**4. Discussion**

In the presented experiment we employed a wide range of biochemical assays to search for a link between OS expressed as antioxidant activity/concentration and oxidative damage products and salivary gland function in patients with HT in euthyreosis. To the best of our knowledge, this study is the first to assess the function of salivary glands and their redox balance in patients with HT in euthyreosis.

OS is the result of an imbalance between ROS production and neutralization, which leads to oxidative damage to tissues. In the case of autoimmune thyroiditis, OS is considered to be the result of a deficiency of thyroid hormones as well as autoimmunity and the associated inflammation. During the synthesis of thyroid hormones, iodine is oxidized by nicotinamide adenine dinucleotide phosphate oxidase (NOX). Hydrogen peroxide (H2O2) formed in this reaction is used for the production of thyroid hormones [37]. Excess of H2O2, e.g., due to excessive iodine substitution or deficiency of glutathione peroxidase (GPx, an enzyme involved in the neutralization of hydrogen peroxide and protecting thyroid tissue from oxidative damage), leads to apoptosis and necrosis of thyrocytes. Interestingly, in the presented study we demonstrated a 45% decrease of GPx activity in blood erythrocytes of HT patients, which, in our opinion, largely contributes to the observed increase in the generation of oxygen free radicals (↑76% TOS in serum). In the situation of increased concentration of ROS, particularly H2O2, increased immunogenicity of thyroid specific antigens (thyroglobulin and thyroid peroxidase) and intensified intercellular adhesion molecule-1 (ICAM-1) expression are observed, which results in increased generation of antibodies and autoimmune response as well as raises in inflammation [38]. The latter, in turn, directly increases H2O2 in thyroid epithelial cells and activates NOX in T and B lymphocytes, which further boosts ROS production [3].

The existence and extent of OS can be assessed based on the behavior of numerous biomarkers, including the measurement of the concentration/activity of antioxidants as well as the evaluation of the concentrations of oxidative modification products [39]. In our research, we used a wide panel of biomarkers of oxidative stress, because there is a belief that a single parameter does not reflect the size and severity of the OS phenomenon [40]. The parameters helpful in the assessment of OS are also total antioxidant capacity (TAC), total oxidative status (TOS) and oxidative stress index (OSI) [41]. It has been evidenced that TAC expresses the efficiency of both enzymatic and non-enzymatic antioxidant defense mechanisms, whereas TOS is the sum of all oxidants present in the sample. OSI illustrates the relationship between antioxidant mechanisms and oxidative molecules [42].

Our results showed a decrease in antioxidant potential (82%↓TAC) and increased production of free radicals as well as all products of oxidative modifications in the serum of patients with HT in euthyreosis compared to the controls, which confirms a shift in the redox balance towards oxidation and the existence of general OS. Decreased concentration of GSH deserves special attention as the reduction of GSH concentration is considered a causative factor in the development of autoimmune diseases by inhibition of IL-1 and T-cell receptor-mediated transduction signaling [1]. Increased concentration of GSH has been evidenced to be the result of GSH use through oxidation, conjugation or extrusion from the cell, and indicates highly raised OS [40]. Interestingly, we demonstrated a negative correlation between anti-TPO antibody and GSH concentrations, which is consistent with the results of Rostami et al. [2]. According to these authors, their findings prove that the presence of antibodies is a causative factor for excessive ROS production and that GSH is capable of inhibiting complement-mediated damage in HT. They also presume that GSH deficiency may indicate the occurrence of processes leading to oxidative stress activation and the development of immune intolerance.

Interestingly, changes in the salivary redox balance seem to reflect those observed in the blood, but it should be noted that the changes in saliva are more intense. Only for UA we see different directionality. Its reduced salivary concentration in patients with HT vs. control suggests that in the oral cavity UA

behaves as an antioxidant. It has been shown, that the increased concentration of UA, that we observe in the serum of patients with HT, shifts the redox balance towards the oxidation reaction and OS and may be a factor predisposing to cardiovascular diseases [43]. Moreover, the performed analyses did not reveal any correlation between the redox balance parameters in blood and saliva. The lack of correlation and saliva/blood ratio analysis may suggest that oxidative stress in the salivary glands is independent of general oxidative stress in the course of HT. What's more, the higher saliva/blood ratio with respect to some of the parameters studied, in the HT women group compared to the control, indicates that the observed changes are the result of processes occurring in the salivary glands, and are not the result of their passive blood transitions.

The large salivary glands together produce about 90% of the total saliva volume. The largest of them—the parotid gland—produces saliva mainly in response to the applied stimulation, hence any changes in stimulated saliva composition and amount are considered to reflect the function of the parotid gland. At rest, 2/3 of the total saliva amount is produced by the submandibular glands. Therefore, any variations in the composition and amount of NWS reflect the functioning of the latter glands [44,45].

Our study demonstrated a reduction in the antioxidant potential of parotid and submandibular salivary glands, confirmed by a 34% decrease in TAC in NWS and a 53% decrease in TAC in SWS, which manifests the inefficiency of antioxidant systems of these glands to eradicate ROS. Although we observed significantly increased CAT and Px activities in NWS and SWS, which may—to some extent—prove an adaptive mechanism of the salivary glands in response to excessive ROS production and be of great importance to oral health. Px and CAT maintain concentration salivary H2O2 at a level 8 to 14 μM [46], which is non-toxic to oral fibroblasts [47] and epithelial cells [48].

In the euthyreosis status in the course of HT, the causes of reduced antioxidant response should be seen as a result of depletion of the resources in the process of ROS neutralization or of the oxidative modification of polypeptide chains rather than as a result of decreased synthesis caused by reduced production of thyroid hormones. The impaired antioxidant defense may also be caused by non-enzymatic glycation of these enzymes, which explains the negative correlation between SOD activity in SWS and plasma glucose concentration in HT group [49], despite the fact that diabetes as well as insulin resistance were excluded in HT patients. Moreover, the negative correlation between TAC in NWS and SWS and the serum concentration of anti-TPO as well as between GSH and IL-1β concentrations in SWS confirm that the exhaustion of antioxidant sources in salivary glands is related to an elevated oxidative stress level due to autoimmunity-related inflammation and that increased concentration of autoimmune antibodies is a key factor for the enhancement of ROS production.

As our results show, the impaired saliva antioxidant barrier results in an increase in oxidative modifications of salivary proteins and lipids. So it is advisable to use exogenous antioxidants supporting endogenous antioxidant mechanisms. The results of our study revealed a significantly greater percentage increase in the concentration of TOS and OSI in SWS (↑201% and ↑588%, respectively) compared to NWS (↑81% and ↑191%, respectively) in HT female patients. A significant reduction of GSH concentration in SWS (↓57%) vs. NWS (↓23%) in HT women as well as a negative correlation between AOPP and GSH levels in SWS may be helpful in understanding the intensification of oxidative modifications to proteins in the parotid vs. submandibular salivary glands. It has been evidenced that the main role of GSH is in maintaining thiol groups of proteins at a reduced level, i.e., protecting proteins against oxidation [50,51]. We also found a positive correlation between MDA in NWS and anti-TG, LOOH in SWS as well as between anti-TG and LOOH in SWS and IL-1β concentration in SWS, which proves that oxidative damage to the lipids contained in salivary glands is boosted with an increase in autoimmunity-related inflammation in the course of HT.

The main physiological difference between the parotid and submandibular saliva is the fact that the former type of saliva is secreted mainly during eating, whereas the latter type is produced continuously and is responsible for maintaining the integrity of oral structures [22]. Maintaining the appropriate rate of both types of secretion is therefore equally important for the functioning of the body. Despite a higher intensity of OS and a more pronounced decrease in antioxidant defense in the parotid glands, the rate of stimulated saliva secretion did not differ significantly between the groups. However, our study demonstrated that HT patients lose the unstimulated salivary gland function. The median values of NWS and SWS secretion were within the standard limits assumed for proper saliva secretion. It is noteworthy that in 15 patients, NWS secretion was lower than 0.2 mL/min (but higher than 0.1 mL/min) and 5 HT group patients secreted stimulated saliva at a level lower than 0.7 mL/min, which proves the developed salivary gland failure in these patients, referred to as hyposalivation, and explains the significant intensification of subjective symptoms of reduced saliva secretion reported when completing the survey on a dry mouth. The performed analyses excluded Sjögren's syndrome, but, to be 100% certain, a biopsy of the salivary glands would be advisable. However, we were not granted permission for its performance from the Bioethics Committee. Evidence showed that proinflammatory cytokines, including the evaluated IL-1β, induce the activity of metalloproteinases, which leads to changes in the basement membrane of the salivary glands and the structure of receptors for neurotransmitters associated with saliva secretion [52]. A higher increase in IL-1β concentration in NWS (↑119%) vs. SWS (↑52%) as well as in the observed relationship between NWS and IL-1β concentration may confirm the destruction of acinar cell-basement membrane interaction by excessive production of MMPs followed by a decreased number of secretory units (acini and ducts) [53,54]. It has also been demonstrated that the presence of IL-β in the inflamed environment may inhibit the release of acetylcholine from the residual nerves, resulting in reduced saliva secretion [13,14,55]. The described phenomenon of extracellular matrix remodeling has been demonstrated as a cause of reduced saliva secretion in Sjögren's syndrome patients [56] and reduced activity of muscarinic neurotransmitters/receptors in the submandibular glands of diabetic patients [55].

Changes in salivary amylase activity are considered a determinant of sympathetic nervous system (SNS) activity. The stimulation ofβ-adrenergic receptor activity results in increased production/secretion and activity of salivary α-amylase and other salivary proteins [57]. In our study, we observed decreased salivary α-amylase activity with a simultaneous increase in protein concentration in both SWS and NWS, which could suggest a higher activity of this branch of autonomic nervous system. The negative correlation between salivary α-amylase activity and TOS in NWS may suggest that lowered activity of salivary α-amylase results from the use of this enzyme in the elimination of excessive amounts of ROS (↑TOS) or, more likely, from the oxidative modification of its polypeptide chain, leading to a loss of/significant reduction in enzymatic activity.

One of the limitations of this paper is the fact of determining only some parameters characterizing the redox balance. Perhaps the use of other biomarkers would change the results and conclusions, which of course can be considered a weak point of this experiment.
