Metabolic Signals That Lead to Control of CBB Gene Expression in Rhodobacter capsulatus

Abstract

Various mutant strains were used to examine the regulation and metabolic control of the Calvin-Benson-Bassham (CBB) reductive pentose phosphate pathway in Rhodobacter capsulatus. Previously, a ribulose 1,5-bisphosphate carboxylase/oxygenase (RubisCO)-deficient strain (strain SBI/II) was found to show enhanced levels of cbb_I and cbb_II promoter activities during photoheterotrophic growth in the presence of dimethyl sulfoxide. With this strain as the starting point, additional mutations were made in genes encoding phosphoribulokinase and transketolase and in the gene encoding the LysR-type transcriptional activator, CbbR_II. These strains revealed that a product generated by phosphoribulokinase was involved in control of CbbR-mediated cbb gene expression in SBI/II. Additionally, heterologous expression experiments indicated that Rhodobacter sphaeroides CbbR responded to the same metabolic signal in R. capsulatus SBI/II and mutant strain backgrounds.

The Calvin-Benson-Bassham (CBB) reductive pentose phosphate pathway is the primary pathway by which plants, algae, and most photosynthetic bacteria accomplish the fixation of carbon dioxide into organic carbon for subsequent cell metabolism and growth (59). There exist two key enzymes unique to the CBB cycle: ribulose 1,5-bisphosphate (RuBP) carboxylase/oxygenase (RubisCO), which catalyzes the carboxylation of RuBP, and phosphoribulokinase (PRK), which catalyzes a reaction in which the CO₂ acceptor molecule, RuBP, is generated via the phosphorylation of ribulose 5-phosphate with ATP. CO₂ fixation via the CBB pathway has two major physiological roles in nonsulfur purple photosynthetic bacteria. Under photoautotrophic or chemoautotrophic growth conditions, the key enzymes of the CBB pathway are maximally expressed, with CBB-dependent CO₂ fixation being the major synthetic pathway for the synthesis of organic carbon. Substantially lower levels of CBB cycle enzymes, however, are synthesized under photoheterotrophic growth conditions. Under such conditions, basal levels of CBB pathway enzymes perform the important metabolic function of enabling the cell to employ CO₂ as a preferred electron sink for excess reducing equivalents generated during photoheterotrophic metabolism (58). Thus, when carbon substrates such as l-malate and succinate are metabolized, the CBB cycle facilitates balancing of the oxidation-reduction potential (redox poise) of the cell (17, 37, 69). Given its different roles in photoheterotrophic and photoautotrophic metabolism, it is not surprising that multiple levels of control influence expression of the CBB system (10, 22, 46, 59, 67).

Enzymes of the CBB pathway are encoded by the cbb genes, which are organized in regulons in both Rhodobacter capsulatus (44, 45) and Rhodobacter sphaeroides (19, 23, 24). For both organisms, there are two major cbb operons: the form I (cbb_I) operon, containing form I RubisCO genes (cbbL cbbS), is predominant under autotrophic growth conditions, and the form II (cbb_II) operon, containing the form II RubisCO gene (cbbM), is expressed under all growth conditions, albeit at enhanced levels during autotrophic growth. In R. sphaeroides, a LysR-type transcriptional activator gene (cbbR) is situated upstream and is divergently transcribed from the cbb_I operon, with CbbR activating expression of both cbb_I and cbb_II promoters (21). By contrast, R. capsulatus contains a separate cbbR gene upstream and divergently transcribed from each cbb operon (44, 45), with each operon being independently regulated by its cognate CbbR activator (46, 67). CbbR has also been identified and in some cases shown to control cbb expression in many other autotrophic organisms that employ the CBB pathway. These organisms include Rhodospirillum rubrum (16), Thiobacillus ferrooxidans (34), Xanthobacter flavus (39), Ralstonia eutropha (Alcaligenes eutrophus) (71), Chromatium vinosum (65), and Nitrobacter vulgaris (56). LysR-type transcriptional regulators generally utilize a metabolite or coinducer produced by the pathway they regulate (52). The nature of the signal(s) for CbbR has remained unclear, although some “organism-dependent specificity in the coinducer” has been suggested (59). For example, NADPH enhances the binding of X. flavus CbbR (64) while exhibiting no effect on CbbR from R. eutropha (26), R. capsulatus (P. Vichivanives and F. R. Tabita, unpublished results), or R. sphaeroides (J. M. Dubbs and F. R. Tabita, unpublished results).

Previous studies with R. capsulatus suggested that a metabolic signal(s) generated by enzymes encoded by the cbb_II genes, downstream of cbbP (PRK) but before cbbM (form II RubisCO), might contribute to the regulation of cbb expression in response to photoheterotrophic and photoautotrophic growth conditions (46, 62). For example, the RubisCO (cbbLS cbbM)-deficient strain SBI/II maximally expressed cbb promoter activities during photoheterotrophic growth which favor activation of the dimethylsulfoxide reductase (DMSOR) system, while a PRK (cbbP)-deficient strain exhibited wild-type control (62). SBI/II thus presented a useful tool to examine the nature of the metabolic signal(s) that could influence CbbR-mediated cbb expression. This work presents in vivo studies that suggest that a product generated by PRK contributes to CbbR-mediated control of cbb expression in an R. capsulatus SBI/II background.

MATERIALS AND METHODS

Bacterial strains and plasmids.

Relevant characteristics of R. capsulatus and Escherichia coli strains as well as plasmids used or constructed in this study are listed in Table 1. Insertional mutagenesis of cbbP, cbbT, and cbbR_II in R. capsulatus SBI/II and SBI is illustrated in Fig. 1.

TABLE 1.

Relevant characteristics of bacterial strains and plasmids used in this study

Strain or plasmid	Relevant characteristic(s)	Source or reference
R. capsulatus
SB1003	Rifr derivative of wild-type strain B10	74
SBI	cbbLS::Spr, derivative of SB1003	46
SBII	cbbM::Spr, derivative of SB1003	46
SBI/II	cbbLS::Spr, cbbM::Kmr derivative of SB1003	46
SBI/IIP	cbbP::Gmr derivative of SBI/II	This work
SBI/IIT	cbbT::Gmr derivative of SBI/II	This work
SBI/IIRII	cbbRII::Gmr derivative of SBI/II	This work
SBI/P	cbbP::Gmr derivative of SBI	This work
E. coli
JM109	recA1 endA1 gyrA96 thi hsdR17 supE44 relA1 Δ(lac-proAB)	73
S17-1λpir	Strr Tpr Spcr, λpir and mobilizing factors on chromosome	47
Plasmids
pUC19	Apr	68
pK18, pK19	Kmr, pUC derivatives	49
pUC1813, pUC1318	Apr, pUC derivative with hybrid multiple cloning site	33
pRK2013	Kmr, helper plasmid for triparental conjugation	18
pXLB	Tcr, R. capsulatus cbbI translational promoter fusion to lacZ	46
pXFB	Tcr, R. capsulatus cbbII translational promoter fusion to lacZ	46
pVKC1	Tcr, R. sphaeroides cbbI translational promoter fusion to lacZ, lacking cbbR	9
pVKD1	Tcr, R. sphaeroides cbbI translational promoter fusion to lacZ, containing cbbR	9
pRPS-1	Tcr, broad-host range expression vector containing Rhodospirillum rubrum cbbM promoter and cbbR gene	15
pRPS::FII(S2)-I	pRPS-1 containing the R. capsulatus form II RubisCO gene, cbbM	45
pRPS::429	pRPS-1 containing the R. sphaeroides aldolase and form I RubisCO genes, cbbAIcbbL cbbS	15
pRPS::329NΔ	pRPS-1 containing the R. sphaeroides form I RubisCO genes, cbbL cbbS	D. L. Falcone and F. R. Tabita, unpublished data
pRPS-MCS-3	Broad-host range expression vector containing the Rhodosprillum rubrum cbbM promoter and cbbR gene	Smith and Tabita, unpublished
pVK-cbbR	Complementation vector containing R. capsulatus cbbRII	46
pK18FIIB2.3	pK18 containing the R. capsulatus cbbRII′, cbbF, and cbbP′ genes on a 2.3-kb BamHI fragment	45
pK18FIIS4.4	pK18 containing the R. capsulatus pgm, qor, cbbRII, cbbF, and cbbP′ genes on a 4.4 SalI fragment	45
pK18FIIS4	pK18 containing R. capsulatus cbbP′, cbbT, cbbG, and cbbA′ genes on a 4-kb SalI fragment	45
pUC1813::RsPIA	pUC1813 containing the R. sphaeroides cbbPI gene on a 1-kb AvaI fragment	46
pTC5603	Tcr, mobilizable suicide vector containing multiple EcoRI sites	46
pJPTC	Tcr, mobilizable suicide vector containing a single EcoRI site located in the multiple cloning site	Y. Qian and F. R. Tabita, unpublished data
pUC1318-Gm	pUC1318 containing a Gmr cassette cloned as a 1.6 kb HindIII-SphI fragment from plasmid pFRK-1	Smith and Tabita, unpublished
pRPS-cbbP	pRPS-MCS-3 containing the R. sphaeroides cbbPI gene cloned as a 1-kb AvaI-SmaI fragment from plasmid pUC1813::RsPIA into XhoI-blunt-ended PstI sites	This work
pRPS-cbbTG	pRPS-MCS-3 containing the R. capsulatus cbbT and cbbG genes cloned as a 4-kb SalI fragment from plasmid pK18FIIS4 into XhoI site	This work
pRPS-cbbG	pRPS-MCS-3 containing the R. capsulatus cbbG gene cloned as a 1.8-kb KpnI-SalI fragment from plasmid p19-cbbGA′ into KpnI-XhoI sites	This work
p19-cbbGA′	pUC19 containing the R. capsulatus cbbT′, cbbG, and cbbA′ genes cloned as a 1.8-kb BamHI-SalI fragment from plasmid pK18FIIS4	This work
pK18-cbbPGm	pK18FIIB2.3 with a Gmr cassette cloned into the XhoI site of cbbP as a 1.6-kb SalI fragment from pUC1318-Gm	This work
pTC5603-cbbPGm	3.8-kb KpnI-XbaI fragment of plasmid pK18-cbbPGm containing the disrupted cbbP gene cloned into pTC5603	This work
p19-cbbT′	pUC19 containing the R. capsulatus cbbP′ and cbbT′ genes cloned as a 2.0-kb BamHI fragment from plasmid pK18FIIS4	This work
p19-cbbTGm	p19-cbbT′ with a Gmr cassette cloned into the XhoI site of cbbT as a 1.6-kb SalI fragment from pUC1318-Gm	This work
pJPTC-cbbTGm	p19-cbbTGm cloned into vector pJPTC by linearizing with EcoRI	This work
pK18-cbbRGm	pK18FIIS4.4 with a Gmr cassette cloned into the EcoRV site of cbbRII as a 1.6-kb SmaI fragment from pUC1318-Gm	This work
pJPTC-cbbRGm	pK18-cbbRGm cloned into vector pJPTC by linearizing with EcoRI	This work

FIG. 1.

cbb_I and cbb_II gene clusters of R. capsulatus. The sites of gene disruptions in the mutant strains are indicated. Solid arrows indicate direction of transcription; dotted arrows indicate orientation of the gentamicin resistance cassette used to inactivate the respective genes. The genes encoding the enzymes of the CBB pathway in R. capsulatus are organized in two operons, each preceded by a separate cbbR gene (encoding a LysR-type transcriptional regulator). cbbL cbbS encodes form I RubisCO, while cbbM encodes form II RubisCO. cbbM is clustered with genes encoding other enzymes of the CBB pathway: cbbF (fructose 1,6/sedoheptulose 1,7-bisphosphatase), cbbP (phosphoribulokinase), cbbT (transketolase), cbbG (glyceraldehydes-3-phosphate dehydrogenase), cbbA (fructose 1,6-bisphosphate aldolase), and cbbE (epimerase). The cbb genes are oriented with respect to other familiar genes in this organism: pgm, phosphoglucomutase; qor, quinol oxidoreductase; anfA, regulator for the alternative nitrogenase genes.

Media and growth conditions.

E. coli strains were grown aerobically on LB medium (1) at 37°C with appropriate antibiotic selection. Photoheterotrophic (0.4% [wt/vol] dl-malate) and photoautotrophic (1.5% CO₂, 98.5% H₂) cultures of R. capsulatus were grown anaerobically in Ormerod's medium as previously described (42) supplemented with 1 μg of thiamine/ml (15, 43). The nitrogen source during photoheterotrophic growth was provided as either 30 mM ammonia or 6.8 mM l-glutamate. The antibiotics used for selection of the R. capsulatus strains were as follows: rifampin, 100 μg/ml; kanamycin, 5 μg/ml; spectinomycin, 10 μg/ml; gentamicin, 10 μg/ml; and tetracycline, 3 μg/ml for stock cultures or 0.5 μg/ml for plasmid maintenance during photosynthetic growth and 0.3 μg/ml for screening exconjugant sensitivity. For E. coli, the concentrations of antibiotics for plasmid maintenance were as follows: ampicillin, 100 μg/ml; kanamycin, 10 μg/ml; gentamicin, 10 μg/ml; and tetracycline, 10 μg/ml. Dimethyl sulfoxide (DMSO) was utilized at 30 mM.

Cell extracts and enzyme assays.

Culture samples (10 to 30 ml) were harvested in late log phase (optical density at 660 nm, 0.9 to 1.2), washed in buffer (25 mM Tris-HCl, 1 mM EDTA [pH 8.0]), and disrupted by sonication. The resultant cell debris was removed by centrifugation for 15 min at 4°C. β-Galactosidase activity was measured as previously reported (46); the production of o-nitrophenol (40) was continuously measured by monitoring the increase in absorbance at 405 nm over 10 min. RubisCO activity was assayed as previously described by measuring RuBP-dependent ¹⁴CO₂ fixation into acid-stable material (70). PRK activity was measured by previously described protocols (57) except that ribulose 5-phosphate was generated from ribose 5-phosphate by the addition of 5 U of phosphoriboisomerase (Sigma Chemical, St. Louis, Mo.). The following CBB enzymatic activities were measured spectrophotometrically on a Beckman DU-70 spectrophotometer by using a coupled enzyme assay following established protocols: fructose 1,6-bisphosphatase (FBPase) (20), transketolase (60), and ribulose 5-phosphate-3-epimerase (35). Protein concentrations were determined with a protein assay dye binding reagent (Bio-Rad Laboratories, Hercules, Calif.) with bovine serum albumin as a standard.

Western immunoblot analysis of form I and form II RubisCO.

Samples were prepared and subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis as previously described (36). A low cross-linker ratio (acrylamide to bisacrylamide, 150/1) was utilized in 18% polyacrylamide gels. Following electrophoresis, proteins were transferred to polyvinylidene difluoride membranes (Immobilon-NC; Millipore, Bedford, Mass.) (63) using a Bio-Rad Transblot semidry cell. Polyclonal antibodies raised against Synechococcus sp. PCC 6301 RubisCO and R. sphaeroides form II RubisCO at a 1:3,000 dilution were utilized to follow the synthesis of form I and form II RubisCO, respectively. Western blots were developed by following protocols of the Vistra ECF fluorescent detection system (Amersham Corporation, Little Chalfont, Buckinghamshire, England). Blots were visualized with a Storm 840 imaging system (Molecular Dynamics, Sunnyvale, Calif.).

DNA manipulations and conjugation techniques.

All routine DNA manipulations were carried out following standard protocols (1). R. capsulatus chromosomal DNA was isolated as previously described (25). Southern blotting procedures were carried out with Hybond N+ (Amersham, Arlington Heights, Ill.) membranes following previously established protocols supplied with GeneScreen Plus membranes (NEN, Dupont, Boston, Mass.). Hybridizations, labeling of probes, and development of blots were conducted by following protocols of the Vistra ECF fluorescent detection system (Amersham Corporation). Blots were visualized with a Storm 840 imaging system. For complementation experiments, E. coli strain JM107 (73) containing the mobilizable helper plasmid pRK2013 (18) was used in triparental matings. To generate mutant strains, derivatives of plasmid pJP5603 were conjugated into R. capsulatus strain SBI/II or SBI using E. coli S17-λpir (11, 47).

Construction of R. capsulatus mutant strains by insertional mutagenesis. (i) Strain SBI/IIP (cbbLS cbbM cbbP).

cbbP was inactivated by insertion of a gentamicin resistance cassette as a 1.56-kb SalI fragment isolated from plasmid pUC1318-Gm into the unique XhoI site of cbbP in plasmid pK18FIIB2.3, generating plasmid pK18-cbbPGm. The 3.8-kb KpnI-XbaI fragment of pK18-cbbPGm, containing the disrupted cbbP gene, was cloned into the suicide vector pTC5603, generating plasmid pTC5603-cbbPGm. pTC5603-cbbPGm was mobilized into R. capsulatus strain SBI/II from E. coli strain S17-λpir. Of the 1,350 Gm^r exconjugants screened, 4 were found to be Tet^s. Southern blot analysis of chromosomal DNA indicated that each of the Gm^r Tet^s isolates resulted from double recombination.

(ii) Strain SBI/IIT (cbbLS cbbM cbbT).

In order to isolate a unique XhoI site in the cbbT gene, a 2.0-kb BamHI fragment containing a truncated cbbT gene was subcloned from plasmid pK18FIIS4 into plasmid pUC19, generating plasmid p19-cbbT′. Inactivation of cbbT was achieved by insertion of a gentamicin resistance cassette as a 1.56-kb SalI fragment isolated from pUC1318-Gm into the XhoI site of cbbT in p19-cbbT′, generating p19-cbbTGm. A cointegrate of p19-cbbTGm and the suicide plasmid pJPTC was formed after linearizing the DNA with EcoRI, resulting in plasmid pJPTC-cbbTGm. E. coli strain S17-λpir was used to mobilize pJPTC-cbbTGm into R. capsulatus strain SBI/II. Of the 700 Gm^r exconjugants screened, 1 was found to be Tet^s and confirmed to be a double crossover by Southern blot analysis of chromosomal DNA.

(iii) Strain SBI/IIRII (cbbLS cbbM cbbR_II).

Inactivation of cbbR_II was accomplished by insertion of a gentamicin resistance cassette as a 1.56-kb SmaI fragment from plasmid pUC1318-Gm into the unique EcoRV site of cbbR_II in plasmid pK18FIIS4.4, generating plasmid pK18-cbbRGm. pJPTC-cbbRGm was generated by forming a cointegrate between pK18-cbbRGm and the suicide vector pJPTC, after linearizing the DNA with EcoRI. pJPTC-cbbRGm was mobilized into R. capsulatus strain SBI/II by using E. coli strain S17-λpir. Of the 500 Gm^r exconjugants, 162 were found to be Tet^s; 8 were chosen for Southern blot analysis of chromosomal DNA, and all were confirmed to be double crossovers. One of these strains was chosen for all subsequent experiments.

(iv) Strain SBI/P (cbbLS cbbP).

Inactivation of cbbP was achieved by mobilizing plasmid pTC5603-cbbPGm from E. coli strain S17-λpir into R. capsulatus strain SBI. Of the 4,100 Gm^r exconjugants screened, 1 was found to be Tet^s. Southern blot analysis of chromosomal DNA indicated that the Gm^r Tet^s isolate resulted from double recombination.

RESULTS

Construction and phenotypes of cbb mutant strains.

Strains SBI/IIP (cbbLS cbbM cbbP), SBI/IIT (cbbLS cbbM cbbT), and SBI/IIRII (cbbLS cbbM cbbR_II) were constructed as described in Materials and Methods, with the orientation of the gentamicin resistance cassette as shown in Fig. 1. This cassette does not contain a transcription terminator. Thus, provided that the cassette is in the correct orientation, expression of genes downstream of the cassette can occur. The orientation of the Gm^r interposon inserted into cbbP was such that transcriptional read-through could occur into the cbb genes located directly downstream of cbbP (Fig. 1). To address the issue of polarity of the Gm interposon insertion in cbbP on downstream gene expression, a cbbP mutant was constructed in R. capsulatus form I (cbbLS) RubisCO deletion strain SBI (see Materials and Methods). For this strain, levels of form II RubisCO protein and activity were examined in order to detect expression of genes downstream of the Gm^r interposon insertion site. The levels of RubisCO and PRK activities in wild-type strain SB1003 were comparable to those of SBI (Fig. 2B, lanes 3 and 4) during photoheterotrophic growth with DMSO. In addition, form II RubisCO was synthesized and no form I RubisCO polypeptides were detected in extracts of either strain (Fig. 2, lanes 3 and 4). Strain SBI/P exhibited levels of RubisCO activity similar to those of wild-type strain SB1003 (Fig. 2B, lanes 3 and 5), and form II RubisCO polypeptides were detected during photoheterotrophic growth with DMSO (Fig. 2B, lane 5). As expected, no PRK activity or form I RubisCO polypeptide was detected in SBI/P (Fig. 2, lane 5). These data suggested transcriptional read-through of genes downstream of the Gm^r interposon insertion site in cbbP.

FIG. 2.

Immunoblot analysis and levels of RubisCO and PRK activities in strain SBI/P during photoheterotrophic growth. Antibodies raised against form I Synechococcus 6301 RubisCO (A) and R. sphaeroides form II RubisCO (B) were utilized. Approximately 5 μg of protein was added per lane (A and B). Levels of RubisCO activity (below the lanes) and PRK activity (below the lanes, in parentheses) (both in nanomoles per minute per milligram) were determined from three independent cultures grown photoautotrophically (1.5%CO₂-98.5% H₂) (lane 1) or photoheterotrophically in the presence of DMSO (lanes 3 to 5). Lanes 1, extract from wild-type strain SB1003; lanes 2, purified Synechococcus sp. strain 6301 RubisCO (A) and R. sphaeroides form II RubisCO (B); lanes 3, extract from SB1003; lanes 4, extract from SBI; lanes 5, extract from SBI/P. ND, not determined.

Under photoheterotrophic growth conditions, strains SBI/IIP, SBI/IIT, and SBI/IIRII behaved similarly to parent strain SBI/II; i.e., no growth was achieved in the absence of DMSO with ammonia as the nitrogen source. Additionally, no growth was achieved in SBI/P under photoheterotrophic conditions when ammonia was used as the nitrogen source; however, when cultures were supplemented with DMSO, growth of SBI/P comparable to that of parent strain SBI was observed.

Metabolic control of the CBB pathway in R. capsulatus.

Previous physiological studies suggested that a metabolic signal produced by enzymes encoded by the R. capsulatus cbb_II genes, in particular the product of a gene downstream of cbbP, but before cbbM (form II RubisCO), contributes to regulation of the CBB pathway (46, 62). To address this issue, insertional mutagenesis was performed on strain SBI/II, whereby specific genes (see orientation in Fig. 1) encoding PRK (cbbP), transketolase (cbbT), and the LysR-type transcriptional activator CbbR_II (cbbR_II) were inactivated (see Materials and Methods).

Form I RubisCO is not synthesized under photoheterotrophic growth conditions when malate is used as the carbon source in wild-type R. capsulatus (19, 45, 55). Accordingly, wild-type strain SB1003 exhibited no cbb_I promoter activity for all photoheterotrophic growth conditions examined. Strain SBI/II exhibited cbb_I promoter activity during photoheterotrophic growth and levels were substantially enhanced during growth with DMSO (Fig. 3A) as previously reported (62). Indeed, promoter activities in SBI/II approached the levels obtained in the wild type under photoautotrophic growth conditions, which is known to be a growth condition that favors maximum cbb transcription (46, 61, 62). However, when a knockout mutation in the cbbP gene was constructed in this strain (creating strain SBI/IIP), as observed in the wild type, no cbb_I promoter activity was noted for all photoheterotrophic growth conditions (Fig. 3A). Strain SBI/IIRII exhibited levels of cbb_I promoter activity approximately fourfold lower than those of strain SBI/II for malate-ammonia-DMSO and malate-glutamate growth conditions and no promoter activity for growth with malate-glutamate-DMSO (Fig. 3A). Compared to strain SBI/II, strain SBI/IIT also exhibited high levels of cbb_I promoter activity during growth with malate-ammonia-DMSO. By contrast, however, levels were greatly enhanced compared to wild-type levels during growth in the absence of DMSO (malate-glutamate) and decreased 4.8-fold during malate-glutamate-DMSO growth (Fig. 3A). For cbb_II promoter activity, for all photoheterotrophic growth conditions tested, activity was readily detected in wild-type strain SB1003. However, SBI/II exhibited a 13- to 14-fold induction of cbb_II promoter activity when cultures were supplemented with DMSO (Fig. 3B) (62). Like the cbb_I promoter results, SBI/IIP regained wild-type control of its cbb_II promoter activity during photoheterotrophic growth, although slightly less activity was observed with malate-glutamate-DMSO than for the wild-type (Fig. 3B). These results were similar to what occurred in PRK deletion strains of wild-type R. capsulatus (62). SBI/IIT enhanced cbb_II promoter activity to levels comparable to those for SBI/II under photoheterotrophic growth conditions when cultures were supplemented with DMSO. However, for this strain, enhanced levels of promoter activity were obtained in cultures grown in the absence of DMSO (malate-glutamate) (Fig. 3B). Levels of cbb_II promoter activity in strain SBI/IIRII were considerably reduced for all the photoheterotrophic growth conditions tested (Fig. 3B) and were at levels observed with R. capsulatus cbbR_II strains (46, 67).

FIG. 3.

cbb_I::lacZ (A) and cbb_II::lacZ (B) promoter activity during photoheterotrophic growth of R. capsulatus strains. β-Galactosidase activities were determined from three to nine independent cultures assayed in duplicate. NG, no growth during photoheterotrophic conditions with ammonia as the nitrogen source in the absence of DMSO.

The promoter-fusion studies suggested that the levels of activity of individual enzymes encoded by the cbb_II operon genes might also be affected in the different strains. Thus, under photoheterotrophic growth conditions in the presence of DMSO, strain SBI/II possessed levels of PRK activity that were comparable to the induced levels observed in wild-type strain SB1003 after photoautotrophic (1.5% CO₂-98.5% H₂) growth (Table 2). Enhanced levels of PRK activity were also found in strain SBI/IIT; moreover, SBI/II and SBI/IIT also had enhanced levels of FBPase activity compared to SB1003 after photoheterotrophic growth in the presence of DMSO (Table 2). However, like the cbb_II promoter activity results (Fig. 3B), a knockout of cbbP (SBI/IIP) resulted in wild-type levels of FBPase and transketolase activity during photoheterotrophic growth with DMSO; decreased levels of FBPase activity were observed in SBI/IIRII, and near-wild-type levels of transketolase activity were also observed (Table 2). The wild-type level of transketolase activity in SBI/IIRII is probably due to the fact that R. capsulatus and R. sphaeroides contain an additional chemoheterotrophic transketolase, TktA, which plays a role in the CBB pathway in the absence of CbbT (5, 7); thus, perhaps TktA contributes to the levels of transketolase activity in SBI/IIRII. Somewhat lower levels of epimerase activity were observed in SBI/II and SBI/IIP than in SB1003 during photoheterotrophic growth with DMSO (Table 2); whether the levels of epimerase activity reflect transcription from a third cbb promoter (43) or activity from an additional epimerase remains to be established.

TABLE 2.

CBB enzymatic activities during photoheterotrophic growth with ammonia and DMSO

Strain	Sp act (nmol/min/mg)a
	FBPase (cbbF)	PRK (cbbP)	Transketolase (cbbT/tktA)	RubisCO (cbbM)	Epimerase (cbbE)
SB1003	19 ± 5	101 ± 24	230 ± 42	39 ± 8	431 ± 80
SB1003b	ND	1,446 ± 148	ND	153 ± 15	ND
SBI/II	211 ± 39	1,122 ± 303	748 ± 159	0 ± 0	199 ± 27
SBI/II::pRPSFIIS2-1	69 ± 15	185 ± 36	228 ± 57	11 ± 4	208 ± 15
SBI/IIP	15 ± 4	0 ± 0	203 ± 46	0 ± 0	316 ± 69
SBI/IIP::pRPScbbP	16 ± 2	31 ± 5	ND	0 ± 0	ND
SBI/IIP::pRPScbbTG	17 ± 3	0 ± 0	353 ± 87	0 ± 0	ND
SBI/IIP::pRPScbbG	14 ± 4	0 ± 0	ND	0 ± 0	ND
SBI/IIP::pRPS-429	20 ± 4	0 ± 0	ND	5 ± 1	ND
SBI/IIP::pRPS-329NΔ	9 ± 3	0 ± 0	ND	7 ± 4	ND
SBI/IIT	424 ± 64	1,362 ± 249	ND	0 ± 0	ND
SBI/IIRII	7 ± 2	4 ± 3	269 ± 54	0 ± 0	ND

^aValues are averages ± standard errors from three to five independent cultures. ND, not determined.

^bPhotoautotrophically grown (1.5% CO₂-98.5% H₂-ammonia).

Vectors containing R. sphaeroides or R. capsulatus cbb genes on a broad-host-range plasmid (containing the Rhodospirillum rubrum cbbM promoter and cbbR gene) were constructed (Table 1). When each plasmid (pRPS-cbbP, pRPS-cbbTG, pRPS-cbbG, pRPS-429, and pRPS-329NΔ), containing the respective cbb gene(s), was mated into SBI/IIP, levels of FBPase activity were not significantly altered during photoheterotrophic growth with DMSO (Table 2). SBI/II, containing pRPSFIIS2-1, with the cbbM gene from R. capsulatus (encoding form II RubisCO), possessed decreased levels of PRK and transketolase activity, to near wild-type levels; moreover, levels of FBPase activity decreased significantly (Table 2). These results indicated that the restoration of a functional CBB pathway in SBI/II (pRPSFIIS2-1) (containing R. capsulatus form II RubisCO) diminished the accumulation of some important metabolite that influenced regulation when RubisCO activity was blocked. This putative compound would appear to substantially contribute to up-regulated gene expression in SBI/II during photoheterotrophic growth with DMSO (62).

R. capsulatus SBII, which has no form II RubisCO, had been shown to synthesize form I RubisCO during photoheterotrophic growth, unlike wild-type strains (46). Thus, in order to compensate for the absence of form II RubisCO, a signal is processed to somehow allow SBII to synthesize compensatory levels of form I RubisCO. Consistent with these observations, cbb_I promoter activity became detectable in SBII after photoheterotrophic growth with ammonia (Table 3). This cbb_I promoter activity was observed under photoheterotrophic growth conditions with either ammonia or glutamate as the nitrogen source in the absence or presence of DMSO. Wild-type strain SB1003, as previously shown (62), exhibited no cbb_I promoter activity (Table 3). SBII enhanced levels of cbb_II promoter activity 7.2-fold compared to SB1003 during malate-ammonia growth, while levels were enhanced 3.4-fold during malate-ammonia-DMSO growth and 2.4- to 2.5-fold during malate-glutamate growth regardless of the presence of DMSO (Table 3). SBII also showed enhanced levels of PRK activity (2.5- to 2.6-fold) compared to SB1003 during photoheterotrophic growth in the presence or absence of DMSO (data not shown). Similar to the situation with SBI/II (Table 2), upon the restoration of a complete CBB pathway via complementation with the R. capsulatus cbbM gene (encoding form II RubisCO) using plasmid pRPS::FII(S2)-I, decreased levels of PRK activity were obtained (data not shown).

TABLE 3.

Levels of cbb_I and cbb_II promoter activities in strain SBII during photoheterotrophic growth

Promoter	Strain	β-Galactosidase activity (nmol/min/mg) ona
		MA	MAD	MG	MGD
cbbI::lacZ	SB1003	0	0	0	0
	SBII	10 ± 3	8 ± 1	10 ± 2	5 ± 1
cbbII::lacZ	SB1003	137 ± 6	187 ± 82	197 ± 43	154 ± 39
	SBII	985 ± 230	629 ± 58	496 ± 12	364 ± 45

^aMA, malate-ammonia; MAD, malate-ammonia-DMSO; MG, malate-glutamate; MGD, malate-glutamate-DMSO. Data are means ± standard errors from three independent cultures.

R. sphaeroides CbbR responds to the R. capsulatus SBI/II background.

Control mechanisms involved in phototrophic redox homeostasis and cbb expression in R. capsulatus differ from those in R. sphaeroides (61, 62, 67). Although the cbb_I and cbb_II operons are independently regulated in R. sphaeroides, CbbR has been shown to be involved in the regulation of both operons (21, 32). By contrast, each cbb operon is independently regulated by its cognate CbbR activator in R. capsulatus (46, 67). Moreover, in R. sphaeroides, cbb_I and cbb_II promoters are both expressed during photoheterotrophic growth (9, 72), while in R. capsulatus, only the cbb_II promoter is expressed during photoheterotrophic growth with malate as the carbon source (45, 46, 62). In view of these differences, and to determine whether the R. sphaeroides cbb_I promoter and its cognate cbbR gene were subject to the same regulatory signals as the R. capsulatus system, heterologous expression experiments were performed. An R. sphaeroides translational cbb_I::lacZ promoter fusion vector (also containing the R. sphaeroides cbbR gene) (9) was examined in the R. capsulatus SBI/II and SBI/II mutant strain background. β-Galactosidase activity levels were enhanced three- to fourfold in the presence of DMSO when R. capsulatus SBI/II was grown under photoheterotrophic growth conditions with glutamate as the nitrogen source (Fig. 4). Likewise, no significant R. sphaeroides cbb_I promoter activity was found in wild-type strain SB1003 and strain SBI/IIP under the tested photoheterotrophic growth conditions (Fig. 4). In contrast to the lack of R. capsulatus cbb_I promoter activity (Fig. 3A), β-galactosidase activity was obtained with the R. sphaeroides cbb_I promoter fusion in SBI/IIRII during photoheterotrophic growth with malate-glutamate-DMSO, to similar levels during growth with malate-ammonia-DMSO. Levels of activity were decreased 12- to 13-fold in SBI/IIRII in the absence of DMSO (Fig. 4). No significant β-galactosidase activity was detected in any strain when R. sphaeroides cbb_I promoter fusion vector pVKC1 (lacking a complete R. sphaeroides cbbR gene) (9) was used (data not shown). Aside from indicating that neither endogenous CbbRI or CbbRII R. capsulatus proteins could activate transcription of the R. sphaeroides cbb_I promoter, these results are important because they show that the R. sphaeroides cbb_I promoter and its cognate CbbR protein were subject to the same control as the R. capsulatus cbb promoters in the R. capsulatus SBI/II background. Thus, the metabolic signal produced in SBI/II is able to affect both homologous (R. capsulatus) and heterologous (R. sphaeroides) CbbR-mediated cbb transcription.

FIG. 4.

Promoter activity using an R. sphaeroides cbbR-cbb_I::lacZ promoter fusion (pVKD1) (9) in R. capsulatus strains during photoheterotrophic growth. β-Galactosidase activities were determined from three independent cultures assayed in duplicate. NG, no growth in the absence of an alternative electron acceptor.

Finally, levels of PRK activity were monitored after complementation experiments were performed with SBI/IIRII during photoheterotrophic growth in the presence of DMSO (Table 4). Consistent with previous findings (Table 2), levels of PRK activity were enhanced (12-fold) in SBI/II compared to the wild type under photoheterotrophic growth conditions in the presence of DMSO. Considerably reduced levels were observed in SBI/IIRII, and of course no activity was found in SBI/IIP. R. sphaeroides cbbR was shown to complement and restore the ability of SBI/IIRII to provide SBI/II levels of PRK activity, since the addition of pVKD1 (with a complete R. sphaeroides cbbR gene), but not pVKC1 (with a truncated R. sphaeroides cbbR gene), yielded high levels of PRK activity. Likewise, complementation with R. capsulatus cbbR_II resulted in high levels of PRK activity while addition of the Rhodospirillum rubrum cbbR gene (Table 4) or the R. capsulatus cbbR_I gene did not (data not shown).

TABLE 4.

Complementation of R. capsulatus strain SBI/IIRII during photoheterotrophic growth with DMSO (malate-ammonia-DMSO)

Strain	PRK activity (nmol/min/mg)a
SB1003	76 ± 12
SBI/II	929 ± 195
SBI/IIP	0
SBI/IIRII	4 ± 2
SB1003::pVKC1b	90 ± 15
SBI/II::pVKC1	824 ± 160
SBI/IIP::pVKC1	0
SBI/IIRII::pVKC1	4 ± 3
SB1003::pVKD1c	40 ± 4
SBI/II::pVKD1	724 ± 191
SBI/IIP::pVKD1	0
SBI/IIRII::pVKD1	1,086 ± 204
SBI/IIRII::pVK-CbbRd	628 ± 93
SBI/IIRII::pRPS-MCS3e	4 ± 1

^aPRK activity was determined from three to five independent cultures.

^bContains truncated R. sphaeroides cbbR gene.

^cContains intact R. sphaeroides cbbR gene.

^dContains intact R. capsulatus cbbR_II gene.

^eContains intact Rs. rubrum cbbR gene.

DISCUSSION

Insight as to metabolic factors that might control the function of the CBB pathway is limited due to the complexity of the system; the CBB cycle involves up to 13 enzymes acting on 16 metabolites in a complex network of reactions, and it is also dependent on a ready source of ATP and reducing equivalents, as well as the removal of metabolites to be used in further biosynthetic pathways involved in cell growth (48). PRK has been suggested to be the target enzyme for in vivo control of the rate of CO₂ fixation (8); indeed, changes in the flux of CO₂ fixation are thought to be regulated in part through changes in the levels of RuBP, the product of PRK activity, in leaves of sugar beet (53).

LysR-type transcriptional activators generally require a coinducer, and this effector ligand is often a metabolite of the pathway regulated (52). With respect to the cbb operons of R. capsulatus, previous studies have shown that these operons belong to independent CbbR regulons (44, 46, 67). It has also been suggested that under phototrophic growth conditions, different inducer molecules could bind to the respective CbbR proteins or, alternatively, that CbbR_I and CbbR_II could bind the same inducer molecule but with different affinities (46). The present investigation strongly indicated a potential regulatory role of a metabolite generated by PRK function (either ADP, RuBP, or a derivative thereof). Presumably, this compound might act as a specific effector in mediating the ability of CbbR_I and CbbR_II to influence control of cbb expression in SBI/II. Previous physiological studies (46, 62) and genetic studies reported here have thus stimulated us to direct attention to the ability of such metabolites to influence CbbR function in vitro. It is clear, however, from the present studies that the enhanced levels of cbb_II promoter activity in SBI/II were dependent on CbbR_II, since inactivation of the cbbR_II gene in SBI/II resulted in decreased basal levels of cbb_II expression. Levels of PRK activity in SBI/IIRII were approximately 4% of wild-type levels during photoheterotrophic growth in the presence of DMSO (Table 2). Perhaps, the lower levels of PRK activity in SBI/IIRII contributed to the approximate fourfold reduction of cbb_I promoter activity in this strain during photoheterotrophic growth. It is also apparent from mutagenesis studies presented here and elsewhere (62) that a separate metabolic signal (besides RuBP, ADP, or a derivative thereof) might be involved in the control of cbb_II expression during photoheterotrophic growth, since SBI/IIP exhibited wild-type control of cbb_II expression (Fig. 3B); additionally, cbbP strains generated by a polar mutation exhibited wild-type control of cbb_II promoter activity during photoheterotrophic growth (62). Obviously, there exists an alternative signal that must be involved in CbbR_II-mediated regulation of cbb_II expression during photoheterotrophic growth. Presumably, this hypothetical signal is generated via processes separate from the function of structural genes downstream of and including cbbP. Again, if one assumes that the regulatory compound is generated by the CBB pathway, a potential metabolite candidate could possibly be the product generated by CbbF (fructose 1,6-bisphosphatase), namely, fructose 6-phosphate, since the cbbF gene is upstream of cbbP and is not inactivated by the polar effect of a cbbP mutation. Additional work obviously needs to be completed to reinforce these hypotheses. Certainly, the ability of LysR-type transcriptional activators to bind two separate inducer molecules is not unprecedented. For example, BenM, a LysR-type transcriptional activator involved in the regulation of benzoate degradation in Acinetobacter sp. strain ADP1, has been shown to utilize two compounds of the pathway, benzoate and cis,cis-muconate, as effector molecules (6).

It is clear that both cbb operons are independently regulated in R. capsulatus 46, 62; this study) and R. sphaeroides (24, 27, 28). The present study shows that cbb_I and cbb_II promoter activities of R. capsulatus are both substantially enhanced during photoheterotrophic growth in the presence of DMSO in the RubisCO-deficient strain SBI/II. Indeed, the enhanced levels observed reached those obtained by the wild-type during photoautotrophic (1.5% CO₂-98.5% H₂) growth (62). It is thus apparent that a signal was generated in SBI/II that led to the deregulation and control of cbb_I and cbb_II expression under photoheterotrophic growth conditions in the presence of DMSO. Furthermore, this enhanced cbb_I and cbb_II promoter activity was lost when cbbP was inactivated (Fig. 3 and Table 2), causing a regain of the wild-type pattern of control. Consequently, we hypothesize that a metabolic signal generated by PRK function (ADP, RuBP, or a metabolite derived from RuBP) serves as a positive effector to enable CbbR_I and CbbR_II to enhance cbb gene expression in SBI/II. Such an effector also provides a common metabolic signal for communication between the two operons. Because effector compounds for LysR type regulators are usually metabolites unique to the pathway, we suggest that RuBP, or something directly derived from RuBP, is the most likely candidate for the coinducer or effector. The coincidence of the levels of cbb expression with levels of PRK function in SBI/II is certainly supportive of this idea. Moreover, upon restoration of a functional CBB pathway (by complementation with the R. capsulatus form II RubisCO gene), levels of PRK activity decreased 6.1-fold, which presumably diminished the intracellular levels of a metabolite that accumulated when RubisCO activity was blocked; this could contribute to the 3.1-fold decrease in FBPase activity and the 3.3-fold decrease in transketolase activity exhibited by SBI/II containing a functional form II RubisCO (Table 2). Additionally, levels of PRK activity appeared to be coincident with cbb induction in response to a photoautotrophic growth environment (1.5% CO₂-98.5% H₂) in SBI/II containing form II RubisCO (data not shown). From these findings, it appears that PRK function, and the levels of RuBP (or a metabolite derived from RuBP), appeared to contribute to CbbR-mediated cbb_I and cbb_II expression in SBI/II.

Further communication between the two cbb operons is illustrated by R. capsulatus SBII; unlike the wild type, SBII synthesized form I RubisCO under photoheterotrophic growth conditions with malate as the carbon source (46). This compensatory increase in the expression of the cbb_I operon in SBII was reflected by enhanced levels of cbb_I promoter activity, compared to the wild type, during photoheterotrophic growth with either ammonia or glutamate as the nitrogen source in the absence or presence of DMSO (Table 3). Somehow a signal was received in SBII to allow the organism to compensate for the absence of form II RubisCO. This compensation resulted in the synthesis of form I RubisCO under conditions where the cbb_I operon was normally not functional. It should be noted, however, that no signal for compensatory form I RubisCO synthesis was evident in various cbbP mutant strains, where a polar effect on cbbM also causes form II RubisCO not to be synthesized (46, 62). SBII also exhibited enhanced levels of cbb_II promoter and enhanced PRK activity levels compared to the wild type under photoheterotrophic growth conditions (Table 3; also data not shown). Again, it is tempting to speculate that a signal for the compensatory effect on cbb_I expression in SBII is, in part, mediated by enhanced levels of a metabolite generated by PRK function during photoheterotrophic conditions, as is the case for SBI/II. In R. sphaeroides, the intracellular level of RuBP or a metabolite derived from RuBP also appears to be important for cbb expression control. Thus, similar to R. capsulatus, when cbbM was disrupted by a polar mutation in cbbA_II, levels of PRK activity were enhanced over wild-type levels during photoheterotrophic growth in R. sphaeroides (27; D. L. Falcone, J. L. Gibson, and F. R. Tabita, unpublished data). Finally, recent experiments performed by S. A. Smith in this laboratory indicated that enhanced levels of cbb_I and cbb_II promoter activities were manifest in an R. sphaeroides RubisCO deficient strain during photoheterotrophic growth in the presence of DMSO (malate-ammonia-DMSO) (S. A. Smith and F. R. Tabita, unpublished observations). Since we have shown that R. sphaeroides cbb_I promoter activity was up-regulated and dependent on R. sphaeroides CbbR in an R. capsulatus SBI/IIRII background, similar to R. capsulatus cbb_I, there appears to be a common and important role for the product of PRK in both organisms. The different abilities of R. sphaeroides CbbR, R. capsulatus CbbR_I or CbbR_II, and Rhodospirillum rubrum CbbR to complement R. capsulatus SBI/IIRII might reflect differences in the specificity of the coinducer(s) used, the nature of the DNA binding motifs, or affinities for binding motifs or might even reflect the requirement for specific interactions with additional factors. Obviously additional studies need to be performed in order to resolve these multiple possibilities; however, preliminary in vitro studies indicate that RuBP might have differential capabilities to influence the binding of CbbR_I and CbbR_II to its cognate promoter sites (66).

Finally, there are both known and unknown additional regulatory components and factors that are involved in controlling expression of the CBB system (9, 10, 67). For example, there is the RegB/A(PrrB/A) two-component signal transduction system, originally shown to control transcription of key operons involved in photosystem biosynthesis in R. capsulatus (30, 41), R. sphaeroides (13), and probably other nonsulfur purple photosynthetic bacteria (38) via a mechanism that shows organism-dependent specificity (3, 13, 14, 41, 54). More recently, RegB/A(PrrB/A) was found to function as a global two-component regulatory system (31), as it controls, among other processes, CO₂ fixation in R. capsulatus (67) and R. sphaeroides (10, 50), nitrogen fixation in R. capsulatus (12, 61), R. sphaeroides (31, 51), Rhodospirillum rubrum (31), and Bradyrhizobium japonicum (4), H₂ oxidation in R. capsulatus (12), and the regulation of formaldehyde oxidation in R. sphaeroides (2). With regard to the cbb system of R. capsulatus, Reg/Prr is involved in activation of cbb_I promoter expression as well as maximal expression of the cbb_II promoter under photoautotrophic growth conditions (67). It was further suggested that alteration of the oxidation-reduction potential of the ubiquinone pool, by activation of the DMSOR system, enables a redox signal to be transmitted to the Reg/Prr system, which in turn regulates the expression of key operons involved in phototrophic metabolism, including the cbb regulons of R. capsulatus (62). Perhaps, the enhanced levels of cbb expression observed during photoheterotrophic growth with DMSO in SBI/II is a result of the cooperative regulation by both the LysR-type regulatory activators CbbR_I and CbbR_II and the global response regulator RegA. It is thus conceivable that a synergistic interaction between CbbR and RegA occurs at the respective cbb_I and cbb_II promoters in SBI/II; the binding of RuBP (or a derivative thereof) to CbbR_I and CbbR_II could then somehow enhance the effects of RegA control at these promoters. Synergistic control of gene expression by a response regulator of a two-component regulatory system and a LysR-type transcriptional regulator is certainly not unprecedented and has been shown, for example, to be important for control of the eps operon of Ralstonia solanacearum (29). Ultimately, further in vitro studies will need to be performed to address these and other possibilities.