The bacterial strains used in this study are summarized in Table 5. E. coli strains were grown at 37°C in TY 23. Lactococcus lactis was grown in M17 at 30°C and L. plantarum WCFS1 was grown at 37°C in de Man-Rogosa-Sharp (MRS) or in Chemically Defined Medium (CDM) 24.

All molecular biology techniques were performed following established protocols by Sambrook 23. DNA was digested according to the conditions recommended by the commercial suppliers of the restriction enzymes (Boehringer, Breda, The Netherlands). In all cases, DNA was eluted from 0.7% agarose gels using the Purification Kits from Promega (Leiden, The Netherlands). For a Polymerase chain reaction (PCR) we used: 1 μl template DNA (10 to 100 ng), 2 μl of each primer combination (50 ng/ml), 1 μl dNTP's (100 nM), 1 μl Pwo-polymerase (5 U/μl) and 10 μl polymerase buffer (5×) (Roche, Woerden, The Netherlands). The reaction mixtures were adjusted to 50 μl with deionized H₂O.

Previously it has been established that the nisin controlled expression system can be functionally implemented in L. plantarum, by chromosomal integration of the regulatory module encoding genes nisRK 25. However, the strain described by Pavan et al. harbors an erythromycin resistance marker. Therefore, a chromosomal lp_0076:::nisRK gene replacement without resistance marker was constructed. For this purpose, lp_0076 upstream and downstream flanking regions were amplified by proofreading PCR using genomic DNA of L. plantarum WCFS1 as template. The primers used to amplify the 5'- and 3'-regions of lp_0075 and lp_0077 respectively were lp_0075-FORW and lp_0075-REV; lp_0077-FORW and lp_0077-REV, shown in Table 6. The PCR amplicons obtained were digested with EcorI-BamhI, and BamhI-XbaI (sites introduced in the primers used are underlined in Table 6) and cloned into a EcoRI-XbaI digested pUC18Em 26. The genetic organization of the resulting plasmid was verified by PCR and was designated pNZ7130. To introduce additional endonuclease restriction sites in pNZ7130 the multiple cloning sites containing fragment was amplified by PCR using the primers EM1 and EM2 (Table 6). The resulting amplicon (35 bp) was digested with BamhI and BglII and cloned in BamhI digested pNZ7130, yielding pNZ7131. The 2.6 kb HindII fragment of pNZ7131 that contains the lp_0076 flanking regions separated by the multiple cloning site and the erythromycin resistance encoding gene, was subcloned in vector pNZ84, yielding the lp_0076 replacement vector pNZ7132. The lp_0076::nisRK replacement plasmid pNZ7133 was constructed by cloning of the 2.4 kb HpaII-PstI fragment of pNZ9521 27 into the NsaI-NsiI digested pNZ7132. The plasmid pNZ7133 was introduced into L. plantarum WCFS1 and primary integrants were selected on basis of erythromycin resistance. The anticipated configuration of the pNZ7133 plasmid integration in the lp_0076 locus was confirmed by PCR and Southern blotting (data not shown). One of the integrants was cultured for 140 generations without antibiotic selection. Subsequently, erythromycin sensitive (Em^S) colonies were identified by plating on media without erythromycin followed by replication plating on erythromycin containing plates. In the Em^S colonies identified, a candidate lp_0076::nisRK replacement mutant was identified by PCR using internal primers in NisK and Orfx (data not shown). In a selected candidate lp_0076::nisRK strain, the anticipated genetic organization of the lp_0076::nisRK locus was further confirmed by additional PCR and Southern blot analyses (data not shown). The lp_0076::nisRK derivative of L. plantarum WCFS1 was designated NZ7100.

In the experiments with the nisin inducible promoter we used as control strain NZ7606, a L. plantarum NZ1700 strain containing the pNZ8150 vector. On the other hand, for the chemostat studies strain NZ7607 was used as control. Strain NZ7607 is L. plantarum WCFS1 containing the pNZ7021 vector.

Nisin induction was done as previously described by Pavan et al 25 using 50 ng/ml nisin for induction.

Total RNA was isolated from exponentially growing L. plantarum WCFS1 cultures with the High Pure RNA Isolation Kit (Roche, Woerden, The Netherlands). To eliminate genomic DNA contamination a 60 min DNAse I treatment was included in the isolation procedure. Furthermore, the quality and quantity of the RNA was confirmed using the RNA 6000 Nano Assay (Agilent Technologies, Amstelveen, The Netherlands) using the Agilent 2100 expert Bioanalyzer. Cultures of the bacterium were grown in anaerobic jars on CDM containing 5 mM Diamide at 37°C, or CDM containing 5 mM DTT at 37°C and 30°C. In all cases 0.5% (w/v) glucose was added as carbon source. For the reverse transcription reaction 200 ng RNA was incubated at 65°C for 5 minutes with 270 ng random nonamers, and 1 μl of 10 mM dNTP mixture. After 5 minutes on ice, the following was added: 5 μl 5 × First strand Buffer, 2 μl 0.1 M DTT, 1 μl Superscript III (200 UNITS), 1 μl RNASEout (40 UNITS) (all Invitrogen) and water to a final volume of 20 μl. The reaction was incubated at 25°C for 5 min and then at 50°C for 60 min. The reaction was inactivated by heating at 70°C for 15 min. Generated cDNA samples were stored at -20°C until use. Quantitative PCR amplification was performed in 96-well plate on a 7500 Fast System (Applied Biosystems), using SYBR green for product detection. Each well contained 10 μl SYBR green Master Mix (Applied Biosystems), 200 nM of Reverse and forward primers, and 1 μl of 10-fold or 100-fold diluted RT product as template. The qPCR gene-specific probes of L. plantarum WCFS1 of trxA1 (0.3 kb), trxA2 (0.3 kb), trxA3 (0.3 kb), trxH (0.3 kb), trxB1 (0.3 kb), and trxB2 (0.6 kb) were amplified using specific primer combinations trxA1-FORW and trxA1-REV; trxA2-FORW and trxA2-REV; trxA3-FORW and trxA3-REV; trxH-FORW and trxH-REV; trxB1-FORW and trxB1-300; trxB2-truncFORW and trxB2-truncREV. All primers are specified in Table 6. Amplification was initiated at 95°C for 10 min, followed by 40 cycles of 95°C for 15 sec, 53°C for 30 sec and 60°C for 60 sec for probes trxA1, trxA3 and trxH. For the other probes (trxA2, trxB1, and trxB2) the step of 53°C for 30 sec was changed to 55°C for 30 sec. All samples were measured in duplicate. Control PCRs were included to detect background contamination (no-template control) and remaining chromosomal DNA (RT reactions in which enzyme superscript III was not added). PCR specificity and product detection were checked post amplification by examining the dissociation curves of the PCR products. These melting curve profiles were generated by first heating the samples to 95°C and then cooling them to 60°C and slowly heating then at 2°C/min to 95°C for detection of SYBR green fluorescence. In each run, five standards of the gene of interest were included with appropriate dilutions of the cDNA, to determine the cDNA concentration in the samples. All q-PCRs amplified a single product as determined by the melting curve analysis and the releative expression level is given as arbitrary units (Au).

Growth curves. For the growth experiments, 96-well plates were used. Each well was filled with 200 μl medium and 10 μl growing cells (OD₆₀₀ of 1.0). The 96-well plates were inoculated at 37°C and cell density was measured by detecting the turbidity of the cultures at 600 nm every 10 min. To study oxidative stress response, either diamide (5 mM) or hydrogen peroxide (3.5 mM) was added to the medium.

Cultures were grown at 37°C in MRS until OD₆₀₀ of 0.4. At this point, 2.5 ml of culture was plated by mixing with 50 ml of 0.7% Agarose MRS at 40°C. After the agar/culture plates hardened, perforations of 6 mm diameter were made. In each of these apertures, 30 μl of solution (1 M, 0.5 M, 0.25 M and 0.10 M) of diamide or (1 M, 0.5 M) hydrogen peroxide was added. Plates were incubated overnight at 37°C. The growth inhibition towards the oxidative stress agents was measured as the radius of growth inhibition expressed in centimeters.

Cells of a growing culture OD₆₀₀ of 1.0 (approx. 10 ml) were harvested by centrifugation at 12.000 × g for 10 minutes at 4°C. The pellet was washed twice with 1 M Tris-HCl buffer pH = 7.5. Next, the cells were suspended in 1 ml of 1 M Tris-HCl buffer pH = 7.5 and disrupted by Fastprep (Qbiogene Inc., Illkirch, France) in four treatments of 40 seconds at speed 4.0. The suspension was centrifuged 2 min at maximum speed and the supernatant or CFE was transferred to a new vial and used directly. Protein content of the CFE was determined through the standard BCA Protein Assay kit (Pierce, Etten-Leur, The Netherlands).

The reaction mixture (RM) was prepared by mixing (in ice) 200 μl of 1 M Tris-HCl, pH = 7.5, 500 μl of insulin (10 μg/μl), 2.5 μl EDTA pH = 7.5 (0.2M), and 40 μl NADPH (40 mg/ml). For each enzymatic reaction 40 μl RM were used together with 20 μl TRX (60 μM) from E. coli (Sigma), and 80 μl CFE. Each reaction was incubated for 25 minutes at 37°C; after this 500 μl of 6M Guanidine HCl/0.2 M Tris-HCl, pH = 7.5/1 mM DTNB was added. The final absorbance at 412 nm was measured at 25°C. Note that the extinction coefficient of DTNB reduction is 13.6 mM^-1 and that there are two molecules of TNB per molecule DTNB being reduded. The background signal in the assay was determined in the absence of TRX in one of the assays. TR activity was expressed as nmol of DTNB·(min·mg protein)^-1 and calculated as follows: Δabs₄₁₂·(0.620)·(protein content)^-1·2·(εDTNB)^-1·10^{^6}.

The assay mixture contained: triethanolamine-HCl buffer (pH 7.6), 100 mM; ATP, 1 mM; EDTA, 1 mM; MgSO4, 1.5 mM; NADH, 0.15 mM; phosphoglycerate kinase 22.5 U (Boehringer, Breda, The Netherlands); and CFE. The reaction was started with 5 mM 3-phosphoglycerate. The absorption of the assay mixture was monitor at 340 nm (E₃₄₀ nm of reduced pyridine-dinucleotide cofactors is 6.3 mM^-1). Enzyme activity is expressed as μM of amount of NADH converted per (min·mg protein)^-1. All assays were performed with two concentrations of cell extract to confirm that reaction rates were proportional to the amount of cell extract added. Protein content of the CFE was determined through the standard BCA Protein Assay kit.

Cultures were grown at 37°C in CDM 24 which was supplemented with 100 mM glucose and 10 μg/ml of chloramphenicol. A 1 Liter bioreactor (Applikon Dependable Instruments) was inoculated with cells from an overnight culture to an initial OD₆₀₀ of 0.1 in 500 ml medium. A pH of 5.5 was maintained by the addition of 5 M NaOH and the stirrer speed was set at 200 rotations per minute (rpm). The headspace of the fermentors and medium vessel were full at all times with nitrogen gas at a flow rate of 520 ml·min^-1. The cultures were kept at the dilution rate of 0.1 h^-1 and steady state was assumed after 5 volume changes. At steady state or (t₀), samples were taken for transcriptomics, HPLC analysis, dry weight, and enzyme samples. In addition, at steady state, 50 ml of hydrogen peroxide was added to the fermentors resulting in a final concentration of 3.5 mM in the fermentor. After 30 min (t₃₀), samples were taken for transcriptomics, HPCL analysis, and enzymatic analysis.

Determination of regulatory motifs from a list of genes was based on predictions of upstream transcriptional units (TU's) made using an in-house Python program, alignments of these TU's using the MEME 33 and determining of nucleotide sequences using the MAST software 34 tools. The MEME software determines through alignments small motifs which are present in the analyzed sequences and assigns a significant value, E_value, and score per position. This significant value represents how well preserved the motif is, the number of TU's that this motif has, and the conserved position of the motif in the analyzed TU's. The MAST software allows searching for the motif in the genome of interest.