Pyruvate oxidase contributes to the aerobic growth efficiency of Escherichia coli

Ahmed M. Abdel-Hamida,1, Margaret M. Attwood1 and John R. Guest1

The Krebs Institute for Biomolecular Research, Department of Molecular Biology and Biotechnology, University of Sheffield, Western Bank, Sheffield S10 2TN, UK1

Author for correspondence: John R. Guest. Tel: +44 114 2224406. Fax: +44 114 2222787. e-mail: j.r.guest{at}sheffield.ac.uk


   ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
The metabolic importance of pyruvate oxidase (PoxB), which converts pyruvate directly to acetate and CO2, was assessed using an isogenic set of genetically engineered strains of Escherichia coli. In a strain lacking the pyruvate dehydrogenase complex (PDHC), PoxB supported acetate-independent aerobic growth when the poxB gene was expressed constitutively or from the IPTG-inducible tac promoter. Using aerobic glucose-limited chemostat cultures of PDH-null strains, it was found that steady-states could be maintained at a low dilution rate (0·05 h-1) when PoxB is expressed from its natural promoter, but not at higher dilution rates (up to at least 0·25 h-1) unless expressed constitutively or from the tac promoter. The poor complementation of PDH-deficient strains by poxB plasmids was attributed to several factors including the stationary-phase-dependent regulation of the natural poxB promoter and deleterious effects of the multicopy plasmids. As a consequence of replacing the PDH complex by PoxB, the growth rate (µmax), growth yield (Ymax) and the carbon conversion efficiency (flux to biomass) were lowered by 33%, 9–25% and 29–39% (respectively), indicating that more carbon has to be oxidized to CO2 for energy generation. Extra energy is needed to convert PoxB-derived acetate to acetyl-CoA for further metabolism and enzyme analysis indicated that acetyl-CoA synthetase is induced for this purpose. In similar experiments with a PoxB-null strain it was shown that PoxB normally makes a significant contribution to the aerobic growth efficiency of E. coli. In glucose minimal medium, the respective growth rates (µmax), growth yields (Ymax) and carbon conversion efficiencies were 16%, 14% and 24% lower than the parental values, and correspondingly more carbon was fluxed to CO2 for energy generation. It was concluded that PoxB is used preferentially at low growth rates and that E. coli benefits from being able to convert pyruvate to acetyl-CoA by a seemingly wasteful route via acetate.

Keywords: pyruvate metabolism, pyruvate dehydrogenase complex, growth rate, biomass yield and energetics, acetyl-CoA synthetase, continuous culture

Abbreviations: ACK, acetokinase; ACS, acetyl-CoA synthetase; PDH, pyruvate dehydrogenase; PDH-E1p, E1 component of the pyruvate dehydrogenase complex; PDHC, pyruvate dehydrogenase complex; PFL, pyruvate formate-lyase; PoxB, pyruvate oxidase; PTA, phosphotransacetylase; ThDP, thiamin diphosphate

a Present address: Botany Department, Faculty of Sciences, Minia University, Minia, Egypt.


   INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
The pyruvate dehydrogenase complex (PDHC) and pyruvate formate-lyase (PFL) are essential enzymes for the oxidation of pyruvate to acetyl-CoA by Escherichia coli during aerobic respiratory growth and anaerobic fermentative growth, respectively. They also perform anabolic functions by providing acetyl units for biosynthetic purposes and their absence generates a requirement for acetate to support aerobic or anaerobic growth on glucose (the Ace- phenotype). In contrast pyruvate oxidase (PoxB), which converts pyruvate directly to acetate, is reported to be a non-essential aerobic stationary phase enzyme of uncertain physiological function (Chang & Cronan, 1983 ; Grabau & Cronan, 1984 ; Chang et al., 1994 ). Two types of PoxB-deficient mutant have been detected by screening for impaired production of radioactive CO2 from labelled pyruvate. The poxA mutants retained 10–15% of normal pyruvate oxidase activity and grew slightly slower than the parental strain in acetate-supplemented rich and minimal media (Chang & Cronan, 1982 ). They were originally thought to lack a regulator of poxB expression but poxA mutations are now known to generate an extensive pleiotropic phenotype by inactivating additional lysyl-tRNA synthetases which are needed for efficient translation of the poxB transcript in Salmonella typhimurium and E. coli (Van Dyk et al., 1987 ; Kaniga et al., 1998 ). Other mutants (poxB) having defects in the pyruvate oxidase structural gene retained <1 to 20% of normal PoxB activity (Chang & Cronan, 1983 ). The poxB mutants were originally isolated as double mutants by screening for derivatives using a PDH null strain ({Delta}aroP-pdhR-aceEF) that had lost the ability to produce micro-colonies after prolonged incubation (6–8 d) in the absence of acetate. Endogenous acetate production has been detected in the PDH null strain (LeMaster & Cronan, 1982 ) but it is blocked by poxB (and poxA) inactivation. Clearly this source of acetate is insufficient to support normal aerobic growth in the absence of the PDH complex, and it is not required in the presence of an active PDHC because poxB single mutants grew normally (Chang & Cronan, 1983 ).

PoxB is a peripheral membrane flavoprotein which catalyses the decarboxylation of pyruvate to acetate and CO2 with the reduction of flavin adenine dinucleotide, FAD (Gennis & Hager, 1976 ). The enzyme contains four identical subunits (62 kDa) each with single tightly bound FAD and loosely bound thiamin diphosphate (ThDP) and Mg2+. PoxB is strongly activated by a variety of phospholipids which increase the maximum velocity about 20-fold and lower the concentration of pyruvate needed for enzyme saturation about 10-fold (Gennis & Hager, 1976 ; Mather et al., 1982 ). Moreover, this activation can be mimicked by limited chymotryptic proteolysis leading to the cleavage of a 3 kDa peptide from the C-terminus of the enzyme (Russell et al., 1977 ; Recny et al., 1985 ). The poxB gene has been cloned and sequenced (Grabau & Cronan, 1984 , 1986a ) and this has greatly facilitated recent studies on both the structure–function relationships (including the lipid activation) of PoxB (Grabau et al., 1989 ; Wang et al., 1991 ; Chang & Cronan, 1997 ) and on the transcriptional regulation of the poxB gene (Chang et al., 1994 ). The regulatory studies indicated that PoxB activity and poxB-lacZ expression reach maximal values in early stationary phase and both are completely dependent on the rpoS-encoded sigma factor (RpoS, {sigma}38 or {sigma}S) that is required for transcribing many genes that are induced in stationary phase. PoxB was also shown to be expressed (albeit at a lower rate) and active, during anaerobic growth. It was suggested that PoxB might be important for survival during the transition between exponential and stationary phases and that it might serve as a source of acetyl units under microaerobic conditions where both PDHC and PFL would function poorly (Chang et al., 1994 ).

The present physiological studies were initiated to (a) establish whether an up-regulated PoxB can replace the PDH complex in providing the acetyl units needed for good aerobic growth in unsupplemented medium, (b) assess the energetic consequences of such a replacement, and (c) determine whether and to what extent PoxB normally contributes to the aerobic growth efficiency of E. coli.


   METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Bacterial strains, plasmids, and molecular-genetic methods.
The strains of E. coli K-12 and plasmids used in this work are listed in Table 1. The strains include: W3110, the parental strain; DH5{alpha}, for routine transformation; TG1 the lacIq strain used to repress the tac promoter in multicopy situations; JC7623, for chromosomal gene replacement; YYC877, the source of an insertionally inactivated poxB gene (poxB::kanR); and isogenic derivatives of W3110 specifically constructed for the present work (see below). Unless stated otherwise, the isogenic derivatives of W3110 and plasmids behaved reproducibly and retained their original properties after culturing in the experiments to be described. The cloning vectors were pUC18, pBR322, pSU38 and pMAK705 and the sources of specific genes, promoters and antibiotic resistance cassettes were: pCG5 (poxB+); ptac-85 (Ptac); pGS367 (Ptac-aceEF-lpdA); pUC4K (kanR); and pHP45 (camR). The isolation and manipulation of DNA and other basic molecular-genetic procedures are described by Sambrook et al. (1989) . Two proof-reading polymerases, Pfu and Pwo, were used for PCR amplification (Promega) with specific primers synthesized in the Krebs Institute. P1vir-mediated transduction was performed according to Miller (1992) .


View this table:
[in this window]
[in a new window]
 
Table 1. Bacterial strains and plasmids

 
Substitution of the chromosomal poxB promoter (PpoxB) by Ptac.
The natural poxB promoter was replaced by the IPTG-inducible tac promoter in JRG3356 (Ptac-poxB) using a multi-step procedure for constructing the upox-kanR-Ptac-poxB' cassette containing a selectable kanR-Ptac fragment flanked by a promoterless upox fragment (derived from the region immediately upstream of the poxB promoter) and the promoterless 5' end of the poxB structural gene (poxB') in pGS860 (Fig. 1a), for ultimate transfer to W3110. The 2·0 kb upstream fragment (upox) extending from the vector junction to position -45 in the poxB promoter was first PCR-amplified from pCG5 DNA with primers incorporating flanking BamHI and PstI sites, and cloned in pUC18. A 1·3 kb PstI fragment containing a pUC4K-derived kanR cassette was inserted at the corresponding site of the intermediate plasmid to generate pGS858 from which a 3·3 kb BamHI–SphI upox-kanR fragment could be recovered for subsequent subcloning (Fig. 1a). A promoterless 0·95 kb poxB' fragment was likewise PCR-amplified from pCG5 DNA with primers incorporating BamHI and SalI sites, and cloned immediately downstream of the tac promoter of ptac-85 to generate pGS859 from which a 1·05 kb SphI–SalI fragment containing the Ptac-poxB' region could be recovered (Fig. 1a). The two composite fragments were simultaneously joined and subcloned in a unique orientation between the BamHI and SalI sites in the multicloning site of pMAK705 to generate pGS860 containing the upox-kanR-Ptac-poxB' cassette (Fig. 1a). The pMAK705 vector was chosen because its thermosensitive replication offered an alternative strategy for chromosomal replacement to that involving JC7623. The structure of pGS860 was verified by diagnostic restriction analysis during which the orientation of the kanR gene was defined by locating the asymmetric XhoI site (Fig. 1a). Chromosomal replacement of Ppox by the kanR-Ptac construct was achieved according to Oden et al. (1990) by transforming JC7623 (recBC sbcC) with pGS860 and screening KnR transformants for the desired KnR CmS products. A P1vir lysate of one such strain was used to generate KnR transductants of W3110 and one representative strain designated JRG3356 was chosen out of several that (a) generated a 3·3 kb (kanR-Ptac-poxB+) amplification product having the predicted BamHI+SalI and PstI restriction digest patterns, with primers spanning the entire poxB region, rather than a 1·9 kb (poxB) product, and (b) expressed an IPTG-inducible PoxB activity when assayed in cell-free extracts. The PoxB specific activities [µmol (mg protein)-1 h-1] were 43 and 4 (JRG3356) compared to 1 and 1 (W3110) in extracts of aerobic L-broth cultures grown with and without IPTG (50 µM), respectively. SDS-PAGE further confirmed that a polypeptide corresponding to the PoxB subunit (62 kDa) was amplified in extracts of induced cultures.



View larger version (28K):
[in this window]
[in a new window]
 
Fig. 1. Organization of the genes encoding pyruvate oxidase (a) and the pyruvate dehydrogenase complex (b) in plasmids used for restructuring the corresponding regions of the E. coli chromosome (see Methods). The kanR and camR antibiotic resistance cassettes were used for insertional inactivation of the poxB and aceE genes (respectively) and the kanR cassette also provided a selectable marker for chromosomal incorporation of the Ptac-regulated poxB gene. The genes are drawn to scale and the polarities of gene transcription are indicated; upox refers to an uncharacterized segment of chromosomal DNA located upstream of the poxB gene. Relevant restriction sites are abbreviated: B, BamHI; Bg, BglII; P, PstI; S, SalI; Sp, SphI; and X, XhoI, where the asterisk (*) denotes restriction sites derived from flanking vector DNA or PCR primers, an engineered SalI site downstream of the pdh operon and the XhoI site in the kanR cassette.

 
Inactivation of the chromosomal aceE gene by camR insertion.
The PDH complex was inactivated by replacing the 0·45 kb BglII fragment of the aceE gene which encodes the pyruvate dehydrogenase (E1p) component of a PDH expression plasmid, pGS367 (Ptac-aceEF-lpdA), by a 3·5 kb BamHI fragment containing a pHP45-derived camR cassette, to generate pGS960 (Fig. 1b). After verifying the structure of pGS960 by diagnostic restriction analysis, the inactive aceE::camR gene was transferred to the chromosome by transforming JC7623 (recBC sbcC) with pGS960 and selecting for CmR transformants. These transformants were screened initially for CmR ApS products and ultimately for Ace- derivatives. No CmR ApS transformants were found amongst >2000 CmR colonies selected on rich media (L agar with and without glucose and acetate supplements). However, when CmR transformants were selected on glucose minimal plates supplemented with acetate (2 mM), approximately 30% had the desired CmR ApS Ace- phenotype. A P1vir lysate of a representative strain was then used to transduce the aceE::camR region to JRG3356 (kanR-Ptac-poxB+) and W3110 to generate JRG3445 (aceE::camR kanR-Ptac-poxB+) and JRG3456 (aceE::camR), respectively. Nutritional and enzymological tests confirmed that both mutants had the predicted Ace- CmR phenotypes and lacked PDH-E1p and PDHC activities. The mutants were further characterized by PCR analysis using chromosomal DNA and primers that flank the two BglII sites in the aceE gene, which generated 3·5 kb fragments corresponding to the camR cassette, compared to the 0·45 kb aceE fragment amplified from parental DNA. The same P1vir lysate was used to construct an analogous derivative of TG1 (JRG4076; aceE::camR lacIq).

Inactivation of the poxB gene by kanR insertion.
A PoxB null strain (JRG3434, poxB::kanR) was constructed by selecting KnR transductants of W3110 with a P1vir lysate of YY877 (Chang et al., 1994 ) which has a kanR cassette inserted at the SalI site in the poxB structural gene (Fig. 1a). The same lysate was used to select KnR transductants of JRG3456 and JRG4077 in order to construct strains lacking pyruvate dehydrogenase (E1p) and PoxB, JRG3931 (aceE::camR poxB::kanR) and JRG4077 (aceE::camR poxB::kanR lacIq). Phenotypic characterization confirmed that JRG3434 was Ace+ and retained PDHC and PDH-E1p activities whereas JRG3931 was Ace- and had neither PDHC nor PDH-E1p plus PoxB activities.

Selection of a PDH-negative PoxB-constitutive strain.
A PoxB-constitutive derivative of the PDH-E1p null strain, JRG3456 (aceE::camR), was selected by acetate-starvation of a glucose-limited chemostat culture and designated JRG3980 (aceE::camR poxBc); see Results. The strain was characterized by the possession of an Ace+ phenotype, being totally deficient in PDHC activity, and maintaining a constant PoxB activity at twice the normal stationary-phase level throughout the growth cycle.

Multicopy poxB plasmids.
Several multicopy plasmids encoding the poxB gene and a derivative (poxBc{Delta}) lacking 23 C-terminal codons (Abdel-Hamid, 1999 ), and others that express the two genes from the tac promoter (Ptac-poxB and Ptac-poxBc{Delta}), were constructed in addition to those already described. The 4·8 kb BglII–SphI fragment of pCG5 was cloned in pBR322 and pSU38 to generate medium-copy and low-copy poxB+ plasmids (pGS1276 and pGS1277) having a smaller insert than pCG5 (Fig. 1). An analogous pair of poxBc{Delta} plasmids (pGS1275 and pGS1279) was constructed in the same way except that the downstream 1·8 kb SalI–SphI pox'B segment was replaced by terminally deleted segments flanked at one end by engineered sites (0·7 kb; SalI–BamHI or XbaI, respectively). Two high-copy expression plasmids (pGS927, Ptac-poxB; and pGS1197, Ptac-poxBc{Delta}) were constructed by combining the Ptac-poxB' segment of pGS859 (1·1 kb EcoRI–SphI–SalI; Fig. 1) and either the 1·8 kb SalI–SphI pox'B segment from pCG5 or the 0·7 kb SalI–BamHI pox'Bc{Delta} segment, in pUC18 and using TG1 as the host for transformation in order to repress the tac promoter.

Microbiological and analytical methods.
The rich medium used for routine subculture was L broth (Lennox, 1955 ) containing glucose (5·6 mM) and supplements of acetate (2 mM), chloramphenicol (15 µg ml-1) and kanamycin (50 µg ml-1), as required. The minimal media used for nutritional tests and for both batch and continuous culture have been described previously (Davé et al., 1995 ). The batch and continuous culture media contained glucose (20 mM) as the major carbon source or growth-limiting substrate, and supplements of acetate (2 mM) and IPTG (generally 50 µM) where indicated. Batch and continuous cultures were grown in chemostat vessels (nominal volumes, 1 l; working volumes, 605 and 650 ml). The pH was maintained at 7·0±0·1 by automatic titration with KOH (2 M). Dissolved O2 tensions were monitored with a galvanic oxygen electrode, and kept above 50% air saturation by adjusting the stirring speed (500–750 r.p.m.) at a constant air-flow of 600 ml min-1. All inocula were cultured overnight, washed, and resuspended in the corresponding experimental medium.

Batch cultures were used to determine doubling times and hence µmax by measuring the OD430 (Pye Unicam SP6-250 spectrophotometer) and then deriving the slopes of log OD430 versus time plots in the exponential region by regression analysis. Small samples (2–5 ml) were taken at specified times, centrifuged (4500 g for 15 min) at room temperature, and the culture supernatants stored at -20 °C prior to enzymic analysis of glucose (Bergmeyer & Bernt, 1974 ) and acetate (Acetic Acid Kit; Digen). Continuous cultures were established at steady-states over a range of different dilution rates from 0·05 to 0·25 h-1 and sampled for metabolic analysis. Substrate consumption, O2 consumption and CO2 production rates (qs, qO2, qCO2, respectively) were determined and the carbon conversion efficiencies (percentage of substrate carbon fluxed into biomass; numerically 695D.qS-1%, assuming that the carbon content of the biomass is 50%, w/w) and carbon balances [numerically (695D+16·7qCO2).qS-1%] were calculated as described by Brooke et al. (1989) .

Cell-free extracts were prepared for enzyme assay using bacteria from samples (20–50 ml) of batch and steady-state continuous cultures, harvested, washed and resuspended at 1 g per 4 ml cold potassium phosphate buffer (20 mM, pH 7·0) containing MgSO4 . 7H2O (5 mM), disodium EDTA (2 mM), PMSF (1 mM) and benzamidine hydrochloride (1 mM). Suspensions at 4 °C were disrupted by ultrasonic treatment (MSE disintegration 150 W) at 15 microns for three periods of 30 s with 30 s intervals for cooling. Intact bacteria and cell debris were sedimented (25000 g for 20 min at 4 °C) and the supernatants were maintained at 0 °C whilst measuring enzyme activities. Standard spectrophotometric procedures were used for measuring PDHC activity with 3-acetyl pyridine adenine dinucleotide as the electron acceptor (Russell & Guest, 1990 ) and PoxB (EC 1 . 2 . 2 . 2) activity with sodium ferricyanide as the electron acceptor (Chang & Cronan, 1986 ). The PoxB assay detects the pyruvate dehydrogenase component of the PDH complex (PDH-E1p; EC 1 . 2 . 4 . 1), albeit sub-optimally, and the analogous ferricyanide-based assay for PDH-E1p (Russell & Guest, 1990 ), used here only for strain characterization, likewise detects PoxB sub-optimally (Abdel-Hamid, 1999 ). Acetyl-CoA synthetase (EC 6 . 2 . 1 . 1) and acetokinase (EC 2 . 7 . 2 . 1) activities were measured according to Brown et al. (1977) . The reaction rates were proportional to the amounts of added cell-free extract and linear for at least 3 min. Specific activities are quoted as µmol of product formed or molar equivalent of electron acceptor reduced (mg protein)-1 h-1 at 37 °C and the protein concentrations of cell-free extracts were assayed by the Lowry method with BSA as standard.


   RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
In order to study the effects of replacing the PDH complex by PoxB and to assess the contribution of PoxB to the aerobic growth efficiency of E. coli, an isogenic set of W3110 derivatives containing appropriately modified pdh operons and poxB genes was constructed by a combination of genetic manipulation (in vitro) and chromosomal replacement (in vivo); see Methods. The set included strains in which one or both of the aceE (PDH-E1p) and poxB (PoxB) genes are disrupted by antibiotic cassettes, a strain in which the natural poxB promoter is replaced by the IPTG-inducible tac promoter linked to a kanR cassette (kanR-Ptac), and another, designated poxBc, in which the poxB gene is expressed constitutively (Table 1; Fig. 1).

Functional replacement of the PDH complex by PoxB in an aceE null strain
Several approaches were used to investigate conditions under which PoxB might replace the function normally attributed to the PDH complex viz. supplying the acetyl units (acetate) needed to support good aerobic growth in unsupplemented glucose minimal medium.

PoxB expression from a Ptac-regulated poxB gene. The possibility that PoxB could replace the PDH complex for supporting good acetate-independent growth on glucose was first investigated by constructing a PDH-E1p null strain (JRG3445; aceE::camR Ptac-poxB+) in which the chromosomal poxB gene is expressed from the IPTG-inducible tac promoter rather than the natural promoter. Plate growth tests using glucose minimal medium supplemented with either 2 mM acetate or 50 µM IPTG showed that JRG3445 exhibits an Ace- phenotype which can be satisfied by either acetate or IPTG, whereas the PDH-E1p null strain (JRG3456; aceE::camR) responds to acetate but not IPTG, and the PoxB null strain (JRG3434; poxB::kanR) resembles the Ace+ parental strain (W3110) in neither requiring nor being affected by either supplement (Table 2). It was concluded that when induced throughout the growth cycle, PoxB can support good acetate-independent growth in the absence of the PDH complex. Further studies with the Ptac-poxB+ strain (JRG3445) in controlled batch culture showed that the maximum specific growth rate (µmax) increases with IPTG concentration to a maximum value of 0·58 h-1 (i.e. 60% of the parental rate) with 30 µM IPTG (Fig. 2). Under these conditions the synthesis of PoxB and hence the metabolic flux to acetate (acetyl) is controlled via the lacI-encoded repressor and the concentration of exogenously supplied IPTG. It may be significant that the IPTG/PoxB-dependent growth rate fails to reach that of the parental strain, because in comparable experiments with a strain in which the entire pdh operon (Ppdh-pdhR-aceEF-lpdA) is replaced by an engineered Ptac-aceEF-lpdA operon, the highest IPTG/PDHC-dependent growth rate (observed at 25 µM IPTG) exceeded by 25% the highest growth rate both of the parental strain and of the acetate-supplemented engineered strain (Guest et al., 1996 ).


View this table:
[in this window]
[in a new window]
 
Table 2. Phenotypes and maximum growth rates (µmax) of W3110 and isogenic derivatives

 


View larger version (28K):
[in this window]
[in a new window]
 
Fig. 4. Growth profiles and substrate utilization of W3110 (parent) and isogenic derivatives. Batch cultures of all strains were grown aerobically at pH 7·0 and 37 °C in minimal medium containing glucose (20 mM) and acetate (2 mM). Growth ({bullet}) was monitored by measuring the OD430 and the glucose ({square}) and acetate ({triangleup}) concentrations (mM) of culture filtrates were assayed. Mean values from at least three independent cultures are shown: (a) W3110 (aceE+ poxB+); (b) JRG3434 (aceE+ poxB); (c) JRG3456 (aceE poxB+); (c') JRG3445 (aceE Ptac-poxB+) in the absence of IPTG; (c') JRG3931 (aceE poxB); (d) JRG3445 (aceE Ptac-poxB+) in the presence of IPTG (50 µM); and (e) JRG3980 (aceE poxBc).

 


View larger version (12K):
[in this window]
[in a new window]
 
Fig. 2. Growth of W3110 (parent) and JRG3445 (aceE::camR Ptac-poxB+) in the presence of increasing IPTG concentrations. The two strains were grown in batch culture on glucose-minimal medium supplemented with different concentrations of IPTG. Samples were collected hourly for the measurement of the optical density (OD430) and calculation of the maximum growth (µmax) rate.

 
PoxB expression from the natural poxB promoter (at a low growth rate) and from a constitutive poxB gene (at higher growth rates). The observation that exogenously induced PoxB activity can support good acetate-independent growth in a PDH null strain is consistent with the earlier report that ‘uninduced’ PoxB can support micro-colony formation by prolonged incubation of cultures that are essentially Ace- (LeMaster & Cronan, 1982 ). The same observation also prompted attempts to detect PoxB-supported acetate-independent growth using chemostat cultures of JRG3456 (aceE::camR), a PDH null strain in which the poxB gene is expressed from the natural promoter. By starting with glucose-limited cultures that had reached steady-state in the presence of 2 mM acetate (Fig. 3, phase 1; D=0·05 h-1, OD430=2·68) it was found that acetate could be withdrawn from the feed, whereupon a new steady-state was established at a slightly lower culture density without affecting the CmR Ace- phenotype of the resident bacteria (Fig. 3, phase 2; D=0·05 h-1, OD430=2·28). This indicates that poxB expression from the natural promoter can support acetate-independent growth at a low rate (µ=0·05 h-1). Steady-states could not be established in phase 2 if the initial dilution rate was set at a higher rate, presumably because the native poxB gene cannot satisfy the demand for a correspondingly higher rate of acetate production. However after 14 h (8 h in phase 2), it was invariably found that the dilution rate could be increased to 0·10 h-1 and a new steady-state established at a much higher culture density (Fig. 3, phase 3; D=0·10 h-1, OD430=5·36). Furthermore, this was accompanied by the appearance of CmR Ace+ bacteria; first detected in the 15 d samples but completely replacing the original CmR Ace- strain as the culture reached the new steady-state (as indicated in Fig. 3). The change in phenotype was presumed to be due to the selection of poxB constitutive mutants and a representative colony from such a culture was purified and designated JRG3980 (aceE::camR poxBc). This strain exhibited a stable CmR Ace+ phenotype characterized by good growth in acetate-free glucose minimal medium on plates (Table 2), in batch cultures, and in carbon-limited continuous cultures which could be maintained in steady-states at dilution rates up to 0·25 h-1 (the highest rate tested) without further strain selection. Enzymological tests further showed that the constitutive strain remains totally deficient in PDHC activity but expresses PoxB throughout the growth cycle at a constant rate equivalent to at least 3·8-fold higher than the normal lag- and exponential-phase rates, and ultimately twice the normal stationary-phase rate (see below). The procedure for isolating poxB constitutive strains closely resembled the classical use of lactose-limiting chemostat cultures for isolating lac constitutive and lac hyper strains (Horiuchi et al., 1962 ). Typically lac constitutives completely replaced the inducible parent within 20 generations of applying the selection (as observed for poxBc), whereas the less stable lac hyper strains containing multiple gene duplications only emerged after several hundred generations. The simplest explanation for the continuous synthesis of PoxB throughout the growth cycle would be altering the natural poxB promoter (or coupling the poxB structural gene) to one that is no longer stationary-phase/RpoS-dependent, but other possibilities were not excluded. It is interesting to note that constitutive mutants were not recovered when the phase-3 dilution rate was set higher than 0·10 h-1. On the contrary, such transitions were always accompanied by washout. This indicates that too few poxBc mutants had accumulated in the culture at the end of phase 2 to prevent washout along with the parental strain, even though the mutants can grow at dilution rates >0·10 h-1. Presumably, if phase 2 had been extended for longer periods, constitutive mutants would accumulate to levels that would readily establish steady-states at higher dilution rates in phase 3. Likewise they should ultimately predominate in phase 2 and then produce a new steady-state culture of CmR Ace+ bacteria at the original dilution rate.



View larger version (19K):
[in this window]
[in a new window]
 
Fig. 3. Growth of JRG3456 (aceE::camR poxB+) in chemostat culture. The culture was first grown to steady-state in glucose-limiting medium supplemented with acetate (2 mM) at a dilution rate of 0·05 h-1 (phase 1). It was then supplied with acetate-free medium until a second steady-state was reached at the same dilution rate (phase 2) and the dilution rate was finally increased to 0·10 h-1 (phase 3). Samples of the culture were taken daily for OD430 measurement and loopfuls were streaked on appropriate plates for phenotype tests: chloramphenicol resistance (CmR) and acetate-dependence/independence (Ace-/Ace+). As indicated, Ace+ bacteria were not detected until day 15 (phase 3) after which they became the predominant organism.

 
The results clearly demonstrate that PDH function can be replaced by PoxB when the poxB gene is expressed throughout the growth cycle from the tac promoter or a constitutive gene, and also from the natural poxB promoter provided that growth is maintained at a low growth/dilution rate in chemostat cultures.

Growth profiles, substrate utilization and enzyme activities of isogenic strains
A detailed analysis of the effects of altering the capacity for aerobic pyruvate metabolism was made with controlled batch cultures of the parental (wild-type) strain and various isogenic mutants. The medium contained glucose as the major carbon source and a supplement of acetate (2 mM), added to permit comparisons between both Ace- and Ace+ strains. Growth profiles including patterns of glucose utilization and acetate production (Fig. 4a–e) and the corresponding specific-activity profiles of PoxB (and/or PDH-E1p) and of the PDH complex (Fig. 5a–e), were constructed. Three distinct patterns were apparent from the former profiles (Fig. 4a–e).



View larger version (26K):
[in this window]
[in a new window]
 
Fig. 5. Growth profiles and enzyme activities of isogenic strains. Aerobic batch cultures were grown at pH 7·0 and 37 °C in minimal medium containing glucose (20 mM) and acetate (2 mM). Growth ({bullet}) was monitored by measuring the OD430 and cell-free extracts were assayed for PDH complex ({blacktriangleup}) and PoxB and/or PDH-E1p subunit ({square}) activities expressed as µmol substrate transformed or equivalent h-1 (mg protein)-1. Mean values for at least three independent cultures are shown: (a) W3110 (aceE+ poxB+); (b) JRG3434 (aceE+ poxB); (c) JRG3456 (aceE poxB+); (d) JRG3445 (aceE Ptac-poxB+) in the presence of IPTG (50 µM); and (e) JRG3980 (aceE poxBc).

 
The first pattern was observed with W3110 (aceE+ poxB+) and the isogenic poxB mutant, JRG3434 (Fig. 4a, b). Here the acetate concentration increased initially, reached a maximum as the growth rate started to decline, and then fell. Enzyme analysis showed that PDHC activity increased from low stationary-phase levels to high levels that were maintained into either late-exponential phase (W3110) or mid-exponential phase (JRG3434, poxB) before falling back to the stationary-phase level, more rapidly in W3110 than in the poxB mutant (Fig. 5a, b). In contrast, the combined activity of PoxB and PDH-E1p, both of which contribute to the ferricyanide-dependent oxidation of pyruvate, exhibited markedly different profiles in the two strains. A high activity was maintained throughout the W3110 growth cycle except for a temporary decline, which coincided with the initial decline in PDHC activity in early stationary phase (Fig. 5a). In the mutant lacking PoxB (JRG3434), the PDH-E1p activity increased in the exponential phase and then declined, albeit somewhat later than the PDHC activity (Fig. 5b). It was thus deduced that the restoration of a high PoxB plus PDH-E1p activity in stationary-phase cultures of W3110 is due to PoxB.

The second pattern of growth and substrate utilization was observed with the PDH-E1p null strain, JRG3456 (aceE::camR poxB+) (Fig. 4c). This strain requires acetate as a growth supplement and, as expected, glucose and acetate were used simultaneously. However, soon after the acetate was exhausted (at 7 h, with almost half of the glucose remaining), the acetate concentration slowly returned to the initial level (Fig. 4c). The PoxB specific activity remained relatively constant until 8 h when it started to increase, slowly at first and then rapidly, as the culture entered the stationary phase (Fig. 5c). These observations indicate that PoxB is responsible for the accumulation of acetate by stationary-phase cultures of the PDH-E1p null strain. This conclusion was further supported by experiments with cultures of two strains lacking PoxB and PDH-E1p activities: JRG3445 (aceE::camR Ptac-poxB+) grown without IPTG (Fig. 4c'); and JRG3931 (aceE::camR poxB::kanR) the double mutant (Fig. 4c'). In both cases glucose and acetate were utilized simultaneously, there was no subsequent period of acetate accumulation, and growth stopped when the exogenous acetate was exhausted. These two strains have the unique property of not accumulating acetate in aerobic stationary-phase cultures.

The third pattern was observed with the PDH-E1p null strain containing the IPTG-regulated poxB gene (JRG3445; aceE::camR Ptac-poxB+) grown with the inducer, and the PDH-E1p null strain having a constitutive poxB gene (JRG3980; aceE::camR poxBc) (Fig. 4d, e). Here the onset of acetate excretion was delayed relative to the aceE+ strains (first pattern) and then followed by the accumulation of high but stable rather than declining levels of acetate (13 and 10 mM, respectively), presumably due to the presence of higher PoxB activities. Indeed, the PoxB specific activities reached by the IPTG-induced and constitutive strains in stationary phase [60 and 1·2 µmol (mg protein)-1 h-1, respectively; Fig. 5d, e] were approximately 100- and 2-fold higher than those of strains expressing the poxB gene from the natural promoter [0·6 µmol (mg protein)-1 h-1; Fig. 5a, c]. The unchanging specific activity of PoxB exhibited by the constitutive strain is entirely consistent with the view that the poxB gene is expressed from a promoter that is no longer subject to stationary-phase (RpoS-dependent) control or to any other major regulatory constraint. It should be noted that for most of the growth cycle, the PoxB activity of the constitutive strain is at least 3·8-fold higher than that of the parental aceE::camR strain, before declining to 2·0-fold in stationary phase (Fig. 5c, e).

It is significant that when the substrate (glucose) was completely consumed, the extents of growth of the aceE+ strains (Fig. 4a, b) were invariably greater than those of the aceE strains, irrespective of whether they are poxB+, poxB or poxBc (Fig. 4c, c', e). In contrast, poxB inactivation appeared to have little effect on the growth attained by aceE+ cultures (Fig. 4a, b), although it was subsequently found to lower the growth yield in continuous culture (see below; Table 5). The effect of poxB inactivation on the extent of growth was more apparent in the aceE background (Fig. 4c versus c', c'), and reversed by IPTG-induced or constitutive poxB expression (Fig. 4d, e). The maximum growth rates (µmax) calculated from the growth curves of the controlled batch cultures containing glucose and acetate showed that growth rate is lowered when either of the aceE or poxB genes are inactivated (Table 2). Inactivation of the aceE gene of W3110 lowered the rate by 33% (0·69 to 0·46 h-1) whereas inactivation of poxB caused a 16% reduction (0·69 to 0·58 h-1), and the combined effects of inactivating both genes was a 48% reduction (0·69 to 0·36 h-1) which is almost additive. Furthermore, induction of PoxB activity by adding IPTG to JRG3445 (aceE::camR Ptac-poxB+) restored the maximum growth rate to that of aceE strains containing either wild-type or constitutive poxB genes (0·46 h-1). These observations demonstrate that PoxB has a very significant effect on the growth rate of E. coli both in the presence and absence of PDHC (and this applies even when the anabolic function of the PDH complex is satisfied by the provision of an acetate supplement).


View this table:
[in this window]
[in a new window]
 
Table 5. Steady-state analysis of W3110 (aceE+ poxB+) and JRG3434 (aceE+ poxB)

 
Comparative studies with plasmid-transformed strains
The observation that a declining 3·8- to 2·0-fold amplification of PoxB activity supported a good Ace+ nutritional phenotype in the constitutive strain, JRG3980 (aceE::camR poxBc), contrasts sharply with previous findings that poxB plasmids which increased PoxB synthesis 3–8-fold either fail to restore, or only partially restore, acetate-independence to PDH-null strains (Grabau & Cronan, 1984 ). More precisely, pCG5 (poxB+) transformants of PDH-null derivatives of HfrC and Hfr6 failed to grow without acetate supplementation whereas those of another strain, UB1005 (which had been chosen for its unusually good growth on acetate as sole carbon source), grew without acetate but only at 20% of the rate observed with added acetate. This could mean that the continuous expression of a relatively low level of PoxB from the single chromosomal gene of the constitutive strain supports a higher rate of acetate-independent growth than potentially higher enzyme levels expressed from multiple copies of the natural stationary-phase-dependent gene. It also suggests that overproduction of PoxB and/or the product of some other cloned gene might impair the growth of the multicopy transformants. In the earlier work PoxB was probably assayed in stationary-phase cultures grown in rich medium rather than throughout the growth cycle with cultures grown in glucose minimal medium, acetate-dependence was tested in succinate rather than glucose minimal medium, and the PDH lesion was more extensive ({Delta}aroP-pdhR-aceEF not aceE::camR). Several poxB multicopy plasmids (see Methods and Table 1) were therefore tested for their ability to complement the Ace- lesion of the aceE::camR derivative of W3110 (JRG3456) in glucose minimal media (without antibiotic). All of the plasmids expressing poxB from the natural promoter, pCG5, pGS1276 (having a smaller poxB insert) and pGS1277 (smaller insert and low copy number), complemented the Ace- nutritional lesion but not to the extent observed with added acetate. This was detectable on plates and in liquid medium. The mean specific growth rates in glucose minimal medium (h-1) and the PoxB specific activities of comparable stationary-phase cultures grown in rich medium [µmol (mg protein)-1 h-1] were: 0·23 h-1 (50%) and 2·4 (4-fold) for pCG5; 0·17 h-1 (34%) and 1·2 (2-fold) for pGS1276; 0·12 h-1 (26%) and 0·9 (1·5-fold) for pGS1277; and 0·46 h-1 (100%) and 1·2 (2-fold) for single-copy poxBc (the figures in parentheses are quoted relative to the rate of growth with added acetate and the PoxB activity of the parental strain). Two analogous plasmids (pGS1275 and pGS1279) encoding a C-terminally deleted but catalytically active protein (PoxBc{Delta}) were totally incapable of supporting acetate-independent growth. These findings with the PDH-null strain of W3110 supported those observed previously with the UB1005 derivative (Grabau & Cronan, 1984 ) and they also confirmed that the C-terminal segment is needed for PoxB to function in vivo (Grabau & Cronan, 1986b ).

The PoxB activities of the transformants were not amplified to the levels expected with multicopy plasmids, suggesting that the poxB gene is strongly regulated, or that the overproduction of PoxB (or the product of another cloned gene) is deleterious. Indeed, growth impairment by the poxB and poxBc{Delta} plasmids (but not the vectors) was observed when transformants of JRG3456 (aceE::camR) were grown under permissive conditions, i.e. with exogenous acetate. Furthermore, two high-copy expression plasmids, pGS927 (Ptac-poxB) and pGS1197 (Ptac-poxBc{Delta}), not only failed to complement the Ace- phenotype of JRG3456 (aceE::camR) in the presence or absence of IPTG, they either impaired growth (pGS1197 Ptac-poxBc{Delta}) or completely inhibited growth (pGS927 Ptac-poxB) in the presence of added acetate, and both plasmids impaired the growth of W3110 under all conditions. The different degrees of toxicity suggest that although excess functional enzyme is more deleterious than the non-functional enzyme, there must be some other inhibitory factor. The two expression plasmids had the same growth inhibiting and non-complementing effects in the lacIq background of TG1 and two TG1 derivatives, JRG4076 (aceE::camR) and JRG4077 (aceE::camR poxB::kanR) which more tightly controls the tac promoter in multicopy situations. In fact TG1 had to be used in order to recover both Ptac-pox plasmids during their construction. Very high PoxB and PoxBc{Delta} activities in the range 480–600 µmol (mg protein)-1 h-1 (i.e. 600–1000-fold amplifications) were induced by IPTG in freshly transformed cultures. However, these plasmids proved to be unstable, their ability to overexpress the corresponding enzyme declining rapidly during propagation under all growth conditions. Such instability or host intolerance was only observed with transformed strains containing multicopy plasmids. It was therefore concluded that the continuous expression of a moderate level of PoxB activity from a single constitutive or tac regulated gene (3·8–100-fold amplification) is tolerated and allows full complementation of the nutritional defect of PDH-deficient mutants. In contrast, the weak complementation exhibited by plasmid transformants probably represents a compromise between several factors where the provision of acetate is limited by strong temporal regulation of the natural gene and the elimination of deleterious multicopy plasmids.

Replacing the PDH complex by PoxB lowers growth efficiency and growth yield
Carbon (glucose)-limited chemostat cultures were used to make a detailed quantitative investigation of the energetic consequences of replacing the PDH complex by PoxB. Steady-state cultures of JRG3980 (aceE::camR poxBc) and JRG3445 (aceE::camR kanR-Ptac-poxB+) having constitutive and IPTG-inducible poxB genes (respectively) were established at four different dilution rates in the range 0·05–0·25 h-1 and compared with analogous cultures of the wild-type (W3110) and PDH-null (JRG3456; aceE::camR) strains. The rates of glucose and O2 utilization (qs and qO2, respectively), CO2 production (qCO2) and the biomass formed (x) were measured. These values were used to calculate the percentage carbon fluxed into biomass (i.e. carbon conversion efficiency) and the percentage carbon balance (see Methods) and representative results obtained at one dilution rate (D=0·15 h-1) are shown in Table 3. At all steady-states the substrate carbon was almost fully accounted for by the total amount of carbon in the biomass formed and the CO2 produced (93–98%). The aceE::camR mutants with normal or constitutive poxB genes exhibited higher rates of glucose consumption and lower carbon conversion efficiencies than the parental strain (W3110), consistent with an impaired metabolism. A similar impairment was also evident from the low carbon conversion efficiency observed when the poxB gene was continuously expressed from the IPTG-inducible tac promoter, but it is not clear why the rate of glucose consumption was also low under these conditions. Overall, the carbon conversion efficiencies were lowered by 29–39% as a consequence of inactivating the PDH complex.


View this table:
[in this window]
[in a new window]
 
Table 3. Physiological parameters of W3110 and isogenic derivatives in glucose-limited chemostat cultures

 
The maximum growth yields (Ymax) were calculated from the reciprocals of the slopes obtained by plotting qs (glucose utilization rates) versus dilution (growth) rates for each strain. The Ymax values for the aceE::camR mutants were between 75% and 91% of the parental yield (Table 3). The lowest yield was observed with JRG3980 (aceE::camR poxBc) where the poxB gene was expressed constitutively. These results confirm that PDHC activity can be replaced by PoxB, but there are adverse energetic consequences, presumably because acetate (the poxB product) has to be converted to acetyl-CoA for further metabolism. The specific activities of acetyl-CoA synthetase (ACS) and acetokinase (ACK) were accordingly compared in cell-free extracts prepared from steady-state cultures of the isogenic strains grown at five different rates (see Table 4 for representative results obtained at D=0·15 h-1). There was a 13–20-fold increase in ACS activity in the PDH-E1p null mutants relative to that of the PoxB null strain (JRG3434, poxB::kanR), which is far greater than the corresponding 1·5–2·1-fold increase in ACK activity (Table 4), and strongly suggests that ACS is primarily responsible for catalysing the conversion of PoxB-derived acetate to acetyl-CoA. In contrast, the specific activities of both ACS and ACK increased only 2·5-fold (approx.) in W3110 relative to those of the PoxB null strain, which is consistent with the view that the PDH complex is normally the prime source of acetyl-CoA in the parental strain (Table 4). Comparable results were obtained at all of five dilution rates in the range 0·05–0·25 h-1 (data not shown). These results are in complete agreement with the suggestion that the extra energy required to convert PoxB-derived acetate to acetyl-CoA probably accounts for the adverse energetic consequences observed when the complex is replaced by PoxB. It is also interesting to note that cell-free extracts of steady-state cultures of W3110 maintained at the lowest dilution rate (0·05 h-1) had no detectable PDHC activity, <0·02 µmol h-1 (mg protein)-1, whereas the combined activity of PDH-E1p and PoxB was at its highest, 1·41 µmol h-1 (mg protein)-1 (data not shown). This confirmed the earlier conclusion based on studies with controlled batch cultures (Fig. 5) that PoxB rather than the PDH complex is the major catalyst for pyruvate oxidation at low growth rates.


View this table:
[in this window]
[in a new window]
 
Table 4. Enzyme activities of W3110 and isogenic derivatives

 
PoxB inactivation diverts carbon utilization from biomass to energy production
The results obtained with batch cultures indicated that poxB inactivation lowers the specific growth rates (µmax) of wild-type and PDH-E1p null strains by 16 and 22%, respectively (Table 2). The extent to which PoxB normally contributes to the aerobic growth efficiency of E. coli was investigated in greater detail by comparing the physiological parameters (x, qs, qO2 and qCO2) of glucose-limited steady-state cultures of W3110 (ace+ poxB+) and the PoxB null strain (JRG3434, ace+ poxB::kanR) at different dilution rates (Table 5). The results indicated that at the same growth rates and comparable carbon balances, inactivation of the poxB gene lowered the carbon conversion efficiency (percentage carbon flux to biomass) by a mean of 24% and increased the amount of carbon used for energy metabolism (qCO2) by a mean of 23% (Table 5). The maximum growth yield (Ymax) also fell by 14% from 93 to 80 g dry weight mol-1 when the poxB gene was inactivated (Table 5). These results together with the 16% fall in µmax that accompanies poxB inactivation (Table 2) clearly establish that PoxB normally makes a significant contribution to the aerobic growth efficiency of E. coli.


   DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Whilst the PDH complex and PFL are accepted as the major enzymes catalysing pyruvate oxidation to acetyl-CoA under aerobic and anaerobic conditions (respectively), pyruvate oxidase is regarded as a non-essential and potentially wasteful enzyme of uncertain metabolic function. However, the data reported here demonstrate that PoxB makes a significant contribution to the aerobic growth efficiency in glucose minimal medium. It is therefore rather surprising that a strain like JRG3456 (aceE::camR) which lacks PDHC activity but retains PoxB cannot grow in glucose minimal medium unless supplemented with acetate. One explanation is that the PoxB activity is initially too low to produce enough acetate to support good growth (Henning & Herz, 1964 ). In fact, Chang & Cronan (1983) reported that a deletion strain lacking the PDH-E1p and PDH-E2p subunits does indeed form very small colonies on glucose minimal medium lacking acetate, but only after prolonged growth (6–8 d). They also showed that growth is abolished by poxB mutations, indicating PoxB is the source of the acetate that supports the weak growth of the deletion strain.

Here it was shown that PoxB is able to replace the PDH complex and support good acetate-independent growth: (i) when the natural poxB promoter is replaced by the IPTG-inducible tac promoter; (ii) when a strain having the natural promoter is grown at low growth rates in chemostat culture; and (iii) when PoxB activity is expressed constitutively. The acetate-independent but IPTG-dependent growth of JRG3445 (aceE::camR Ptac-poxB+) shows that PoxB can replace the PDH complex by oxidizing pyruvate to acetate at the rate needed to support good growth on glucose. The relationship between the maximum growth rate of JRG3445 and IPTG concentration further indicated that the growth-limiting factor under these conditions is the carbon flux through PoxB. Grabau & Cronan (1984) had earlier shown that poxB+ transformants of one out of three PDHC-deficient strains grew without acetate, but only at 20% of the rate observed with added acetate, despite being able to produce (at least in the stationary phase) as much or more PoxB activity than required to restore normal growth to the constitutive strain studied here. They suggested that the Ace- phenotype of strains containing PoxB but lacking PDHC activity could be due to the inefficiency of the pathways responsible for converting PoxB-derived acetate to acetyl-CoA, rather than to insufficient PoxB. However, the present demonstration that the Ptac-poxB+ strain (JRG3445) grows without acetate in the presence of IPTG suggests that growth is not limited by the acetate utilization pathways but by a combination of insufficient PoxB activity early in the growth cycle and the growth inhibitory effects of deleterious multicopy plasmids. This view is also supported by the acetate-independence of the poxBc strain (JRG3980) which synthesizes PoxB throughout the growth cycle. Moreover, a 3·8-fold amplification of PoxB activity early in the growth cycle is sufficient to raise the permitted growth rate from 0·05 h-1 to greater than 0·25 h-1.

Although the exact nature of the constitutive mutation was not defined, the simplest explanation is that the poxB promoter has lost its RpoS-dependence. The stability and facile selection of such mutants indicates that they are not hyper strains containing multiple poxB genes, nor are they likely to have arisen by the activation of a novel pyruvate oxidase or a cryptic gene, because no revertants of aceEF or poxB mutants have ever been reported to possess a novel activity, nor does the E. coli genome contain an obvious candidate. There is no evidence for factor-mediated positive or negative regulation of poxB, e.g. by the pyruvate-sensitive regulator (PdhR), and no PdhR-site is associated with the poxB gene. Moreover, regulator or operator mutations are unlikely to alter the stationary-phase dependence of poxB gene expression.

Studies with the poxB plasmids indicated that compared to the Ptac-poxB+ and poxBc strains the presence of multiple copies of the poxB region impairs growth and hence compromises the transformant’s ability to grow well without acetate. Here it was also shown that expressing poxB from the natural promoter can sustain acetate-independent aerobic growth in a PDH null strain, albeit at a low growth rate. Indeed, the specific activity of PoxB was at its maximum in the parental strain (W3110) during growth at a low dilution rate (0·05 h-1) in glucose-limited chemostat culture whereas PDHC activity was undetectable. These observations show that expression of the poxB gene from its RpoS ({sigma}38)-dependent promoter can be induced at low growth (dilution) rates, a finding which is entirely consistent with those of Notley & Ferenci (1996) , who have concluded that growth at low dilution rates in glucose-limited chemostat culture induces the expression of RpoS ({sigma}38)-dependent genes in E. coli.

The major physiological consequences stemming from the use of PoxB to support the growth of the various aceE::camR strains are lower maximum growth rates in batch cultures (33% decrease), lower maximum growth yields (9–25% decrease) and carbon conversion efficiencies (29–39% decrease) in steady-state chemostat cultures, and correspondingly greater use of the carbon substrate for energy generation, compared to the parental strain. Thus, although PoxB can support aerobic growth of E. coli in the absence of the PDH complex it is energetically less efficient. This can to some extent be explained by the need for energy to convert the PoxB-derived acetate to acetyl-CoA. E. coli possesses two pathways for this conversion, the acetokinase-phosphotransacetylase (ACK-PTA) and the acetyl-CoA synthetase (ACS) pathways, both of which require ATP (Brown et al., 1977 ). When PoxB was inactivated the activities of ACK and ACS were both relatively low, 18 and 0·12 µmol (mg protein)-1 h-1, respectively. However with an active PoxB, the ACK activity increased 1·5–2·5-fold both in the presence or absence of the PDH complex, and the ACS activity likewise increased 2·5-fold in the presence of the PDH complex, but in the PDH-E1p null strains the increase was far greater, 13–20-fold. This strongly suggests that ACS is induced to convert PoxB-derived acetate to acetyl-CoA for further metabolism. It may also be significant that the affinity of ACS for acetate is very much higher than that of ACK; Km=0·2 mM for ACS and 0·3 M for ACK (Brown et al., 1977 ; Rose, 1962 ).

Previous studies on the effects of PoxB deficiency had indicated that poxB mutants grew at normal rates (Chang & Cronan, 1983 ). In marked contrast, the poxB::kanR mutation of JRG3434 was here shown to lower the maximum growth rate by 16% relative to the parental strain. This discrepancy might have been due to the use of point mutants in the earlier study, some of which may retain some residual activity in vivo. The current findings were further supported by observing that the flux of carbon into biomass was 24% lower and the maximum yield (Ymax) 14% lower in the poxB mutant relative to the parental strain. These results clearly indicate that poxB contributes to the aerobic growth efficiency of E. coli. It is not known how this is achieved but PoxB may function as a safety valve to convert excess pyruvate to acetate rather than to acetyl-CoA, thus maintaining the intracellular CoA pool for other metabolic functions including the conversion of 2-oxoglutarate to succinyl-CoA and hence citric acid cycle activity. Likewise, the exclusive use of PDH complex may have a detrimental effect on cellular redox balance by lowering the NAD+/NADH ratio. It was anticipated that the involvement of PoxB in pyruvate oxidation would lower the growth efficiency rather than being beneficial. This is because direct coupling to the respiratory chain via the PoxB flavin cofactor would bypass the energy-conserving step associated with NADH oxidation as well as imposing the need to return acetate to the citric acid cycle via an ATP-dependent route. The energy conservation derived from NADH oxidation depends upon the extent to which each of the two NADH dehydrogenases (NADH:quinol oxidoreductases), NdhI and NdhII, are used. The first, NdhI or complex 1 (product of the nuo operon), is a proton-translocating enzyme complex whereas NdhII (the ndh gene product) is a non-proton-translocating flavoprotein (Gennis & Stewart, 1996 ). Although both enzymes are expressed aerobically and repressed in fermentation, NdhI is essential for fumarate respiration and it is used preferentially during low aeration, in the late exponential and stationary phases of the growth cycle and during carbon limitation, when energy conservation is most needed, whereas NdhII is used preferentially during aerobic and nitrate respiration (Bongaerts et al., 1995 ; Calhoun & Gennis, 1993 ; Calhoun et al., 1993 ; Green & Guest, 1994 ; Tran et al., 1997 ). Clearly the deficit in respiratory-chain derived energy will be directly related to the extent to which NdhI is used relative to NdhII (which is equivalent to PoxB) under the carbon-limiting growth employed in the work. But why does inactivation of the PoxB route impair the aerobic growth efficiency? It would appear that energy conservation per se is not a major factor. On the contrary, E. coli must derive some benefit from being able to convert pyruvate to acetyl-CoA by the seemingly wasteful route via acetate. It has been suggested that PoxB functions in the stationary-phase conversion of cell wall D-Ala-D-Ala dipeptide to acetate via pyruvate (Lessard et al., 1998 ) but it seems unlikely that this could account for the physiological perturbations observed here. Although further studies are needed to elucidate exactly how PoxB contributes to the metabolic economy of E. coli, it is evident from the adverse energetic consequences of its inactivation that PoxB must perform a significant beneficial role in the aerobic growth efficiency of E. coli.


   ACKNOWLEDGEMENTS
 
We thank Dr J. E. Cronan Jr for very kindly supplying the poxB plasmid and mutant (pCG5 and YYC877) and Dr J. P. van Dijken for his interest and helpful discussions. The work was supported by a studentship from the Egyptian Education Bureau (A.M.A-H.) and a project grant from the Biotechnology and Biological Sciences Research Council (J.R.G. and M.M.A.).


   REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES
 
Abdel-Hamid, A. M. (1999). Molecular and physiological studies on the role of pyruvate oxidase in E. coli. PhD thesis, University of Sheffield.

Bergmeyer, H. U. & Bernt, E. (1974). Determination of D-glucose with glucose oxidase and peroxidase. In Methods of Enzymatic Analysis , pp. 1205-1215. Edited by H. U. Bergmeyer. New York & London:Academic Press.

Bongaerts, J., Zoske, S., Weidner, U. & Unden, G. (1995). Transcriptional regulation of the proton translocating NADH dehydrogenase genes (nuoA-N) of Escherichia coli by electron acceptors, electron donors and gene regulators. Mol Microbiol 16, 521-534.[Medline]

Brooke, A. G., Watling, E. M., Attwood, M. M. & Tempest, D. W. (1989). Environmental control of metabolic fluxes in thermo-tolerant methylotrophic Bacillus strains. Arch Microbiol 151, 268-273.

Brown, T. D. K., Jones-Mortimer, M. C. & Kornberg, H. L. (1977). The enzymic interconversion of acetate and acetyl-coenzyme A in Escherichia coli. J Gen Microbiol 102, 327-336.[Medline]

Calhoun, M. W. & Gennis, R. B. (1993). Demonstration of separate genetic loci encoding distinct membrane-bound respiratory NADH dehydrogenases of Escherichia coli. J Bacteriol 175, 3013-3019.[Abstract]

Calhoun, M. W., Oden, K. L., Gennis, R. B., Teixeira de Mattos, M. J. & Neijssel, O. M. (1993). Energetic efficiency of Escherichia coli: effects of mutations in components of the aerobic respiratory chain. J Bacteriol 175, 3020-3025.[Abstract]

Chang, Y.-Y. & Cronan, J. E.Jr (1982). Mapping non-selectable genes of Escherichia coli by using transposon Tn10: location of a gene affecting pyruvate oxidase. J Bacteriol 151, 1279-1289.[Medline]

Chang, Y.-Y. & Cronan, J. E.Jr (1983). Genetic and biochemical analyses of Escherichia coli strains having a mutation in the structural gene (poxB) for pyruvate oxidase. J Bacteriol 154, 757-762.

Chang, Y.-Y. & Cronan, J. E.Jr (1986). Molecular cloning, DNA sequencing and enzymatic analysis of two Escherichia coli pyruvate oxidase mutants defective in activation by lipids. J Bacteriol 167, 312-318.[Medline]

Chang, Y. Y. & Cronan, J. E.Jr (1997). Thiol chemistry detects three conformations of the lipid binding region of Escherichia coli pyruvate oxidase. Biochemistry 36, 11564-11573.[Medline]

Chang, Y.-Y., Wang, A.-Y. & Cronan, J. E.Jr (1994). Expression of Escherichia coli pyruvate oxidase (PoxB) depends on the sigma factor encoded by the rpoS (katF) gene. Mol Microbiol 11, 1019-1028.[Medline]

Davé, E., Guest, J. R. & Attwood, M. M. (1995). Metabolic engineering in Escherichia coli: lowering the lipoyl domain content of the pyruvate dehydrogenase complex adversely affects growth rate and yield. Microbiology 141, 1839-1849.[Abstract]

Fellay, R., Frey, J. & Krisch, H. (1987). Interposon mutagenesis of soil and water bacteria: a family of DNA fragments designed for in vivo insertional mutagenesis of Gram-negative bacteria. Gene 52, 147-154.[Medline]

Gennis, R. B. & Hager, L. P. (1976). Pyruvate oxidase. In The Enzymes and Biological Membranes , pp. 493-504. Edited by A. Martonosi. New York:Plenum.

Gennis, R. B. & Stewart, V. (1996). Respiration. In Escherichia coli and Salmonella: Cellular and Molecular Biology, pp. 217–261. Edited by F. C. Neidhardt and others. Washington, DC: American Society for Microbiology.

Grabau, C. & Cronan, J. E.Jr (1984). Molecular cloning of the gene (poxB) encoding the pyruvate oxidase of Escherichia coli, a lipid-activated enzyme. J Bacteriol 160, 1088-1092.[Medline]

Grabau, C. & Cronan, J. E.Jr (1986a). Nucleotide sequence and deduced amino acid sequence of Escherichia coli pyruvate oxidase, a lipid-activated flavoprotein. Nucleic Acids Res 14, 5449-5460.[Abstract]

Grabau, C. & Cronan, J. E.Jr (1986b). In vivo function of Escherichia coli pyruvate oxidase specifically requires a functional lipid binding site. Biochemistry 25, 3748-3751.[Medline]

Grabau, C., Chang, Y. Y. & Cronan, J. E.Jr (1989). Lipid binding by Escherichia coli pyruvate oxidase is disrupted by small alterations of the carboxyl-terminal region. J Biol Chem 264, 12510-12519.[Abstract/Free Full Text]

Green, J. & Guest, J. R. (1994). Regulation of transcription at the ndh promoter of Escherichia coli by FNR and novel factors. Mol Microbiol 12, 433-444.[Medline]

Guest, J. R., Quail, M. A., Davé, E., Cassey, B. & Attwood, M. M. (1996). Regulatory and other aspects of pyruvate dehydrogenase complex synthesis in Escherichia coli. In Biochemistry and Physiology of Thiamin Diphosphate Enzymes , pp. 326-333. Edited by H. Bisswanger & A. Schellenberger. Prien:Intemann.

Hamilton, C. M., Aldea, M., Washburn, B. K., Babitzke, P. & Kushner, S. R. (1989). New method for generating deletions and gene replacements in Escherichia coli. J Bacteriol 171, 4617-4622.[Medline]

Henning, U. & Herz, C. (1964). Ein Struckturgen-Komplex fur den Pyruvat-Dehydrogenase-Komplex von Escherichia coli K12. Z Vererbungsl 95, 260-275.[Medline]

Horiuchi, T., Tomizawa, J.-I. & Novick, A. (1962). Isolation and properties of bacteria capable of high rates of ß-galactosidase synthesis. Biochim Biophys Acta 55, 152-163.[Medline]

Kaniga, K., Compton, M. S., Curtiss, R.III & Sundaram, P. (1998). Molecular and functional characterization of Salmonella enterica serovar Typhimurium poxA gene: effect on attenuation of virulence and protection. Infect Immun 66, 5599-5606.[Abstract/Free Full Text]

LeMaster, D. M. & Cronan, J. E.Jr (1982). Biosynthetic production of 13C-labeled amino acids with site-specific enrichment. J Biol Chem 257, 1224-1230.[Abstract/Free Full Text]

Lennox, E. S. (1955). Transduction of linked genetic characters of the host by bacteriophage P1. Virology 1, 190-206.

Lessard, I. A. D., Pratt, S. D., McCafferty, D. G., Bussiere, D. E., Hutchins, C., Wanner, B. L., Katz, L. & Walsh, C. T. (1998). Homologs of the vancomycin resistance D-Ala-D-Ala dipeptidase VanX in Streptomyces toyocaensis, Escherichia coli and Synechocystis: attributes of catalytic efficiency, stereoselectivity and regulation with implications for function. Chem Biol 5, 489-504.[Medline]

Marsh, P. (1986). ptac-85, an E. coli vector for expression of non-fusion proteins. Nucleic Acids Res 14, 3603.[Medline]

Martinez, E., Bartolemé, B. & De la Cruz, F. (1988). pACYC184-derived cloning vectors containing the multiple cloning site and lacZ{alpha} reporter gene of pUC8/9 and pUC18/19 plasmids. Gene 68, 157-162.

Mather, M., Blake, R., Koland, J., Schrock, H., Russell, P., O’Brien, T., Hager, L. P. & Gennis, R. B. (1982). Escherichia coli pyruvate oxidase: interaction of a peripheral membrane protein with lipids. Biophys J 37, 87-88.

Miller, J. H. (1992). Preparation and use of P1vir lysates. In A Short Course in Bacterial Genetics: a Laboratory Manual and Handbook for E. coli and Related Bacteria, pp. 268–272. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Notley, L. & Ferenci, T. (1996). Induction of RpoS-dependent functions in glucose-limited continuous culture: what level of nutrient limitation induces the stationary phase of Escherichia coli? J Bacteriol 178, 1445-1468.[Abstract]

Oden, K. L., De Veaux, L. C., Vibat, C. R. T., Cronan, J. E.Jr & Gennis, R. B. (1990). Genomic replacement in Escherichia coli K-12 using covalently closed circular plasmid DNA. Gene 96, 29-36.[Medline]

Recny, M. A., Grabau, C., Cronan, J. E.Jr & Hager, L. P. (1985). Characterization of the {alpha}-peptide released upon protease activation of pyruvate oxidase. J Biol Chem 260, 14287-14291.[Abstract/Free Full Text]

Rose, I. A. (1962). Acetate kinase. In The Enzymes , pp. 115-118. Edited by P. D. Boyer, H. Lardy & K. Myrbäck. New York & London:Academic Press.

Russell, G. C. & Guest, J. R. (1990). Overexpression of reconstructed pyruvate dehydrogenase complexes and site-directed mutagenesis of a potential active-site histidine residue. Biochem J 269, 443-450.[Medline]

Russell, G. C., Machado, R. S. & Guest, J. R. (1992). Overproduction of the pyruvate dehydrogenase multienzyme complex of Escherichia coli and site-directed substitutions in the E1p and E2p subunits. Biochem J 287, 611-619.[Medline]

Russell, P., Hager, L. P. & Gennis, R. B. (1977). Lipid activation and protease activation of pyruvate oxidase. J Biol Chem 252, 7883-7887.[Medline]

Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning: a Laboratory Manual, 2nd edn. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Tran, Q. H., Bongaerts, J., Vlad, D. & Unden, G. (1997). Requirement for the proton-pumping NADH dehydrogenase I of Escherichia coli in respiration of NADH to fumarate and its bioenergetic implications. Eur J Biochem 244, 155-160.[Abstract]

Van Dyk, T. K., Smulski, D. R. & Chang, Y.-Y. (1987). Pleiotropic effects of poxA regulatory mutations of Escherichia coli and Salmonella typhimurium, mutations conferring sulfometuron methyl and {alpha}-ketobutyrate hypersensitivity. J Bacteriol 169, 4540-4546.[Medline]

Wang, A. Y., Chang, Y. Y. & Cronan, J. E.Jr (1991). Role of the tetrameric structure of Escherichia coli pyruvate oxidase in enzyme activation and lipid binding. J Biol Chem 266, 10959-10966.[Abstract/Free Full Text]

Received 19 January 2001; accepted 9 February 2001.