Low Ozone Concentrations Affect the Structural and Functional Features of Jurkat T Cells

: Autohemotherapy is the most used method to administer O 2 -O 3 systemically. It consists in exposing a limited amount of blood to a gaseous O 2 -O 3 and reinfusing it, thus activating a cascade of biochemical pathways involving plasma and blood cells that gives rise to antioxidant and anti-inﬂammatory responses. The therapeutic effects strictly depend on the O 3 dose; it is therefore necessary to understand the relationship between the O 3 concentration and the effects on blood cells involved in antioxidant and immune response. Here we performed a basic study on the effects of the low O 3 concentrations used for autohemotherapy on the structural and functional features of the human T-lymphocyte-derived Jurkat cells. Ultrastructural, biomolecular, and bioanalytic techniques were used. Our ﬁndings showed that 10, 20, and 30 µ g O 3 concentrations were able to trigger Nrf2-induced antioxidant response and increase IL-2 secretion. However, viability and proliferation tests as well as ultrastructural observations revealed stress signs after treatment with 20 and 30 µ g O 3 , thus designating 10 µ g O 3 as the optimal concentration in combining cell safety and efﬁcient antioxidant and immune response in our in vitro system. These data offer novel evidence of the ﬁne regulatory role played by the oxidative stress level in the hormetic response of T lymphocytes to O 2 -O 3 administration.

The most used method to administer O 2 -O 3 systemically is the so-called autohemotherapy. It classically consists in drawing a limited volume of blood (100-200 mL), exposing it ex vivo to a gaseous mixture of O 2 -O 3 for a few minutes and finally re-infusing it into the patient. This procedure activates a cascade of biochemical pathways involving both plasma and blood cells that gives rise to antioxidant and anti-inflammatory responses (reviewed in [3,5]), which account for the therapeutic effects of the O 2 -O 3 treatment. These effects strictly depend on the O 3 dose; in fact, high dosages, by causing a strong oxidative stress, activate the nuclear transcriptional factor kappa B and induce inflammation and tissue damage, whereas low dosages activate, through mild oxidative stress, the nuclear factor erythroid 2-related factor 2 (Nrf2) and, in turn, several antioxidant enzymes, thus promoting a cytoprotective and anti-inflammatory response [4,5]. Accordingly, in the last years, O 3 dosages have been progressively lowered [6], relying on their therapeutic efficacy for the hormetic dose-response relationship [7,8]. By contrast, some authors consider the O 3 -dependent activation of Nrf2 as potentially detrimental in oncologic patients since it has been reported that hyperactivated Nrf2 may support cancer progression through many pathways (e.g., [4,[9][10][11][12]). However, recent experimental evidence exists that low O 3 concentrations are unable to alter motility and proliferation of cancer cells in vitro [13] and the question remains controversial.
The clinical practice highlighted a strong heterogeneity in the antioxidant and antiinflammatory response of patients treated with autohemotherapy. For an effective and targeted application of O 2 -O 3 therapy, it is therefore necessary to understand the relationship between the O 3 concentration and the effects on blood cells involved in antioxidant and immune response.
In this view, we performed a basic study by investigating the effects of the low O 3 concentrations mostly used for major autohemotherapy on the structural and functional features of the human T-lymphocyte-derived Jurkat cells [14]. Jurkat cells are a suitable immortalized model to analyze T cell responses associated with signaling pathways; moreover, unlike other T cell lines, they can be activated and led to synthesize and secrete IL-2 and IFN-γ (e.g., [15,16]). This in vitro model allowed us to investigate the molecular and cellular changes induced by O 3 treatment under controlled experimental conditions, avoiding the complex interactions with multiple factors as it occurs in a living organism. In addition, a stabilized T cell line ensures standardized samples in comparison to primary cultured lymphocytes, thus making our model fully compliant with the highly sensitive ultrastructural, biomolecular, and bioanalytic techniques used in the present work.
As previously reported [12], samples of 4 × 10 6 cells suspended in 10 mL medium were collected into a 20 mL polypropylene syringe, then 10 mL of gas was added in the syringe by using a sterile filter (Alnitec s.r.l.). The cell suspension was gently mixed with the gas for 10 min to allow the full reaction of cells with O 3 [21]. Some samples were exposed to pure O 2 to distinguish between O 3 and O 2 effects, while other samples were exposed to air as controls.
After gas exposure, Jurkat cells were seeded in plastic flasks and analyzed at increasing times to evaluate some structural and functional features (see below).
In order to verify if the activation state of T lymphocytes may influence the effects of gas treatment on cytokine secretion, Jurkat cells were pre-incubated with 1 µg/mL phytohemagglutinin (PHA) + 50 ng/mL phorbol 12-myristate 13-acetate (PMA) for 24 h, as previously reported [22][23][24]. Then, the cells were treated as described above.

Cell Viability
To assess the effect of gas treatment on cell viability, the trypan blue exclusion test was used. This test is based on the principle that the dye is excluded from live cells, which have intact plasma membranes, whereas it easily enters dead cells [25]. Cell death rate was estimated after 2 h, 24 h, and 48 h from gas treatment by incubating the cells for 2 min with 0.1% trypan blue in the culture medium. The cells were counted in three randomly chosen microscope fields (20×) with a Leica DM IL inverted microscope (Leica Microsystems, Wetzlar, Germany), and the percentage of cells stained with trypan blue was estimated. Three independent experiments were performed.
For a proper interpretation of data on cytokine secretion, cell viability was also assessed for PHA/PMA-activated Jurkat cells.

Cell Growth
To evaluate cell growth, 4 × 10 5 cells/mL were seeded in T25 plastic flasks and the total cell number was evaluated after 2 h, 24 h and 48 h from the treatment by counting viable cells in a Burker hemocytometer in the presence of 0.1% trypan blue in the culture medium. Three independent experiments were performed.
For a proper interpretation of data on cytokines secretion, cell growth was also evaluated for PHA/PMA-activated Jurkat cells.

Transmission Electron Microscopy
Morphological and immunocytochemical analyses were carried out at transmission electron microscopy in order to analyze the effects of the exposure to low O 3 concentrations on the fine cell features and Nrf2 nuclear translocation. Based on our previous investigations [19], the effects were evaluated after 24 h from treatment in order to clearly detect morphological changes and Nrf2 translocation.
To avoid possible morphological alterations due to pre-fixation cell handling, Jurkat cells were fixed in their culture medium by adding an equal amount of a fixative solution made of 5% glutaraldehyde and 4% paraformaldehyde in phosphate buffered saline (PBS). In detail, cell samples were fixed for 2 h at 4 • C; then, cell suspensions were centrifuged, the supernatant was removed, and the samples were processed as cell pellets. After rinsing with PBS, the samples were post-fixed with 1% OsO 4 and 1.5% K 4 Fe(CN) 6 for 2 h at 4 • C, dehydrated in acetone, and embedded in Epon resin.
For ultrastructural morphology, ultrathin sections were collected on copper grids coated with a Formvar layer and stained with Reynolds lead citrate.
For immunocytochemistry, ultrathin sections were collected on nickel grids coated with a Formvar-carbon layer. Before immunostaining, the sections were treated with a 0.2 M aqueous solution of NaIO 4 for 60 min to improve the immune reaction [26,27]. Sections were then briefly floated on normal goat serum diluted 1:100 in PBS, incubated overnight at 4 • C with the anti-Nrf2 antibody (Abcam #ab62352, Cambridge, United Kingdom) diluted 1:5 with PBS containing 0.1% bovine serum albumin (Fluka, Buchs, Switzerland) and 0.05% Tween 20. Sections were then floated on normal goat serum and incubated for 30 min with a goat anti-rabbit IgG secondary antibody conjugated with 12-nm gold particles (Jackson ImmunoResearch Laboratories Inc., West Grove, PA, USA), diluted 1:20 in PBS. After rinsing with PBS and water, the sections were finally air-dried. As immunostaining controls, the primary antibody was omitted.
The samples were observed in a Philips Morgagni transmission electron microscope (FEI Company Italia Srl, Milan, Italy) operating at 80 kV; a Megaview III camera (FEI Company Italia Srl) was used for image acquisition.
Quantitation of anti-Nrf2 immunolabeling was performed by estimating the gold particle density on sections treated in the same run: the area of nucleoplasmic regions and resin regions (as an intra-sample negative control) was measured on 30 micrographs (28,000×) per sample. In samples treated with 30 µg O 3, obviously necrotic cells were not considered. Background evaluation was performed in sections processed for immunocytochemistry without the primary antibody. In each measured area, the gold particles were counted manually and the labeling density (i.e., the number of gold particles/µm 2 of nucleoplasm) was calculated.

IL-2 and IFN-γ Secretion
The amount of interleukin (IL)-2 and interferon (IFN)-γ secretion was evaluated in the culture medium after 24 h from the gas treatment, when the peak concentration after PHA/PMA stimulation is known to occur [23,28,29]. For each sample, 4 × 10 5 cells/mL were treated with gas as described above; experiments were performed four times per sample. The samples were centrifuged at 200 g for 8 min, the media were collected from the cell pellet, centrifuged again at 1500 g for 15 min, and the supernatants were finally stored at −80 • C. Quantitation of IL-2 and IFN-γ was conducted on a Luminex Biorad Bio-Plex 100 instrument (Bio-Rad Laboratories S.r.l., Segrate, MI, Italy) coupled to Bio-Plex Manager software v6.0, which allows measuring multiple proteins in a single well. Briefly, 50 µL aliquots of undiluted cell medium were put in a 96-well plate (samples were run in duplicate). Superparamagnetic microspheres (beads) conjugated with fluorophores and antibodies against IL-2 and IFN-γ were added to the assay wells. Incubation and washing steps were performed as per manufacturer's recommendations, then the plate was loaded into the Luminex system for reading and signal quantitation.

Statistical Analyses
Data for each variable were presented as mean ± standard deviation of the mean (SD). Statistical comparison was performed by either the analysis of variance (ANOVA) test for linear trend (for RT-PCR) or the one-way ANOVA followed by Bonferroni's post-hoc test (for all other variables); p ≤ 0.05 was considered as statistically significant.

Cell Growth
The number of Jurkat cells significantly decreased after 24 and 48 h from treatment with 20 µ g O3 and 30 µ g O3 in comparison to the respective controls ( Figure 2).

Cell Growth
The number of Jurkat cells significantly decreased after 24 and 48 h from treatment with 20 µg O 3 and 30 µg O 3 in comparison to the respective controls ( Figure 2). ure 1. Diagrams show the mean value ± SD of percentage of dead cells in the samples after 2 h, 24 h, and 48 h from gas tment. The asterisk (*) indicates significant difference with the respective control (CT).

Cell Growth
The number of Jurkat cells significantly decreased after 24 and 48 h from treatment with 20 µ g O3 and 30 µ g O3 in comparison to the respective controls ( Figure 2).

Transmission Electron Microscopy
At TEM, control Jurkat cells showed a roundish shape, many microvilli on their surface, and one large nucleus of irregular shape (Figure 3a). In the cytoplasm, ovoid mitochondria with lamellar cristae were numerous, the Golgi apparatus and the rough endoplasmic reticulum were well developed, and large amounts of free ribosomes were scattered in the cytosol (Figure 3a'). Conversely, the smooth endoplasmic reticulum and secondary lysosomes/residual bodies were quite rare. After treatment with O2, the cells appeared unaltered, apart from the presence of some swollen mitochondria with scarce cristae (Figure 3b

Transmission Electron Microscopy
At TEM, control Jurkat cells showed a roundish shape, many microvilli on their surface, and one large nucleus of irregular shape (Figure 3a). In the cytoplasm, ovoid mitochondria with lamellar cristae were numerous, the Golgi apparatus and the rough endoplasmic reticulum were well developed, and large amounts of free ribosomes were scattered in the cytosol (Figure 3a'). Conversely, the smooth endoplasmic reticulum and secondary lysosomes/residual bodies were quite rare. After treatment with O 2 , the cells appeared unaltered, apart from the presence of some swollen mitochondria with scarce cristae (Figure 3b with dense matrix and hardly visible cristae, whereas the other cytoplasmic organelles did not undergo evident alterations (Figure 3e,e',f). In addition, some necrotic cells were found (Figure 3g), whereas apoptotic cells were never observed. To assess whether the O3 treatment might affect the nuclear distribution of Nrf2, we investigated the ultrastructural immunostaining of Jurkat cell nuclei. In all samples, Nrf2 was specifically located along the borders of the heterochromatin clumps (Figure 4a-e). Quantitative evaluation revealed a statistically significant increase in nucleoplasmic anti-Nrf2 labeling density (p < 0.001, one-way ANOVA) in all cell samples treated with O3 in comparison to control, while O2-treated cells showed values similar to control (Figure 4f). Background was negligible (0.03 ± 0.09 gold particles/µ m 2 ).

Quantitative Real-Time Polymerase Chain Reaction
qPCR demonstrated a significant upregulation of Hmox1 in Jurkat cells treated with O3 in comparison to control and O2-treated cells. In particular, Hmox1 increased in proportion to the O3 concentration, as demonstrated by the ANOVA test for linear trend in Figure 5.

IL-2 and IFN-γ Secretion
The amount of IL-2 and IFN-γ was assessed in the culture medium of both non-activated and activated Jurkat cells ( Figure 6). In non-activated cells, the amount of IL-2 significantly increased in the medium of all treated cells in comparison to control. After activation, the amount of IL-2 detected in the medium of activated control cells was significantly higher in comparison to the non-activated control samples (p < 0.001), but no significant modification due to gas treatment was detected among the activated cells samples.
The amount of IFN-γ was lower than the detection limit in all the samples. For a correct interpretation of the results on IL-2 and IFN-γ secretion, cell death and growth were assessed also after PHA/PMA activation. The results were quite similar to those obtained with non-activated cells.

IL-2 and IFN-γ Secretion
The amount of IL-2 and IFN-γ was assessed in the culture medium of both nonactivated and activated Jurkat cells ( Figure 6).

IL-2 and IFN-γ Secretion
The amount of IL-2 and IFN-γ was assessed in the culture medium of both non-activated and activated Jurkat cells ( Figure 6). In non-activated cells, the amount of IL-2 significantly increased in the medium of all treated cells in comparison to control. After activation, the amount of IL-2 detected in the medium of activated control cells was significantly higher in comparison to the non-activated control samples (p < 0.001), but no significant modification due to gas treatment was detected among the activated cells samples.
The amount of IFN-γ was lower than the detection limit in all the samples. For a correct interpretation of the results on IL-2 and IFN-γ secretion, cell death and growth were assessed also after PHA/PMA activation. The results were quite similar to those obtained with non-activated cells.
The death rate of PHA/PMA-activated Jurkat cells did not differ in control and treated samples (O2, 10 µ g O3, 20 µ g O3, 30 µ g O3) at all the considered time points (2, 24, In non-activated cells, the amount of IL-2 significantly increased in the medium of all treated cells in comparison to control. After activation, the amount of IL-2 detected in the medium of activated control cells was significantly higher in comparison to the nonactivated control samples (p < 0.001), but no significant modification due to gas treatment was detected among the activated cells samples.
The amount of IFN-γ was lower than the detection limit in all the samples.

Discussion
In this study, Jurkat cells were treated with gaseous O2-O3 mixtures at the O3 concentrations currently used for systemic administration by major autohemotherapy, with the aim of investigating their differential effects on some structural and functional features of T lymphocytes. Jurkat cells are from a cell line derived from a patient affected by acute lymphoblastic leukemia and leukemic transformed non-Hodgkin lymphoma [14]; however, during the last several decades it has been widely used as reliable in vitro model for scientific studies not only on T cell leukemia but also on T cell signaling and response to various stimuli, including antioxidant and anti-inflammatory compounds e.g., [30][31][32][33][34] as well as ozonation [21]. On this basis, Jurkat cells were chosen for the present study as a

Discussion
In this study, Jurkat cells were treated with gaseous O2-O3 mixtures at the O3 concentrations currently used for systemic administration by major autohemotherapy, with the aim of investigating their differential effects on some structural and functional features of T lymphocytes. Jurkat cells are from a cell line derived from a patient affected by acute lymphoblastic leukemia and leukemic transformed non-Hodgkin lymphoma [14]; however, during the last several decades it has been widely used as reliable in vitro model for scientific studies not only on T cell leukemia but also on T cell signaling and response to various stimuli, including antioxidant and anti-inflammatory compounds e.g., [30][31][32][33][34] as well as ozonation [21]. On this basis, Jurkat cells were chosen for the present study as a

Discussion
In this study, Jurkat cells were treated with gaseous O 2 -O 3 mixtures at the O 3 concentrations currently used for systemic administration by major autohemotherapy, with the aim of investigating their differential effects on some structural and functional features of T lymphocytes. Jurkat cells are from a cell line derived from a patient affected by acute lymphoblastic leukemia and leukemic transformed non-Hodgkin lymphoma [14]; however, during the last several decades it has been widely used as reliable in vitro model for scientific studies not only on T cell leukemia but also on T cell signaling and response to various stimuli, including antioxidant and anti-inflammatory compounds e.g., [30][31][32][33][34] as well as ozonation [21]. On this basis, Jurkat cells were chosen for the present study as a suitable in vitro model to investigate the molecular and cellular effects of low O 3 concentrations, with special focus on the capability of O 3 to influence IL-2 and IFN-γ secretion in both nonactivated and activated T cells. In fact, the knowledge of the relationship between O 3 dose and T lymphocyte response in term of antioxidant activity and immune activation may give a significant contribution to improve the curative potential of major autohemotherapy.
The trypan blue exclusion test revealed that, under our experimental conditions, the gas exposure (O 2 , 10 µg O 3 , 20 µg O 3 , and 30 µg O 3 ) did not induce significant alteration in Jurkat cell viability at both short (2 h) and medium (24 h) term. However, at 48 h post-treatment, exposure to 30 µg O 3 resulted in a significant increase of cell death, whereas the other gases (O 2 , 10 µg O 3 , and 20 µg O 3 ) confirmed their safety at long term. These data demonstrate that the treatment with 10 or 20 µg O 3 does not induce appreciable cell death in T lymphocytes in vitro, consistently with previous findings in different cultured cells [17,18,35] as well as in explanted adipose tissue [20]. In addition, these data suggest that cell death induced by the exposure to 30 µg O 3 is not an acute event but probably occurs due to the cellular failure to repair the damages caused by the initial stress. This hypothesis is consistent with the finding that in samples treated with 30 µg O 3 , the cell population significantly decreased after 24 h from the gas exposure (indicating a reduced cell proliferation in the absence of cell death) and further declined after 48 h (when the increased death rate contributed to cell loss).
Interestingly, the treatment with 20 µg O 3 also induced a significant decrease in cell population after both 24 and 48 h, suggesting the occurrence of some cell stress responsible for a reduction in cell proliferation but not in cell viability. Our data on cell proliferation are consistent with previous findings on Jurkat cells [21] as well as on peripheral blood mononuclear cells [36] demonstrating that proliferation rate is inversely related to O 3 concentration.
According to the results on cell viability and proliferation, TEM observations showed that, after 24 h post-treatment with 20 µg O 3 , Jurkat cells were vital but showed mitochondrial alterations, while after exposure to 30 µg O 3 , many vacuoles also occurred in most cells. Moreover, in 30 µg O 3 -treated samples some necrotic cells were found, demonstrating cell stress and damage, whereas no sign of apoptosis was found consistently with previous studies excluding apoptogenic effects of low O 3 concentrations [17,18,20].
Mitochondrial damage usually occurs when respiratory capability is reduced or altered [32], and reactive oxygen species regulate mitochondrial functions (review in [37,38]). Therefore, the oxidative stress due to 20 µg O 3 and 30 µg O 3 could induce various degrees of functional alterations that could also explain the lowering in cell population observed in 20 µg O 3 and 30 µg O 3 -treated samples at long incubation times. Some mitochondrial alterations were also found in cells treated with pure O 2 , probably due again to an excessive oxidative stress; however mitochondrial damage was less frequent than in O 3 -treated cells and was not accompanied by a decrease in cell viability or proliferation at any time point considered. On the other hand, no ultrastructural alteration was ever found in 10 µg O 3 -treated cells up to 48 h, indicating that, under our experimental conditions, this O 3 concentration was the safest for T lymphocytes. Moreover, experimental evidence proved that appropriate O 3 concentrations might even improve mitochondrial activity [39,40].
Nrf2 is a transcription factor and its presence along the border of heterochromatin clumps is consistent with its functional role since this nuclear region contains perichromatin fibrils i.e., the in situ form of transcription, splicing, and 3 -end processing of pre-mRNAs [41][42][43]. Interestingly, the amount of Nrf2 markedly increased in the nuclei of all O 3 -treated cell samples in comparison to control, whereas no modification occurred in O 2 -treated samples. It has been demonstrated that the mild oxidative stress caused by the exposure to low O 3 concentrations induces the dissociation of Nrf2 from its negative regulator Keap1; then, Nrf2 rapidly moves from the cytoplasm into the nucleus, where it activates the expression of antioxidant response elements (ARE)-driven genes [19,44]. This mechanism ensures an efficient transcription of antioxidant genes through the mobilization of the cytoplasmic Nrf2, avoiding its de novo synthesis. The exposure to 10, 20, and 30 µg O 3 therefore proved to stimulate in Jurkat cells a stronger nuclear translocation of Nrf2 than treatment with pure O 2 . The expression of Hmox1 i.e., the gene encoding for the stressprotective protein heme oxygenase-1 involved in the regulatory pathways of antioxidant response, inflammation, apoptosis and angiogenesis [45], proved to be upregulated following exposure to low O 3 concentrations [28]. Consistently, Hmox1 expression significantly increased in O 3 -treated Jurkat cells in dose-dependent manner in comparison to control, whereas pure O 2 had no effect. This finding is consistent with the O 3 dose-dependent increase of some antioxidant enzyme activity previously reported in Jurkat cells [21].
Gas treatments proved also to increase the secretion of IL-2 in Jurkat cells. IL-2 is a cytokine secreted by antigen-activated T cells and is characterized by a wide range of actions, including the ability to stimulate the proliferation of T cells [46], improve the cytolytic activity of natural killer (NK) cells and tumor-infiltrating lymphocytes, increase immunoglobulin production by activated B cells, and act on innate lymphoid cells, memory T cells, effector T cells, and monocytes (review in [47]). In particular, IL-2 plays a key role in the control of the immune response by opposite functions: it promotes effector T cells (as an immunostimulatory factor) and maintains homeostatic proliferation of regulatory T cells (as an immunoinhibitory factor). In this view, both pure O 2 and O 2 -O 3 mixtures with low ozone concentrations may be envisaged as a stimulating factor of T lymphocytes, likely due to their ability to induce oxidative stress. It is in fact known that reactive oxygen species, through the Nrf2-dependent activation of ARE-dependent genes [44], may act as regulators of the inflammatory process by triggering several signaling cascades, which involve the expression of several antioxidant enzymes, the inhibition of pro-inflammatory and the stimulation of anti-inflammatory cytokines and, in particular, the activation of T cells [48][49][50]. This finding is consistent with clinical and experimental evidence that the O 2 -O 3 therapy contributes in modulating inflammatory cytokines in various pathological conditions (recent reviews in [51][52][53][54][55][56][57]).
In the frame of its multiple roles in immune response, IL-2 is involved also in wound healing, by acting both systemically and at the wound site. In particular, IL-2 signaling is fundamental at the early phase by drawing the immune mediators necessary to start the wound healing, by limiting microbial infections and promoting epithelial and endothelial cell proliferation, while later the decrease of IL-2 plays an anti-inflammatory effect by attracting and stimulating regulatory T cells, thus avoiding scar formation (review in [58]). The stimulating effect on IL-2 secretion by O 2 -O 3 gas treatments therefore provides a scientific support to the success of O 2 -O 3 therapy in improving healing in various pathological conditions, such as skin and oral diseases, diabetic foot, and peripheral arterial disease (reviews in [53,57,[59][60][61][62]).
On the other hand, gas treatments were unable to affect the secretion of IFN-γ, whose levels remained always below the detection limit. Probably, the level of reactive oxygen species necessary for stimulating Jurkat cells to secrete IFN-γ must be higher than the mild oxidation induced in our experiments, as suggested by experiments with ionizing radiations [63].
In sum, the present findings provide evidence that the cytoprotective mechanisms promoted by a mild oxidative stress includes the activation of T lymphocytes, which are stimulated to secrete IL-2, but not IFN-γ. However, if Jurkat cells were previously activated with PHA/PMA, gas treatments had no effect on IL-2 and IFN-γ secretion. It is worth noting that activated Jurkat cells behaved similarly to non-activated cells in response to gas treatments as they did to cell viability and growth, suggesting that activation does not alter T lymphocyte sensitivity to oxidative stress.
As expected, PHA/PMA-activated control Jurkat cells secreted significantly higher amounts of IL-2 than non-activated control, but all gas-treated activated Jurkat cells showed a secretory activity similar to activated control. This suggests that, under our experimental conditions, the mild oxidative stress caused by gas treatments was unable to promote further IL-2 secretion by already activated T cells, thus avoiding exacerbation of the immune response. As for IFN-γ, the PHA/PMA stimulation was unable to increase the secretion of this cytokine by Jurkat cells, according to previous observations [28]. Similarly, gas treatments were ineffective in stimulating IFN-γ secretion by activated cells, probably because of the too weak oxidative stress induced as supposed for IL-2.

Conclusions
In our in vitro study, we tested on Jurkat cells the effects of the O 2 -O 3 mixtures usually administered for major autohemotherapy in humans. Our findings demonstrate that 10, 20, and 30 µg O 3 concentrations are all able to trigger the Nrf2-induced activation of antioxidant cytoprotective mechanisms. Moreover, all O 2 -O 3 mixtures tested as well as pure O 2 proved to increase the secretion of IL-2, suggesting that T lymphocyte activation may be induced by a mild oxidative stress irrespective of the presence of O 3 . Importantly, the mild oxidative stress induced by these gas treatments do not exert stimulating effects in already activated Jurkat cells; this evidence suggests that the O 3 -induced lymphocyte hyperactivation is unlikely to occur.
However, the viability and proliferation tests as well as the fine morphological analysis at transmission electron microscopy revealed cell alterations and damage after exposure to 30 µg O 3 and, at a lesser extent, to 20 µg O 3 , thus designating 10 µg O 3 as the most suitable O 3 concentration to combine cell safety and efficient antioxidant and immune response in Jurkat cells.
Although in the patient, the effects of O 3 are mediated by many blood and tissue factors, our basic study offers novel evidence of the fine regulatory role played by the oxidative stress level in the hormetic response of T lymphocytes to O 2 -O 3 administration. Further research is necessary to fully elucidate the correlation between O 3 concentration and the modulation of immune cell activity with the aim of obtaining the best therapeutic effects while avoiding cell damage. Achieving this knowledge would pave the way to the development of specific protocols for different inflammatory conditions and to personalized applications of O 2 -O 3 therapy.