Antibacterial blue light is a promising tool for inactivating Escherichia coli in the food sector due to its low risk of cross-stress tolerance

Abstract

Background

Escherichia coli is an integral part of the colonic microflora, though its pathogenic intestinal strains can contaminate animal and plant products and cause significant challenges in the food industry. Thermal processing is one of the most common methods used to preserve food. Nevertheless, non-thermal antibacterial methods, such as antibacterial blue light (aBL), are attracting more interest due to the growing demand for minimally processed products. Thus, the current study was aimed at assessment whether the risk of co-selection for these two food processing approaches exist.

Results

The development of E. coli tolerance to both selective factors was observed after repeated exposure to sublethal doses of heat and aBL, and the observed adaptations were confirmed to be phenotypically stable. The results demonstrated that populations with increased tolerance to aBL also exhibited increased tolerance to temperature, while the sensitivity of temperature-tolerant populations to aBL did not change. We also identified 11 genes that could be involved in cross-stress tolerance. Neither adaptation changed the antibiotic sensitivity of the tolerant strains. Finally, short- and long-term pre-incubation at elevated temperatures significantly increased the tolerance of E. coli BW25113 to aBL.

Conclusions

The results obtained clearly demonstrate that aBL may serve as a complementary approach in food industry lacking resistance development and exerting no impact on microbial drug susceptibility. Nevertheless, the phenomenon of cross-tolerance should be considered an issue when designing food processing including sequential use of aBL and high temperature.

Graphical Abstract

Introduction

Foodborne diseases are a serious challenge to public health worldwide and are associated with significant morbidity and mortality. These diseases can lead to human suffering, hinder socioeconomic development, burden health systems and harm the national economy, tourism and trade. According to estimates, as many as 600 million people fall ill each year after eating contaminated food, resulting in 420,000 deaths. In low- and middle-income countries, where the problem is the most serious, expenditures related to the consumption of contaminated food reach 110 billion USD annually [1]. Foodborne pathogens are particularly dangerous for immunocompromised individuals, such as infants and small children, pregnant women, elderly people and people with immunodeficiencies. Statistics show that as many as 40% of cases occur in children under 5 years of age, even though these children constitute 9% of the world’s population [1]. One of the most common bacterial pathogen that cause foodborne diseases is E. coli, which is the subject of current work [2].

Pathogenic strains (pathotypes) of E. coli can be found in various food products, including meat and animal products; plant products, such as fruits and vegetables, leaves, seeds and sprouts; and bakery products, such as flour and raw dough [3,4,5]. At each stage of the food production process, microorganisms encounter various selective pressure factors, which may negatively affect their growth and/or survival; however, these factors can lead to mutations that help bacteria adapt to unfavorable environmental conditions.

Thermal stress, which can be induced by high and low temperatures, is widely used in the food industry. High temperatures are used to disinfect food and production lines while low temperatures are most often used to extend the shelf life of food, after which food is still safe to eat [6]. Bacteria in composted food waste may be exposed to temperatures between 40 and 70 °C depending on the stage of the process [7, 8].

As thermal food processing causes undesirable effects, such as changes in the taste, structure and appearance of some products and loss of food ingredients, minimal food processing is preferred; thus, researchers are searching for nonthermal bactericidal agents and methods with potential application in the food industry [9]. Promising modern technologies include high-pressure processing, pulsed electric field, cold plasma, and light-based technologies, such as antimicrobial blue light (aBL), which is the focus of this paper [10]. Many studies have confirmed the effectiveness of aBL in the eradication of numerous foodborne bacterial pathogens, such as Bacillus cereus, Listeria monocytogenes, Staphylococcus aureus, Salmonella enterica, Shigella sonnei, Campylobacter jejuni and E. coli [11,12,13]. Moreover, irradiation does not significantly affect the physicochemical quality of food products, i.e., color, pigment, antioxidant or vitamin content [14]. However, each methods used, or is likely to be used, introduces various stressors to foodborne pathogens, and research on bacterial responses to these alternative stressors is crucial to ensuring food safety. Furthermore, scientists dealing with this topic emphasize that further exploration of the antibacterial mechanism of phototreatment in food industry is needed [15]. The aim of the current work was to assess the risk of co-selection to aBL and thermal stress in E. coli. Co-selection studies help determine if exposure to one type of antimicrobial treatment, such as blue light, could potentially influence the resistance of microorganisms to other treatments like high temperatures. Understanding these dynamics is crucial for ensuring that food safety protocols effectively minimize the risk of microbial contamination and spoilage. Unfortunately, regular use of certain antimicrobial treatments can lead to tolerance among pathogens. By understanding co-selection mechanisms, food scientists can design treatment protocols that minimize the risk of developing resistant strains, thereby maintaining the efficacy of antimicrobial strategies over time. Finally, by preventing the spread of foodborne illnesses through effective antimicrobial strategies, the food industry plays a critical role in safeguarding public health. Co-selection studies are a key component in understanding how to best achieve this goal.

Materials and methods

Strains and culture conditions

Experiments were performed using E. coli BW25113 obtained from the Keio collection [16]. All of its variants were stored at − 80 °C with 20% glycerol and cultured in Luria–Bertani (LB) broth (BTL, Łódź, Poland) or LB agar medium (A&A, Gdansk, Poland) at 37 °C under aerobic conditions. Overnight cultures were prepared in an orbital incubator (Innova 40, Brunswick, Germany) at 150 rpm for 16–20 h.

In the analysis on the risk of tolerance development, and feature stability confirmation, the aBL and temperature-tolerant populations (day 10) were inoculated into LB media directly from frozen glycerol stocks for characterization. In experiments performed to evaluate the aBL and temperature sensitivity of E. coli (day 0), overnight cultures were inoculated from single colonies grown on LA plates using glycerol stocks frozen at − 80 °C.

Sixty-four aBL-hypersensitive mutants from the Keio collection were cultured in the presence of 15 µg/ml kanamycin [16]. The strains were stored in 96-deep-well plates filled with LB, and 15% glycerol medium was used for storage at − 80 °C. Before the cells were used, they were freshly stamped into new microtiter plates filled with LB medium and incubated overnight (16–20 h) in an orbital incubator at 150 rpm.

Chemicals

Trimethoprim, ciprofloxacin, tigecycline, cefuroxime, ceftazidime, and piperacillin antibiotics were purchased from Sigma Aldrich (Darmstadt, Germany). Gentamycin and ampicillin were purchased from Carl Roth (Karlsruhe, Germany), while meropenem was purchased from Thermo Fisher Scientific (UK). Kanamycin sulfate was purchased from Gibco (Paisley, UK). All stock solutions at a concentration of 10 mg/l were stored at − 20 °C.

Light source

Irradiation was performed using an LED light source emitting blue light (λ_max 415 nm, irradiance 25 mW/cm²; Cezos, Gdynia, Poland) [17].

Determination of the temperature leading to E. coli BW25113 tolerance development

Overnight cultures in triplicate were adjusted to an optical density (OD) of 0.5 McF (approx. 5 × 10⁷ CFU/ml). Aliquots (150 µl) were transferred to 1.5 ml tubes and incubated at 52 °C for 90 min in a TB-941 T/TB-941 U incubator (JW Electronic, Warsaw, Poland). Every 5 min, 10 µl aliquots were serially diluted and streaked horizontally on LB agar plates. After overnight incubation at 37 °C, the colonies were counted to estimate the survival rate and compared to that of the untreated cultures. For tolerance development analysis, a temperature dose leading to a decrease in the survival rate of 1.2 log₁₀ CFU/ml was chosen (20 min of incubation at 52 °C).

Determination of the aBL dose leading to E. coli BW25113 tolerance development

The aBL dose used for tolerance analysis was determined according to our previously published work [18].

Determination of tolerance development following repeated sublethal exposure to aBL and temperature

Three biological replicates of aBL tolerance development were prepared exactly as described previously [17] with 10 days of sublethal exposure of 100 µl of overnight cultures to aBL (32.4 J/cm²).

Temperature-tolerant strains were prepared as follows: overnight cultures of E. coli BW25113 in triplicate were diluted to an OD of 0.5 McF. Then, 1.5 ml aliquots were transferred to 1.5 ml Eppendorf tubes and incubated at 52 °C for 20 min. Following exposure, 10 μl aliquots of the incubated samples were collected to determine the survival rate. Sample aliquots of 50 μl were transferred to fresh LB medium (5 ml) for regrowth overnight. The next day, after 16–20 h of incubation, the treatment was repeated under the same conditions. The cycle of exposure—regrowth—exposure was repeated 10 times. Simultaneously, control samples were prepared in the same manner but without exposure to temperature or aBL.

Potential reductions in susceptibility to temperature and aBL were examined after the 5th and 10th consecutive cycles at higher doses of light (up to 86.4 J/cm²) and compared to those of the control strain (day 0) and (day 10th).

Stability of the acquired tolerance to aBL or temperature

The experiments were performed using the samples that were taken from the 10th consecutive cycle of aBL or temperature treatment (the cycle in which a significant decrease in susceptibility was observed) and transferred to fresh LB medium and cultured overnight. The cycles of transfer—regrowth—transfer were repeated 5 times. After the 5th cycle, the cultures were diluted to an OD of 0.5 McF, and 100 μl of the bacterial suspensions were irradiated with 415 nm light at a dose up to 86.4 J/cm², or 150 µl of bacterial suspension was incubated at 52 °C for up to 90 min. The resulting suspensions were compared with the initial samples and with the untreated controls.

aBL and temperature sensitivity of E. coli BW25113

Analysis of aBL sensitivity was performed as described previously [18]. Different light doses ranging from 0 to 86.4 J/cm² were tested. To test temperature sensitivity, overnight cultures of temperature-tolerant populations (day 10) and controls (days 0 and 10) were diluted to an OD of 0.5 McF. A total of 150 µl of bacterial suspension was incubated at 50 °C, 52 °C, 54 °C, and 56 °C for up to 90 min, and 10 µl was collected every 5 min, diluted and streaked horizontally onto LB-agar plates to assess growth reduction. The experiments were performed in triplicate.

Assessment of the impact of short-term pre-incubation at increased temperatures on the sensitivity of E. coli BW25113 to aBL

Three overnight cultures of the E. coli BW25113 strain were diluted to an OD of 0.5 McF. Then, 350 µl aliquots were pre-incubated at 50 °C or 52 °C for 10 or 20 min in 1.5 ml Eppendorf tubes. Next, 100 µl of each sample was transferred to a 96-well plate, which was subsequently irradiated to assess the aBL sensitivity, diluted and incubated overnight at 37 °C, after which the surviving colonies were counted.

Assessment of the impact of long-term (overnight) incubation at increased temperatures on the sensitivity of E. coli BW25113 to aBL

Three cultures of E. coli BW25113 were incubated overnight in ThermoMixer C (Eppendorf, Hamburg, Germany) at 37 °C or 42 °C. Then, aBL sensitivity profiles were investigated as described previously.

Assessment of the sensitivity of 64-aBL-hypersensitive E. coli BW25113 single-gene mutants to increased temperature

Overnight cultures of single gene mutants maintained with kanamycin (15 mg/ml) and the wild-type strain were diluted 1:100 in LB media. The OD₆₀₀ of the initial cultures was measured using an EnVision Multilabel Plate Reader (PerkinElmer, USA). Then, the mutants were incubated for 16 h at 37 °C (control) or 40 °C in an orbital incubator (Innova 42, Brunswick, Germany) at 150 rpm, and the OD₆₀₀ was measured after 16 h. A heat shock temperature of 40 °C, which causes a greater than 50% reduction in the growth of the wild-type strain, was investigated. To determine growth defects, the ΔOD₆₀₀ was calculated (OD_{600, 16 h}–OD_{600, 0 h}). The sensitivity of each mutant was calculated according to the following formula described by Krewing et al. [19]:

$${\text{Grown}}\;{\text{defect}}\;\left[ \% \right] = 100 - 100 \times {\text{OD}}_{{{\text{wt}}}} \times {\text{OD}}^{ - 1}_{{{\text{wt}},\;{\text{stressed}}}} \times \left( {{\text{OD}}_{{{\text{mutant}}}} \times {\text{OD}}^{ - 1}_{{{\text{mutant}},\;{\text{stressed}}}} } \right)^{ - 1} .$$

Antimicrobial susceptibility testing of aBL and temperature-tolerant strains

The MICs of trimethoprim, ciprofloxacin, tigecycline, cefuroxime, ceftazidime, piperacillin, gentamycin and meropenem were tested by the microbroth dilution method according to the European Committee for Antimicrobial Susceptibility Testing [20] for the temperature-tolerant strain (day 10), aBL-tolerant strain (day 10) and the control (day 0) and control (day 10) strains.

Bioinformatics and statistical analysis

All the statistical analyses and figures were created using GraphPad Prism version 9.0 (GraphPad Software, Inc., CA, USA). The significant differences between the groups were calculated using two-way analysis of variance (ANOVA) with P < 0.05 and Tukey’s or Dunnett’s multiple comparison tests. The graphic figures were prepared with the use of BioRender.com (accessed on 31 July 2024).

The experimental workflow is presented in Fig. 1.

Fig. 1

The experimental workflow

Results

Establishment of MDK₉₉ conditions for assessing tolerance

The MDK₉₉ parameter should be determined to perform the tolerance study [21, 22]. This parameter demonstrates that the treatment conditions resulted in a reduction in cell viability of < 2 log₁₀. The determination of treatment doses leading to a 99% reduction in microbial viability must be determined to ensure a reliable experimental design, as sufficient number of cells of the treated bacterial population must remain alive for further analysis. After 20 min of incubation at 52 °C, a 1.2 CFU/ml reduction in the bacterial survival rate was observed. For aBL tolerance development, a light dose of 32.4 J/cm² was chosen, as described in our previous studies, and this dose also causes a 1.2 CFU/ml reduction in bacterial growth [18].

Phenotypically stable temperature and aBL tolerance were observed after 5 cycles of selective pressure

Compared to the control passaged without selective pressure, significant differences in the bacterial survival rate were observed starting on the 5th day of consecutive bacterial treatment. An increase in bacterial survival of up to 3.06 log₁₀ CFU/ml was observed for the temperature treatment. Consecutive aBL pressure increased bacterial survival up to 2.74 log₁₀ CFU/ml. This effect was also observed on the 10th day of bacterial treatment for both temperature and aBL treatment, causing an increase of up to 3.24 log₁₀ CFU/ml in the survival rate (Fig. 2A, C).

Fig. 2

Temperature and aBL tolerance and stability. Response of E. coli populations to 52 °C. The survival rates of the control populations (days 0 and 10) and temperature-tolerant populations (day 5 and 10) were assessed (A). The stability of temperature tolerance after 5 passages without selective pressure (B). Response of E. coli populations to aBL. The survival rates of the control populations (days 0 and 10) and aBL-tolerant populations (day 5 and 10) were assessed (C). Stability of aBL tolerance after 5 passages without selective pressure (D). The plots present the reduction in log₁₀ units of CFU/ml. The experiment was performed in biological triplicates. Statistical comparison between the tolerant population (day 10) and the control (day 10) (A, C) and between the tolerant population 10th day and passaged strains (B, D). Significance at the respective P values is marked with asterisks (ns P > 0.05; *P < 0.05; **P < 0.01; ***P < 0.001; ****P ≤ 0.0001)

The stability of the acquired adaptations was assessed by passaging strains from the 10th day of factor treatment for 5 subsequent cycles without selective pressure. This step was performed to assess whether the observed adaptations result from genetic alterations caused by repeated exposure to sublethal stressor doses or are a phenotypic change caused by temporary changes in the gene expression profile, which return to the original state after selective pressure is removed. After 5 cycles of incubation without selective factors, the sensitivity of the tolerant populations to temperature and aBL was investigated and compared to that of the tolerant populations from the 10th day of subsequent treatments. The results are presented in Fig. 2B, D. For the temperature test, a statistically significant reduction in bacterial survival of up to 0.55 log₁₀ CFU/ml was observed in the passaged population without selective pressure in the 5th cycle compared to the initial culture from the 10th day. This may result from changes in gene expression profiles in response to a lack of selective pressure. However, the statistical significance remained the same for the difference between the tested population (day 10 + 5 passages) and the control (initial culture, day 10), suggesting that the acquired tolerance is stable.

No significant differences were observed between aBL (day 10) and aBL (day 10 + 5) (Fig. 2D), suggesting that the developed tolerance is caused mainly by stable genetic alterations due to multiple sublethal aBL irradiation cycles.

Compared to the WT strain, the 52 °C temperature-tolerant strain is less sensitive to higher temperatures

After the E. coli population treated with a sublethal temperature was found to exhibit reduced sensitivity to heat stress, further studies were carried out to assess changes in sensitivity to different temperatures, i.e., 53 °C, 54 °C and 56 °C. The population from day 10 was used in this analysis because its tolerance level was constant and the accumulation of genetic changes resulting from prolonged selection pressure was potentially greater than that of other populations. The experimental results are presented in Fig. 3.

Fig. 3

Temperature sensitivity profile of temperature-tolerant E. coli. A 53 °C, B 54 °C, C 56 °C. The plots present the reduction in log₁₀ units of CFU/ml. The experiment was performed in biological triplicates. Statistical analysis between the temperature-tolerant strain (day 10) and the control strain (day 0) (upper asterisks) and between the temperature-tolerant strain (day 10) and the control strain (day 10) (lower asterisks). Significance at the respective P values is marked with asterisks (ns P > 0.05; *P < 0.05; **P < 0.01; ***P < 0.001; ****P ≤ 0.0001)

Compared to both control groups, the 52 °C-tolerant population showed significantly increased tolerance to all other tested temperatures (53 °C, 54 °C and 56 °C) (survival increased by ≤ 2.37, 2.18 and 3.05 log₁₀, respectively) (Figs. 3A, 5B, C).

The temperature-tolerant population does not exhibit increased tolerance to aBL

The obtained E. coli populations with stable temperature and aBL tolerance were used to assess the risk of co-selection. The first experiment was performed to determine the survival rate of the temperature-tolerant population under aBL stress conditions. The experimental results were compared with the results of the passaged control (day 10) and unpassaged control (day 0) for comparative analysis and are presented in Fig. 4A.

Fig. 4

Co-selection studies. aBL sensitivity profile of the temperature-tolerant population and controls (A). Statistical comparison between the temperature-tolerant population and the control strain (day 10). Heat sensitivity profiles of the control (day 0), control (day 10) and aBL (day 10). B 52 °C, C 53 °C, D 56 °C. Statistical analysis between aBL (day 10th) and control (day 0)—upper asterisk and aBL (day 10th) and control (day 10th)—lower asterisk

In Fig. 4A, the aBL sensitivity profiles of controls and temperature-tolerant population are compared. A statistically significant decrease in aBL sensitivity was observed for the temperature-tolerant population in comparison to the passaged control population with light doses of 21.6, 36, 43.2, 64.8, and 79.2 J/cm². This nonstable pattern of aBL sensitivity suggests that the temperature tolerance and aBL sensitivity are not correlated.

The aBL-tolerant population exhibits increased tolerance to elevated temperature

When higher temperatures (52 °C, 53 °C and 56 °C) were applied, significant differences in the average survival of the studied populations were observed. At all tested temperatures, the aBL-tolerant population exhibited greater mean survival rates than that of the passaged control group at some time points (survival increases of ≤ 0.97, 1.22, 0.64, and 1.01 log₁₀) (Fig. 4B–D). However, it is important to emphasize that the temperature sensitivity of the aBL-tolerant strain did not reach the levels observed in the temperature-tolerant population.

The results obtained suggest that a co-selection phenomenon occurs in which acquired tolerance to aBL may confer a certain degree of cross-stress tolerance to temperature stress. Importantly, although co-selection occurs, the resulting temperature tolerance is significantly lower than that of populations selected for this feature.

The plots present the reduction in log₁₀ units of CFU/ml. The experiment was performed in biological triplicates. Significance at the respective P values is marked with asterisks (ns P > 0.05; *P < 0.05; **P < 0.01; ***P < 0.001; ****P ≤ 0.0001). Statistical comparison between the temperature-tolerant population and the control population (day 10).

Short-term pre-incubation at 50 °C increased E. coli tolerance to aBL

This experiment was performed to evaluate whether short-term thermal pre-incubation affects E. coli sensitivity to blue light treatment. In this study, E. coli were exposed to 50 °C or 52 °C for 10 or 20 min and then examined for their sensitivity to aBL in comparison to an untreated control (Fig. 5).

Fig. 5

aBL sensitivity of temperature-pretreated bacteria. A 50 °C, B 52 °C. Statistical comparison between untreated and temperature-treated bacteria. The plots present the reduction in log₁₀ units of CFU/ml. The experiment was performed in biological triplicates. Significance at the respective P values is marked with asterisks (ns P > 0.05; *P < 0.05; **P < 0.01; ***P < 0.001; ****P ≤ 0.0001). The upper asterisk indicates 10 min of pretreatment, while the lower asterisk indicates 20 min of pretreatment

Under all conditions tested, populations subjected to short-term pre-incubation exhibited significantly reduced sensitivity to aBL. After pre-incubation at 50 °C, increases in survival of ≤ 2.55 log₁₀ and ≤ 1.50 log₁₀ were observed for exposure times of 10 min and 20 min, respectively. After pre-incubation at 52 °C, increases in survival of ≤ 2.07 log₁₀ and ≤ 1.32 log₁₀ were observed for exposure times of 10 and 20 min, respectively. These results indicate that at both temperatures, a shorter pre-incubation time (10 min) resulted in a greater increase in survival during light treatment, and the highest tolerance to aBL exposure was observed for the bacterial population exposed to a lower temperature (50 °C) for 10 min.

Long-term incubation at 42 °C results in increased E. coli tolerance to aBL

The next experiment was performed to investigate whether long-term (overnight) culturing of E. coli at an elevated temperature (42 °C) affects its sensitivity to aBL. The aBL sensitivity profiles of bacteria cultured at 37 °C (optimal) and 42 °C were compared according to the results presented in Fig. 6. The population cultured at higher temperatures was less sensitive to aBL at light doses ranging from 36 to 64.8 J/cm² than the population grown at standard temperatures (survival increase of 0.58 to 1.19 log₁₀). This observation confirms that the temperature significantly impacts the response of the tested bacterium to aBL treatment.

Fig. 6

aBL sensitivity profiles of bacteria cultured at 37 °C and 42 °C. The plots present the reduction in log₁₀ units of CFU/ml. The experiment was performed in biological triplicates. Significance at the respective P values is marked with asterisks (ns P > 0.05; *P < 0.05; **P < 0.01; ***P < 0.001; ****P ≤ 0.0001)

Temperature and aBL adaptation mechanisms do not significantly affect the susceptibility of E. coli to antibiotics

This step was performed to assess whether the adaption of E. coli to aBL and increased temperature affects the susceptibility of E. coli to antimicrobial agents, which would support the hypothesis that genetic alterations occur during the development of tolerance. The MIC values are presented in Table 1. The most commonly reported change is a twofold increase in the MIC, and the greatest susceptibility (although still insignificant) was observed for ciprofloxacin upon the development of aBL tolerance. Nevertheless, the impact of tolerance development on drug susceptibility is limited because it results in a twofold increase in MICs and no change in the categorization of antimicrobial susceptibility.

Table 1 Antimicrobial susceptibility testing

Cross-tolerance to heat and aBL can be regulated by certain genes

At the end of the studies, aBL-hypersensitive single-gene mutants of E. coli were cultured under heat shock conditions for 16 h; this was performed to assess the role of a single gene in bacterial survival at relatively high temperatures and the co-occurrence of aBL and temperature protection. The growth defect in comparison to the wild-type strain was observed only for 11 mutants lacking: atpE (ATP synthase Fo complex—subunit c), atpG (ATP synthase F1 complex subunit γ), atpH (ATP synthase F1 complex subunit δ), dnaJ (chaperone protein DnaJ), nuoN (NADH:quinone oxidoreductase subunit N), tolA (Tol–Pal system protein TolA), yccM (putative electron transport protein YccM), ydcE/pptA (tautomerase PptA), ydcX (orphan toxin OrtT), yfgL/bamB (outer membrane protein assembly factor BamB), yncA/mnaT (l-amino acid N-acyltransferase), while the most of proteins encoded by these genes are plasma membrane proteins (www.biocyc.org). Proteins encoded by these genes can play an important role in cellular protection against aBL and higher temperatures. For the remaining 53 mutants, growth defects under high temperature were not observed (according to the results presented in Fig. 7).

Fig. 7

Growth defects of single-gene deletion mutants hypersensitive to aBL. The values are the means of five biological repetitions

Discussion

Thermal treatment is the most frequently used method in food processing and is performed to ensure food safety by eliminating pathogens [23]. Different microorganisms may have different sensitivities to heat stress. The term used to determine the thermal resistance of bacteria is the decimal reduction time (D-value), which corresponds to the time needed to reduce 90% of bacterial cells at a given temperature (a 1 log₁₀ decrease in survival rate) [24]. The E. coli strain discussed in this paper is a relatively heat-sensitive bacterium. According to the literature, the D_60°C for E. coli K12 is approximately 0.1–0.3 min [25]; however, the development of tolerance may lead to a multiple increase in this value. For instance, highly heat-resistant E. coli strains isolated from a beef processing facility were characterized by D_60°C values ranging from 15 to even 71 min [26]. Although much research has been conducted to understand the mechanisms by which heat treatment inactivates microorganisms, as in the case of aBL, this issue is not fully understood. It is believed that, similar to aBL, thermal stress has a broad spectrum of action and causes damage to many cellular structures; when the accumulation damages pass above a critical threshold, cellular death ultimately occurs [27]. Exposure to elevated temperatures induces significant morphological and structural changes in the outer membrane of gram-negative bacteria and leads to the loss of its lipopolysaccharide (LPS) component [28]. As a consequence, the permeability of the membrane is altered, leading to a loss of intracellular substances and disruption of osmotic and pH homeostasis [29,30,31,32]. Genetic material is also exposed to thermal damage. DNA is a cellular component with relatively high thermal stability, as temperatures close to 100 °C or higher are needed to denature DNA [33]. However, lower temperatures can also lead to damage, which mainly leads to an increased frequency of mutations [34]; these mutations can subsequently result in adaptation to thermal stress.

Recently, innovative approaches aimed at extending the shelf life of food products without the need for thermal treatment have been developed in the food industry. However, adapting these technologies to meet the requirements of sustainable and economic food preservation is challenging [35]. Therefore, light-emitting diode (LED)-based technologies are becoming increasingly desirable as environmentally friendly alternatives for the thermal sterilization of food [36]. The bactericidal effect of UV light is likely the best-tested thus far among these technologies. However, accidental exposure to this light poses serious health risks, leading to serious eye and skin diseases, hair loss, and an increased risk of skin cancer [37]. Therefore, aBL is attracting increased attention as an alternative that is safer for workers in industrial settings. The results of studies in animal models and volunteers suggest that exposure to blue light, at an effective antimicrobial dose, does not cause significant DNA damage to keratinocytes, inflammatory reactions or skin burns [38, 39]. The effectiveness of E. coli eradication using aBL has been demonstrated for nonpathogenic and pathogenic strains, including the seven main Shiga-like toxin-producing E. coli (STEC) pathotypes, which are among the most serious threats in the food industry [40, 41]. Promising in vitro research results have prompted many researchers to test the effects of blue light on pathogens found directly in food products. It has been confirmed that aBL exhibits antimicrobial activity in milk and dairy products, which helps extend the shelf life of products without decreasing nutritional content [42,43,44]. Researchers have also demonstrated that aBL is effective in eliminating microorganisms and maintaining the sensory quality of various fruits and vegetables without significantly changing the composition of these products [10, 45]. Moreover, researchers have shown that LED light at a wavelength of 405 nm can improve the quality of stored strawberries by increasing the antioxidant activity of the enzymes within the fruit [46]. In the storage of cabbage, blue light contributed to increasing the levels of chlorophyll, polyphenols and vitamin C, which shows the potential of aBL to positively impact the appearance and nutritional value of food [47]. Furthermore, the effectiveness of aBL against pathogenic bacteria has also been demonstrated on materials such as stainless steel and polyethylene. These materials are often used in the food industry, serving as work surfaces and food packaging [48]. Less satisfactory results were obtained when the impact of aBL on meat, fish and seafood was examined. The observed decreases in bacterial survival were often sublethal, i.e., 1–2 log₁₀ CFU/ml, which may result from limited light penetration due to the opacity and irregular surface of these products [12, 49,50,51]. This result indicates that aBL should be optimized for industrial applications to increase its effectiveness [49,50,51]. Photodynamic treatment opens up new possibilities for maintaining cleanliness and safety in the food industry. Nevertheless, bacteria may develop tolerance or resistance to many stress factors in addition to those used in food processing. In the environment, bacteria are exposed to different stressors (e.g., antibiotics, acids, organic solvents, heavy metals, and oxidative stress). Different environmental contaminants provide selective pressure for bacteria to mutate, evolve, and develop mechanisms to tolerate and resist such stressors [52]. It is essential to determine the factors that contribute the emergence of bacterial adaptation in environmental reservoirs, and studies of the co-selection phenomenon to other factors and/or to therapeutic approaches should also be performed. For instance, many studies indicate that bacteria gain cross-resistance to antibiotics, disinfectants, and heavy metals due to their antimicrobial properties [53,54,55,56]. Associations between specific patterns of antibiotic resistance and the levels or types of metal contamination suggest that several mechanisms underlie the co-selection process. These phenomena increase bacterial persistence and resistance and should be considered when developing strategies to reduce antimicrobial resistance [57]. Gram-negative bacteria, including E. coli, can adapt to many physical, chemical and environmental stressors, such as UV radiation [58], ionizing radiation [59], organic solvents [60], heavy metals [61], acids [62], antibiotics [63], aBL and aPDI [18, 64]. Ramteke [65] demonstrated that 90% of 448 coliform isolates were resistant to one or more antibiotics but showed tolerance to multiple metals. Rowe and Kirk [66] investigated the phenomenon of cross-protection in E. coli O157:H7 and found that salt or heat tolerance was increased when the bacteria were prestressed with acid, indicating that this procedure could affect food processing. Other studies have shown that adaptation to increased temperature often leads to a reduction in the sensitivity of microorganisms to other stressors, and vice versa. For example, Isohanni et al. [67] observed that Arcobacter butzleri (a pathogenic bacterium found in food) exhibited significantly reduced sensitivity to acidic conditions (pH 4.0) after a 2-h incubation at 48 °C. In addition, experiments conducted by Liao et al. [68] showed that exposure of some Staphylococcus aureus isolates from food to organic acids increases their tolerance to heat treatment. Bacterial co-selection for heat stress can be induced by factors other than acidic environments, such as nutrient deficiency, high salinity and ethanol. Jenkins et al. [69] showed that tolerance to heat (57 °C) increased in E. coli populations deprived of access to glucose or nitrogen. Similarly, Pumirat et al. [70] reported that the growth of Burkholderia pseudomallei on media rich in NaCl (150 and 300 mmol/l) significantly increased the survival of these bacteria when they were exposed to thermal stress (15 min at 50 °C). Yang et al. [71] reported a sevenfold increase in the survival rate of Tetragenococcus halophilus in response to ethanol exposure after prior exposure to high temperature (45 °C for 1.5 h). Moreover, research conducted by St. Denis et al. [72] has shown that the exposure of Escherichia coli to a sublethal dose of aPDI leads to significant changes in the levels of heat stress-related proteins. A sevenfold increase in the heat shock protein GroEL and a threefold increase in DnaK were observed. These results support the possibility of cross-adaptation to thermal stress and therapies based on the photodynamic mechanism, including aBL, which was examined in this study. Due to its antimicrobial properties, aBL could be considered as an advantageous tool in the prevention of foodborne illnesses. If this method were to be introduced to the food industry in the future, there is a risk that bacteria could be exposed to aBL (e.g. during surface disinfection) and then transferred to a food product that will be heat treated. It is also likely that bacteria that have already developed tolerance to heat stress could be exposed to aBL during food processing.

Our previously published studies [18, 64] revealed that sublethal aPDI and aBL treatment leads to tolerance development in Staphylococcus aureus and that sublethal aBL treatment leads to tolerance development in E. coli, Klebsiella pneumoniae and Pseudomonas aeruginosa. The observed results suggest that the perceived accelerated mutation rate results directly from ROS-induced DNA damage, as evidenced in a previously published report [73]. Moreover, in another study, we performed screening analysis using the Keio knockout collection and identified 64 aBL-protective genes that could be involved in the development of tolerance or in E. coli [17]. Analogously, Murata et al. [74, 75] identified 72 thermotolerant genes in E. coli by screening the Keio collection grown at 47 °C. About 60% of mutants lacking genes involved in heat response were also H₂O₂ hypersensitive. In the present study, we cultured the selected aBL-hypersensitive single-gene mutants of E. coli under heat shock conditions to assess the role of a single gene in bacterial survival at high temperatures and risk of the co-selection of aBL and temperature. We observed growth defects in 11 of the 64 mutants lacking atpE, atpG, atpH (encoding ATP synthase subunits), dnaJ (encoding a chaperone protein), nuoN, tolA, yccM, ydcX, yfgL/bamB (encoding plasma membrane proteins), ydcE/pptA and yncA/mnaT (enzymes). The presence of ATPase subunits and NADH: quinone oxidoreductase subunit N involved in ATP synthesis and coupled electron transport among the proteins encoded by deleted genes from mutants identified as co-sensitive to heat and aBL suggests that ATP is required in these two types of stress. One hypothesis could be that during stress all cellular repairs need energy, and the second hypothesis could be that due to the disruption of the cell membrane ATP is exposed to leakage and the lack of fully functional systems to restore ATP deficiencies causes greater sensitivity of bacteria to stress. Modification of the tRNA with a sulfur relay system encoded by yccM seems to be important in protection against heat and aBL. Murata et al. [75] confirmed that complementation of an appropriate single-gene mutant with yccM restored the insensitive phenotype, confirming the role of selected gene in the response to heat stress. Five of eleven genes encode proteins localized in the plasma membrane. In study by Ruiz et al. [76], it was observed that mutations in yfgL alter outer membrane permeability, which may contribute to greater sensitivity to membrane stress generated by heat and aBL treatments. TolA is also essential for outer membrane stability [77].

Therefore, proteins encoded by these genes can play an important role in cellular protection against aBL and higher temperatures and may be responsible for the cross-stress tolerance that was observed. These findings and insight into the function of the deleted genes suggest that heat stress contributes to cell membrane disruption and electron leakage, which can lead to ROS generation. This step is also present in the aBL mode of action, which partially explains that the same genes may be involved in the cellular response to heat and aBL and overexpression of shared protecting genes may lead to the risk of developing cross-stress tolerance to both stressors. Despite the similar effects of these two approaches, such as membrane disruption, ion and electron leakage, macromolecule damage resulting in bacterial death, heat stress and aBL have different initial cell killing strategies in which different genes are involved. According to current research, 53 of all 64 single-gene mutants hypersensitive to aBL were not hypersensitive to heat. The proposed mechanism of aBL and heat cross-stress tolerance is shown in Fig. 8.

Fig. 8

The proposed mechanism of aBL and heat cross-stress tolerance. aBL leads to the excitation of endogenous photosensitizers, which results in the generation of ROS causing membrane stress and affecting the respiratory chain. Heat stress leads to membrane destabilization and affects the respiratory chain, which leads to electron and ion leakage and consequently to generation of ROS. As a result of both aBL and heat stress, damage to cellular macromolecules accumulates, resulting in bacterial cell death. Possible mechanisms of cross-stress tolerance development may relate to membrane stabilization, ROS detoxification and activation of the macromolecule repair system. The figure is based on Murata et al. [75]

Research has shown that populations that are tolerant to high temperatures do not exhibit different sensitivity profiles to aBL. In addition, aBL-tolerant populations were less sensitive to temperatures ranging from 52 to 56 °C. This result confirms that co-selection of thermal stress and aBL can occur with E. coli, but this process depends on the order of exposure to these selection factors. The results obtained suggest that a co-selection phenomenon occurs in which acquired tolerance to aBL may lead to cross-stress tolerance to temperature stress; however, the resulting temperature tolerance is significantly lower than that in a population selected for this feature. Fortunately, increased tolerance to temperature and aBL did not significantly affect the antibiotic sensitivity of the tolerant strains.

Our results indicate that both increased culture temperature (42 °C, 16–20 h) and short-term pre-incubation (50 °C or 52 °C, 10 or 20 min) at a high temperature affect the sensitivity of E. coli cultures to aBL. The aBL sensitivity profiles of bacteria cultured at 37 °C (optimal) and 42 °C were compared, and the population cultured at higher temperatures was less sensitive to aBL. Similarly, in the case of short-term pre-incubation, a significantly reduced sensitivity to aBL was observed for all tested parameters (50 °C or 52 °C; 10 min or 20 min), however, the greatest tolerance to aBL exposure was observed for the bacterial population exposed to shorter pre-incubation time (10 min, 50 °C). This is probably related to the heat shock response (HSR), which in Escherichia coli is a very rapid process involving the immediate transcription of heat shock proteins (HSPs) upon exposure to heat stress. The regulator of HSR (σ32 factor) is immediately stabilized and activated. Initiation of transcription of HSP genes takes place already in the first two minutes after exposure, and after five minutes of exposure, significant levels of mRNA for heat shock proteins can be detected. The HSPs’ mRNA translation begins within 5–10 min. Which means that after 10–15 min, protective proteins accumulate in the cell [78]. Most likely, 10 min of pre-exposure to elevated temperature was enough for E. coli cells to accumulate sufficient amounts of HSPs’ mRNA and proteins in bacterial cells exposed to high temperature, which enabled increased tolerance to aBL. Apparently, as much as 20 min of exposure to pre-treatment with heat stress was less beneficial for the cells, perhaps it was associated with greater damage in the cell. Longer exposure to heat could also have caused that HSP proteins, instead of accumulating in the cell, were mostly already used in response to prolonged heat stress and in the case of aBL treatment there could have been significantly less of HSP accumulated in the cell.

Our result corresponds to the findings obtained by St. Denis et al. [72] for sublethal toluidine blue O-mediated (TBO-mediated) aPDI (λ 635 nm) and pretreatment with heat (50 °C for 30 min). Short-term pre-incubation reduced E. coli sensitivity to TBO-mediated aPDI by 2 log₁₀. The scientists also noted that protective responses, such as HSPs, were induced after TBO-aPDI treatment (GroEL increased sevenfold and DnaK threefold) [72]. The results obtained by Kitagawa et al. [79] suggest that small HSPs (IbpA and IbpB) is also involved in resistance to oxidative stress and heat. Moreover, research conducted by Bolean et al. [80] demonstrated that GroEL levels were increased following rose bengal-mediated (RB-mediated) aPDI of Streptococcus mutans. The expression of HSP after RB-aPDI was similar to that induced by osmotic stress (1 mol/l NaCl). HSR genes encode chaperones, proteases and other stress-related proteins that play important roles in cellular responses to stress associated with elevated temperatures and responses to several other environmental stresses, such as insufficient nutrients, DNA damage, oxidative stress and heavy metals [81].

Conclusions

In conclusion, as in all studies on aBL conducted so far, bacteria did not develop resistance to this method, consistent with all studies on aBL conducted thus far. After repeated sublethal treatment with aBL, the drug susceptibility profile of the tolerant populations did not change significantly. Moreover, heat-tolerant strains can still be eradicated using aBL, and aBL is a safe method for the food sector, as many studies have indicated. The results obtained by our team do not invalidate the use of aBL in the food industry but indicate that researchers should consider adaptation or co-selection when establishing rules for the safe use of this method. Further research is needed on the long-term effects of aBL use and interactions with other stressors found in the food industry to develop protocols that ensure the greatest safety for consumers.