Download FnrP interactions with the Pasteurella haemolytica leukotoxin promoter

FEMS Microbiology Letters 186 (2000) 73^77
www.fems-microbiology.org
FnrP interactions with the Pasteurella haemolytica
leukotoxin promoter
Gaylen A. Uhlich
a
a;1
, Peter J. McNamara b , John J. Iandolo c , Derek A. Mosier
a;
*
Department of Diagnostic Medicine/Pathobiology, College of Veterinary Medicine, Kansas State University, 1600 Denison Ave., Manhattan, KS 66506, USA
b
Department of Medical Microbiology and Immunology, University of Wisconsin School of Medicine, 404 Service Memorial Institute, Madison,
WI 53706, USA
c
Department of Microbiology and Immunology, University of Oklahoma Health Science Center, P.O. Box 26901, Oklahoma City, OK 73190, USA
Abstract
Pasteurella haemolytica FnrP is homologous to Fnr, the global transcriptional regulator of anaerobic respiration in Escherichia coli. To
investigate the role of O2 in the expression of P. haemolytica leukotoxin, we tested a lktC: :lacZ fusion constructed in E. coli for a FnrPmediated regulatory effect under aerobic and anaerobic growth conditions. Both E. coli Fnr and FnrP suppressed leukotoxin transcription
under aerobic conditions. Under anaerobic conditions, Fnr suppressed transcription, while FnrP increased transcription. These results were
confirmed using FnrP*, a mutant form of FnrP that activates anaerobically inducible genes under aerobic conditions. In mobility shift
assays, partially purified FnrP bound to a potential regulatory site in a P. haemolytica lktC promoter fragment. ß 2000 Published by
Elsevier Science B.V. All rights reserved.
Keywords : FnrP; Leukotoxin ; Fnr
1. Introduction
Pasteurella haemolytica A1 is the agent most commonly
isolated from severe bronchopneumonia associated with
the bovine respiratory disease complex [1]. Infection with
P. haemolytica results in necrosis and the production of a
¢brinocellular exudate within the lung [2]. Within such an
environment, alterations in O2 concentrations could in£uence bacterial growth and gene expression.
Leukotoxin is an important P. haemolytica virulence
factor as demonstrated by the reduced virulence of a leukotoxin-de¢cient mutant [3]. Leukotoxin is encoded in an
operon of four genes ordered lktCABD [4]. The second
gene, lktA, encodes an inactive form of leukotoxin that
is activated by the product of the ¢rst gene, LktC. The
remaining two genes, lktB and lktD, appear to be required
for transmembrane secretion of LktA [4].
Transcription of the leukotoxin operon is initiated from
* Corresponding author. Tel. : +1 (785) 532-4410;
Fax: +1 (785) 532-4039; E-mail: [email protected]
1
Present address: Roman L. Hruska U.S. Meat Animal Research Center, ARS, USDA, P.O. Box 166, Clay Center, NE 68933, USA.
two promoters (P1 and P2 ), located 3290 and 330 bp
from the translational start of LktC [5]. Immediately upstream of the P1 promoter is lapT, a divergently transcribed gene that encodes the periplasmic binding protein
component of the leukotoxin-associated arginine permease
gene cluster [6]. A model for co-regulation of the leukotoxin and arginine permease gene complex operons includes both static and protein-induced DNA bends, and
one or more regulatory protein(s) that bind(s) within the
region between lktC and lapT [7]. One regulatory protein
that can induce conformational changes in the DNA of
this region is integration host factor (IHF) [7,8].
P. haemolytica A1 fnrP is a homolog of Escherichia coli
fnr, a global transcriptional regulator that controls enzymes involved in the transition from aerobic to anaerobic
respiration [9]. Like fnr, fnrP trans-complements the anaerobic de¢ciencies of E. coli lacking fnr and increases the
transcription of the gene encoding fumarate reductase
(frdA) under anaerobic conditions [9]. During anaerobic
growth, Fnr can either activate or repress transcription
of genes by binding DNA at speci¢c sites or half-sites
with the consensus sequence TTGATö öATCAA
[10,11].
In this study, we investigated oxygen-dependent altera-
0378-1097 / 00 / $20.00 ß 2000 Published by Elsevier Science B.V. All rights reserved.
PII: S 0 3 7 8 - 1 0 9 7 ( 0 0 ) 0 0 1 1 9 - 1
FEMSLE 9354 12-4-00
Downloaded from http://femsle.oxfordjournals.org/ by guest on October 20, 2016
Received 3 August 1999 ; received in revised form 3 March 2000; accepted 6 March 2000
74
G.A. Uhlich et al. / FEMS Microbiology Letters 186 (2000) 73^77
tions in leukotoxin transcription mediated by FnrP. To
test for a FnrP-dependent regulatory e¡ect, we measured
L-galactosidase activity under aerobic and anaerobic conditions from a lktC: :lacZ promoter fusion in a fnr-de¢cient E. coli that expresses P. haemolytica FnrP. We also
determined the ability of P. haemolytica FnrP to bind
within the promoter region of leukotoxin.
testine alkaline phosphatase were used as recommended by
the manufacturers (Promega, Madison, WI, USA). Plasmids were introduced into E. coli and analyzed using
standard techniques [14]. Procedures for bacteriophage
P1 and M13 propagation and transduction have been previously described [15,14]. PCR was performed using described conditions [14].
2.3. Site-directed mutagenesis of fnrP
2. Materials and methods
2.1. Bacterial strains, growth conditions, protein
puri¢cation and antibodies
2.2. Recombinant techniques and polymerase chain reaction
(PCR)
Restriction endonucleases, T4 DNA ligase, and calf in-
2.4. L-Galactosidase assays
Bacterial cultures were grown at 37³C to an optical
Table 1
Bacterial strains and plasmids
Bacterium or plasmid
E. coli:
GAU370
GAU460
JRG1728
JRG1787
SH370
SH460
Plasmids
pAAUBlue
pAAU6
pGT3
pGT3L
pG4X4
pHSS6
pMJU6
Genotype/phenotype
MC4100
MC4100
MC1000
MC1000
MC4100
MC4100
Source or reference
v(gal-GV )b2V : :[P(lktC-lacZ‡ ) 1.4-kb BclI^BglII lacY‡ ] v(tyrR, fnr, rac, trg)
v(gal-GV )b2V : :[P(lapT-lacZ‡ ) 695-bp HindII^EcoRV lacY‡ ] v(tyrR, fnr, rac, trg)
v(lac)X74 galU galK rpsL v(ara leu) v(tyrR, fnr, rac, trg)
v(lac)X74 galU galK rpsL v(ara leu) v(tyrR, fnr, rac, trg) VP(frdA: :-lacZ)
v(gal-GV )b2V : :[P(lktC-lacZ‡ ) 1.4-kb BclI^BglII lacY‡ ]
v(gal-GV )b2V : :[P(lapT-lacZ‡ ) 695-bp HindII^EcoRV lacY‡ ]
pT7Blue: :1.8-kb EcoRI^HindIII fragment of pGT3, Apr
pHSS6: :1.8-kb EcoRI^HindIII fragment of pGT3, Knr
pBluescript II KS(+): :1852-bp Sau3A^XhoI fragment carrying P. haemolytica fnrP, Apr
pACYC184 : :2254-bp PvuII fragment of pGT3
pBluescript II KS(+): :1087 bp, Fnr‡ , 3P deletion clone of pGT3
ColE1 replicon, Knr
pHSS6: :1.8-kb, FnrP*‡ , fragment of pGT3, Knr
FEMSLE 9354 12-4-00
This study
This study
[19]
[19]
[7]
[7]
This study
This study
[9]
[9]
[9]
[20]
This study
Downloaded from http://femsle.oxfordjournals.org/ by guest on October 20, 2016
Bacterial strains and plasmids are listed in Table 1. The
fnr deletion of E. coli strain JRG1728 was transferred to
E. coli strains SH370 and SH460 by co-transduction with
chloramphenicol resistance using bacteriophage P1vir and
resulted in strains GAU370 and GAU460, respectively.
These fusion constructs contain the P. haemolytica lapTlktC intergenic sequence with either lktC (GAU370) or
lapT (GAU460) fused to the gene encoding L-galactosidase [7].
E. coli was cultured in Luria^Bertani (LB) broth or on
LB agar plates supplemented with 20 Wg ml31 chloramphenicol, 10 Wg ml31 tetracycline, 25 Wg ml31 kanamycin
or 100 Wg ml31 ampicillin. Anaerobic cultures were grown
at 37³C in stationary, screw-topped, polypropylene tubes
overlaid with sterile mineral oil [12].
FnrP was expressed and puri¢ed as previously described
[9]. IHF polyclonal antibodies and IHF protein, overexpressed and puri¢ed from E. coli strain K5746, were provided by Dr. Steven Goodman [13].
FnrP* was produced by site-directed mutagenesis [16].
Substitution of a histidine residue for a leucine residue at
position 28 in E. coli Fnr increased expression of the E.
coli nar operon in the presence of oxygen [17]. The corresponding mutation was created in FnrP by altering codon
27 from CTA to CAT in fnrP. A phosphorylated oligonucleotide with the sequence 5P-GCAGTATTAGCCAACATTGCTTACC that corresponds to bases 65^89 of
fnrP (GenBank accession No. AFO33119) was synthesized
(Research Genetics, Huntsville, AL, USA). Singlestranded template was obtained from E. coli strain
CJ236 transfected with M13mp19 carrying the 1.8-kb
KpnI^XbaI insert from plasmid pGT3. The mutagenic oligonucleotide was annealed to the template DNA at a 3:1
molar ratio, extended in vitro using T4 DNA polymerase
and ligase, and the resulting product was used to transfect
E. coli XL1-Blue (Stratagene, Cedar Creek, TX, USA).
The expected mutation in fnrP was identi¢ed by DNA
sequence analysis using the primer 5P-CTTTCACTTAAGGTATAGGC. An XbaI/KpnI fragment containing
the altered fnrP was transferred from M13mp19 into similar sites in plasmid pT7Blue (Novagen, Madison, WI,
USA). Transfer of the pT7Blue insert to plasmid pHSS6
using HindIII and EcoRI created pMJU6.
G.A. Uhlich et al. / FEMS Microbiology Letters 186 (2000) 73^77
75
density at 600 nm of 0.3^0.6, permeabilized using chloroform and sodium dodecyl sulfate, and tested for L-galactosidase activity [15]. Speci¢c activity (SA) was calculated
for each strain from the results of three independent trials
with three replicates per trial. Data were analyzed by oneway analysis of variance and Fisher's least signi¢cant difference using SAS (SAS Institute, Cary, NC, USA).
2.5. Mobility shift assays
3. Results
3.1. The e¡ect of Fnr and FnrP on a lktC: :lacZ promoter
fusion
Under aerobic conditions, a comparison of three strains
of E. coli containing a lktC: :lacZ chromosomal fusion
and di¡ering in their ability to produce Fnr showed that
L-galactosidase activity was reduced in strains with fnrP or
fnr when compared to a fnr-null mutant (Fig. 1). Compared to strain GAU371 (fnr3), chromosomally encoded
fnr (GAU373) resulted in an approximately 3.5-fold reduction (P 6 0.0001) in L-galactosidase activity. An approximately 2.5-fold decrease in L-galactosidase activity
(P 6 0.0001) occurred with a plasmid-encoded copy of
fnrP (GAU372) compared to GAU371. Anaerobically, signi¢cantly less L-galactosidase activity was present in
GAU373 (fnr+) than in GAU371 (P 6 0.0001). However,
in GAU372, fnrP was associated with a signi¢cant increase in L-galactosidase activity (P 6 0.0001) when compared to GAU371.
Fig. 1. The e¡ect of E. coli Fnr and P. haemolytica FnrP on leukotoxin
transcription under aerobic (white) and anaerobic (black) growth conditions. The production of L-galactosidase, expressed as SA, from
lktC: :lacZ chromosomal fusions in E. coli strains SH370 (Fnr) with
pACYC184, GAU370 (Fnr3) with pACYC184 and GAU370 with
pGT3L (FnrP) is shown.
3.2. The e¡ect of FnrP* on frdA, lktC, and lapT promoter
fusions
To facilitate comparison of active and inactive forms of
FnrP, a mutant FnrP that activates target genes in the
presence of oxygen was created (FnrP*). The activity of
FnrP* was tested in JRG1787, a fnr-de¢cient strain of E.
coli that contains a frdA: :lacZ chromosomal fusion, as
described [9]. Under aerobic conditions, L-galactosidase
activity induced by FnrP carried on plasmid pGT3
(SA = 27 029.6, S.D. = 9482.7) was not signi¢cantly di¡erent than production by the pBluescript II KS(+) (Stratagene) control (SA = 16 540.8, S.D. = 9300.5). However,
FnrP* (pMJU6) induced a signi¢cant increase
(P 6 0.0001) in L-galactosidase activity (SA = 83 995.5,
S.D. = 23 551.6) compared to the control, con¢rming that
the mutant allele can activate a Fnr-dependent gene in the
presence of oxygen.
The activity of L-galactosidase from the lktC: :lacZ fusion in the fnr-null mutant strain GAU370 with plasmids
pMJU6 (FnrP*), pAAU6 (FnrP) or pHSS6 (control) was
compared aerobically. Activity with FnrP* (SA = 13 309.8,
S.D. = 3716.3) was signi¢cantly greater than with the control (SA = 9266.4, S.D. = 2464.9) (P 6 0.0001). However, Lgalactosidase
activity
with
FnrP
(SA = 2692.5,
S.D. = 952.7) was signi¢cantly less than the control
(P 6 0.0001). This con¢rmed the opposing a¡ects of the
active and inactive forms of FnrP on leukotoxin expression.
The e¡ect of FnrP* and FnrP on lapT, the divergently
encoded gene located 5P to the leukotoxin promoter region, was tested in GAU460 grown under aerobic conditions. GAU460 containing FnrP* induced signi¢cantly
greater
L-galactosidase
activity
(SA = 12 240.8,
S.D. = 1774.3) than either GAU460 containing FnrP
(SA = 7887.1, S.D. = 1082.2) (P 6 0.0001) or GAU460 harboring a vector control (SA = 5177.8, S.D. = 583.3)
(P 6 0.0001). FnrP-induced L-galactosidase activity, like
FEMSLE 9354 12-4-00
Downloaded from http://femsle.oxfordjournals.org/ by guest on October 20, 2016
A 43-bp DNA fragment containing the potential protein
binding sites, the P1 leukotoxin promoter, and £anking
HindIII sites was generated by synthesizing complementary deoxyoligonucleotides identical to the 3309 to 3267
promoter sequence of P. haemolytica lktC (GenBank accession No. M20730). The sequence for the sense strand
with the restriction sites underlined was 5P-AGCTTAAACACATCAAATTGAGATGTAGTTTCTCAATCCTCTTGATTCA. The oligonucleotides were annealed [14],
cloned into the HindIII site of plasmid pBC SK (+) (Stratagene) to form plasmid pFPS and propagated in E. coli
XL1-Blue (Stratagene). HindIII-digested pFPS was endlabeled using [K-32 P]dATP (Dupont, Boston, MA, USA)
and the Klenow fragment of DNA polymerase I (Promega) [14]. The labeled mix was puri¢ed as described [18].
Mobility shift reactions consisted of 45 ng labeled
DNA, 5 or 16 Wg partially puri¢ed proteins, 50 mM
EDTA and 5% glycerol in a ¢nal volume of 15 Wl of
1UKGB bu¡er [18]. The reaction mixtures were incubated
for 20 min at 35³C and subjected to electrophoresis
through a 5% polyacrylamide gel in TBE bu¡er [14].
76
G.A. Uhlich et al. / FEMS Microbiology Letters 186 (2000) 73^77
that of FnrP*, was signi¢cantly greater than the control
(P 6 0.0001).
3.3. DNA mobility shift assays
A DNA fragment with sequence upstream of lktC was
tested for interactions with FnrP and IHF in DNA mobility shift assays (Fig. 2). When 5 Wg of partially puri¢ed
FnrP or puri¢ed IHF was used, single bands were shifted
relative to the untreated control reaction. In the case of
FnrP, the shifted band corresponded to the higher band
produced when 16 Wg of partially puri¢ed FnrP was used
in the reaction mixture. For IHF, the shifted band corresponded in position with the lower band produced with 16
Wg of partially puri¢ed FnrP.
4. Discussion
This study demonstrated that Fnr and FnrP produced
in E. coli suppress leukotoxin expression in aerobic conditions. Reduction of O2 concentration reversed FnrPmediated leukotoxin suppression, but not suppression
mediated by Fnr. Studies with the active mutant, FnrP*,
suggest that the mechanism of this e¡ect may be an independent activation of leukotoxin expression by activated
FnrP, rather than by a simple relief of aerobically induced
suppression.
The conserved nature of the putative binding regions of
Acknowledgements
We thank J.R. Guest (University of She¤eld, She¤eld,
UK), G.C. Stewart (Kansas State University, Manhattan,
KS, USA) and S.K. Highlander (Baylor College of Medicine, Houston, TX, USA) for providing bacterial strains,
plasmids or bacteriophage. We also thank S. Goodman
(University of Southern California Dental School, Los
Angeles, CA, USA) for providing IHF protein and IHF
polyclonal antibodies. We thank J. Rosch for assistance
with manuscript preparation. This work was supported in
FEMSLE 9354 12-4-00
Downloaded from http://femsle.oxfordjournals.org/ by guest on October 20, 2016
Fig. 2. Mobility shift of the lktC promoter fragment with FnrP protein.
The P. haemolytica lktC promoter was carried on plasmid pFPS and
propagated in E. coli. Histidine-tagged P. haemolytica FnrP and E. coli
IHF proteins were overexpressed and puri¢ed from E. coli. Lane 1, no
protein. Lane 2, 16 Wg partially puri¢ed FnrP. Lane 3, 5 Wg partially
puri¢ed FnrP. Lane 4, 5 Wg puri¢ed IHF protein. The positions of vector DNA, free DNA and DNA/protein complexes are marked.
FnrP and Fnr indicates that both proteins may recognize
similar DNA binding sequences [9]. Three potential binding sites or half-sites were identi¢ed in the sequence upstream of the P. haemolytica leukotoxin P1 promoter. A
partially palindromic sequence centered 276 bp upstream
of the translational start site of lktC di¡ers from the consensus site by ¢ve additional bases in the central variable
region. Overlapping the palindromic sequence is a known
IHF binding site located 281^268 bp upstream of the
translational start site of lktC [7]. The 3P end of the IHF
binding site contains a sequence identical to the consensus
Fnr half-site. An additional Fnr half-site overlaps the 310
region of the P1 leukotoxin promoter. Binding of FnrP at
either half-site could a¡ect leukotoxin expression by excluding transcription initiated at the leukotoxin P1 promoter. Binding at the downstream half-site could inhibit
binding of another regulatory protein at the overlapping
near-IHF site.
Mobility shift assays con¢rmed binding of FnrP in the
leukotoxin promoter region containing these sites. The
lower shifted band produced at high protein concentrations corresponded in position to the IHF control and
may re£ect binding of IHF protein. IHF polyclonal antibody demonstrated reactivity at a position distinct from
the hyper-expressed FnrP protein during puri¢cation (results not shown).
Regulatory elements within the lktC^lapT intergenic
area are positioned such that they could allow interaction
with either the lapT or lktC promoter regions. Transcription from lapT in the presence of FnrP during aerobic
growth was slightly but signi¢cantly greater than the control. However, transcription of lapT in the presence of
FnrP* was over 2-fold greater than the control. Therefore,
FnrP has a regulatory e¡ect on lapT that is more typical
of the Fnr-like proteins and results in activation of expression during anaerobic growth.
This study demonstrates that O2 concentration can alter
leukotoxin transcription in E. coli in vitro, due to the
regulatory e¡ects of FnrP. In reduced O2 environments,
similar to conditions that could exist in regions of a pneumonic lung, the FnrP regulatory pathway of P. haemolytica could play an important role in regulation of leukotoxin production.
G.A. Uhlich et al. / FEMS Microbiology Letters 186 (2000) 73^77
part by the Kansas Agricultural Experiment Station Animal Health (Section 1433) Funds. Published as contribution #99-446-J of the KAES.
[10]
References
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
(1999) Cloning and characterization of the gene encoding Pasteurella
haemolytica FnrP: A regulator of the Escherichia coli silent hemolysin
SheA. J. Bacteriol. 181, 3845^3848.
Eiglmeier, K., Honorë, N., Iuchi, S., Lin, E.C.C. and Cole, S.T.
(1989) Molecular genetic analysis of FNR-dependent promoters.
Mol. Microbiol. 3, 869^878.
Sharrocks, A.D., Green, J. and Guest, J.R. (1991) FNR activates and
represses transcription in vitro. Proc. R. Soc. Lond. B 245, 219^226.
McInnes, J.I., Kim, J.E., Lian, C.-J. and Soltes, G.A. (1990) Actinobacillus pleuropneumoniae hlyX gene homology with the fnr gene of
Escherichia coli. J. Bacteriol. 172, 4587^4592.
Nash, H.A., Robertson, C.A., Flamm, E., Weisberg, R.A. and Miller,
H.I. (1987) Overproduction of Escherichia coli integration host factor, a protein with nonidentical subunits. J. Bacteriol. 169, 4124^
4127.
Ausubel, F.M., Brent, R., Kingston, R.E., Moore, D.D., Seidman,
J.G., Smith, J.A. and Struhl, K. (1992) Short Protocols in Molecular
Biology. Green Publishing Associates and John Wiley and Sons, New
York.
Miller, J.H. (1972) Experiments in Molecular Genetics. Cold Springs
Harbor Laboratory Press, Cold Spring Harbor, NY.
Kunkel, T.A. (1985) Rapid and e¤cient site-speci¢c mutagenesis
without phenotypic selection. Proc. Natl. Acad. Sci. USA 82, 488^
492.
Kiley, P.J. and Rezniko¡, W.S. (1991) Fnr mutants that activate gene
expression in the presence of oxygen. J. Bacteriol. 173, 16^22.
Lazazzera, B.A., Bates, D.M. and Kiley, P.J. (1993) The activity of
the Escherichia coli transcription factor FNR is regulated by a change
in oligomeric state. Genes Dev. 7, 1993^2005.
Spiro, S. and Guest, J.R. (1987) Activation of the lac operon of
Escherichia coli by a mutant FNR protein. Mol. Microbiol. 1, 53^58.
Seifert, H.S., Chen, E.Y., So, M. and He¡ron, F. (1986) Shuttle
mutagenesis: a method of transposon mutagenesis for Saccharomyces
cerevisiae. Proc. Natl. Acad. Sci. USA 83, 735^739.
FEMSLE 9354 12-4-00
Downloaded from http://femsle.oxfordjournals.org/ by guest on October 20, 2016
[1] Wilkie, B.N. and Shewen, P. (1988) De¢ning the role that Pasteurella
haemolytica plays in shipping fever. Vet. Med. Oct., 1053^1058.
[2] Confer, A.W., Panciera, R.J., Clinkenbeard, K.D. and Mosier, D.A.
(1990) Molecular aspects of virulence of Pasteurella haemolytica.
Can. J. Vet. Res. 54, S48^S52.
[3] Tatum, F.M., Briggs, R.E., Sreevatsan, S.S., Zehr, E.S., Hsuan, S.L.,
Whiteley, L.O., Ames, T.R. and Maheswaran, S.K. (1998) Construction of an isogenic leukotoxin deletion mutant of Pasteurella haemolytica serotype 1: characterization and virulence. Microb. Pathogen.
24, 37^46.
[4] Highlander, S.K., Chidambaram, M., Engler, M.J. and Weinstock,
G.M. (1989) DNA sequence of the Pasteurella haemolytica leukotoxin gene cluster. DNA 8, 15^28.
[5] Strathdee, C.A. and Lo, R.Y.C. (1989) Regulation of expression of
the Pasteurella haemolytica leukotoxin determinant. J. Bacteriol. 171,
5955^5962.
[6] Highlander, S.K., Wickersham, E.A., Garza, O. and Weinstock,
G.M. (1993) Expression of the Pasteurella haemolytica leukotoxin is
inhibited by a locus that encodes an ATP-binding cassette homolog.
Infect. Immun. 61, 3942^3951.
[7] Highlander, S.K. and Weinstock, G.M. (1994) Static DNA bending
and protein interactions within the Pasteurella haemolytica leukotoxin promoter region : development of an activation model for leukotoxin transcriptional control. DNA Cell Biol. 13, 171^181.
[8] Craig, N.L. and Nash, H.A. (1984) E. coli integration host factor
binds to speci¢c sites in DNA. Cell 39, 707^716.
[9] Uhlich, G.A., McNamara, P.J., Iandolo, J.J. and Mosier, D.A.
77