of reverse primer, 2 mM concentrations of dATP, dGTP, dCTP, and dTTP, 25 mM MgCl₂, and 2 U of Tfl DNA polymerase in a buffer system formulated by the supplier of the polymerase (Biozym) and were placed in a thermal cycler (model PTC-100; MJ Research, Watertown, Mass.). The mutagenized PCR product was purified by electrophoresis in low-melting-point agarose. It was used as the primer in a second PCR (30 cycles at 95°C for 30 s, 50°C for 60 s, and 72°C for 60 s) together with a universal primer (24-mer) to accomplish the amplification of the complete BglII-HinfI fragment. If insufficient amounts of the product (330 bp) were obtained in the second PCR, it was reamplified in a third PCR (the same conditions as in the first PCR) by means of the universal and reverse primers. The product of the final PCR was digested with either EcoRI and HindIII or XbaI and PstI prior to being cloned into correspondingly cleaved pUC19 or pBluescript KS, respectively. DNA sequencing of the respective plasmids pM1 through pM14 (not listed) confirmed the specific mutations. Finally, the inserts of these plasmids were excised with XbaI and PstI and recloned into correspondingly digested pBK. The resulting plasmids, pBKM1 through pBKM14, carried the cbb_c::lacZ transcriptional fusions with differently modified P_L promoters.

Construction of suicide plasmid pNHG1. The suicide vector pLO1 (7,322 bp [23]) has been designed as a tool for gene replacement mutagenesis in gram-negative bacteria, especially in R. eutropha (24). However, because of the inherently low sensitivities of R. eutropha H16 and particularly of strain HF210 to kanamycin, an additional, easily selectable marker in these strains was incorporated into pLO1. For this purpose, the tetracycline resistance gene tet was excised from pBBR1MCS-3 as a 1,484-bp DdeI-VspI DNA fragment. Plasmid pLO1 was partially digested with BamHI, and the recessed 3' ends of the linearized plasmid and of the DdeI-VspI fragment were filled in by treatment with the Klenow fragment of DNA polymerase I. The ligation of the two DNA fragments resulted eventually in the isolation of the new suicide vector pNHG1 (Fig. 1).

View larger version (14K): [in this window] [in a new window]   FIG. 1.   Map of suicide vector pNHG1. The plasmid encompasses 8,810 bp and was constructed by inserting a fragment carrying the tetracycline resistance gene tet from pBBR1MCS-3 into pLO1. Only unique cleavage sites of restriction endonucleases are indicated within the multiple cloning site (MCS). neo, the gene encoding kanamycin resistance; oriT, the origin of transfer replication; sacB, the gene encoding levansucrase.

Construction of cbbR deletion mutants. To construct isogenic cbbR deletion mutants of R. eutropha H16 and HF210, a 1.8-kb HindIII-SalI fragment containing cbbR was first cloned into pUC18, yielding pAEC1200. The plasmid was linearized with BglII and further digested with HindIII after filling in the recessed 3' ends of the BglII site by treatment with Klenow fragment. In a separate reaction pAEC1200 was cut with HindIII and SmaI. The resulting 3.2-kb HindIII-BglII and 0.4-kb HindIII-SmaI fragments, respectively, were ligated to produce pAEC1200, which lacked the internal 891 bp of cbbR (total length of the gene, 954 bp). To reclone the 0.9-kb HindIII-SalI fragment, pAEC1200 was digested with HindIII, treated with Klenow fragment to fill in the 3' ends of the site, and cleaved with XbaI. The 0.9-kb fragment carrying cbbR was finally ligated to pNHG1 digested with XbaI and PmeI, generating pNHG110.

Enzyme assays. The activity of RubisCO in R. eutropha was determined in a radiometric, whole-cell assay based on the fixation of ¹⁴CO₂ as described previously (22). A colorimetric assay was used for -galactosidase (25), as well as a fluorimetric assay for -glucuronidase (14), employing crude cell extracts prepared from R. eutropha or E. coli. One unit of activity represents the amount of enzyme catalyzing the formation of 1 µmol of product per min. Cultures of the strains were grown to an optical density at 436 nm of 2 to 3. The cells were harvested, resuspended in the appropriate buffer, and disrupted by sonication. Cell extracts were obtained after centrifugation at 14,000 × g for 20 min to remove unbroken cells and cell debris. Protein concentrations in the extracts were estimated by the method of Bradford (7).

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Mutational modification of P_L. The proposed ⁷⁰-type P_L promoters of the two cbb operons of R. eutropha H16 are believed to have the following structure on the nontemplate DNA strand: [35] TTTACC-N₁₇-TATACC [10] (Table 2). The resemblance of P_L to the canonical ⁷⁰-dependent promoter of E. coli (12) is evident for the hexameric 35 and 10 boxes as well as the length (17 bp) of the spacer region. In order to study the functional significance of the P_L substructures, various mutations were introduced into the chromosomal wild-type promoter cloned within plasmid pBH2241. Specifically designed oligonucleotides were used to direct the PCR-based mutageneses. Fourteen different mutant P_L promoters were produced that carried single or multiple sequence alterations (Table 2). The 10 and 35 boxes of mutants M1 and M6, respectively, matched the corresponding elements of the E. coli consensus promoter. Mutant M9, which was generated from M1, contains the consensus sequence in both regions. The spacer mutants included clones with either a single nucleotide deletion (M12), an insertion (M13), or a double substitution (M14). P_L::lacZ transcriptional fusions were constructed for each mutant (pM1 through pM14) to enable the subsequent determination of the different promoter activities in R. eutropha as well as in E. coli.

Isolation and phenotype of cbbR deletion mutants. Because the cbb promoter activities were intended to be determined in both the wild-type and a CbbR-free background, cbbR null mutants of R. eutropha wild-type H16 and pHG1-cured HF210 were constructed by gene replacement mutagenesis. The derived mutants, HB14 and HB15, respectively, carried an 891-bp in-frame deletion within cbbR, as verified by the sequencing of corresponding PCR-generated fragments and by Southern hybridizations (data not shown). The potential product of cbbR is a small protein of 20 amino acid residues that lacks 297 residues, including the helix-turn-helix structural motif in the NH₂-terminal domain, of the authentic CbbR. A cbbR-derived product with binding capacity to the cbb operator region could therefore not be formed in the mutants. Both strains concurrently lost the ability to grow autotrophically and to derepress or induce the cbb operons. They regained these properties by in trans complementation with cbbR present on plasmid pUW7. The transconjugant HB14(pUW7) showed RubisCO activities similar to that of wild-type H16 after mixotrophic growth on fructose plus formate or heterotrophic growth on fructose, whereas no activity was detected in cells grown on pyruvate (Fig. 2). These data corroborated the function of CbbR as an activator of the cbb operons and the suitability of the mutants for the subsequent promoter assays.

View larger version (24K): [in this window] [in a new window]   FIG. 2.   Activities of RubisCO in R. eutropha H16, HB14, and HB14(pUW7) grown on fructose plus formate, fructose, or pyruvate. The enzyme activity was determined by means of a whole-cell assay. Sp act, specific activity.

Activities of mutant P_L in R. eutropha. The -galactosidase reporter activities originating from the various mutant P_L::lacZ fusions were first determined in transconjugants of wild-type R. eutropha H16 grown under lithoautotrophic (H₂-CO₂) or heterotrophic (fructose or pyruvate) conditions. Confidence in the significance of the individual data on promoter activities was ensured by obtaining them from at least three independent cultures. Promoters with a closer match of the 10 (M1, M2, and M3) or 35 (M5, M6, and M7) box of P_L to those of the ⁷⁰ consensus promoter of E. coli showed increased activity relative to that of the wild-type promoter present in pBK2241 (Table 3). The increase depended on the position and number of base substitutions but occurred independently of the growth substrate. This was particularly evident for mutants M1 (10 consensus), M6 (35 consensus), and M9 (full consensus), whose activities in autotrophic cells reached a level about twofold that of the wild type. More significant, however, was the drastic increase in heterotrophic cells, even after growth on the normally strongly repressing pyruvate. The basal activity of P_L M9 in pyruvate-grown cells was actually higher than that of the wild-type promoter under autotrophic conditions. At least M9 can thus be considered to be nearly constitutive. Nevertheless, all improved variants of P_L still exhibited the principal regulatory pattern of the wild-type promoter characterized by derepression or induction in autotrophically or fructose-grown cells. Diminishing the resemblance of either the 10 (M4) or 35 (M8) box to the consensus by replacing the conserved distal bases TA or TT with CC resulted in a nearly total loss of promoter activity. The substitution of the distal T of the 35 box in the full-consensus M9 converted the constitutive into a repressible promoter (M10). Replacing one consensus base in the 10 box of wild-type P_L and simultaneously introducing a 1-base match in the 35 box gave a promoter (M11) which was slightly more active than the wild type and also retained the activation. In contrast, the deletion of one base within the spacer region (N₁₇N₁₆) largely inactivated P_L (M12), whereas the insertion of one base (N₁₇N₁₈) caused only a partial loss of activity (M13). An exchange of two adjacent bases (GTTA) within the spacer had no significant effect on the promoter (M14).

Activities of mutant P_L promoters in E. coli. Because the rationale to mutagenize P_L of R. eutropha was based on the consensus structure of an E. coli ⁷⁰ promoter, the activities of the various P_L::lacZ fusions were also determined in E. coli transformants. Wild-type P_L showed very low but reproducible activity in the foreign host as did most of the mutant promoters (Fig. 3). A significant increase was observed with mutant M5, but only M1, M6, and M9 yielded high P_L activities. The full-consensus mutant M9 was most active and reached the same level as the activated wild-type P_L in autotrophically grown R. eutropha. These data basically supported the findings made on the activities of mutant P_L promoters in the authentic host.

View larger version (15K): [in this window] [in a new window]   FIG. 3.   Activities of mutant PL promoters determined by PL::lacZ fusions (pBKM1 through pBKM14) in transformants of E. coli JW1 grown in Luria-Bertani medium. The strains harboring pBK and pBK2241 served as background and wild-type references, respectively. The specific activity (sp act) values represent mean values obtained from at least three independent determinations. They varied up to ±5 mU/mg in the low-level activity range (2 to 60 mU/mg) and up to ±30 mU/mg in the high range (510 to 930 mU/mg).

P_R activities associated with mutant P_L. The presumed cbbR promoter P_Ra active in autotrophically growing R. eutropha overlaps P_L. Its 35 box is located within the spacer region of the cbb operon promoter (21). Mutations in P_L were thus thought to potentially influence P_R activity. The -glucuronidase reporter activities of the P_R::gusA fusions were determined for P_L mutants M1, M6, and M9 in transconjugants of R. eutropha H16 and HF210 as well as of the corresponding cbbR mutants HB14 and HB15 grown under heterotrophic or mixotrophic conditions. The previously observed, relatively low P_R activity in pyruvate-grown cells (19, 21) was confirmed for the wild-type P_L region and for mutants M6 and M9 (Table 5). Surprisingly, mutant M1 showed a strongly enhanced activity in all four transconjugants. Determinations with the remaining P_L variants revealed no significant influence of the mutations on P_R activity (data not shown). Growth on fructose plus formate led to significantly higher P_R activities, although the increase was less pronounced in HF210 and HB15 (Table 5). This putative activation or derepression of P_R, which apparently occurs in the wild type as well as in a CbbR-free background, parallels the findings obtained for P_L. P_R activities were not detected in transformants of E. coli.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

cbb promoter activities in a wild-type background. The present work reports on a detailed mutational study of the cbb operon promoter of R. eutropha. Its intention was to verify the proposed location of P_L and identify structural attributes which are of importance for the activity and regulation of the promoter. Our previous supposition of P_L as a ⁷⁰-type promoter located at the predicted position within the cbb control region of R. eutropha H16 was confirmed by the data. The resemblance of P_L to the E. coli ⁷⁰ consensus promoter is supported by the low but significant level of expression of the RubisCO genes cbbLS in the heterologous host (13). In order to investigate structure-function relationships of P_L, alterations were introduced into the presumed promoter elements to either increase or diminish their similarity to the consensus promoter. Mutations that strongly reduced or abolished the activity of P_L were due to changes deviating from the consensus nucleotides within the 10 or 35 box. The modification of the length of the spacer region between the boxes by a 1-bp deletion or insertion also resulted in drastic decreases of activated promoter activity, corroborating both the structure and location of P_L. The optimal length of the spacer (N₁₇) rather than its sequence is a critical parameter because the exchange of two base pairs in P_L M14 had no effect on the promoter. This finding is consistent with the equivalent activities of the chromosomal and pHG1-borne P_L promoters of R. eutropha H16, which differ only in the two spacer bases (GG versus TA) located directly downstream of the mutated bases in M14 (21).

cbb promoter activities in a CbbR-free background. A CbbR-deficient mutant, strain HB13, of R. eutropha has already been isolated (37). In this mutant cbbR is disrupted by a Tn5 insertion within the 3' third of the gene (21), potentially permitting the production of a truncated CbbR protein, which might still be able to bind to the operator in the cbb control region. Therefore, to provide a noninterfering CbbR-free background for the cbb promoter analysis, the cbbR mutant strains HB14 and HB15 were constructed.