- Email: [email protected]
The Bvg accessory factor (Baf) enhances pertussis toxin expression in Escherichia coli and is essential for Bordetella pertussis viability
FEMS Microbiology Letters 193 (2000) 25^30
www.fems-microbiology.org
The Bvg accessory factor (Baf) enhances pertussis toxin expression in Escherichia coli and is essential for Bordetella pertussis viability Gwendolyn E. Wood 1 , Richard L. Friedman * Department of Microbiology and Immunology, University of Arizona Health Sciences Center, 1501 N. Campbell Avenue, Box 245049, Tucson, AZ 85724, USA Received 17 August 2000; accepted 25 September 2000
Abstract Pertussis toxin expression in the Gram-negative respiratory pathogen, Bordetella pertussis, is regulated by the BvgAS two-component system. Previous studies suggested that an additional gene encoding a Bvg accessory factor (Baf) was required, along with BvgAS, for expression of a ptx-lacZ fusion in Escherichia coli grown in rich medium. However, other studies showed that BvgAS is sufficient for ptxlacZ expression in minimal medium. Here we show that Baf acts with BvgAS to further increase ptx-lacZ expression in E. coli grown in minimal media and this is concomitant with a two-fold increase in BvgA protein levels. Gene replacement experiments show that baf is essential for viability of B. pertussis, suggesting that Baf affects the expression of other genes in addition to ptx. ß 2000 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved. Keywords : BvgAS; Pertussis toxin; Bvg accessory factor; Two-component system ; Bordetella
1. Introduction Pertussis toxin is an important virulence factor of Bordetella pertussis, the causative agent of pertussis or whooping cough [1]. The BvgAS two-component system controls the expression of pertussis toxin (ptx) and a number of other B. pertussis virulence factors including genes for ¢lamentous hemagglutinin (fhaB), BvgAS (bvgAS), and adenylate cyclase toxin (cya) [2]. Certain growth conditions (addition of magnesium sulfate or nicotinic acid or growth at 25³C) reversibly inhibit the BvgAS phosphorelay, a phenomenon called phenotypic modulation [2]. BvgASmediated regulation of several B. pertussis virulence factors has been examined in Escherichia coli. In rich medium the addition of bvgAS in trans increases fhaB-lacZ expression to high levels but has no e¡ect on ptx-lacZ expression. This suggests that ptx requires an unidenti¢ed factor in addition to BvgAS for optimal expression [3]. Since this initial observation, BvgAS-dependent ptx-lacZ expression * Corresponding author. Tel. : +1 (520) 626-6061; Fax: +1 (520) 626-2100; E-mail: [email protected] 1 Present address: Division of Allergy and Infectious Disease, University of Washington, Harborview Medical Center, P.O. Box 359779, 325 Ninth Avenue, Seattle, WA 98104, USA.
in rich medium has been achieved in E. coli by altering the supercoiling of the cloned ptx operon [4] or by introducing the B. pertussis baf (Bvg accessory factor) gene [5]. Baf is not similar to other proteins with known function and lacks identi¢able protein motifs that might aid in determining its function [5]. Several lines of evidence suggest that activation of the ptx promoter requires higher BvgA levels than other BvgA-regulated genes such as fhaB or bvg [6^9]. Scarlato et al. [7] found that the fhaB and bvg promoters are activated within 10 min of a shift from modulating to nonmodulating growth conditions, when levels of phosphorylated BvgA (BvgAVP) are relatively low [7]. In contrast, ptx and cya are expressed 2 h after this shift when BvgAVP levels are much higher. Furthermore, growth of E. coli strains in minimal media, which results in slow growth and concomitant increases in BvgA levels, allows ptx-lacZ expression by BvgAS in the absence of baf [10]. A number of in vitro studies suggest that high levels of BvgA are required to bind to the ptx promoter and activate transcription, while the bvg and fhaB promoters require much lower levels of BvgA [6,8,9]. Thus, two di¡erent models currently describe ptx regulation by BvgA. The ¢rst model proposes that BvgA is su¤cient for ptx promoter activation, but higher concentrations of phosphorylated BvgA are required than for early gene activation
0378-1097 / 00 / $20.00 ß 2000 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved. PII: S 0 3 7 8 - 1 0 9 7 ( 0 0 ) 0 0 4 4 8 - 1
FEMSLE 9667 16-11-00
26
G.E. Wood, R.L. Friedman / FEMS Microbiology Letters 193 (2000) 25^30
[6,8,9]. The second model proposes that another factor (Baf) assists BvgA in ptx promoter activation [5]. The work described in this paper shows that baf acts with BvgAS to further activate ptx transcription in E. coli grown in minimal medium without a¡ecting the growth rate, and suggests that functional Baf is required for viability of B. pertussis. 2. Materials and methods 2.1. Bacterial strains, plasmids, media and growth conditions The bacterial strains and plasmids used in this study are described in Table 1. E. coli cultures were routinely grown in Luria^Bertani (LB) broth (1% tryptone, 0.5% yeast extract, 170 mM NaCl, pH 7.5) or on LB agar plates. M9 minimal medium (M9) and Stainer^Scholte (SS) medium were made as described previously [10]. B. pertussis was grown on Bordet^Gengou (BG) agar supplemented with 10% de¢brinated sheep blood. Antibiotics were used at the following concentrations : ampicillin (Amp), 100 Wg ml31 ; chloramphenicol (Cam), 30 Wg ml31 ; gentamicin (Gent), 15 Wg ml31 ; kanamycin (Kan), 25 Wg ml31 ; nalidixic acid (Nal), 50 Wg ml31 ; streptomycin (Strep), 100 Wg ml31 ; and tetracycline (Tet), 15 Wg ml31 .
2.2. Construction of a baf mutant The plasmid (pBK1013) used to make the baf: :kan mutant was constructed in the following manner. The kanamycin cassette of pHP456-Km [11] was cloned on a 2.2-kb BamHI fragment into the BglII site within baf on pDD201 [5] so that the kanamycin resistance gene is transcribed in the opposite orientation as baf. This kanamycin cassette contains transcriptional and translational terminators at both ends to prevent polar e¡ects on adjacent genes [11]. The resulting baf: :kan allele was then cloned on a 4.4-kb NotI fragment into the NotI site of the allelic exchange vector pSS1129 [12] yielding pBK1013. Conjugations and allelic exchange were performed as previously described [12] using E. coli S17-1 as the donor and B. pertussis BP339 (pLAFR3) or BP339 (pLAFRBAF) as recipients. The presence of the baf: :kan mutation was con¢rmed by Southern hybridization (data not shown). To test for loss of pLAFRBAF from BP339, GWB1 or GWB1 containing the pBBR1MCS-5 derivatives (i.e., pBAF45, pBAF55, pBAF65 and pBAF75), strains were repeatedly passed (up to seven times) on BG, BG Kan, or BG Kan Gent agar, respectively. After each passage, isolated colonies (10^20) were picked to the same media containing Tet to test for Tet sensitivity (loss of pLAFRBAF).
Table 1 Bacterial strains and plasmids used Strain or plasmid E. coli DH5K JFME3 CR430 S17-1 B. pertussis BP339 GWB1 Plasmids pACYC184 pUW1004 pUWBAFB1 pUWBAFC2 pBBR1MCS-5 pBAF45 pBAF55 pBAF65 pBAF75 pLAFR3 pLAFRBAF pSS1129 pBK1013 pDD100 pDD100BC pDD100R pDD100SN pHP456Km
Relevant characteristics
Reference or source
K-complementation, high frequency transformation V lysogen containing ptxA-lacZ transcriptional fusion V lysogen containing bvgA-lacZ transcriptional fusion conjugation pro¢cient, RP4-2-Tet: :Mu-Kan: :Tn7
Gibco BRL [3] [14] [17]
Wild-type B. pertussis, StrepR BP339, baf: :kan, pLAFRBAF
[18] This study
p15A cloning vector, TetR , CamR
New England Biolabs [19] This study This study [15] This study This study This study This study [20] This study [12] This study [5] [5] [5] This study [11]
pACYC184 containing 14.7-kb bvgAS fragment pUW1004 containing the 1.2-kb PvuII bafBC fragment from pDD100BC in the EcoRV site upstream of bvgR pUW1004 containing the 1.2-kb PvuII baf fragment from pDD100 in the EcoRV site upstream of bvgR Broad host range cloning vector, GentR pBBR1MCS-5 containing the 2.2-kb PvuII-SalI baf fragment from pDD201 pBBR1MCS-5 containing the 2.2-kb PvuII-SalI bafBC fragment from pDD202 pBBR1MCS-5 containing the 1.1-kb SalI-PvuII baf fragment from pDD100R pBBR1MCS-5 containing the 1.1-kb PvuII-EcoRI bafSN fragment from pDD100SN Broad host range vector, TetR pLAFR3 containing the 6.5-kb SalI baf fragment from pDD201 Suicide vector for allelic exchange, GentR , AmpR , rpsL pSS1129 containing baf: :kan, KanR , AmpR pUC18 containing the 1.1-kb Sau3AIbaf fragment, AmpR pDD100 containing a 4-bp insertion at the BglII site of baf, AmpR pUC18 containing a 2.8-kb Sau3A baf insert, AmpR pDD100 with a 4-bp insertion at the StyI site of baf, AmpR Kan cassette vector, AmpR
FEMSLE 9667 16-11-00
G.E. Wood, R.L. Friedman / FEMS Microbiology Letters 193 (2000) 25^30
27
2.3. Southern hybridizations Southern hybridizations were performed using the Genius 1 kit for non-radioactive DNA labeling and detection of nucleic acids (Boehringer Mannheim). Chromosomal DNA was isolated from B. pertussis grown on plates using a hexadecyltrimethylammonium bromide extraction procedure [13]. The baf probe consisted of a 650-bp BamHI^ BglII fragment derived from pDD100 [5] and was labeled with digoxigenin-11-dUTP using random hexanucleotide primers according to the procedure described in the Genius manual. 2.4. Western blots E. coli cultures were grown overnight in LB broth then subcultured to 20 ml M9 medium and grown for 20 h. Bacteria were centrifuged at 7500Ug for 10 min then lysed by boiling for 10 min in PBS with 1% SDS. SDS-PAGE and Western blots were performed as described previously using 100 Wg total protein [5]. Band intensity was quantitated with the AlphaImage 2000 Documentation and Analysis system software using the spot densitometry function (AlphaInnotech). BvgA protein was detected with monoclonal antibody 63/8, a gift from Dr. Scott Stibitz, Center for Biologics Evaluation and Research, Food and Drug Administration, Bethesda, MD, USA. 3. Results and discussion 3.1. The e¡ect of Baf on ptx-lacZ expression in E. coli grown in minimal media Previous experiments demonstrated that expression of ptx-lacZ in E. coli grown in rich medium requires Baf in addition to BvgAS [5]. Recently, BvgAS-dependent ptxlacZ expression was achieved in E. coli strain JFME3 without Baf when the cultures were grown in M9 minimal salts medium (M9) or SS medium [10]. To determine the e¡ect of Baf on ptx-lacZ expression in minimal media, the wild-type baf gene and the bafBC allele were cloned onto pUW1004 which carries bvgAS. The bafBC allele contains a 4-bp insertion in the BglII site of baf, resulting in a frameshift in the C-terminus of Baf. The resulting plasmids were named pUWBAFC2 (baf ) and pUWBAFB1 (bafBC ), respectively, and their e¡ect on ptx-lacZ expression in E. coli JFME3 [3] was determined. pUW1004-activated ptx-lacZ expression in M9 (Fig. 1a) and SS medium (Fig. 1b) in agreement with previously published results [10]. However, pUWBAFC2 increased ptx-lacZ expression 3.5-fold in M9 medium (P 6 0.01 by Student's t-test) and 8.7-fold in SS medium (P 6 0.001) as compared to pUW1004 alone. Thus, maximal ptx-lacZ expression in E. coli requires both Baf and BvgAS. Interestingly, pUWBAFB1 (carrying bvgAS and the bafBC allele) reduced ptx-
Fig. 1. Activation of ptx-lacZ expression by BvgAS and Baf in E. coli grown in minimal medium. JFME3 (ptx-lacZ) [3] containing various plasmids was inoculated into LB medium and grown overnight (20 h). Forty microliters of overnight culture was used to inoculate 3 ml M9 (a) or SS (b) medium. Overnight cultures were assayed for L-galactosidase activity [21]. The mean and standard error of three independent experiments assayed in triplicate are shown.
lacZ expression to 10.6% (P 6 0.01) of pUW1004 control levels in SS medium (Fig. 1b). A reduction in ptx-lacZ expression was also observed in M9 medium (Fig. 1a) but this di¡erence was not statistically signi¢cant (P 6 0.1). 3.2. The e¡ect of Baf on BvgA expression in E. coli Since growth conditions that increase BvgA levels in E. coli also increase ptx-lacZ expression [10], we hypothesized that baf might have a similar e¡ect. We therefore determined the e¡ect of baf on BvgA levels in E. coli JFME3 containing pUW1004, pUWBAFB1 or pUWBAFC2 via Western blot and densitometry analysis. pUWBAFC2 (baf ) increased BvgA levels approximately two-fold as compared to pUW1004 in E. coli JFME3 (Fig. 2). pUWBAFB1 (bafBC ) did not a¡ect BvgA levels when the band intensity was corrected for di¡erences in background con-
FEMSLE 9667 16-11-00
28
G.E. Wood, R.L. Friedman / FEMS Microbiology Letters 193 (2000) 25^30
Fig. 2. Western blot analysis of whole cell lysates of E. coli JFME3 containing various plasmids grown in M9 medium. Whole cell lysates (100 Wg of protein) were electrophoresed through an SDS polyacrylamide gradient gel (5^20% polyacrylamide) and immunoblotted to detect BvgA using monoclonal antibody 63/8. The experiment was repeated four times with results similar to those presented.
sistent with its inability to increase ptx-lacZ expression. Modulation with 40 mM MgSO4 resulted in loss of BvgA expression by JFME3 containing pUW1004, pUWBAFB1 or pUWBAFC2 (data not shown). Thus, Baf-induced increases in ptx-lacZ expression in E. coli are concomitant with increases in BvgA levels. Conditions which slow the growth of E. coli have been shown to increase BvgA levels [10], therefore the e¡ect of Baf on the growth rate of E. coli JFME3 was determined. JFME3 (pUW1004) grew at the same rate as JFME3 (pUWBAFB1) or JFME3 (pUWBAFC2) in M9 and LB with similar ¢nal optical densities (data not shown). Growth of JFME3 in SS medium was poor but similar ¢nal optical densities were achieved for this strain carrying the various plasmids (data not shown). Thus, the mechanism by which Baf increases BvgA and ptx-lacZ expression in E. coli grown in minimal medium is not due to a decrease in the growth rate. To determine if Baf increases the transcription of bvgA in E. coli, pUW1004, pUWBAFB1, and pUWBAFC2 were introduced into E. coli CR430 which contains a chromosomal bvgA-lacZ transcriptional fusion [14]. As shown in Fig. 3, bvgAS alone (pUW1004) was able to activate bvgA-lacZ transcription, consistent with previous observations [14]. Wild-type baf with bvgAS (pUWBAFC2) had no e¡ect on bvgA-lacZ transcription in M9 medium (P 6 0.5) as compared to bvgAS alone (pUW1004) (Fig. 3a). However, in SS medium, Baf increased BvgAS-dependent bvgA-lacZ expression over BvgAS alone (P 6 0.01) (Fig. 3b). This suggests that Baf can increase BvgAS-dependent bvgA-lacZ transcription under certain growth conditions. The bafBC allele (pUWBAFB1) had a negative e¡ect on bvgAS autoregulation, reducing bvgA-lacZ expression to 30.5 and 36.4% in M9 and SS medium, respectively (P 6 0.01) as compared to pUW1004. These data suggest that Baf may be a Bvg accessory factor which acts by assisting BvgA in its role as a transcriptional activator either directly or indirectly. The observations that the bafBC allele decreases ptx-lacZ and bvgA-lacZ expression (Figs. 1 and 3) are consistent with a direct interaction between Baf and the BvgAS system. One possible explanation is that the BvgA and Baf proteins interact directly, but the mistranslated C-terminus of
BafBC prevents a productive interaction of BvgA with the bvg and ptx promoters. Alternatively, BafBC may prevent phosphorylation of BvgA by BvgS or inhibit Bvg-dependent promoter activation by interfering with the activation of RNA polymerase by BvgA. Previous studies showed that overproduction of Baf can partially restore the decrease in pertussis toxin expression caused by overproduction of the RNA polymerase K subunit (RpoA) [5]. These data, coupled with the results shown here, suggest that Baf interacts with RpoA and enhances stimulation of RNA polymerase by BvgA. Alternatively, Baf may enhance phosphorylation of BvgA by BvgS or prevent dephosphorylation of BvgA. Either of these scenarios are predicted to increase ptx-lacZ expression. 3.3. baf is an essential B. pertussis gene To explore the role of baf in ptx expression in B. pertussis, experiments were done to construct a B. pertussis baf mutant. Numerous attempts to construct a baf mutant
Fig. 3. Activation of bvgA-lacZ expression by BvgAS and Baf in E. coli grown in minimal medium. CR430 (bvgA-lacZ) [14] containing various plasmids was inoculated into LB medium and grown overnight (20 h). Forty microliters of overnight culture was used to inoculate 3 ml M9 (a) or SS (b) medium. Overnight cultures were assayed for L-galactosidase activity [21]. The mean and standard error of three independent experiments assayed in triplicate are shown.
FEMSLE 9667 16-11-00
G.E. Wood, R.L. Friedman / FEMS Microbiology Letters 193 (2000) 25^30
Fig. 4. Restriction maps of the various baf-containing fragments cloned on pBBR1MCS-5 and introduced into GWB1 (pLAFRBAF). The upper line shows the position of baf on the 6.5-kb SalI fragment relative to the LPS biosynthetic genes [22].
by allelic exchange and Campbell integration were unsuccessful ([5] and data not shown) which suggested that baf may be essential for B. pertussis viability. To investigate this possibility, we attempted to integrate a baf: :kan allele into the chromosome of B. pertussis BP339 in the presence (pLAFRBAF) or absence (pLAFR3) of plasmid-encoded wild-type baf. As expected, it was not possible to construct a baf: :kan mutation in BP339 (pLAFR3). However, the chromosomal baf allele was easily replaced with the baf: : kan mutation in BP339 (pLAFRBAF) and this strain was renamed GWB1. Southern blots con¢rmed the presence of wild-type baf on pLAFRBAF and the replacement of the chromosomal baf with the baf: :kan allele (data not shown). It follows that if baf is an essential gene, it should not be possible to cure GWB1 of pLAFRBAF without losing viability. Therefore, GWB1 (pLAFRBAF) and BP339 (pLAFRBAF) were repeatedly passed on BG agar without tetracycline to allow the loss of pLAFRBAF. While 100% of BP339 (pLAFRBAF) colonies had lost the plasmid after four passages on BG agar, no colonies of GWB1 (pLAFRBAF) lost pLAFRBAF even after seven passages (data not shown). This suggests that pLAFRBAF cannot be lost from GWB1 without losing cell viability. Finally, to rule out the possibility of polar e¡ects on nearby genes, various deletions of pLAFRBAF were made and cloned onto pBBR1MCS-5 [15]. The resulting plasmids (pBAF45, pBAF55, pBAF65 and pBAF75; Fig. 4) were then introduced into GWB1 (pLAFRBAF) and passed on BG agar without tetracycline to allow the loss of pLAFRBAF. Only plasmids containing an intact copy of baf (pBAF45 and pBAF65) allowed the loss of pLAFRBAF from GWB1. In contrast, introduction of pBBR1MCS-5, pBAF55 (bafBC ) or pBAF75 (bafSN ) did not allow loss of pLAFRBAF (data not shown). The bafBC and bafSN alleles have 4-bp insertions at the BglII and StyI sites, respectively (Fig. 4), resulting in frameshifts which should not a¡ect the expression of nearby genes. These data con¢rm that inactivation of baf itself is respon-
29
sible for the loss of viability and suggests that baf is an essential B. pertussis gene involved in the expression of other genes in addition to pertussis toxin (pertussis toxin is not essential for in vitro viability). In this respect, Baf resembles the Tex protein of B. pertussis [16]. Tex is a member of a new class of transcription-associated factors identi¢ed by its ability to restore toxin expression in B. pertussis mutants that overexpress RpoA [16]. The tex gene is also essential but di¡ers from baf in that it cannot be overexpressed in B. pertussis without inhibiting growth [16]. Further studies are in progress to characterize the role of Baf in virulence gene activation as well as its essential role in B. pertussis metabolism. Acknowledgements The authors thank David DeShazer for construction of plasmid pDD100SN. We also thank Scott Stibitz for helpful discussions and James Moulder and Derek Wood for critical review of the manuscript. This work was supported by Grant AI37600 to R.L.F. from the National Institutes of Health.
References [1] Ui, M. (1988) The multiple biological activities of pertussis toxin. In: Pathogenesis and Immunity to Pertussis (Wardlaw, A.C. and Parton, R., Eds.), pp. 121^145. John Wiley and Sons Ltd., New York. [2] Coote, J.G. (1991) Antigenic switching and pathogenicity: environmental e¡ects on virulence gene expression in Bordetella pertussis. J. Gen. Microbiol. 137, 2493^2503. [3] Miller, J.F., Roy, C.R. and Falkow, S. (1989) Analysis of Bordetella pertussis virulence gene regulation by use of transcriptional fusions in Escherichia coli. J. Bacteriol. 171, 6345^6348. [4] Scarlato, V., Arico, B. and Rappuoli, R. (1993) DNA topology affects transcriptional regulation of the pertussis toxin gene of Bordetella pertussis in Escherichia coli and in vitro. J. Bacteriol. 175, 4764^ 4771. [5] DeShazer, D., Wood, G.E. and Friedman, R.L. (1995) Identi¢cation of a Bordetella pertussis regulatory factor required for transcription of the pertussis toxin operon in Escherichia coli. J. Bacteriol. 177, 3801^3807. [6] Boucher, P.E. and Stibitz, S. (1995) Synergistic binding of RNA polymerase and BvgA phosphate to the pertussis toxin promoter of Bordetella pertussis. J. Bacteriol. 177, 6486^6491. [7] Scarlato, V., Arico, B., Prugnola, A. and Rappuoli, R. (1991) Sequential activation and environmental regulation of virulence genes in Bordetella pertussis. EMBO J. 10, 3971^3975. [8] Ste¡en, P., Goyard, S. and Ullman, A. (1996) Phosphorylated BvgA is su¤cient for transcriptional activation of virulence-regulated genes in Bordetella pertussis. EMBO J. 15, 102^109. [9] Zu, T., Manetti, R., Rappuoli, R. and Scarlato, V. (1996) Di¡erential binding of BvgA to two classes of virulence genes of Bordetella pertussis directs promoter selectivity by RNA polymerase. Mol. Microbiol. 21, 557^565. [10] Uhl, M.A. and Miller, J.F. (1995) BvgAS is su¤cient for activation of the Bordetella pertussis ptx locus in Escherichia coli. J. Bacteriol. 177, 6477^6485. [11] Fellay, R., Frey, J. and Krisch, H. (1987) Interposon mutagenesis of
FEMSLE 9667 16-11-00
30
[12] [13]
[14]
[15]
[16]
[17]
G.E. Wood, R.L. Friedman / FEMS Microbiology Letters 193 (2000) 25^30 soil and water bacteria : a family of DNA fragments designed for in vitro insertional mutagenesis of Gram-negative bacteria. Gene 52, 147^154. Stibitz, S. (1994) Use of conditionally counterselectable suicide vectors for allelic exchange. Methods Enzymol. 235, 458^465. Ausubel, F.M., Brent, R., Kingson, R.E., Moore, D.D., Seidman, J.G., Smith, J.A. and Struhl, K. (1988) Current Protocols in Molecular Biology. John Wiley and Sons, Inc., New York. Roy, C.R., Miller, J.F. and Falkow, S. (1990) Autogenous regulation of the Bordetella pertussis bvgABC operon. Proc. Natl. Acad. Sci. USA 87, 3763^3767. Kovach, M.E., Elzer, P.H., Hill, D.S., Robertson, G.R., Farris, M.A., Roop II, R.M. and Peterson, K.M. (1995) Four new derivatives of the broad-host-range cloning vector pBBR1MCS, carrying di¡erent antibiotic-resistance cassettes. Gene 166, 175^176. Fuchs, T.M., Deppisch, H., Scarlato, V. and Gross, R. (1996) A new gene locus of Bordetella pertussis de¢nes a novel family of prokaryotic transcriptional accessory proteins. J. Bacteriol. 178, 4445^4452. Simon, R., Priefer, U. and Puhler, A. (1983) A broad host range
[18]
[19]
[20]
[21] [22]
mobilization system for in vivo genetic engineering: transposon mutagenesis in Gram-negative bacteria. BioTechnology 1, 784^789. Bannan, J.D., Moran, M.J., MacInnes, J.I., Soltes, G.A. and Friedman, R.L. (1993) Cloning and characterization of btr, a Bordetella pertussis gene encoding an FNR-like transcriptional regulator. J. Bacteriol. 175, 7228^7235. Stibitz, S., Weiss, A.A. and Falkow, S. (1988) Genetic analysis of a region of the Bordetella pertussis chromosome encoding ¢lamentous hemagglutinin and the pleiotropic regulatory locus vir. J. Bacteriol. 170, 2904^2913. Peet, R.C., Lindgren, P.B., Willis, D.K. and Panopoulos, N.J. (1986) Identi¢cation and cloning of genes involved in phaseolotoxin production by Pseudomonas syringae pv. `phaseolicola'. J. Bacteriol. 166, 1096^1105. Miller, J. (1972) Experiments in Molecular Genetics. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Allen, A. and Maskell, D. (1996) The identi¢cation, cloning and mutagenesis of a genetic locus required for lipopolysaccharide biosynthesis in Bordetella pertussis. Mol. Microbiol. 19, 37^52.
FEMSLE 9667 16-11-00
www.fems-microbiology.org
The Bvg accessory factor (Baf) enhances pertussis toxin expression in Escherichia coli and is essential for Bordetella pertussis viability Gwendolyn E. Wood 1 , Richard L. Friedman * Department of Microbiology and Immunology, University of Arizona Health Sciences Center, 1501 N. Campbell Avenue, Box 245049, Tucson, AZ 85724, USA Received 17 August 2000; accepted 25 September 2000
Abstract Pertussis toxin expression in the Gram-negative respiratory pathogen, Bordetella pertussis, is regulated by the BvgAS two-component system. Previous studies suggested that an additional gene encoding a Bvg accessory factor (Baf) was required, along with BvgAS, for expression of a ptx-lacZ fusion in Escherichia coli grown in rich medium. However, other studies showed that BvgAS is sufficient for ptxlacZ expression in minimal medium. Here we show that Baf acts with BvgAS to further increase ptx-lacZ expression in E. coli grown in minimal media and this is concomitant with a two-fold increase in BvgA protein levels. Gene replacement experiments show that baf is essential for viability of B. pertussis, suggesting that Baf affects the expression of other genes in addition to ptx. ß 2000 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved. Keywords : BvgAS; Pertussis toxin; Bvg accessory factor; Two-component system ; Bordetella
1. Introduction Pertussis toxin is an important virulence factor of Bordetella pertussis, the causative agent of pertussis or whooping cough [1]. The BvgAS two-component system controls the expression of pertussis toxin (ptx) and a number of other B. pertussis virulence factors including genes for ¢lamentous hemagglutinin (fhaB), BvgAS (bvgAS), and adenylate cyclase toxin (cya) [2]. Certain growth conditions (addition of magnesium sulfate or nicotinic acid or growth at 25³C) reversibly inhibit the BvgAS phosphorelay, a phenomenon called phenotypic modulation [2]. BvgASmediated regulation of several B. pertussis virulence factors has been examined in Escherichia coli. In rich medium the addition of bvgAS in trans increases fhaB-lacZ expression to high levels but has no e¡ect on ptx-lacZ expression. This suggests that ptx requires an unidenti¢ed factor in addition to BvgAS for optimal expression [3]. Since this initial observation, BvgAS-dependent ptx-lacZ expression * Corresponding author. Tel. : +1 (520) 626-6061; Fax: +1 (520) 626-2100; E-mail: [email protected] 1 Present address: Division of Allergy and Infectious Disease, University of Washington, Harborview Medical Center, P.O. Box 359779, 325 Ninth Avenue, Seattle, WA 98104, USA.
in rich medium has been achieved in E. coli by altering the supercoiling of the cloned ptx operon [4] or by introducing the B. pertussis baf (Bvg accessory factor) gene [5]. Baf is not similar to other proteins with known function and lacks identi¢able protein motifs that might aid in determining its function [5]. Several lines of evidence suggest that activation of the ptx promoter requires higher BvgA levels than other BvgA-regulated genes such as fhaB or bvg [6^9]. Scarlato et al. [7] found that the fhaB and bvg promoters are activated within 10 min of a shift from modulating to nonmodulating growth conditions, when levels of phosphorylated BvgA (BvgAVP) are relatively low [7]. In contrast, ptx and cya are expressed 2 h after this shift when BvgAVP levels are much higher. Furthermore, growth of E. coli strains in minimal media, which results in slow growth and concomitant increases in BvgA levels, allows ptx-lacZ expression by BvgAS in the absence of baf [10]. A number of in vitro studies suggest that high levels of BvgA are required to bind to the ptx promoter and activate transcription, while the bvg and fhaB promoters require much lower levels of BvgA [6,8,9]. Thus, two di¡erent models currently describe ptx regulation by BvgA. The ¢rst model proposes that BvgA is su¤cient for ptx promoter activation, but higher concentrations of phosphorylated BvgA are required than for early gene activation
0378-1097 / 00 / $20.00 ß 2000 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved. PII: S 0 3 7 8 - 1 0 9 7 ( 0 0 ) 0 0 4 4 8 - 1
FEMSLE 9667 16-11-00
26
G.E. Wood, R.L. Friedman / FEMS Microbiology Letters 193 (2000) 25^30
[6,8,9]. The second model proposes that another factor (Baf) assists BvgA in ptx promoter activation [5]. The work described in this paper shows that baf acts with BvgAS to further activate ptx transcription in E. coli grown in minimal medium without a¡ecting the growth rate, and suggests that functional Baf is required for viability of B. pertussis. 2. Materials and methods 2.1. Bacterial strains, plasmids, media and growth conditions The bacterial strains and plasmids used in this study are described in Table 1. E. coli cultures were routinely grown in Luria^Bertani (LB) broth (1% tryptone, 0.5% yeast extract, 170 mM NaCl, pH 7.5) or on LB agar plates. M9 minimal medium (M9) and Stainer^Scholte (SS) medium were made as described previously [10]. B. pertussis was grown on Bordet^Gengou (BG) agar supplemented with 10% de¢brinated sheep blood. Antibiotics were used at the following concentrations : ampicillin (Amp), 100 Wg ml31 ; chloramphenicol (Cam), 30 Wg ml31 ; gentamicin (Gent), 15 Wg ml31 ; kanamycin (Kan), 25 Wg ml31 ; nalidixic acid (Nal), 50 Wg ml31 ; streptomycin (Strep), 100 Wg ml31 ; and tetracycline (Tet), 15 Wg ml31 .
2.2. Construction of a baf mutant The plasmid (pBK1013) used to make the baf: :kan mutant was constructed in the following manner. The kanamycin cassette of pHP456-Km [11] was cloned on a 2.2-kb BamHI fragment into the BglII site within baf on pDD201 [5] so that the kanamycin resistance gene is transcribed in the opposite orientation as baf. This kanamycin cassette contains transcriptional and translational terminators at both ends to prevent polar e¡ects on adjacent genes [11]. The resulting baf: :kan allele was then cloned on a 4.4-kb NotI fragment into the NotI site of the allelic exchange vector pSS1129 [12] yielding pBK1013. Conjugations and allelic exchange were performed as previously described [12] using E. coli S17-1 as the donor and B. pertussis BP339 (pLAFR3) or BP339 (pLAFRBAF) as recipients. The presence of the baf: :kan mutation was con¢rmed by Southern hybridization (data not shown). To test for loss of pLAFRBAF from BP339, GWB1 or GWB1 containing the pBBR1MCS-5 derivatives (i.e., pBAF45, pBAF55, pBAF65 and pBAF75), strains were repeatedly passed (up to seven times) on BG, BG Kan, or BG Kan Gent agar, respectively. After each passage, isolated colonies (10^20) were picked to the same media containing Tet to test for Tet sensitivity (loss of pLAFRBAF).
Table 1 Bacterial strains and plasmids used Strain or plasmid E. coli DH5K JFME3 CR430 S17-1 B. pertussis BP339 GWB1 Plasmids pACYC184 pUW1004 pUWBAFB1 pUWBAFC2 pBBR1MCS-5 pBAF45 pBAF55 pBAF65 pBAF75 pLAFR3 pLAFRBAF pSS1129 pBK1013 pDD100 pDD100BC pDD100R pDD100SN pHP456Km
Relevant characteristics
Reference or source
K-complementation, high frequency transformation V lysogen containing ptxA-lacZ transcriptional fusion V lysogen containing bvgA-lacZ transcriptional fusion conjugation pro¢cient, RP4-2-Tet: :Mu-Kan: :Tn7
Gibco BRL [3] [14] [17]
Wild-type B. pertussis, StrepR BP339, baf: :kan, pLAFRBAF
[18] This study
p15A cloning vector, TetR , CamR
New England Biolabs [19] This study This study [15] This study This study This study This study [20] This study [12] This study [5] [5] [5] This study [11]
pACYC184 containing 14.7-kb bvgAS fragment pUW1004 containing the 1.2-kb PvuII bafBC fragment from pDD100BC in the EcoRV site upstream of bvgR pUW1004 containing the 1.2-kb PvuII baf fragment from pDD100 in the EcoRV site upstream of bvgR Broad host range cloning vector, GentR pBBR1MCS-5 containing the 2.2-kb PvuII-SalI baf fragment from pDD201 pBBR1MCS-5 containing the 2.2-kb PvuII-SalI bafBC fragment from pDD202 pBBR1MCS-5 containing the 1.1-kb SalI-PvuII baf fragment from pDD100R pBBR1MCS-5 containing the 1.1-kb PvuII-EcoRI bafSN fragment from pDD100SN Broad host range vector, TetR pLAFR3 containing the 6.5-kb SalI baf fragment from pDD201 Suicide vector for allelic exchange, GentR , AmpR , rpsL pSS1129 containing baf: :kan, KanR , AmpR pUC18 containing the 1.1-kb Sau3AIbaf fragment, AmpR pDD100 containing a 4-bp insertion at the BglII site of baf, AmpR pUC18 containing a 2.8-kb Sau3A baf insert, AmpR pDD100 with a 4-bp insertion at the StyI site of baf, AmpR Kan cassette vector, AmpR
FEMSLE 9667 16-11-00
G.E. Wood, R.L. Friedman / FEMS Microbiology Letters 193 (2000) 25^30
27
2.3. Southern hybridizations Southern hybridizations were performed using the Genius 1 kit for non-radioactive DNA labeling and detection of nucleic acids (Boehringer Mannheim). Chromosomal DNA was isolated from B. pertussis grown on plates using a hexadecyltrimethylammonium bromide extraction procedure [13]. The baf probe consisted of a 650-bp BamHI^ BglII fragment derived from pDD100 [5] and was labeled with digoxigenin-11-dUTP using random hexanucleotide primers according to the procedure described in the Genius manual. 2.4. Western blots E. coli cultures were grown overnight in LB broth then subcultured to 20 ml M9 medium and grown for 20 h. Bacteria were centrifuged at 7500Ug for 10 min then lysed by boiling for 10 min in PBS with 1% SDS. SDS-PAGE and Western blots were performed as described previously using 100 Wg total protein [5]. Band intensity was quantitated with the AlphaImage 2000 Documentation and Analysis system software using the spot densitometry function (AlphaInnotech). BvgA protein was detected with monoclonal antibody 63/8, a gift from Dr. Scott Stibitz, Center for Biologics Evaluation and Research, Food and Drug Administration, Bethesda, MD, USA. 3. Results and discussion 3.1. The e¡ect of Baf on ptx-lacZ expression in E. coli grown in minimal media Previous experiments demonstrated that expression of ptx-lacZ in E. coli grown in rich medium requires Baf in addition to BvgAS [5]. Recently, BvgAS-dependent ptxlacZ expression was achieved in E. coli strain JFME3 without Baf when the cultures were grown in M9 minimal salts medium (M9) or SS medium [10]. To determine the e¡ect of Baf on ptx-lacZ expression in minimal media, the wild-type baf gene and the bafBC allele were cloned onto pUW1004 which carries bvgAS. The bafBC allele contains a 4-bp insertion in the BglII site of baf, resulting in a frameshift in the C-terminus of Baf. The resulting plasmids were named pUWBAFC2 (baf ) and pUWBAFB1 (bafBC ), respectively, and their e¡ect on ptx-lacZ expression in E. coli JFME3 [3] was determined. pUW1004-activated ptx-lacZ expression in M9 (Fig. 1a) and SS medium (Fig. 1b) in agreement with previously published results [10]. However, pUWBAFC2 increased ptx-lacZ expression 3.5-fold in M9 medium (P 6 0.01 by Student's t-test) and 8.7-fold in SS medium (P 6 0.001) as compared to pUW1004 alone. Thus, maximal ptx-lacZ expression in E. coli requires both Baf and BvgAS. Interestingly, pUWBAFB1 (carrying bvgAS and the bafBC allele) reduced ptx-
Fig. 1. Activation of ptx-lacZ expression by BvgAS and Baf in E. coli grown in minimal medium. JFME3 (ptx-lacZ) [3] containing various plasmids was inoculated into LB medium and grown overnight (20 h). Forty microliters of overnight culture was used to inoculate 3 ml M9 (a) or SS (b) medium. Overnight cultures were assayed for L-galactosidase activity [21]. The mean and standard error of three independent experiments assayed in triplicate are shown.
lacZ expression to 10.6% (P 6 0.01) of pUW1004 control levels in SS medium (Fig. 1b). A reduction in ptx-lacZ expression was also observed in M9 medium (Fig. 1a) but this di¡erence was not statistically signi¢cant (P 6 0.1). 3.2. The e¡ect of Baf on BvgA expression in E. coli Since growth conditions that increase BvgA levels in E. coli also increase ptx-lacZ expression [10], we hypothesized that baf might have a similar e¡ect. We therefore determined the e¡ect of baf on BvgA levels in E. coli JFME3 containing pUW1004, pUWBAFB1 or pUWBAFC2 via Western blot and densitometry analysis. pUWBAFC2 (baf ) increased BvgA levels approximately two-fold as compared to pUW1004 in E. coli JFME3 (Fig. 2). pUWBAFB1 (bafBC ) did not a¡ect BvgA levels when the band intensity was corrected for di¡erences in background con-
FEMSLE 9667 16-11-00
28
G.E. Wood, R.L. Friedman / FEMS Microbiology Letters 193 (2000) 25^30
Fig. 2. Western blot analysis of whole cell lysates of E. coli JFME3 containing various plasmids grown in M9 medium. Whole cell lysates (100 Wg of protein) were electrophoresed through an SDS polyacrylamide gradient gel (5^20% polyacrylamide) and immunoblotted to detect BvgA using monoclonal antibody 63/8. The experiment was repeated four times with results similar to those presented.
sistent with its inability to increase ptx-lacZ expression. Modulation with 40 mM MgSO4 resulted in loss of BvgA expression by JFME3 containing pUW1004, pUWBAFB1 or pUWBAFC2 (data not shown). Thus, Baf-induced increases in ptx-lacZ expression in E. coli are concomitant with increases in BvgA levels. Conditions which slow the growth of E. coli have been shown to increase BvgA levels [10], therefore the e¡ect of Baf on the growth rate of E. coli JFME3 was determined. JFME3 (pUW1004) grew at the same rate as JFME3 (pUWBAFB1) or JFME3 (pUWBAFC2) in M9 and LB with similar ¢nal optical densities (data not shown). Growth of JFME3 in SS medium was poor but similar ¢nal optical densities were achieved for this strain carrying the various plasmids (data not shown). Thus, the mechanism by which Baf increases BvgA and ptx-lacZ expression in E. coli grown in minimal medium is not due to a decrease in the growth rate. To determine if Baf increases the transcription of bvgA in E. coli, pUW1004, pUWBAFB1, and pUWBAFC2 were introduced into E. coli CR430 which contains a chromosomal bvgA-lacZ transcriptional fusion [14]. As shown in Fig. 3, bvgAS alone (pUW1004) was able to activate bvgA-lacZ transcription, consistent with previous observations [14]. Wild-type baf with bvgAS (pUWBAFC2) had no e¡ect on bvgA-lacZ transcription in M9 medium (P 6 0.5) as compared to bvgAS alone (pUW1004) (Fig. 3a). However, in SS medium, Baf increased BvgAS-dependent bvgA-lacZ expression over BvgAS alone (P 6 0.01) (Fig. 3b). This suggests that Baf can increase BvgAS-dependent bvgA-lacZ transcription under certain growth conditions. The bafBC allele (pUWBAFB1) had a negative e¡ect on bvgAS autoregulation, reducing bvgA-lacZ expression to 30.5 and 36.4% in M9 and SS medium, respectively (P 6 0.01) as compared to pUW1004. These data suggest that Baf may be a Bvg accessory factor which acts by assisting BvgA in its role as a transcriptional activator either directly or indirectly. The observations that the bafBC allele decreases ptx-lacZ and bvgA-lacZ expression (Figs. 1 and 3) are consistent with a direct interaction between Baf and the BvgAS system. One possible explanation is that the BvgA and Baf proteins interact directly, but the mistranslated C-terminus of
BafBC prevents a productive interaction of BvgA with the bvg and ptx promoters. Alternatively, BafBC may prevent phosphorylation of BvgA by BvgS or inhibit Bvg-dependent promoter activation by interfering with the activation of RNA polymerase by BvgA. Previous studies showed that overproduction of Baf can partially restore the decrease in pertussis toxin expression caused by overproduction of the RNA polymerase K subunit (RpoA) [5]. These data, coupled with the results shown here, suggest that Baf interacts with RpoA and enhances stimulation of RNA polymerase by BvgA. Alternatively, Baf may enhance phosphorylation of BvgA by BvgS or prevent dephosphorylation of BvgA. Either of these scenarios are predicted to increase ptx-lacZ expression. 3.3. baf is an essential B. pertussis gene To explore the role of baf in ptx expression in B. pertussis, experiments were done to construct a B. pertussis baf mutant. Numerous attempts to construct a baf mutant
Fig. 3. Activation of bvgA-lacZ expression by BvgAS and Baf in E. coli grown in minimal medium. CR430 (bvgA-lacZ) [14] containing various plasmids was inoculated into LB medium and grown overnight (20 h). Forty microliters of overnight culture was used to inoculate 3 ml M9 (a) or SS (b) medium. Overnight cultures were assayed for L-galactosidase activity [21]. The mean and standard error of three independent experiments assayed in triplicate are shown.
FEMSLE 9667 16-11-00
G.E. Wood, R.L. Friedman / FEMS Microbiology Letters 193 (2000) 25^30
Fig. 4. Restriction maps of the various baf-containing fragments cloned on pBBR1MCS-5 and introduced into GWB1 (pLAFRBAF). The upper line shows the position of baf on the 6.5-kb SalI fragment relative to the LPS biosynthetic genes [22].
by allelic exchange and Campbell integration were unsuccessful ([5] and data not shown) which suggested that baf may be essential for B. pertussis viability. To investigate this possibility, we attempted to integrate a baf: :kan allele into the chromosome of B. pertussis BP339 in the presence (pLAFRBAF) or absence (pLAFR3) of plasmid-encoded wild-type baf. As expected, it was not possible to construct a baf: :kan mutation in BP339 (pLAFR3). However, the chromosomal baf allele was easily replaced with the baf: : kan mutation in BP339 (pLAFRBAF) and this strain was renamed GWB1. Southern blots con¢rmed the presence of wild-type baf on pLAFRBAF and the replacement of the chromosomal baf with the baf: :kan allele (data not shown). It follows that if baf is an essential gene, it should not be possible to cure GWB1 of pLAFRBAF without losing viability. Therefore, GWB1 (pLAFRBAF) and BP339 (pLAFRBAF) were repeatedly passed on BG agar without tetracycline to allow the loss of pLAFRBAF. While 100% of BP339 (pLAFRBAF) colonies had lost the plasmid after four passages on BG agar, no colonies of GWB1 (pLAFRBAF) lost pLAFRBAF even after seven passages (data not shown). This suggests that pLAFRBAF cannot be lost from GWB1 without losing cell viability. Finally, to rule out the possibility of polar e¡ects on nearby genes, various deletions of pLAFRBAF were made and cloned onto pBBR1MCS-5 [15]. The resulting plasmids (pBAF45, pBAF55, pBAF65 and pBAF75; Fig. 4) were then introduced into GWB1 (pLAFRBAF) and passed on BG agar without tetracycline to allow the loss of pLAFRBAF. Only plasmids containing an intact copy of baf (pBAF45 and pBAF65) allowed the loss of pLAFRBAF from GWB1. In contrast, introduction of pBBR1MCS-5, pBAF55 (bafBC ) or pBAF75 (bafSN ) did not allow loss of pLAFRBAF (data not shown). The bafBC and bafSN alleles have 4-bp insertions at the BglII and StyI sites, respectively (Fig. 4), resulting in frameshifts which should not a¡ect the expression of nearby genes. These data con¢rm that inactivation of baf itself is respon-
29
sible for the loss of viability and suggests that baf is an essential B. pertussis gene involved in the expression of other genes in addition to pertussis toxin (pertussis toxin is not essential for in vitro viability). In this respect, Baf resembles the Tex protein of B. pertussis [16]. Tex is a member of a new class of transcription-associated factors identi¢ed by its ability to restore toxin expression in B. pertussis mutants that overexpress RpoA [16]. The tex gene is also essential but di¡ers from baf in that it cannot be overexpressed in B. pertussis without inhibiting growth [16]. Further studies are in progress to characterize the role of Baf in virulence gene activation as well as its essential role in B. pertussis metabolism. Acknowledgements The authors thank David DeShazer for construction of plasmid pDD100SN. We also thank Scott Stibitz for helpful discussions and James Moulder and Derek Wood for critical review of the manuscript. This work was supported by Grant AI37600 to R.L.F. from the National Institutes of Health.
References [1] Ui, M. (1988) The multiple biological activities of pertussis toxin. In: Pathogenesis and Immunity to Pertussis (Wardlaw, A.C. and Parton, R., Eds.), pp. 121^145. John Wiley and Sons Ltd., New York. [2] Coote, J.G. (1991) Antigenic switching and pathogenicity: environmental e¡ects on virulence gene expression in Bordetella pertussis. J. Gen. Microbiol. 137, 2493^2503. [3] Miller, J.F., Roy, C.R. and Falkow, S. (1989) Analysis of Bordetella pertussis virulence gene regulation by use of transcriptional fusions in Escherichia coli. J. Bacteriol. 171, 6345^6348. [4] Scarlato, V., Arico, B. and Rappuoli, R. (1993) DNA topology affects transcriptional regulation of the pertussis toxin gene of Bordetella pertussis in Escherichia coli and in vitro. J. Bacteriol. 175, 4764^ 4771. [5] DeShazer, D., Wood, G.E. and Friedman, R.L. (1995) Identi¢cation of a Bordetella pertussis regulatory factor required for transcription of the pertussis toxin operon in Escherichia coli. J. Bacteriol. 177, 3801^3807. [6] Boucher, P.E. and Stibitz, S. (1995) Synergistic binding of RNA polymerase and BvgA phosphate to the pertussis toxin promoter of Bordetella pertussis. J. Bacteriol. 177, 6486^6491. [7] Scarlato, V., Arico, B., Prugnola, A. and Rappuoli, R. (1991) Sequential activation and environmental regulation of virulence genes in Bordetella pertussis. EMBO J. 10, 3971^3975. [8] Ste¡en, P., Goyard, S. and Ullman, A. (1996) Phosphorylated BvgA is su¤cient for transcriptional activation of virulence-regulated genes in Bordetella pertussis. EMBO J. 15, 102^109. [9] Zu, T., Manetti, R., Rappuoli, R. and Scarlato, V. (1996) Di¡erential binding of BvgA to two classes of virulence genes of Bordetella pertussis directs promoter selectivity by RNA polymerase. Mol. Microbiol. 21, 557^565. [10] Uhl, M.A. and Miller, J.F. (1995) BvgAS is su¤cient for activation of the Bordetella pertussis ptx locus in Escherichia coli. J. Bacteriol. 177, 6477^6485. [11] Fellay, R., Frey, J. and Krisch, H. (1987) Interposon mutagenesis of
FEMSLE 9667 16-11-00
30
[12] [13]
[14]
[15]
[16]
[17]
G.E. Wood, R.L. Friedman / FEMS Microbiology Letters 193 (2000) 25^30 soil and water bacteria : a family of DNA fragments designed for in vitro insertional mutagenesis of Gram-negative bacteria. Gene 52, 147^154. Stibitz, S. (1994) Use of conditionally counterselectable suicide vectors for allelic exchange. Methods Enzymol. 235, 458^465. Ausubel, F.M., Brent, R., Kingson, R.E., Moore, D.D., Seidman, J.G., Smith, J.A. and Struhl, K. (1988) Current Protocols in Molecular Biology. John Wiley and Sons, Inc., New York. Roy, C.R., Miller, J.F. and Falkow, S. (1990) Autogenous regulation of the Bordetella pertussis bvgABC operon. Proc. Natl. Acad. Sci. USA 87, 3763^3767. Kovach, M.E., Elzer, P.H., Hill, D.S., Robertson, G.R., Farris, M.A., Roop II, R.M. and Peterson, K.M. (1995) Four new derivatives of the broad-host-range cloning vector pBBR1MCS, carrying di¡erent antibiotic-resistance cassettes. Gene 166, 175^176. Fuchs, T.M., Deppisch, H., Scarlato, V. and Gross, R. (1996) A new gene locus of Bordetella pertussis de¢nes a novel family of prokaryotic transcriptional accessory proteins. J. Bacteriol. 178, 4445^4452. Simon, R., Priefer, U. and Puhler, A. (1983) A broad host range
[18]
[19]
[20]
[21] [22]
mobilization system for in vivo genetic engineering: transposon mutagenesis in Gram-negative bacteria. BioTechnology 1, 784^789. Bannan, J.D., Moran, M.J., MacInnes, J.I., Soltes, G.A. and Friedman, R.L. (1993) Cloning and characterization of btr, a Bordetella pertussis gene encoding an FNR-like transcriptional regulator. J. Bacteriol. 175, 7228^7235. Stibitz, S., Weiss, A.A. and Falkow, S. (1988) Genetic analysis of a region of the Bordetella pertussis chromosome encoding ¢lamentous hemagglutinin and the pleiotropic regulatory locus vir. J. Bacteriol. 170, 2904^2913. Peet, R.C., Lindgren, P.B., Willis, D.K. and Panopoulos, N.J. (1986) Identi¢cation and cloning of genes involved in phaseolotoxin production by Pseudomonas syringae pv. `phaseolicola'. J. Bacteriol. 166, 1096^1105. Miller, J. (1972) Experiments in Molecular Genetics. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Allen, A. and Maskell, D. (1996) The identi¢cation, cloning and mutagenesis of a genetic locus required for lipopolysaccharide biosynthesis in Bordetella pertussis. Mol. Microbiol. 19, 37^52.
FEMSLE 9667 16-11-00