Regulation of the hydrogenase-4 operon of Escherichia coli by the sigma(54)-dependent transcriptional activators FhlA and HyfR.

The hyf locus (hyfABCDEFGHIJ-hyfR-focB) of Escherichia coli encodes a putative 10-subunit hydrogenase complex (hydrogenase-4 [Hyf]); a potential sigma(54)-dependent transcriptional activator, HyfR (related to FhlA); and a putative formate transporter, FocB (related to FocA). In order to gain insight into the physiological role of the Hyf system, we investigated hyf expression by using a hyfA-lacZ transcriptional fusion. This work revealed that hyf is induced under fermentative conditions by formate at a low pH and in an FhlA-dependent fashion. Expression was sigma(54) dependent and was inhibited by HycA, the negative transcriptional regulator of the formate regulon. Thus, hyf expression resembles that of the hyc operon. Primer extension analysis identified a transcriptional start site 30 bp upstream of the hyfA structural gene, with appropriately located -24 and -12 boxes indicative of a sigma(54)-dependent promoter. No reverse transcriptase PCR product could be detected for hyfJ-hyfR, suggesting that hyfR-focB may be independently transcribed from the rest of the hyf operon. Expression of hyf was strongly induced ( approximately 1,000-fold) in the presence of a multicopy plasmid expressing hyfR from a heterologous promoter. This induction was dependent on low pH, anaerobiosis, and postexponential growth and was weakly enhanced by formate. The hyfR-expressing plasmid increased fdhF-lacZ transcription just twofold but did not influence the expression of hycB-lacZ. Interestingly, inactivation of the chromosomal hyfR gene had no effect on hyfA-lacZ expression. Purified HyfR was found to specifically interact with the hyf promoter/operator region. Inactivation of the hyf operon had no discernible effect on growth under the range of conditions tested. No Hyf-derived hydrogenase or formate dehydrogenase activity could be detected, and no Ni-containing protein corresponding to HyfG was observed.

Expression of the Escherichia coli yfiD gene responds to intracellular pH and reduces the accumulation of acidic metabolic end products.

The YfiD protein of Escherichia coli has been reported to be an acid-inducible protein. Here it is shown that expression of a yfiD::lac reporter fusion is enhanced up to 3 small middle dot5-fold during acidic growth. The anaerobic transcription factor FNR was confirmed as the major regulator of yfiD expression, and ArcA was found to enhance anaerobic yfiD expression, probably by displacing a repressing FNR dimer in the -93 small middle dot5 region of the promoter. Moreover, the pyruvate sensor PdhR was shown to act as a minor anaerobic repressor of yfiD expression. On the basis of its strong homology to the C-terminal region of pyruvate formate-lyase (PFL) it was predicted that YfiD would be a radical-containing enzyme. The YfiD radical was found to be introduced by the PFL-activase enzyme, but unlike PFL, AdhE did not deactivate radicalized YfiD. The extent of radical activation of YfiD was enhanced by low intracellular pH, and thus it was concluded that both yfiD expression and YfiD activity are affected by growth at low pH. The yfiD mutant strain JRG4033 excreted increased levels of organic acids compared to the parental strain when grown in chemostat culture under oxygen-starved conditions, consistent with the acid-inducibility of yfiD expression and the recently reported ability of YfiD to rescue the activity of oxygenolytically cleaved PFL.

Pyruvate oxidase contributes to the aerobic growth efficiency of Escherichia coli.

The metabolic importance of pyruvate oxidase (PoxB), which converts pyruvate directly to acetate and CO(2), 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 (mu(max)), growth yield (Y(max)) 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 CO(2) 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 (mu(max)), growth yields (Y(max)) and carbon conversion efficiencies were 16%, 14% and 24% lower than the parental values, and correspondingly more carbon was fluxed to CO(2) 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.

Enzymological and physiological consequences of restructuring the lipoyl domain content of the pyruvate dehydrogenase complex of Escherichia coli.

The core-forming lipoate acetyltransferase (E2p) subunits of the pyruvate dehydrogenase (PDH) complex of Escherichia coli contain three tandemly repeated lipoyl domains although one lipoyl domain is apparently sufficient for full catalytic activity in vitro. Plasmids containing IPTG-inducible aceEF-IpdA operons which express multilip-PDH complexes bearing one N-terminal lipoyl domain and up to seven unlipoylated (mutant) domains per E2p chain, were constructed. Each plasmid restored the nutritional lesion of a strain lacking the PDH complex and expressed a sedimentable PDH complex, although the catalytic activities declined significantly as the number of unlipoylated domains increased above four per E2p chain. It was concluded that the extra domains protrude from the 24-meric E2p core without affecting assembly of the E1p and E3 subunits, and that the lipoyl cofactor bound to the outermost domain can participate successfully at each of the three types of active site in the assembled complex. Physiological studies with two series of isogenic strains expressing multilip-PDH complexes from modified chromosomal pdh operons (pdhR-aceEF-IpdA) showed that three lipoyl domains per E2p chain is optimal and that only the outermost domain need be lipoylated for optimal activity. It is concluded that the reason for retaining three lipoyl domains is to extend the reach of the outermost lipoyl cofactor rather than to provide extra cofactors for catalysis.

Metabolic engineering in Escherichia coli: lowering the lipoyl domain content of the pyruvate dehydrogenase complex adversely affects the growth rate and yield.

Isogenic strains of Escherichia coli W3110 containing pyruvate dehydrogenase complexes with three (wild-type), two or one lipoyl domains per lipoate acetyltransferase (E2p) chain, were constructed. The maximum growth rates (mumax) for batch cultures growing in minimal medium containing different carbon sources showed that reducing the number of lipoyl domains adversely mumax value of the mutant containing one lipoyl domain per E2p chain was restored by the presence of compatible multicopy plasmids encoding PDH complexes with either one or three lipoyl domains per E2p chain. In glucose-limited chemostat cultures the protein contents of all strains were similar and substrate carbon was totally accounted for in the biomass and CO2 produced. However, the carbon efficiencies (percentage carbon conversion to biomass) were significantly lower when the lipoyl domain content of the E2p subunit was reduced from three to one. Similarly, the cellular maintenance energy (m(e)) and the maximum growth yield (Ymax) were lower in bacteria containing PDH complexes with fewer than three lipoyl domains per E2p chain. Wild-type values were restored by supplementing the medium with either casamino acids (0.01%) or acetate (up to 0.1 mM). The lower growth efficiencies of the mutants were further confirmed in competition experiments where equal numbers of genetically marked (NalR) mutant and wild-type bacteria were used to inoculate glucose-limited chemostat cultures (dilution rate 0.075 h-1). The mutants with one or two lipoyl domains per E2p chain were washed out, whereas in controls, the initial ratio of wild-type (Nals) to reconstructed wild-type (NalR) bacteria was maintained over 50 generations.