Contribution of exofacial thiol groups in the reducing activity of Lactococcus lactis D. Michelon 1 , S. Abraham 1 , B. Ebel 1 , J. De Coninck 1 , F. Husson 1 , G. Feron 2 , P. Gervais 1 and R. Cachon 1 1 Laboratoire de Ge ´ nie des Proce ´ de ´ s Microbiologiques et Alimentaires, AgroSup Dijon, Universite ´ de Bourgogne, Dijon, France 2 Unite ´ Mixte de Recherche 1129 FLAVIC ENESAD-INRA-UB, Dijon, France Introduction Electrochemical measurement of the oxidoreduction potential (E h ) in liquid media has been used for eight decades [1]. Recent studies have emphasized that the characterization of redox conditions is of both scien- tific and practical significance in water [2], food prod- ucts [3–6], soils [7] and sediments [8]. In such complex natural ecosystems, both the mechanisms involved in the redox equilibrium and the possible influence of microbial reductive activities on the measured redox potential remain poorly investigated. The most important reaction catalyzed by microbial cells is energy generation from the oxidation of organic substrates and the corresponding dehydrogenation steps in the glycolysis and citric acid cycles that gener- ate reduced electron carriers (NADH, FADH 2 ). The latter are re-oxidized by electron transfer to oxidants (i.e. respiratory metabolism) or to metabolic intermedi- ates (i.e. fermentative metabolism). The use of external oxidants such as electron acceptors and, in some cases, the production of reduced compounds, might explain the reducing capacity (i.e. decrease in E h ) measured in cultures of microorganisms. The implication of microorganisms in the redox equilibrium varies according to whether they are aerobic, facultative anaerobic or obligate anaerobic [9]. Strict aerobic bacteria specifically use oxygen as Keywords Lactococcus lactis; proton motive force; redox; reducing activity; thiol groups Correspondence R. Cachon, Laboratoire de Ge ´ nie des Proce ´ de ´ s Microbiologiques et Alimentaires, AgroSup Dijon, Universite ´ de Bourgogne, site INRA, 17 Rue Sully, 21065 Dijon, France Fax: +33 3 80 69 32 29 Tel: +33 3 80 69 33 73 E-mail: remy.cachon@u-bourgogne.fr (Received 17 July 2009, revised 22 February 2010, accepted 8 March 2010) doi:10.1111/j.1742-4658.2010.07644.x Lactococcus lactis can decrease the redox potential at pH 7 (E h7 ) from 200 to )200 mV in oxygen free Man–Rogosa–Sharpe media. Neither the con- sumption of oxidizing compounds or the release of reducing compounds during lactic acid fermentation were involved in the decrease in E h7 by the bacteria. Thiol groups located on the bacterial cell surface appear to be the main components that are able to establish a greater exchange current between the Pt electrode and the bacteria. After the final E h7 ()200 mV) was reached, only thiol-reactive reagents could restore the initial E h7 value. Inhibition of the proton motive force showed no effect on maintaining the final E h7 value. These results suggest that maintaining the exofacial thiol (–SH) groups in a reduced state does not depend on an active mechanism. Thiol groups appear to be displayed by membrane proteins or cell wall- bound proteins and may participate in protecting cells against oxidative stress. Abbreviations AMdIS, 4-acetamido-4¢-maleimidylstilbene-2,2¢-disulfonic acid, disodium salt; BIAM, N-(biotinoyl)-N ¢-(iodoacetyl)ethylenediamine; CCCP, carbonyl cyanide m-chlorophenyl; DCCD, N,N ¢-dicyclohexylcardiimide; DTNB, 5,5¢-dithiobis(2-nitrobenzoic acid); Dsb, disulfide bond formation protein in Escherichia coli; E 0 ¢, midpoint oxidation reduction potential; E h , redox potential; E h7 , redox potential at pH 7; FNR, transcription factor fumarate nitrate reductase; GSH, glutathione; MRS, Man–Rogosa–Sharpe; NEM, N-ethylmaleimide; PMF, proton motive force; PMSF, phenylmethanesulfonyl fluoride. 2282 FEBS Journal 277 (2010) 2282–2290 ª 2010 The Authors Journal compilation ª 2010 FEBS terminal electron acceptors in respiration, which restricts the range of redox potentials to values close to the oxi- dant values [10]. Anaerobes have higher reducing capacities; they can decrease the E h from )200 to )600 mV [9]. They can reduce external terminal elec- tron acceptors such as NO 3 ) ,SO 4 2) , Mn(III ⁄ IV) and Fe(III) [11] but, in many cases, their reducing capaci- ties can be explained by the production of strongly reducing end-products, such as H 2 (midpoint oxidation reduction potential, E 0 ¢ = )420 mV) [12]. Enterobac- teria such as Escherichia coli can produce H 2 during mixed fermentation, regardless of the pH [13,14]. For this bacterium, it has also been suggested that, under aerobic conditions, other reducing mechanisms might also be involved [15]. Lactic acid bacteria have no respiratory chain and no strong peroxidase activity (catalase ) ), but can partly tolerate oxygen. They obtain most of their energy from lactic acid fermentation; reducing equiva- lents (NADH) produced during glycolysis are used to reduce pyruvate to lactic acid. Among the lactic acid bacteria, some species have low reducing capacities (e.g. Lactobacillus bulga ricus, Streptococcus thermophilus) [3,16], whereas higher reducing species such as Lactococcus lactis are able to decrease the E h7 to )200 mV (E h7 : E h calculated at theoretical pH 7) [3]. L. lactis can eliminate oxygen by water-forming NADH oxidase [17] using NADH produced during glycolysis, thus leading to a decrease in E h . Neverthe- less, oxygen disruption by L. lactis cannot explain the decrease in E h from oxidant to reducing values. Removing oxygen from a liquid media using nitrogen gas does not decrease the E h to reducing values [4,18]. According to the Nernst equation, E h is decreased by 59 mV for one log of oxygen concentration, and it is generally observed that degassing the medium only decreases the E h by 100–150 mV. These results strongly suggest that other mechanisms may be involved in the reducing activity of L. lactis, and that reducing molecules leading to an E h7 of )200 mV may be implicated. This bacterium does not produce hydro- gen or H 2 S, which are the main reducing molecules produced by microorganisms; thus, it is an attractive model for investigating new mechanisms involved in the reducing activity of microorganisms. This was the aim of the present study. Results Redox activity of L. lactis Figure 1 presents the evolution over time of E h7 and pH in Man–Rogosa–Sharpe (MRS) media under anaerobic conditions. The initial pH and E h7 values were respectively 6.5 ± 0.1 and 204 ± 34 mV. E h7 was the first parameter that changed, with a maximal reducing rate equal to )367 mVÆh )1 . After 5.5 h, the reduction stage was finished and the E h7 remained sta- ble at approximately )200 mV until the end of fermen- tation. Acidification began 1 h after the start of the reduction step, with a final pH of 4.7. These results were obtained in oxygen free med- ium. The reducing activity of L. lactis has only been previously reported for cultures in static aerobic batch conditions [3]. We can thus conclude that the aptitude of L. lactis to decrease E h is not dependent on the presence of oxygen. Moreover, the final E h7 values were the same ()200 mV) despite the fact that the culture media (MRS ⁄ milk) had different initial E h7 values (100–240 mV in MRS; 230 mV in milk); consequently, L. lactis reduction is not so much characterized by the amplitude of the decrease in E h7 but rather by the final E h7 . Lastly, a major part of the acidification occurred once the final E h7 had already been reached. During both the acidification stage and after the final pH was reached, the final E h7 was stable. Implication of cell components in the decrease in E h The bacterial cells were removed from the culture media by filtration after the minimal E h7 ()200 mV) was reached (Table 1). For the three stains of L. lactis, the E h7 measured in the filtrate was not significantly different from the initial E h7 of the sterile MRS media. Moreover, maximal care was taken in the experiment to avoid introducing oxygen into the filtrate. Conse- quently, the restoration of the initial E h7 in the filtrate 4.0 4.5 5.0 5.5 6.0 6.5 7.0 –250 –150 –50 50 150 250 01 4 23 567 8 910 pH E h7 (mV) Time (h) (A) (B) Fig. 1. Time course evolution of (A) E h7 and (B) pH during culture of Lactococcus lactis (TIL46) under anaerobic conditions (N 2 ). Experiments were performed in triplicate; average curves are shown (SD for Eh: ± 22 mV; SD for pH: ± 0.1 pH units). D. Michelon et al. Thiol groups in the reducing activity of L. lactis FEBS Journal 277 (2010) 2282–2290 ª 2010 The Authors Journal compilation ª 2010 FEBS 2283 was provoked by removing the bacterial cells from the culture media rather than by the dissolved oxygen. These results led us to assume that the drop in E h7 in the L. lactis culture was mediated by the whole cells, thus demonstrating that the decrease in the redox potential to a low value was not caused by the produc- tion of end-products with reducing metabolisms. L. lactis is able to maintain a low E h7 until the end of fermentation and for 24 h [16]. The role of the active mechanisms in maintaining a low reducing E h7 was thus investigated. Only the effects of compounds that could modify the cell activity were investigated. The activity of L. lactis is dependent on the mainte- nance of the proton motive force (PMF), which is involved in the membrane transport systems. This PMF is composed of a pH gradient and an electro- chemical gradient. Four energetic inhibitors were used to target PMF activity: nigericin (a K + ⁄ H + exchan- ger) and valinomycin (an ionophore) were used together to destroy the PMF; carbonyl cyanide m-chlorophenyl (CCCP) (a protonophore) was used to cancel the pH gradient; and N,N¢-dicyclohexylcardii- mide (DCCD) acts as a specific proton pump F 0 F 1 -ATPase inhibitor. DCCD, CCCP and a nigeri- cin ⁄ valinomycin mixture were added just after the end of reduction (E h7  )200 mV) (Fig. 2). Using such inhibitors, the PMF, ATP synthesis and primary and secondary transport systems collapsed, and the glyco- lytic flux and acidification of the medium by lactic acid synthesis stopped. Despite inhibition of the metabolism and, consequently, the decrease in ATP and NADH levels, the low reducing potential remained stable (Fig. 2). Thiol groups and the decrease in E h The role of thiol groups in the decrease in E h was investigated using thiol-reactive reagents [N-ethylmale- imide (NEM), 4-acetamido-4¢-maleimidylstilbene- 2,2¢-disulfonic acid, disodium salt (AMdiS)]. They contain maleimide, which can bind with thiol groups in an irreversible reaction that may suppress the con- tribution of thiol groups to the redox equilibrium. NEM can diffuse across a cytoplasmic membrane. Consequently, thiol groups on both sides of the mem- brane and in the cytosol are neutralized by this thiol reagent. By contrast, AMdiS can only neutralize acces- sible thiol groups exposed on the external bacterial surface. The addition of NEM or AMdiS to a reducing culture of L. lactis rapidly increased the E h7 and restored the initial E h7 value (Fig. 3). These data show the implication of thiol groups in the decrease in E h mediated by L. lactis and their exofacial localization. As shown in Fig. 4, thiol groups were labeled with a membrane impermeable fluorescent thiol-reactive reagent and observed using an upright fluorescent microscope. The latter procedure allowed us more par- ticularly to visualize the fluorescent rim on the surface of L. lactis, confirming the presence of exofacial thiol groups. Table 1. Effect of filtration on the E h7 value of MRS media reduced by three different strains of Lactococcus lactis. Filtration was carried out at pH 6. E h7 i, redox potential in degassed sterile MRS with (n = 81); E h7 r, redox potential in culture media when the minimal E h7 was reached (for each strain, n = 3); E h7 f, redox potential in filtrate (for each strain, n = 3). Strains E h7 (mV) E h7 iE h7 rE h7 f L. lactis subsp cremoris TIL 46 183 ± 81 b )212 ± 2 a 212 ± 40 a,b L. lactis subsp cremoris SK11 183 ± 81 b )223 ± 3 a 156 ± 20 a,b L. lactis subsp lactis SL03 183 ± 81 b )205 ± 8 a 256 ± 18 a,b a ANOVA (P < 0.05, n = 3), values in a column with the same superscript letter are not significantly different. b ANOVA test (P < 0.05, n = 3), values in a row with the same superscript letter are not significantly different. 3 3.5 4 4.5 5 5.5 6 6.5 7 –250 –200 –150 –100 –50 0 50 100 150 200 250 012345678 pH E h7 (mV) Time (h) (A) (C) (D) (B) Fig. 2. Effect of inhibitors on E h7 and pH during lactic acid fermen- tation by Lactococcus lactis TIL 46 (typical curves). Nigericin and valinomycin mixture or DCCD or CCCP were added when the mini- mal E h7 was reached. Curves (A) and (B) are the pH and E h7 in the experiments with the addition of inhibitor (the arrow indicates the time of addition); curves (C) and (D) are the pH and E h7 in the control experiment. Thiol groups in the reducing activity of L. lactis D. Michelon et al. 2284 FEBS Journal 277 (2010) 2282–2290 ª 2010 The Authors Journal compilation ª 2010 FEBS Evolution of exofacial thiol groups during medium reduction by L. lactis The concentration of accessible exofacial thiol groups was monitored during the growth of L. lactis (Fig. 5A). Before the reduction phenomenon began, the concentration of thiol groups was below 1 lm, and increased to 12 lm at the end of growth. The latter thiol concentration was correlated to the reducing activity. Indeed, a decrease in E h7 was linked to an increase in exofacial thiol groups (Fig. 5A, phase 1) and the E h7 ceased to decrease and remained stable when the maximal amount of thiol groups was reached (Fig. 5A, phase 2). During the reduction phase, the amount of the exofacial thiol groups was correlated with growth, with a value of 7.7 attomolÆcell )1 (Fig. 5B, phase 1). When the final E h7 was reached, growth had not yet finished and the amount of exofa- cial thiol groups per cell decreased to 5.1 attomol Æ cell )1 (Fig. 5B, phase 2). These results confirm that the decrease in E h7 was directly related to exofacial thiol groups. Exofacial protein thiols Bacterial cells of L. lactis were labeled with a biotiny- lated cell impermeable thiol reagent (BIAM) [19,20] targeting only thiol groups on the external face of the membrane. The membrane protein fraction of the sam- ples treated with BIAM only or the samples pre-trea- ted with NEM and then treated with BIAM were loaded (20 lg) onto each lane. After western blotting, BIAM-labeled SH groups were mainly detected in the lanes loaded with BIAM-treated samples only (Fig. 6), in contrast to the lanes loaded with NEM-pre-treated samples. These results confirmed that the thiol groups were located on the external face of L. lactis, on pro- teins. The thiol groups were mainly on cysteine resi- dues in proteins. Consequently, the results obtained suggest that exoproteins (i.e. membrane proteins or cell wall proteins) were involved in the decrease in redox potential. Discussion The capacity of L. lactis to decrease E h to a reducing value is known, although the mechanism involved is not understood [3,16]. An interesting redox property of L. lactis is the final reducing redox value of E h7  )200 mV, regardless of the culture medium and the initial E h7 value [3]. Moreover, this reducing E h7 value remained very stable until the end of fermenta- tion. External reducing E h stability suggests that reversible redox systems might be involved [21]. We showed that thiol-reactive reagents were able to cancel this reducing E h stability under anaerobic conditions; therefore, a thiol–disulfide couple (Eqn 1) is likely to play a major role in maintaining the external reducing –300 –200 –100 0 100 200 300 E h7 (mV) E h7 i E h7 r E h7 f,n,a a a Fig. 3. Effect of thiol-reactive reagents and filtration on the decrease in E h7 by Lactococcus lactis TIL 46 in MRS media. E h7 i = E h7 in sterile MRS media (n = 12); E h7 r = E h7 after reduction, (n = 12); E h7 f, n, a = E h7 after filtration (f) or the addition of NEM (n) or AMdiS (a). Each different treatment (filtration, NEM or AMdiS) was performed in a separate experiment and each experi- ment was carried out in triplicate. ANOVA (P = 0.05) was used for the statistical analysis and significant differences are shown by an ‘a’ above the column. A B Fig. 4. Labeling with Oregon Green â 488 maleimide of surface thiol groups on the bacterial cell surface of Lactococcus lactis TIL 46. Images were realized using bright field (A), or laser excitation at 480 nm (B) to confirm that all the bacterial cells were fluorescent. D. Michelon et al. Thiol groups in the reducing activity of L. lactis FEBS Journal 277 (2010) 2282–2290 ª 2010 The Authors Journal compilation ª 2010 FEBS 2285 E h [glutathione (GSH); E 0 ¢ = )240 mV and cysteine; E 0 ¢ = )340 mV] [22,23]. RSSR+2H þ +2e  $ 2RSH ð1Þ It could be suggested that the E h was stabilized when the concentration of thiol molecules compared to the bacterial cell density was sufficient to establish a more intense current between the thiol–disulfide redox couple and the Pt electrode than other redox couples in the culture medium [24]. The thiol concentration was directly related to the bacterial cell concentration, which means that a minimum cell density around the Pt electrode surface was required for an optimal exchange current between the thiol–disulfide molecules and the Pt electrode. This might explain the same final E h values in different complex media (MRS ⁄ milk). In aerated E. coli and Bacillus subtilis cultures, a sharp decrease in E h during the transition from the active growth phase to the stationary phase was observed and was related to a transitory increase in thiol groups in both the culture medium and on the cell surface [15]. The E h of the periplasm of E. coli AB Fig. 6. Labeling of membrane protein fraction with selective bioti- nylated thiol reagent (BIAM). Twenty micrograms of proteins were loaded in each lane and transferred to a nitrocellulose membrane for western blotting. Lane A, membrane protein fraction from NEM-untreated sample; lane B, membrane protein fraction from NEM-treated sample. 0 2 4 6 8 10 12 14 16 18 20 22 24 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 A B 0312 45678910 Growth (10 8 cells.mL –1 ) Exofacial thiol groups (µ M) Time (h) 6 8 10 12 14 16 18 1 2 E h7 E h7 1 2 0 2 4 0 24 6 8 10 12 14 16 18 20 22 24 Exofacial thiol groups (µ M ) Cell concentration (10 11 cells. L –1 ) y = 0.77x ( r 2 = 0.9896) Fig. 5. Evolution in the concentration of the exofacial thiol groups during reduction by Lactococcus lactis TIL 46. (A) Time course evolution of growth ( ) and time course evolution of concentration of exofacial thiol groups ( ). (B) Evolution according to the amount of exofacial thiol groups per cell during growth of L. lactis. Phase 1 is the reduction phase and phase 2 is the phase when the E h7 is stabilized at )200 mV. Exofacial thiol groups were measured using Ellman’s method. Thiol groups in the reducing activity of L. lactis D. Michelon et al. 2286 FEBS Journal 277 (2010) 2282–2290 ª 2010 The Authors Journal compilation ª 2010 FEBS depends on the presence of thiol–disulfide proteins (Dsb) and GSH, and was maintained at )165 mV [25]. Moreover, the standard redox state (E 0 ¢) of thioredox- in superfamily proteins characterized by two active-site cysteine residues separated by two amino-acids (CX 1 X 2 C) was in the range )125 mV (DsbA) to )270 mV (thioredoxin ⁄ DsbB) [26]. A method based on the protein–protein redox equilibrium enabled the E 0 ¢ of two thiol–disulfide oxidoreductases of E. coli: glut- aredoxin 1 and 3, to be determined ()233 and )198 mV, respectively) [27]. It was strongly suggested that the release of extracellular GSH participates in modulating the thiol–disulfide ratio in the medium and on the cell surface in response to a variation in the intracellular pH [15]. GSH is not synthesized in L. lactis [28] and a decrease in E h is mainly associated with an increase in accessible thiol groups on the cell surface. The results obtained in the present study clearly identify the implication of exofacial thiol groups in the decrease in E h and suggest that these thiol groups are located on proteins (exoproteins: membrane proteins, cell wall-bound proteins). Thiol groups are known to play a central role in protection against oxidative stress and contribute to detoxifying the reactive oxygen spe- cies by reversible thiol oxidation to bound disulfide [29]. One or several proteins might be implicated; for example, an arginine–ornithine antiporter in L. lactis was characterized by reactive exofacial thiol groups displayed on the outer surface of the cytoplasmic membrane [30]. The identification of proteins located on the extracellular surface and involved in the decrease in E h would be of interest for increasing our understanding of the mechanisms involved as well as the reducing activity of L. lactis. A Gram-positive bacterium such as L. lactis has a thick cell wall composed of mainly peptidoglycan and teichoic acids. Proteins present on the external surface are mainly anchored to the cytoplasmic membrane, in contrast to Gram-negative bacteria. Despite these major structural differences, thiol–disulfide oxidoreduc- tases characterized by thioredoxin-like sequence motifs (CXXC) that form the core of the active site [31], and with a similar Dsb function, are present in vegetative forms of B. subtilis. The latter and L. lactis are mem- bers of the same phylogenetic class and the analysis of the L. lactis genome revealed that the conserved thio- redoxin-like motifs are present in numerous ORFs encoding repair or stress response proteins [32]. Homo- log disulfide bond formation proteins in B. subtilis such as BdbB, BdbC or CcdA may also be involved in the display of exofacial thiol groups and their role in decreasing the E h can be implied. Bacteria are able to sense the extra or intracellular environmental redox state with redox sensing mecha- nisms related to the thiol–disulfide balance and adapt their cell activity [29]. Two genes encoding transcrip- tion factor fumarate nitrate reductase (FNR)-like pro- teins (flpA and flpB) with a potential for mediating the dithiol–disulfide regulatory switch, were discovered in L. lactis [33]. In E. coli, the FNR protein plays a major role in altering gene expression under aerobic and anaerobic conditions [29]. Thereby, as demon- strated in Bacillus cereus, FNR-like proteins can act coordinately with another redox response regu- lator such as ResDE, which is composed of a mem- brane sensor and a cytoplasmic regulator [34,35]. Thiols might be used as ligands to coordinate such redox-responsive clusters [29]. In conclusion, the present study has shown that a decrease in anaerobiosis and the reduction phenomenon are not coupled to an accumulation of reducing end-products in the environment or the consumption of oxidizing compounds, as is mainly observed for other bacterial species. The exofacial thiol groups play a cen- tral role in decreasing the E h , and this E h reduction appears to be linked to the density of cells around the Pt electrode. Thiol groups displayed on proteins on the bacterial cell surface could establish a reducing microen- vironment around the cell. Maintaining a low reducing potential was not directly related to metabolic activity, whereas reducing equivalents such as NADH or thio- redoxin are likely to be involved in the formation of exofacial thiol groups during the reducing phase. Materials and methods Chemicals CCCP, DCCD, n-dodecyl-l-maltoside, EDTA, NEM, 5,5¢-dithiobis(2-nitrobenzoic acid) (DTNB), N -acetyl-l-cys- teine, phenylmethylsulfonyl fluoride (PMSF) nigericin and valinomycin were purchased from Sigma (St Quentin Falla- vier, France). Oregon Green Ò 488 maleimide, AMdiS and BIAM were supplied by Invitrogen (Cergy-Pontoise, France). Bacterial strains and culture conditions The L. lactis subsp. cremoris TIL46 derived from L. lactis NCDO763 cured of its 2 kb plasmid (National Collection of Food Bacteria, Shinfield, Reading, UK) and SK11 pro- vided from the CNRZ collection of INRA, as used in the present study, were kindly provided by Dr M. Yvon. The L. lactis subsp. lactis SL03 strain was obtained from the collection of the French Association for Research in the D. Michelon et al. Thiol groups in the reducing activity of L. lactis FEBS Journal 277 (2010) 2282–2290 ª 2010 The Authors Journal compilation ª 2010 FEBS 2287 Dairy Industry (Paris, France). A concentrated stock cell suspension was stored in MRS media supplemented with glycerol (10%) at )80 °C. Cultures were grown in static conditions in MRS media (Difco, Elancourt, France) at 30 °C. The cells were then harvested and concentrated by centrifugation (3500 g for 15 min) and resuspended in buffer 7 (0.1 m potassium phos- phate, pH 7) for inoculation. General methods: L. lactis growth and data acquisition As with L. lactis, there is an oxygen-responsive FNR-like transcriptional regulator [33,36,37], and all experiments were carried out in a specific anaerobic chamber (Bactron I; Sheldon Manufacturing, Cornelius, OR, USA) to prevent oxygen having an effect on the medium’s redox properties and any oxygen-induced oxidative stress. MRS media was prepared as: meat extract (10 gÆL )1 ), yeast extract (5 gÆL )1 ), bacto trypton (10 gÆL )1 ), sodium acetate (5 gÆL )1 ), sodium citrate (2 gÆL )1 ), KH 2 PO 4 (2 gÆL )1 ), MgCl 2 (0.25 gÆL )1 ), MnSO 4 (0.05 gÆL )1 ) and Tween 80 (0.1%). Sterilized lactose was added to MRS media after sterilization (120 °C for 20 min) at a final concentration of 10 gÆL )1 . MRS media was divided into flasks equipped with two sensors; E h and pH were measured with a combined autoclavable redox electrode and a combined autoclavable pH electrode (405- DXK-SC K8S ⁄ 120 and 4805-DXK-S8 ⁄ 120; Mettler Toledo SARL, Paris, France). The next steps were performed in an anaerobic chamber. Oxygen was degassed by nitrogen bub- bling for 12 h. MRS media was inoculated with 1 · 10 7 or 1 · 10 8 cellsÆmL )1 . Data were collected on-line using a multi channel interface (ELIT 8088 multi-channel pH-meter ⁄ redox-meter Computer interface; Bioblock, Ilkirch, France). The measured redox values (E m ) collected via the interface were converted to E h values according to the equation E h = E m + E ref , where E h is the redox potential of the stan- dard hydrogen electrode, E m is the measured potential and E ref is the standard potential of the reference electrode when the latter is different from hydrogen. In our case, the stan- dard potential of the reference Ag ⁄ AgCl electrode at 30 °C (temperature of analysis) is E ref = 204 mV. To correct the Nernstian effect of pH on the E h values, we calculated E h7 values. E h7 values correspond to E h at pH 7. E h7 is calculated by applying the Leitsner and Mirna equation [38]. E h7 = E h –[a(7 – pH)]. An E h -pH correlation factor of a =30mVÆpH U )1 was determined in MRS media at 30 °C. Filtration Filtration and heat treatment were performed at the end of the reduction stage (when the minimal E h7 was reached). The bacterial cultures were filtered using 0.22 lm MILLEX Ò -GV poly(vinylidene fluoride) (Millipore, Carrigtwohill, Ireland) and the redox potential of the filtrate was measured. Syn- thetic membrane filters [poly(vinylidene fluoride)], charac- terized by very low protein absorption, were used and were degassed by three nitrogen injections beforehand. Thiol-reactive reagents and energetic inhibitors L. lactis was cultivated in MRS media. Thiol-reactive reagents (NEM, AMdiS) and inhibitors (DCCD, CCCP, nigericin and valinomycin) were added at the end of the reduction stage. As a control, a stock solution of 1 m NEM was prepared in methanol : water (3 : 1) and degassed, and the equivalent volume of a methanol : water mixture was added. The final concentration of the NEM batch was 25 mm. A 65.2 mm AMdiS stock solution was prepared in water with a final concentration of 9 mm. A 0.2 m DCCD stock solution was prepared in acetonitrile, with a final DCCD batch concentration of 9 mm. A 140 mm CCCP stock solution was prepared in methanol with a final CCCP batch concentration of 152 lm. A nigericin : valinomycin stock solution was prepared in methanol with a concentra- tion of 1.6 and 12 mm, respectively, with a final nigeri- cin:valinomicyn concentration of 10 and 75 lm, respectively. For each experiment, controls were carried out using equivalent volumes of the solution used for diluting the chemical compounds (methanol, acetonitrile and water). Titration of free accessible exofacial thiol groups Exofacial (accessible) thiol groups were measured using Ellman’s method. DTNB is membrane impermeable, and only the thiol groups on the bacterial cell surface can react with the reagent. Cells were collected by centrifugation for 15 min at 3500 g, and they were dislocated with 1 mL of buffer 8 (0.1 m potassium phosphate buffers, pH 8) con- taining 10 lLof6mm DTNB. After 30 min of incubation in the dark at room temperature, the cell suspension was centrifuged for 15 min at 3500 g. The supernatants were filtered through a 0.45 lm filter (Millipore). A 412 of the filtrate was measured and the concentration of accessi- ble free thiol groups was calculated using the N-acetyl-l- cysteine standard curves. The standard curves were in the range 5–60 lm. Fluorescent thiol labeling on the bacterial cell surface The bacterial culture was centrifuged for 15 min at 3500 g and the supernatant was eliminated. Cell pellets were trea- ted with Oregon Green Ò 488 maleimide (40 lm for 30 min at 37 °C in the dark). The cells were then washed three times with buffer 7 and mounted in Fluorsave reagent (Calbiochem, San Diego, CA, USA) to avoid rehydration and to reduce fluorescence decay. The slides were dried Thiol groups in the reducing activity of L. lactis D. Michelon et al. 2288 FEBS Journal 277 (2010) 2282–2290 ª 2010 The Authors Journal compilation ª 2010 FEBS overnight in the dark at room temperature and analysed using bright field or at a wavelength of 480 nm under an upright fluorescent microscope (Axioplan 2i; Carl Zeiss, Jena, Germany). Images were acquired using axiovision 4.8 software and an AxioCam MRm digital camera (Carl Zeiss). Protein extraction and blotting analysis Bacterial cells of L. lactis TIL 46 were produced as previ- ously described. Cells were collected by centrifugation for 15 min at 3500 g at room temperature when the minimal E h7 value was reached (A 600  1.5, pH  6). The cells were resuspended with an A 600  60 in phosphate buffer 7.5 (0.05 m potassium phosphate buffers, pH 7.5). Part of the cells was incubated with 100 mm NEM for 30 min to block all free thiol groups. NEM-treated and NEM-untreated cells were incubated with 0.9 m m BIAM for 30 min at 30 °C. The labeling was stopped by the addition of 200 lm NEM and the cells were washed three times in buffer 7.5. The cells were resuspended with an A 600  120 and treated with 5 mgÆmL )1 of lyzozyme for 1 h at 37 °C. The protop- lasts were then centrifuged for 15 min at 21 000 g and the supernatant was removed. The pellet was resuspended in buffer 7.5 containing 0.2 mm PMSF and glass beads (diam- eter < 50 lm) were added corresponding to 20% (w ⁄ w) of the final mixture. The latter were homogenized with Fast- Prep Ò (6.0 mÆs )1 for 45 s). Subsequently, the homogenate was centrifuged for 1 min at 800 g to eliminate the glass beads. The membranes were pelleted by centrifugation 1 h at 300 000 g and resuspended in buffer [Tris-HCL, 5 mm; EDTA, 20 mm, PMSF, 0.2 mm; n-dodecyl-l-maltoside, 0.5% (w ⁄ v)]. The membranes were solubilized by shaking for 1 h at room temperature and lastly the insoluble mate- rial was removed by centrifugation for 1 h at 21 000 g and the supernatant was stored at )20 °C until electrophoresis. Protein titration was carried out using the Bio-Rad Protein assay (Bio-Rad, Marnes-la-coquette, France). Twenty micrograms of protein from each sample was sub- jected to a short SDS-PAGE using 12.5% polyacrylamide (the samples were boiled for 5 min in Laemmli sample buffer prior to loading on the gel). The proteins were trans- ferred to nitrocellulose membranes (Bio-Rad) using humid electroblotting (Mini Trans-Blot Ò Electrophoretic Transfert Cell; Bio-Rad) and revealed by exposing the blot to avidin- horseradish peroxidase conjugate (Bio-Rad) followed by development with 3,3¢-diaminobenzidine Color Develop- ment Solution (Bio-Rad) and hydrogen peroxide. Statistical analysis Data were analysed using the statistical analysis software Minitab13 Ò (Minitab Inc., State College, PA, USA). Data plots were compared by analysis of variance (anova) with a £ 0.055. Acknowledgements This work was supported by a EUREKA Research grant (R!3562-LABREDOX). We would like to thank Catherine Vergoignan (INRA) for her technical aid as well as all members of the LABREDOX project. References 1 Hewitt LF (1936) Oxidation-Reducing Potentials in Bacteriology and Biochemistry. London County Council, London. 2Ku ¨ sel K & Dorsch T (2000) Effect of supplemental elec- tron donors on the microbial reduction of Fe(III), sul- fate, and CO2 in coal mining–impacted freshwater lake sediments. Microb Ecol 40, 238–249. 3 Cachon R, Jeanson S, Aldarf M & Divies C (2002) Characterisation of lactic starters based on acidification and reduction activities. Le Lait 82, 281–288. 4 Alwazeer D, Delbeau C, Divies C & Cachon R (2003) Use of redox potential modification by gas improves microbial quality, color retention, and ascorbic acid sta- bility of pasteurized orange juice. Int J Food Microbiol 89, 21–29. 5 Abraham S, Cachon R, Colas B, Feron G & De Coninck J (2007) Eh and pH gradients in Camembert cheese during ripening: measurements using micro- electrodes and correlations with texture. Int Dairy J 17, 954–960. 6 Giroux HJ, St-Amant JB, Fustier P, Chapuzet J-M & Britten M (2008) Effect of electroreduction and heat treatments on oxidative degradation of a dairy beverage enriched with polyunsaturated fatty acids. Food Res Int 41, 145–153. 7 Fiedler S, Vepraskas MJ, Richardson JL & Donald LS (2007) Soil redox potential: importance, field measure- ments, and observations. In Advances in Agronomy, pp. 1–54. Academic Press, London. 8 Schulz HD (2000) Redox fundamentals, processes and applications Redox measurement in marine sediments. In Redox Fundamentals, Processes and Applications (Schu ¨ ring J, Schulz HD, Ficher WR, Bo ¨ ttcher J & Duijnisveld WHM eds), pp 235–246. Academic Press Inc, London. 9 Jacob HE (1970) Redox potential. In Methods in Microbiology (Norris JR & Ribbons DW eds), pp 91–123. vol 2, Academic Press, London & New York. 10 Brown M & Emberger O (1980) Oxidation -reduction potential. In Microbial Ecology of Food: Factors Affecting Life and Death of Micro-organisms (Silliker JH, Elliot PP & Baird Parker AC eds), pp 112– 125. Academic Press, New York. 11 Chapelle FH (1993) Groundwater Microbiology and Geochemisrty. Wiley, New York. D. Michelon et al. Thiol groups in the reducing activity of L. lactis FEBS Journal 277 (2010) 2282–2290 ª 2010 The Authors Journal compilation ª 2010 FEBS 2289 12 Chapelle FH, McMahon PB, Dubrovsky NM, Fujii RF, Oaksford ET & Vroblesky DA (1995) Deducing the distribution of terminal electron-accepting processes in hydrologically diverse groundwater sytems. Water Resour 31, 359–371. 13 Kirakosyan G, Bagramyan K & Trchounian A (2004) Redox sensing by Escherichia coli: effects of dithiothreitol, a redox reagent reducing disulphides, on bacterial growth. Biochem Biophys Res Commun 325, 803–806. 14 Bagramyan K, Mnatsakanyan N, Poladian A, Vassilian A & Trchounian A (2002) The roles of hydrogenases 3 and 4, and the F0F1-ATPase, in H2 production by Escherichia coli at alkaline and acidic pH. FEBS Lett 516, 172–178. 15 Obtyabr’skii OD & Smirnova GV (1989) Changes in the redox potential and number of accessible SH groups in cultures of Escherichia coli and Bacillus subtilis during transient processes. Biokhimiia. 53, 2042–2050. 16 Brasca M, Morandi S, Lodi R & Tamburini A (2007) Redox potential to discriminate among species of lactic acid bacteria. J Appl Microbiol 103, 1516–1524. 17 Lopez de Felipe F & Hugenholtz J (2001) Purification and characterisation of the water forming NADH- oxidase from Lactococcus lactis . Int Dairy J 11, 37–44. 18 Ignatova M, Leguerinel I, Guilbot M, Prevost H & Guillou S (2008) Modelling the effect of the redox potential and pH of heating media on Listeria monocyt- ogenes heat resistance. J Appl Microbiol 105, 875–883. 19 Laragione T, Gianazza E, Tonelli R, Bigini P, Mennini T, Casoni F, Massignan T, Bonetto V & Ghezzi P (2006) Regulation of redox-sensitive exofacial protein thiols in CHO cells. Biol Chem 387, 1371–1376. 20 Dominici S, Valentini M, Maellaro E, Del Bello B, Paolicchi A, Lorenzini E, Tongiani R, Comporti M & Pompella A (1999) Redox modulation of cell surface protein thiols in U937 lymphoma cells: the role of gamma-glutamyl transpeptidase-dependent H 2 O 2 production and S-thiolation. Free Radic Biol Med 27, 623–635. 21 Glaster H (2000) Technique of measurement, electrode processes and electrode treatment. In Redox Fundamen- tals, Processes and Applications (Schu ¨ ring J, Schulz HD, Ficher WR, Bo ¨ ttcher J & Duijnisveld WHM eds), pp 13–23. Vol 10, Heidelberg, Springer-verlag Berlin. 22 Schafer FQ & Buettner GR (2001) Redox environment of the cell as viewed through the redox state of the glu- tathione disulfide ⁄ glutathione couple. Free Radic Biol Med 30, 1191–1212. 23 Sober HA (1968) Handbook of Biochemistry, Selected Data for Molecular Biology, The Chemical Rubber Co., Cleveland. 24 Peiffer S, Klemm O, Pecher K & Hollerung R (1992) Redox measurements in aqueous solutions - a theoreti- cal approach to data interpretation, based on electrode kinetics. J Contam Hydrol 10, 1–18. 25 Messens J, Collet JF, Van Belle K, Brosens E, Loris R & Wyns L (2007) The oxidase DsbA folds a protein with a nonconsecutive disulfide. J Biol Chem 282, 31302–31307. 26 Sevier CS & Kaiser CA (2002) Formation and transfer of disulphide bonds in living cells. Nat Rev Mol Cell Biol 3, 836–847. 27 Aslund F, Berndt KD & Holmgren A (1997) Redox potentials of glutaredoxins and other thiol-disulfide oxidoreductases of the thioredoxin superfamily determined by direct protein-protein redox equilibria. J Biol Chem 272, 30780–30786. 28 Li Y, Hugenholtz J, Tjakko A & Molenaar D (2003) Glutathione protects Lactococcus lactis against oxida- tive stress. Appl Environ Microbiol 69, 5739–5745. 29 Green J & Paget MS (2004) Bacterial redox sensors. Nat Rev Microbiol 2, 954–966. 30 Driessen AJ & Konings WN (1990) Reactive exofacial sulfhydyl-groups the arginine-ornithine antiporter of Lactococcus lactis. Biochemica and Biophysica Acta 1015, 87–95. 31 Moller M & Hederstedt L (2006) Role of membrane- bound thiol-disulfide oxidoreductases in endospore- forming bacteria. Antioxid Redox Signal 8, 823– 833. 32 Cesselin B, Ali D, Gratadoux JJ, Gaudu P, Duwat P, Gruss A & El Karoui M (2009) Inactivation of the Lactococcus lactis high affinity phosphate transporter confers oxygen and thiol resistance and alters metal homeostasis. Microbiology 155, 2274–2281. 33 Akyol I & Shearman CA (2008) Regulation of flpA, flpB and rcfA promoters, Lactococcus lactis. Curr Microbiol 57 , 200–205. 34 Duport C, Zigha A, Rosenfeld E & Schmitt P (2006) Control of enterotoxin gene expression in Bacillus cereus F4430 ⁄ 73 involves the redox-sensitive ResDE signal transduction system. J Bacteriol 188, 6640–6651. 35 Zigha A, Rosenfeld E, Schmitt P & Duport C (2007) The redox regulator Fnr is required for fermentative growth and enterotoxin synthesis in Bacillus cereus F4430 ⁄ 73. J Bacteriol 189, 2813–2824. 36 Gostick DO, Griffin HG, Shearman CA, Scott C, Green J, Gasson M & Guest JR (1999) Two operons that encode FNR-like proteins, Lactococcus lactis. Mol Microbiol 31 , 1523–1535. 37 Scott C, Guest JR & Green J (2000) Characterization of the Lactococcus lactis transcription factor FlpA and demonstration of an in vitro switch. Mol Microbiol 35, 1383–1393. 38 Leistner L & Mirna A (1959) Das redoxpotential von Po ¨ kellaken. Die Fleischwirtschaft 8, 659–666. Thiol groups in the reducing activity of L. lactis D. Michelon et al. 2290 FEBS Journal 277 (2010) 2282–2290 ª 2010 The Authors Journal compilation ª 2010 FEBS . course evolution of concentration of exofacial thiol groups ( ). (B) Evolution according to the amount of exofacial thiol groups per cell during growth of L. lactis. . decrease in ATP and NADH levels, the low reducing potential remained stable (Fig. 2). Thiol groups and the decrease in E h The role of thiol groups in the decrease