The Response to Oxidative Stress in Listeria monocytogenes Is Temperature Dependent.

The stress response of 11 strains of Listeria monocytogenes to oxidative stress was studied. The strains included ST1, ST5, ST7, ST6, ST9, ST87, ST199 and ST321 and were isolated from diverse food processing environments (a meat factory, a dairy plant and a seafood company) and sample types (floor, wall, drain, boxes, food products and water machine). Isolates were exposed to two oxidizing agents: 13.8 mM cumene hydroperoxide (CHP) and 100 mM hydrogen peroxide (H2O2) at 10 °C and 37 °C. Temperature affected the oxidative stress response as cells treated at 10 °C survived better than those treated at 37 °C. H2O2 at 37 °C was the condition tested resulting in poorest L. monocytogenes survival. Strains belonging to STs of Lineage I (ST5, ST6, ST87, ST1) were more resistant to oxidative stress than those of Lineage II (ST7, ST9, ST199 and ST321), with the exception of ST7 that showed tolerance to H2O2 at 10 °C. Isolates of each ST5 and ST9 from different food industry origins showed differences in oxidative stress response. The gene expression of two relevant virulence (hly) and stress (clpC) genes was studied in representative isolates in the stressful conditions. hly and clpC were upregulated during oxidative stress at low temperature. Our results indicate that conditions prevalent in food industries may allow L. monocytogenes to develop survival strategies: these include activating molecular mechanisms based on cross protection that can promote virulence, possibly increasing the risk of virulent strains persisting in food processing plants.


Introduction
The bacterium Listeria monocytogenes is ubiquitous, and able to survive and grow at a wide range of temperatures, and in alkaline or acid media and high osmolality conditions [1]. Most of these stressful conditions are common in food processing environments (FPE) and inside the human host during infection [2]. L. monocytogenes is exposed to acid and high osmolality within food matrices (e.g. in dairy products after fermentation or in brine tanks and after addition of food preservatives) [3]. Likewise, gastric acid provides a harsh environment [4].
L. monocytogenes is exposed to diverse stresses in FPEs and during infection. Refrigeration to preserve food products both in production facilities and in consumers' fridges imposes low temperatures, and oxidative stress is caused by sanitizer agents, especially disinfectant application and antibiotic treatments [5,6]. Disinfectants based on quaternary ammonium compounds are the most common bactericidal agents used in the food industry, and chlorine derivates or peracetic acid are also applied to prevent L. monocytogenes spread within facilities [7]. Hydrogen peroxide (H 2 O 2 ) is a non-toxic, hydro-soluble and bacteriostatic or bactericidal agent also commonly used as a disinfectant [8].
Oxidizing agents cause several types of damage in cells, affecting the peptidoglycan wall and cell membrane, denaturing proteins and disrupting nucleic acid structure [6,9]. L. monocytogenes can sense stressful conditions through molecular signalling [10] and activates survival strategies to reduce oxidative damage; these strategies include expression of sigB, cold and heat shock proteins (cspABCD), proteases (clpC, clpP, groEL) and genes related to oxidative response notably superoxide dismutase (sod), perR and catalase (kat) [11][12][13]. sigB acts on genes related to stress (GRS) and virulence genes such as inlA and LIPI-1 [5]. L. monocytogenes virulence can increase under stress conditions: prf A is regulated by a sigB-depedent promoter, and clpC expression influences some genes responsible for adherence [13,14]. This relation between virulence and the stress response illustrates how L. monocytogenes may protect itself in different stressful conditions, being able to survive in environments with multiple stress factors [15].
The first aim of this study was to analyse the effect of oxidizing agents on the growth of L. monocytogenes at optimal and refrigeration temperatures. The second was to study changes in hly and clpC expression to investigate the relationship between virulence and the oxidative stress response. Table 1 collects the information of the eleven strains used in this study that were previously isolated and characterized in Manso et al. and Melero et al. [3,16,17]. The strains of L. monocytogenes belonged to eight sequence types (ST) (ST1 (n = 1), ST5 (n = 2), ST6 (n = 1), ST7 (n = 1), ST9 (n = 3), ST87 (n = 1), ST199 (n = 1) and ST321 (n = 1)). Moreover, they were isolated from three food processing plants: six strains from a poultry meat factory, four from a dairy plant and one from a seafood company. They were found on non-food contact surfaces (n = 6), food contact surfaces (n = 2) and food (n = 3) samples. They were grown on Chromogenic Listeria Agar ISO (Oxoid, United Kingdom) at 37 • C for 48 hours. One single colony from each OCLA plates was streaked onto Tryptone Soya Agar (TSA, Oxoid) plates supplemented with 0.6% yeast extract (YE, Pronadisa, Madrid, Spain) and incubated at 37 • C for 24 hours. A single colony from each plate was used to inoculate 5 mL of Brain Heart Infusion broth (BHI broth) (Oxoid) and incubated statically overnight at 37 • C.

Oxidative Stress Assay
L. monocytogenes strains were grown in RPMI broth medium (1× RPMI-1640 Medium, HyClone ™ and 2.05 mM L-Glutamine, GE Healthcare Life Sciences) [18] supplemented with oxidative agents according to Rea et al. [19] but with some modifications. Approximately 10 9 cfu/mL of the overnight culture was inoculated to 50 mL fresh medium (BHI broth) and incubated until mid-exponential phase (OD 600~0 .8) with shaking. After that, the overnight culture was distributed in 10 mL and centrifuged at 14,600× g for 5 min at room temperature. The bacterial pellets were collected, washed with Ringer solution (Oxoid), and centrifuged again as previously. The pellets were then resuspended in 10 mL of RPMI medium containing 8 mg/mL ferric citrate (Sigma, San Luis, Misuri, USA) and 13.8 mM cumene hydroperoxide (CHP) (Aldrich) [20], or 100 mM hydrogen peroxide (H 2 O 2 ) (VWR Chemicals), or with no added agent (controls). These Listeria cultures were incubated at 10 • C and 37 • C for 4 hours, and two aliquots were taken after 2, 3 and 4 hours for enumeration and RNA extraction. Serial decimal dilutions were streaked onto TSAYE plates and were incubated at 37 • C for 24 hours to calculate the susceptibility of L. monocytogenes strains against oxidative stress conditions. All the experiments were performed in triplicate.

RNA Extraction and Gene Expression
Five out of the eleven L. monocytogenes strains (ST87, ST5 and ST9 -from a meat industry-; ST9 from a dairy plant and the ST321 from a seafood company) were chosen to perform RNA extraction using the RNA Pure Link ™ RNA Mini Kit (Invitrogen, Carlsbad, California, USA) following the manufacturer's recommendations (Table 1). RNA samples were reverse transcribed using the ImProm-II ™ Reverse Transcription System (Promega, USA) as described previously [21]. Resulting cDNAs were diluted 1:20 and used as templates for specific real-time PCR assays as previously described [21] in a StepOne Real-Time PCR System (Applied Biosystems, Foster City, California, USA). Expression of hly (listeriolysin O gene) [22] and clpC (endopeptidase Clp ATP binding chain gene) [23] was studied and ldh (lactate dehydrogenase gene) [21] was used for normalization results following the 2 -∆∆Ct quantification method.

Statistical Analysis
A multifactor analysis of variance was used to determine the correlation between the response to each temperature and oxidizing agents in all L. monocytogenes strains. Fisher's least significant difference (LSD) procedure was used to determine any significant differences (p values < 0.05) amongst the means between the results for the oxidative stress at 37 • C and that at 10 • C. (Stat Graphics Centurion XVI software, Stat Graphics Centurion, Madrid, Spain). Table 1 shows the results of the oxidative stress in the L. monocytogenes strains tested. L. monocytogenes strains, regardless of their origin or genetic background, were significantly (p < 0.05) more tolerant to oxidizing agents (CHP and H 2 O 2 ) at 10 ºC than at 37 ºC. The stress response was also significantly different (p < 0.05) between CHP and H 2 O 2 , and the mean H 2 O 2 effect was significantly higher (p < 0.05) ( Table 1).

Response to Oxidative Stress
The response to oxidative stress differed between strains at the same temperature depending on the food industry origin ( Table 1). The oxidative stress response to CHP at 37 ºC differed between ST5 and ST9 strains depending on the sample types and site of isolation (Table 1) although the differences overall between strains at 37 • C were not significant (p = 1). The ST9 strain isolated from a floor in a meat processing plant was more resistant to CHP at 37 • C during the first hour (reduction of 6.07 log units); however, ST9 strains isolated from a wall in the same meat factory and from cheese crumbs showed higher count reductions (6.75 and 7.23 log unit, respectively) (Table 1). Similarly, the count reduction during the first hour for the ST5 strain isolated from the meat processing plant was lower than that for the ST5 strain isolated from the dairy plant (5.17 vs 5.70 log units) and it continued to survive after 3 h ( Table 1). The ST321 strain from the seafood facility and the ST9 strain from the meat factory wall were not detectable after 2 h of incubation, whereas strains from the meat processing plant, belonging to ST1 (7.88 log unit decline), ST87 (8.20 log unit decline) and ST5 (9 log unit decline), survived for 3 h (Table 1). Similarly, the stress response to CHP at 10 • C was different within ST5 and ST9 strains; the reduction for the isolates from cheese crumbs was lower than those for the isolates from the meat processing plant: 1.93 and 2.24 log unit reduction vs. 2.35 and 2.60 and 3.18 log unit reduction after 3h of incubation, respectively ( Table 1). The LSD Test indicated that the ST9 strain from the wall sample (meat processing) and the ST199 strain were significantly the most susceptible (p = 0.0129) to all the other strains at 10 • C; both were the most susceptible strains at refrigeration temperature. By contrast, ST1 and ST6 strains were the most resistant to CHP at 10 • C (Table 1).
Similar to our observations for CHP, lower temperature moderated the effect of the oxidative stress; L. monocytogenes strains were more tolerant to H 2 O 2 at 10 • C than at 37 • C regardless the origin or genetic background of the strains. The oxidative stress response in L. monocytogenes to H 2 O 2 at 37 • C was higher than to CHP. No colonies were found after just 1 hour of incubation in H 2 O 2 at 37 • C, except for the ST5 strain from the dairy plant and ST87 −7.26 and 6.67 log unit declines, respectively-(Data not shown). However, after incubation at 10 • C for 3 hours, H 2 O 2 was less toxic than CHP for the L. monocytogenes strains: count reductions were between 1.20 (ST87) and > 9.30 (ST9) log units (Table 1). Figure 1 shows analysis of hly and clpC expression under oxidative stress conditions: hly expression was upregulated by oxidative stress (in both CHP and H 2 O 2 ), and clpC was downregulated. ST9 isolated from meat and ST321 only expressed hly and clpC for the first hour in CHP at 37 • C, and there was a tendency for hly downregulation ( Figure 1A).

Gene Expression in Oxidative Stress Conditions
Strains incubated in CHP showed higher hly expression at 10 • C than at 37 • C ( Figure 1B). By contrast, clpC was downregulated in all strains tested during exposure to CHP regardless of the temperature (37 • C or 10 • C) ( Figure 1A,B), with the exception of ST87 that showed clpC overexpression after CHP incubation for 3 hours at 10 • C ( Figure 1B). We have also studied the relation between hly and clpC expression during CHP exposure at 10 • C and 37 • C, and it can be observed that hly expression was significantly higher when exposed to CHP at 10 • C the 37 • C, whereas clpC expression was lower under oxidative conditions regardless of the temperature ( Figure 1C).  It was not possible to analyse the gene expression of hly and clpC in L. monocytogenes strains treated with H 2 O 2 at 37 • C, because L. monocytogenes counts were below the detection limit in less than one hour. Only ST87 and ST321 survived exposure to H 2 O 2 at 10 • C: hly was upregulated and clpC was downregulated after 3 h ( Figure 1D). The relation between hly and clpC expression during H 2 O 2 exposure at 10 • C and 37 • C was also studied ( Figure 1E): only ST87 and ST321 were able to survive these oxidative conditions, but strains treated with H 2 O 2 showed hly upregulation, while clpC was downregulated although its expression was slightly higher at 10 • C.

Discussion
L. monocytogenes is a foodborne bacterium commonly found in food processing plants and is able to withstand adverse conditions [1,11]. In food processing environments (FPE), various stressful conditions can influence L. monocytogenes growth and survival, especially refrigeration temperatures, osmotic, acid and oxidative stresses [24,25]. L. monocytogenes is also exposed to stressful conditions in hosts and some are common to FPE stresses: gastric acids provide acid and both invasion of phagolysosomes or macrophages and antibiotic treatments can cause oxidative stress [26,27]. L. monocytogenes is able to increase its tolerance to stressful conditions over time following repeated exposure to sub-lethal doses.
We report here that oxidative stress is temperature and oxidizing compound dependent: the effect was significantly lower at lower temperatures (10 • C vs 37 • C), and for CHP than H 2 O 2 . The role of temperature in oxidative stress has also been studied by other authors who reported that lower temperature increased the response to oxidative stress and that there was similar damage to nucleic acids and cell membranes in both stressful conditions [28,29]. In general, detergents and disinfectants (H 2 O 2 , paracetic acid and QAC compounds such as NaOCl, NH 4 OH 4 ) used in food industries are applied at refrigeration temperature; this may favour the development in L. monocytogenes of resistance to these sanitizer agents over long period time [24].
Our comparison of H 2 O 2 and CHP confirms previous reports showing that H 2 O 2 is a more effective listericidal agent [27,30]. The presence of molecular oxygen, growth phase and serovar may all affect the response to oxidative conditions [26,31]. From the point of phenotyping strains results, we found that genotypes belonging to STs of Lineage I (ST5, ST6, ST87, ST1) were more resistant to oxidative stress than those of Lineage II (ST7, ST9, ST199 and ST321), with the exception of ST7 that showed tolerance to H 2 O 2 at 10 • C, however the differences between strains were not statistically significant. This pattern has been observed previously: L. monocytogenes serovar 1/2a (Lineage II) is more sensitive than 4b strains (Lineage I) to 0.6 % H 2 O 2 [24,32].
The differences between lineages may be due to difference in transcription of genes regulating and encoding oxidative responses. L. monocytogenes expresses molecular mechanisms based on stress regulator genes (sigB, ctsR, hrcA, lexA or recA) and response genes (fri, kat, perR or sod) against oxidative stress [28,32]. Most of the genes involved in stress responses in L. monocytogenes are regulated by sigB factor and they include ctsR, that is the clp operon repressor during optimal conditions. Clp family proteins (chaperones and proteases) are generally influenced by temperature [33] and stressful conditions. Likewise, clpC is also implicated in the responses to oxidative or high osmolality stresses and iron starvation [13,34]. However, kat is frequently considered the most relevant gene oxidative stress response together with sod, even at low temperature as Azizoglu & Kathariou (2010) [35] described how kat mutant strains showed smaller colonies size and less tolerance to refrigeration or freeze temperatures. In addition, the response to H 2 O 2 by kat could be interfered by the enzymatic reaction from food products [36]. Wherefore, the present study was focused on the expression of clpC as representative gene in response to oxidative stress combined with different temperature incubation. It is well known that L. monocytogenes virulence is found in island LIPI-1, regulated by prf A [37]. Listeriolyin O is encoded by hly and its expression could be modified during exposure to range of temperature, osmotic and oxidative environmental conditions [38,39]. However, some studies supported the connection between stress conditions with virulence due to the relation among sigB and prf A, as prf A has three significant promotors dependent of sigA and sigB [14]. This inter-genetic relation could explain why stressed L. monocytogenes strains could increase their virulence, although the reaction against the stress could be different depending on possible prf A promotor sequence [40].
This study reported that clpC was overexpressed in some L. monocytogenes strains at 37 • C in the presence of both of oxidizing agents and its expression was downregulated in H 2 O 2 at 10 • C; these findings implicate clpC in the responses to oxidative and heat stresses. Similar results were described by Ochiai et al. [28]. The relationship between stress exposure and virulence in L. monocytogenes has been studied previously. Van der Veen and Abee. [13] reported that clpC mutant strains (∆clpC) can survive inside of macrophages and other host cells; Chastanet et al. [33] found that clpP mutants were unable to grow intracellularly; and the promotors of prf A (pPrf A 1 and pPrf A 2 ) and sigB (sigA and sigB) are intrinsically regulated [34].
In conclusion, this study describes for the first time the effect of two different oxidizing agents at two temperatures (optimal growth temperature and the refrigeration temperature in food industries) at the same time on different genotypes of L. monocytogenes. The oxidative effect is temperature dependent, being lower at 10 • C than 37 • C. The virulence LIPI-1 genes were more strongly expressed when oxidative agents were applied at refrigeration temperatures.

Conflicts of Interest:
The authors declare no conflict of interest.