Effects of venlafaxine on the expression level and methylation status of genes involved in oxidative stress in rats exposed to a chronic mild stress

Abstract Recent human and animal studies indicate that oxidative and nitrosative stress may play a role in the aetiology and pathogenesis of depression. This study investigates the effect of chronic administration of the serotonin‐norepinephrine reuptake inhibitor, venlafaxine, on the expression and methylation status of SOD1, SOD2, GPx1, GPx4, CAT, NOS1 and NOS2 in the brain and blood of rats exposed to a chronic mild stress (CMS) model of depression. Separate groups of animals were exposed to CMS for 2 or 7 weeks; the second group received saline or venlafaxine (10 mg/kg/d, IP) for 5 weeks. After completion of both stress conditions and drug administration, the mRNA and protein expression of selected genes and the methylation status of their promoters were measured in peripheral mononuclear blood cells (PBMCs) and in brain structures (hippocampus, amygdala, hypothalamus, midbrain, cortex, basal ganglia) with the use of TaqMan Gene Expression Assay, Western blot and methylation‐sensitive high‐resolution melting techniques. CMS caused a decrease in sucrose consumption, and this effect was normalized by fluoxetine. In PBMCs, SOD1, SOD2 and NOS2 mRNA expression changed only after venlafaxine administration. In brain, CAT, Gpx1, Gpx4 and NOS1 gene expression changed following CMS or venlafaxine exposure, most prominently in the hippocampus, midbrain and basal ganglia. CMS increased the methylation of the Gpx1 promoter in PBMCs, the second Gpx4 promoter in midbrain and basal ganglia, and SOD1 and SOD2 in hippocampus. The CMS animals treated with venlafaxine displayed a significantly higher CAT level in midbrain and cerebral cortex. CMS caused an elevation of Gpx4 in the hippocampus, which was lowered in cerebral cortex by venlafaxine. The results indicate that CMS and venlafaxine administration affect the methylation of promoters of genes involved in oxidative and nitrosative stress. They also indicate that peripheral and central tissue differ in their response to stress or antidepressant treatments. It is possible that that apart from DNA methylation, a crucial role of expression level of genes may be played by other forms of epigenetic regulation, such as histone modification or microRNA interference. These findings provide strong evidence for thesis that analysis of the level of mRNA and protein expression as well as the status of promoter methylation can help in understanding the pathomechanisms of mental diseases, including depression, and the mechanisms of action of drugs effective in their therapy.


| INTRODUC TI ON
Depression is the most common of all mental disorders, with around 350 million sufferers Worldwide. According to WHO estimation, depression will become the second most common health problem in the World by 2020. 1,2 Furthermore, patients with depression are characterized by an increased risk of the development of somatic disease, including cardiovascular disease, 3 obesity, 4 diabetes, 5 cancer 6 and cognitive impairment. 7 Despite intensive research, the ethology and pathogenesis of depression remain unclear. 8 Moreover, around 30% of depressed patients do not benefit from antidepressant drug treatment. 9,10 Unfortunately, despite extensive studies, the pathophysiology of depression is not fully known, because the condition is not associated with any gross alternations of brain. 11 On the other hand, a rapidly growing body of evidence suggests the involvement of various factors in the pathophysiology of depression, including inflammation, DNA damage and impairment of tryptophan transformation. There are also reports suggesting that affective disorders, including depression, are characterized by disturbances of the oxidant/antioxidant balance, that in consequence lead to damage of cellular macromolecules. [12][13][14][15][16][17] Reactive oxygen and nitrogen species can cause changes in cell membranes, receptor and enzymes functions, as well as activity of genes, and these imbalances can be influenced by antidepressant treatments. 14,18,19 Among the signs of antioxidant defects, increased levels of 8-F2-isoprostane (8-iso-PGF2α), malondialdehyde (MDA) and 8-hydroxy-2-deoxyguanosine have been found in depressed patients and are also believed to be markers of the effectiveness of an antidepressant treatment. [20][21][22][23] For example, an enhanced concentration of MDA is normalized after 3 months of SSRI treatment. 24,25 Similarly, chronic mild stress (CMS) model was associated with increased protein peroxidation-carbonyl-level in prefrontal, hippocampus, striatum and cortex as compared to control rats. 26 Moreover, MDA level was increased in cerebellum and striatum of stressed rats. 26 However, in another study, bupropion or sertraline increased 8-iso-PGF2α in patients with major depression and the severity of the disease was strongly correlated with an increased level of F2 isoprostane. 27 Decreased levels of enzymatic and non-enzymatic antioxidants can also play a crucial role in the mechanisms of depression. For example, a low glutathione peroxidase (Gpx) and superoxide dismutase (SOD) activity have been observed in depressed patients, with this decrease being correlated with the severity of the disease [28][29][30][31] ; however, opposite effects have also been reported. 25,32,33 Additionally, antidepressant therapy may lead to partly improve of antioxidant response. 33 Similarly, previous studies found that chronic ozone inhalation induced depression-like symptoms, including anxiety, and reduced cortical and hippocampal SOD and CAT activity. 34,35 Interestingly, next study showed that the antidepressant therapy caused a reduction myeloperoxidase activity in the amygdala, hippocampus, amygdala and prefrontal cortex while SOD and CAT activity is believed to increase in nucleus accumbens of rat brain after antidepressant therapy. 36 Another important oxidative enzyme associated with depression is xanthine oxidase (XO); a post-mortem study had shown an increased level of XO in serum and the thalamus. 37,38 In addition, patients with depression have been characterized by decreased levels of vitamin E, vitamin C, zinc, uric acid, coenzyme Q10 and glutathione (GSH), although the latter did not change following electroconvulsive therapy. 19,32,[39][40][41][42][43][44] Moreover, an animal study showed that the level of GSH was reduced in cerebral cortex, hypothalamus and brain stem regions of stressed mice, although it was not changed in cerebellum. 45 Nitric oxide (NO) is another important chemical associated with the mechanisms of depression. Low levels of NO help facilitate the release of dopamine and noradrenalin, but high concentrations lead to nitration and hypernitrosylation of amino acids and proteins. This results in the generation of highly reactive substances: NO-tyrosine, NO-tryptophan and NO-arginine. 46,47 It was found that the levels of IgM against NO-aspartate, NO-phenylalanine and NO-tyrosine are increased in serum of depressed patients and the inhibition of NO synthesis can induce an antidepressant-like effect. 48 Thus, a NO inhibitor can increase the effectiveness of serotonergic antidepressants (SNRIs) and can be applied to patients with treatment-resistant depression. 48 Furthermore, an animal study suggests that the depression may be associated with increased activity of endothelial NOS (eNOS) and increased nNOS (neuronal NOS; NOS1) protein and mRNA expression in the hippocampus. 49 However, Yoshino et al 50 found that antidepressant treatment increased nNOS mRNA by other forms of epigenetic regulation, such as histone modification or microRNA interference.
These findings provide strong evidence for thesis that analysis of the level of mRNA and protein expression as well as the status of promoter methylation can help in understanding the pathomechanisms of mental diseases, including depression, and the mechanisms of action of drugs effective in their therapy.

K E Y W O R D S
brain structures and blood cells, chronic mild stress model, gene expression and methylation, oxidative and nitrosative stress, rat, venlafaxine expression in hippocampus, midbrain, cerebellum and olfactory bulb, and iNOS (inducible NOS, NOS2) mRNA expression in frontal cortex and midbrain, and decreased eNOS mRNA expression in most brain regions.
The above data suggest that the mechanisms of depression can be associated with disturbances in the balance between oxidants and antioxidants. Thus, antioxidant agents may be an effective antidepressant therapy. Molecular hydrogen has antioxidative activities, and the mice after inhalation of hydrogen were characterized by decreased pathological damage, neuronal apoptosis and BBB disruption and reversed the cognitive decline. 51 Similarly, Gao et al 52 found that that repeated inhalation of hydrogen-oxygen mixed gas decreased both the acute and chronic stress-induced depressive-and anxiety-like behaviours of mice. The next antioxidant compound-vanillin-inhibits the protein oxidation and lipid peroxidation in hepatic mitochondria. Thus, many previous studies showed that the vanillin relieved symptoms of CMS and it may be a potential antidepressant. [53][54][55] Moreover, Amira et al 55 found that CMS procedure caused an increase of lipid peroxidation and a decrease of GSH and serotonin in the brain. Sesamol is another antioxidant agent, which exerted antidepressant-like effects, since it reversed the unpredictable chronic stress-induced behavioural, including increased immobility period and reduced sucrose preference and biochemical parameters (increased lipid peroxidation and nitrite levels; decreased GSH levels, SOD and catalase activities) in stressed mice. 41 Human studies also confirmed that antioxidants, including N-acetylcysteine, may relieve symptoms of depression. 56 On the other hand, a growing body of evidence suggests that antidepressants, including SSRIs, serotonin norepinephrine reuptake inhibitors (SNRIs) and tricyclic antidepressants (TCAs), may have antioxidant action. 57 Therefore, a chronic treatment of imipramine increased SOD and CAT activity and decreased lipid and protein damage in prefrontal cortex and hippocampus of rats. 58 Similarly, Zafir et al 59 found that the activities of SOD, CAT, GST, GR and GSH levels in the rat brain increased after fluoxetine and venlafaxine administration. Additionally, the therapy prevented lipid and protein oxidative damage induced by stress.
Therefore, this study aimed to investigate whether: (a) the CMS procedure, used as an validated animal model of depression 59-61 changes the expression of SOD1, SOD2, GPx1, GPx4, CAT, NOS1 and NOS2 at the mRNA and protein levels in peripheral blood mononuclear cells (PBMCs) and is selected brain structures (hippocampus, amygdala, midbrain, hypothalamus, cerebral cortex and basal ganglia); (b) chronic administration of serotonin-norepinephrine reuptake inhibitor, venlafaxine, affects the expression of these genes; (c) the CMS procedure and venlafaxine administration cause epigenetic changes, that is methylation level of these gene promoters; (d) a degree to which these changes in methylation affect the genes expression; and (e) the changes observed in PBMCs can serve as markers of similar changes in the brain. The last point has an important clinical implication, as there is a great need for peripheral markers that would allow earlier diagnosis, more precise prognosis of pharmacotherapy outcome, and more personalized therapies of the mood disorders. All of the genes analysed in our study are located on chromosomes significantly associated with depression (Table S1).

| Animals
The study was carried out on male Wistar Han rats (Charles River), brought into the laboratory for adaptation to housing conditions 1 month before the start of the experiment. Except as described below, the animals were singly housed with freely available food and water, and maintained on a 12-hour light/dark cycle (lights on at 8.00) in a constant temperature (22 ± 20°C) and humidity (50 ± 5%) conditions. All procedures used in this study conform to the rules and principles of Directive 86/609/EEC and have been approved by the Bioethical Committee at the Institute of Pharmacology, Polish Academy of Sciences, Krakow, Poland.

| Chronic mild stress procedure
The CMS procedure was conducted as previously described. [60][61][62] Briefly, the animals (200-220 g) were first trained to consume a 1% sucrose solution in seven once weekly baseline tests, in which sucrose solution was presented for 1 hour following 14-hour food and water deprivation in the home cage. Subsequently, sucrose consumption was monitored once weekly, under similar conditions, until the end of the study. On the basis of their sucrose intakes in the final baseline test, the animals were divided into two matched groups.
One group was exposed to the stress procedure for a period of 2 or 7 weeks. Each week of the stress regime consisted of two periods of food or water deprivation, two periods of 45-degree cage tilt, two periods of intermittent illumination (light on and off every 2 hours), two periods of soiled cage (250 mL water in sawdust bedding), one period of paired housing, two periods of low intensity stroboscopic illumination (150 flashes/min), and three periods of no stress. The duration of all stressors was 10-14 hours, and they were used individually and continuously, day and night. The control, non-stressed animals were housed in separate rooms, with no contact with the stressed animals. They were deprived of food and water for 14 hours before each weakly sucrose test, but otherwise food and water were available ad libitum. In all CMS studies conducted in our laboratory, the stressed animals display a gradual decrease in the consumption of the sucrose solution to approximately 40% of pre-stress values.
When this effect stabilized, that is after initial 2 weeks of stress, the animals were either decapitated (see below) or further divided into matched subgroups, and for subsequent 5 weeks, they received daily administration of vehicle (1 mL/kg, IP) or venlafaxine (10 mg/kg, IP).
The drug was administered to all animals (ie control and stressed) at approximately 10.00, and the weekly sucrose tests were carried out 24 hours after the last dose.
Twenty-four hours after the sucrose test conducted following initial 2 weeks of stress, that is before venlafaxine administration was commenced (groups: Control and Stressed) and after the final sucrose test conducted following 7 weeks of stress, that is after completion of 5 weeks administration (groups: Control/Venla, Stressed/ Saline, Stressed/Venla), the animals were decapitated and the blood and brain samples were collected (see below).

| Specimen collection and RNA/DNA isolation from animal brain
After blood harvesting, hippocampus, amygdala, midbrain, hypothalamus, cerebral cortex and basal ganglia were removed, rapidly frozen in liquid nitrogen and stored at −80°C until further analysis.
A sufficient volume of PBS was added to each samples, which were then homogenized by a FastGene ® Tissue Grinder (Nippon Genetics Europe). The homogenized samples were sonicated, centrifuged and rinsed with PBS using a commercial kit (ISOLATE II RNA/DNA/ Protein Kit; Bioline), according to the manufacturer's instructions.
Finally, the total concentration and purity of the RNA and DNA isolated from each sample was determined using a spectrophotometer as a 260/280 nm OD ratio with expected values between 1.8 and 2.0. Finally, the RNA and DNA samples were stored at −20°C until further analysis.

| Reverse transcription and gene expression
The reverse transcription reaction was performed using a Highcapacity cDNA Reverse Transcription Kit (Applied Biosystems).
Briefly, the following components were mixed to form a 20 µL reaction volume: nuclease-free water; 10xRT Buffer; 10xRT Random Primers; 25xdNTP Mix (100 mM); total RNA (0.5 ng/µL) and MultiScribe ® Reverse Transcriptase. The reaction tubes were incubated for 10 minutes at 25°C, then for 120 minutes at 37°C, and then finally for 5 minutes at 85°C to inactivate reverse transcriptase.

| Methylation and HRM analysis
The methylation status of gene promoters was specified by methylation-sensitive high-resolution melting. 64,65 Of the gene promotors studied in this assay, only the NOS2 promoter did not contain any CpG islands. In the case of genes expression regulated by many promoters, primers were designed for promoters containing CpG islands. Next, we designed primers using the Methyl Primer experiments; these were prepared following the same procedure as the blood and brain DNA samples.
In order to control for the sensitivity of methylation detection, serial dilutions were performed: non-methylated, 10% methylated, 25% methylated, 50% methylated, 75% methylated and 100% methylated DNA ( Figure S1). These methylation analyses were performed using the Bio-Rad CFX96 Real-Time PCR Detection System and analysed in HRM Powered by Precision Melt Analysis™ Software (Bio-Rad Laboratories Inc). Each PCR contained 5× HOT FIREPol ® EvaGreen ® HRM Mix (no ROX) (Solis BioDyne), 500 nM of each primer and 10 ng of DNA after bisulphite modification (theoretical calculation) and was performed as three replicates. The parameters for amplification and HRM analyses included initial activation for 12 minutes at 95℃, 45 cycles of 95℃ for 15 seconds; annealing at optimal primer temperatures (tested experimentally) for 20 seconds (see Table 1 for characteristics of the primers), and elongation at 72℃ for 20 seconds. The HRM analysis consisted of denaturation at 95℃ for 15 seconds, reannealing at 60℃ for 1 minute and melting from 60 to 95℃ at a ramp rate of 0.2℃.

| Protein samples isolated from animal brains and the Western blot procedure
For the analysis, we selected proteins that are crucial for the regulation of oxidative stress, occurred in various cell locations (cytoplasm, mitochondria and cell nucleus) and have not been investigated in animal models of depression so far. The protein fractions isolated from the tissue of the rat brains were used for Western blotting. 98 Frozen samples of the brain parts were homogenized using a FastGene ® Tisue Grinder (Nippon Genetics Europe), in a RIPA buffer (10 mM Tris-HCl PH 8.0, 1 mM EDTA, 1% Triton X-100, 0.1% sodium deoxycholate, 0.1% SDS, 10 mM NaCl) containing 1 mM phenylmethylsulfonyl fluoride (PMSF, serine protease inhibitor). The homogenates were then sonicated twice for 15 seconds each in an ice bath and centrifuged at 2500 g for 5 minutes at 4℃. The supernatant was then collected. The protein content of the homogenate tissue fractions was estimated by means of the Lowry procedure, 71

| Drugs
Venlafaxine HCl (Carbosynth Ltd) was dissolved in 0.9% sterile saline, which was used for vehicle administration, and administered IP in a volume of 1 mL/kg of body weight, at the dose of 10 mg/kg, as used previously. 61,62

| Statistical analysis
The effect of initial 2 weeks of CMS on sucrose consumption was analysed by test t when the data were normally distributed or Mann-Whitney rank-sum Test when the data were not normally distributed. The sucrose intakes, gene expression and methylation data were analysed using an one-way analysis of variance (one-way ANOVA), followed by post hoc Tukey's test, where F ratios were significant for the following groups: Controls/Vehicle, Stressed/Vehicle and Stressed/Venlafaxine when the data were normally distributed. Otherwise, the data were analysed using a Kruskal-Wallis one-way analysis of variance on ranks followed by post hoc Student-Newman-Keuls test, where H ratios were significant for the studied groups. Student's t test was used to analyse differences between blood and brain samples. P values <.05 were considered significant. Statistics were calculated using Statistica 12 (StatSoft), SigmaPlot 11.0 (Systat Software Inc) and GraphPad Prism 5.0 (GraphPad Software, Inc).

| Sucrose intakes and body weights of animals exposed to CMS and venlafaxine
As shown in Table 2, before the stress was initiated (Week 0), the consumption of 1% sucrose solution was comparable in all groups. Following initial 2 weeks of stress (Week 2), the intakes decreased by approximately 40% (F = 6.76, fd = 2, P < .01, Tukey's test P = .01) and remained low until the end of experiment (Week TA B L E 2 Sucrose intakes in animals exposed to chronic mild stress (CMS) for 2 wk (Week 2) and in animals exposed to CMS for 7 wk (Week 7) and administered vehicle (1 mL/kg) or venlafaxine (10 mg/kg) for 5 wk

| Gene expression in PBMCs
The expression of Gpx1, Gpx4 and CAT mRNA in PBMCs did not differ between the two groups exposed to the control and stress F I G U R E 2 mRNA expression of CAT, Gpx1, Gpx4, SOD1, SOD2 and NOS1 in brain structures of animals exposed to chronic mild stress (CMS) for 2 wk (Control, Stressed) and in animals exposed to CMS for 7 wk and administered vehicle (1 mL/kg) or venlafaxine (10 mg/kg) for 5 wk (Control/Venla, Stressed/Saline, Stressed/Venla). Relative gene expression levels were estimated using a 2 −ΔCt (Ctgene-Ct18S) method. Data represent means ± SEM. N = 6. *P < .05; relative to Stressed/Saline group. # P < .05; relative to Control group conditions for 2 weeks (Table S2), as well as between the control and stressed animals treated with saline (Figure 1.). Chronic (5 weeks) administration of venlafaxine had no effect in control animals but significantly increased the expression of SOD1 (F = 6.49, df = 2, P < .001, Tukey's test P < .001), SOD2 (F = 7.20, df = 2, P < .01, Tukey's test P < .05) and NOS2 (F = 6.49, df = 2, P < .001) Tukey's test P < .001) genes in the stressed rats after venlafaxine therapy.

| Gene expression in brain
As shown in Figure

| Comparison of gene expressions in PMBCs and brain
In order to compare effects of CMS and venlafaxine on the expression of CAT, Gpx1, Gpx4, SOD1, SOD2 and NOS1 genes in the brain and in blood, all data measured in the brain structures were pooled and compared with the changes detected in PMBCs. As shown in Figure 3, in all groups the expression of Gpx1 was higher, and the expression of GPx4 lower, in the PBMCs than in the brain structures (P < .01 and P < .05, respectively). The expression of CAT and SOD2 were lower in the brain tissue of the stressed rats treated with venlafaxine (P < .001). In addition, the expression of SOD1 in brain tissue was higher than that observed in PBMCs of the stressed rats and the stressed rats treated with saline and venlafaxine (P < .01).

| Effect of venlafaxine on the expression of genes in PMBCs and brain structures
Venlafaxine significantly decreased the expression of CAT and Gpx4 in amygdala (P < .05) and midbrain (P < .01), when compared to the expression of these genes in the PMBCs (Figure 4).

Administration of venlafaxine caused a down-regulation of Gpx1
and SOD1 expression ( Figure 4) in hippocampus (P < .01, P < .05, respectively), amygdala (P < .01), midbrain (P < .001, P < .01, respectively) and cerebral cortex (P < .05, P < .01, respectively), as compared to its effects in PBMCs. No differences in the effect of venlafaxine on SOD2 expression was found between the blood and brain tissue.

| Methylation in PBMCs
The only significant change that was found in PBMCs was an increased methylation of Gpx1 promoter region in animals exposed to the CMS procedure for 2 weeks (H = 12.83, df = 2, P < .05, Tukey's test P < .05; Figure 5).
No significant differences between groups were observed for promoters of other genes in PBMCs (Table S3).

| Methylation in brain
As shown in Figure 6, the CMS procedure significantly increased methylation level of the second promoter of Gpx4 in midbrain

| Comparison of the methylation status of gene promoters in PBMCs and brain
In order to compare the effects of CMS and venlafaxine on the brain and peripheral expression of CAT, Gpx1, Gpx4, SOD1, SOD2 and NOS1 genes, all results obtained for the brain structures were pooled and compared with changes detected in the PBMCs ( Figure 7).
The untreated and venlafaxine-treated control animals showed an increased methylation in the promoter region of CAT and the third Gpx4 promoter in brain tissue, compared to PBMCs (P < .05). In the stressed animals treated with venlafaxine, the methylation level of the second Gpx4 promoter was lower in the brain than in PBMCs (P < .01). In all groups, the third promoter of NOS1 demonstrated lower methylation status in brain tissue than in PBMCs (P < .001).
No other significant differences were observed for other promoters (Table S5). Gpx1 (B), Gpx4 (C), SOD1 (D) and SOD2 (E) gene expression between brain tissue and PBMCs of animals exposed to chronic mild stress (CMS) for 2 wk (Control, Stressed) and in animals exposed to CMS for 7 wk and administered vehicle (1 mL/ kg) or venlafaxine (10 mg/kg) for 5 wk (Control/Venla, Stressed/Saline, Stressed/Venla). The mRNA relative expression levels of all five genes measured in hippocampus, amygdala, hypothalamus, midbrain, cortex and basal ganglia were pooled and compared with the data detected in PBMCs. Relative gene expression levels were estimated using a 2 −ΔCt (Ct gene-Ct 18S) method. Data represent means ± SEM. N = 6. *P < .05, **P < .01, ***P < .001

| Effect of venlafaxine on the methylation status of gene promoters in PBMCs and brain structures
Compared to the effect of venlafaxine in PBMCs, the drug caused a stronger increase in the methylation level of CAT promoter in amygdala and hippocampus (P < .05) and the NOS1 seventh promoter in hippocampus and midbrain (P < .01) (Figure 8).

| Gene expression on the protein level
As shown in Figure (Table S6).

| D ISCUSS I ON
The present study is the first to investigate effects of repeated administration of venlafaxine on the expression of selected genes in the PBMCs of rats exposed to the CMS model of depression.
Our results also demonstrate the CMS-and venlafaxine-induced Western blot or immunosorbent assays.
The first studied gene encodes catalase (CAT), an important antioxidative enzyme that decomposes hydrogen peroxide into water and oxygen. 68 It is believed to be involved in the mechanisms of depression but the exact nature of this involvement remains unclear.
Our findings indicate that CAT expression in hippocampus was lowered in animals exposed to the CMS procedure, and this finding is  Gpx1 (B), Gpx4 (C), SOD1 (D) and SOD2 (E) in PBMCs and in brain structures of animals exposed to chronic mild stress (CMS) for 2 wk (Control, Stressed) and in animals exposed to CMS for 7 wk and administered vehicle (1 mL/kg) or venlafaxine (10 mg/kg) for 5 wk (Control/Venla, Stressed/Saline, Stressed/Venla). The effects are presented as fold change (2 −ΔΔCt method 51 ). Data represent means ± SEM. N = 6. *P < .05, **P < .01, ***P < .001; relative to PBMCs F I G U R E 5 Methylation level of Gpx1 promoter in PBMCs of animals exposed to chronic mild stress (CMS) for 2 wk (Control, Stressed) and in animals exposed to CMS for 7 wk and administered vehicle (1 mL/kg) or venlafaxine (10 mg/kg) for 5 wk (Control/ Venla, Stressed/Saline, Stressed/Venla). Data represent median and maximum-minimum values. N = 6. *P < .05 is present only in the cytosol, Gpx4 is located in the membrane and mitochondria. 73 Our findings indicate that the mRNA expression of Gpx1 was increased in midbrain of the CMS rats. In contrast, Eren et al 74 found that Gpx activity in the cortex is lowered in rats subjected to the CMS procedure and that venlafaxine administration normalized this effect. In humans, Gpx1 activity was found to be decreased in the haemolysed erythrocytes of depressed women 32 and increased in the erythrocytes of patients with melancholia. 75 Moreover, the changes in the Gpx1 concentration have been found to be negatively correlated with the severity of depressive symptoms. 76 Our present findings indicate that the midbrain and hippocampal expression of Gpx1 mRNA was significantly lower in the venlafaxine-treated stressed animals than in the vehicle-treated stressed rats. Accordingly, the activity of Gpx in serum was found to decrease over the course of 3-month administration of different SSRIs. 75 In the present study, higher Gpx1 mRNA expression was observed in PBMCs than in brain tissue in all of the stress and treatment conditions, and the methylation of the Gpx1 promoter in PBMCs was lower in the control than the CMS rats: a key novel finding of other crucial regulation of mRNA gene expression.
This study is also the first to examine the changes in the level of Gpx4 isoform in an animal model of depression. The CMS procedure increased the expression of Gpx4 mRNA in midbrain and basal ganglia, and this effect was normalized by administration of venlafaxine.
Interestingly, the methylation levels of the second Gpx4 promoter in these two brain regions were also higher in the stressed group. This however did not affect the level of mRNA expression, because the Gpx4 gene contains six promoters and only one of them was found to have higher methylation status in animals exposed to the CMS procedure. CMS also increased the Gpx4 protein level in hippocampus, and venlafaxine caused a further increase in Gpx4 protein expression while the opposite trend was observed in cerebral cortex.
The differences between the expression of Gpx4 mRNA and protein could be a result of epigenetic or posttranscriptional modifications, for example the influence of microRNA. Moreover, the presence of a high level of Gpx4 mRNA expression and a high methylation status F I G U R E 6 Methylation levels of Gpx4 promoter 2 in midbrain (A) and basal ganglia (B), SOD1 (C) and SOD2 (D) promoter in hippocampus of animals exposed to chronic mild stress (CMS) for 2 wk (Control, Stressed) and in animals exposed to CMS for 7 wk and administered vehicle (1 mL/kg) or venlafaxine (10 mg/kg) for 5 wk (Control/Venla, Stressed/Saline, Stressed/Venla). Data represent means ± SEM (A) or median and minimum-maximum values (B-D). N = 6. *P < .05, **P < .01, ***P < .001 this finding is consistent with our observation of increased SOD1 F I G U R E 7 Differences in methylation level of CAT (A), Gpx4 promoter 2 (B), Gpx4 promoter 3 (C) and NOS1 promoter 3 (D) between brain tissue and PBMCs of animals exposed to chronic mild stress (CMS) for 2 wk (Control, Stressed) and in animals exposed to CMS for 7 wk and administered vehicle (1 mL/kg) or venlafaxine (10 mg/kg) for 5 weeks (Control/Venla, Stressed/Saline, Stressed/Venla). The methylation level of all four genes measured in hippocampus, amygdala, hypothalamus, midbrain, cortex and basal ganglia were pooled and compared with the data detected in PBMCs. Data represent median and maximum-minimum values. N = 6. *P < .05, **P < .01, ***P < .001 Similarly, while SOD2 activity was previously found to be elevated in brain tissue of animals treated with venlafaxine, 59 our findings indicate that hippocampal SOD2 mRNA expression is decreased after venlafaxine therapy. There is evidence that overexpression of SOD2 could play a crucial role in reducing oxidative stress and preventing neurodegenerative disease 83 and that reduced SOD2 expression in the cerebral cortex, cerebellum and basal ganglia may lead to neurodegeneration. 84 The differences between blood and brain in rats after therapy may be result of different tissue response to treatment. Probably, enzyme activity might have been more susceptible to antidepressant therapy in blood than in the brain. Moreover, these changes in SOD2 concentration may be responsible for the reductions in the volume of prefrontal cortex and hippocampus that are characteristic of patients with depression. 85 Therefore, the reduction in brain structure volume may cause irreversible changes in antioxidant capacity, resulting in no response to treatment in brain.
Our present results show that the SOD2 promoter region displayed higher methylation status in the stressed groups, suggesting that F I G U R E 8 Differences in methylation level of CAT (A), Gpx1 (B), Gpx4 promoter 2 (C), Gpx4 promoter 3 (D), SOD1 (E), SOD2 (F), NOS1 promoter 3 (G) and NOS1 promoter 7 (H) between brain tissue and PBMCs of animals exposed to chronic mild stress (CMS) for 2 wk (Control, Stressed) and in animals exposed to CMS for 7 wk and administered vehicle (1 mL/kg) or venlafaxine (10 mg/kg) for 5 wk (Control/Venla, Stressed/Saline, Stressed/Venla). The methylation level of all eight genes measured in hippocampus, amygdala, hypothalamus, midbrain, cortex and basal ganglia were pooled and compared with the data detected in PBMCs. Data represent means ± SEM. N = 6. *P < .05, **P < .01; relative to PBMCs F I G U R E 9 mRNA expression of CAT in brain structures of animals exposed to chronic mild stress (CMS) for 2 wk (Control, Stressed) and in animals exposed to CMS for 7 wk and administered vehicle (1 mL/kg) or venlafaxine (10 mg/kg) for 5 wk (Control/Venla, Stressed/Saline, Stressed/Venla). A, Representative Western blot analysis of the effects in midbrain and cerebral cortex. C-controls, S-stressed for 2 wk, C/V-Control/Venlafaxine, S/S-Stressed/Saline, S/V-Stressed/Venlafaxine. B, Level of CAT proteins measured in hippocampus, amygdala, hypothalamus, midbrain, cortex and basal ganglia. Samples containing 25 μg of proteins were resolved by SDS-PAGE. The intensity of bands corresponding to CAT was analysed by densitometry. Integrated optical density (IOD) was normalized by protein content and a reference sample (see the Methods section for details). The graphs show the mean IODs of the bands from all analysed samples. The IOD gene /IOD ACTB method was used to estimate the relative protein expression levels in the analysed samples. Data represent means ± SEM. N = 6. **P < .01, ***P < .001; relative to Stressed/Saline group Overproduction of NO can lead to oxidative and antioxidative imbalance, which is believed to be involved in pathomechanisms of neurodegenerative diseases, including anxiety and depression. 86,87 Previous studies confirmed that depressed patients are characterized by an increased level of NO in plasma. 88,89 In line with this, the CMS rats and mice showed an increased level of NO in cortex. 74,90 Our results also indicate that the NOS1 mRNA expression in midbrain and basal ganglia was higher in rats exposed to the CMS procedure. However, no changes were observed in the stressed animals administered venlafaxine, which is inconsistent with the results ob- The presence of elevated NO levels in depressed patients could be a result of an increased expression of cellular NOS1 in suprachiasmatic nucleus, cornu ammonis area 1 and subiculum regions. 92 The serum NO level was significantly decreased in depressed patients after 8 weeks of SSRI treatment, 37 while an earlier animal study indicated that CMS caused a deformation of neurons in the hippocampus, while fluoxetine therapy led tonormalization of these neurons by inhibiting NOS1. 93 Hence, it is likely that CMS could cause a nitrosative stress that is inhibited by venlafaxine-induced reduction in NO generation. However, the level of NO in cortex of the CMS animals was found to be reduced following venlafaxine treatment. 85 In the present study, no significant differences were detected between any groups with regard to NOS1 protein level in the brain. Talarowska et al (2015) found F I G U R E 1 0 mRNA expression of Gpx4 in brain structures of animals exposed to chronic mild stress (CMS) for 2 wk (Control, Stressed) and in animals exposed to CMS for 7 wk and administered vehicle (1 mL/kg) or venlafaxine (10 mg/kg) for 5 wk (Control/Venla, Stressed/ Saline, Stressed/Venla). A, Representative Western blot analysis of the effects in hippocampus and cerebral cortex. C-controls, S-stressed for 2 wk, C/V-Control/Venlafaxine, S/S-Stressed/Saline, S/V-Stressed/Venlafaxine. B, Level of Gpx4 proteins measured in hippocampus, amygdala, hypothalamus, midbrain, cortex and basal ganglia. Samples containing 25 μg of proteins were resolved by SDS-PAGE. The intensity of bands corresponding to CAT was analysed by densitometry. Integrated optical density (IOD) was normalized by protein content and a reference sample (see the Methods for details). The graphs show the mean IODs of the bands from all analysed samples. The IOD gene / IOD ACTB method was used to estimate the relative protein expression levels in the analysed samples. Data represent means ± SEM. N = 6. **P < .01, ***P < .001; relative to Stressed/Saline group. **P < .01, ***P < .001; relative to Stressed/Saline group. # P < .05; relative to Control group no significant difference in NOS2 levels between patients with recurrent depressive disorders and those in a first episode of depression. 95 However, our findings indicate that in the CMS animals venlafaxine did increase mRNA expression of NOS2 in PBMCs compared to the untreated stressed group.
This study is the first to report the effect of CMS and antidepressant treatment on the expression and methylation status of genes involved in oxidative and nitrosative stress in both the blood and the brain in the same experimental setting. Such comparative analysis is not possible to carry out in humans, because it would require too many subjects, and the brain tissue would be available post-mortem only. Hence, the use of a validated animal model of depression, which can pinpoint most promising genes that can be then verified in depressed patients, seems to be most reasonable for identification peripheral markers of the central nervous system pathology. We found that the mRNA expression of Gpx1 was significantly higher, and that of Gpx4 significantly lower, in PBMCs than the brain structures in all groups. Also, the expression of SOD1 mRNA was higher in the brains of all control and stressed groups, except for the stressed animals successfully treated with venlafaxine, which showed higher expression of this gene in the PBMCs. Similar enhanced expression in the PBMCs of the venlafaxine-treated stressed animals was observed for the CAT and SOD2 genes; however, in other groups these genes either did not change (CAT) or changed regardless of the conditions (SOD2).
The present findings and those of previously reported by others [28][29][30][31]50,95 suggest that the lack of sufficient enzymatic antioxidant defence in PBMCs and brain are involved in effects observed in animals exposed to the CMS procedure, a conclusion that could be cautiously extrapolated to the mechanisms of human depression. We can speculate that this inadequate antioxidant defence leads to increased production of toxic free radicals which, in excessive amounts, imply damage to biomolecules. In consequence, prolonged exposure to free radicals may cause a cell death, neuronal and glia atrophy leading to decreased grey matter volume, reduced cell numbers and low glucose metabolism observed in depressed patients. 96,97 On the other hand, antidepressant therapy may modulate the activity of enzymes restoring this antioxidant defence system to its proper functioning. Accordingly, our findings confirm that venlafaxine reduces the mRNA expression of the genes encoding the SOD1, SOD2, Gpx1, Gpx4, CAT and NOS1 enzymes in the brain, but increases that of SOD1, SOD2 and NOS2 in PBMCs. Only brain expression of Gpx1, Gpx4 and NOS1 mRNA was reduced to control lev- Additionally, the disease genes for depression are unknown, and thus, many transgenic models and models based on behavioural selection might not represent valid models of the disease. 11

| CON CLUS ION
Our findings indicate that oxidative and nitrosative stress are in-

CO N FLI C T O F I NTE R E S T
The authors declare no conflicts of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.