Bacterial Sensor Kinase/Response Regulator Systems PDF

BACTERIAL SENSOR KINASE/RESPONSE REGULATOR SYSTEMS 437 IV. ELECTRON TRANSPORT REGULATION Aerobic-anaerobic gene regulation in Escherichia coil: control by the Ar...

BACTERIAL SENSOR IV. ELECTRON TRANSPORT REGULATION Aerobic-anaerobic gene regulation control by the ArcAB and Fnr Summary A variety of pathways for carbon and electron flow in the bacterium Escherichia coli and in other enteric bacteria are differentially expressed depend- i n g c~n......u.r. h..o..t.h o r ama Ln~lJ~z ~e, ~, lLaav, environment. This review briefly summarizes the metabolic pathways operative during aerobic versus an:,crobi~ cell growth, and provides a regulatory overview for how the cell controls expression of the many genes involved in these processes. The cell p.as two distinctly different transcriptional regu- lators, consisting of the Fnr and the ArcA/ArcB regulatory proteins to accomplish this task. Together, they coordinate gene expression to adjust carbon flow with electron flow and energy generation so that cells can balance growth in an efficiently coupled manner. Introduction Escherichia coli, like many other enteric bacteria, has evolved the ability to grow on a variety of differ- ent carbon compounds including many mono- and (*) Correspondingauthor. 438 12th Pathways of carbon and electron flow during aerobic and anaerobic conditions During aerobic cell growth, glucose, and other mono and dL~.ccharides that may be directly ed to it are channeled into the major pathway for sugar utilization, the Embden-Meyerhof-Parnas pathway, that converts them to pyruvate with generation of two ATP and two NADH per mole glu- cose (fig. 1). Pyruvate is then converted CoA and CO 2 by pyruvate dehydrogenase with generation of additional reducing equivalent in the form of NADH. Acetyl-CoA is then subsequently oxidized by the tricarboxyfic acid cycle (TCA enzymes to CO 2. Many of the enzymes for these reactions and their corresponding genes (except for the glycolytic pathway enzymes) are fisted in table I. Depending on what other substrate(s) may be avail- able to the cell, one or more additional enzymes be synthesized to convert the substrate to mediate of t~.e above pathways so that it can be sub- sequently degraded (e.g., lactate; fig. 1). When oxygen becomes limiting or absent from the cell environment, the cell can utilise any ber of alternative electron acceptors (nitrate, trimethyl-amine-N-oxide (TMAO), dimethyl-sul- phoxide (DMSO), or fumarate) as a respiratory sub- strate in lieu of oxygen for electron transport-linked phosphorylation reactions. The acceptors (figure 2) and the oxidoreductases employed for their reduc- tion pins the corresponding genes are summarized in table I. NADH serves as the major electron donor during growth with glucose: other electron donors also couple electron flow to the quinone pool via their specific dehydrogenases (fig. 2). Ubiquinone (Q) is the intermediate electron carder during respiration with oxygen or nitrate, whereas menaquinone (MK) functions under anaerobic conditions with anaerobic terminal oxidoreductase enzymes. els of these two fipophilic carriers in the membrane also vary depending on the availability of oxygen (see below). If neither oxygen nor any of the alternative anaerobic electron acceptors are present, E. coii must resort to a fermentative mode of carbon catabolism also called mixed-acid fermentation (fig. 3). The reducing equivalents (i.e., NADH) generated during conversion of glucose to pyruvate cannot be reoxi- dized by electron transport reactions and must be used ~ t e a d to reduce partially oxidized carbon com- pounds derived from glucose. Thus, accumulation of ethanol, lactate and succinate results, along with ac- cumulation of the other oxidation products that in- c!ude acetate, formate, hydrogen and CO 2. The energy obtained by the cell using this mode of metabolism is limited to the 2 to 3 ATP per mole glu- cose obtained by substrate level phosphorylation reactions. This relatively low energy yield per BACTERIAL SENSOR KINASE/RESPONSE Table I. Enzymes and genes involved in Reaction Enzyme Aerobic carbon flow 1 pyruvate dehydrogenase 2 isocitrate synthase 3 aconitase 4 citrate dehydrogenase 5 ~-ketoglutarate dehydrogenase 6 succinate thiolkinase 7 succinate dehydrogenase 8 fumarase A (aerobic) 9 malate dehydrogenase 10 L-lactate dehydrogenase (aerobic) Anaerobic carbon flow (fermentation) 1 PEP carboxylase 2 lactate dehydrogenase (anaerobic) 3 pyruvate formate lyase 4 formate hydrogen lyase 5 acetaldehyde dehydrogenase 6 alcohol dehydrogenase 7 phosphotransacetylase 8 acetate kinase 9 malate dehydrogenase 10 fumarase B (anaerobic) 11 fumarate reductase Respiratory pathways (terminal oxidoreductases) 1 cytochrome o oxJdase 2 cytochrome d oxidase 3 nitrate reductase 4 nitrite reductase 5 DMSO/TMAO reductase 6 TMAO reductase 7 fumarate reductase Other 1 NADH dehydrogenase (aerobic) 2 formate dehydrogenase-N 3 glycerol-P dehydrogenase (aerobic) 4 glycerol-P dehydrogenase (anaerobic) 5 nitrite efflux 6 fumarate uptake (anaerobic) 7 nickel uptake 8 molybdate uptake 9 lactate uptake 10 superoxide dismutase Plus symbol (~ indicates positive control plus symbol ((~) indicates provisional Fnr indicates negative control (Le., transcriptional spaces indicate that either no control exists determined. 440 ~.2th Condition Electron Donors glycerol.~ Q fat~ adds NOH / glycerol formate hydrogen lfig. 2. Electron flow during The alternative electron donors, for aerobic (plus oxygen) and during for each terminal oxidoreductase reaction menaquinone. glu,cose I I | I I PEP pyruvate acetyl-CoA acetaldehyde a~l-P Fig. 3. Carbon flow during The fermentative reactions, indicated absence of the ~ternative electron acceptors. table I with the corresponding gene enclosed by boxes. BACTERIAL SENSOR KiNASE/RESPONSE while apparently superior in oxygen binding ability, is somewhat less efficient in generating a chemios- motic potential, and is the predominant enzyme un- der micro-aerophilic cell culture conditions (Rice and Hempfling, 1978). During anaerobic conditions, the cell can synthesize a respiratory nitrate reductase, en- coded by narGHJI for reducing nitrate, and a nitrite reductase (nirB) to reduce nitrite (Cole, 1988), although the physiological contribution of the latter enzyme is not well understood. When TMAO, an osmolyte common in many marine organisms, is available, it can be reduced by either of two TMAO reductase enzymes encoded by the dmsABC and torA genes (Gunsalus, 1992). DMSO plus other sulphox- ides and amine-N-oxides can also be reduced by the broad substrate specificity enzyme encoded by dmsABC (Weiner et al., 1988). A fumarate reduc- tase encoded byfrdABCD is synthesized only when both oxygen and nitrate are unavailable to The preference for utilization of these alternative electron acceptors by E. coli has been examined (reviewed in Gunsalus, 1992) and is in the order of oxygen, nitrate, TMAO, DMSO and fumarate, where little distinction is made by the cell for use of TMAO, DMSO and fumarate. The synthesis of these alter- native anaerobic respiratory enzymes is modulated in response to oxygen, nitrate and iron availability (Cotter and Gunsalus, 1992). The latter is an essen- tial cofactor required for formation of multiple Fe-S centres contained in each of these anaerobic func- tion oxidoreductases. Depending on the availability of other respirato- ry substrates such as formate and hydrogen, which can serve as electron donors for anaerobic electron transport-linked phosphorylation reactions, addi- tional enzymes can be synthesized to oxidize these compounds. The metabolism of formate and hydro- gen is complex, as there are three distinct hydrogenase activities and two formate de- hydrogenases plus associated proteins required for nickel, selenium, and molybdenum cofactor processing. Other aerobic/anaerobic controlled functions The levels of several other proteins in the cell also vary depending on whether aerobic or anaerobic growth conditions are encountered. These include proteins that mediate uptake of some of the carbon compounds used as aerobic or anaerobic substrates (e.g., lactate, glycerol) or as anaerobic electron ac- ceptors such as nitrate (DeMoss and Hsu, 1991), fumarate (Engel et al., 1992), and presumably TMAO and DMSO. Other membrane-associated transporter proteins exist for uptake of metals such as nickel required for synthesis of certain anaerobic enzymes including hydrogenase. The levels of en- zymes for synthesis of the ubiquinone and menaqui- 442 12th dehydrogenas~, includir,g L-lactate dehydrogenase, D-lactate dehydrogev~e, D-amino acid dehydro- genase and pyruvatc: dehydrogenase, enzymes in- volved in reactions supplying carbon precursors to the TCA, were a!so abnormally elevated. Finally, enzymes in fatty acid degradation (acyl-CoA de- hydrogenase and 3-hydroxyacyl-CoA dehydrogenase) and an enzyme in the glyoxylate shunt (isocitrate lyase) were elevated in the mutant. The regulatory gene responsible for the apparent anaerobic repres- sion of these aerobically expressed structural genes was designated arcA for aerobic respiratory control. Gene mapping and complementation studies locat- ed the areA gene nearby 0 min on the E. coli chro- mosome and showed that arcA was identical to the dye gene previously identified by another phenotype, i.e., loss of resistance to toluidine blue and loss of sex factor expression (Buxton and Drury, 1984). DNA sequence analysis of arcA (Drury and Buxton, 1985) indicated that the gene encodes a 29,000 dalton polypeptide of 238 amino acids (fig. 4A). Compari- son with other bacterial proteins revealed that arcA encodes a regulatory protein of a two-component regulatory system which is prevalent in many prokaryotic organisms (Iuchi et ai., 1990a; reviewed in Stock et al., 1989). Subsequently, a second gene which had a similar pleiotropic phenotype, but which was unlinked to the arcA gene, was located at 69.5 min of E. coli chromosome, and was designat- ed arcB (luchi et aL, 1989). DNA sequence analysis A ! ArcB ~m B i Fm- ! 6 Sensor Fig. 4. Domains of the ArcA/ArcB The boxed regions represent to the N-terminal amino acid. Functions ed above or below the boxed regions. DNA recognition domain. The vertical cysteine; H, histidine; D, aspartate. BACTERIAL SENSOR KINASE/RESPONSE cognate sensor protein, for example, in CheY and in NtrC (Sanders et al., 1989, 1992). Among the sensor proteins, there is a subgroup to which ArcB belongs that has both a transmitter domain and a receiver domain (fig. 4A). This subgroup includes FrzE of Myxococcus xanthus (McCieariy and Zus- man, 1990), RcsC of E. coil for cell-capsule forma- tion (Stout and Gottesman, 1990), BvgS of Bordetella pertussis for toxin production (Miller et al., 1992) and VirA for Agrobacterium tumefaciens for host recog- nition and transformation (Stachel and Zambryski, 1986). The mechanism for ArfB/ArcA signal transduc- tion has been recently demonstrated to involve a transphosphorylation event that occurs in a some- what different manner from other two-component systems. Using purified ArcA protein and a truncat- ed ArcB protein ('ArcB) that lacks the transmem- brane region, Iuchi and Lin (1992a) demonstrated that 'ArcB was autophosphorylated when incubat- Aerobic State A ArcA, ArcB Regulator t~ membrane -H D I O I inactive i B Fnr Regulator inactivr I Fig. 5. Model for ArcA/ArcB (panel The cytoplasmic membrane is states of Fnr and ArcA are indicated. phate group. The coordination of A~J 12th mined using phosphorylation assays with the wild- type, ArcB aspartate substitution, and the ArcB receiver domain deletion protein mutant (luchi, 1993). This study demonstrated that ArcB can trans- fer a phosphate group from histidine 292 to aspar- tare 576 in the ArcB receiver domain and to aspartate 54 of ArcA (fig. 5A). The phosphate group at aspar- tare 276 of the ArcB receiver domain can also b~ released by an intrinsic phosphatase activity. The ~: ~al that ArcB responds to is unknown. Operatioaally, one can easily see from the in vivo data that the ~ e s controlled by the ArcA/ArcB regulators vary in their level depending on whether oxygen is a~ilable to the cell. However, evidence has not yet been obtained to implicate molecular oxygen directly in the ArcB sensing process. Iuchi (1993) re- cently proposed that the presence of reduced carbon compounds in the cell may be detected by ArcB. The presence of lactate decreased phosphatase activity of ArcB in vitro, thus moving the equilibrium towards elevated phosphorylation of ArcB at aspartate 576. This study also suggests that the receiver domain in ArcB may facifitate phosphate group transfer be- tween ArcB and ArcA if its aspartate residue is phos- phorylated. ArcB interferes with that process in an unphosphorylated state. W'nether reduced carbon compounds such as lactate can somehow generate a signal directly for detection by ArcB or whether it indirectly leads to the formation of some is not dear, as crude cell membranes were employed in these stucFles. The in vitro phosphorylation of Iuchi 0uchi, 1993) also showed that cellular metabofites accumulated during anaerobic growth (for example, lactate, acetate, pyruvate and NADH) increased the amount of in vitro phosphorylation of ArcB by inhibiting phosphatase activity of ArcB. By impfication, the regulator portion of ArcB may somehow be involved in sensing or attenuating the environmental signal in vivo. Thus, whatever signal is used by the ArcB/ArcA regulatory modu- Ion, the predominant effect is to modulate the synthesis of many enzyrnes in response to oxygen availability. Since the initial studies of Iuchi and Lin (1988) that documented the effect of arcA mutations on derepression of sdh-lacZ expression under anaerobic conditions, transcriptional control by ArcA has been documented for the cydAB and cyoABCDE operons which encode the cytochrome o and cytochrome d oxidase enzyme complexes (Iuchi et al., 1990; Cot- ter and Gunsalus, 1992), the sodA gene that encodes the manganese superoxide dismutase (Compan and Touati, 1993; Hasan and Sun, 1992), and the pfl gene which encodes the pyruvate formate lyase enzyme in- volved in pyruvate cleavage during fermentative con- ditions (Sawers and Suppmann, 1992). ArcA control of many other aerobic carbon pathway genes (fig. 1, table 1) is presumed to be mediated directly by ArcA tein. The amino-end of Fur contains four cysteine residues of which three, at positions 20, 23 and 29, are essential for Fur function (Melville and Gunsa- lus, 1990; Sharrocks et al., 1990; Spiro and Guest, 1988)" a fourth essential cysteine residue occurs at position 154 (fig. 4B). Unlike Crp, Fur is unable to bind c-AMP (Unden and Duchene, 1987). With the demonstration that the C-terminal region of Fur con- tains a helix-turn-helix motif for recognition and binding to DNA (Spiro and Guest, 1990), one may operationally distinguish the region specifying DNA binding from the cysteine-rich region required for signal detection (fig. 4B). The Fur protein was first isolated as a protein of Mr 28,000 lacking the 9 amino-terminal amino acids 0dnden and Guest, 1985; Trageser et al., 1990) due to proteolytic cleavage during isolation. Later, using a strain with reduced protease activity, Green et al. (1991) succeed- ed in isolating an Fur protein of Mr 30,000 which exhibited specific recognition of a consensus Fur site (as revealed by DNase I footprinting). Subsequent purifications of the Fur protein have been reported. They have allowed a preliminary analysis of DNA binding, lability to sulphydryl reactive reagents, and of metal binding (described below). However, the iso- lation of Fur protein in a biologically active form that binds DNA in an oxygen responsive manner has not yet been achieved. Very little is yet known about how Fur senses and respond to anaerobiosis. In vivo, the ability of me- tal chelating agents to deprive the cell of divalent me- tals including iron was correlated with the loss of Fur-dependent gene regulation (Spiro et al., 1989; Trageser and Unden, 1989). The four cysteine l l ~ , ~ l l. g U l ~ , t ~ ll.lL / ' l l l binding domain based on their reactivity to sul- phydryl reagents. Metal chelating studies showed that iron addition was the most effective in restoring Fur- dependent gene expression (Engel et al., 1991 ; Shar- rock et al., 1991). Activity of Fur was also modulat- ed by changing the oxidation/reduction potential of the cell environment by addition of hexacyanoferrate (III) (Unden et al., 1990). When the midvolt poten- tial of the culture medium was raised above +400 mV, Fur-dependent activation of f r d A - l a c Z expression was impaired even when cells w,.re or,,,~,,..~ anaerobically. This in vivo experiment suggests that molecular oxygen is not required for interconversion of Fur from an active to an inactive form. Iron has been detected in highly purified Fur preparations at a ratio of 0.2 to 1.1 molecules of iron per monomer of Fur (Mr = 30,000), which was inversely related to free sulphydryl content (Green et al., 1991). However, there is no evidence for the presence ferredoxin-like [Fe-SI centre in Fur as determined by UV-visible or by epr spectroscopy. Reduction of bound iron from the Fe 3+ to the Fe z+ oxidation state has been proposed to allow Fur to sense the anaerobic state to activate the DNA binding func- reductase: Cotter and Gunsalus, 1989), and frdABCD (fumarate reductase: Jones and Gunsalus, 1987) (reviewed in Gunsalus, 1992). It also works as an anaerobic activator of the pfl genes for pyruvate formate lyase (Sawers and Suppmann, 1992), f d n G H l for formate dehydrogenase N (Berg and Stewart, 1990), aspA for aspartase (Woods and Guest, 1987), fumB for the anaerobic fumarase B (Woods and Guest, 1987; Bell et al., 1989), and glpA for glycerol-3-phosphate dehydrogenase (Kuritzkes et ai., 1984); table I. For several promoters ing those for cyoABCDE (cytochrome o oxidase: CoRer et ai., 1990), cydAB (cytochrome d oxidase: Cotter et al., 1990), ndh (aerobic NADH de- hydrogenase II gene: Spiro et al., 1989), sdhCDAB for succinate dehydrogenase (Iuchi and Lin, 1988; Park and Gunsalus, 1994), and sodA for superoxide dismutase, Fnr functions as a transcriptional repressor. The Fnr DNA binding site was proposed to be a 22-base pair sequence having partial two-fold sym- metry (Spiro and Guest, 1990) based on comparison of Fnr-dependent promoters: subsequent tests have been performed in vitro with mutated natural and synthetic binding sites (Bell et al., 1989; Jayaraman et ai., 1989; Li and DeMoss, 1988). Direct binding of Fnr to a synthetic Fur binding site called FFmelR OWGATgtacATCAA) and to the ndh promoter region which contains an Fnr repression site was ac- complished using in vitro footprinting approaches and has confirmed the predicted Fur binding se- quence (Green et ai., 1991 ; Sharrock et al., The location for Fnr binding at Fur-activated e r ~ r c ~r~no2re tn ~ ;r~u~r~.~nt. TI'GATnnnATCAA sequence recognized by Fnr is centred at position - 40 to - 42 and adjacent to the signa-70-containing RNA polymerase domain on the DNA. It has also been shown that by substituting a single nucleotide in the core motif of an Fnr half site of TTGAT to GTGAT, an Fnr-dependent promoter became Crp-cAMP dependent (Bell et al., 1989). These studies support the notion that the Fnr pro- tein acts much like Crp in its ability to activate tran- scription, presumably by aiding RNA polymerase in formation of a DNA open complex. Unlike Crp, no Fur binding sites for gene activation have been lo- cated further upstream than position - 40 relative to the start of transcription. Thus, Fnr and Crp clearly differ in their DNA binding specificity and in their means for signal detection. For Fnr-repressed promoters, the sequence and location(s) of Fnr binding sites is not yet well estab- fished. Sharrocks et ai. (1991) obtained an Fnr foot- print at the ndh promoter which overlaps a perfect Fnr consensus site centred at position - 50 (atypical location of Fnr activator sites) and also identified a consensus Fur half site (TTGAT) located well up- stream between position - 83 and - 105. The pro- BACTERIAL SENSOR KINASE/RESPONSE a 12-fold activation of cydAB expression during a shift from oxygen-rich to oxygen-limiting (microaerophilic) conditions. As the availability of oxygen is further limited, Fnr is then converted into an active form and represses cydAB expression by about 4-fold from the peak level of expression seen during microaerophilic g r o ~ h conditions. The net effect on cydAB expression is an overall 3-fold in- crease from oxygen-saturating growth conditions to the complete absence of oxygen, conditions where the cell is now poised with elevated levels of cytochrome d oxidase for respiration to oxygen should it be re- encountered. This model predicts different thresholds for sensing of a common signal by Fnr and ArcA/ArcB, or alternatively, the detection of two differer~t signals (one by Fnr and another by ArcA/ArcB) that report alternative states that exist in the cell during microaerophilic to anaerobic conditions. This model would rationalize the seem- ingly redundant need for two anaerobic sensor- regulators in E. coil The pattern for ArcA and Fnr control of cyoABCDE expression differs from that observed for cydAB as both regulators function as repressors of transcription (Cotter and Gunsalus, 1992; Fu et al., 1991). Again by the general model of Cotter, cyoABCDE is predicted to be regulated in two stages depending on the availability of oxy- gen: upon shift to microaerophilic conditions, ArcA would act first to suppress cyoABCDE transcription by about one half, while Fnr then represses cyoABCDE expression a further 20-fold when oxy- gen is fully depleted. The two regulators appear to function independently of one another in their con- trol of cyoABCDE expression (Cotter and Gunsalus, 1992), in contrast to cydAB where a functional protein is required for Fnr to work normally. This two-stage level of control for cyoABCDE and cydAB expression would insure that E. coli appropriately justs the synthesis of each aerobic respiratory com- plex to allow rapid adaptation for energy during transitions to and from microaerophilic growth. An experimental test of this model is in progress. The sdhCDAB genes that encode the TCA cycle pathway enzyme, succinate dehydrogenase, are also regulated by both aerobic/anaerobic regulators. Suc- cinate dehydrogenase is a membrane-bound enzyme that functions as a member of the aerobic electron transport pathway to donate electrons to ubiquinone (fig. 2). The expression of the sdhCDAB operon is repressed by ArcA under both aerobic and anaero- bic conditions (Park and Gunsalus, 1994): a 10-fold repression is seen under anaerobic conditions, ~vhereas a 2-fold repression is seen under oxygen saturating conditions. Far also works as of sdhCDAB expression when the cells are growing anaerobically (Park and Gunsalus, 1994). Whether the sdhCDAB genes are expressed at an intermedi- ate level under microaerophilic conditions, like that References redox regulation of the Cob operon in Salmonella typMmurimn. Mol. MicrobioL, 6, 1491-1494. A.v~!ey,M.J. & Stadtman, T.C. (1988), Anaerobic induc- tion of Escherichia coil formate dehydrogenase is en- hanced by gyrase inactivation. Proc. Natl. Acad. Sci. USA, 85, 1023-1027. Bell, A.I., Gaston, K.L. & Busby, S.J.W. (1989), Cloning of binding sequences for the Escherichia coli transcrip- tional activator Fnr and Crp location of bases involved in discrimination between Fnr and Crp. Nucl. Acids Res., 17, 3865-3874. Bentley, R. & Mcganthan, R. (1982), Biosynthesis of vita- min K (menaquinone) in bacteria. Microbiol. Rev., 46, 241-280. Berg, B.L. & Stewart, V. (1991), Structural genes for nitrate-inducible formate dehydre.genase in Es- cherichia coil K-12. Genetics, 125, 691-702. Bonnefoy, V., Pascal, M.-C.~ Ratouchniak, J. & Chip- pauxm, M. (1986), Autorcgulation of the nar operon encoding nitrate reductase in Escherichia coil Mol. Gen. Genet., 204, 180-184. Buxton, R~S. & Drury, L.S. (1984), Identification of the dye gene product, mutational loss of which alters en- velop protein composition and also affects sex fac- tor F expression in Escherichia coli K-! 2. Mol. Gen. Genet., 194, 241-247. Cole, J. (1988), A~imilatory and dissimilatory, reduction of nitrite to ammonia, in "Society for General Microbiology Symposium 42. The Nitrogen and Sul- phur Cycles" (Cole, J. & Ferguson, Cambridge University Press, Cambridge. Compan, I. & Touati, D. (1993), Interaction of six global transcriptional regulators in expression of manganese superoxide dismutase in Escherichia coli K-12. J. Bac- terioL, i75, 1687-1696. Cotter, P.A., Chepuri, V., Gennis, R.B. & Gunsalus, R.P. (1990), Cytochrome o (cyoABCDE) and d (cydAB) oxidase gene expression in Escherichia coli is regu- lated by oxygen, pH, and thefnr gene product. J. Bac- terioi., 172, 6333-6338. Cotter, P.A. & Gunsalus, R.P. (1989), Oxygen, nitrate and molybdenum regulation of dmsABC gene expression in Escherichia coli. J. Bacterioi., 171, 3817-3823. Cotter, P.A. & Gunsalus, R.P. (1992), Contribution of the f n r and arcA gene products in coordinate regulation of the cytochrome o (cyoABCDE) and d (cydAB) oxidase genes in Escherichia coli. FEMS Microbiol. Lett., 91, 31-36. DeMoss, J.A. & Hsu, P.Y. (1991), NarK enhances nitrate uptake and nitrate excretion in Escherichia coil J. Bacteriol., 173, 3303-3310. Dorman, C.J., Burr: G.C., Bhriain, N.N. & Higgins, C.F. (1988), DNA supercoiling and the anaerobic and growth phase regulation of tonB gene expression. J. Bacteriol., 170, 2816-2826. Drury, L.S. & Buxton, R.S. (1985), DNA sequence analy- sis of the dye gene of Escherichia coli reveals amino acid homology between Dye and OmpR proteins. J. Biol. Chem., 260, 4236-4242. Engel, P., Kramer, R. & Unden, G. (1992), Anaerobic fumarate transport in Escherichia coli by a fnr- dependent dicarboxylate uptake system which is BACTERIAL SENSOR KINASE/RESPONSE utilizing delcfic...... lac fusions of cyo and cyd. J. Bacteriol., 1;'2.... 0-6025. Jayaraman, P.S., Gaston, K.L., Cole, J.A. & Busby, S.J.W. (1989), l~:utational analysis of sequences at the Fnr-dependent nirB promoters in Esci~eric,~iacoll. Nucl. Acid Res., 17, 137-145. Jones, H.M. & Gunsalus, R.P. (1987), Regulation of Es- cherichia coil fumarate reductase (frdABCD) operon expression by respiratory electron acceptors and the f n r gene product. J. BacterioL, 169, 3340-3349. Kiley, P.J. & Reznikoff, W. (1991), Fnr mutants that ac- tivate gene expression in the presence of oxygen. Knappe, J. & Sawers, G, (1990), A radical-chemical route to Acetyl-CoA: the anaerobically induced pyruvate formate lyase system of Escherichia coli. FEMS MicrobioL Rev., 75, 383-398. Kuritzkes, D.R., Zhang, X.-¥'. & of ¢(glp-lac) in studies of respiratory Escherichia coli anaerobic sa-glycerol-3-phosphate de- hydrogenase genes (glpAB). J. Bacteriol., 157, 591-598. Lambden, P.R.,& Guest, J.R. (1976), Mutants of Es- cherichia c~Mi KI2 unab!e to use fumarate as an anaerobic electron acceptor. J. Gen. Microbiol., 97, 145-160. Lazazzera, B.A., Bates, D.M. & Kiley, P.J. (1993), The activity of the Escherichia coli transcription factor Fnr is regulated by a change in oligomeric state. Genes & Dev., 7, i993-2005. Li, S.F. & DeMoss, JoA. (1988), Location of sequences in the nar promoter of Escherichia coli required for regu- lation by Fnr and NarL. J. Biol. Chem., 263, 13700-13705. McCleary, W.R. & Zusman, D.E. (1990), Purification and characterization of the Myxococcus xanthus FrzE pro- tein shows that it has autophosphorylation activity. J. Bacteriol., 172, 6661-6668. Melville, S.M. & Gunsalus, R.P. (1990), Mutations infnr that alter anaerobic regulation of electron transport- associated genes in Escherichia coil J. Biol. Chem., 265, !8733-i8736. Miller, J.F., Johnson, S.A., Black, W.J., Beattie, D.T., Mekalanos, J.J. & Falkow, S. (1992), Constitutive sensory transduction mutations ha the Bordetellaper- tussis bvgS gene. J. Bacteriol., 174, 970-979. Ninfa, A.J. & Bennett, R.L. (1991), Identification of the site of autophosphorylation of the bacterial protein kinase/phosphatase NRII. J. Biol. Chem., 266, 6888-6893. Park, S.J., Cotter, P.A. & Gunsalus, R.P. (1992), Au- toregulatic, n of the arcA gene of Escherichia coil Amer. Soc. Microbiol. (Abstracts of the annual Meet- ing), 92, 207. Park, S.-J. & Gunsalus, R.P. (i994), Regulation of suc- cinate dehydrogenase (sdhCDAB) operon expression in Escherichia coli in response to carbon supply and anaerobiosis: role of ArcA and Fnr. Poole, R.K. & Ingledew, W.J. (1987), Pathways of elec- trons to oxygen, pp. 170-200, in "Escherichia Salmonella typhimurium: cellular and molecular biology" (F.C. Neidhardt). American Society for Microbiology, Washington, DC. Rezinkoff, W.S. (1992), Catabolite gene activator protein activation of lac transcription. J. Bacteriol., 174, 655-658. 450 component regulator of capsule synthesis in Es- cherichia coil J. Bacteriol., 172, 659-669. Tardat, B. & Touati, D. (1993), Iron and oxygen regula- tion of Escherichia coli MnSOD expression: compe- tition between the global regulators Fur and ArcA for binding to DNA. Mol. Microbiol., 9, 53-63. Trageser, M., Spiro, S., Duchene, A., Kojro, E., Fahren- holz, F., Guest, J.R. & Unden, G. (1990), Isolation of intact FNR ~pmtein(Mr 30,000) of Escher/ch/a coll. Moi. M'~biol~; 4, 21-27. Trageser, M. & Unden, G. (1989), Role of cysteine residues and of metal ions in the regulatory functioning of FNR, the transcri[~tional regulator of anaerobic respi- ration in ~ CoIL Mol. Microbiol., 3, 593-599. Unden, G, (1988), I~Terential roles for menaqulnone and demethytmenaquinone in anaerobic electron transport of E. co/i and theirfm'-independent expression. Arch. Microbiol., 150, 499-503. Unden, G. & Duchene, A. (1987), On the role of cyclic AMP and the Fnr protein in Escherichia coli grow- ing anaerobically. Arch. Microbiol., 147, 195-200. Dual interacting two-component regulatory mediate nitrate- and nitrite-regulated in Escherichia coli Sections o f Microbiology and Genetics & Enterobacteria are facultative aerobes. These or- ganisms efficiently Use oxygen as a terminal electron acceptor for respiration, but they also synthesize a variety of anaerobic respiratory chains that allow for respiration with diverse electron donors and accep- tors. The control of gene expression in response to oxygen is reviewed in the chapter by Gunsalus and Park in this symposium. This chapter summarizes an additional level of anaerobic respiratory gene expres- sion, in response to the alternate electron acceptors nitrate (NO~) and nitrite (NO~). During anaerobiosis, utilization of respiratory electron aceeptors is subject to hierarchical regula- tion. Nitrate is the most efficient anaerobic electron acceptor, and the presence of nitrate in the culture medium induces synthesis of the nitrate respiratory chain enzymes, formate dehydrogenase-N (encoded KINASE/RESPONSE REGULATOR SYSTEMS 437 in Escherichia coil: regulons R.P. Gunsalus t*) and S.-J. P a r k Department of Microbiology and Molecular Genetics, University of California, Los Angeles, CA 90024 (USA) disaccharides, amino acids, fatty acids and short chain alcohols or acids. It can derive energy via elec- tron transport-linked phosphorylafion reactions dur- ing aerobic as well as during anaerobic conditions when any of the alternative electron acceptors are ;~ p resent ; -- ,. ha ~ l a ra~ o x y g e n AO ,. l,,till ,,~IA ~vQ]tlfllJlg;. I llg;; ~g;ll g,ggll gU~lIJ ~Pgg|IUPlIll ~lllllJlg; l~---im~tl- tation of sugars in order to yield ATP by substrate level phosphorylation when no electron acceptors are present in the cell environment. Thus, cell survival and growth are possible over a wide range of condi- tions. To choose which of the alternative pathways for carbon and electron flow to use, the cell must first detect the presence of the alternative substrates, and in turn, "switch" on or off the appropriate path- way(s). This control is, in part, due to transcription- al regulation of the genes that encode the individual enzymes of the aerobic and anaerobic pathways. It ensures that unneeded enzymes are not synthesized, whereas others that are needed are present in ~=-~fi- cient amounts. This control aids in optimizing cellular energy generation for support of macro- molecular biosynthesis, assembly and cell division. This review discusses the roles of the ArcA/ArcB and Fnr regulators in the aerobic/anaerobic control processes. FORUM IN MICROBIOLOGY glucose I I I I convert- l~mte..~ pyruvate lO the net to acetyl- oxaloacetate citrate cycle) 8 may an inter- ~7 a-ketoglutatate succinate 5 ~ of a num- nitrite, ~ succinyl-CoA ~['~ Fig. 1. Carbon flow during aerobic cell growth in E. coli. The reactions that convert glucose to pyruvate are in- dicated by the dashed arrow, while the solid arrows denote the cleavage of pyruvate to acetyl-CoA and CO2 (reac- tion 1), the steps of the TCA cycle (reactions 2-9) and utili- zation of lactate by conversion to pyruvate (reaction 10). The enzyme names and their corresponding genes are list- ed in table I. The boxed compounds indicate the products resulting from aerobic catabolism of glucose. any of the The lev- cytoplasmic molecule of glucose utilized during fermentation is in marked contrast to the numerous ATP molecules generatcd by electron transport linked phosphoryla- tion reactions coupled to acrobic as well as anaero- bic respiration when any of the alternative electron acceptors are present. Respiratory enzymes E. coli can synthesize two distinct respiratory en- zymes for reduction of oxygen. These are the cytochrome o oxidase and the cytochrome d oxidase complexes that are synthesized under oxygen-rich and oxygen-limited conditions, respectively. The cytochrome o oxidase, encoded by cyoABCDE, is energetically more efficient than cytochrome d oxi- dase (cydAB), but has a weaker affinity for oxygen (Poole and Ingledew, 1987). Cytochrome d oxidase, REGULA TOR SYSTEMS 439 aerobic/anaerobic carbon flow and in cellular respiration. Genes Fnr ArcA aceEF, lpd gltA (@) acn (0) icd (@) sucAB (@) sucCD tO) sdhCDAB ® ® fumA mdh G) lctD ppc ldhA Pfl @ @ fdhF, hyc acd adhE pta ackA mdh (@) fumB @ frdABCD @ cyoABCDF~ @ cydAB (~) narGHJl @ nirB @ dmsABC (~ torA (@) frdABCD ndh @ fdnGHI @ glpD @ glpACB @ narK @ ? (@) hydC modABCD lctP sodA @ @ (i.e., transcriptional activation of gene expression) by Fnr or by ArcA, while an enclosed or ArcA control based on assay of enzymes wild-type and mutant strains. A minus e symbol repression of gene expression) of the indicated genes by Fnr or by ArcA. Blank due to Fnr and/or ArcA, or that the contributions of the two gene regulators are not yet F O R U M I N M I C R O B I O L O G Y Quinone Electron Acceptors ! 2 ~ oxygen 43 ~nitrate MK nitrite 5 - DMSO/TMAO 6 _ TMAO - fumarate aerobic and anaerobic respiratory conditions. quinone carriers and electron acceptors used by E. coli are shown anaerobic respiratory conditions (minus oxygen). The enzymes and their genes are listed in table I. Q, ubiquinone; MK, t~~ ox~l~tate ~ ~ ~ isuccinateI 9 I0 II I ' llac' l ~ formate ~ anaerobic cell growth (fermentation conditions). by the solid arrows for glucose utilization, occur in the The enzymes indicated by the numbers are listed in designations. The endproducts of glucose breakdown are REGULATOR SYSTEMS 441 none cofactors of the electron transport chains also appear to be modulated in response to aerobic versus anaerobic cell growth: the Q and MK levels are known to vary (Bently and Meganathan, 1982; Unden, 1988; Gibert et al., 1988). Although E. coil does not synthesize vitamin Bt2 , the related emetic bacterium, Salmonella typhimurium, controls the synthesis of B~2 in response to lack of oxygen (An- dersson, 1992). Lastly, the level of the manganese- superoxide dismutase varies in the cell depending on conditions of oxygen availability (Hassan and Fridovich, 1977). Aerobically, the enzyme is synthe- sized at elevated levels to aid in detoxifying superox- ide radicals generated by intra- or extracellular processes. Control of gene expression by switch to anaerobiosis the cell. Two independent regulatory systems have been identified in E. coli that serve major roles in modulat- ing expression of the pathways for carbon and elec- tron flow in response to oxygen availability. They are the Fur and the ArcA/ArcB regulons. Each element can respond independently to the shift between aerobic and anaerobic cell growth conditions, and can act as either a positive or a negative effector of gene expression. Whereas Fnr is active only under anaerobic conditions, ArcA/ArcB functions under both aerobic and anaerobic conditions. These two regulatory elements act in combination to coordinate the pathways for carbon flow from pyruvate to the various end products of aerobic and anaerobic metabolism, to electron flow via the respiratory path- ways. A partial list of the genes controlled by the Fnr and/or ArcA/ArcB regulatory circuits is s h o ~ -in table I. A short description of these two aero- bic/anaerobic transcriptional regulators follows. Regulation by the ArcA/ArcB proteins In a genetic search for an anaerobic regulator of the TCA cycle enzyme, succinate dehydrogenase, Iuchi and Lin (1988) isolated arcA mutants that ex- hibited abnormally elevated levels of the enzyme dur- ing anaerobic cell culture, conditions where it is normally very low. This search employed a ¢(sdh- lacZ) fusion to screen for the expression of the sdhCDAB operon for succinate dehydrogenase. The areA mutation had pleiotropic effects on expression of many other aerobic function enzymes including other TCA cycle enzymes, fatty acid degradation enzymes, some flavoprotein dehydrogenases and ubiquinone oxidase (Iuchi and Lin, 1988). The TCA cycle enzymes include citrate synthase, acommse, isocitrate dehydrogenase, 0t-ketoglutarate dehydro- genase, succinate dehydrogenase, fumarase and malate dehydrogenase. In addition, levels of several F O R U M I N M I C R O B I O L O G Y of arcB revealed the idenlLityof the second member of the two-component system; the deduced protein contains 778 amino acids (Mr 87,900) and is a trans- membrane protein as predicted by hydrophobic profile analysis (fig. 4A). The cellular location of ArcB was subsequently shown to be membrane- associated (Iuchi et al., 1990a). ArcB serves the role of a sensor protein to detect an environmental signal, while ArcA is the receiver regulator protein of the two-component system which modulates gene expression in response to that environmental signal. The members of the family of prokaryotic pro- teins called two-component system (reviewed in Stoc)c et ai., 1989) show striking homology in certain pro, tein domains, even thoug~ they recognize differedt environmental signals. Inmost examples, the system is composed of two proteins, a transmembrane sen- sor protein and a response regulator protein. The sen- sor protein contains a transmitter domain with one conserved histidine residue (e.g., histidine 292 of ArcB) which was shown to be the site of covalent phosphorylation in NRII (NtrB) and CheA (Ninfa and Bennett, 1991 ; Hess et al., 1988). The response regulator protein usually contains a helix-turn-helix DNA binding domain at the carboxyl terminus and a signal, receiver, domain, a~ the amino-terminus. The receiver domain includes a conserved aspartate residue (e.g~. aspartate 54 of ArcA), which has be(~fi demonstra-~ed to accept the phosphate from the Sensor Transmitter Receiver Domain 778 "l 6 t 238.~a'cA I I Receiver DNA Binding 298 I - DNA Binding (panel A) and the Fnr (panel B) regulatory proteins. the indicated protein with the numbering of amino acids relative assigned to the various domains of the proteins are indicat- The thin bars below the ArcA and Fro"proteins represent the bars in ArcB represent the two transmembrane regions. C, REGULATOR SYSTEMS 443 ed with 32p-y-ATP. When ArcA was incubated with phosphorylated 'ArcB, the phosphate group was rapidly transferred to ArcA (fig. 5A). Based on the differential stability of covalent phosphate groups toward acid, it was shown that AreA contains only an aspartyl-linked phosphate, whereas 'ArcB con- tains both aspartyl-phosphate and histidyl-phosphate linkages (fig. 5A). Mutational studies of the arcB gene indicate that histidine 292~ontaining transmit- ter domain of ArcB is the site for phosphorylation and that it is indispensable for ArcA/ArcB signal transduction process. In addition, either a deletion of the receiver domain in ArcB (fig. 4A) or the replacement of the aspartate residues located at po- sitions 533 and 576 within the receiver domain criti- cally affected ArcA-dependent gene expression in vivo (Iuchi and Lin, 1992b). This suggests that the C-terminal domain of ArcB is somehow involved in regulating the signal transduction event. In vitro, the role of this receiver domain of ArcB has been exa- Anaerobic State I I I I I ~ N I v' d I ATP ADP ATP ADP (J II -H-~ -H -H-® I I ~',, D| o-® ,o-® ,3 ,,3 ,2 Ii I ! active I I I I I ~-, I active I A) and Fnr (panel B) functioning during aerobic and anaero- bic cell growth conditions. indicated by the two thin horizontal lines. The active and inactive The circled P symbol represents a covalently attached phos- the iron [Fe] in Fnr and its oxidation state(s) is presently unknown. F O R U M I N M I C R O B I O L O G Y binding to DNA at their ArcA-controlled promoters. Inspection of the many ArcA/ArcB responsive genes has not revealed a potential DNA consensus bind- ing site for the ArcA protein. If ArcA acts like many other regulater~ of the two-component systems, one would expect ArcA to recognize (and protect) a 6- to 9-base pair site near each promoter it regulates. A partially purified ArcA protein was recently report- ed that could bind at the sodA promoter and pro- tect a 60- to 65-bp region of DNA extending from position - 2 7 to + 24 relative to the start of sodA transcription (Tardat and Touati, 1993). Hydroxy radical footprint analysis further revealed the ArcA- DNA interactions were on one face of the DNA and consisted of 5 to 6 tandem protected regions which span the - 35 to + 10 sites where RNA polymerase binds. The binding of ArcA to the DNA would thus occlude polymerase binding and transcription of sodA. If similar ArcA protection data can be ob- tained for other ArcA repressed promoters (e.g., cyoABCDE, sdhCDAB, etc.), it would suggest that ArcA has an atypical mode of protein-DNA inter- actions compared to other two-component regula- tors. Since ArcA can also activate gene expression, as in the case of the cydAB and pfl promoters, locali- zation of ArcA binding sites either adjacent to or upstream of the RNA polymerase binding site for those promoters is anticipated by analogy to the NtrC and OmpR systems. other signal studies Transcriptional regulation by Fnr The Fnr protein is the second major aero-."........" -- --a..... !1 UIU/i:ILII~CIUUIU [~UIdLU[ IUI [UIILIUIIIII~ ~UIIU UAI.)I u~- sion in E. coli. It has been shown to function as both an anaerobic activator and repressor of many anae- robically controlled genes (see table I: reviewed in Spiro and Guest, 1990; Gunsalus, 1992). Mutants defective in f n r were first isolated as strains lacking fumarate and nitrate reductase activity (Lambden and Guest, 1976). The fnr gene was mapped to 29 min on the E. coli chromosome, cloned, and its DNA sequence was determined (Shaw and Guest, 1982a,b). Comparison of the deduced amino acid se- quence of Fnr protein with other bacterial proteins revealed a striking similarity to the Crp cyclic-AMP binding protein of E. coli (Shaw et al., 1983). Sub- sequent protein modeling studies of Fnr have pro- posed a similar three-dimensional structure to Crp (J. Perry, personal communication), and together with the other information accumulated about Fnr (reviewed by Spiro and Guest, 1990; Unden and Trageser, 1991), suggests that Fnr activates gene ex- pression by a mechanism similar to that employed by Crp (Reznikoff, 1992) but with several distinc- tions. Fnr differs from Crp in that Fnr has an addi- tional 26 amino acids at the amino-terminus and 13 amino acids in the carboxyl-terminus of the pro- BACTERIAL SENSOR KINASE/RESPONSE REGULATOR SYSTEMS ~3 tion (Unden et al., 1990). Alternatively, a dissociation-reassociation of iron onto Fur could drive a conformational change in the protein to allow specific DNA binding only under anaerobic condi- tions (Green et al., 1991). Lastly, Fur has been pro- posed to respond to anaerobiosis by acylation of an essential thiol group or by reductive acylation of a disulphide bridge within the protein (Spiro and Guest, 1988). More recently, Fur purified under aerobic condi- tions was shown to stimulate transcription of a semi- synthetic promoter in an iron-dependent fashion (Green and Guest, 1993). Although this transcrip- tional process is not responsive to either anaerobio- sis or the presence of oxygen which would allow a dissection of the determinants of anaerobic sensing, it establishes a viable direction for future studies. f u r mutants which show star-activity, i.e., enhanced ability to perform Fur-dependent gene activation un- der aerobic conditions, have been isolated (Kiley and Reznikoff, 1991 ; Melville and Gunsalus, 1990). Some of the mutant Fur proteins were purified and shown to exhibit the ability to dimerize, although the wild- type protein did not under identical conditions (Laar- zazzera et al., 1993). A model was proposed where- by the ability of Fur to bind DNA is regulated by a change in equilibrium in vivo between monomeric Fur (inactive) and dimeric Fur (active). The active form of Fur would then bind to DNA to regulate gene expression during anaerobic cell conditions. Resolu- tion of these alternative models for Fur activation must await the purification and biochemical analy- sis of anaerobic responsive Fur protein. llO.V~; ILl"~7~ll 1 J l WJl..lllJ~lk,l[ L U lIJllll ¢'1 111~1[.1¢1 l - The level of Fur in the cell appears to be main- tained within a relatively narrow range: f u r expres- sion was shown to vary by about 25 % under aerobic versus anaerobic conditions as measured using a sin- gle copyfnr-lacZ reporter fusion (Jones and Gunsa- lus, 1987). A direct estimate of the amount of Fur present in the cell was 2,400 molecules per cell, as- suming a monomeric protein 0dnden and Duchene, 1990). Studies of Engel et al. (1991) demonstrated that reversible interconversions between active and inactive Fur occur independently of whether new Fur ~ A ~ VVAA protein is synthesized. Together, the above findings rule out several alternative mechanisms for controll- ing anaerobic pathway genes whereby (1) synthesis of new Fur is required under anaerobic conditions to replace protein inactivated irreversibly under aer- obic conditions, or (2) Fur levels are controlled by some other cellular regulator that detects the anaero- bic state (i.e., a cascade-type mechanism). of a Fur functions as an anaerobic activator of genes for the anaerobic terminal oxidoreductases includ- ing n a r G H J l (nitrate reductase: Bonnefoy et al., 1986; Li and DeMoss, 1988), nirB (nitrite reductase: Harbourne et al., 1992), d m s A B C (DMSO/TMAO 12th FORUM I N MICROBIOLOG Y tection data are consistent with an Fnr dimer bind- ing at each site, although a mechanism that accounts for the observed transcriptional repression at the ndh promoter is not yet evident. Neither site is sufficiently close to the promoter - 3 5 - 1 0 region to prevent RNA polymerase from binding to the DNA. Further studies are needed to explain how Fro- accomplishes negative control at the ndh and other Fur-controlled genes. includ- Genes under dual control by Fnr and ArcA/ArcB Several genes involved in anaerobic and aerobic metabolism are regulated by both the Fnr and the ArcA/ArcB regulatory proteins. One of these, the pfl gene, encodes pyruvate formate lyase which cata- lyses formation of acetyl-CoA and formate from pyruvate under anaerobic conditions (Knappe and Sawers, 1990): fig. 3. Synthesis of pyruvate formate lyase is increased about 10-fold upon a shift to anaerobic conditions (Sawers and Bock, 1988). This expression occurs from seven promoters (Sawers and Bock, 1989) where Fur is a major activator of transcription from each. ArcA is responsible for transcriptional activation from the two upstream promoters (Sawers and Suppmann, 1992). At least one accessory protein, IHF, is also required for this control (Sirko et al., 1993). The molecular basis for this unusual gene regulation appears complex. 1991). The genes for the two cytochrome oxidases for tl~ *t,r~;,~ed oxygen utilisation, cyoABCDE and cydAB, are also under dual control by Fnr and ArcA/ArcB. Cytochrome o oxidase encoded by cyoABCDE is the predominant enzyme under oxygen-rich growth con- ditions, in keeping with its low affinity for oxygen, whereas the cytochrome d oxidase (cydAB) exhibits high affinity for oxygen and is synthesized optimal- ly under microaerophilic conditions (Rice and Hemp- fling, 1978; Poole and Ingledew, 1987). The synthesis of the two enzymes is coordinated transcriptionally to ensure that each is present at optimal levels de- pending on the availability of oxygen. Fnr works as a repressor for expression of both cyoABCDE and cydAB under anaerobic conditions (Cotter et aL, 1990; Cotter and Gunsalus, 1992). In contra.--2, ArcA functions as an anaerobic repressor of cyoABCDE expression, while it serves as an activator of cydAB expression during aerobic and anaerobic conditions (Iuchi et aL, 1990b; Cotter and Gunsalus, 1992). Using shake flask culture techniques, Fu et al. (1991) found that a cydA-lacZ fusion was expressed max- imally under microaerophilic conditions. They pro- posed that Fnr also functions as an activator of cydAB expression, in contrast to the findings of Cot- ter and Gunsalus (1992). By the model of Cotter (Cotter and Gunsalus, 1992), ArcA is responsible for REGULATOR SYSTEMS g47 seen for the cydAB genes, is unknown. Lastly, tran- scription of the sodA gene that encodes manganese superoxide dismutase was found to be repressed by both Fnr and ArcA proteins under anaerobic condi- tions (Compan and Touati, 1993). Interestingly, Fnr is unable to function as a repressor if the ArcA pro- tein is not present in the cell (Hassan arid Sun, 1992). Whereas the regulatory pattern for Fnr and ArcA control of the cyoABCDE, sodA a~d sdhCDAB genes is similar (negative control by both Fnr and ArcA), the regulatory pattern differs for control of the pfl and cydAB genes (positive control by ArcA and either positive or negative control by Fnr). Each operon has acquired a somewhat different way to uti- rise the two aerobic/anaerobic regulators. Because in- tegration host factor (IHI~ protein is also required for expression of the sodA (Compan and Touati, 1993), pfl (Sirko et al., 1993) and cyoABCDE genes (Albrecht and Gunsalus, unpublished), a DNA bend- ing or looping event may be involved in bringing the transcriptional regulatory proteins together to effect their control. Is there aerobic/anaerobic control occurring in the cell that cannot be accounted for by Fnr or ArcA? Strains deleted for arcA andfnr exhibited a r~idual 1.5- to 2-fold aerobic/anaerobic difference in cydAB and cyoABCDE expression (Cotter and Gunsalus, 1992), whereas the frdABCD and narGHlI operons exhibit a 3- to 10-fold residual control, respectively (Jones and Gunsalus, 1987; Schroeder et al., 1993). The molecular basis for this control is not known, but it may involve DNA supercoiling, as has been proposed for aerobic/anaerobic control of other bac-........ '.... J " - - !.... 1988" r~.... ArcA aL, 1988). ad- Conclusions and prospects for future study generation Many questions remain concerning how E. coli controls the metabolic shift between the aerobic and anaerobic growth. Although two cell regulatory ele- ments have been identified, li~le is known about, what environmental signals are detected by each regula- tor or how the regulator is converted into an active form that transcriptionally regulates gene expression. Many microorganisms such as E. coli can adapt to continually changing conditions of oxygen availabil- ity. They have tailored the controls for their alter- native carbon and energy generation pathways accordingly to coordinate and optimize cell metabolism for cell growth. a repressor Portions of the studies described herein were supported by research grants from the National Institutes of Health (A121678 and GM49694) and the National Science Foundation (PCM-8402974). 12th F O R U M I N M I C R O B I O L O G Y different from the aerobic dicarboxylate uptake sys- tem. J. 8acteriol., 174, 5533-5539. Engel, P., Trageser, M. & Unden, G. (1991), Reversible.~defsson, D.I. (1992), Involvement of the Arc system in interconversions of the functional state of the gene regulator Fnr from Escherichia coli in vivo by 0 2 and iron availability. Arch. Microbiol., 156, 463-470. Fu, H.A., Iuchi, S. & Lin, E.C.C. (1991), The requirement of ArcA and Fnr for peak ~'xpression of the cyd ope- ron in Escherichia coli under microaerophilic condi- tions. MoL Gen. Genet., 2?6, 209-213. Gibert, K., Llagostera, M. & Barbe, J. (1988), Regulation of ubiG gene expression in Escherichia coll. J. Bac- terioL, 170, 1346-1349. Green, J., Trageser, M., Six, S., Unden, G. & Guest, J.R. (1991), Characterization of the Fnr protein of Es- cherichia coli, an iron-binding transcriptional regu- lator. Proc. R. Soc. Lond. B, 244, 137-144. Gunsalus, R.P. (1992), Control of electron flow in E~- cherichia coli: coordinated transcription of respira- tory pathway genes. J. Bacteriol., 174, 7069-7074. Harborne, N.R., Griffiths, L., Busby, S.J.W. & Cole, J.A. (1992), Transcriptional control, translation and func- tion of the products of the five operon reading frames of the Escherichia coli nir operon. Mol. Microbiol., 6, 2805-2813. Hassan, H.M. & Fridovich, I. (1977), Regulation of the synthesis of superoxide dismutase in Escherichia coli; induction by methyl viologen. J. Biol. Chem., 252, 7667-7672. Hassan, H.M. & Sun, H.C.H. (1992), Regulatory roles of Fnr, Fur and AcrA in expression of manganese su- peroxide dismutase in Escherichia coli. Proc. Natl. S.) (pp. 281-330). Acad. Sci. USA, 89, 3217-3221. Hess, J.F., Bourret, R.B. & Simon, M.I. (1988), Histidine phosphorylation and phosphoryl group transfer in bacterial chemotaxis. Nature (Lond.), 336, 139-143. Iuchi, S. (1993), Phosphorylation/dephosphorylation of the receiver moduie at the conserved aspartate residue controls transphosphorylation activity of histidine kinase in sensor protein ArcB of Escherichia coli. J. Biol. Chem., 268, 23972-23980. Iuchi, S. & Lin, E.C.C. (1988), arcA (dye), a global regula- tory gene in Escherichia coli mediating repression of enzymes in aerobic pathways. Proc. Natl. Acad. Sci. USA, 85, 1888-1892. Iuchi, S. & Lin, E.C.C. (1992a), Purification and phos- phorylation of the Arc regulatory components of Es- cherichia coli. J. Bacteriol., 174, 5617-5623. Iuchi, S. & Lin, E.C.C. (1992b), Mutational analysis of signal transduction by ArcB, a membrane sensor pro- tein responsible for anaerobic repression of operons involved in the central aerobic pathways in Escherichia coil J. Bacteriol., 174, 3972-3980. Iuchi, S., Cameron, D.C. & Lin, E.C.C. (1989), A second global regulatory gene (arcB) mediating repression of enzymes in aerobic pathways of Escherichia coli. J. Bacteriol., 171,868-873. Iuchi, S., Matsuda, Z., Fujiwara, T. & Lin, E.C.C. (1990a), The arcB gene of Escherichia coli encodes a sensor-regulator protein for anaerobic repression of the arc modulon. Moi. Microbiol., 4, 715-727. luchi, S., Chepuri, V., Fu, H.-A., Gennis, R.B. & Lin, E.C.C. (1990b), Requirement for terminal cytochromes in generation of the aerobic signal for the arc regulatory system in Escherichia coli: study REGULATOR SYSTEMS 449 Rice, C.W. & Hempff-lg, W.P. (1978), Oxygen-limited contir-lous culture and respiratory energy conserva- tion in E~cherichia coil J. Bacteriol.. 134, 115-124. Sanders, D.A., Gillece-Castro, B.L., Burlingame, A.L. & Koshland, D.E. Jr. (1992), PhosphoryPation site of NtrC, a protein phosphatase whose covalent inter- mediate activates tr -anscr;,:ption. J. Bacteriol., 174, 5117-5122. Sanders, D.A., Gillece-Castro, B.L., Stock, A.M., Burlin- game, A.L. & Koshland, D.E. Jr. (1989), Identifica- tion of the site of phosphorylation of the chemotaxis response regulator protein, CheY. J. Biol. Chem.,.I. Bacteriol., 173, 16-22. 264, 21770-21778. Sawers, G. & Bock, A. (1988), Anaerobic regulation of pyruvate formate lyase from Escherichia coli K-12. J. BacterioL, 170, 5330-5336. Sawers, G. & Bock, A. (1989), Novel transcriptional con- Lin, E.C.C. (1994), Use trol of the pyruvate formate lyase gene: upstream regulation of the regulatory sequences and multiple promoters reguLale anaerobic expression. J. Bacteriol., 171, 2485-2498. Sawers, G. & Suppmann, B. (1992), Anaerobic induction of pyruvate formate lyase gene expression is mediat- ed by the ArcA and Fnr proteins. J. Bacteriol., 174, 3474-3478. Schrfder, I., Daire, S. & Gunsalus, R.P. (1993), Activa- tion of the Escherichia coli nitrate reductase (narGHJl) operon by NarL and Fnr requires Integra- tion Host Factor. J. Biol. Chem., 268, 771-774. Sharrocks, A.D., Green, J. & Guest, J.R. (1990), In vivo and in vitro mutants of FNR the anaerobic tran- scription~ regulator in E. coil FEBS Leg.. 270, 119-122. Sharrocks, A.D., Green, J. & Guest, J.R. (1991), Fnr ac- tivates and represses transcription in vitro. Proc. R. Soc. Lona. B, 245, 219-226. Shaw, D.J. & Guest, J.R. (1982a), Amplification and product identification of the fnr gene Escherichia coli. J. Gen. MicrobioL, 128, 2221-2228. Shaw, D.J. & Guest, J.R. (1982b), Nucleotide sequence of thefnr gene and primary structure of the Fnr pro- rein of Escherichia coil Nucl. Acids Res., 10. Shaw, D.J., Rice, D.W. & Guest, J.R. (1983), Homology between CAP and Fnr, a regulator anaerobic respi- ration in Escherichia coli. J. MoL Biol., 166, 241-247. Sirko, A., Zehelein, E., Freundlich, M. & Sav,~,ers, G. (1993), Integration host factor is required for anaero- bic pyruvate induction ofpfl operon expression in Es- cherichia coli. J. Bacteriol., 175, 5769-5777. Spiro, S. & Guest, J.R. (1988), Inactivation of the protein of Escherichia coli by ~ g e t e d mutagenesis in the N-terminal region. Mol. Microbiol., 2, 701-707. Spiro, S. & Guest, J.R. (1990), ~ and its role hi oxygen- regulated gene expression in Escherichia coli. FEMS Microbiol. Rev., 75, 399-428. Spiro, S., Roberts, R.E. & Guest, J.R. (1989), FNR- dependent repression of the ndh gene of Escherichia coli and metal ion requirement for FNR-regulated gene expression. Mol. Microbiol., 3, 601-608. coli and Stachel, S.E. & Zambryski, P.C. (1986), virA and virG con- trol the plant-induced activation of the T-DNA trans- fer process of A. tumefaciens. Ceil, 46, 325-333. Stock, J.B., Ninfa, A.J. & Stock, A.M. (i989), Protein phosphorylation and regulation of adaptive respon.~es in bacteria. Microbiol. Rev., 53, 450-490. Stout, V. & Gottesman, S. (1990), RscB and RscC: a two 12th F O R U M I N M I C R O B I O L O G Y Unden, G. & Guest, J.R. (1985), Isolation and characteri- zation of the Fnr protein, the transcriptional regula- tor of anaerobic electron transport in Escherichia coll. Eur. J. Biochem., 146, 193-199. Unden, G. & Trageser, M. (1991), Oxygen-regulated gene expression in Escherichia coli: control of anaerobic respiration by the FNR protein. Antonie van Leeuwenhoec, 59, 65-76. Unden, G., Trageser, M. & Duchene, A. (1990), Effect of positive redox potentials (> + 400 mV) on the expres- sion of anaerobic respiratory enzymes in E. colio Mol. Microbiol., 4, 315-319. Weiner, J.H., Maclsaac, D.P., Bishop, R.E. & Bilous~ P.T. (1988), Purification and properties of Es- cherichia coli dimethyl sulfoxide reductase, an iron- sulfur molybdoenzyme with broad substrate specific- ity. J. BacterioL, 170, 1505-1511. Woods, S.A. & Guest, J.R. (1987), Differential role of the Escherichia coli fumarase and fnr-dependent expres- sion of furnarase B and aspartase. FEMS Microbiol. Lett., 48, 219-224. systems gene expression V. Stewart Development, Cornell University, Ithaca N Y 14853-8101 (USA) by thefdnGHI operon) and nitrate reductase (encod- ed by the narGHJI operon). Additionally, nitrate in- hibits the synthesis of enzymes that are involved in respiration of alternate electron acceptors: formate- dependent nitrite reductase (cytochrome c552, encod- ed by the nrfABCDEFG operon), a broad-specificity dimethyl sulphoxide/trimethylamine N-oxide reduc~ tase (encoded by the dmsABC operon) and fumarate reductase (encoded by the f r d A B C D operon). This regulatory pattern helps to ensure that the organism appropriately synthesizes respiratory enzyme systems in response to shifting availability of electron accep- tors. Physiological aspects of anaerobic respiration have been previously reviewed (Stewart, 1988). Nitrite, the reduction product of nitrate, is also an efficient electron acceptor, and nitrite regulates the synthesis of certain respiratory enzyme systems.

Bacterial Sensor Kinase/Response Regulator Systems PDF

Document Details

Tags

Related

Summary

Full Transcript

Upgrade to continue