Research in Microbiology 156 (2005) 225–232 www.elsevier.com/locate/resmic
Unique stress response to the lactoperoxidase-thiocyanate enzyme system in Escherichia coli Jan Sermon, Kristof Vanoirbeek, Philipp De Spiegeleer, Rob Van Houdt, Abram Aertsen, Chris W. Michiels ∗ Laboratory of Food Microbiology, Katholieke Universiteit Leuven, Kasteelpark Arenberg 22, 3001 Leuven (Heverlee), Belgium Received 11 August 2004; accepted 30 September 2004 Available online 19 November 2004
Abstract Using a differential fluorescence induction approach, we screened a promoter trap library constructed in a vector with a promoterless gfp gene for Escherichia coli MG1655 promoters that are induced upon challenge with the antimicrobial lactoperoxidase-thiocyanate enzyme system. None of the thirteen identified lactoperoxidase-inducible open reading frames was inducible by H2 O2 or by the superoxide generator plumbagin. However, analysis of specific promoters of known stress genes showed some of these, including recA, dnaK and sodA, to be inducible by the lactoperoxidase-thiocyanate enzyme system. The results show that the lactoperoxidase-thiocyanate enzyme system elicits a distinct stress response different from but partly overlapping other oxidative stress responses. Several of the induced genes or pathways may be involved in bacterial defense against the toxic effects of the lactoperoxidase-thiocyanate enzyme system. 2004 Elsevier SAS. All rights reserved. Keywords: Oxidative stress; Stress response; Escherichia coli; Lactoperoxidase; Differential fluorescence induction; Antimicrobial enzyme system
1. Introduction Lactoperoxidase (LP) is a heme peroxidase that catalyzes oxidation by hydrogen peroxide of a wide range of substrates. It is found in vertebrate secretions such as milk, saliva, tears and airway mucus, and plays an important role in innate host defense against pathogenic microorganisms [19,32,49]. Vertebrates also produce other related, heme peroxidases with a similar function, such as myeloperoxidase which contributes to microbial killing in phagocytes [20,21]. There is growing interest in the application of LP and other heme peroxidases as biopreservatives in foods and in industrial, cosmetic and pharmaceutical products [6,26]. Because of its occurrence and function in the human body, and its natural presence in a number of foods, LP is assumed to present little or no toxicological con* Corresponding author.
E-mail address:
[email protected] (C.W. Michiels). 0923-2508/$ – see front matter 2004 Elsevier SAS. All rights reserved. doi:10.1016/j.resmic.2004.09.012
cern when used in foods provided that the uptake remains within the range of normal food exposure. Although the effect of the LP system is generally bacteriostatic, application of the system even at low doses can strongly increase bacterial inactivation in combination with other treatments, and this opens up interesting perspectives for novel combined preservation approaches based on hurdle technology [13,47]. The major physiological substrate for LP is thiocyanate (SCN− ), which is oxidized to hypothiocyanate (OSCN− ) as a primary reaction product. Bromide and iodide anions can also be oxidized by LP, but unlike myeloperoxidase, LP is unable to oxidize chloride ions [2,12,31]. The antimicrobial activity of these enzyme systems is non-specific and stems from the oxidative power of their reaction products. The primary cellular targets of the LP/SCN− system are believed to be sulfhydryl groups in cellular proteins and probably also low molecular weight components of the cytoplasmic thiol pool, which are oxidized into disulfides (–S–S–), sulfenyl thiocyanates (–S–SCN) or sulfenic acids (–S–OH) [3,30].
226
J. Sermon et al. / Research in Microbiology 156 (2005) 225–232
As a result, bacteria exposed to the LP/SCN− system become functionally damaged, showing loss of pH gradient, K+ leakage, inhibition of respiration, and/or inhibition of protein, RNA and DNA synthesis [9,22,33,35]. As opposed − to oxidants such as H2 O2 and O− 2 (superoxide), OSCN has not been reported to cause DNA damage, and is considered not to be toxic for the vertebrate host cells producing it. Bacteria have developed specific mechanisms to cope with oxidative challenge. Escherichia coli produces specific H2 O2 and O− 2 sensors which function as transcriptional activators of specific oxidative stress regulons. OxyR responds to H2 O2 by means of a thiol-disulfide switch, and activates a number of genes encoding, for instance, hydroperoxidase I (katG) and alkyl hydroperoxide reductase (ahpCF) which are involved in eliminating peroxides, and glutaredoxin 1 (grxA), glutathione reductase (gorA) and thioredoxin 2 (trxC), which help maintain the thiol-disulfide balance. The other sensor, SoxR, responds to O− 2 by means of an iron–sulfur cluster. Among the soxRS regulon are genes encoding manganese superoxide dismutase (sodA), which deals directly with superoxide, endonuclease IV (nfo), which plays a role in repair of DNA damage that may be inflicted by superoxide, and glucose-6-phosphate dehydrogenase (zwf ), which generates reducing power for the cell to cope with oxidants [40]. In addition to these regulons a number of other functions which are not under SoxRS or OxyR control are known to play a role in defense against H2 O2 and O− 2 . Furthermore, the soxRS regulon also responds to ‘nitrosative’ stress by nitrogen monoxide (NO) [28]. This is relevant since macrophages use both O− 2 and NO as a defense against pathogenic bacteria. A NO-specific bacterial defense system that is SoxRS-independent is flavohemoglobin (Hmp), which neutralizes NO by dioxygenation to NO− 3 [16]. In spite of the importance of the LP/SCN− system in host defense against pathogens and its potential application as a food biopreservative, the bacterial stress response to this specific form of oxidative stress has not been characterized to date. During the preparation of this manuscript, a DNA microarray study appeared on the E. coli stress response to the Curvularia haloperoxidase in combination with bromide as a substrate, showing the induction of only a limited number of genes, among them ibpA, ibpB and cpxP, known to be induced by protein denaturation stress [15]. In this study, we used differential fluorescence induction (DFI) [44] to screen for promoters that are induced by the LP/SCN− system in E. coli, and we demonstrate that the LP/SCN− system induces a unique stress response that is different from the SoxRS- and OxyR-dependent responses. This work will contribute to a better understanding of the cellular mode of action of the LP/SCN− enzyme system and of the bacterial stress response to this system in host– pathogen interactions.
Table 1 Plasmids used in this work Plasmid
Relevant characteristics
Reference
pFPV25
bla, mob, promoterless gfpmut3a, Ampr
[44]
pFPV25PrecA pFPV25PkatE pFPV25PgrxA pFPV25PdnaK pFPV25PsodA pFPV25Pzwf
pFPV25 + recA promoter upstream gfp pFPV25 + katE promoter upstream gfp pFPV25 + grxA promoter upstream gfp pFPV25 + dnaK promoter upstream gfp pFPV25 + sodA promoter upstream gfp pFPV25 + zwf promoter upstream gfp
[5] [5] [5] [5] [5] [5]
pJS115 pJS116 pJS117 pJS118 pJS119 pJS120
pFPV25 + cysJ promoter upstream gfp pFPV25 + ydjM promoter upstream gfp pFPV25 + ydjN promoter upstream gfp pFPV25 + fruB promoter upstream gfp pFPV25 + ybjG promoter upstream gfp pFPV25 + hscB promoter, hscB and hscA upstream gfp
This study This study This study This study This study This study
2. Materials and methods 2.1. Bacterial strains, plasmids and culture conditions E. coli K-12 MG1655 was used as a host strain in this work. Plasmid constructs (except for clones from the promoter trap library) are listed in Table 1. Overnight cultures were grown in Luria–Bertani (LB) broth for 21 h at 37 ◦ C under well-aerated conditions, with 100 µg/ml ampicillin (Ap100 ) (Applichem, Darmstadt, Germany) when appropriate. 2.2. Enzymes and chemicals A stock solution of lactoperoxidase (Sigma–Aldrich, Bornem, Belgium) (10 mg/ml) was prepared in a 50% glycerol solution in phosphate-buffered saline (2.87 mM KH2 PO4 , 7.12 mM K2 HPO4 , 0.151 M NaCl, pH 6.0). A stock solution of plumbagin (5-hydroxy-2-methyl-1,4naphtoquinone) (Sigma–Aldrich) (20 mM) was prepared in dimethylsulfoxide. Both stock solutions were stored at −18 ◦ C. Potassium thiocyanate (KSCN) and hydrogen peroxide (H2 O2 ) were stored at 4 ◦ C as a 25 mM stock solution, prepared freshly every week. 2.3. Identification of promoters inducible by the LP system The construction and validation of the promoter trap library in E. coli MG1655 was described earlier [1]. Briefly, 1–3-kb genomic DNA fragments of a partial Sau3AI digest were ligated into the BamHI site of pFPV25 [44], upstream of the promoterless gfp gene. Approximately 15 000 independent clones were purified on LB agar + Ap100 and stored at −80 ◦ C individually in microplate wells. To identify promoters inducible by the LP system, clones of the promoter probe library were individually grown in 200 µl LB in microplates for 21 h at 37 ◦ C on a rotary shaker (200 rpm), diluted 1/60 in 300 µl fresh prewarmed LB and
J. Sermon et al. / Research in Microbiology 156 (2005) 225–232
grown until an optical density at 600 nm (OD600 ) of 0.65 measured with a Multiskan RC microplate reader (Thermo Labsystems, Brussels, Belgium) which corresponds to a spectrophotometer value of 0.8. One half of the culture volume was then transferred to a new microplate and induced by addition of 100 µl of a mixture of enzyme and substrates to obtain final concentrations of 5 µg/ml lactoperoxidase, 0.50 mM H2 O2 and 0.50 mM KSCN. To the other half of the culture, 100 µl water was added as a control treatment. After incubation for 16 h at 37 ◦ C on a rotary shaker (200 rpm), fluorescence at 520 nm was measured in a fluorescence microplate reader at 37 ◦ C (Fluoroskan Ascent FL, Thermo Labsystems), using an excitation wavelength of 485 nm. Clones showing a fluorescence signal significantly higher than the corresponding water control were retested in triplicate with continuous monitoring of fluorescence during 16 h at 37 ◦ C. To identify clones inducible by H2 O2 and/or KSCN in the absence of LP enzyme, an additional control treatment was included in this confirmation experiment consisting of induction by a mixture of 0.50 mM H2 O2 and 0.50 mM KSCN. To determine the effect of the LP system, H2 O2 and/or KSCN on bacterial growth in these experiments, growth curves under identical conditions were recorded using a Bioscreen C Automatic Growth Analyzer (Thermo Labsystems). Plasmids from clones that were specifically induced by the LP enzyme system were isolated and sequenced by MWG Biotech AG (Ebersberg, Germany) using gfp-pR (Table 2) in the 5 -region of gfp as sequencing primer. 2.4. PCR cloning of specific promoters Specific gfp transcriptional fusions were constructed in pFPV25 with selected promoters found to be inducible by the LP enzyme system (cysJ, ydjM, ydjN, fruB, ybjG and fdx). Promoter fragments were amplified using Platinum Pfx DNA polymerase (Invitrogen, Merelbeke, Belgium) using the primers specified in Table 2 (synthesized by Eurogentec, Seraing, Belgium). It should be noted that for fdx, the cloned fragment spans two upstream located genes, hscB and hscA, and the hscB promoter, because these genes are believed to form an operon with fdx. All six PCR products were cloned as EcoRI/BamHI fragments upstream of gfp in pFPV25, resulting in pJS115, pJS116, pJS117, pJS118, pJS119, pJS120, respectively (Table 1). Restriction endonucleases were obtained from Roche Diagnostics (Vilvoorde, Belgium). Constructs were verified by PCR using Taq polymerase (Fermentas, St. Leon-Rot, Germany) with the corresponding upstream primer used in the initial amplification step, and the primer gfp-pR (Table 2). 2.5. Analysis of induction by flow cytometry The induction of LP/SCN− responding clones and PCRcloned promoters in pFPV25 was studied by flow cytomet-
227
Table 2 PCR primers used in this work Primers
Sequencea,b,c
cysJ-pF cysJ-pR ydjN-pF ydjN-pR ydjM-pF ydjM-pR ybjG-pF ybjG-pR fruB-pF fruB-pR hscB-pF fdx-pR gfp-pR
5 -ATCGGAATTCTTTGCTAAAACCTCGCTGGT-3 5 -GATCGGATCCTTACGGGTTCAACGGAAGCAACG-3 5 -TCAGGAATTCACAGTCTGCTGGCGGTATTT-3 5 -TACGGGATCCTTACACCACGATGTTCGCAATTA-3 5 -CTGAGAATTCCCGAAAAACTGCCTTACAGC-3 5 -GACTGGATCCTTAGGCAAATACCGCACAAGCAAT-3 5 -ACTGGAATTCTCGTTGCCGATATAGGTTGA-3 5 -GTACGGATCCTTATTTAGCAATAAAAATCGCCAAC-3 5 -CTAGGAATTCGGTCGATCACTGGAAGAGGA-3 5 -TCAGGGATCCTTACGGATGGATGTCCTGTACG-3 5 -TGACGAATTCCCATTGCGGACTATAAAAGCA-3 5 -ACGTGGATCCTTAACCGCTATTAGCTTCCAGAACA-3 5 -GACAAGTGTTGGCCATGGAACAGGT-3
a GAATTC: EcoRI recognition site. b GGATCC: BamHI recognition site. c TTA: complement of translation stop codon on reverse primer.
ric analysis of gfp expression, using a culturing and induction procedure specifically optimized for this purpose. Overnight cultures were diluted 1/40 in prewarmed LB and incubated for 30 min at 30 ◦ C to bring them to the exponential growth phase. Cultures were subsequently treated with either the entire LP system (5 µg/ml LP enzyme, 0.50 mM H2 O2 and 0.50 mM KSCN), H2 O2 (0.50 mM), KSCN (0.50 mM), or plumbagin (40 µM), and incubated for 4 h at 30 ◦ C (200 rpm). Flow cytometric analysis was then performed using a FACSCalibur apparatus (Beckton– Dickinson, Aalst, Belgium) equipped with a 15 mW argon laser emitting at 488 nm. Previously optimized instrument settings were used and logarithmic amplifiers were applied for forward scatter, side scatter and fluorescence [1]. Fluorescence emission intensities of 50 000 bacteria per sample were recorded and plotted as histograms using WinMDI 2.8 (http://facs.scripps.edu). Induction ratios were calculated by dividing the geometric mean of fluorescence values for treated samples by the geometric mean of the H2 O treated control.
3. Results 3.1. Identification of genes inducible by the LP system Some preliminary experiments were conducted to determine suitable experimental conditions for the isolation of clones from a gfp promoter trap library that are induced by the LP/SCN− system. Parameters that were varied include growth stage and size of inoculum, dosage of enzyme system, medium to grow inoculum, and medium used for challenge with LP system. Since no LP-inducible genes were known at the start of the study, we first screened part of the library using a few different conditions. Two clones that were clearly induced in these experiments (later identified as cysJ– and ydjN–gfp fusions) were selected and subsequently used for further optimalization of the experimental
228
J. Sermon et al. / Research in Microbiology 156 (2005) 225–232
conditions. The results of these optimization experiments are not shown, but the selected experimental conditions are described in Section 2. Using these optimized conditions, 188 out of 11 500 tested clones were identified in a first selection round, showing an increase in fluorescence after treatment with the LP system compared to the water treated control. Retesting of these initially selected clones in triplicate further reduced the selection to seventeen clones showing a reproducible and significant (P < 0.05) increase in fluorescence after treatment with the complete LP system, and not after treatment with water or H2 O2 + KSCN only. Typical results for two clones are shown in Fig. 1 (A and B). Fig. 1C represents a growth curve of one of the clones treated with water, H2 O2 + KSCN only, or with the full LP system, showing that growth was only slightly affected by the LP system under our experimental conditions. Sequencing revealed that 12 out of the 17 selected clones contained a fusion of the promoterless gfp gene with a known E. coli open reading frame (ORF) in the sense orientation. One of these ORFs (sirA) occurred two times. In two clones, the gfp gene was fused directly to a promoter region in sense orientation. In the remaining three clones, there was no correctly oriented promoter or ORF in the vicinity, and these clones were not further taken into account. An overview of the 13 genes identified to be inducible by the LP system is shown in Table 3. 3.2. Induction of selected LP-responsive clones in exponential growth phase by the LP system, H2 O2 and superoxide
Fig. 1. Fluorescence response curves of clones pASV_BTC2 (A) and pASV_BUC3 (B) after induction with the complete LP/KSCN/H2 O2 system (P), with H2 O2 + KSCN only (1) or with water only (E). Growth (C) of strain containing clone pASV_BTC2 treated in the same way. Growth of other clones was similar. OD: optical density measured in a Bioscreen C Automatic Growth analyzer.
For the screening of the promoter trap library above, we used stationary phase cells and measured Gfp production 16 h after treatment with the LP system. Although this procedure resulted in the best sensitivity due to the high cell densities reached and due to the accumulation of Gfp in the cells over a long expression time, as also reported previously [1], it is more conventional to study stress responses in
Table 3 Genes identified to be inducible by the LP system Clone
Gene or promoter upstream of gfp
b no.
Function
pASV_AOD12 pASV_CIC2 pASV_DGD7 pASV_AYF7 pASV_BTC2 pASV_ACB1 pASV_ZD1 pASV_DOD4 pASV_TE7 pASV_UA11 pASV_AHF11 pASV_BUC3 pASV_ATC9 pASV_CYD2
mgtA corA fruB promoter region pstS cysJ rstA fdx yjeQ intB sirA sirA ydjN promoter region ybjG yhdA
b4242 b3816 b2169 b3728 b2764 b1608 b2525 b4161 b4271 b3470 b3470 b1729 b0841 b3252
Mg2+ -transporter: PhoP/Q regulated Mg2+ -transporter Fructose transporter Phosphate transpoter Alfa-subunit of NADPH sulfite reductase Transcriptional regulatory protein, activated by phosphorylated RstB: PhoP/Q regulated Fe–S cluster synthesis GTP-ase Prophage P4 integrase Possible RNA-binding protein Possible RNA-binding protein Possible sodium-dicarboxylate symporter Possible permease, PhoP/Q and EvgA/S regulated Unknown
J. Sermon et al. / Research in Microbiology 156 (2005) 225–232
exponential phase cells and only during the first hours after stress challenge. However, due to the background fluorescence of LB medium and the lower cell densities under such conditions, this type of study could not be successfully conducted with a fluorescence microplate reader. Therefore, we switched to flow cytometry analysis, which allows detection of fluorescence at the single cell level, thus eliminating the problem of low cell densities. Exponential phase cells of the fourteen clones identified to be responsive to the LP system and for which a probable promoter had been identified (Table 3) were analyzed by flow cytometry 4 h after induction with KSCN or the LP system. To investigate the specificity of induction, clones were also induced with H2 O2 and plumbagin. Whereas H2 O2 induces the oxyR-regulon [40], plumbagin is a redox-cycling agent that diverts electrons from the electron transport system to O2 , thereby causing formation of superoxide radicals [10]. In accordance with this mode of action, treatment of bacteria with plumbagin was found to induce the soxRSregulon [27]. Results are shown in Table 4. For ten clones, Table 4 Flow cytometric analysis of LP-inducible clones selected from gfp promoter trap library Clone
Promoter Fold induction after treatment with upstream of gfp H O KSCN LP system Plumbagin 2 2
pASV_BUC3 pASV_BTC2 pASV_DGD7 pASV_ATC9 pASV_TE7 pASV_AYF7 pASV_ACB1 pASV_ZD1 pASV_UA11 pASV_CIC2 pASV_AOD12 pASV_AHF11 pASV_CYD2 pASV_DOD4
ydjN cysJ fruB ybjG intB pstS rstA fdx sirA corA mgtA sirA yhdA yjeQ
1.1 1.0 1.1 1.2 1.0 1.0 1.2 1.2 1.0 0.9 1.0 1.0 1.0 1.0
1.0 1.0 1.0 1.2 1.0 1.0 1.2 1.1 1.0 0.9 1.0 1.0 1.0 1.0
7.6 7.3 4.8 3.3 3.2 2.8 2.5 2.5 2.4 2.2 1.9 1.5 1.4 1.2
1.0 1.1 1.8 1.2 1.2 1.2 1.3 1.3 1.4 1.2 0.9 1.1 1.1 1.0
229
LP induction ratios higher than 2 were obtained. H2 O2 or KSCN at the same concentration as used in the LP enzyme system did not induce any of the LP inducible clones (induction ratios 1.2), indicating that the observed inductions are due to the reaction products of the LP enzyme system. Treatment with plumbagin, caused weak induction of the fruB promoter carrying clone pASV_DGD7 (induction ratio 1.8), but not of the other clones (induction ratio 1.4). 3.3. Further study of the specificity of the LP stress response using gfp fusions of selected PCR-cloned promoters From five clones showing LP-specific induction, the putative LP-inducible promoters were cloned as gfp fusions in pFPV25 for further analysis. The selection included the fusions with ydjN, cysJ, fruB and ybjG because these clones showed the highest LP induction, and also the fdx fusion because of the possible link between Fe–S containing proteins and oxidative stress. Upstream of ydjN a known SOS response gene, ydjM is located [11]. Since we could not exclude the possibility that ydjN induction would result from read-through activity of the ydjM promoter, we constructed separate gfp-fusions for both promoters. The induction of the six cloned promoters, and of six known stress-responsive promoters (also cloned in pFPV25 [5]) by H2 O2 , the complete LP system and plumbagin were then studied by flow cytometry analysis (Table 5). The calculated induction ratios after LP treatment were high for the cysJ (14.0) and ydjN (12.1) promoters, moderate for the fruB promoter (3.1), and low for the ybjG (1.9), ydjM (1.9) and fdx (1.5) promoters. None of the promoters was induced by H2 O2 (induction ratios 1.2) or plumbagin (induction ratios 1.3). In general, and although there are quantitative differences, these results confirm the qualitative induction pattern obtained with the clones from the promoter trap library (Table 4), except perhaps for the weak plumbagin induction of the fruB containing clone which could not be reproduced with the cloned fruB promoter.
Table 5 Flow cytometric analysis of transcriptional fusions of cloned promoters Plasmid
Promoter
Size of cloned promoter fragmenta
Fold induction after treatment with H2 O2
KSCN
LP system
Plumbagin
pJS115 pJS117 pJS118 pJS119 pJS116 pJS120
cysJ ydjN fruB ybjG ydjM fdx
From –584 to +45 From –497 to +33 From –364 to +33 From –386 to +96 From –410 to +51 From –2857 to +62
1.0 1.0 1.2 1.1 1.2 1.1
1.0 1.0 0.9 1.0 1.1 1.0
14.0 12.1 3.1 1.9 1.9 1.5
1.1 1.2 1.3 0.9 1.1 1.0
pFPV25PsodA pFPV25PrecA pFPV25PdnaK pFPV25PkatE pFPV25PgrxA pFPV25Pzwf
sodA recA dnaK katE grxA zwf
From –345 to +35 From –279 to –1 From –221 to +4 From –217 to +45 From –316 to +51 From –276 to +21
1.3 1.4 1.1 1.1 3.9 1.2
1.0 1.0 1.1 1.0 1.0 1.0
4.4 3.9 3.5 2.6 1.6 1.3
8.5 1.1 0.9 1.1 9.8 8.0
a Relative distance to translation start.
230
J. Sermon et al. / Research in Microbiology 156 (2005) 225–232
Of the six tested known stress-inducible promoters, those of sodA (4.4), recA (3.9), dnaK (3.5) and katE (2.6) were moderately induced by the LP system, and grxA (1.6) weakly. None of the promoters was induced by KSCN. As expected, the grxA promoter was induced by H2 O2 (3.9), and the sodA (8.5) and zwf (8.0) promoters were induced by plumbagin. Unexpectedly, we also observed induction (9.8) of the grxA promoter by plumbagin.
4. Discussion Using DFI, we have identified a set of thirteen ORFs that are specifically induced by oxidative stress caused by the LP/SCN− enzyme system. Interestingly, none of these has been previously found to be inducible by H2 O2 or superoxide (see for example gene array studies by [29,50]), and our own experiments with these compounds confirmed this. There is also no correspondence with the genes involved in nitrosative stress responses [16,28]. These results indicate that the LP/SCN− system elicits a distinct stress response in E. coli and suggest that the type of oxidative damage caused by this system is different from that caused by H2 O2 or superoxide. Besides these newly identified genes, we also found several previously known stress genes to be induced by the LP system. At least some of the gene products induced by H2 O2 or superoxide are involved in repair of the damage caused by oxidation (e.g., GrxA, glutaredoxin 1; the DNA repair endonuclease IV), degradation of the oxidants (e.g., KatG, a catalase; SodA, a superoxide dismutase) or the production of oxidation resistant cell components (e.g., AcnA, a superoxide resistant isozyme of aconitase; FumC, a superoxide resistant isozyme of fumarase) [39]. Assuming a similar situation for the LP stress response, analysis of the function of the LP-induced genes identified in this work may provide novel insight into the nature of the cellular stress induced by this system. A first observation in this respect is the induction of recA and dnaK, key genes of the SOS and the heat shock regulon, respectively. The SOS response is induced by DNA damage as caused for example by exposure to UV or to mutagenic chemicals. Oxidative DNA damage is usually ascribed to OH• radicals produced from H2 O2 and Fe2+ in the Fenton reaction, but it is not known if and how the LP system might promote this reaction, for example by enhancing the endogenous formation of H2 O2 or by solubilizing Fe2+ that is bound to cellular proteins. Although weak (induction factor 1.9), the upregulation of the LexA-dependent ydjM promoter [11] may be an additional indication of SOS induction by the LP system. DnaK is primarily known for its role, together with other chaperones, in assisting protein folding and export, but interestingly, it has also been proposed to prevent oxidation of some cellular proteins by acting as a protective molecular shield [7]. Induction of cellular protein damage was also suggested for the Curvularia haloperoxidase enzyme system, based on a DNA microarray study
showing upregulation of a putative chaperone-encoding gene (cpxP) and two other genes that are known to be induced by protein damage (ibpA and ibpB) [15]. One of the most strongly induced promoters in our study is from cysJ, the first gene of the cysJIH operon. Together with CysI, CysJ forms the NADPH sulfite reductase which catalyzes a key step in the production of sulphydryl (–SH) compounds such as cysteine, glutathione and coenzyme A. Induction of this operon may therefore reflect depletion of the extracellular and/or intracellular sulfhydryl resources due to oxidation by the OSCN− formed by the LP system, and consequently the need of the cell to boost its own sulfhydryl production. The antagonistic effect of extracellularly added cysteine against the antibacterial action of the LP system has been known for some time: preincubation of cell suspensions with cysteine leads to an increased cellular sulfhydryl content and not only provides buffering capacity to the oxidative action of the LP system, but also allows reversion of oxidative damage by converting disulfides, sulfenic acids and sulfenyl thiocyanates back to sulfhydryls [41,42]. The activation of the sulfite reduction pathway thus seems to be a dedicated protective response to LP challenge. In line with this view is the finding that a NADPH-dependent cystine reductase may account for increased resistance against the LP system in Streptococcus agalactiae [24]. Several more recent studies have already implicated the cysteine biosynthesis pathway in response to, or resistance to H2 O2 and superoxide. Mutants defective in CysJ were found to be resistant to paraquat, because CysJ functions as a paraquat reductase and consequently acts as the source of O− 2 in vivo [14]. cysK, coding for a O-acetylserine (thiol)-lyase A, was shown to be transcriptionally upregulated after challenge with H2 O2 or O− 2 [29,50]. Although the transcriptional upregulation could not be verified with lacZ-fusions [23], a cysK knockout mutant was hypersensitive towards paraquat, indicating its relevance in resistance to O− 2 stress [50]. Taken together, all these findings indicate a role of the cysteine biosynthesis pathway in oxidative stress and in the response to LP stress in particular. Finally, the Curvularia haloperoxidase enzyme system induced several genes involved in both anorganic and organic sulfate assimilation including sbp, a periplasmic sulfate-binding protein, tauA, a periplasmatic taurine transport system protein and cbl, a transcriptional regulator of several members of the cys regulon and of the tauABCD operon. These genes are normally induced in sulfate-limiting conditions [15,45], but interestingly, are also controlled by another regulator, CysB, which also regulates the cysJIH operon found in our study to be inducible by the LP system. Together, these results indicate that the pathways for sulfate assimilation and cysteine biosynthesis are part of a stress response induced by different peroxidase enzyme systems. The other strongly LP-induced promoter was from ydjN. This ORF is located downstream of ydjM, but is itself not preceded by a LexA box, and is probably not cotranscribed with ydjM [11,46]. The differential induction
J. Sermon et al. / Research in Microbiology 156 (2005) 225–232
of both promoters by the LP system seems to confirm this. The hypothetical protein YdjN shows strong homology to the sodium:dicarboxylate symporter family (PFAM 00375; http://www.sanger.ac.uk/Software/Pfam/), and has homologs in other Enterobacteriaceae, but its function is not known. A link to oxidative stress caused by the LP system is not immediately evident, but the strong induction which seems to be specific, since no induction by H2 O2 and superoxide has been reported [29,50], indicates the unique nature of the LP stress. Mutant analysis will be necessary to elucidate the potential role of YdjN in LP sensitivity or resistance. An interesting subset of LP-induced genes includes corA, mgtA, rstA and ybjG. In E. coli and many other bacteria corA encodes a Mg2+ transporter [36,38], recently identified to be a homotetramer [48]. corA is upregulated after UV treatment as well as after heat shock in E. coli [4,34], and may play a role in virulence of E. coli and Salmonella typhimurium. A S. typhimurium corA mutant showed a markedly attenuated virulence after oral administration in mice and, interestingly, was also significantly more heat- and H2 O2 -sensitive than the wild-type strain [18]. Another Mg2+ transporter is encoded by mgtA. In contrast to corA, this gene is induced at low Mg2+ concentrations together with other genes belonging to the Mg2+ stimulon. The histidine kinase sensor encoded by rstA also belongs to this stimulon [25]. While this manuscript was in preparation, a signal transduction cascade between the EvgA/EvgS and PhoP/PhoQ twocomponent systems was demonstrated [8]. Several genes, including mgtA, rstA and ybjG share the same PhoP binding site (consensus sequence (T/G)GTTTAnnnnn(T/G)GTTTA) and are regulated by an interaction between the EvgA/EvgS and PhoP/PhoQ systems upon an unknown environmental stimulus, even at high Mg2+ concentrations. MgtA levels are also highly increased in S. typhymurium upon invasion of epithelial cells but neither the stimulus nor the exact function of this increase are known [37]. Taking these findings together with our results, it might be speculated that the LP system or another peroxidase enzyme system produces the stimulus that activates these genes during host infection and thereby elicits a response resulting in increased resistance to the peroxidase enzyme system. This would also explain the attenuated virulence of a S. typhymurium strain containing a corA knockout. The induction of the fdx gene, although rather weak (Tables 4 and 5), may also be meaningful, because its gene product is part of a complex iron sulfur cluster assembly machinery [17,43] and iron sulfur cluster-containing proteins are important targets of O− 2 and H2 O2 [39]. This result therefore suggests that these proteins may also be targeted by the LP system. In conclusion, this work demonstrates the existence of a specific genetic response in E. coli upon exposure to the antibacterial LP/SCN− system, which differs from the known oxidative stress responses regulated by OxyR and SoxRS and the nitrosative stress responses. Initial analysis of this response reveals a number of pathways which may be involved
231
in antagonizing the toxic effects of the enzyme system and increasing cellular resistance, and which may be involved in virulence in pathogenic strains. Although the screening of the promoter trap library was not exhaustive, these results give a first insight into the specific stress response caused by the LP system and highlight the need for further investigation on LP-specific responses and resistance in bacteria.
Acknowledgements This work was conducted in the framework of research projects financed by the K.U. Leuven Research Fund (OT/01/35) and the Fund for Scientific Research Flanders (F.W.O. G.0195.02). Author R.V.H. is a research assistant of the Fund for Scientific Research Flanders (F.W.O.Vlaanderen).
References [1] A. Aertsen, K. Vanoirbeek, P. De Spiegeleer, J. Sermon, K. Hauben, A. Farewell, T. Nyström, C.W. Michiels, Heat shock protein-mediated resistance to high hydrostatic pressure in E. coli, Appl. Environ. Microbiol. 70 (2004) 2660–2666. [2] T.M. Aune, E.L. Thomas, Accumulation of hypothiocyanite ion during peroxidase-catalyzed oxidation of thiocyanate ion, Eur. J. Biochem. 80 (1977) 209–214. [3] T.M. Aune, E.L. Thomas, Oxidation of protein sulfhydryls by products of peroxidase-catalyzed oxidation of thiocyanate ion, Biochemistry 17 (1978) 1005–1010. [4] J. Courcelle, A. Khodursky, B. Peter, P.O. Brown, P.C. Hanawalt, Comparative gene expression profiles following UV exposure in wildtype and sos-deficient E. coli, Genetics 158 (2001) 41–64. [5] P. De Spiegeleer, J. Sermon, A. Lietaert, A. Aertsen, C.W. Michiels, Source of tryptone in growth medium affects oxidative stress resistance in E. coli, J. Appl. Microbiol. 97 (2004) 124–133. [6] J.N. de Wit, A.C.M. Van Hooydonk, Structure, functions and applications of lactoperoxidase in natural antimicrobial systems, Neth. Milk Dairy J. 50 (1996) 227–244. [7] P. Echave, M.A. Esparza-Cerón, E. Cabiscol, J. Tamarit, J. Ros, J. Membrillo-Hernández, E.C.C. Lin, DnaK dependence of mutant ethanol oxidoreductases evolved for aerobic function and protective role of the chaperone against protein oxidative damage in E. coli, Proc. Natl. Acad. Sci. USA 99 (2002) 4626–4631. [8] Y. Eguchi, T. Okada, S. Minagawa, T. Oshima, H. Mori, K. Yamamoto, A. Ishihama, R. Utsumi, Signal transduction cascade between EvgA/EvgS and PhoP/PhoQ two-component systems of E. coli, J. Bacteriol. 186 (2004) 3006–3014. [9] B. Ekstrand, in: V.M. Dillon, R.G. Board (Eds.), Natural Antimicrobial Systems and Food Preservation, Cab International, Wallingford, 1994, pp. 15–63. [10] S.B. Farr, D.O. Natvig, T. Kogoma, Toxicity and mutagenicity of plumbagin and the induction of a possible new DNA repair pathway in E. coli, J. Bacteriol. 164 (1985) 1309–1316. [11] A.R. Fernández de Henestrosa, T. Ogi, S. Aoyagi, D. Chafin, J.J. Hayes, H. Ohmori, R. Woodgate, Identification of additional genes belonging to the LexA regulon in E. coli, Mol. Microbiol. 35 (2000) 1560–1572. [12] P.G. Furtmüller, W. Jantschko, G. Regelsberger, C. Jakopitsch, J. Arnhold, C. Obinger, Reaction of lactoperoxidase compound I with halides and thiocyanate, Biochemistry 41 (2002) 11895–11900.
232
J. Sermon et al. / Research in Microbiology 156 (2005) 225–232
[13] C. García-Graells, C. Valckx, C.W. Michiels, Inactivation of E. coli and Listeria innocua in milk by combined treatment with high hydrostatic pressure and the lactoperoxidase system, Appl. Environ. Microbiol. 66 (2000) 4173–4179. [14] P. Gaudu, M. Fontecave, The NADPH:sulfite reductase of E. coli is a paraquat reductase, Eur. J. Biochem. 226 (1994) 459–463. [15] E.H. Hansen, M.A. Schembri, P. Klemm, T. Schäfer, S. Molin, L. Gram, Elucidation of the antibacterial mechanism of the Curvularia haloperoxidase system by DNA microarray profiling, Appl. Environ. Microbiol. 70 (2004) 1749–1757. [16] A. Hausladen, A.J. Gow, J.S. Stamler, Nitrosative stress: Metabolic pathway involving the flavohemoglobin, Proc. Natl. Acad. Sci. USA 95 (1998) 14100–14105. [17] Y. Kakuta, T. Horio, Y. Takahashi, K. Fukuyama, Crystal structure of E. coli Fdx, an adrenodoxin-type ferredoxin involved in the assembly of iron–sulfur clusters, Biochemistry 40 (2001) 11007–11012. [18] D.G. Kehres, M.E. Maguire, Structure, properties and regulation of magnesium transport proteins, Biometals 15 (2002) 261–270. [19] C. Kiser, J. Caterina, J.A. Engler, B. Rahemtulla, F. Rahemtulla, Cloning and sequence analysis of the human salivary peroxidaseencoding cDNA, Gene 173 (1996) 261–264. [20] S.J. Klebanoff, Iodination of bacteria: A bactericidal mechanism, J. Exp. Med. 126 (1967) 1063–1078. [21] S.J. Klebanoff, Myeloperoxidase–halide–hydrogen peroxide antibacterial system, J. Bacteriol. 95 (1968) 2131–2138. [22] V.M.E. Marshall, B. Reiter, Comparison of the antibacterial activity of the hypothiocyanite anion towards Streptococcus lactis and E. coli, J. Gen. Microbiol. 120 (1980) 513–516. [23] R.G. Martin, J.L. Rosner, Genomics of the marA/soxS/rob regulon of E. coli: Identification of directly activated promoters by application of molecular genetics and informatics to microarray data, Mol. Microbiol. 44 (2002) 1611–1624. [24] M.N. Mickelson, A.J. Anderson, Cystine antagonism of the antibacterial action of lactoperoxidase-thiocyanate–hydrogen peroxide on S. agalactiae, Appl. Environ. Microbiol. 47 (1984) 338–342. [25] S. Minagawa, H. Ogasawara, A. Kato, K. Yamamoto, Y. Eguchi, T. Oshima, H. Mori, A. Ishihama, R. Utsumi, Identification and molecular characterization of the Mg2+ stimulon in E. coli, J. Bacteriol. 185 (2003) 3696–3702. [26] A.S. Naidu, in: A.S. Naidu (Ed.), Natural Food Antimicrobial Systems, CRC Press, Boca Raton, FL, 2000, pp. 103–132. [27] T. Nunoshiba, E. Hidalgo, C.F. Amábile Cuevas, B. Demple, Twostage control of an oxidative stress regulon: The E. coli SoxR protein triggers redox-inducible expression of the soxS regulatory gene, J. Bacteriol. 174 (1992) 6054–6060. [28] T. Nunoshiba, T. DeRojas-Walker, J.S. Wishnok, S.R. Tannenbaum, B. Demple, Activation by nitric oxide of an oxidative-stress response that defends E. coli against activated macrophages, Proc. Natl. Acad. Sci. USA 90 (1993) 9993–9997. [29] P.J. Pomposiello, M.H.J. Bennik, B. Demple, Genome-wide transcriptional profiling of the E. coli responses to superoxide stress and sodium salicylate, J. Bacteriol. 183 (2001) 3890–3902. [30] K.M. Pruitt, B. Reiter, in: K.M. Pruitt, J. Tenovuo (Eds.), The Lactoperoxidase System, Chemistry and Biological Significance, Dekker, New York, 1985, pp. 143–178. [31] K.M. Pruitt, J. Tenovuo, R.W. Andrews, T. McKane, Lactoperoxidasecatalyzed oxidation of thiocyanate: Polarographic study of the oxidation products, Biochemistry 21 (1982) 562–567.
[32] B. Reiter, J.D. Oram, Bacterial inhibitors in milk and other biological fluids, Nature 216 (1967) 328–330. [33] B. Reiter, J.P. Perraudin, in: J. Everse, M.B. Grisham, K.E. Everse (Eds.), Peroxidases in Chemistry and Biology, CRC Press, Boca Raton, FL, 1990, pp. 143–180. [34] C.S. Richmond, J.D. Glasner, R. Mau, H. Jin, F.R. Blattner, Genome wide expression profiling in E. coli K-12, Nucleic Acids Res. 27 (1999) 3821–3835. [35] K. Shin, H. Hayasawa, B. Lönnerdal, Inhibition of E. coli respiratory enzymes by the lactoperoxidase-hydrogen peroxide-thiocyanate antimicrobial system, J. Appl. Microbiol. 90 (2001) 489–493. [36] R.L. Smith, J.L. Banks, M.D. Snavely, M.E. Maguire, Sequence and topology of the CorA magnesium transport systems of S. typhimurium and E. coli, J. Biol. Chem. 268 (1993) 14071–14080. [37] R.L. Smith, M.T. Kaczmarek, L.M. Kucharski, M.E. Maguire, Magnesium transport in S. typhimurium: Regulation of mgtA and mgtCB during invasion of epithelial and macrophage cells, Microbiology 144 (1998) 1835–1843. [38] R.L. Smith, M.E. Maguire, Distribution of the CorA Mg2+ transport system in Gram-negative bacteria, J. Bacteriol. 177 (1995) 1638–1640. [39] G. Storz, J.A. Imlay, Oxidative stress, Curr. Opin. Microbiol. 2 (1999) 188–194. [40] G. Storz, R. Hengge-Aronis (Eds.), Bacterial Stress Responses, ASM Press, Washington, DC, 2000. [41] E.L. Thomas, T.M. Aune, Susceptibility of E. coli to bactericidal action of lactoperoxidase, peroxide, and iodide or thiocyanate, Antimicrob. Agents. Chemother. 13 (1978) 261–265. [42] E.L. Thomas, K.A. Pera, K.W. Smith, A.K. Chwang, Inhibition of S. mutans by the lactoperoxidase antimicrobial system, Infect. Immun. 39 (1983) 767–778. [43] U. Tokumoto, Y. Takahashi, Genetic analysis of the isc operon in E. coli involved in the biogenesis of cellular iron–sulfur proteins, J. Biochem. 130 (2001) 63–71. [44] R.H. Valdivia, S. Falkow, Bacterial genetics by flow cytometry: Rapid isolation of S. typhimurium acid-inducible promoters by differential fluorescence induction, Mol. Microbiol. 22 (1996) 367–378. [45] J.R. Van der Ploeg, M.A. Weiss, E. Saller, H. Nashimoto, N. Saito, M.A. Kertesz, T. Leisinger, Identification of sulfate starvationregulated genes in E. coli: A gene cluster involved in the utilization of taurine as a sulfur source, J. Bacteriol. 178 (1996) 5438–5446. [46] T.K. Van Dyk, E.J. DeRose, G.E. Gonye, LuxArray, a high-density, genomewide transcription analysis of E. coli using bioluminescent reporter strains, J. Bacteriol. 183 (2001) 5496–5505. [47] I. Van Opstal, S.C.M. Vanmuysen, C.W. Michiels, High sucrose concentration protects E. coli against high pressure inactivation but not against high pressure sensitization to the lactoperoxidase system, Int. J. Food. Microbiol. 88 (2003) 1–9. [48] M.A. Warren, L.M. Kucharski, A. Veenstra, L. Shi, P.F. Grulich, M.E. Maguire, The CorA Mg2+ transporter is a homotetramer, J. Bacteriol. 186 (2004) 4605–4612. [49] C. Wijkstrom-Frei, S. El-Chemaly, R. Ali-Rachedi, C. Gerson, M.A. Cobas, R. Forteza, M. Salathe, G.E. Conner, Lactoperoxidase and human airway host defense, Am. J. Respir. Cell. Mol. Biol. 29 (2003) 206–212. [50] M. Zheng, X. Wang, L.J. Templeton, D.R. Smulski, R.A. LaRossa, G. Storz, DNA microarray-mediated transcriptional profiling of the E. coli response to hydrogen peroxide, J. Bacteriol. 183 (2001) 4562– 4570.
