The experiments were conducted employing the MEDICAL 99 IR Multiossigen device (Gorle, Bergamo, Lombardy, Italy), provided by SIOOT. The device is certified by the Notified Body of the Ministry of Health as Class IIa under Directive 93/42/EEC.

For the experiments, we used the MRSA USA300 reference strain from our collection of isolates derived from clinical specimens. The growth and inactivation of MRSA were conducted in Tryptic Soy Broth (TSB, OXOID) overnight (O/N) at 37 °C. The O/N bacterial culture was diluted 1:100 in fresh TSB broth and incubated until the exponential growth phase was reached (OD₆₀₀ = 0.6). The bacterial culture was then processed using two protocols, namely one-step and multi-step, prior to O₃-induced bacterial killing. In the one-step approach, 5 mL of the bacterial culture (OD₆₀₀ of 0.6) was washed and resuspended in phosphate-buffered saline (PBS) (v/v) before proceeding with O₃ killing. The bacterial culture was treated, in independent experiments, with three O₃ concentrations (20 µg/mL, 40 µg/mL and 60 µg/mL, respectively) for 5 minutes (MRSA O₃I). To verify whether bacterial killing was successful, 10-fold serial dilutions of the bacterial suspension from each treatment condition, were prepared, plated on Tryptic Soy Agar (TSA; Oxoid Ltd., Basingstoke, Hampshire, UK) and incubated at 37 °C O/N. The absence of bacterial growth was taken as evidence of the efficacy of the killing method.

In the multi-step protocol, 5 mL of the bacterial culture (OD₆₀₀ of 0.6), was washed twice in RPMI 1640 and resuspended in ozonated RPMI 1640 (v/v) obtained by a prior O₃ treatment at 60 µg/mL for 20 minutes. RPMI 1640 ozonation was necessary to reduce the antioxidant effect of the medium. The bacterial suspension in ozonated RPMI 1640 was subsequently treated with an O₃ concentration of 60 µg/mL at 20-minute intervals for at least 1 hour, depending on the initial bacterial concentration. After O₃ treatment, 200 µL of the bacterial suspension was plated at serial dilutions either on TSA or on Chromid MRSA SMART plates (bioMérieux SA, Marcy-l’Étoile, France) to assess killing activity.

In addition, two further bacterial-killing experiments were performed with a bacterial cell density (10⁷ colony-forming units (CFU)/mL) using heat inactivation or gamma-ray irradiation. In the first experiment, the bacterial culture was washed in PBS and heat-inactivated at 100 °C for 10 minutes. In the second experiment, bacterial culture was treated with O₃ for up to five hours according to the multi-step protocol, and the surviving rare colonies were subsequently subjected to gamma-ray irradiation over the weekend at a dose of 2700 Gy [46] as described in more detail in the Results.

Gamma irradiations from a ¹³⁷Cs source were performed at the Istituto Superiore di Sanità (ISS, Rome, Italy) using the Gammacell Exactor 40 (Nordion) [47].

The MRSA bacterial culture (10⁷ CFU/mL) was heat-inactivated at 100 °C, to prepare the control (MRSA HI) [48]. To confirm the lack of viable bacteria following heat inactivation and gamma-ray irradiation, 200 µL of the bacterial suspension was plated on TSA and incubated at 37 °C O/N.

The study was approved by the Ethics Committee of the Italian National Institute of Health (Istituto Superiore di Sanità, ISS), Rome, Italy (Protocol PRE BIO CE n. 0041072, November 24, 2021). Peripheral blood mononuclear cells (PBMCs) were isolated from healthy adult volunteers who provided written informed consent, in accordance with the Agreement between the Umberto I University Hospital (AOU Policlinico Umberto I, Rome) and the ISS regarding the “transfer of blood and its derivatives for laboratory use”, established under Resolution No. 0000853 of August 18, 2020. Only primary PBMCs were employed to ensure a physiologically relevant model; no immortalized or transformed cell lines were used.

PBMCs were isolated from buffy coats obtained from healthy donors (n = 3/5) released by the Transfusion Center of Rome Policlinico Umberto I, as previously described [49]. PBMCs were cultured in RPMI 1640 with 10% fetal bovine serum (FBS) and stimulated with MRSA at a multiplicity of infection (MOI) of 1, using bacteria inactivated by the methods that yielded the most effective bactericidal results. After two hours of stimulation, PBMCs were treated with O₃ at 40 µg/mL to investigate, in vitro, the immunomodulatory and anti-inflammatory actions observed in vivo. Appropriate negative controls and Staphylococcal enterotoxin B (SEB) from Sigma-Aldrich (St. Louis, Missouri, USA; catalog no. S4881) positive controls were included in all experiments, conducted with both MRSA -O₃-inactivated (MRSA O₃I) and MRSA HI bacteria.

Our experiments were conducted using PBMCs isolated from freshly obtained buffy coats, ensuring that blood donations were processed within 24 hours of collection. This short interval between blood withdrawal and PBMC isolation is unlikely to allow the emergence of detectable changes in cellular morphology, surface marker expression, or mycoplasma contamination.

Cytokine production (TNF-$\alpha$, IL-10, IL-6, IL-1$\beta$ and IL-8) was quantified in culture supernatants collected from PBMCs before and 24 h after stimulation with various bacterial preparations, with or without O₃, as well as from untreated PBMCs. Cell samples were analyzed using a specific Cytometric Bead Array (CBA) Human Inflammatory Cytokine Kit (BD Biosciences, San Jose, CA, USA; cat. BDB551811). Samples were acquired on a FACS Canto II flow cytometer (BD Biosciences, San Jose, CA, USA), and data were analyzed with FCAP Array software v3.0 (BD Biosciences, San Jose, CA, USA) according to the manufacturer’s instructions.

For intracellular cytokine analysis, we used a dried commercial panel (Duraclone IF T Helper Cell kit #C04666, Beckman Coulter, Brea, CA, USA) as the backbone, supplemented with liquid-format antibodies against CD8, IL-10, and a dead-cell exclusion marker, see Table 1 for details of the antibodies. Briefly, PBMCs from healthy donor buffy coats were cultured under three conditions: unstimulated (NT), stimulated with SEB at 2 µg/mL as a positive control, or stimulated with MRSA HI or MRSA O₃I (at a 1:1 MRSA:PBMC ratio). After 2 hours, samples were treated with O₃ (40 µg/mL or 20 µg/mL) and then incubated for 24 hours with Brefeldin A and Monensin (Becton Dickinson), which were added ~1-hour post-O₃ treatment.

The next day, cells were stained for dead cell exclusion, followed by staining with the Duraclone antibody cocktail, which incorporated anti-CD8 and anti-IL-10 monoclonal antibodies. Samples were acquired on a FACS Gallios instrument (Beckman Coulter, Brea, CA, USA), and data were analyzed using the Kaluza Analysis Software v.1.3 (Beckman Coulter Brea, CA, USA).

The gating strategy for multiparameter intracellular cytokine analysis within the CD3+, CD4+, and CD8+ T gates is shown in Supplementary Fig. 1, beginning with singlet and live-cell selection followed by T cell subset identification and cytokine detection (IFN-$\gamma$, IL-4, IL-17 and IL-10). Representative dot plots are shown for CD3+ T cells under NT and SEB-stimulated conditions. The same analysis was applied to CD4+ and CD8+ T cells as well as to samples treated with MRSA alone and MRSA + O₃.

Total RNA was isolated from 3 $\times$ 10⁶ PBMCs, either untreated or stimulated for 24, 48 or 72 hours with MRSA HI or MRSA O₃I and subsequently exposed to O₃ in vitro using TRIzol reagent (Invitrogen, Thermo Fischer Scientific, Waltham, MA, USA). Extracted RNA was quantified with a NanoDrop OneC spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA), and RNA quality was assessed using an established 260/280 absorbance ratio cutoff of ~1.8. Reverse-transcription was performed using the Vilo reverse transcription kit (Invitrogen, Thermo Fisher Scientific, Waltham, MA, USA). The expression of genes NFE2L2 (NRF2 protein), NLRP3 (NLRP3 protein), NFKB1 and NFKB2 genes (encoding the inactive precursors p105 and p100 of the NF-$\kappa$B complex) was measured by qRT-PCR using the corresponding TaqMan assays and TaqMan Universal Master Mix II (Applied Biosystems, Thermo Fisher Scientific, Waltham, MA, USA) on a ViiA7 Instrument (Applied Biosystems, Thermo Fisher Scientific, Waltham, MA, USA). The housekeeping gene TATA-box-binding protein (TBP) served as the endogenous normalizer. Real-time reactions were performed at least in duplicate. Sample values for each mRNA were normalized to the selected housekeeping gene using the 2^-Δ⁢Ct method.

Among the various methods available to assess ROS production or oxidative stress in cells, we used the probe 2^′,7^′-Dichloro-dihydro-fluorescein diacetate (DCFH-DA). This is a lipophilic, cell-permeable, non-fluorescent probe that, upon hydrolysis by intracellular esterases, yields DCFH, which fluoresces upon oxidation by intracellular ROS. Oxidation of the probe results from the action of several types of peroxynitrite, hydroxyl radicals, nitric oxide, metal-derived oxidants and other peroxides, and can be detected by flow cytometry. PBMCs were seeded at 1 $\times$ 10⁶/mL, and incubated with bacterial preparations including O₃-killed bacteria (20–60 µg/mL O₃) and heat-killed MRSA as a control. After incubation at 37 °C, with 5% CO₂, for 2 hours, we added 2^′,7^′-Dichlorofluorescein (DCF) (Sigma) to a final concentration of 5 µM to detect intracellular ROS and incubated the cells for at least 1 hour, before the subsequent treatment with 40 µg/mL O₃. We used the MultiOssigen O₃ technology - Medical 99IR (MultiOssigen S.p.a) as an O₃ delivery system.

After ozonation, the samples were centrifuged at 1200 rpm at room temperature for 10 minutes and the pellets were then suspended in PBS containing the vital dye Sytox Blue (Invitrogen™ Cat. n. S34857) at 1:1000 for flow cytometric analysis of ROS using the CytoFLEX LX (Beckman Coulter) with excitation and detection wavelengths of 488 nm and 535 nm, respectively. As negative controls, we used PBMCs and PBMCs plus DCF, with and without O₃ insufflation, to estimate the number of unstained cells with autofluorescence in the green emission range and to quantify basal ROS levels in our sample. As a positive control, we incubated cells with 0.1 mM H₂O₂, with and without O₃ insufflation.

ROS production was assessed every 20 minutes for 2 hours following O₃ administration, with a final measurement taken at 24 hours. The gating strategy for ROS measurement with DCF is described in Supplementary Fig. 2.

PBMCs (2 $\times$ 10⁶/mL) were incubated in the absence or in the presence of the different stimuli at 37 °C for 2 hours. O₃ (40 µg/mL) was then added and, after a further 22 hours incubation, cells were washed, resuspended in an appropriate volume of PBS, and lysed (1:2 v/v) in RIPA buffer (50 mM Tris HCl, pH 8.0, 150 mM NaCl, 0.5% sodium deoxycholate, 0.1% SDS) containing protease inhibitors (10 µg/mL, leupeptin, 10 µg/mL aprotinin, 0.15 mM PMSF). After 10 minutes in ice, cell lysates were diluted 1:4 (v/v) in PBS. An aliquot of cell lysate was treated with ice-cold 20% trichloroacetic acid (1:2 v/v) for 10 min on ice, then centrifuged at 15,000 rpm, at 4 °C for 10 min. The cleared acidic supernatants, used for glutathione measurement, and the residual cell lysate volume, used for protein measurement and GPx activity evaluation, were stored at –80 °C and used within one week. The glutathione and the glutathione peroxidase assay kits (Cayman Chemicals) were used to measure glutathione concentration and GPx activity, respectively, according to the manufacturer’s instructions. Protein content was measured using the BCA Protein Assay kit (Millipore).

We computed the between-samples Pearson’s correlation coefficients to check the consistency among different subjects (samples) across all the measured variables. The results indicated strong consistency among independent subjects, providing confidence in the reliability of the data (Table 2). Therefore, given the similarity among profiles, we considered the pattern reproducible despite the small sample size (n = 3–5) used. A two-tailed paired Student’s t-test was then performed according to the experimental design, with p 0.05 considered statistically significant. Raw data are presented in Supplementary Tables 1,2.

The relatively small sample size in this study is justified by the preliminary nature of the analysis, aimed at exploring specific in vitro effects of O₃ exposure. This initial investigation provides valuable insights that can inform the design of future, larger-scale studies to achieve a more comprehensive understanding. While the in vitro model does not fully replicate the complexity of living organisms, a useful platform is provided to examine the associated cellular and molecular responses to ozone.

Our choice to use inactivated bacteria was a deliberate methodological decision guided by the primary objective of our study. The aim of the study was not to replicate an active infection, but to isolate and investigate the direct effect of ozone on the immunomodulatory and antioxidant responses of host cells to bacterial components without the additional complexities and unpredictable variables introduced by live bacterial replication and metabolism. Thus using inactivated bacteria allowed us to isolate specific response mechanisms and ensure reproducibility, which is essential for reliable comparative analysis across experimental groups.

The immunomodulatory effects of O₃ were assessed in PBMCs using functional assays to measure cytokine production, ROS, and oxidative stress. In addition, we performed qRT-PCR analysis on selected pathway regulators including key genes such as NFKB1, NFKB2, and NLRP3 (involved in immune activation) at early stages of O₃ treatment, and NRF2 (from gene NFE2L2) (associated with antioxidant and anti-inflammatory responses) at a later stage. This approach reflects the hypothesis-driven nature of our study, which focuses on the involvement of well-established genes in immune response, rather than on a broad, data-driven exploration. Meanwhile, qRT-PCR was selected for the associated high sensitivity and accuracy of the assay in quantifying expression changes in the genes of interest, thereby supporting our hypotheses. Although RNA-sequencing could aid broader preliminary analyses, this method was deemed unnecessary for the specific goals of this study.

Together, these functional and molecular evaluations provide robust evidence of the biological impact of O₃. Nevertheless, further studies employing more complex models, larger sample sizes, and in vivo validation will be necessary to substantiate and expand upon these findings.

The bactericidal activity of O₃ against MRSA was first assessed using a single-step exposure protocol. MRSA suspensions (10⁶–10⁷ CFU/mL) in PBS were exposed to O₃ for 5 minutes, which was sufficient to achieve maximal bacterial killing (data not shown). For subsequent PBMC stimulation experiments, heat-inactivated MRSA (MRSA HI) was used as a control, providing a non-replicating bacterial stimulus while preserving antigenic properties (Table 3).

When using RPMI medium, a multi-step protocol was implemented to provide a closer clinical replication of relevant exposure conditions. This protocol involves repeated O₃ treatments at 60 µg/mL every 20 minutes, consistent with clinical protocols for managing patients with antibiotic-resistant bacterial infections, where O₃ is administered twice weekly for MRSA treatment [11]. As shown in Table 3C, MRSA suspensions (10⁶ CFU/mL) exposed to three sequential O₃ applications over 1 hour exhibited a complete inhibition of bacterial growth.

To further evaluate the efficacy of O₃ in the presence of antioxidant components within RPMI, additional experiments were performed using high-titer MRSA suspensions (10⁷–10⁹ CFU/mL) and the multi-step exposure scheme. Table 4, summarizes three experiments in which bacterial suspensions were treated every 20 minutes for 3–5 hours. A time-dependent and concentration-dependent logarithmic decline in CFU counts was observed in all cases. After approximately 3 hours of repeated O₃ exposure, a five-log reduction in viable bacteria was achieved, confirming the potent bactericidal effect of O₃ even under conditions that partially mimic physiological environments.

However, at very high bacterial densities, S. aureus tends to form aggregates, which can transiently reduce O₃ efficacy due to the protective effect of outer-layer killed cells, meanwhile those within the clusters remain protected, a phenomenon also reported for Pseudomonas aeruginosa exposed to conventional antibiotics [50].

To ensure complete sterility before use in stimulation assays with PBMCs, residual bacterial colonies were eliminated by gamma irradiation at 2700 Gy for 3 days. No bacterial growth was observed in any of the preparations after 24 hours. This approach is well documented for decontaminating S. aureus in food matrices and has the advantage of preserving protein epitopes, thereby maintaining antigenic integrity [51, 52].

Both MRSA-O₃I and MRSA HI preparations were subsequently used to stimulate PBMCs isolated from 3/5 healthy donors.

Cruciani et al. [48] showed that S. aureus induced a specific panel of regulatory cytokines (IL-23, IL-12, and IL-10) and pro-inflammatory mediators (TNF-$\alpha$, IL-6, and IL-1$\beta$) in human primary dendritic cells, thereby promoting Th1 and Th17 differentiation. To analyze the impact of O₃ treatment on the secreted cytokines, PBMC supernatants were collected after 24 hours of O₃ treatment and stimulation with both MRSA O₃I and MRSA HI (Fig. 1). By CBA (Fig. 1A–E), we observed a reduction in the immunoregulatory cytokines (TNF-$\alpha$, IL-6, IL-1$\beta$, and IL-10) whereas IL-8 slightly increased in the MRSA HI samples. In particular, reductions in IL-1$\beta$ (Fig. 1B) and IL-6 (Fig. 1C) levels were statistically significant, with p-values of 0.022 and 0.04, respectively, based on experiments performed with PBMCs from four independent donors in the MRSA O₃I group. Stimulation with the SEB positive control induced a significant increase in all tested cytokines compared with the untreated control.

Fig. 1.

Cytokine detection in culture supernatants by CBA. The bar graphs depict cytokine levels in culture supernatants from PBMCs collected 24 hours after treatment, with or without O₃ added after 2 hours of stimulation with MRSA O₃I or MRSA HI. Cytokines are expressed in pg/mL on a logarithmic scale. Cytokine production of TNF$\alpha$ (A), IL-1$\beta$ (B), IL-6 (C), IL-8 (D), and IL-10 (E) was measured over 24 hours using CBA. Negative and positive controls (SEB) were included. Data are presented as the means $\pm$ standard error of mean (SEM) from 4 independent experiments for MRSA O₃I and from 3 independent experiments for MRSA HI (n = 4/3 different healthy subjects). Statistical analysis was performed using paired t-tests, with p-values indicated (*p $\le$ 0.05, **p $\le$ 0.01, NS, not significant). CBA, Cytokine Bead Array; PBMCs, peripheral blood mononuclear cells; MRSA, methicillin-resistant Staphylococcus aureus.

The intracellular production of cytokines associated with the activation of T lymphocytes was assessed by intracellular cytokine staining in PBMCs stimulated with MRSA O₃I and MRSA HI. The gating strategy is shown in Supplementary Fig. 1. A reduction in IFN-$\gamma$ levels was observed in the CD3+, CD4+, and CD8+ T lymphocyte gated populations following in vitro O₃ treatment of MRSA O₃I-stimulated PBMCs, compared with the untreated MRSA O₃I sample. In MRSA HI samples treated in vitro with O₃, the decrease in IFN-$\gamma$ expression in the CD3+ and CD4+ gated populations was considerably less evident (Fig. 2A).

Fig. 2.

Intracellular cytokines in CD3, CD4, and CD8 lymphocyte gates. Bar graphs of the percentage of intracellular cytokines IFN-$\gamma$ (A), IL-4 (B), IL-10 (C), and IL-17 (D), measured by ICS, within CD3+, CD4+, and CD8+ T cell subsets of PBMCs stimulated for 2 hours with MRSA O₃I or MRSA HI, and subsequently treated with O₃ or untreated for 24 hours. Negative and positive controls (SEB) are also included. Data represent the means $\pm$ SEM of cytokine-positive cells from three independent experiments with three different healthy subjects for MRSA O₃I and 2 independent experiments with two different healthy subjects for MRSA HI. Statistical significance was determined by paired t-test (*p $\le$ 0.05, **p $\le$ 0.01, ***p $\le$ 0.001, NS, not significant). INF-$\gamma$, Interferongamma.

The O₃-induced reduction in IL-4 levels was comparable across all T cell subsets, regardless of the bacterial inactivation method (Fig. 2B).

Conversely, IL-10 was increased in CD3+ and CD8+ T cells following O₃ treatment of cells stimulated with either MRSA O₃I or MRSA HI. In contrast, in lymphocytes in the CD4+ gated population, IL-10 remained unchanged in MRSA O₃I-treated samples, while the IL-10 levels were slightly decreased in the MRSA HI-stimulated samples (Fig. 2C).

An O₃-induced decrease in IL-17 levels was detected exclusively within the CD3+ and CD4+ T cell populations, whereas a slight increase of this cytokine was observed in CD8+ T cells. This trend was consistent across both MRSA strains (Fig. 2D). The increase in IL-17 in CD8+ T cells may be partially correlated with the rise in IL-10 within the same gate in cells exposed to both bacterial strains. Although not statistically significant, these data suggest a trend toward differential expression of intracellular cytokines, except for SEB, which induces a significant increase in all tested cytokines, except IL-17, compared with untreated control samples.

To assess the effects of O₃ on key regulators of inflammation, we used qRT-PCR to measure the effect of O₃ on prooxidant- (NLRP3, NF-$\kappa$B) and antioxidant NRF2 (NFE2L2) gene expressions in PBMCs, treated with MRSA O₃I and MRSA HI and subsequently treated in vitro with O₃, compared to unstimulated control cells (Fig. 3).

Fig. 3.

Gene expression of key O₃ regulators. Graphs of the gene expression time course for key O₃ regulators assessed by RT-PCR from 24 to 72 hours: (A) antioxidant NRF2, (B) pro-oxidant NLRP3, (C) NF-$\kappa$B, and its subunits NFKB1 and NFKB2 (D). A comparison is made between MRSA HI (left) and MRSA O₃I (right), before (white circles) and after (black circles) O₃ treatment. Results are expressed as fold changes in gene expression.

NRF2 expression, a key regulator of the antioxidant response, increased following O₃ treatment, with the most pronounced upregulation observed after 48 hours in PBMCs exposed to O₃-killed bacteria in PBS and treated with O₃. A more modest increase was noted in cells treated with heat-killed MRSA HI and subsequently exposed to O₃ (Fig. 3A), indicating that O₃ enhances antioxidant defenses, particularly in the context of O₃I bacterial stimulation.

In contrast, NLRP3, a proinflammatory gene, was strongly induced by MRSA O₃I at 24 and 72 hours. O₃ treatment attenuated this induction, reducing NLRP3 expression is reduced to approximately 1.8-fold after 24 hours and decreases further at 48 hours, and by approximately 50% at 72 hours, while a similar reduction was observed in unstimulated cells at 48 hours. PBMCs stimulated with MRSA HI showed only minimal changes in NLRP3 expression (Fig. 3B).

Similarly, NFKB1 expression induced by MRSA O₃I was affected by O₃ addition at 24 hours and 72 hours while no changes were observed at 48 hours with NFKB1 below the threshold (Fig. 3C) Likewise, but to a lesser extent NFKB1 induced by MRSA HI at 24 hours was downregulated by O₃, while no substantial variations in the expression of NFKB1 were observed in MRSA HI stimulated PBMCs at 48 and 72 hours before and following O₃ treatment. For NFKB2, an increase was observed at 24 and 48 hours only in MRSA O₃I stimulated cultures after O₃ treatment, whereas expression decreased at 72 hours. A slight reduction was also measured in MRSA HI-treated PBMCs following O₃ stimulation (Fig. 3C).

NFKB1 expression followed a similar pattern: NFKB1 was induced by MRSA O₃I at 24 and 72 hours and downregulated by O₃ treatment, whereas at 48 hours expression remained below the threshold. In PBMCs stimulated with MRSA HI, NFKB1 expression showed only a slight reduction at 24 hours following O₃ exposure, with negligible changes at 48 and 72 hours (Fig. 3C).

Correlation analyses revealed minimal changes in NFKB1 and NFKB2 expression at 24 hours, while at 48 and 72 hours, strong correlations were observed among NRF2, NLRP3, NFKB1, and NFKB2 (Table 2). These findings indicate that O₃ maintains the associated antioxidant and modulatory effects over time, supporting sustained regulation of both pro- and anti-inflammatory pathways for up to 72 hours.

As expected, NRF2 was not induced by SEB at any time point (24–72 hours), consistent with the role of NRF2 in anti-inflammatory signaling. Similarly, NLRP3, a key component of the inflammasome, was not upregulated at 48–72 hours, suggesting that additional inflammatory stimuli are likely required to induce NLRP3. In contrast, the proinflammatory transcription factors NFKB1 and NFKB2 were both induced by SEB at 24 and 48 hours. However, by 72 hours, only NFKB2 remained upregulated, while NFKB1 returned to baseline levels (Supplementary Fig. 3).

Data are presented as the means $\pm$ SEM from three independent experiments up to 24 hours and two independent experiments from 48 to 72 hours. mRNA levels were normalized using the 2^-Δ⁢Ct method with GAPDH as the housekeeping gene.

ROS comprise a wide range of chemical species derived from O₂ that are more reactive than solely O₂, exhibiting distinct properties, reactivities and interactions.

When ROS are generated in vivo, numerous antioxidants are activated, delaying, preventing or removing oxidative damage to target molecules [53].

Thus, we evaluated DCF detection in lymphocytes and monocytes separately across the total PBMC population and observed similar kinetics in both cell types within 2 hours and after 24 hours.

As shown in Fig. 4, the presence of the inactivated MRSA in the culture slightly increased the amount of intracellular ROS released by lymphocytes (Fig. 4A) and monocytes (Fig. 4B). Nevertheless, the O₃ treatment initially increased ROS levels, showing an initial oxidant effect of the highly reactive gas. The highest ROS levels were observed within the first 20–40 minutes after ozonation and gradually decreased reaching an antioxidant status comparable to that of the corresponding untreated samples within 2 hours, and even falling below control levels after 24 hours in lymphocytes and monocytes stimulated with MRSA HI, where significant p values were found (Supplementary Table 2). Meanwhile, a similar trend to MRSA HI was observed for those treated with MRSA O₃I and exposed to O₃ (data not shown). This confirms that initial exposure to low doses of O₃ produced only a partial, reversible oxidizing effect, resulting in a more stable antioxidant outcome. Because O₃ is a highly reactive gas composed of three oxygen atoms, we compared the effects of the two gases using the same protocol described previously, applying pure oxygen (O₂) and O₃ treatments. The experiment was performed using MRSA HI and DCF fluorescence intensity was measured after 20 minutes. As expected, since Staphylococcus aureus is an aerobic bacterium, O₂ MRSA HI had an oxidizing effect. As shown in Fig. 4, enhanced O₃ derived ROS production in both lymphoid (Fig. 4C) and monocytic (Fig. 4D) subpopulations is higher in the presence of O₃. Furthermore, it must be considered that O₂ does not inactivate bacteria unlike O₃.

Fig. 4.

Measurement of reactive oxygen species (ROS) via 2^′,7^′-Dichlorofluorescein (DCF) assay. Flow cytometry histograms of the ROS levels in PBMCs from MRSA-infected and non-infected samples treated with O₃. ROS production is measured at multiple time points (0–2 hours and 24 hours post-treatment) using the DCF assay. H₂O₂ was used as the positive control. Data are shown as the fluorescence intensity profiles for lymphocyte (A) and monocyte (B) populations. In panels (C) and (D) the comparison of ROS levels in MRSA-infected and non-infected PBMCs treated with O₃ or oxygen (O₂) was reported. ROS production has been assessed over time with the DCF assay. Statistical analysis was performed to compare the effects of O₃ and O₂ on ROS generation in lymphocytes (C) and monocytes (D), respectively. The experiments were conducted with five healthy subjects (see Raw data), but the figure presents descriptive statistics for a representative sample. Statistical analysis was performed to compare the effects of O₃ on ROS generation. The p-values for three experiments were significant in MRSA HI in both lymphocytes and monocytes. Data are described in (Supplementary Table 2). The p-value for three independent experiments, conducted by three different donors, was 0.075 (two-tailed) for paired samples comparing MRSA O₃I and DCF at 20 minutes, compared with MRSA O₃I and DCF at 24 hours (data not shown).

GSH plays a key role in maintaining the cellular redox balance by directly neutralizing ROS and serving as a cofactor for various enzymes. During this process, GSH is oxidized to oxidized glutathione (GSSG), which is then reduced back to GSH by specific enzymes. GSH is also essential for GPx activity, which protects cells from oxidative damage by reducing peroxides using GSH as an electron donor [54].

The increase in ROS formation in MRSA-treated PBMCs and the protective effects of O₃ prompted us to investigate the status of intracellular antioxidant defenses. It is well known that increased oxidative stress alters intracellular redox status due to the depletion of antioxidant defenses. Thus, we investigated the ability of O₃ to modulate the intracellular concentration of glutathione and the activity of the antioxidant enzyme GPx in unstimulated PBMCs or those activated with SEB as a positive control, MRSA O₃I, and MRSA HI (Fig. 5A). The GSH concentration was significantly decreased by SEB and by both MRSA O₃I and MRSA HI (Fig. 5A), indicating that the thiol was consumed in counteracting oxidant species formed in activated PBMCs. Interestingly, the O₃ treatment restored GSH levels to levels comparable to controls when PBMCs were activated with MRSA O₃I and MRSA HI, whereas GSH levels were even higher than controls when activated with SEB.

Fig. 5.

Restoration of redox status in MRSA-treated PBMCs by ozone. (A) Graphs of the GSH and GSSG concentrations in PBMCs treated with SEB, MRSA O₃I, and MRSA HI. O₃ treatment is shown to restore redox balance to control-like levels. (B) GPx activity levels in PBMCs under various treatments, with data on the restorative effect of O₃ treatment. The results are presented as the mean $\pm$ SEM (n = 3), with statistical significance assessed by a paired t-test (p-value indicated). Data have been normalized to protein content (using the BCA protein assay). “*” indicates statistical significance determined by paired t-test (*p $\le$ 0.05).

The GSSG intracellular concentration was unaffected by the exposure of PBMCs to SEB or to MRSA O₃I; meanwhile, the GSSG intracellular concentration was significantly decreased by the MRSA HI treatment (Fig. 5A). The 40 µg/mL O₃ treatment did not significantly modify the GSSG concentration in any experimental condition tested with respect to the unstimulated cells or the corresponding treatment group.

The activity of GPx was significantly reduced in PBMCs activated with SEB or with MRSA O₃I and MRSA HI by 31, 25, and 35%, respectively, with respect to the unstimulated cells (Fig. 5B). In MRSA O₃I + O₃ versus MRSA O₃I, the difference was significant with a p-value of 0.026 for the two-tailed paired sample analysis. Interestingly, the 40 µg/mL O₃ treatment, not only restored but rather increased the enzyme activity in PBMCs activated with SEB or with MRSA O₃I and MRSA HI by 24, 37 and 10%, respectively, compared to the unstimulated cells (Fig. 5B). Overall, these results show that O₃, can protect cells from MRSA-dependent oxidative stress, decreasing the levels of ROS formed and restoring the intracellular redox equilibrium.