Both E. coli W3110 and S. Typhimurium SL1344 were purchased from ATCC. All of the stains were cultivated at 37°C on Luria-Bertani agar plates containing 10 g/L tryptone 5 g/L yeast extract, 10 g/L NaCl and 15 g/L agar. Subsequently, a single colony was selected and inoculated into 5 mL of LB broth, then incubated at 37°C with shaking at 200 rpm. Cells were harvested once the optical density at 600 nm (OD₆₀₀) reached the desired level for further investigation.

The procedure of aBL irradiation was carried out using a method previously reported (Wu et al., 2018). The aBL irradiation was performed using a 415 nm LED array (Omnilux clear-UTM, Photo Therapeutics, Inc.) with a peak wavelength of 415 nm and a bandwidth of 20 nm. The LED array was positioned at a fixed height to ensure an irradiation intensity of 16.7 mW/cm². The cell suspension with OD₆₀₀ at 2.0 was prepared for aBL treatment testing. Five mL of cell suspension was transferred to a well of a 6-well clear flat-bottom plate (36 mm diameter, 16.5 mm height) and placed on a magnetic stirrer at 30 rpm with gentle stirring. The bacterial suspensions were illuminated with aBL, and the survival rates were measured (Lipovsky et al., 2009; Wu et al., 2018). Aliquots of 20 µL were taken from the aBL-treated cell suspensions at the irradiation doses of 0, 10, 20, 30, 60, 90, and 120 J/cm² respectively. Visible cell counts of the aBL-treated bacterial population were measured and the bacterial cell reduction was calculated by comparison with untreated control group. Cells exposed to the sub-lethal aBL doses of 15 J/cm² and 30 J/cm² were harvested separately to lipid extraction, as detailed in the procedure below.

Lipid extraction from S. Typhimurium and E. coli cells following aBL treatment was performed using a modified protocol based on the method of (Folch et al., 1957). Thirty mL of cell suspensions were centrifuged at 4,000 × g for 10 min at 4 °C. The cell pellets were washed thrice with ultrapure water, with each wash followed by 5-min of vibration. After washing, the bacterial cells were subjected to three freeze-thaw cycles by alternately immersing them in liquid nitrogen for 1 min and thawing on ice for 3 min to promote lysis. The lysed cells were resuspended in 6 mL of ultrapure water and gently vortexed for 30 s. For lipid extraction, 900 μL of pre-cooled chloroform/methanol (2:1, v/v) was added to the suspension. The mixture was vortexed for 30 s and centrifuged at 4,000 × g for 10 min at 4 °C. The lower organic phase was carefully collected. A second round of extraction was performed using the same solvent system, and both organic phases were combined. The combined extracts were dried under a gentle stream of nitrogen gas. The resulting lipid residues were reconstituted in a mixture of acetonitrile/isopropanol and chloroform/methanol (1:1, v/v), and stored at –20°C until LC-MS analysis.

The Lipid extracts were separated using a UHPLC system (Ultimate 3000, Thermo Scientific, Bremen, Germany) coupled to a ThermoFisher Q Exactive™ Plus Hybrid Quadrupole-Orbitrap™ mass spectrometer. Chromatographic separation was performed using two mobile phases: mobile phase A, composed of acetonitrile and water (60:40, v/v) with 10 mM ammonium acetate, and mobile phase B, composed of isopropanol and acetonitrile (90:10, v/v) with 10 mM ammonium acetate. The sample was injected into a Waters CSH C18 column (2.1 mm × 100 mm, 1.7 μm) and separated using a linear gradient, transitioning from 60% mobile phase A to 95% mobile phase B over 20 minutes, followed by re-equilibration to initial conditions. The flow rate was maintained at 0.25 mL/min. The MS analysis was performed in the negative ion mode with an electrospray voltage of 2.8 kV, a capillary temperature of 350°C, and a sheath gas flow rate of 35 units. MS and MS/MS data acquisition were acquired over an m/z range of 80-1200 (Schwalbe-Herrmann et al., 2010).

Lipid species were identified using LipidSearch software based on retention time, accurate mass, and MS/MS spectra. Specific ion forms were used for lipid identification, including [M-2H]^2- for CL and [M-H]^- for PE, PG, and PS. MS/MS spectra further characterized lipids, particularly for those containing two fatty acid chains. For example, PE(17:0-18:0) was identified by characteristic m/z signals corresponding to its head group and fatty acid carboxylate ions. Normalized data sets were analyzed using MetaboAnalyst 5.0. Partial least squares discriminant analysis (PLS-DA) revealed clear clustering between aBL-treated and control groups, differentiating lipidomic profiles under varying aBL irradiation conditions. All experiments were conducted in triplicate, and results were expressed as mean ± standard deviation (SD). Statistical analyses were performed using Origin with one-way analysis of variance (ANOVA), followed by appropriate post hoc testing to evaluate differences between treatment groups. A p-value less than 0.05 was considered statistically significant. In Figures 1 and 5, lowercase letters above the bars indicate statistically significant differences between specific time points:

The procedure of malondialdehyde (MDA) exact and analysis was identical to that previously reported by our group (Wu et al., 2018). The absorbance of the supernatant at 532 nm was measured and MDA content was calculated using a standard curve. The detailed procedure followed the manufacturer’s protocol (Kit #S0103, Beyotime, China). The intracellular ROS was detected based on our previous report (Wu et al., 2018).

OM permeability was assessed using the N-phenyl-α-naphthylamine (NPN) uptake method (Qiao et al., 2021). Following aBL irradiation at different doses (0, 15, and 30 J/cm²), the cells were resuspended in 1.92 mL of phosphate buffer solution (OD₆₀₀ = 0.5). Subsequently, 80 μL of NPN solution (1 mM) was added to the cell suspension, and the fluorescence intensity was immediately measured using a spectrofluorometer (Hitachi, Tokyo, Japan). The excitation wavelength was set at 350/10 nm, and the emission wavelength was set at 420/10 nm. Potassium leakage was determined by measuring extracellular K⁺ concentration in the supernatant using a flame photometer after centrifugation at 12,000 × g for 5 min (Wu et al., 2018).

The aBL exhibited efficient bactericidal effect against E. coli W3110 and S. Typhimurium SL1344, while two strains exhibited differential sensitivity to aBL in a dose-dependent manner. As shown in Figure 2, when aBL dose exceeded 30 J/cm², the population of E. coli W3110 began to decrease, and when irradiation was extended to 120 J/cm², it was further reduced by 3.76 log CFU/mL. Previous studies have shown that aBL at various wavelengths produced similar bactericidal effects against E. coli isolates. For examples, 410-nm aBL irradiation at 412.50 J/cm² dose killed E. coli ST10 beyond 3.7 logCFU and E. coli ST648 at 2.8 logCFU, respectively (Dos Anjos et al., 2019). Furthermore, E. coli ST131 and O127:H7 became non-culturable following aBL exposure (Dos Anjos et al., 2019). Another study reported that 455-nm aBL illumination at 120 J/cm² reduced 1.5-2.5 log CFU/mL of E. coli DH5α and 1 log CFU/mL of E. coli MG1655 (Abana et al., 2017). In the present study, 415-nm aBL illumination at 120 J/cm² achieved comparable inactivation efficiency. These variations of bactericidal efficiency of aBL were likely due to differences in strain-specific sensitivity, initial bacterial concentration, culture conditions, and light exposure parameters such as wavelength and irradiance uniformity. In our previous report on S. Typhimurium SL1344 (Hu et al., 2021a), even a single mutation of lipopolysaccharide located at OM significantly changed its photosensitivity, thus for these E. coli strains mentioned above, substantial differences in genetic background and OM properties plausibly account for the observed discrepancies across studies.

The inactivation of S. Typhimurium SL1344 was observed when aBL dose exceeded 60 J/cm² (Figure 2). Viable cell counts decreased by 1.03 log CFU/mL after 90 J/cm² and by 2.02 after 120 J/cm². In contrast, our previous study reported 383 J/cm² aBL inactivated SL1344 by 4 log CFU/mL (Hu et al., 2021a). Similarly, other studies reported that 405 nm aBL reduced S. enterica serovar Enteritidis by 1.4 log CFU/mL after 739.6 J/cm² exposure (Endarko et al., 2012), and reduced S. Enteritidis and S. Saintpaul by 2.0 log CFU/mL and 1.0 log CFU/mL, respectively, after 288 J/cm² treatment (Kim and Yuk, 2017).

In summary, our study demonstrated the bactericidal efficacy of 415-nm aBL against both E. coli and S. Typhimurium, while revealing species-specific susceptibility to aBL treatment. At the maximum aBL dose at 120 J/cm², viable cell counts decreased by 3.76 log CFU/mL for E. coli W3110 and 2.02 log CFU/mL for S. Typhimurium SL1344. Notably, aBL could be generated using simple LED-based devices without requiring exogenous photosensitizers, and exert bactericidal effect against a broad spectrum of foodborne pathogens. These findings supported aBL as a promising non-thermal disinfection strategy for foodborne pathogens.

Similarly, in S. Typhimurium, 74 lipid species across 5 classes were identified using PLS-DA (Supplementary Figure S1B). PE and PG were the predominant lipids class, comprising 23 PE species and 12 PG species, all of which were identical to those found in E. coli, respectively. For other lipid classes such as LPE, PS and CL, there are 3, 6, 2 species similar to those in E. coli. These 46 lipids mentioned above maybe conserved in E. coli and S. Typhimurium.

Similar to its impact on E. coli, aBL irradiation induced notable changes in the lipidome of S. Typhimurium. All lipids were listed in Supplementary Table S3, and some lipids were demonstrated in Supplementary Table S4. Following exposure to 15 J/cm² aBL treatment, changes were observed in the types and intensities of PE, PG, and CL (Figure 3B). Increasing the dose to 30 J/cm² further altered the lipid profile and content, as detailed below.

Under aBL irradiation, S. Typhimurium exhibited more variable lipid changes compared to E. coli., with 18 species showing erratic lipid changes (Figure 6; Table 4). Eleven PE species initially decreased at a 15 J/cm² dose, and subsequently increased at 30 J/cm². Comparatively, the changes in PE lipids showed a similar trend in both E. coli and S. Typhimurium (Tables 2A, 2B). However, in S. Typhimurium, a notable increase in saturated lipids and a decline in unsaturated lipids were observed, particularly within the PE class (Figure 6; Table 5).

More PG species were detected in S. Typhimurium than in E. coli, although fewer of them exhibited changes during aBL irradiation (Figure 7A). Seven PG species exhibited continuous increases or decreases, whereas 16 PG species displayed erratic patterns. Additionally, CL species in S. Typhimurium differed from those found in E. coli. Notably, CL(18:1-16:1-16:0-18:1) increased in S. Typhimurium but decreased in E. coli.

However, the overall extent of lipid changes, particularly in PE and PG species, was less pronounced and less consistent in S. Typhimurium compared to E. coli, as fewer lipid species exhibited statistically significant or directional changes across irradiation doses. This raises the question of whether such differences in lipid fluctuation correlate with S. Typhimurium’s lower sensitivity to aBL irradiation.

Next, MDA level, OM permeability and potassium leakage in S. Typhimurium exposed to aBL were determined. As illustrated in Figure 4B, the MDA levels increased from 4.37 ± 0.71 µM/OD₆₀₀ before aBL irradiation to 8.37 ± 0.97 µM/OD₆₀₀ after 15 J/cm², then declined to 4.77 ± 1.69 µM/OD₆₀₀ at after 30 J/cm² dose. Compared to E. coli, where MDA levels increased to 22.64 ± 3.28 µM/OD₆₀₀ at 30 J/cm², S. Typhimurium exhibited lower MDA levels (p < 0.001). Besides, OM permeability in S. Typhimurium increased markedly, rising by 101.65% at 15 J/cm² and reaching 174.38% at 30 J/cm² (Figure 6B), likely due to the oxidative stress-induced alterations in membrane composition (p < 0.001). Additionally, although S. Typhimurium exhibited high background leakage, it still showed a significant increase along with aBL irradiation.

This study demonstrates the profound impact of aBL irradiation on the lipidomic profiles of S. Typhimurium and E. coli. Our findings reveal substantial diversity and variability in lipid composition across sublethal aBL irradiations. Notably, we observed significant alteration membrane lipids—particularly unsaturated species—likely driven by aBL-induced oxidative stress. These insights are critical for deciphering the bactericidal mechanisms of aBL and uncovering the molecular basis of strain-specific difference in aBL sensitivity.