Transcription of glpT of Escherichia coli K12 is regulated by anaerobiosis and fnr

FEMS MicrobiologyLetters 94 (19q2} 15-18 © 1992 Federation of European Microbiological Societies 0378-1097192/$05.00

Publishedby Elsevier

FEMSLE 04915

Transcription of glpT of Escherichia coli K12 is regulated by anaerobiosis and fnr K.K. Wong i and H.S. Kwan Department of Biology, The Chinese Universityof Hong Kong, Shatin. Hong Kong Received 12 February 1992 Revision received t~ March 1992 Accepted 24 Mar~,h Iq92

Key words: Anaerobiosis; Glycerol 3-phosphate permease; FNR; Escherichiacoli 1. SUMMARY The transcriptional expression of the snglycerol 3-phosphate transport gene, gipT, was studied with a glpT-lac fusion contained on a low-copy-number plasmid pFZY1. The fusion was isolated during a 'shot gun' cloning of promoters expressed under anaerobic growth conditions. It was shown that glpT was induced six-fold by anaerobiosis. Furthermore, aerobic expression of glpT was induced by the substrate glycerol 3phosphate, while the anaerobic expression of glpT was decreased by nitrate, glucose and a fnr mutation. However, nitrate repression on glpT expression was relieved if the upstream region from -290 of the transcription site was removed.

2. INTRODUCTION In Escherichia coli the glycerol regulon consists of five operons, regulated by a common Correspondence to: K.K. Wong. Present address: California Institute of Biological Research, ! 1099 N. Torrey Pines Rd., La Jolla, CA 92037, USA.

repressor GLPR and is involved in the metabolism of glycerol, sn-glyceroi 3-phosphate or glycerophosphodiesters [1]. The major upta:'e pathway for G-3-P in E. coli is via the sn-glycerol 3-phosphate transport protein, a cytoplasmic membrane protein encoded by the glpT gene [2]. glpT is adjacent to glpQ at 49 rain and is divergently transcribed from the glpABC operon which encodes sn-glyceroi 3-phosphate dehydrogenase [3]. Transcription of the glpABC operon is ind~aced by anaerobiosis and glycerol. A mutation in fnr, which encodes a positive regulator for many oxygen-regulated genes [4], prevents the full induction of glpABC [5,6]. The intergenic region between the start codons of the gipT and gipA genes consists of 272 bp [7]. CRP, cyclic adenosine Y,5'-monophosphate receptor protein, which is required to active transcription of many genes that encode enzymes for catabolism and energy metabolism [8], is also required for the expression of glpT [3]. However, the anaerobic regulation of gipT is unknown. Hence, we have examined the transcriptional regulation of glpT by constructing a glpT-lacZ fusion on the low. copy-number plasmid pFZY1, and measuring its expression in response to anaerobic growth.

3. MATERIAL AND METHODS

Bacterial strains and media Strain MC1061-5 [9], a derivative of E. coil K12, was used for transformation and preparation of promoter-probe plasmid pFZY1 [9] which is a low-copy-number plasmid with oriF and a promoterless iacZ gene. P1 transductions were performed by the method of Miller [10] for constructing isogenic strains fnrl / f n r + zci604 : : Tn 10. The constructed strains were checked for fnr phenotypes. Rich medium LB was used for the growth of bacteria. Minimal medium was as previously described [11]. For the screening of promoter clones, 2 ml of 2% X-gal (Sigma) was added per litre LB-amp 5cj medium. Reagents Restriction enzymes were obtained from New England Biolabs. Sequencing of ds plasmid or ss M13 subclones was performed by the dideoxy chain-termination method using a T7 polymerase sequencing Kit (Pharmacia). Construction of a promoter-clone library from Escherichia coIi KI2 Sau3A1 partially digested DNA fragments of about 2 kb were used for ligation with BamHl.di-

gcsted and dephosphorylated plasmid pFZYI. Electroporation was employed for transformation [12]. Transformants were selected on LB-X-galamp s° plates. The plates were incubated in an anaerobic jar at 37°C for 24 h. The plates were then left in 4°C for 34 h for blue color development.

fl-Galactosidase assay and growth conditions The assay procedure was similar to that described by Miller [10]. Growth conditions were as described previously [13]. The expression of glpT was studied with the glpT fusion plasmids pFE29 or pFE33 in strain MCl061-5, while the effect of fnr mutation was studied with isogenic strains (MC1061-5 fnrl /fnr + zci-606:: TnlO). Assays were carried out in triplicate with three independent transformants carrying the glpT fusion pinsmid.

4. RESULTS AND DISCUSSION In total, 92 anaerobically expressed promoter 'shot gun' clones from E. coli were isolated as described in MATERIALS AND METHODS. The strains were purified and then grown under aerobic and anaerobic conditions in LB-amp 5° broth

Table 1 Expression of Plasmid a

glpT from

plasmids pFE29 and pFE33 under various growth conditions

Medium b

/3-Galactosidase activity (U) ~ +0 2

-0 a

Ratio

(-0:,/+0:)

pFE29

LBE LBE + sn-glycerol 3-phosphate LBE+glucose LBE+nitrate LBE+thiosulfate LBE + fumarate MM + sn-glycerol 3-phosphate + fumarate

200+_ 50 1600_+ 120 400-r 40 220_+ 30 210+_ 80 420 +- 40 500 + 40

1100± 91) ! 700 + 160 230+_ 50 370+_ 40 1500+- 120 i 300 +_1l0 1 100_+ 80

5.5 i. 1 0.6 1.7 7.1 3.1 2.2

pFE33

LBE LBE + sn-glycerol 3-phosphate LBE + glucose LBE + nitrate

170_ 20 1700 +_180 380_+ 40 480 _+ 60

1200_+ 70 1200_+ 140 220 +_ 20 1400 + 210

7.1 0.7 0.6 2.9

a The host strain is MC1061-5, pFE33 was derived from pFE29 by deleting a 0.9-kb fragment at the 3' end of the DNA insert. b LBE is a buffered rich medium. The concentration of all supplements in the medium was 0.5%. sn-glycerol 3-phosphate, a-glycerol-3-phosphate. MM is a minimal salt medium [11]. /3-Galactosidase activity was expressed as nmol of o-nitrophenol produced per minute per ml per ODhsl~,

for 6 h prior to cell harvest and/3-galactosid,se assay. Among the 92 clones, two had a strih 'g anaerobic induction ratio (anaerobic activity/ aerobic activity). The two promoter clones pFE29 and pFEII7 had inserts of 1.6 kb and 2.9 kb, respectively. The DNA sequence of the inserts was determined and used for homology search in the GenBank and EMBL database with the FASTA-MAIL program in the 131ONET system. From the search, the lac proximal end of pFE29 corresponded to the position of 385 bp from the transcriptional initiation site of gipT, while that of pFEII7 corresponded to 829 bp from the transcriptional initiation site of glpT. Thus, both pFE29 and pFE117 contained the same transcriptional glpT-lac fusion. By the same approach, we have cloned and identified several different anaerobic inducible promoters from Salmonella typhimurium LT2 (unpublished results). Plasmid pFE29 (with 1.6-kb insert) containing the glpT-lac transcriptional fusion was used for the transcriptional studies./3-Galactosidase activities of E. coil strain MC1061-5 harboring pFE29 were assayed from cell culture grown in different media and under differing culture conditions (Table 1). The expression of glpT was induced about six-fold under anaerobic conditions. However, the aerobic expression of gipT could be increased to the same expression level under anaerobic conditions by addition of the substrate glycerol 3-phosphate. Thus, it was possible that anaerobic indue-

tion might be an indirect effect of the accumulation of glycerol 3-phosphate under anaerobic growth in a rich medium. We tested this possibility by assaying the aerobic and anaerobic expression of glpT with minimal glycerolphosphatefumarate medium, and anaerobic induction was still observable (Table 1). Nitrate is known to repress the expression of many cellular activities [14]. it was shown that glpT was also repressed by nitrate and glucose under anaerobic conditions. It was argued that the effect of nitrate may be indirect by allowing utilization of glycerol in the LB broth. However, in the presence of other electron aeceptors, the expression of glpT was not affected (Table 1). When a 0.9-kb fragment from the 5' end of the 1.6-kb insert was removed from pFE29, the resuiting 0.72-kb fragment in pFE33, the resulting glpT-lac fusion still contained the necessary elements for anaerobic and glycerol 3-phosphate induction and for glucose repression. However, the expression of gipT from pFE33 (Table 1) was no longer repressed by nitrate under anaerobic growth conditions. It is suggested that an upstream region of the glpT promoter might be necessary for the nitrate response. The E. coil gene fizr gene (also called nirA and nirR) [4] is involved in regulating the expression of several anaerobically induced respiratory enzymes. Since glpT was induced by anaerobiosis as demonstrated in this study, we tested the el'-

Table 2 Effect of a ~r mula!i*,qon ,he expressionof glpT-hw Medium~

LBE LBE+ sn-glycerol 3-phosphate LBE+ glucose LBE+ nitrate

/3-Galactosidaseactivity(U) b Fnr ~ +O, -O, -O2/~-O~ 90__. 5 5[}0_+lO 5.0

Fnr+O, 90+_21)

-O, 190_+20

-O2/+O 2 2.1

440± 40 40+30 80_+20

470+_10 50__. 5 50+10

!.20_+30 30_+15 1.60+20

0.3 0.6 3.2

350+_20 30+20 160__.20

0.8 0.8 2.0

" The host strainsare HKE501(MCI06I-5,fizr*, :ci-604::Tn10) and HKE502(MCl061-5.fitrl, zci-604::Tn 10).Mediumwas the same as Table 1. b/3-Oalactosidase activitywas expressedas nmolof o-nitrophenolproducedper minutepar ml per ODd50.

feet of fnr mutation on glpT expression with isogenic strains (MC1061-5, fnr-/fnr + zci604::Tni0). The anaerobic fl-galactosidase activity of glpT-lacZ fusion was decreased by a fnr mutation even in the presence of sn-glycerol 3phosphate (Table 2). In conclusion, we demonstrate that glpT is regulated by anaerobiosis, nitrate and glucose and that the anaerobic induction is mediated by the fnr gene product. The 272-bp intergenic region between the start codons of the glpT and glp,4 contains a putative FNR binding site (81-102 bp from the transcriptional initiation site) [15] and a CRP binding site (31-52 bp from the transcriptional initiation site) [16]. The FNR binding site shows poor homology with the consensus FNR binding site [4]; however, there was a reasonable half site (AAA---TGAT---). It will be of interest to determine whether these sites control the regulation of both glpT and and glpA. Furthermore, the proximity of the putative FNR and CRP sites may allow an interaction between the FNR and CRP proteins, similar to that suggested by CRP-AraC interaction in their respective multiple activator regulatory systems [17].

ACKNOWLEDGEMENTS We thank Dr. S. Bourgeois for the plasmid pFZY1, and Prof. E.C.C. Lin and Dr. M. McClelland for discussions and :,aluable suggestions. This work was partially supported by a research grant from UPGC of Hone Kong.

REFERENCES [I] Lin, E.C.C.(1976) Annu. Rev. Mierobiol.30, 535-578. [2] Larson, T.L, Schumacher, G. and Boos, W. (1982) J. Bacteriol. 152, 1008-1021. [3] Ehrmann, M., Boos, W., Ormseth, E., Schwizcr,H. and Larson, T.L (1987) J. Bacteriol. 169, 526-532. [4] Spiro, S. and Guest. J.R. 0990) FEMS Mierobiol. Rev. 75, 399-428. [5] luchi, S., Coil, S.T, and Lin, E.C.C.(1990) J. Bacteriol, 172, 279-184. [6] Kuritzkes, D.R., Zhang, X.-Y. and Lin, E.C.C. (1984) ,I. Bacteriol. 157, 591-598. [7] Cote, S.T., Eiglmeier,K., Ahmed, S., Honore, N., EImes, L., Anderson, W.F. and Weiner, J.H. (1988) J, Bacteriol. 170, 2448-2456. [8] DeCrombrugghe,B., Busby, S. and Buc, H. (1984) Science 224, 831-838. [9] Koop, A.H., Hartley, M.E. and Bourgeois, S. (1987) Gene 52, 245-256. [10] Miller, J.H. (1972) Experiments in Molecular Genetics. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY. Ill] Neidhardt, F.C., Bloch, P.L. and Smith, D.F. (197~.~ J. Bacteriol. 119, 736-747. [12] Dower, W.J., Miller, J.F. and Ragsdale, C.W. (1988) NucleicAcids Res, 16, 6127-6145. [13] Wong, K.K., Suen, K.L. and Kwan, H.S. (1989) J. Bacteriol. 171,4900-4905. [14] Stewart,V. (1988) Microbiol. Rev. 52, 190-232. [15] Eiglmeier, K., Honore, N., luchi, S., Lin, E.C.C. and Cole, S.T. (1989) Mol. Microbiol. 3, 869-878, [16] Eigimeier,K., Boos, W. and Cole, S.T. (1987) Mol. Microbiol. 1,251-258. [17] Collado-Vides,J., Majasanik, B. and Gralla, J.D. 0991) Microbiol. Rev. 55, 371-394.