Regulation of fumarase (fumB) gene expression in Escherichia coli in response to oxygen, iron and heme availability:
role of the arcA, fur, and hemA gene products
Ching-Ping Tseng *
Institute of Biological Science and Technology, National Chiao Tung University, Hsinchu, Taiwan, ROC Received 30June 1997; revised 15 September 1997; accepted 17 September 1997
Abstract
Three distinct fumarases, FumA, FumB and FumC, have been reported in Escherichia coli. While the fumA and fumC gene products are expressed under aerobic cell growth conditions, the FumB fumarase appears to be more abundant during anaerobic growth. To study the transcriptional regulation of the fumB gene, a fumB-lacZ operon fusion was constructed and analyzed in a single copy under a variety of cell culture conditions. Expression of fumB-lacZ was fourfold higher under anaerobic than aerobic growth conditions. This anaerobic response is modulated by the ArcA and Fnr proteins, which function independently as anaerobic activators of fumB gene expression. Cellular iron limitation in a fur mutant caused fumB-lacZ expression to decrease sevenfold while cellular heme limitation decreased fumB gene expression twofold. In addition, fumB- lacZ expression was shown to vary depending on the DNA superhelicity. This study further delineates the regulation of the fumB gene in cell growth.
Keywords: Fumarase gene expression; fumB; Aerobic vs. anaerobic growth; Escherichia coli
1. Introduction
Escherichia coli contains three fumarase genes, fumA, fumB, and fumC, that catalyze the intercon- version of fumarate and L-malate [1,2]. The fumA and fumC genes are located at 35.5 min of the E.
coli linkage map [3] while the fumB gene is located at 93.5 min [1]. The physiological function of each E.
coli fumarase has been studied by using a triple mu- tant transformed with a plasmid containing one of the three fumarase genes. The FumA enzyme ap-
peared to be a component of the tricarboxylic acid (TCA) cycle since it was synthesized predominantly under aerobic conditions [4]. Expression of the fumA gene was lowest during anaerobic cell growth sug- gesting a role for FumA as an aerobic fumarase.
Anaerobic expression of fumA-lacZ from the fumA promoter was derepressed in both an arcA and an fnr mutant indicating that both ArcA and Fnr func- tion as anaerobic repressors [5]. The FumB enzyme was shown to be more abundant under anaerobic conditions, especially during anaerobic respiratory growth with glycerol plus fumarate [4]. Furthermore, the anaerobic expression of the fumB gene is reduced in an fnr mutant suggesting that Fnr is a transcrip-
0378-1097 / 97 / $17.00 ß 1997 Federation of European Microbiological Societies. Published by Elsevier Science B.V.
PII S 0 3 7 8 - 1 0 9 7 ( 9 7 ) 0 0 4 5 5 - 2
* Tel.: +886 (35) 729289; Fax: +886 (35) 729288;
E-mail: [email protected]
FEMS Microbiology Letters 157 (1997) 67^72
tional activator [4]. Transcription of the fumC gene was shown to be complex: it was dependent on both the fumA and fumC promoters [5]. It has been re- ported that synthesis of FumCwas increased by the addition of oxidizing agents, and this increase was assumed to be dependent on the soxRS gene prod- ucts [6]. Recently, it has been proved that both the superoxide control and the iron control of fumC ex- pression require the SoxR regulatory protein. FumC was thus proposed to substitute for FumA when environmental iron is limiting or when superoxide radicals accumulate [5]. In this study, we further ex- amined the regulation of fumB gene using a fumB- lacZ operon fusion.
2. Materials and methods
2.1. Bacterial strains, bacteriophages, and plasmids The genotypes of the E. coli K-12 strains, plasmids and the bacteriophage are listed in Table 1. The arcA, fur, topA and ¢s strains were constructed by introducing the indicated mutation into strain MC4100 VCPT7 (fumB-lacZ) by P1 transduction fol-
lowed by selection for resistance to the appropriate drug [7]. The PC2 (fnr) VCPT7 lysogen and PC10 (hemA) VCPT7 lysogen were constructed by infecting PC2 and PC10 with a high titer VCPT7 lysate as previously described [8].
2.2. Construction of fumB-lacZ operon fusion To construct the fumB-lacZ fusion, the 3.16-kb HindIII-HpaI fragment of fumB gene was cloned from the Kohara library into pTZ19 to give plasmid pCPT4. The 1.49-kb BegII-PstI fragment was then inserted into M13 mp19 to give M13CTP5 [9]. By using oligonucleotide-directed mutagenesis, a new EcoRI site was introduced into the fumB gene at position +86 relative to the start of fumB translation to give M13CPT6. A 896-bp BglII-EcoRI fragment containing the 5P end of the fumB gene and the up- stream 810 bp was isolated from M13CPT6 and in- serted into the BamHI-EcoRI sites of pRS1247 to generate the fumB-lacZ operon fusion designated pCPT7. The junction between the fumB promoter regions and the lacZ gene was con¢rmed by dou- ble-strand DNA sequencing analysis [10]. The fusion was transferred to RZ5 to generate the correspond-
Table 1
Escherichia coli strains, phages and plasmids
Strain, phage, plasmid Derivation Genotype or phenotype Reference or source
Strains
MC4100 F3araD139 v(argF3lac)U169rpsL 150
relA1 £b-5301 deoC1 ptsF25 rbsR [11]
PC2 MC4100 fnr [13]
C35 MC4100 arcA Kanr [13]
PC40 MC4100 hemA41 Kanr [5]
SJP 2 MC4100 fur: :Tn5 [5]
SJP 4 MC4100 ¢s-767Kanr [5]
SJP 6 PC2 fnr arcA Kanr [5]
RS 2 MC4100 pyrF287topAgal125rps V3 [16]
Phages
VRZ5 [11]
VCPT7 pCPT7 x(fumB-lacZ)lacYlacA This study
M13mp19 [9]
M13CPT5 M13mp but with 1.48-kb
BglII-PstI fragment This study
M13CPT6 M13 CPT5 but with an
EcoRI site This study
Plasmids
pRS1247 lacZ lacYlacA [11]
pCPT4 pTZ19 fumB This study
pCPT7 pRS1247 x(fumB-lacZ)lacYlacA This study
ing VCPT7, which was then introduced into the MC4100 chromosome as previously described [11].
Single lysogen was identi¢ed and puri¢ed for subse- quent study. The wild-type locus of fumB was stably integrated at the lambda attachment site on the chro- mosome.
2.3. Cell growth
For strain manipulation and maintenance, cells were grown in Luria broth or on solid medium.
When required, ampicillin and kanamycin were added to the medium at concentrations of 100 and 50 Wg ml31, respectively. For L-galactosidase assay, cells were grown in glucose (40 mM) minimal me- dium (pH 7.0) [12], unless otherwise indicated [13].
Aerobic and anaerobic growth were performed as previously described [13]. Flasks or tubes containing medium were inoculated from the overnight cultures grown under the same conditions, and the cells were allowed to double four of ¢ve times while in mid- exponential phase prior to harvest for analysis (opti- cal density at 600 nm of 0.4^0.5; Kontron Uvikon 810 spectrophotometer). Trimethylamine-N-oxide (TMAO), sodium nitrate, or fumarate was added at an initial concentration of 40 mM [12]. Where indicated, N-aminolevulinic acid (N-ALA), 2,2-dipyr- idyl or ferrous sulfate was added at ¢nal concentra- tions of 24 Wg ml31 or 150 or 80 WM, respectively.
2.4. L-Galactosidase assay
L-Galactosidase assays were performed as previ- ously described [12]. One unit of L-galactosidase is de¢ned as the hydrolysis of 1 nmol o-nitrophenyl-L-
D-galactopyranoside (ONPG) per min per mg pro- tein. All values shown represent the average of at least three determinations and did not vary more than 10% from the mean.
3. Results
3.1. E¡ect of oxygen and other electron acceptors on fumB-lacZexpression
To examine the e¡ect of anaerobic respiratory conditions on fumB expression, an E. coli wild-type strain containing a fumB-lacZ fusion was grown in the presence and absence of the alternative electron acceptors oxygen, nitrate, TMAO, or fumarate (Ta- ble 2). Consistent with the earlier observations of Woods and Guest [4], fumB expression was lowest during aerobic cell growth and elevated fourfold when cells were grown anaerobically under fermen- tative conditions (minimal glucose medium). When cells were grown anaerobically with nitrate present, fumB-lacZ expression was twofold lower than when no electron acceptors were added. Interestingly, fumB expression was highest when fumarate and TMAO were present as electron acceptors during anaerobic growth in a glucose minimal medium.
When glycerol was substituted for glucose as the carbon source, conditions in which the cells derive energy from electron transport-linked phosphoryla- tion reactions, fumB-lacZ expression was lowest in the presence of oxygen. It was elevated four- to sev- en-fold when nitrate, TMAO or fumarate was added in glycerol minimal medium (Table 2). The results of these fumB-lacZ studies suggest that FumB is oper-
Table 2
E¡ect of alternative electron acceptors on fumB-lacZ expression
Electron acceptor addeda L-Galactosidase activity (U)
Minimal glucose Minimal glycerol
None 120 NGb
Oxygen 30 32
Nitrate 74 140
TMAO 150 190
Fumarate 140 230
aCells were grown in 40 mM glucose or glycerol minimal medium either aerobically or anaerobically as described in the text. Sodium nitrate, TMAO, or fumarate was added at an initial concentration of 40 mM.
bNG, no growth.
ative during conditions of anaerobic respiration since fumB expression is elevated during these conditions.
3.2. E¡ect of the arcA, fnr, topA and ¢s gene products on fumB-lacZ expression
The aerobic/anaerobic control of the fumA, fumB and fumC genes is mediated by the fnr gene product [4,5]. Expression of the fumA and fumC genes is also negatively regulated by the ArcA protein [5]. To ex- amine the individual and combined e¡ects of muta- tions in these two genes, the VCPT7 fusion was in- troduced into isogenic fnr, arcA, and fnr arcA strains. Single lysogens were identi¢ed and puri¢ed for subsequent study. A fourfold di¡erence in fumB- lacZ expression occurred in response to oxygen availability in the wild-type parent strain (Table 3).
Fnr acts as a positive regulator of fumB-lacZ expres- sion during anaerobic conditions as previously pro- posed [4]. Expression of fumB-lacZ was decreased by 50% in an fnr mutant. The results showed that ArcA was also a positive regulator of fumB gene expres- sion. The level of fumB-lacZ expression in the arcA deletion strain decreased 30% under anaerobic growth conditions relative to the wild-type strain, whereas aerobic expression was una¡ected. In addi- tion, fumB-lacZ expression decreased fourfold in the fnr and arcA double mutant strain anaerobically.
Thus, the Fnr and ArcA proteins appear to function independently of each other to regulate fumB gene expression.
There is increasing evidence that DNA supercoil-
ing varies in response to environmental signals such as osmolarity or anaerobic growth [14,15]. In a topA mutant strain, expression of fumB-lacZ decreased threefold under aerobic conditions and lowered 30% under anaerobic conditions (Table 3). When the topA allele was introduced into arcA and fnr arcA double mutant strains [16], fumB-lacZ expres- sion was lowered about fourfold and ¢vefold, respec- tively. These results indicate that a change of DNA superhelicity negatively regulates fumB gene expres- sion. The Fis protein is known to be involved in regulation of a variety of genes in E. coli [17]. Under the conditions tested, a ¢s mutation activated fumB gene expression about 30% during anaerobic condi- tions relative to the parent strain (Table 3).
3.3. E¡ect of iron and heme availabilityon fumB-lacZ expression
Iron is an essential component of the FumA and FumB fumarase activities [18]. Because cellular iron limitation stimulates fumC gene expression [5], the e¡ect of if iron limitation on fumB gene expression was tested. When cells were grown in the presence of the iron chelator 2,2-dipyridyl to limit iron, fumB- lacZ expression was decreased sevenfold during anaerobic growth (Table 4). When iron was added in excess, fumB-lacZ returned to the level seen when no 2,2-dipyridyl was present. Under aerobic condi- tions, expression of fumB-lacZ was elevated twofold when iron was added in excess to the medium. To test whether a fur mutant which is defective for iron
Table 3
E¡ect of arcA, fnr, topA, himA, and ¢s mutations on fumB-lacZ gene expressiona
Genotype L-Galactosidase activity (U)
Aerobic Anaerobic
Wild-type 30 120
fnr 30 56
arcA 28 85
fnr arcA 18 35
topA 12 94
fnr topA 12 46
arcA topA 12 31
fnr arcA topA 14 26
¢s 45 160
aCells were grown in glucose minimal medium under aerobic or anaerobic conditions as described in the text.
Table 4
E¡ect of iron availability on fumB-lacZ expression Additiona L-Galactosidase activity (U)
O2 Dip Fe2 Wild-type fur
+ 3 3 30 20
+ + 3 15 14
+ 3 + 57 22
+ + + 50 15
3 3 3 120 28
3 + 3 16 15
3 3 + 134 30
3 + + 125 32
aCells were grown in a minimal glucose medium aerobically or anaerobically as described in the text. Dipyridyl (Dip) and ferrous sulfate (Fe2) were added at initial concentrations of 150 WM and 80 WM, respectively, as indicated.
regulation and uptake a¡ects fumB-lacZ expression, a fur allele was introduced into the fumB-lacZ fusion strain [19]. Anaerobic levels of L-galactosidase were reduced to the level seen under aerobic conditions (Table 4). Addition of excess iron to the medium did not restore fumB-lacZ expression. These results indicate that Fur is involved in the observed iron regulation.
Heme is a cofactor of several enzymes needed for energy generation during anaerobic growth [20]. We examined the e¡ects of heme availability on fumB- lacZ expression since it a¡ects fumA and fumC gene expression [5]. In the hemA mutant strain which is defective for heme biosynthesis, fumB-lacZ expres- sion was reduced twofold during anaerobic growth (Table 5). Expression was partially restored to wild- type levels by providing the cells exogenously with N- ALA, the end product of the reaction catalyzed by the hemA gene product. Whether this regulatory ef- fect is direct or indirect is unclear.
4. Discussion
Fumarase participates in the TCA cycle during aerobic growth, and in the reductive pathway from oxaloacetate to succinate during anaerobic growth.
It was previously shown that the fumA and fumC genes are most highly expressed during aerobic con- ditions [5], whereas the fumB gene was expressed at a higher level under anaerobic cell growth conditions [4]. As the other TCA cycle genes are regulated by ArcA or both ArcA and Fnr repressors [5,19,21], ArcA and Fnr also function as negative regulators of the fumA and fumC genes while Fnr is an anae- robic activator of the fumB gene [4]. In this study, we
further demonstrate that this anaerobic regulation of fumB gene expression is mediated by both ArcA and Fnr proteins that function independently as anaero- bic activators of fumB-lacZ expression. It is similar to their e¡ects on the respiratory genes cydAB, frdABCD, dmsABC and narGHJK [22]. The fact that anaerobic cell growth in a glycerol medium with the electron acceptor fumarate present results in a relatively high level of fumB gene expression suggested that FumB contributes to the fermentative and non-cyclic TCA pathway [4].
Iron is an essential component of both the FumA and FumB fumarases, while FumC does not require iron for its activity [23]. However, iron limitation only a¡ects the fumC gene but has no e¡ects on fumA gene expression [5]. In contrast to the obser- vation that fumC gene expression was activated by 2,2-dipyridyl, fumB-lacZ expression decreased when 2,2-dipyridyl was added. This repression can parti- ally be restored by adding excess iron (Table 4).
Since iron is required for Fnr activity [24], the de- crease of fumB-lacZ expression by the chelator may be due to lowered Fnr activity. FumA and FumB are members of the iron-dependent hydrolases [3,23].
Earlier studies suggested that the fumA gene is acti- vated by Fur whereas the fumB and fumC genes were not signi¢cantly a¡ected by Fur under aerobic con- ditions [25]. In this study, we found that the expres- sion of the fumB gene was abolished in the fur dele- tion strain under anaerobic conditions where fumB is more highly expressed. Thus, iron control of fumB expression was fur-dependent. Although the fumA and fumB genes are homologous and their encoded products are structurally similar, FumA has a higher a¤nity for fumarate than for L-malate whereas FumB exhibits the reverse pattern of a¤nities for these two compounds [23]. Therefore, a hierarchical control of fumA and fumB genes by ArcA and Fnr during anaerobic growth suggested that the synthesis of the FumB fumarase under anaerobic conditions gives E. coli the £exibility to meet its requirement for diverse environmental conditions.
Acknowledgments
We thank Dr. Robert P. Gunsalus for providing strains for this study. This work was supported by
Table 5
E¡ect of heme availability on fumB-lacZ gene expression Additiona L-Galactosidase activity (U)
Oxygen N-ALA Wild-type hemA
+ 3 30 22
+ + 28 25
3 3 120 63
3 + 110 92
aCells were grown in a minimal glucose medium aerobically or anaerobically as described in the text. N-ALA was added at an initial concentration of 24 Wg ml31.
Grant NSC84-2321-B-009-001 from the National Science Council of the Republic of China.
References
[1] Guest, J.R., Miles, J.S., Robert, R.E. and Woods, S.A. (1985) The fumarase genes of Escherichia coli: location of the fumB gene and discovery of a new gene (fumC). J. Gen. Microbiol.
131, 2971^2984.
[2] Guest, J.R. and Roberts, R.E. (1983) Cloning and mapping and expression of the fumarase gene of Escherichia coli K-12.
J. Bacteriol. 153, 588^596.
[3] Miles, J.S. and Guest, J.R. (1984) Complete nucleotide se- quence of the fumarase gene fumA of Escherichia coli. Nucleic AcidsRes. 12, 3631^3642.
[4] Woods, S.A. and Guest, J.R. (1988) Di¡erential roles of the Escherichia coli fumarases and fnr-dependent expression of fumarase B and aspartase. FEMS Microbiol. Lett. 48, 219^
224.
[5] Park, S.-J. and Gunsalus, R.P. (1995) Oxygen, iron, carbon, and superoxide control of the fumarase fumA and fumC genes of Escherichia coli: role of the arcA, fnr, and soxR gene prod- ucts. J. Bacteriol. 177, 6255^6262.
[6] Liochevand, S.I. and Fridovich, I. (1992) Fumarase C, the stable fumarase of Escherichia coli, iscontrolled by the soxRS regulon. Proc. Natl. Acad. Sci. USA 89, 5892^5896.
[7] Miller, J.H. (1972) Experimentsin Molecular Genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.
[8] Silhavy, T., Besman, M. and Enquist, L. (1984) Experiments with Gene Fusions. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY.
[9] Messing, J. and Vieira, J. (1982) A new pair of M13 vectors for selecting either DNA strand of double digest restriction fragments. Gene 19, 269^276.
[10] Sanger, F., Nicklen, B.J. and Coulson, A.R. (1977) DNA sequencing with chain-terminating inhibitors. Proc. Natl.
Acad. Sci. USA 74, 5463^5467.
[11] Simons, R.W., Houman, R. and Kleckner, N. (1987) Im- proved single and multicopy lac-based cloning vectors for protein and operon fusions. Gene 53, 85^96.
[12] Cotter, P.A. and Gunsalus, R.P. (1989) Oxygen, nitrate, and molybdate regulation of dmsABC gene expression in Esche- richia coli. J. Bacteriol. 171, 3817^3823.
[13] Cotter, P.A. and Gunsalus, R.P. (1992) Contribution of the fnr and arcA gene productsin coordinate regulation of the
cytochrome o and d oxidase (cyoABCDE and cydAB) genes in Escherichia coli. FEMS Microbiol. Lett. 91, 31^36.
[14] Dorman, C.J., Barr, G.C., Bhriain, N.N. and Higgins, C.F.
(1988) DNA supercoiling and the anaerobic and growth phase regulation of tonB gene expression. J. Bacteriol. 170, 2816^
2826.
[15] Higgins, C.F., Dorman, C.J., Waddell, D.A., Booth, I.R., May, G. and Bremer, E. (1988) A physiological role for DNA supercoiling in the osmotic regulation of gene expres- sion in S. typhimurium and E. coli. Cell 52, 569^584.
[16] Miller, W.G. and Simons, R.W. (1993) Chromosomal super- coiling in Escherichia coli. Mol. Microbiol. 10, 675^684.
[17] Tsui, P., Haung, C. and Freudlich, M. (1991) Integration host factor binds speci¢cally to multiple sites in the ompB pro- motersof Escherichia coli and inhibitstranscription. J. Bac- teriol. 173, 5800^5807.
[18] Flint, D.H., Emptage, M.H. and Guest, J.R. (1992) Fumarase A from Escherichia coli: puri¢cation and characterization as an iron-sulfur cluster containing enzyme. Biochemistry 31, 10331^10337.
[19] Park, S.-J., Tseng, C.P. and Gunsalus, R.P. (1995) Regulation of the Escherichia coli succinate dehydrogenase (sdhCDAB) operon in response to anaerobiosis and medium richness:
role of the ArcA and Fnr proteins. Mol. Microbiol. 15, 473^482.
[20] Darie, S. and Gunsalus, R.P. (1994) E¡ect of heme and oxy- gen availability on hemA gene expression in Escherichia coli:
role of the fnr, arcA, and himA gene products. J. Bacteriol.
176, 5270^5276.
[21] Park, S.-J., Cotter, P.A. and Gunsalus, R.P. (1995) Regula- tion of malate dehydrogenase (mdh) gene expression in Esche- richia coli in response to oxygen, carbon, and heme availabil- ity. J. Bacteriol. 177, 6652^6656.
[22] Gunsalus, R.P. (1992) Control of electron £ow in Escherichia coli: coordinated transcription of respiratory pathway genes.
J. Bacteriol. 174, 7069^7074.
[23] Woods, S.A., Schwartzbach, S.D. and Guest, J.R. (1988) Two biochemically distinct classes of fumarases in Escherichia coli.
Biochim. Biophys. Acta 954, 14^26.
[24] Engel, P., Trageser, M. and Unden, G. (1991) Reversible in- terconversion of the functional state of the gene regulator FNR from Escherichia coli in vivo by O2and iron availability.
Arch. Microbiol. 156, 463^470.
[25] Gruer, M.J. and Guest, J.R. (1994) Two genetically distinct and di¡erentially regulated aconitases (AcnA and AcnB) in Escherichia coli. Microbiology 140, 2531^2541.