Modifications of Blood Molecular Components after Treatment with Low Ozone Concentrations

The ex vivo treatment of a limited volume of blood with gaseous oxygen–ozone (O2–O3) mixtures and its rapid reinfusion into the patient is a widespread medical procedure. O3 instantly reacts with the blood’s antioxidant systems, disappearing before reinfusion, although the molecules formed act as messengers in the organism, inducing multiple antioxidant and anti-inflammatory responses. An appropriate dose of O3 is obviously essential to ensure both safety and therapeutic efficacy, and in recent years, the low-dose O3 concept has led to a significant reduction in the administered O3 concentrations. However, the molecular events triggered by such low concentrations in the blood still need to be fully elucidated. In this basic study, we analysed the molecular modifications induced ex vivo in sheep blood by 5 and 10 µg O3/mL O2 by means of a powerful metabolomics analysis in association with haemogas, light microscopy and bioanalytical assays. This combined approach revealed increased oxygenation and an increased antioxidant capacity in the O3-treated blood, which accorded with the literature. Moreover, original information was obtained on the impact of these low O3 concentrations on the metabolic pathways of amino acids, carbohydrates, lipids and nucleotides, with the modified metabolites being mostly involved in the preservation of the oxidant–antioxidant balance and in energy production.


Introduction
Ozone (O 3 ) is an unstable gas occurring in the Earth's atmosphere, where it naturally forms from oxygen (O 2 ) due to the action of ultraviolet light and electrical discharges and then rapidly decomposes into O 2 .Its high oxidation power makes O 3 harmful to organisms; however, if appropriately applied, it may be used as a therapeutic agent.The beneficial properties of O 3 have been known since the 18th century and, starting from the 19th century, its use in medicine has progressively expanded for the treatment of an increasing number of diseases through different administration routes [1].
The ex vivo treatment of a limited volume (100-200 mL) of whole blood with gaseous O 2 -O 3 mixtures and rapid reinfusion into the patient via the venous route is a widespread medical procedure first described by the Austrian physician H. Wolff in the 1970s [2].This procedure was initially applied on an empirical basis but, starting from the 1990s, with the advent of O 3 generators able to produce photometrically-measured O 3 concentrations in real time and in a given gas volume, the administration protocols were refined via precisely determining the actually administered O 3 dosages.Notably, this also allowed conducting accurate scientific investigations into the effect of O 3 on blood components, thus unveiling the basic molecular mechanisms involved in its therapeutic potential [3][4][5].
Scientific research has demonstrated that, if administered to blood at appropriate (low) doses, the deleterious oxidising effect of O 3 can be neutralised by the plasmatic antioxidant factors, such as uric acid, ascorbic acid, glutathione (GSH), albumin and lipophilic molecules, as well as by the great reservoir of GSH and other antioxidant enzymes located in the erythrocytes [4,6].In the blood, the O 2 -O 3 mixture immediately dissolves into the water of the plasma: O 2 fully oxygenates haemoglobin while O 3 instantly reacts with hydrophilic antioxidants and polyunsaturated fatty acids (PUFA), giving rise to the formation of H 2 O 2 and a variety of lipid oxidation products (LOPs) [7][8][9].Since these reactions take place in a few seconds, the O 3 present in the administered gas mixture disappears in 2-5 min from the blood contained in the bottle before its reinfusion, thus never entering circulation after blood reinfusion.However, as a result of these reactions, O 3 causes a small and transient depletion of antioxidants and a plasmatic increase in reactive oxygen species (ROS) and LOPs [10], which after blood reinfusion act as messengers in the whole organism, activating biochemical and immunological pathways and initiating cascades of biological events (e.g., production of growth factors and cytokines, upregulation of many antioxidant enzymes) in various tissues and organs (exhaustive reviews in [3,4,6,11]).Therefore, the therapeutic potential of blood ozonation observed in multiple diseases (recent publications in [11][12][13][14][15][16][17]) relies on biological events triggered by O 3 ex vivo in the drawn blood and then accomplished in the organism by the O 3 derivatives generated as a physiological response to the mild oxidative stress [3][4][5]18].
An appropriate dose of gaseous O 3 is essential to ensure both safety and therapeutic efficacy because it must not exceed the blood antioxidant potential but, at the same time, it must generate enough molecular messengers to induce beneficial biological effects.A concentration window of 20-80 µg O 3 /mL O 2 was identified as suitable in the 1990s [19,20] but, during the last decade, the clinical experience and the scientific evidence led to a significant reduction in the concentration of the administered O 3 while maintaining its therapeutic efficacy, according to the low-dose O 3 concept [21,22], although the molecular effects of such low concentrations on blood still need to be fully elucidated.
In light of this, in order to provide additional knowledge of the molecular events triggered by O 3 in blood, we conducted a basic study analysing the molecular modifications induced ex vivo by gaseous O 2 -O 3 mixtures at O 3 concentrations of 5 µg and 10 µg.To ensure highly controlled experimental conditions when administering such low O 3 concentrations, a next-generation O 3 generator was used and a powerful metabolomics analysis was applied in association with haemogas analysis, light microscopy and bioanalytical assays.

Haemogas Analysis
The haemogas analysis showed that both 5 µg O 3 and 10 µg O 3 induced a significant increase in pO 2 (p = 0.036 and p = 0.027, respectively) (Figure 1a) and pH (p = 0.028 and p = 0.026, respectively) (Figure 1c) compared to the controls.On the contrary, a decrease in pCO 2 was observed in the 5 µg O 3 -and 10 µg O 3 -treated samples in comparison to the controls (p = 0.028 for both concentrations) (Figure 1b).

Haemolysis
The blood haemolysis assessment, carried out via a colorimetric assay for th globin evaluation, revealed no statistically significant difference in the O2-and treated samples in comparison to the control values, while a significant increas molysis was found in the blood sample treated with 5 µg O3 (p = 0.008) (

Haemolysis
The blood haemolysis assessment, carried out via a colorimetric assay for the haemoglobin evaluation, revealed no statistically significant difference in the O 2 -and 10 µg O 3 -treated samples in comparison to the control values, while a significant increase in haemolysis was found in the blood sample treated with 5 µg O 3 (p = 0.008) (Figure 2).
The blood smears showed in all the samples well-preserved erythrocytes with no evident morphological sign of damage (Figure 3).
The blood haemolysis assessment, carried out via a colorimetric assay for the haemoglobin evaluation, revealed no statistically significant difference in the O2-and 10 µg O3treated samples in comparison to the control values, while a significant increase in haemolysis was found in the blood sample treated with 5 µg O3 (p = 0.008) (Figure 2).The blood smears showed in all the samples well-preserved erythrocytes with no evident morphological sign of damage (Figure 3).

Total Antioxidant Capacity
The total antioxidant capacity, evaluated via a colorimetric assay for the measurement of the antioxidant proteins and/or small molecules, significantly increased in the 5 µg O3-and 10 µg O3-treated blood samples in comparison to the controls (p = 0.004 for both concentrations); conversely, O2 significantly reduced the antioxidant capacity in comparison to the control (p = 0.011) (Figure 4).In addition, the antioxidant capacity of the 5 µg O3-and 10 µg O3-treated samples was significantly higher than in the O2-treated samples (p = 0.003 and p = 0.004, respectively).

Interleukins
The amounts of the interleukins IL-2 and IL-6 were found to be unchanged by any treatment in comparison to the control samples (Figure 5).

Total Antioxidant Capacity
The total antioxidant capacity, evaluated via a colorimetric assay for the measurement of the antioxidant proteins and/or small molecules, significantly increased in the 5 µg O 3 -and 10 µg O 3 -treated blood samples in comparison to the controls (p = 0.004 for both concentrations); conversely, O 2 significantly reduced the antioxidant capacity in comparison to the control (p = 0.011) (Figure 4).In addition, the antioxidant capacity of the 5 µg O 3 -and 10 µg O 3 -treated samples was significantly higher than in the O 2 -treated samples (p = 0.003 and p = 0.004, respectively).The blood smears showed in all the samples well-preserved erythrocytes with no evident morphological sign of damage (Figure 3).

Total Antioxidant Capacity
The total antioxidant capacity, evaluated via a colorimetric assay for the measurement of the antioxidant proteins and/or small molecules, significantly increased in the 5 µg O3-and 10 µg O3-treated blood samples in comparison to the controls (p = 0.004 for both concentrations); conversely, O2 significantly reduced the antioxidant capacity in comparison to the control (p = 0.011) (Figure 4).In addition, the antioxidant capacity of the 5 µg O3-and 10 µg O3-treated samples was significantly higher than in the O2-treated samples (p = 0.003 and p = 0.004, respectively).

Interleukins
The amounts of the interleukins IL-2 and IL-6 were found to be unchanged by any treatment in comparison to the control samples (Figure 5).

Interleukins
The amounts of the interleukins IL-2 and IL-6 were found to be unchanged by any treatment in comparison to the control samples (Figure 5).

Metabolomics
Untargeted metabolomic analysis showed that both the 5 µg O3 and 10 µg O3 treatments induced significant molecular changes compared to the controls.Principal component analysis (PCA) and the hierarchical clustering heatmap (Figure 6) clearly highlighted the presence of a metabolic signature associated with O3 treatment.In addition, both PCA and clustering analysis showed the high reproducibility of the treatment, since all the resultant replicates were well grouped together.

Metabolomics
Untargeted metabolomic analysis showed that both the 5 µg O 3 and 10 µg O 3 treatments induced significant molecular changes compared to the controls.Principal component analysis (PCA) and the hierarchical clustering heatmap (Figure 6) clearly highlighted the presence of a metabolic signature associated with O 3 treatment.In addition, both PCA and clustering analysis showed the high reproducibility of the treatment, since all the resultant replicates were well grouped together.

Metabolomics
Untargeted metabolomic analysis showed that both the 5 µg O3 and 10 µg O3 treatments induced significant molecular changes compared to the controls.Principal component analysis (PCA) and the hierarchical clustering heatmap (Figure 6) clearly highlighted the presence of a metabolic signature associated with O3 treatment.In addition, both PCA and clustering analysis showed the high reproducibility of the treatment, since all the resultant replicates were well grouped together.A total of 572 metabolites were identified (Table S1).Since O 3 was administered as an O 2 -O 3 mixture, the metabolites showing a difference in comparison to the control after treatment with pure O 2 were not considered.
In comparison to the control, 25 metabolites showed statistically significant modifications after treatment with 5 µg O 3 , and 89 metabolites after treatment with 10 µg O 3 ; 31 metabolites showed a statistical difference with the control after treatment with both 5 µg O 3 and 10 µg O 3 .Among these metabolites, only those showing a fold change ≥ 1.3 or ≤0.769 were considered.Moreover, metabolites corresponding to exposomes or specific food components were excluded.
As the final result of this selection process, a list of molecules involved in protein, carbohydrate, lipid and nucleotide metabolism was obtained: some of them underwent a decrease while others increased (Table 1).

Discussion
The molecular modifications observed in the whole blood exposed to 5 µg O 3 or 10 µg O 3 demonstrate that these low O 3 concentrations are able to significantly act on multiple chemical features and metabolic pathways.

The Impact on Blood Oxygenation and Haemolysis
According to the well-known effects of blood ozonation, pO 2 increased in both the O 3treated blood samples [23][24][25], whereas pCO 2 decreased [25].The increase in the pH value in the O 3 -treated samples is consistent with the concomitant increase in pO 2 .The reason for the higher pO 2 value observed in the O 3 -treated samples in comparison to the O 2 -treated samples remains unclear.However, it should be underlined that the hyperoxygenation of ozonated blood occurring in the bottle is considered clinically irrelevant because this limited amount of blood (100-200 mL) is reinfused via the venous route during the following 15-20 min and is heavily diluted in the venous blood, which has a pO 2 of about 40 mmHg, so that the final venous pO 2 is hardly modified [3][4][5]24].
A limited but statistically significant increase in haemolysis was found in the blood samples treated with 5 µg O 3 , in the absence of microscopically detectable damage to the erythrocytes.Accordingly, a slight increase in haemolysis has been frequently reported after O 3 treatment and has been always considered negligible [10,20,23,24,26].

The Increase in Blood Antioxidant Capacity
The increase in the antioxidant capacity has been considered for decades the rationale behind the therapeutic efficacy of blood ozonation [5,19], and the evidence in our O 3treated blood samples is consistent with previous results obtained in different experimental models.It has been reported that the treatment of blood with therapeutic doses of O 3 , after causing an initial slight and transient decrease in the antioxidant capacity of the plasma (fully reconstituted within a few minutes) [5], gives rise to a prompt plasmatic and cellular antioxidant response [4,18] that is likely responsible for the increased antioxidant capacity in our O 3 -treated samples.In particular, it has already been experimentally demonstrated that an O 3 concentration of 10 µg is able to stimulate an antioxidant cytoprotective response through the activation of the Keap1-dependent nuclear factor erythroid 2-related factor 2 (Nrf2) pathway [27][28][29][30][31][32][33][34][35].Interestingly, in the present study, we found that also the very low concentration of 5 µg O 3 is able to induce a significant increase in the antioxidant capacity.On the other hand, pure O 2 not only is unable to increase the antioxidant capacity but even induces its significant decrease, demonstrating that the O 2 -derived oxidative stress cannot activate an antioxidant response as low O 3 doses actually do.

The Unchanged Levels of IL-2 and IL-6
Under our experimental conditions, O 3 treatment did not modify the plasmatic amounts of IL-2 and IL-6, both characterized by a wide range of actions in the immune response.The literature data on the effect of therapeutic doses of O 3 on cytokine secretion are heterogeneous: some studies described a stimulating action on leukocytes with increased IL-2 and IL-6 secretion [10,33,36], whereas others reported unchanged levels of IL-2 and IL-6 [37].These discrepancies could be related to the different experimental conditions used, such as the O 3 concentration, the treatment in vitro or in vivo, and the detection techniques.Moreover, it has been reported that the ability of low O 3 concentrations to stimulate IL-2 secretion from T lymphocytes is related to their activation status [33].

The Modifications of Amino Acid, Carbohydrate, Lipid and Nucleotide Metabolites
Metabolomic analysis provided original demonstrations of still unknown effects of low O 3 concentrations on various amino acid, carbohydrate, lipid and nucleotide metabolites, whose modifications may be interpreted in the frame of the biological effects induced by O 3 in blood.As discussed below, some of these metabolites underwent a quantitative decrease after O 3 treatment, maybe due to their involvement as "sacrificial" molecules in the maintenance of the oxidant-antioxidant balance [5] or as substrates for energy production; on the other hand, other metabolites showed a significant increase after O 3 treatment as factors required for the protective response against oxidative stress.
The modifications of the molecules following O 3 exposure are mainly due to the oxidation of amino acids in the free amino group [38,39] and their quantitative changes in the plasma are often related to the role of these molecules in the antioxidant response triggered by the oxidative effect of O 3 .N-acetyl-L-alanine is a biologically available N-terminal capped form of L-alanine, an amino acid that exerts an antioxidant action by promoting the expression of the proteins heme oxygenase-1 (HO-1) and ferritin [40]; therefore, the increase in N-acetyl-L-alanine after O 3 treatment could be due to its antioxidant role.Similarly, tryptophan metabolites play a role in the cellular redox response: 5-hydroxytryptophan was found to increase after O 3 treatment, and it has been demonstrated that this compound acts as an antioxidant by transferring electrons to free radicals and directly scavenging H 2 O 2 , which is the precursor of −OH in the Fenton reaction system [41,42].Tryptophan plays an important regulatory role in restoring the antioxidant system, enhancing the levels of GSH and glutathione peroxidase (GPx) in tissues [43], and the observed increase in picolinic acid, indoleacetic acid and indolelactic acid after O 3 exposure is probably due to the enhanced catabolism of tryptophan [35].The increase in ornithine and L-glutamic acid may contribute to the GSH synthesis, which is a key player in the cellular antioxidant response; its crucial role in the blood antioxidant response to O 3 treatment is well-known [3,4,18] and an increase in GSH was reported after treatment with low O 3 doses [44].Modifications of the blood levels of L-threonine and 2-butenedioic acid (also known as maleic acid) have been reported under treatment with anti-inflammatory agents acting through the Nrf2/HO-1 signalling pathway [45]; on this basis, the increase in these metabolites found in the O 3 -treated samples could be related to the anti-inflammatory action of low O 3 doses [5,46].Similarly, S-carboxymethyl-L-cysteine (also known as carbocysteine) exhibits free-radical scavenging and anti-inflammatory properties [47,48] and shows an increase in ozonated blood.Finally, proline and its related metabolite, pyrrole-2-carboxylic acid, are involved in redox homeostasis, and the increased levels of these molecules after O 3 treatment are likely linked to their protective effects against oxidative stress [49,50].
The modifications of molecules involved in carbohydrate metabolism are probably related to the stimulating effect of O 3 on the glycolytic pathways.In fact, the D-glucose decrease agrees with the O 3 ability to transiently increase the glycolysis rate with a consequent increase in the intracellular adenosine triphosphate (ATP) in erythrocytes [51].Accordingly, also the glucose derivative glucuronolactone decreases in ozonated blood.Deoxypentitol and arabinofuranose are sugar-derived plant compounds, and their decrease may again be related to their use as energetic substrates.In this way, the reduction in pentanedioic acid (also known as glutaric acid) and tyrosine could be due to their consumption to supply the tricarboxylic acid cycle.The increase in malic acid, an intermediate of the Krebs cycle, may be due to the increased energy metabolism; in addition, malic acid is able to exert an antioxidant action by inhibiting the superoxide anion and downregulating the tumour necrosis factor α [52,53].Aceturic acid (also known as N-acetyl glycine) is known to play a role in glucose metabolism, with its increase in blood having a strong correlation with oxidative metabolism, GSH biosynthesis and monosaccharide metabolism [54]; this could explain its increase in our O 3 -treated samples.
Lipid molecules as well can be used as energy sources in the ATP production process, so the reduction in some lipid metabolites observed after O 3 treatment may be due to their use as energy substrates.For instance, behenic and stearic acid (dietary fatty acids), campesterol (a phytosterol with a chemical structure similar to cholesterol), benzenepropanoic acid, and 3,5-bis(1,1-dimethylethyl)-4-hydroxy-methyl ester (a carboxylic ester derivative of a fatty acid) undergo reductions in O 3 -treated blood.Moreover, it is known that plasmatic unsaturated fatty acids are optimal substrates for O 3 , generating lipid peroxides that act as messengers to activate the Nrf2 response [11].In this context, we found an increase in 2-hydroxy-3-methylbutyric acid (also known as 3-hydroxyisovaleric acid), involved in lipid peroxidation and β-oxidation [55], and in pimelic acid, a product of partially β-oxidized dietary odd-chain fatty acids [56].Moreover, increased fatty acid oxidation has been found to increase the serum levels of oleic acid and linoleic acid [57], thus providing an explanation also for the observed increase in 9(E),11(E)-conjugated linoleic acid, trimethylsilyl ester (a linoleic acid derivative) in the O 3 -treated samples.Many lipid metabolites are involved in the antioxidant and anti-inflammatory pathways.Erythrono-1,4-lactone is an erythronicacid-derived compound that is known to increase under oxidation conditions [58], and we can hypothesise that its increase in the O 3 -treated samples is related to this mechanism.Similarly, the increase in 3-octenoic acid (also known as carylic acid) could be related to its anti-inflammatory properties [59].Butanoic acid, 2,4-bis[(trimethylsilyl)oxy]-, trimethylsilyl ester [60] and N-isobutyrylglycine [61] have been found to be upregulated under oxidative stress conditions, similarly to what was observed in our O 3 -treated samples.A decrease in (8Z,11Z,14Z)-icosa-8,11,14-trienoate (also known as 8,11,14-eicosatrienoic acid) has already been reported under oxidant-antioxidant altered conditions [62,63], probably due to its antioxidant activity [64,65].Similarly, 4-hydroxybenzeneacetic acid, a carboxylic acid (also known as mandelic acid) has been demonstrated to have antioxidant properties [66], and 9-octadecen-1-ol (also known as linolenyl alcohol), a long chain fatty primary alcohol, acts as an antibacterial [67] and anti-inflammation agent [68]: all these molecules decreased in ozonised blood.Arachidic acid (also known as eicosanoid acid) is a lipid mediator involved in haemodynamics and inflammation with both anti-inflammatory and protective properties [69,70] and lanopalmitic acid (also known as hydroxyhexadecanoic acid) is involved in the Nrf2/HO-1 signalling pathway [45]: both of them have been found to decrease in O 3 -treated blood.Similarly, the tartaric acid reduction may be due to its utilisation as an antioxidant agent [71].On the other hand, oleic acid, which has anti-inflammatory and antioxidant properties [72], and decanoic acid (also known as capric acid) able to reduce inflammatory cytokine production and oxidative stress [73] were found to increase in our O 3 -treated blood samples.Also, pentanoic acid (also known as valeric acid) that has antioxidant and anti-inflammatory properties [74] was found to increase in ozonated blood.
Concerning the nucleotide metabolites, our results showed a decrease in niacinamide (also known as nicotinamide or vitamin B3) and adenine, which are both precursors of nicotinamide adenine dinucleotide (NAD).More specifically, NAD may generate from two different pathways, one using tryptophan and the other using niacinamide [75].NAD has the capability of heightening the production of nicotinamide adenine dinucleotide phosphate (NADPH), in order to have reducing equivalents to enhance the antioxidant capacity and increase the GSH levels [76].The role of niacinamide in the blood antioxidant response has already been reported [3], and its decrease in ozonated blood could also be related to its antioxidant property [77].Glyceric acid, which we found to increase in the O 3 -treated samples, is a NAD/NADPH upregulator [78].It has been demonstrated that 9H-purin-6-ol (also known as hypoxantine) tends to increase in hypoxic conditions [79]; since in our O 3 -treated samples hypoxantine decreased, we may assume that this effect could be due to the hyperoxygenation induced by O 3 .The uridine decrease may be related to its capability to stabilise energy metabolism as well as to its antioxidant capacity [80,81].

Blood Ozonation
Due to the high amount of blood required for each experimental treatment (100 mL/sample), commercially available blood was used for this study.Sterile sheep blood was purchased from Microbiol Snc (Uta, CA, Italy).Coagulation was prevented via mechanical defibrination and the blood was used within 24 h from the collection.Before experimentation, the blood was heated to a temperature of 37 • C in a sterile incubator and then submitted to gaseous treatments.Aliquots of 100 mL were exposed to O 2 -O 3 gas mixtures produced from medical-grade O 2 by using an ECO3 apparatus (Ozoline, Brandizzo, TO, Italy).This apparatus operates at room temperature and allows photometric real-time control of the O 3 concentration.The apparatus is connected to a system for blood treatment (Tecnoline, Concordia sulla Secchia, MO, Italy), which is composed of (i) a sterile and apyrogenic circuit made of tubes for blood draw, (ii) hollow fibres for continuous mixing of blood with gas, and (iii) a bag for collection of the ozonised blood.The whole system is internally coated with phosphorylcholine to minimise the interaction of blood with plastics.Through a peristaltic pump, the blood was ozonised at a constant flow of 20 mL/minute and then, reversing the rotation direction, recovered from the bag.The whole procedure required 20 min.O 3 was used at the concentrations of 5 µg O 3 and 10 µg O 3 /mL O 2 (for a total of 50 µg O 3 and 100 µg O 3 administered to each blood sample, respectively).Pure O 2 was administered to discern the effect induced by O 3 from the O 2 -induced one.After gas treatment, the blood samples were submitted to different analyses (see below).The control samples did not undergo any treatment and were maintained at room temperature for the same time (20 min) required for the treatment with the O 3 generator.All experiments were conducted in triplicate.

Haemogas Analysis
After gas treatment, 1 mL of blood was collected from each sample and then analysed with an Edan i15 Vet blood gas analyser (Scil Animal Care Company S.r.l., Treviglio, BG, Italy) following the manufacturer's instructions to obtain the pO 2 , pCO 2 and pH values.Each sample was analysed in triplicate.

Haemolysis Assay
To evaluate the haemolysis in the blood samples, the haemoglobin concentration was evaluated using a haemoglobin colorimetric assay kit (ab234046, Abcam, MA, USA) at the end of the gas treatment.Briefly, 20 µL of blood and plasma were incubated with 180 µL of haemoglobin detector buffer at room temperature for 15 min in 96-well plates.The absorbance was detected at 570 nm using a Stat Fax ® 4300 ChroMate ® (Awareness Technology, Inc., Palm City, FL, USA) and the haemolysis was expressed as the percentage of plasma haemoglobin over the total blood haemoglobin.Each sample was analysed in triplicate.

Total Antioxidant Capacity Assay
The antioxidant capacity of the blood samples was measured using a Total Antioxidant Capacity Assay kit (ab65329, Abcam, MA, USA) at the end of the gas treatment.Briefly, blood was diluted 1:2500 in bidistilled water and incubated according to the manufacturer's instructions in 96-well plates.The absorbance was detected at 570 nm using a Stat Fax ® 4300 ChroMate ® (Awareness Technology, Inc., Palm City, FL, USA).The total antioxidant capacity was then calculated based on the absorbance of the samples with the standard curve.Each sample was analysed in triplicate.

Light Microscopy Analysis
Morphological analysis of the erythrocytes was performed on blood smears by means of brightfield microscopy.Ten µL of control and treated blood samples were smeared on a slide and air-dried under a sterile hood.The smears were then stained with the May-Grunwald-Giemsa technique and observed with an Olympus BX51 microscope (Olympus Italia S.r.l., Segrate, MI, Italy) equipped with a QICAM Fast 1394 Digital 116 Camera (QImaging, Surrey, BC, Canada) for image acquisition.Each sample was analysed in triplicate.

Interleukin Assessment
The amounts of IL-2 and IL-6 were evaluated in the defibrinated plasma of the control and treated blood samples after blood cell sedimentation.The samples were maintained in a sterile incubator at 37 • C in a 5% CO 2 humidified atmosphere for 2 h; then, 200 µL aliquots of plasma were stored at −80 • C until analysis.Interleukin quantitation was conducted via enzyme-linked immunosorbent assay (ELISA) on a Victor 3V mod.1420 plate reader (Perkin Elmer, Waltham, MA, USA).Briefly, 200 µL aliquots of plasma were put onto a 96-well plate and the FineTest ELISA kits ESH0013 and ESH0019 (FineTest Biotech Inc., Boulder, CO, USA) were used as per the manufacturer's recommendations to detect the sheep IL-2 and IL-6, respectively.The absorbance was read at 450 nm and the concentration of the target antigen in the sample was quantified.Samples were run in duplicate.

Metabolomics
Metabolomic analyses were performed on plasma aliquots obtained as described in Section 4.6.The plasma samples were prepared as previously reported [83].Briefly, plasma metabolites were obtained via protein precipitation with cold acetonitrile/isopropanol/water, followed by derivatisation with methoxamine and BSTFA.Small molecules were analysed using a LECO Pegasus BT 4D GCXGC/TOFMS instrument (Leco Corp., St. Josef, MI, USA) equipped with a LECO dual stage quad jet thermal modulator.The GC part of the instrument was an Agilent 7890 gas chromatograph (Agilent Technologies, Palo Alto, CA, USA) equipped with a split/splitless injector.The first dimension column was a 30 m Rxi-5Sil (Restek Corp., Bellefonte, PA, USA) MS capillary column with an internal diameter of 0.25 mm and a stationary phase film thickness of 0.25 µm, and the second dimension chromatographic column was a 2 m Rxi-17Sil MS (Restek Corp., Bellefonte, PA, USA) with a diameter of 0.25 mm and a film thickness of 0.25 µm.High-purity helium (99.9999%) was used as the carrier gas, with a flow rate of 1.4 mL/minute.One µL of sample was injected in splitless mode at 250 • C. The temperature programme was as follow: the initial temperature was 100 • C for 2 min, then ramped 20 • C/minute up to 330 • C and then held at this value for 2 min.The secondary column was maintained at +5 • C relative to the GC oven temperature of the first column.The programming rate was the same for the two columns.Electron impact ionisation was applied (70 eV).The ion source temperature was set at 250 • C, the mass range was 25 to 550 m/z with an extraction frequency of 32 kHz.The acquisition rates were 200 spectra/s.The modulation period for the bi-dimensional analysis was 4 s for the entire run.The modulator temperature offset was set at +15 • C relative to the secondary oven temperature, while the transfer line was set at 280 • C.
The chromatograms were acquired in total ion current mode.Peaks with a signal-tonoise (S/N) value lower than 500.0 were rejected.ChromaTOF version 5.31 was used for the raw data processing.Mass spectral assignment was performed by matching with the NIST MS Search 2.3 libraries and the FiehnLib.An in-house library of standards was also used for the small molecule identification.

Statistical Analysis
Data for each variable were presented as the mean ± standard error (SE).The Mann-Whitney test was used for the statistical comparison between the gas-treated samples and controls as well as between the O 3 -treated samples and O 2 -treated ones.Statistical significance was set at p ≤ 0.05.MetaboAnalyst 5.0 software (www.metaboanalyst.orgaccessed on 5 December 2023) was used for the statistical analysis of the metabolomics data.

Conclusions
Our basic study on the molecular modifications induced in blood ex vivo by low O 3 concentrations provides original information on the impact of mild ozonation on multiple metabolic pathways.The modified metabolites are mostly involved in the preservation of the oxidant-antioxidant balance and in energy production, according to the well-known effects of low O 3 doses as enhancers of both the antioxidant/anti-inflammatory response and metabolic activity [5,19].
Remarkably, our findings demonstrate the ability of 5 µg O 3 /mL O 2 to induce modifications in pO 2 , pCO 2 , pH, total antioxidant capacity, and many metabolites similarly to 10 µg O 3 /mL O 2 .The larger number of metabolites affected by 10 µg O 3 in comparison to 5 µg O 3 obviously suggests the stronger action of the higher O 3 concentration; however, the effects of 5 µg O 3 cannot be considered negligible and deserve attention, opening interesting perspectives to investigate, also in vivo, the therapeutic potential of very low O 3 concentrations presently considered unsuitable because of the neutralising antioxidant potential of blood [4].Blood treatment with reduced O 3 concentrations would represent a further advancement of the clinical application of medical O 3 in line with the low-dose O 3 concept [21].Moreover, the use of reduced O 3 concentrations would be especially beneficial for patients with a chronically high level of oxidative stress: in these patients, mild O 3 treatment would allow the restoration of a correct oxidant-antioxidant balance, avoiding the negative effect of high O 3 concentrations on their blood antioxidant capacity.

Figure 1 .
Figure 1.Mean value ± SE of pO2 (a), pCO2 (b) and pH (c) in the blood samples after gas (n = 3).The asterisk (*) indicates a significant difference in comparison to the control (CT

Figure 1 .
Figure 1.Mean value ± SE of pO 2 (a), pCO 2 (b) and pH (c) in the blood samples after gas treatment (n = 3).The asterisk (*) indicates a significant difference in comparison to the control (CTR).

Figure 2 .
Figure 2. Mean value ± SE of the haemolysis rate in the blood samples after gas treatment (n = 3).The asterisk (*) indicates a significant difference in comparison to the control (CTR).

Figure 2 .
Figure 2. Mean value ± SE of the haemolysis rate in the blood samples after gas treatment (n = 3).The asterisk (*) indicates a significant difference in comparison to the control (CTR).

Figure 4 .
Figure 4. Mean value ± SE of the total antioxidant capacity of the blood samples after gas treatment (n = 3).The asterisk (*) indicates a significant difference in comparison to the control (CTR); the symbol # indicates a significant difference in comparison to O2.

Figure 2 .
Figure 2. Mean value ± SE of the haemolysis rate in the blood samples after gas treatment (n = 3).The asterisk (*) indicates a significant difference in comparison to the control (CTR).

Figure 4 .
Figure 4. Mean value ± SE of the total antioxidant capacity of the blood samples after gas treatment (n = 3).The asterisk (*) indicates a significant difference in comparison to the control (CTR); the symbol # indicates a significant difference in comparison to O2.

Figure 4 .
Figure 4. Mean value ± SE of the total antioxidant capacity of the blood samples after gas treatment (n = 3).The asterisk (*) indicates a significant difference in comparison to the control (CTR); the symbol # indicates a significant difference in comparison to O 2 .

16 Figure 5 .
Figure 5. Mean value ± SE of the IL-2 (a) and IL-6 (b) amounts in the blood samples after gas treatment (n = 3).No significant difference was found among the samples.

Figure 5 .
Figure 5. Mean value ± SE of the IL-2 (a) and IL-6 (b) amounts in the blood samples after gas treatment (n = 3).No significant difference was found among the samples.

16 Figure 5 .
Figure 5. Mean value ± SE of the IL-2 (a) and IL-6 (b) amounts in the blood samples after gas treatment (n = 3).No significant difference was found among the samples.

Table 1 .
Quantitative changes (decrease/increase: ↓/↑) in the metabolites of interest, as a consequence of the biological effects of O 3 .