Ozone is a double-faceted gas. It has a crucial protective relevance in partially blocking mutagenic and carcinogenic UV radiations emitted by the sun (wavelengths of 100-280 nm) in the stratosphere [1], while its increasing concentration in the troposphere causes severe pulmonary damage and increased mortality [2, 3]. In spite of this drawback, there are growing experimental and clinical evidences about the medical use of ozone [4–11]. Since XVI Century, Paracelsus had ingeniously guessed that "all things are poison and nothing is without poison and only the right dose differentiates a poison from a remedy". In 2005, Timbrell reiterated the concept in his book: "The poison paradox; chemicals as friends and foes" [12]. During the Earth evolution, harnessing oxygen by metazoans has allowed a fantastic biodiversity and growth but it has also created a slow acting "poison". It is reasonable to believe that the antioxidant system slowly evolved and specialized during the last two billion years for counteracting the daily formation (3-5 g in humans) of anion superoxide in the mitochondria and the release of H₂O₂ by ubiquitous NADPH oxidases. However, there is a general consensus that the physiological production of H₂O₂ is essential for life. Olivieri et al.[13] and Wolff [14] were the first to describe the effect of either low concentrations of radioactive thymidine or of a very low dose of radiation inducing an adaptive response in human cells in comparison to a high dose. Goldman [15] introduced the term "hormesis" to mean "the beneficial effect of a low level exposure to an agent that is harmful at high levels". It goes to the merit of Calabrese [16–19] to have experimentally controlled this concept and to have presented a number of examples of stimulatory responses following stimuli below the toxicological threshold. Until 2002 ozone therapy was pharmacologically conceived as a therapy where low ozone doses were stimulatory, while high doses were inhibitory. This conception, reflecting the classical idea that a low antigen dose is stimulatory, where an antigen overdose is inhibitory, was vague and unsuitable because ozone acts in a complex way and a high dose can still be effective but accompanied by side-effects. Indeed, one of us in 2002 amply delineated the sequence of biochemical reactions elicited ex vivo after the addition of a certain volume of O₂-O₃ gas mixture to an equal volume of human blood [20]. First of all, mixing blood with an oxidant implies a calculated and precise oxidative stress, i.e. a homeostatic change with production of highly reactive messengers. The oxidative stress, like many others, induces a biological response leading to an adaptive phenomenon. The teleological significance of this response is universal, from bacteria to plants and Mammals, and small repetitive stresses induce an extremely useful adaption response represented by the revival of critical defense mechanisms [20–22]. At the same time, Calabrese and Baldwin described the "overcompensation stimulation hormesis" (OCSH) as the result of a compensatory biological process following an initial disruption in homeostasis [17]. After a reviewer's information also Re later on had expressed this possibility [23]. Ozone presents some subtle differences that will be explained by clarifying the biochemical reactions occurring between the organic compounds of plasma and this gas.

The three oxygen atoms in gas-phase ozone form an isosceles triangle with a distance among the equal sides of 1.26 Å, and exist in several mesomeric states in dynamic equilibrium [24]. In terms of oxidation potential (E°), ozone (2.07 V) is the third after fluorine (3.06 V) and hydroxyl radical (2.80 V). Other pertinent oxidants are: hydrogen peroxide (1.77 V), hypochlorous acid (1.49 V) and chlorine (1.36 V). Ozone has a paired number of electrons in the external orbit and, although it is not a radical molecule, it is far more reactive than oxygen and readily generates some of the ROS produced by oxygen. Ozone is very unstable and at 20 °C, with a half-life of about 40 min, it decomposes according to the exothermic reaction:

Such an aspect has generated the idea that ozone will donate its energy to the organism by reacting with specific body compartments [20]. However, after having ascertained the complexity of the mechanism of action, the conclusion is that ozone dissolved in the water of plasma acts as a pro-drug, generating chemical messengers which will accelerate transfer of electrons and the overall metabolism. It goes to the merit of Hans Wolff (1927-1980), a German physician, to have developed the O₃-AHT by insufflating ex vivo a gas mixture composed of medical oxygen (95%) and ozone (5%) into the blood contained in a dispensable ozone-resistant and sterile glass bottle [25].

During the most precise and safe methodological ex vivo O₃-AHT approach, oxygen-ozone mixture dissolves into the water of plasma. Oxygen has a low solubility, but the pO₂ slowly raises up to about 400 mmHg [29]. Hemoglobin become fully oxygenated (Hb₄O₈) but this is hardly relevant because, during the infusion period, it mixes with venous blood which has a pO₂ of about 40 mmHg. On the other hand, ozone behaves quite differently because, by immediately reacting with ions and biomolecules, it does not follow the classical Henry's law in terms of linear solubility variation with pressure. First of all ozone is about tenfold more soluble than oxygen and, as ozone dissolves in the plasmatic water, it instantaneously reacts with hydrophilic antioxidants: by using an ozone concentration of 40 μg/ml, corresponding to 0.84 µ;mol/ml per ml of blood, within five min an average of 78% of Aa has been oxidized to dehydroascorbate and about 20% of uric acid has been oxidized to allantoin [30]. Only about 10% of alpha tocopherol has formed an alpha tocopheryl radical. At the same time the remaining ozone performs the peroxidation of available unsaturated fatty acids, which represent an elective substrate and are mostly albumin-bound. Peroxidation of n-6 PUFA leads to the formation of H₂O₂ and 4-hydroxy-2E-nonenal (4-HNE) [31], while n-3 PUFA leads to the formation of 4-hydroxy-2E-hexenal (4-HHE) [32, 33]:

As all of these reactions happen in a few seconds, ozone, until present in the gas phase, continues to dissolve in the plasmatic water and instantly reacts. Within the canonical 5 min, ozone is fully extinct with both a rather small depletion of hydrosoluble antioxidants and the simultaneous plasmatic increase of ROS and LOP. The ozonated blood is then infused into the donor patient.

First of all the brief life-span of H₂O₂ will be discussed. During the 5 min of mixing blood with the gas ex vivo, H₂O₂ will dynamically increase its concentration: rapid at first and progressively slowing down as ozone is being depleted. With the therapeutically high ozone concentration of 80 μg/ml per ml blood, the H₂O₂ concentration measured in plasma after 2.5 min is at most 40 μM because the rate of synthesis is equilibrated by multiple degradation routes. Some H₂O₂ is reduced by free soluble antioxidants including traces of catalase and GSH-Px. As the hemolysis is negligible (<0.5%), free Fe²⁺ or Cu⁺ are not present and it is unlikely that hydroxyl ions are ever formed by either the Fenton-Jackson or the Haber-Weiss reactions. As H₂O₂ is unionized, it freely diffuse into all blood cells although the bulk is mopped up by erythrocytes. The establishment of a dynamic, yet transitory, H₂O₂ gradient between the plasma and the cytoplasmatic water of blood cells makes this oxidant a very early effector. Its final intracellular concentration may be not higher than 10%, hence 3-4 μmoles, as it has been demonstrated in other studies [34–39]. The smartness of this system is that the H₂O₂ concentration, though small, is enough to trigger several crucial biochemical reactions without toxicity because the internal cell environment contains a wealth of GSH, thioredoxin, catalase and GSH-Px, which do not allow a dangerous increase. In spite of a threshold of only a few micromoles, it has a critical relevance and means that an ozone amount below 0.42 μmol for each ml volume of the gas mixture (medical grade O₂ ≥95% and O₃ ≤5%) reacting in a 1:1 ratio with autologous blood may be ineffective, resulting in a therapeutic failure of O₃-AHT. It is also necessary to remind that the ozonation process greatly differs whether it occurs either in plasma or in blood. In plasma, TAS levels was, as expected, ozone-dose dependent and decreased between 46 and 63% in relation to ozone concentrations of either 0.84 μmol/ml or 1.68 μmol/ml per ml of plasma, respectively. On the other hand, in blood taken from the same donors, after being treated with the same ozone concentrations, TAS only decreased from 11 to 33% in the first minute after ozonation, respectively. Moreover, it was surprising to determine that they both recovered and returned to the original value within 20 min, indicating the capacity of blood cells to quickly regenerate dehydroascorbate and GSH disulfide [34]. It has been also brilliantly demonstrated that, thanks to erythrocytes, dehydroascorbate was recycled back to Aa within 3 min [40]. On the same way, only about 20% of the intraerythrocytic GSH had been oxidized to GSSG within one min after ozonation and promptly reduced to normal after 20 min [41]. Aa, alpha-tocopherol, GSH and lipoic acid undergo an orderly reduction by a cooperative sequence of electron donation continuously supplied by NADPH-reducing equivalents to GSH-Rd and thioredoxin reductase [42] (Figure 1). These data, by showing that the therapeutic ozonation only temporarily and reversibly modifies the cellular redox homeostasis were reassuring regarding the safety of ozone as a medical drug. In summary, the initial disruption of homeostasis due to ozone oxidation is followed by the rapid reestablishment of homeostasis with two main advantages: the first being the value of triggering several biochemical reactions in blood cells and the second mediated by LOP compounds, the induction of an adaptive process due to the up-regulation of the antioxidant enzymes. This is in line with the temporal sequence of the OCSH dose-response relationship.

As shown in Figure 2, LOP production follows peroxidation of PUFA present in the plasma: they are heterogeneous and can be classified as lipoperoxide radicals, alkoxyl radicals, lipohydroperoxides, F₂-isoprostanes, as well as aldehydes like acrolein, MDA and terminal hydroxyl alkenals, among which 4-HNE and 4-HHE. As free radicals and aldehydes are intrinsically deleterious, only precise and appropriate ozone doses must be used in order to generate them in very low concentrations. Among the aldehydes, 4-HNE is quantitatively the most important. It is an amphipathic molecule and it has a brief-half-life in saline solution. On the other hand it reacts with a variety of compounds such as albumin, enzymes, GSH, carnosine, and phospholipids [31, 53]. There is no receptor for 4-HNE but it has been reported that, in concentration above 1 μM in vitro, after binding to more than 70 biochemical targets, it exerts some deleterious activity [31]. On the other hand, during the rapid reaction of ozone with blood, the generated hydroxy-alkenals, will form adducts both with GSH or with the abundant albumin molecules. This possibility is supported by findings which have shown that human albumin, rich in accessible nucleophilic residues, can quench up to nine 4-HNE molecules, the first being Cys34, followed by Lys199 and His146 [27, 28]. Interestingly, when samples of ozonated human plasma were incubated at 37 °C for 9 hours, 4-HNE, most likely bound to albumin, remained stable [54]. These data clarify why a judicious ex vivo ozonation of blood does not harm the vascular system during the infusion into the donor. Aerobic organisms, in order to tolerate the continuous generation of aldehydic compounds have developed detoxifying systems as follows: the first is the dilution of these products in both the plasma and the extracellular fluid involving a volume of about 11 L in humans. The second is the detoxification operated by aldehyde dehydrogenase, aldose reductase and GSH-Tr [55, 56] and the third is the excretion via bile and urine excretion [57–59]. The relevance of these catabolic pathways was appreciated when the half-life of infused alkenals present in ozonated blood in a patient was less than 5 min [60]. The interesting aspect is that albumin can transport 4-HNE in all body tissues, from liver to endocrine glands and the CNS. 4-HNE-Cys adducts, released at many sites, inform a variety of cells of a transient, acute oxidative stress and represent an important biochemical trigger. At submicromolar or picomolar levels, 4-HNE can act as a signaling molecule capable of activating the synthesis of γ-glutamate cysteine ligase, γ -glutamyl transferase, γ -glutamyl transpeptidase, HSP-70, HO-1, and antioxidant enzymes such as SOD, GSH-Px, catalase and last but not least, G6PDH, a critical electron-donor enzyme during erythropoiesis in the bone marrow. There is a wide consensus on the relevance of the induction of protective molecules during small but repeated oxidative stress [20, 61–65]. In other words, the concept that a precisely controlled oxidative stress can strengthen the antioxidant defenses is well accepted today. Once again, the low level of stress by enhancing the fitness of the defense system, is consistent with the hormetic concept. Moreover at the time of ozonated blood infusion, 4-HNE-Cys adduct can also act on the vast expanse of endothelial cells and enhance the production of NO [35]. Such a crucial mediator on its own or as a nitrosothiol, with a trace of CO released with bilirubin via HO-1 activity, allows vasodilation, thus improving tissue oxygenation in ischemic tissues [66]. H₂S is another potentially toxic molecule that, when released in trace amounts, it becomes an important physiological vasodilator like NO and CO [67, 68]. Moreover, as it happens for the mentioned physiological traces of other gases, the small amount of ozone necessary to trigger useful biological effects is in line with the concept of the hormesis theory [69].

Another interesting aspect observed in about 2/3 of patients is a sense of wellness and physical energy throughout the ozonetherapy [70]. It is not yet known whether these feelings are due to the power of the generated ozone messengers which can modify or improve the hormonal secretion. On the other hand, the feeling of euphoria may be due to improved oxygenation or/and enhanced secretion of growth hormone, ACTH-cortisol and dehydroepiandrosterone [26, 71]. Furthermore, when LOP reach the hypothalamic area they may improve the release of serotonin and endorphins, as it was observed after intense dynamic exercise [72]. Experience acquired after thousands O₃-AHT has clarified that there is neither objective nor subjective toxicity, or to use Calabrese's acronyms, there is no observable adverse effects (NOAEL). Moreover, neither structural nor enzymatic damages have been observed in blood components after ozonation of blood within the therapeutic window [73, 74]. On the other hand, patients with more advanced disease during the initial session especially if performed with a high ozone dosage, frequently report to feel very tired and sleepy. This is the lowest observed adverse effect level (LOAEL) that has been observed in about 10% of PAD's patients with stage III and IV of the Leriche-Fontaine's classification. Such a knowledge compels to begin always with low ozone dosage and carefully observe the patient's response.

The metabolic syndrome is recognized as one of the most serious disease in Western countries caused by a number of metabolic alterations such as type-2 diabetes, hypercholesterolaemia, atherosclerosis and renal dysfunction with the common denominator represented by a chronic oxidative stress. Diabetic patients, particularly those with foot ulcers, are critical and today they still have a gloomy prognosis. This is because they need a multiform therapy aiming to eliminate the peripheral ischemia, the neuropathy and the infected skin lesions. The range of ozone concentrations between 15 and 35-50 µ;g/ml is safe also in individuals with a low TAS level and it appears to be particularly effective in PAD [43, 80–85]. Several clinical studies performed in different hospitals seem to establish the validity of the inverted U-shaped curve in this frequent pathology (Figure 3, panel B). In line with "the concept of a beneficial effect within the context of a dose-response study is difficult to determine due to considerable biological complexity and the fact that beneficial effects are often seen with reference to a specific and relative setting" [17], a word of caution is necessary. This is especially true when ozone therapy is performed in different patients within the variety of three PAD's II, III and IV stages, according to the Leriche-Fontaine classification [86]. First of all it is necessary to trust the precision of ozone's dosages used by different clinicians and secondly, ozone activity cannot be compared with that expressed by a single compound (see, eg, Arsenic [76], and homocysteine [77]) in cultured cells. As it has been clarified, the real ozone messengers are H₂O₂ as a ROS and a variety of alkenals as LOP. These messengers act on different cells, have a quite different lifetime and alkenals are intrinsically toxic. Furthermore, each patient has his own medical history and his own psycho-physical reactivity. Consequently, ozone dosages between 0.42-0.84 µ;mol/ml generate less alkenals than dosages in the range 0.84-1.68 µ;mol/ml, and therefore patients with a low antioxidant capacity become more susceptible to a side effect like deep fatigue after the therapy session. Attention must be also paid to the type of pharmacological response achieved in different pathologies as either muscular-orthopedic or autoimmune diseases. So far, in the latter it remains unknown the ozone dosage, if any, able to increase the T-cell regulatory levels and activity. Consequently, at this stage the U-shaped curve remains meaningful only for PAD and only future trials will be able to define the ozone behavior in either stroke or chronic heart disease. Martinez-Sanchez et al. have also reported that the theoretical U-shaped curve fits the ozone therapy results [87]. Blood ozonation, even if performed within the therapeutic range and for a few minutes, represents always a calibrated acute oxidative stress. In order to never harm the patient, the strategy: "start low-go slow" is a golden rule to induce a valid adaptation to the far more dangerous chronic oxidative stress, typical of inflammatory and degenerative diseases [88]. Such an aspect implies that the final therapeutic effect is due to an average of progressively increasing ozone dosages.

The gas mixture medical grade oxygen-ozone can be proficiently used for the ozonation of blood because this incomparable liquid tissue contains an imposing array of antioxidants, which are able to tame not only its oxidant power but also its messengers (ROS and LOP) generated by the reactions with blood components. Therefore, if ozone is judiciously used within the established therapeutic window (0.42-1.68 μmol/ml per ml of autologous blood) in PAD, it can exert better therapeutic effects than the current therapy by prostacyclin analogue. Moreover, regarding the accompanying foot-related problems, both some ozone derivatives like ozonated water and different gradation of standardized ozonated vegetable oils will be used until complete healing [89, 90]. As stroke, heart infarction and PAD are cumulatively the first cause of death and disability, if it will become possible to use ozone therapy in the public hospitals of the developed Countries, it may be possible to enter a phase where ozone will become an extensive remedy. Moreover, there is no doubt that either infective or autoimmune glomerulo-nephritis as well as end stages of renal failure associated with hemodialysis are characterized, to a different extent, by an imbalance between pro- and antioxidative mechanisms [91]. Moreover the kidney does not have the regenerative ability of liver and this is one of the reasons for explaining why too often "nephropaties lack a specific treatment and progress relentlessly to end-stage renal disease" [92]. Another important reason is that till today a valid strategy to reduce oxidative stress in renal diseases is not available. Ozone therapy, not only may correct a chronic oxidative stress, but it may also stimulate untapped resources able to afford some improvement [9, 93]. It appears therefore reasonable to suggest the combination of conventional treatments with mild O₃-AHT in any initial nephropathy for preventing the risk of progression towards a chronic disease.

In several Countries, among others Cuba, Russia, and Ukraine, treatments by ozone are already a reality, although different administration modalities, such as the infusion of ozonated saline and of the rectal insufflations of ozone, are in current use because inexpensive and applicable to thousands of patients every day [94]. Nevertheless, it is hoped that adequate ozone-based therapeutic treatments for patients affected by oxidative-stress related diseases could be implemented in every public hospital.

During the last two decades the paradoxical behaviour of ozone has been clarified: when it is chronically inhaled, it is highly toxic for the pulmonary system because the enormous alveolar surface, unprotected by sufficient antioxidants, is exposed to the cumulative ozone dose, which causes a chronic inflammation. This is not surprising because even for oxygen [95], as well as for glucose and uric acid levels a modification of the physiological concentrations is deleterious.

It remains to be clarified whether some messengers present in the ozonated blood are able to stimulate the release of staminal cells in the patient's bone marrow.

The evaluation of results obtained in several clinical trials performed in PAD has allowed to establish that the dose-response relationship in PAD can be depicted as an inverted U-shaped hormetic model with a brief, initial lack of effect due to the potency of blood antioxidants. A mild acute oxidative stress induced by ozone in blood ex vivo, as several other mild stresses due to either heat or cold exposure, a transient ischemia, other chemicals and physical exercise are able to induce a sort of "preconditioning response" often leading to both a repair and an increased defense capacity well within the "overcompensation stimulation hormesis". This new achievement, added to an increasing wide consensus in carefully using gases as NO, CO, H₂S, N₂O and H₂ as real medical drugs [68], suggests that also ozone may be soon included into this category. One of the basic functions of ozone, after dissolving in the water of plasma is to accelerate the exchange of protons and electrons or, in simple words, to reactivate the metabolism all over the body. In this way, crucial biological functions gone astray can recover indicating that ozone operated as both a biological response modifier and an antioxidant inducer.

It is hoped that this paper will elicit the interest of clinical scientists in evaluating ozone therapy in vascular, renal and diabetic diseases, thus translating the laboratory results to the patient's bed.

VAB, M.D., Emeritus professor of Physiology, Department of Physiology, University of Siena, Viale Aldo Moro, 2, 53100, Siena, Italy

IZ, in charge as post-doc position at the Department of Pharmaceutical Chemistry and Technology, Viale Aldo Moro, 2, 53100, Siena, Italy

VT, Associate professor in Pharmaceutical Technology and Chief of the Post-Graduate School of Hospital Pharmacy, University of Siena, Viale Aldo Moro, 2, 53100, Siena, Italy