NG mutagenesis and isolation of Rif mutants’ active producers of hydrogen peroxide
The selection of mutants with increased production of H2O2 is difficult to overcome due to the phenomenon of "auto-inhibition", therefore, mutants were tested for H2O2 activity before the beginning of auto-inhibition. The experiment showed that the NG mutagenesis increased the frequency and the diversity of pleiotropic rif mutations, which in turn increased the yield of H2O2. Out of three independent experiments, among 150 rifampicin-resistant cultures assessed per hydrogen peroxide synthesis, 3.3 % was found to accumulate relatively high amount of hydrogen peroxide, while among the same amount of rifampicin sensitive NG-treated cultures only 0.67% was active. Thus, the frequency of occurrence of active peroxide producers among Rif mutants was about five times higher than among cultures sensitive to rifampicin.
The activity of selected Rif mutants was examined and confirmed also by microbiological agar disk diffusion method (Fig. 1).
Relatively larger inhibitory zones on the lawn of the pathogen E. coli O157:H7 around the disks soaked in culture liquids of active Rif mutants indicate a high antibacterial activity of the mutants compared to the parental strain (in the center). For further study, the mutants numerated on the plate were designated as follows: 19 - Rif1, 31 - Rif2, 33 - Rif 3, 14 - Rif 4, and 21 - Rif5.
Growth behavior and hydrogen peroxide accumulation by Rif mutants
To study the growth behavior of L. delbrueckii Rif mutants, LAPTg medium was inoculated 1:10 with overnight Rif cultures and incubated at 37 °C under anaerobic and aerobic (agitation 200 rpm) conditions. Biomass (OD600) and H2O2 accumulation was assessed every 30 min (Figures 2 and 3).
The specific growth rate (0.8 ± 0.1 h-1) of the cultures under aerobic and anaerobic conditions, does not differ significantly from one to another, but aerated cultures enter the stationary phase significantly earlier than anaerobic ones (Fig. 2). Aerobic cultures release hydrogen peroxide into the environment, which leads to premature growth inhibition, which occurs at different times of growth depending on the rate of H2O2 accumulation. After autoinhibition, the populations of Rif mutants, as well as the parental strain remain constant for the entire time. As compared to the parental strain, all Rif mutants, except of Rif-3, eventually accumulated approximately the same amount of H2O2, but by a 20-50% lower biomass, as a result of an earlier transition to the stationary phase. When the level of H2O2 reaches to approximately 120 mg/l, the growth, and thus the synthesis of H2O2 gradually cease. Due to auto-inhibition, the maximum level of H2O2 accumulation in all cultures was approximately the same (p ≥ 0.05) (Fig. 3). But in mid logarithmic phase the comparison of the H2O2 concentration revealed significant (p < 0.05) differences between the mutants as well as the parental strain; the Rif mutants produced 30–50% more hydrogen peroxide than the wild strain.
The Rif-3 mutant accumulates the most biomass and H2O2 (P<0,05). In an anaerobically growing culture, neither accumulation of hydrogen peroxide, nor auto-inhibition was observed.
The effect of growth phase and aeration on H2O2 tolerance of L. delbrueckii
In contrast to the H2O2 MIC obtained earlier for non-aerated L. delbrueckii, the extracellular concentration of H2O2 causing auto-inhibition of aerobically growing cultures was significantly lower. Hence, the relative H2O2 MICs of the cultures depending on the growth phase and aeration was studied (Fig. 4).
The MIC of hydrogen peroxide for anaerobically growing L. delbrueckii in stationary and exponentially phases differ from each other. Cells in log phase were about twice more sensitive to hydrogen peroxide than cells in the stationary phase. In turn, significantly less amount of H2O2 was needed to arrest the growth (auto-inhibition) of aerobically growing cultures.
Hydrogen peroxide production by Rif mutants in LAPTg at 5 °C
Since lactobacilli are widely used for food preservation at refrigeration temperature, the behavior of L. delbrueckii mutants at 5°C was studied. In the most of studies the production of hydrogen peroxide by LAB cells was examined in phosphate buffer (Amézquita and Brashears 2002, Ruby et al. 2009). But the formation of hydrogen peroxide must also be studied in complex media, since food, especially raw meat, is rich in nutrients.
Harvested L. delbrueckii cells were inoculated into fresh LAPTg at ~108 CFU/ml and incubated at 5°C. The rate of accumulation of hydrogen peroxide, determined per every other day for five days, is shown in Figure 5.
The accumulation of hydrogen peroxide by wild and mutant strains gradually increased and reached its maximum on the fourth day (Fig. 5) and remained stable for at least ten days of cold incubation (data are not presented in the figure). During the entire period of cold exposure of the cultures, despite the active production of H2O2, no significant changes in the pH and in the live cell population was observed. With the same biomass without the growth Rif mutants have produced significantly more hydrogen peroxide in LAPTg than the wild strain (p <0.05). Despite the high amount of extracellular H2O2 accumulated in the cold, it does not lead to death of host. Interestingly in cold condition the Rif3 mutant lost its superiority over other mutants and became closer to the parental strain.
Antagonism of Rif 4 mutant against E. coli O157:H7 at 5°C
The antagonistic effect of most active Rif-3 mutant against the pathogen E. coli O157:H7 was determined at 5°C co-cultivation at ratios of 10:1, 100:1 and 1000:1 respectively. Data of E. coli O157:H7 survival determined on MacConkey agar are shown in Figure 6.
As can be seen from Figure 6, there is a direct relationship between the number of lactobacilli and the death of E. coli O157:H7. At a ratio of 1000:1, Rif-4 practically eliminated the pathogen (p < 0.05) within 4 days of co-cultivation. At a ratio of 100:1, complete elimination occurred on the 6th day, which is not far behind 1000:1. Whereas, at a ratio of 10:1, the E. coli population decreased only by 1.6 log in 6 days. The optimal ratio of lactic acid bacteria to pathogen is most likely in the range of 1000:1 to 100:1.
On the first day of co-cultivation, when the H2O2 concentration (see fig. 4) had not yet reached to the MIC of E. coli (~ 34 mg/ml), a significant decrease in E. coli viable cell count was observed depending of lactobacilli concentration (Fig5). According to our earlier data (Pashayan and Hovhannisyan 2021), death at sub-inhibitory concentrations of H2O2 can be caused by L. delbrueckii co-aggregation with the pathogen and direct release of H2O2 on it. Further experiments revealed 39 ± 1.5% level of co-aggregation between L. delbrueckii and E. coli.