Early Effects of Hyperbaric Oxygen on Inducible Nitric Oxide Synthase Activity/Expression in Lymphocytes of Type 1 Diabetes Patients: A Prospective Pilot Study

This study aimed at examining the early effects of hyperbaric oxygen therapy (HBOT) on inducible nitric oxide synthase (iNOS) activity/expression in lymphocytes of type 1 diabetes mellitus (T1DM) patients. A group of 19 patients (mean age: 63 ± 2.1) with T1DM and with the peripheral arterial disease were included in this study. Patients were exposed to 10 sessions of HBOT in the duration of 1 h to 100% oxygen inhalation at 2.4 ATA. Blood samples were collected for the plasma C-reactive protein (CRP), plasma free fatty acid (FFA), serum nitrite/nitrate, and serum arginase activity measurements. Expression of iNOS and phosphorylation of p65 subunit of nuclear factor-κB (NFκB-p65), extracellular-regulated kinases 1/2 (ERK1/2), and protein kinase B (Akt) were examined in lymphocyte lysates by Western blot. After exposure to HBOT, plasma CRP and FFA were significantly decreased (p < 0.001). Protein expression of iNOS and serum nitrite/nitrate levels were decreased (p < 0.01), while serum arginase activity was increased (p < 0.05) versus before exposure to HBOT. Increased phosphorylation of NFκB-p65 at Ser536 (p < 0.05) and decreased level of NFκB-p65 protein (p < 0.001) in lymphocytes of T1DM patients were observed after HBOT. Decreased phosphorylation of ERK1/2 (p < 0.05) and Akt (p < 0.05) was detected after HBOT. Our results indicate that exposure to HBO decreased iNOS activity/expression via decreasing phosphorylation of ERK1/2 and Akt followed by decreased activity of NFκB.

Systemic hyperbaric oxygen (HBO) therapy (HBOT) has been proposed as a medical treatment for DM patients with the developed peripheral arterial disease. HBOT is defined as therapeutic inhalation of 100% oxygen in the elevated pressure controlled conditions, which induces micro-and macrovascular hemodynamic changes [11]. An increase in oxygen arterial partial pressure in hyperbaric conditions promotes better solubility of plasma oxygen. This further results in the preservation of vitality of tissues, reversibly damaged by atherosclerosis-induced ischemia, simultaneously with microcirculation restoration [12]. Exposure to HBO decreases iNOS activity/expression followed by reduction of NO generation, which implies that this may be mechanisms responsible for the anti-inflammatory effect of HBOT [13,14]. However, the exact molecular mechanism by which HBOT reduces inflammation remains still unclear.
We hypothesized that HBO affects iNOS regulation through inhibition of the NFκB activation, by a mechanism that involves phosphorylation of ERK1/2 and Akt. The aim of this prospective pilot study was twofold: (1) to investigate the early effects of HBOT on iNOS activity/expression and (2) to investigate the involvement of Akt and ERK1/2 in the regulation of iNOS activity/expression, in T1DM patients exposed to HBO.

Subjects and Methods.
After obtaining the Ethical Committee of Zemun Clinical Hospital approval and written informed consent, 19 type 1 DM (T1DM) patients with no obvious DM ulcers, inflammatory redness on lower extremities, and contraindications for HBOT were included in this prospective pilot study during 2017, in Zemun Clinical Hospital, Zemun, Serbia. The study inclusion criterion was the presence of T1DM with the peripheral arterial disease. A total of 10 sessions of HBOT were applied as one session per day, for two weeks, in the duration of 1 h of 100% oxygen inhalation at 2.4 ATA. The demographic properties and medical history of all patients were recorded on the first visit, including gender, type of diabetes, current diabetes condition, hypertension, and other concomitant therapy of study interest (aspirin, statins, fibrates, antiplatelet therapy, angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, calcium-channel blockers, and beta blockers). The peripheral arterial disease was confirmed by the standard questionnaire. Blood samples were collected, and lymphocytes were isolated and stored at −20°C. Half of the collected blood was transferred to EDTA-containing vacutainer tubes, incubated on ice for one hour, and subsequently centrifuged for 15 min at 2000 ×g. Obtained supernatants (plasma) were stored at −20°C before analysis. Serums were isolated by incubation of blood at room temperature for 30 min without anticoagulants, followed by 15 min centrifugation at 1800 ×g. Serums were stored at −20°C until further biochemical analysis. Glycosylated hemoglobin A1c (HbA 1c ), C-reactive protein (CRP), free fatty acids (FFA), nitrite/nitrate levels, and arginase activity were measured before and after completion of HBOT. Complete blood count, total cholesterol, and triglyceride (TG) were measured before HBOT.

Measurement of Plasma TG and Total Cholesterol
Concentration. The concentration of TG was measured using a commercially available kit according to the manufacturer's guidelines using a Roche cobas c 501 analyzer (Roche Diagnostics, Indianapolis, IN, USA). The concentration of total cholesterol was measured by standardized enzymatic colour test using Beckman Coulter Olympus AU400 analyzer (Brea, CA, USA) [15]. Total cholesterol and TG concentrations were expressed as mmol/L.

Measurement of Plasma CRP and FFA Concentrations.
The concentration of CRP in the plasma was measured by the immunoturbidimetric method using a commercially available kit (system reagent for the quantitative determination of CRP in human plasma), following manufacturer guidelines, on Roche cobas c 501 analyzer (Roche Diagnostics, Indianapolis, IN, USA). CRP concentrations were expressed in mg/L and presented as % of CRP before HBOT.
The FFA concentration was measured in the plasma containing EDTA, using a modified version of the Duncombe method [16]. The principle of the method is that extracted FFA in chloroform, in the presence of an appropriate reagent (aqueous solution of Cu(NO 3 ) 2 × 3H 2 O with triethanolamine, pH 7.8), forms salts of copper, which in contact with diethyldithiocarbamate builds a yellow complex compound with a maximum absorbance at 436 nm [15]. The concentrations of FFA were expressed in mmol/L and presented as % of FFA before HBOT.
2.6. Measurement of Serum Nitrite/Nitrate Concentration and Arginase Activity. The concentration of the serum nitrite and nitrate was measured by using a nitrate/nitrite colourimetric assay kit (Cayman Chemical, Ann Arbor, MI, USA) according to the manufacturer's protocol [15]. The concentrations of nitrite/nitrate were expressed in mmol/L and presented as % of nitrite/nitrate before HBOT.
Arginase activity was measured spectrophotometrically in serum, according to the method of Corraliza et al. [17]. Arginase is the enzyme that converts L-arginine to urea and L-ornithine; thereby, the principle of this method is based on measurement of the amount of generated urea which is directly proportional to arginase activity. Briefly, aliquots of serum were incubated for 10 min at 55°C in complete assay mixture lacking arginine. The reaction was initiated by addition of L-arginine, and incubation was continued at 37°C for 1 h. The reaction was terminated by heating at 100°C for 45 min. Values of arginase activity were expressed in nmol/min/mg of proteins and presented as % of arginase activity before HBOT.

Isolation of Lymphocytes from Peripheral
Blood. The heparinized whole blood samples from each patient were used for the isolation of lymphocytes with LSM, according to the method of Boyum [18]. The lymphocytes forming a layer directly above the LSM were isolated and washed twice with phosphate-buffered saline. Each wash was followed by centrifugation at 1200 ×g for 10 min. Finally, the supernatant was removed and stored at −70°C for further analysis.

Isolation of Proteins from Lymphocytes.
Whole cell extracts were prepared by suspending cell pellets in ice-cold buffer for protein isolation (150 mM sodium chloride (NaCl), 20 mM tris(hydroxymethyl)aminomethane (TRIS), 2 mM ethylenediaminetetraacetic acid (EDTA), 2 mM dithiothreitol (DTT), 1% nonionic detergent Triton X-100, and 10% glycerol) containing phosphatase and protease inhibitor cocktails. Lymphocytes were lysed by rotation at +4°C for a period of 1 h. Lysed samples were then centrifuged for 30 min at 14000 ×g at +4°C. Protein concentration was determined by the Lowry method [19]. The supernatants were stored at −70°C for further analysis.

Statistical Analysis.
Values are expressed as means ± SEM. SPSS 18.0 for Windows (SPSS Inc., Chicago, IL, USA) was used for all statistical calculations. Analysis of data was evaluated using two-tailed Student's t-test. The statistical significance level was set at p < 0 05.

Results
The anthropometric parameters including the metabolic parameters of T1DM patients at admission are presented in Table 1.
Since the promoter of iNOS gene is the binding site for NFκB transcription factor [21], we further examined whether NFκB is involved in HBOT downregulation of iNOS activity/expression by measuring p65 subunit level of NFκB ( Figure 3). Furthermore, we also assessed the effect of HBOT of the phosphorylation state of NFκB-p65 at Ser 536 by the immunoblotting with the phospho-specific NFκB-p65 antibody that recognizes NFκB-p65 only when phosphorylated at Ser 536 .
HBOT leads to a significant decrease of the level of the total form of NFκB-p65 by 0.23-fold (before HBOT = 1-fold; after HBOT = 0.77 ± 0.07-fold, p < 0 01) (Figure 3(a)), as well as the level of the phosphorylated form of NFκB-p65 by 0.3-fold (before HBOT = 1-fold; after HBOT = 0.70 ± 0.13-fold, p < 0 05) (Figure 3(b)). However, HBOT increased the ratio between the phosphorylated and total forms of NFκB-p65 in the lysate of T1DM patients' lymphocytes (before HBOT = 1-fold; after HBOT = 1.42 ± 0.17-fold, p < 0 05) (Figure 3(c)). Since we have previously shown that the stimulation of iNOS also involves activation of ERK1/2 phosphorylation [22] and activation of Akt phosphorylation [23], we next examined whether ERK1/2 and Akt are involved in decreased iNOS activity/expression after the exposure to HBOT. Therefore, we assessed the effect of HBOT of the phosphorylation state of ERK1/2 ( Figure 4) and Akt ( Figure 5) by the immunoblotting with phospho-specific ERK1/2 and Akt antibodies that recognize ERK1/2 and Akt only when ERK is phosphorylated at Thr 202 /Tyr 204 and Akt at Ser 473 , respectively.

Discussion
The principal findings of our study included the observation that in T1DM patients, HBOT downregulates the iNOS activity/expression by decreasing phosphorylation of ERK1/2 and Akt followed by a decreased activation of NFκB. For this study, we have used HBOT as a medical treatment for T1DM patients with the developed peripheral arterial disease, with the purpose to reduce inflammation and to promote microcirculation. Biochemical profile of DM is commonly presented by elevated levels of HbA 1c , CRP, FFA, and NO which provide an environment of oxidative stress and inflammation [24]. HbA 1c is a reliable biomarker for the diagnosis and prognosis of DM and correlates well with the risk of long-term diabetes complications [25]. Results from our study show that the level of HbA 1c was 8.6 ± 0.08%, which indicates a higher risk of DM-associated complications [26].
Elevated FFA, which is common in the diabetic population [27], may trigger systemic inflammation and impair NO bioavailability, leading to impaired endothelium-dependent and insulin-mediated vasodilation [28]. Increase in FFA level is associated with an increase in NO level and upregulation of iNOS mRNA as well as the reduction of insulin output in cultured prediabetic rat pancreatic islets [29]. Also, increased plasma FFA levels are accompanied by increased expression of several NFκB-dependent cytokines, leading to chronic inflammatory processes [30]. In our study, we observed a significant decrease in FFA after HBOT in T1DM patients. These results are consistent with data gained by Teshigawara et al. [31]. Teshigawara et al. showed that FFA concentration was significantly decreased after HBOT, implying that HBOT could have beneficial effects on lipid metabolism. Clinical manifestation of inflammation is characterized by the increase of acute-phase proteins, among which prominent place has CRP [32,33]. The level of CRP is a well-known marker of  Table 1. inflammation [32], and it is reported that CRP may stimulate iNOS expression and NO production [34]. Our results show that exposure to HBOT significantly reduces the level of CRP in T1DM patients. Thus, the results from our study suggest that HBOT leads to a reduction of inflammation and subsequent progression of atherosclerosis [35]. iNOS is an important mediator of inflammation and may be the critical link between metabolic disorders and inflammation [36]. iNOS expression occurs after induction as a response to cytokines and other proinflammatory agents. After induction, iNOS continuously produces 100-to 1000-fold more NO than constitutive NOS [37,38]. Excessive NO production favors the formation of peroxynitrite and thus contributes to tissue injury and reduced NO bioavailability [39,40]. In our study, we have observed that HBOT leads to a decrease in iNOS expression in lymphocytes and consequently decreases the level of the serum nitrite/nitrate concentration. Our results also show that HBOT increased serum arginase activity, and increased arginase activity may reduce the substrate availability of NOS, since NOS and arginase use the same substrate, the amino acid L-arginine, and consequently could reduce nitrite/nitrate concentration [41].
Moreover, numerous literature data reported that activities of iNOS and arginase are regulated reciprocally by cytokines and that the inhibition of iNOS leads to the subsequent decrease in the NO production that is associated with increased arginase activity [42][43][44]. Elevated levels of glucose in diabetes may enhance NO production through increased expression of endothelial NOS (eNOS) and iNOS gene and protein levels [8,[45][46][47]. Although iNOS/eNOS might be responsible for the release of NO from endothelial cells [47], numerous literature data reported that hyperglycaemia increased the NO level mostly through activation of iNOS [6,[48][49][50]. Thus, the decreased serum nitrite/nitrate concentration suggests that decreased iNOS and increased arginase activity after exposure to HBOT in our study may be due to decreased iNOS expression. This is in accordance with other results [51][52][53][54][55][56]. Pedoto et al. have also observed in their study that HBO pretreatment reduces lipopolysaccharide-induced iNOS expression [51]. Also, HBOT reduces the level of exhaled NO for 1 h in human subjects [55]. Similarly, rats exposed to HBOT for 7 days show significantly decreased NOS activity [13], while rabbits exposed to HBOT for 5 days have decreased expression of iNOS [52].
The results from our study reveal that HBOT significantly increased phosphorylation of NFκB-p65 at Ser 536 Table 1. exposure of rats to HBO significantly reduces the activation of the NFκB protein [57]. Mattioli et al. reported that phosphorylation of NFκB-p65 at Ser 536 negatively regulates the import of NFκB-p65 to the nucleus [58]. Additionally, phosphorylation of NFκB-p65 at Ser 536 plays an important role in promoting the proteasomal degradation of NFκB-p65, thus leading to transcriptional termination of NFκB-dependent genes [59], and phosphorylation of NFκB-p65 at Ser 536 regulates the ability of nuclear IκBα to inhibit the NFκB binding to promoters in human leukocytes and thereby inhibit transcription of genes in a selective manner [60]. Furthermore, phosphorylation of NFκB-p65 at Ser 536 increases turnover of NFκB-p65, thereby inhibiting NFκB activity in macrophages and subsequently reducing inflammation [61]. Since iNOS expression is regulated primarily by control of gene transcription, a possible explanation for the decrease of iNOS protein level after HBOT may be that decreased activation of transcription factor NFκB-p65 subunit, which has a binding site on the iNOS gene promoter, consequentially leads to reduced iNOS expression. Here, we demonstrate that HBOT causes downregulation of NFκB-p65 protein and this decrease could be because of the decreased cytokine level after exposure to HBO [62]. Thus, we assume that HBOT downregulates iNOS activity/expression by a mechanism involving NFκB in combination with decreased CRP concentrations. Although CRP by itself did not affect NO synthesis, CRP further enhanced cytokine-evoked increases in iNOS mRNA protein levels [34]. Also, we postulate that decreased iNOS expression is associated with decreased expression of FFA translocase, a cluster of differentiation 36 (CD36), and a reduction in the level of FFA after exposure to HBOT. ERK1/2 is involved in the regulation of iNOS activation [63][64][65][66]. Increased ERK1/2 phosphorylation upregulates both NFκB activity and iNOS protein expression [64,[67][68][69][70], while inhibition of ERK1/2 phosphorylation downregulates the expression of cytokine-induced NFκBactivation [71]. Furthermore, inhibition of ERK1/2 decreases iNOS gene expression [71]. In our study, HBOT significantly decreases the phosphorylation of ERK 1/2, and this decrease could be due to the decreased NFκB activity [9]. Our results are in accordance with the study of Niu et al. [72]. Niu et al. reported that HBO suppressed the phosphorylation of ERK1/2 and decreased the synthesis of NO. Therefore, we assume that decreases in ERK1/2 phosphorylation after exposure to HBOT could lead to the reduction of NFκB activation and consequentially reduced expression of iNOS protein.
Akt is an upstream activator of NFκB, whereby phosphorylated Akt (or ERK1/2) activates NFκB inhibitor kinase  Table 1. and consequently leads to transcription and translation of iNOS [68,73]. Hattori et al. reported that the activation of NFκB is suppressed by inhibitors of Akt [74]. In our study, HBOT decreased Akt phosphorylation, indicating that activation of Akt is involved in mediating the effect of HBOT on the regulation of iNOS. Our results are in accordance with the study of Felderhoff-Mueser et al. who reported that hyperoxia reduces phosphorylation of Akt at Ser 473 [75].
Since the phosphorylation of Akt is necessary for NFκB activation, we suggest that a decrease in Akt phosphorylation leads to NFκB downregulation and a decrease in iNOS protein expression. To explain our results, we proposed the following model ( Figure 6). Altered level of FFA, CRP, and cytokines caused by T1DM induces inflammation which influences the iNOS/NO pathway and thus promotes vascular complications by attenuating insulin action. This is likely a consequence of increased ERK1/2/Akt activation, which further upregulates NFκB. There are several possibilities to explain the effect of HBOT in the modulation of iNOS. It is possible that HBOT leads to the decreased level of CRP, FFA, and cytokines, and consequently this leads to decreased activation of ERK1/2 and Akt activation and thus decreases the activity of NFκB, and subsequently iNOS or NFκB activation by cytokines in T1DM is achieved only through Akt.
In accordance to the promising results of this study that pointed out the early effects of HBOT on some molecular aspects of inflammation in T1DM patients, we are planning to carry out another study that will be performed on a more significant sample size with inclusion of patients suffering from metabolic syndrome and type 2 DM as well healthy control subjects. The currently presented prospective pilot study is based on the number of administered HBOT sessions included in one HBOT course and the exploration of some early inflammatory molecular effects of HBOT after one course of exposure. As the positive effects are detected in our study, the evaluation of long-term HBOT effects (e.g., after 1 or 5 years) regarding the molecular basis of the anti-inflammatory effects of the therapy is needed. Furthermore, there is a possibility to test the effects of different HBOT administration protocols and to select the most appropriate model (i.e., study plan could be performed either through the increase in the number of sessions involved in HBOT course or by shortening of the period between successive HBOT courses). Besides evaluating the molecular mechanisms involved in the reduction of inflammation, it is also necessary to design prediction models for major cardioand cerebrovascular atherosclerotic events and to examine  Table 1. the influence of various confounding variables that will be of great interest and contribution in the planning of primary and secondary prevention of major atherosclerotic accidents in diabetic patients and in improving patients' quality of life. The limitation of our study is a small number of T1DM patients exposed to HBOT. The other limitation of our study is that we did not directly measure the level of cytokines and we did not measure the level of NO in lymphocytes and the level of NFκB in the lymphocyte nucleus.
In conclusion, in the present study, we demonstrate for the first time that HBOT causes downregulation of iNOS activity/expression by a mechanism involving ERK1/2, Akt, and NFκB, in T1DM patients. The results from our study suggest that HBOT has a vital role in the reduction of inflammation in T1DM patients. Future studies on a larger population size are needed to elucidate the molecular mechanism by which HBOT regulates iNOS activity/expression in T1DM patients with vascular complications.

Data Availability
The data used to support the findings of this study are available from the corresponding author upon request.

Disclosure
Results presented in this work are part of the PhD thesis of MSc Ivana Resanovic.

Conflicts of Interest
The authors confirm that this article content has no conflict of interest.