Introduction

Medical ozone is an ozone/oxygen mixture which has demonstrated its clinical efficacy in various diseases such as ischemic syndrome, diabetes and neuro infectious foot, disc hernia, pain and other diseases which have chronic oxidative stress as common pathological characteristic. Ozone oxidative pre/post conditioning is a mechanism that stimulates the endogenous antioxidant defence system consequently regulating a cellular redox balance which became disrupted in the above mentioned disorders [1,2].

Rheumatoid Arthritis (RA) is classified as an autoimmune disease. Reactive Oxygen Species (ROS) are produced to a great extent in arthritic joints [3], so that many studies have associated oxidative stress with RA. Most of these papers have studied the role of ROS overproduction in cartilage and bone injury. Oxidative stress is in fact closely related to RA, and the maintenance of a proper redox balance seems to be a critical step in order to improve the clinical outcome.

Methotrexate (MTX) + medical ozone administered by rectal insufflation is a combined therapy that has increased the anti-inflammatory response and decreased the hepatotoxicity risk as compared with MTX-treated patients [4,5].

When ozone rectal insufflation is administered, the therapeutic schedule generally covers more than one cycle of ozone treatment (each cycle consisting of 20 treatments). The second cycle is able to increase or to maintain an improvement in clinical response to ozone. The cycles are separated by intervals of three months. This is a routine procedure in clinical practice to provide the efficacy and safety that have been demonstrated on a scientific basis [6]. However, no study investigating the clinical and redox characterization at the end of a second or several cycles of ozone treatment in RA patients has yet been performed.

Taking into account the good clinical response achieved in RA patients treated with more than one cycle of MTX+ozone, the aim of this study was to assess the beneficial effects of a second ozone cycle when compared with results from the first cycle using clinical indicators and redox biomarkers and its relationship with innate immune memory at the end of the second ozone cycle in RA patients treated with the MTX+ozone combined therapy.

Materials and Methods

Study design

This controlled clinical study was approved by the joint institutional review board (Scientific and Ethics Committees of the National Institute of Rheumatology, Ministry of Public Health, Cuba, and Pharmacy and Food Institute, University of Havana, Cuba) in accordance with the principles of the Declaration of Helsinki 2005. All patients gave their informed consent to enrolment after receiving adequate information concerning the study (characteristics, benefits and possible side effects). Before enrolment, all participants attended a training program to familiarize them with the study objectives and treatment plans.

To calculate the size of the sample, the Medstat Systems, Inc. (version 2.1, 1989; Fridley, MN, USA) method was used. The statistical difference between the beginning and end of ozone therapy was 0.2 with a type 1 error of 0.05 [7]. The target level of enrolment was determined in 20 patients.

Inclusion criteria

Adult patients (>18 years) of both sexes and different ethnic origins with a diagnosis of RA who fulfilled the revised American Rheumatism Association’s [8] criteria for RA (morning stiffness, swelling of hand joints, swelling of three or more joints, symmetric swelling of joints) were eligible to participate in the study. Patients of the National Institute of Rheumatology, Ministry of Public Health, Cuba, who accomplished the following criteria were chosen: Disease Activity Score 28 (DAS 28 >3.2 y≤5.1) The Health Assessment Questionnaire-Disability Index: (HAQ-DI, according to the validated Spanish version [9], and anti-Cyclic Citrullinate Peptides (anti-CCP >10 U/ml in serum) as well as patients with a disease duration longer than five years were included. The exclusion criteria were: patients with any history of chronic conditions such as liver disease, diabetes mellitus, respiratory disorders, cardiovascular diseases, alcohol abuse and smoking were not included in the study. Patients with overlapping syndrome, cancer, or other associated autoimmune disorders or who were pregnant were also excluded. Those patients who had been receiving corticosteroid agents and were under treatment with disease- modifying anti-rheumatic drugs and anti-TNF or other biological agents for at least 3 months before the study date were also excluded. The patients were treated as follows:

First Cycle of Ozone Treatment

MTX 12.5 mg, intramuscular (i.m.), once/week (every Monday from 9:00-10:00 in the morning) + Ibuprofen (400 mg, oral), one tablet each 8 h + folic acid (5 mg, oral), one tablet/day from Wednesday to Saturday. The ozone was generated by an OZOMED unit, Cuba. The 20 treatments by rectal insufflation (five/week from Monday to Friday) at 25 mg/l to 40 mg/l of ozone in stepped application and in increasing order were administered as follows:

1st week: 25 mg/l, 100 ml; 2nd week: 30 mg/l, 150 ml; 3rd week: 35 mg/l, 200 ml; 4th week: 40 mg/l, 200 ml.

At the end of the first cycle, ozone treatment was withdrawn from the therapeutic plan though RA patients continued with MTX, Ibuprofen and folic acid.

Second Cycle of Ozone Treatment

Three months after the preceding ozone treatment, patients treated in the first cycle received the second, applying the same therapeutic plan as described in the first cycle. Medical personnel were instructed to report all adverse reactions, whether described in the package circulars of the study medications or not.

Evaluation of Disease Activity

Changes in the development of the disease were determined via suitable indices of activity (clinical parameters) as well as anti-CCP antibodies and redox status determination before starting and after completing the first and second cycles of ozone application. Each patient served as their own control as our previous papers [4,10,11] (i.e. before medical ozone treatment). It is important to state clearly that placebo control is not possible to use because ozone is generated through special equipment designed for it. On the other hand, there is no another gas with proven safety and without therapeutic effects.

The main variables considered were:

Clinical parameters: DAS28 [12], HAQ-DI and anti-CCP levels. Pain intensity using a Visual Analogue Scale (VAS) from ‘10’ to ‘100’. This was classified as ‘10’ (minimum pain intensity) and ‘100’ (maximum pain intensity). No pain was considered as “0”.

Secondary variables considered were: (a) Plasma levels of injury markers such as Advanced Oxidation Protein Products (AOPP), Nitric Oxide (NO), Total Hydroperoxides (TH) and Malondialdehyde (MDA). (b) Serum levels of protective redox markers such as reduced Glutathione (GSH), Catalase (CAT) and Superoxide Dismutase (SOD) activities. c) Side effects.

Biochemical determinations

Blood samples for biochemical analysis were obtained after a 12 h overnight fast, at the beginning, and 24 h after the last ozone treatment.

Anti-CCP antibodies were determined using an ELISA kit (DRG, DRG Diagnostics, GmbH, Germany) (sensitivity 90%, specificity 98.3% and diagnostic efficacy 95.3%) Erythrocyte Sedimentation Rate (ESR) was obtained using Westergren’s quantitative method.

Redox parameters were determined by spectrophotometric methods using a reader plate (SUMA, Cuba) and a BOECO Model S 220 Spectrophotometer, Germany.

Superoxide Dismutase (SOD) activity was measured using kits supplied by Randox Laboratories Ltd., Ireland (Cat. No. SD125 and No. RS505). Catalase (CAT) activity was measured by following the decomposition of hydrogen peroxide at 240 nm at 10s intervals for 1 min [13]. After precipitation of thiol proteins using trichloroacetic acid 10%, reduced Glutathione (GSH) was measured according to the method of Sedlak and Lindsay [14] with Ellman's reagent [5′5 dithiobis (2-nitrobenzoic acid)10-2M (Sigma St. Louis, MO, USA)]; absorbance was measured at 412 nm.

Nitrite/nitrate levels as a measure of Nitric Oxide (NO) were determined by the Griess reaction after first converting nitrates to nitrites using nitrate reductase (Boehringer- Mannheim Italy SpA, Milan, Italy) [15]. The Advanced Oxidation Protein Products (AOPP) were measured as the oxidation of iodide anion to diatomic iodine by advanced oxidation protein products [16]. Quantification of Total Hydroperoxides (TH) was measured by Bioxytech H2O2-560 kit (Oxis International Inc., Portland, OR, USA). Concentrations of Malondialdehyde (MDA) were analysed using the LPO-586 kit obtained from Calbiochem (La Jolla, CA).

In the case of biochemical variables, laboratory data for healthy individuals (n=60) were taken as normal reference values. This group of subjects corresponded in terms of age, sex and ethnicity with the patients enrolled in the study.

Statistical analysis

The OUTLIERS preliminary test for detection of error values was initially applied. In order to assess differences among the first and second cycle of ozone treatment for clinical variables, the Friedman non-parametric test (for more than two samples linked) was used. In order to contrast each redox variable at the end of the study between both cycles of ozone treatment, the Mann-Whitney U non-parametric test was applied. The risk of GSH decrease at the end of the study for each cycle of ozone treatment was calculated by Odds ratio and non-parametric Wilcoxon signed rank test for two linked samples. Results are presented as mean ± Standard Deviation (SD). The level of statistical significance used was at least P<0.05.

Results

General characteristics of the patients involved in the study

In relation to the baseline characteristics (Table 1), all patients were women with an average age of 53 ± 8 years receiving the same therapy while the evolution time of RA was 7 ± 2 years. No major ethnic differences were observed.

Picture of clinical variables at the beginning and at the end of each cycle of ozone treatment

At the end of the first cycle of ozone treatment, a significant decrease in all clinical variables was observed. Three months after the first cycle, the period of time during which ozone was removed from the therapeutic plan, RA patients received another cycle of ozone treatment. The second cycle of ozone promoted a final decrease in all clinical variables as compared with the beginning of this cycle. Although there was a decreasing tendency at the beginning of the second ozone cycle, there were no statistic differences between the beginning of either two cycles. However, at the end of the second ozone exposure a stronger therapeutic response from patients was observed compared with the end of the first cycle. In some clinical variables this mainly involved anti-CCP levels (42 ± 1.5 vs 116 ± 23 U/L), DAS-28 (1.8 ± 0.96 vs 2.4 ± 0.7), and HAQ-DI (0.09 ± 0.02 vs 0.2 ± 0.05) respectively. g-Glutamyl transferase activity showed a tendency to decrease in the context of the first cycle, but without statistical differences and remaining within the reference range. Although the pain variable showed no statistical differences, there was a decreasing tendency after the second cycle when compared with the first cycle (Figure 1).

Redox biomarkers at the end of both cycles of ozone treatment

The protective (antioxidants) and injury (pro-oxidants) redox markers in rheumatoid arthritis patients at the end of first and second cycles was determined in the plasma (Figure 2).

After the second cycle of ozone treatment, the redox response was greater than clinical responses. 85.7% of the total redox biomarkers assessed (6/7 biomarkers) returned to normal reference values. Catalase activity did not change and remained below the normal reference range.

In relation to antioxidant defense biomarkers (Figure 2A), there was a greater increase in reduced glutathione (GSH) so that it could achieve the normal reference value, whereas SOD activity showed a considerable decrease and enzymatic activity returned to normal levels.

Injury markers showed impressive changes in line with the above results. Total Hydroperoxides (TH) and lipid peroxidation (MDA) decreased their high levels and reached the normal reference values, as well as the Advanced Oxidation Protein Products (AOPP). Although Nitric Oxide (NO) remained within normal value ranges at the end of both cycles, the NO levels were better in the second cycle.

Reduced glutathione is an important participant in the endogenous antioxidant defense system, and it is depleted in RA. The results of the Wilcoxon test showed a significant GSH increase (p<0.05) at the end of second ozone exposure in RA patients (15/20 patients increased GSH levels at the end of second cycle of ozone treatment expressed as the difference GSH₂-GSH₁ (GSH at the end of the 2^nd versus GSH at the end of the 1^st treatment cycles). Only 4 patients had decreased GSH levels compared with the end of first cycle, and one patient had the same GSH concentration at the end of both cycles (Figure 3).

A low risk of GSH concentration decrease was found at the end of the second cycle of ozone treatment (0.667), whereas the probability of GSH remaining within the normal reference interval increased by a factor of 4 (Table 2).

Discussion

A second ozone exposure in RA patients treated with MTX+ozone combined therapy promoted powerful clinical and redox responses. A reduction in inflammation (DAS-28) and disability (HAQ-DI) was found. These results agreed with the decrease in Anti-CCP levels at the end of the second ozone cycle. Anti-CCP antibodies are actively produced or enriched at the site of inflammation (joints and synovial tissue) and are able to participate actively in the pathogenesis of anti-CCP positive RA by enhancing oxidative stress in the rheumatoid joint [17]. Indeed, high titers (titers US, titres BRIT) of anti-CCP antibodies have been associated with an erosive disease course and outcome in RA [18]. g-Glutamyl Transferase (GGT) concentrations were within the normal reference interval and corresponded to the decrease in anti-CCP levels combined with a clinical and redox improvement of the patients. This enzyme is close connected with the bone destruction observed in RA as mediated by excessive osteoclastogenesis. GGT acts as an endogenous activator of this type of pathological osteclastogenesis, independently of its enzymatic activity. The underlying mechanisms involve Toll-like Receptor 4 (TLR4) which recognizes GGT in order to activate inflammation-associated osteoclastogenesis [19]. The impact of the second ozone exposure on the cellular redox status was greater than the clinical response described above. All redox biomarkers returned to their normal reference interval, except the catalase activity combined with a reduction in the mediators contributing to cartilage and bone injury (TH, MDA, AOPP). These results were expected because the major ozone targets are the redox markers determining antioxidant/prooxidant cellular balance, meaning that the main effect of the second ozone exposure should be on the redox status of the patients as found in this study. It is important to emphasize the complete recovery of GSH as well as the low risk of GSH depletion found after the second ozone exposure. This antioxidant tripeptide is depleted by GGT and it has been proposed as a mediator of signal transduction events associated with regulation of cellular redox balance [20].

The clinical and redox improvement achieved in RA patients treated with combined MTX+ozone therapy suggest that a protective effect took place in the first ozone cycle involving events associated with the activity of the innate immune system such as reduction of oxidative damage, Toll-like receptor 4 activation by GGT, cytokine release, etc. After three months, a second ozone exposure is a stimulus capable of activating the innate memory that strengthens the protective responses against cartilage and bone injury associated with autoimmune disease. On the other hand, memory-induced hyper response may be involved in the pathological sequelae of innate immunity/inflammation as seen in autoimmunity [21].

Last year, important discoveries changed our angle of view in the context of the innate immune system. Features such as specificity and memory, the main traits of the adaptive immune system, are now also to a certain extent included in the context of innate immunity. The discovery of PRRs (Pattern Recognition Receptors) such as Toll-like receptors makes it possible to recognize that innate immune cells can identify different pathogenic stimuli based on their recognition of damage-associated molecular patterns. A very interesting notion is that the innate memory is, at least in part, nonspecific. This implies that an improved defensive response can be obtained by pre-challenging the host with (almost) any kind of agents [22].

The mechanisms underlying innate immune memory have not yet been fully elucidated. Nevertheless, a number of principal events were involved [22] which are closely related to the effects of medical ozone.

TLR4 represents one of the best-characterized PRRS; these are able to recognize bacterial Lipopolysaccharide (LPS) and types of molecular formation known as Damage-Associated Molecular Patterns (DAMPs) [23]. These are closely associated with the induction and development of sterile inflammation observed in diseases such those in an autoimmune context [24].

Ultralow LPS stimulation (implying a long-term slow infection with tissue debilitation) induces an innate memory that results in an enhanced reactivity to subsequent stimuli, necessary for the adequate defence of a weakened tissue [25]. The activity of GGT as regulated by medical ozone [5] may be similar to that of LPS, as LPS stimulates TLR4 and is known to exert a potent osteoclastogenic activity through this receptor [26]. It appears that extracellular GGT is able to activate other types of TLR4-expressing cells in addition, including osteoblasts and macrophages, inducing them to react with pro-inflammatory responses in a way similar to that described for LPS.

Although the underlying mechanism by which extracellular GGT can be released to serve as a TLR4-stimulating DAMP remains unclear, the production and release of GGT from the lymphocytes or inflammatory cells may be stimulated by the inflammation when accompanied by oxidative stress. Previous results have shown that the transcription of GGT is highly sensitive to oxidative stress [27], and that elevated humoral GGT levels have been observed in those diseases associated with chronic inflammation, as found for example in the synovial fluids collected from the arthritic joints [28].

Another mechanism associated with an innate immune memory is found in the type of cells involved. Among the innate immune cells, the most active innate memory cells consist of monocytes/macrophages and NK cells. In addition, innate memory can be induced at the immune stem cell level, i.e. in bone marrow niches in which non-immune cells probably contribute to the induction of stem cell priming [29]. The recent findings that epithelial stem cells retain a memory of previous inflammatory challenges by displaying an enhanced wound healing capacity after skin damage reveal that an innate memory may not be restricted to immune cells [30].

It is very interesting to find that monocytes/macrophages are able to develop different types of memory depending on the type of priming. Thus, monocytes and/or macrophages can develop a memory that makes them less reactive to certain challenges (tolerance, the avoidance of extensive tissue damage) or to an enhancement of their response (training, improved tissue surveillance e.g. against immune diseases). These different types depend on the nature of the first challenge. A low stimulation, similar to that by which medical ozone is administered (low dose), promote an enhanced reactivity to subsequent stimuli [31].

Regulation of a cytokine pattern when mononuclear cells in peripheral blood are activated by medical ozone has been demonstrated [32]. The release of Interleukin 4, 6 and 10, tumour necrosis factor alpha (TNF-α) and IFN-α and IFN-g was dependent on ozone concentrations and incubation time of the cells. Recently, it also has been demonstrated that cytokine combinations can induce memory-like properties in human cells [22,33,34]. A parallel between innate immune memory and its relationship with medical ozone effects associated with immune system are shown in Table 3.

In summary, RA patients treated with combined MTX-ozone therapy increased clinical and redox response after the second cycle of ozone exposure when compared with the results from the first ozone cycle. This potentiation seems to be mediated by an innate immune memory based on the close relationship between the targets revealing an innate immune memory and the regulating effect of medical ozone.

At present, an investigation of those epigenetic markers which have been identified in association with the acquisition of an innate immune memory is in progress.

Figure 1:  Medical ozone effects on clinical variables in RA patients after receiving two cycles of ozone treatment at intervals of three months between each cycle. The mean ± Standard Deviation (SD) is given. The different letters indicate statistical differences  (p<0.05). DAS-28: Disease Activity Score 28; HAQ: DI, Health Assessment Disability Questionnaire; Anti-CCP: Anti-Cyclic Citrullinated Peptides.

Figure 2: (A) Protective and (B) Pathogenic redox biomarkers in RA patients treated with MTX+ozone after receiving two cycles of ozone treatment at intervals of three months between each cycle. The mean ± standard deviation (SD) is given. (^*) mean statistical difference p<0.05. Range marked with dotted lines: normal reference values (n=60 healthy subjects). GSH: Reduced Glutathion; SOD: Superoxide Dismutase; CAT: Catalase Activities; TH: Total Hydroperoxides; MDA: Malone Dialdehyde; AOPP: Advanced Oxidation Protein Products; NO: Nitric Oxide.

Figure 3: Increase of the GSH level (µg/ml) in 15 of 20 patients at the end of second cycle of ozone treatment, expressed as the difference GSH₂-GSH₁ (GSH at the end of the 2^nd versus GSH at the end of the 1^st treatment cycle).  A slight decrease was found in 4 patients and no difference in 1 patient after the second treatment cycle (non-parametric Wilcoxon signed rank test for two linked samples).

1.       León OS, Menéndez S, Merino N, Castillo R, Sam S, et al. (1998) Ozone oxidative preconditioning: a protection against cellular damage by free radicals. Mediators Inflamm 7: 289-294.

2.       León OS (2014) Ozone Therapy: Oxidative Conditioning, Basis for Its Clinical Effectiveness. In: Antioxidant, pro-oxidant balance. 1^st Edition. ODREI, Germany. 1-202 pages.

3.       Stamp LK, Hazlett J, Roberts RL, Frampton C, Highton J, et al. (2012) Adenosine receptor expression in rheumatoid synovium: a basis for methotrexate action. Arthritis Res Ther 14: R138.

4.       León Fernández OS, Viebahn-Haensler R, Cabreja GL, Espinosa IS, Matos YH, et al. (2016) Medical ozone increases methotrexate clinical response and improves cellular redox balance in patients with rheumatoid arthritis. Eur J Pharmacol 789: 313-318.

5.       Oru GT, Viebhan-Haensler R, Cabreja GL, Espinosa IS, Santos BT, et al. (2018) Medical ozone reduces the risk of γ-glutamyl transferase and alkaline phosphatase abnormalities and oxidative stress in rheumatoid arthritis patients treated with methotrexate. SM J Arthritis Res 1: 1004.

6.       Viebahn-Hansler R, León Fernández OS, Fahmy Z (2016) Ozone in Medicine: Clinical Evaluation and Evidence Classification of the Systemic Ozone Applications, Major Autohemotherapy and Rectal Insufflation, According to the Requirements for Evidence-Based Medicine. Ozone: Science & Engineering 38: 322-345.

7.       Levy PS, Lemeshow S (1991) Sampling of populations: methods and applications. 4^th Edition. John Wiley & Sons. New York, USA.

8.       Arnett FC, Edworthy SM, Bloch DA, McShane DJ, Fries JF, et al. (1988) The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum 31: 315-324.

9.       Cardiel MH, Abello-Banfi M, Ruiz-Mercado R, Alarcon-Segovia D (1993) How to measure health status in rheumatoid arthritis in non-English speaking patients: validation of a Spanish version of the Health Assessment Questionnaire Disability Index (Spanish HAQ-DI). Clin Exp Rheumatol 11: 117-121.

10.    Martínez-Sánchez G, Al-Dalain SM, Menéndez S, Re L, Giuliani A, et al. (2005) Therapeutic efficacy of ozone in patients with diabetic foot. Eur J Pharmacol 523: 151-161.

11.    León Fernández OS, Pantoja M, Díaz Soto MT, Dranguet J, García Insua M, et al. (2012) Ozone oxidative post-conditioning reduces oxidative protein damage in patients with disc hernia. Neurol Res 34: 59-67.

12.    Prevoo ML, van 't Hof MA, Kuper HH, van Leeuwen MA, van de Putte LB, et al. (1995) Modified disease activity scores that include twenty-eight-joint counts. Development and validation in a prospective longitudinal study of patients with rheumatoid arthritis. Arthritis Rheum 38: 44-48.

13.    Keesey J (1987) Biochemica Information: A Revised Biochemical Reference Source. Enzymes for Routine, 1^st Edition. Berlin, Germany. 80-81.

14.    Sedlak J, Lindsay RH (1968) Estimation of total, protein-bound, and nonprotein sulfhydryl groups in tissue with Ellman's reagent. Anal Biochem 25: 192-205.

15.    Granger DL, Taintor RR, Boockvar KS, Hibbs JB (1995) Determination of nitrate and nitrite in biological samples using bacterial nitrate reductase coupled with the Griess reaction. Comp Methods Enzymol 7: 78-83.

16.    Witko-Sarsat V, Friedlander M, Nguyen Khoa T, Capeillère-Blandin C, Nguyen AT, et al. (1998) Advanced oxidation protein products as novel mediators of inflammation and monocyte activation in chronic renal failure. J Immunol 16: 2524-2532.

17.    Snir O, Widhe M, Hermansson M, von Spee C, Lindberg J, et al. (2010) Antibodies to several citrullinated antigens are enriched in the joints of rheumatoid arthritis patients. Arthritis Rheum 62: 44-52.

18.    Silveira IG, Burlingame RW, von Mühlen CA, Bender AL, Staub HL (2007) Anti-CCP antibodies have more diagnostic impact than rheumatoid factor (RF) in a population tested for RF. Clin Rheumatol 26: 1883-1889.

19.    Moriwaki S, Into T, Suzuki K, Miyauchi M, Takata T, et al. (2016) γ-Glutamyl transpeptidase is an endogenous activator of Toll-like receptor 4-mediated osteoclastogenesis. Sci Rep 6: 35930.

20.    Viebahn-Hänsler R, León Fernández OS, Fahmy Z (2012) Ozone in Medicine: The Low-Dose Ozone Concept-Guidelines and Treatment Strategies. Ozone: Science & Engineering 34: 408-424.

21.    Arts RJW, Joosten LAB, Netea MG (2018) The potential role of trained immunity in auto-immune and auto-inflammatory disorders. Front Immunol 9: 298.

22.    Töpfer E, Boraschi D, Italiani P (2015) Innate Immune Memory: The Latest Frontier of Adjuvanticity. J Immunol Res 2015: 478408.

23.    Schaefer L (2014) Complexity of danger: the diverse nature of damage-associated molecular patterns. J Biol Chem 289: 35237-35245.

24.    Chen GY, Nunez G (2010) Sterile inflammation: sensing and reacting to damage. Nat Rev Immunol 10: 826-837.

25.    Ifrim DC, Quintin J, Joosten LA, Jacobs C, Jansen T, et al. (2014) Trained immunity or tolerance: opposing functional programs induced in human monocytes after engagement of various pattern recognition receptors. Clin Vaccine Immunol 21: 534-545.

26.    Liu J, Wang S, Zhang P, Said-Al-Naief N, Michalek SM, et al. (2009) Molecular mechanism of the bifunctional role of lipopolysaccharide in osteoclastogenesis. J Biol Chem 284: 12512-12523.

27.    Zhang H, Forman HJ (2009) Redox regulation of gamma-glutamyl transpeptidase. Am J Respir Cell Mol Biol 41: 509-515.

28.    Ishizuka Y, Moriwaki S, Kawahara-Hanaoka M, Uemura Y, Serizawa I, et al. (2007) Treatment with anti-gamma-glutamyl transpeptidase antibody attenuates osteolysis in collagen-induced arthritis mice. J Bone Miner Res 22: 1933-1942.

29.    Mitroulis I, Ruppova K, Wang B, Chen LS, Grzybek M, et al. (2018) Modulation of Myelopoiesis Progenitors is an Integral Component of Trained Immunity. Cell 172: 147-161.

30.    Netea MG, Joosten LA, Latz E, Mills KH, Natoli G, et al. (2016) Trained immunity: A program of innate immune memory in health and disease. Science 352: aaf1098.

31.    Saeed S, Quintin J, Kerstens HH, Rao NA, Aghajanirefah A, et al. (2014) Epigenetic programming of monocyte-to-macrophage differentiation and trained innate immunity. Science 345: 1251086.

32.    Larini A, Bocci V (2005) Effects of ozone on isolated peripheral blood mononuclear cells. T Toxicol In Vitro 19: 55-61.

33.    Romee R, Schneider SE, Leong JW, Chase JM, Keppel CR, et al. (2012) Cytokine activation induces human memory-like NK cells. Blood 24: 4751-4760.

34.    World Medical Association (2013) World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects. JAMA 310: 2191-2194.