Association Between Endothelial Dysfunction and Depression-Like Symptoms in Chronic Mild Stress Model of Depression ELENA V. BOUZINOVA, PHD, RIKKE NORREGAARD, PHD, DONNA M.B. BOEDTKJER, PHD, IRINA A. RAZGOVOROVA, MSC, ANABS M.J. MOELLER, BSC, OLGA KUDRYAVTSEVA, PHD, OVE WIBORG, PHD, CHRISTIAN AALKJAER, DMSC, AND VLADIMIR V. MATCHKOV, DMSC Objective: Cardiovascular diseases have high comorbidity with major depression. Endothelial dysfunction may explain the adverse cardiovascular outcome in depression; therefore, we analyzed it in vitro. In the chronic mild stress model, some rats develop depression-like symptoms (including ‘‘anhedonia’’), whereas others are stress resilient. Methods: After 8 weeks of chronic mild stress, anhedonic rats reduced their sucrose intake by 55% (7%), whereas resilient rats did not. Acetylcholine-induced endotheliumdependent relaxation of norepinephrine-preconstricted mesenteric arteries was analyzed in nonstressed, anhedonic, and resilient rat groups. Results: Small resistance arteries from anhedonic rats were less sensitive to acetylcholine than those of the nonstressed and resilient groups ( p = .029). Pathways of endothelium-dependent relaxation were altered in arteries from anhedonic rats. Nitric oxide (NO)Ydependent relaxation and endothelial NO synthase expression were increased in arteries from anhedonic rats (0.235 [0.039] arbitrary units and 155.7% [8.15%]) compared with the nonstressed (0.135 [0.012] arbitrary units and 100.0% [8.08%]) and resilient (0.152 [0.018] arbitrary units and 108.1% [11.65%]) groups ( p G .001 and p = .002, respectively). Inhibition of cyclooxygenase (COX) activity revealed increased COX-2Ydependent relaxation in the anhedonic group. In contrast, endothelial NO synthaseY and COX-independent relaxation to acetylcholine (endothelium-dependent hyperpolarizationYlike response) was reduced in anhedonic rats ( p G .001). This was associated with decreased transcription of intermediate-conductance Ca2+-activated K+ channels. Conclusions: Our findings demonstrate that depression-like symptoms are associated with reduced endothelium-dependent relaxation due to suppressed endothelium-dependent hyperpolarizationYlike relaxation despite up-regulation of the NO and COX-2Ydependent pathways in rat mesenteric arteries. These changes could affect peripheral resistance and organ perfusion in major depression. Key words: major depression, chronic mild stress, cyclooxygenase, endothelium-dependent hyperpolarization, nitric oxide, resistance arteries.

ACh = acetylcholine; ANOVA = analysis of variance; CMS = chronic mild stress; COX = cyclooxygenase; EDH = endothelium-dependent hyperpolarization; EGTA = ethylene glycol tetraacetic acid; eNOS = endothelial nitric oxide synthase; IKCa = intermediate-conductance Ca2+-activated K+ channels; L-NAME = nitroarginine methyl ester; NE = norepinephrine; NO = nitric oxide; SI = sucrose index; SKCa = small-conductance Ca2+-activated K+ channels; TPR = total peripheral resistance.

INTRODUCTION ajor depression and cardiovascular diseases (CVDs) are two of the most prevalent health problems in Western society, and an association between them is generally accepted (1,2). Various mechanisms for this association have been proposed. Interestingly, endothelial dysfunctionVa factor involved in most CVDs (3)Vhas also been associated with major depression (2,4Y6). Endothelial dysfunction is a complex term describing changes in several endothelium-dependent pathways in blood vessels of different caliber and localization. The changes in endothelial control of vascular tone in small resistance arteries, which are chiefly responsible for total peripheral resistance (TPR), could be critical for tissue perfusion and blood pressure. It is accepted that small resistance artery structure and function are

M

From the Department of Clinical Medicine (E.V.B., R.N., O.W.), Aarhus University Hospital, Aarhus, Denmark; Department of Biomedicine (D.M.B.B., A.M.J.M., O.K., C.A.,V.V.M), MEMBRANES, Aarhus University, Aarhus, Denmark; and Department of General Physiology (I.A.R.), St Petersburg State University, St Petersburg, Russia. Address correspondence and reprint requests to Vladimir V. Matchkov, DMSc, Department of Biomedicine, Aarhus University, Ole Worms Alle bygn. 4, 1163, Aarhus C 8000, Denmark. E-mail: [email protected] Received for publication September 10, 2013; revision received January 29, 2014. DOI: 10.1097/PSY.0000000000000062 268

the important contributors to the development of CVD, and it is therefore of interest to determine the level of endothelial dysfunction in small arteries from individuals experiencing major depression. Three major endothelial-derived relaxing pathways modulate vessel diameter: nitric oxide (NO); vasorelaxant eicosanoids, such as prostacyclin; and endothelium-dependent hyperpolarization (EDH) (3). The contribution of NO and endotheliumdependent relaxation decreases with vessel diameter, whereas EDH exhibits an inverse pattern contributing to a small extent in large arteries but as a major dilatator in small arteries. The molecular pathways responsible for endothelial NO synthase (eNOS)Yand cyclooxygenase (COX)Yindependent relaxation as well as EDH are not fully understood, but they generally share a feature in that they are inhibited by blocking intermediate- and small-conductance Ca2+-activated K+ channels (IKCa and SKCa, respectively) (7). Dysfunction of the endothelium-dependent relaxing pathways is important in pathological conditions such as diabetes mellitus, arterial hypertension, and atherosclerosis (3). It was recently reported that patients with major depression have endothelial dysfunction (8Y11). The mechanisms linking major depression and endothelial dysfunction are not known, but dysregulation of the autonomic nervous system and hypothalamic-pituitaryadrenal axis, inflammation, and oxidative stress could be important risk factors (4). Animal models are a useful approach to study depressionassociated cardiovascular abnormalities. In the chronic mild stress (CMS) model, rodents develop symptoms similar to depression (12,13), such as impaired physical activity and an inability to experience pleasure from normally pleasurable stimuli (anhedonia) measured as a reduction in the consumption of palatable sucrose solution (14,15). The CMS model has Psychosomatic Medicine 76:268Y276 (2014)

0033-3174/14/7604Y0268 Copyright * 2014 by the American Psychosomatic Society

Copyright © 2014 by the American Psychosomatic Society. Unauthorized reproduction of this article is prohibited.

DEPRESSION AND ENDOTHELIAL FUNCTION a high validity and is considered a realistic model to produce depression-like symptoms (13,16,17). Rats exposed to CMS demonstrate a significant elevation of corticosteroids and activation of the hypothalamic-pituitary-adrenal axis (18Y20), as do approximately 50% of patients with major depression (21). In the rat CMS model of depression, anhedonia-like symptoms (reduction in sucrose intake) are detected in approximately 50% of stressed rats. Approximately 30% of CMS rats reduce their sucrose intake in less than 10% and are assigned to the stressresilient group (15,18,22Y24). The remaining 20% of stressed rats do not demonstrate consistent sucrose consumption and are not usually considered in the studies. Despite the significantly reduced cardiac output, blood pressure in CMS rats is unchanged compared with unstressed animals (14,25,26), suggesting that CMS rats have an elevated TPR. We have recently shown that these changes are specific to anhedonic rats (27). The major determinants of TPR are small resistance arteries and arterioles. Elevation of TPR could be caused by structural remodeling of small arteries (28) and/or be caused by functional changes in vascular reactivity leading to increased arterial tone. Factors leading to increased tone could be either increased contractility or decreased endothelium-dependent relaxation of small arteries, or a combination of both. Endothelium function in the CMS model has previously been studied only in large conduit arteries and the aorta (29Y32). The endothelial function of the large vessels differs significantly from that of small arteries (3). Thus, the role of small artery endothelium function in elevating TPR in CMS requires investigation. There are several reports on endothelial function in the CMS model (29Y32), but none of these examined or grouped the animals according to their stress response. In the present study, we further the understanding of endothelium dysfunction in relation to depression-like symptoms by investigating the three major pathways of endothelium-derived relaxation in rat mesenteric small resistance arteries from CMS anhedonic and resilient rats. METHODS Experiments were conducted in accordance with the national guidelines for animal research and after permission from the Animal Experiments Inspectorate of the Danish Ministry of Justice. Subjects were 6-week-old, male Wistar rats purchased from Taconic, Denmark. Their weight was approximately 200 grams when training for sucrose consumption was initiated and approximately 350 grams at the start of the stress regime. The animals were housed individually, except when grouping was applied as a stressor. Food and water were available ad libitum, except when food and/or water deprivation was applied as a stressor. The standard 12-hour light/ dark cycle (light phase 0600Y1800 hours) was only changed in the course of the stress regime. After a 2-week acclimatization to the housing facility, subjects were trained to drink a palatable sucrose solution during the following 2 weeks and then exposed to the sucrose consumption test (SCT) semiweekly. The SCT consisted of a 1-hour exposure to a bottle containing sucrose solution, as described below. Subsequently, animals were exposed to the sucrose solution for 1 hour once a week until the end of the experiment. The CMS regimen was initiated with a group of 200 rats intended for different studies. Subgroups of anhedonic and resilient rats were randomly allocated to the various investigations (33,34). Two subgroups were assigned to the studies reported here, together with an unchallenged control group (control,

n = 12; anhedonic, n = 12; resilient, n = 12). Stability in food and water consumption was controlled during 8 weeks of CMS in randomly chosen subjects. No change was observed between groups.

Sucrose Consumption Test In the SCT, the sucrose intake of individual rats is measured after a 1-hour exposure to a bottle with 1.5% sucrose solution. To achieve a comparable level of thirst and hunger before the SCT, animals were deprived of food and water for 14 hours before the test. Baseline sucrose intake was calculated as an average intake from three consecutive SCTs applied before starting the CMS protocol. For each SCT during the 8 weeks of the experiment, the sucrose index (SI) was calculated as a percentage of the current intake value relative to the baseline intake. Averaged SI was then calculated and used as a parameter to evaluate the individual hedonic status (19).

CMS Protocol The CMS protocol has previously been described in detail (27). In short, the protocol consisted of a 14-day cycle with one period of intermittent illumination, stroboscopic light, grouping, food or water deprivation; two periods of soiled cage and no stress; and three periods of 45- box tilting. During grouping, rats were housed in pairs with different partners alternately serving as resident or intruder. All stressors lasted 10 to 14 hours and continued for 8 weeks.

Isometric Force Measurement After 8 weeks of CMS, the rats (È18 weeks old) were anesthetized with isoflurane and decapitated. Second- and third-order branches of the mesenteric artery were dissected in ice-cold physiological salt solution (in millimolars): NaCl, 119; KCl, 4.7; KH2PO4, 1.18; MgSO4, 1.17; NaHCO3, 25; CaCl2, 1.6; EDTA, 0.026; and glucose, 5.5, gassed with 5% CO2 in air and adjusted to pH 7.4. The arterial segments were mounted in a myograph for isometric force measurements (Danish Myo Technology A/S, Aarhus, Denmark), and the artery circumference was set to a value where maximal active force is obtained (35). The myograph chamber was heated to 37-C, whereas the physiological salt solution was constantly aerated with 5% CO2 in air. Force (in millinewtons) was recorded with a PowerLab and Chart5 acquisition system (ADInstruments Ltd, Dunedin, New Zealand) and converted to wall tension (in newtons per meter) by dividing the force with double the segment length. The contractile responses were normalized to the contraction with 30 KM norepinephrine (NE) (stipulated as 100% contraction). Endothelium function was tested using concentration-response relationships to acetylcholine (ACh). Arteries were first constricted with 6 KM NA, and ACh was added upon the NE-induced contraction in a cumulative manner. Relaxation was expressed in percentage from the preconstricted level (0% of relaxation) to passive wall tension (100% of relaxation). Three parameters were calculated for each rat: sensitivity to ACh (j log EC50), maximal relaxation (% of relaxation observed at maximal ACh concentration), and effect of inhibitors (a difference in areas under concentration-response curves before and after administration of the drug).

Semiquantitative Western Blot Pieces of mesenteric small arteries were lysed in 25 Kl lysis buffer (in millimolars: Tris-HCl, 10; sucrose, 250; EDTA, 1; EGTA, 1; Triton X-100, 2% [pH 7.4]; and 1 tablet protease inhibitor per 10 ml), and protein was isolated as described previously (27). Protein (10 Kg) was loaded on a gel for protein separation. The membranes with electrotransferred proteins were incubated with the primary antibody overnight at 4-C and then with horseradish peroxidaseYconjugated secondary antibody (1:4000; Catalog No. P0448; Dako, Glostrup, Denmark) for 1 hour. Bound antibody was detected by an enhanced chemiluminiscence kit (Amersham, Buckinghamshire, UK). Densitometry analyses were performed using ImageJ software (Rasband; National Institutes of Health, Bethesda, MD). To control for uneven protein loading, either pan-actin expression (semiquantification of eNOS) or A-actin expression (semiquantification of COX-1 and COX-2) was determined, and the corrected band densities were reported after normalization to the average level of nonstressed control. The following primary antibodies were used: pan-actin antibody (1:1000; Catalog No. 4968; Cell Signaling Technology, Danvers, MA), A-actin (1:1000; Catalog No. 3597-100; BioVision, Milpitas, CA), eNOS antibody (1:1000; Catalog

Psychosomatic Medicine 76:268Y276 (2014)

Copyright © 2014 by the American Psychosomatic Society. Unauthorized reproduction of this article is prohibited.

269

E. V. BOUZINOVA et al. No. ab5589; Abcam, Cambridge, UK), COX-1 antibody (1:500; Catalog No.160109; Cayman Chemical, Ann Arbor, MI), and COX-2 antibody (1:500; Catalog No.160126; Cayman Chemical). Phosphorylation of eNOS at Ser-1177 was investigated with a phosphospecific antibody (1:500; Catalog No. Ser1177; Cell Signaling Technology).

Quantitative Polymerase Chain Reaction Immediately after isolation, mesenteric small arteries were stored in RNA later. The RNA isolation was carried out with Qiagen mini kit (VWR, Harlev, Denmark). Complementary DNA was synthesized using Superscript reverse transcriptase (Invitrogen, Taastrup, Denmark) and superase (Ambion Ltd, England, UK) was used for deactivation of RNAse and DNAse. Quantitative polymerase chain reaction (qPCR) was carried out on MX3000P (Stratagene, La Jolla, CA) using Taqman probe (FAM) technology. Gene expression was normalized to GAPDH and transferrin receptor (average Ct value) and presented by a $Ct value. Comparison of gene expression was derived from subtraction of averaged nonstressed control $Ct from other $Ct value, producing $$Ct. Relative gene expression was calculated as 1/(2$$Ct) and thereby standardized to nonstressed control arteries. Primer sets for qPCR analyses of IKCa and SKCa expression and primers for housekeeping GAPDH and transferrin receptor were purchased from Applied Biosystems (Aerum, Denmark). IKCa (target geneVAF156554): sense 5¶-TTGCAAGTGGGTGCTGTACCTGC3¶; antisense 5¶-TGTCAGTCATGAACAGCTGGAC-3¶; probe 5¶-TGTCTTATTGTG GTCTTCCATGCCAAGGAGGTCC-3¶ SKCa (target geneVAF292389): sense 5¶-TCCGGGAAACATGGCTGA TCTA-3¶; antisense 5¶-CCTTTGTTCCATCTTGACACCCCTC-3¶; probe 5¶CCACGCCAAAGTCAGGAAACACCAGAGGAA-3¶ GAPDH (target geneVNM_017008.3): sense 5¶-CACGGCAAGTTCA ACGGCACAG-3¶; antisense 5¶-AGACTCCACGACATACTCAGCACC-3¶; probe 5¶-AGCTGGTCATCAACGGGAAACCCATCACCA-3¶ Transferrin receptor (target geneVNM_022712.1): sense 5¶-AGACAGTA ATTGGATTAGCAAAATTG-3¶; antisense 5¶-GGGTCAACTCAATGTCAT GGGTA-3¶; probe 5¶-CAGCAGAGGTGGCCGGTCAGTTCATTATTAAA-3¶

Data Analysis Results are presented as means (standard error of the mean) for all analyses, tables, and figures. A value of p G .05 was considered to be statistically significant. SCT results were analyzed by repeated-measures analysis of variance (ANOVA) with group as the independent group factor and SI as the repeatedmeasures factor. Bonferonni post hoc test was used for multiple comparisons. Differences in NE and ACh concentration-response curves were analyzed by two-way ANOVA, with group as the independent group factor and agonist as the repeated-measures factor. Differences in the effects of inhibitors were analyzed by one-way ANOVA. Bonferonni post hoc test was used for multiple comparisons in both cases. The effects of sizes calculated as Cohen d values were large for 69% of all pairwise comparisons.

RESULTS Sucrose Consumption Repeated-measures ANOVA of the SCT results indicated significant differences in between-subject variation (F(2,31) = 14.36, p G .001) and within-subject variation (F(6,186) = 6.54, p G .001). The resilient group did not differ in SI from the nonstressed group during the CMS exposure, whereas the anhedonic group had a significantly reduced SI after 1 week of CMS in comparison with the nonstressed (t = 3.036, p = .003) and resilient groups (t = 3.035, p = .003; Fig. 1). ACh-Induced NO-Dependent Relaxation Mesenteric small arteries isolated from nonstressed, resilient, and anhedonic rats had similar contractile responses to NE under control conditions (F(2,12) = 0.17, p = .85; Fig. 2A). 270

Figure 1. The time course for the sucrose index for two groups of CMS-exposed rats and the nonstressed control group. *p G .05 and **p G .01 in anhedonic group (n = 11) versus nonstressed group (n = 12). +p G .05, ++ p G .01, and +++p G .001 in anhedonic group versus resilient group (n = 11). Data are presented as mean (SEM). CMS = chronic mild stress; SEM = standard error of the mean.

The ANOVA analysis of ACh-induced relaxation did not yield a main effect of the groups (F(2,31) = 1.928, p = .16) nor a group-by-concentration interaction (F(16,248) = 1.266, p = .22; Fig. 2B). The maximal relaxation was similar between groups, but arteries isolated from anhedonic rats relaxed significantly less at an intermittent concentration of ACh (300 nM) in comparison with resilient groups (t = 3.226, p = .001). The sensitivity to ACh was significantly reduced in arteries from the anhedonic rats compared with the other groups (F(2,33) = 3.954, p = .029; Table 1), indicating a change in endothelial function in these animals. Pretreatment of arteries with the nonselective NO synthase inhibitor, nitroarginine methyl ester (L-NAME; 100 KM), suppressed ACh-induced relaxation as expected (Fig. 2C). Inhibiting NOS with L-NAME revealed a different significance in the NO component of the endothelium-dependent relaxation between the different experimental groups: ANOVA analysis of the ACh-induced relaxation in the presence of 100 KM L-NAME yielded a main effect of the groups (F(2,31) = 3.381, p = .047) and a group-by-concentration interaction (F(16,248) = 4.255, p G .001). Arteries isolated from anhedonic rats had reduced maximal relaxation in the presence of L-NAME compared with arteries both from the nonstressed (t = 3.611; p G .001) and resilient groups (t = 3.685, p G .001; Fig. 2C). In the presence of L-NAME, the arterial sensitivity to ACh did not differ between the groups (F(2,33) = 0.409, p = .67; Table 1). The ANOVA comparison of the NO-dependent component in ACh-induced relaxation (Fig. 3A) yielded a main effect of group (F(2,33) = 4.474, p = .020). Mesenteric small arteries isolated from anhedonic rats demonstrate stronger NO-dependent component of relaxation than arteries from nonstressed rats (t = 2.824; p = .022), but changes between anhedonic and resilient groups did not reach the applied level of significance (t = 2.290, p = .07). eNOS Expression We tested whether the difference in the NO-dependent component of ACh-induced vasorelaxation was due to changes Psychosomatic Medicine 76:268Y276 (2014)

Copyright © 2014 by the American Psychosomatic Society. Unauthorized reproduction of this article is prohibited.

DEPRESSION AND ENDOTHELIAL FUNCTION

Figure 2. Concentration-response curves under specific physiological conditions: contractile responses to increased concentrations of norepinephrine (A) and relaxations induced by increased concentrations of acetylcholine (BYF). A (arrow), Contractile response to 6 KM norepinephrine used for preconstriction of arteries in the acetylcholine relaxation experiments. B, The acetylcholine concentration-relaxation curves under control conditions. C, 100 KM of L-NAME used to inhibit NO synthase. D, 100 KM L-NAME + 1 KM SC560 used to block NO synthase and COX-1. E, 100 KM L-NAME + 1 KM SC560 + 10 KM NS398 used to block NO synthase, COX-1, and COX-2. F, 100 KM L-NAME + 1 KM SC560 + 10 KM NS398 + 1 KM TRAM-34 + 50 nM apamin used to block all endothelium-dependent relaxation including the EDH-like component. **p G .01 and ***p G .001 in anhedonic (n = 6Y11) versus nonstressed group (n = 7Y12). +p G .05, ++p G .01, and +++p G .001 in anhedonic versus resilient group (n = 7-12). Data are presented as mean (SEM). L-NAME = nitroarginine methyl ester; NO = nitric oxide; EDH = endothelium-dependent hyperpolarization; COX = cyclooxygenase; SEM = standard error of the mean. TABLE 1. Sensitivities to Acetylcholine (Expressed as jlogEC50) of Mesenteric Small Arteries Isolated From Nonstressed, Resilient, and Anhedonic Rats Control

+L-Name

+L-Name + SC560

+L-Name + SC560 + NS398

Nonstressed

6.81 (0.07) (n = 12)

6.11 (0.08; n = 12)

6.18 (0.12; n = 7)

6.14 (0.12; n = 6)

Resilient

6.90 (0.04; n = 11)

6.00 (0.09; n = 11)

6.04 (0.15; n = 7)

6.01 (0.14; n = 6)

Anhedonic

6.64 (0.08; n = 11)*

6.04 (0.10; n = 11)

6.13 (0.10; n = 6)

6.11 (0.32; n = 6)

The jlogEC50 was calculated under control conditions and after treatment with 100 KM L-NAME, 1 KM SC560, and 10 KM NS398. The addition of 1 KM TRAM34 and 50 nM apamin in the presence of eNOS and COX blockers completely abolished relaxation (Fig. 2F); thus, a determination of the jlogEC50 was not possible for this treatment. Data are presented as mean (SEM). * p G .05 versus resilient. Psychosomatic Medicine 76:268Y276 (2014)

Copyright © 2014 by the American Psychosomatic Society. Unauthorized reproduction of this article is prohibited.

271

E. V. BOUZINOVA et al.

Figure 3. The average contributions for the different components of the acetylcholine-stimulated relaxation. Each pathway was quantified as a difference in areas under concentration-response curves before and after administration of corresponding drug. The NO-dependent component was isolated by 100 KM L-NAME inhibition of NOS activity (A), COX-1Ydependent component was isolated by 1 KM SC560 inhibition of COX-1 activity (B), COX-2Ydependent component was isolated by 10 KM NS398 inhibition of COX-2 activity (C), and EDH-like component was isolated by the addition of 1 KM TRAM-34 and 50 nM apamin (D). * p G .05, ** p G .01, and ***p G .001 for anhedonic versus nonstressed group. +p G .05 and ++p G .01 for anhedonic versus resilient group (n = 6Y7). Data represent means (in arbitrary units; SEM). NO = nitric oxide; L-NAME = nitroarginine methyl ester; NOS = nitric oxide synthase; COX = cyclooxygenase; EDH = endothelium-dependent hyperpolarization; SEM = standard error of the mean.

in the expression of eNOS. In keeping with our observations of vasorelaxation, arteries from anhedonic rats expressed higher levels of total eNOS (Fig. 4A, B), with a main statistical effect of group (F(2,12) = 10.47, p = .002). The expression of eNOS was significantly higher in arteries from the anhedonic group in comparison with the nonstressed (t = 4.032; p = .002) and resilient groups (t = 3.675, p = .003). The phosphorylation state of eNOS modulates the enzyme’s activity: when phosphorylated, there is enhanced NO generation. Consistent with the increased levels of total eNOS expression, the expression of the phosphorylated eNOA (p-eNOS) form (Fig. 4A, C) was increased in arteries from anhedonic rats with main effect of group (F(2,10) = 11.20, p = .003). The expression of p-eNOS was significantly higher in arteries from the anhedonic group in comparison with the nonstressed (t = 3.864; p = .003) and resilient (t = 4.136; p = .002) groups. Selective COX-1 Inhibition Two COX enzymes, COX-1 and COX-2, have been shown to produce vasoactive prostanoids in the vascular wall. Incubation with a specific COX-1 inhibitor SC560 (1 KM), in the presence of 100 KM L-NAME, potentiated ACh-induced relaxation (Fig. 2D) but was without an effect on arterial sensitivity to ACh (Table 1). There was no significant difference in the ACh concentration-response curves after NOS and COX-1 inhibition between arteries from nonstressed, resilient, and anhedonic rats: a two-way ANOVA did not yield a main effect of group (F(2,17) = 0.605, p= .56). The ANOVA for the COX-1Ydependent component of ACh-induced relaxation (Fig. 3B) also failed to detect a main effect of the groups (F(2,17) = 0.348, p = .71). Selective COX-2 Inhibition Inhibition of COX-2 with a specific inhibitor NS398 (10 KM), in the presence of 100 KM L-NAME and 1 KM SC560, significantly 272

suppressed the ACh-induced relaxation in anhedonic rat arteries only (Fig. 2E). A two-way ANOVA for the ACh concentrationresponses in the presence of L-NAME, SC560, and NS398

Figure 4. A, Relative expression of eNOS and p-eNOS proteins quantified by Western blot and normalized to a housekeeping protein A-actin. The relative normalized expression (mean [SEM]) for eNOS (B) and p-eNOS (C) in arteries isolated from nonstressed (n = 3Y5), resilient (n = 6), and anhedonic (n = 5Y6) rats. **p G .01. Data are presented as mean (SEM). eNOS = endothelial nitric oxide synthase; p-eNOS = phosphorylated eNOA; SEM = standard error of the mean. Psychosomatic Medicine 76:268Y276 (2014)

Copyright © 2014 by the American Psychosomatic Society. Unauthorized reproduction of this article is prohibited.

DEPRESSION AND ENDOTHELIAL FUNCTION (50 nM apamin and 1 KM TRAM-34) were applied simultaneously (Fig. 2F). Similarly, a complete inhibition of the relaxation was achieved by a combination of 100 KM L-NAME, 3 KM indomethacin, 50 nM apamin, and 1 KM TRAM-34 in each experimental group (data not shown). Analysis of the EDH-like component of relaxation (Fig. 3D) revealed a main effect of rat groups (F(2,17) = 12.580, p G .001), where arteries from anhedonic rats demonstrated less pronounced EDH-like relaxation than arteries from both nonstressed (t = 4.802, p G .001) and resilient (t = 3.881, p = .002) rats. Expression of IKCa and SKCa Channel Messenger RNA The possible role of IKCa and SKCa channels in the functional endothelial changes was investigated using qPCR. A decreased level of IKCa messenger RNA (mRNA) expression was found in arteries from anhedonic rats (Fig. 6A): the ANOVA for relative expression yielded a main effect of groups (F(2,9) = 9.565, p = .006), with differences between the anhedonic and nonstressed (t = 3.148; p = .012) groups as well as the resilient (t = 4.057; p = .003) groups. Analysis of qPCR results did not reveal differences between groups in mRNA expression of SKCa channel (F(2,8) = 0.175, p = .84; Fig. 6B). Figure 5. A, Relative expression for COX-1 and COX-2 proteins by Western blot. The total protein load was normalized to a housekeeping protein A-actin. An averaged relative expression of COX-1 (B; n = 10Y11) and COX-2 (C; n = 7Y8) in the mesenteric arteries from nonstressed rats was set to 100%. Data are presented as mean (SEM). COX = cyclooxygenase; SEM = standard error of the mean.

yielded a main effect of group (F(2,14) = 8.841, p = .003) and a group-by-concentration interaction (F(16,112) = 3.434, p G .0001). Arteries from anhedonic rats had significantly reduced maximal relaxation in the presence of L-NAME, SC560, and NS398 in comparison with the nonstressed (t = 4.390, p G .001) and resilient groups (t = 4.116, p G .001; Fig. 2E). ACh sensitivity was not different between groups in the presence of L-NAME, SC560, and NS398 (Table 1). The main effect of COX-2 on ACh-induced relaxation (Fig. 3C) was confirmed by one-way ANOVA (F(2,17) = 4.750, p = .025). The COX-2 component of relaxation was larger in arteries from anhedonic rats compared with the nonstressed (t = 2.727; p = .016) and resilient groups (t = 2.607; p = .020; Fig. 3E). Expression of COX-1 and COX-2 We tested whether the observed effects of COX inhibition on arterial relaxation were associated with the expression of COX enzymes (Fig. 5). Semiquantitative analysis of protein expression did not reveal any differences in the expression of either COX-1 (F(2,28) = 0.307, p = .74; Fig. 5A, B) or COX-2 (F(2,20) = 2.085, p = .15; Fig. 5A, C) proteins in mesenteric arteries from the three groups of rats. EDH-Like Component of Relaxation ACh-dependent relaxation was blocked when the inhibitors of NO-dependent relaxation (100 KM L-NAME), COX signaling (1 KM SC560 and 10 KM NS398), and EDH-like signal

DISCUSSION The major finding of this study is that mesenteric small arteries from anhedonic rats are less sensitive to the vasodilator action of ACh than arteries from nonstressed and resilient rats. This seems to be caused by differences in the relative contribution of all three major endothelium-dependent relaxing pathways (NO, COX product(s), and EDH) in resistance arteries from rats susceptible to stress. Reduction in the EDH component of relaxation in anhedonic rats cannot be completely compensated by elevation of NO- and prostacyclin-dependent pathways. Such a decrease in the endothelial-derived relaxing components could affect vascular resistance, thereby provoking hypertension and local ischemia

Figure 6. Relative expression of IKCa (A) and SKCa (B) at the mRNA levels quantified by qPCR. Expression was normalized to an averaged level in arteries from nonstressed rats taken as 100%. *p G .05 versus nonstressed groups. +++p G .001 versus resilient groups (n = 5Y12). Data are presented as mean (SEM). IKCa and SKCa = intermediate- and small-conductance Ca2+-activated K+ channels; mRNA = messenger RNA; qPCR = quantitative polymerase chain reaction; SEM = standard error of the mean.

Psychosomatic Medicine 76:268Y276 (2014)

Copyright © 2014 by the American Psychosomatic Society. Unauthorized reproduction of this article is prohibited.

273

E. V. BOUZINOVA et al. (2,5). This is in line with previous studies suggesting that CMS is associated with TPR elevation (14,25,26). Indeed, we have previously shown that increased TPR is associated with anhedonia but not the stress resilience (27). CMS rats were divided into resilient and anhedonic groups. This allowed us to distinguish the vascular effects associated with stress in general and with stress-induced depression in particular. The finding that CMS followed by anhedonia was associated with differences in endothelial function compared with the resilient rats would indicate that these differences in endothelial function either predispose to anhedonia or develop as a consequence of anhedonia. The nature of the signals responsible for the association between the changes in endothelial function and the susceptibility to CMS is unclear, and numerous factors could play a role. Depression-associated vascular and organ inflammation, associated oxidative stress, and chronic glucocorticoid elevation are well-known factors affecting the vascular endothelium (3,4,7,36). Chronic exposure to elevated cortisol/corticosterone, that is, specific for stress-induced depression (20), could affect the expression and activity of eNOS, COX, and endothelial K+ channels (4,37). Moreover, depression is associated with low-grade inflammation, which, in turn, also affects the endothelial function (38). Glucocorticoids and proinflammatory agents are known causes for overproduction of reactive oxygen species, which would further modify the expression and activity of endothelial enzymes and channels (37). Our results indicate an increased NO-dependent component of ACh-induced relaxation in anhedonic rats. An elevated expression of eNOS, specifically the active phosphorylated form (p-eNOS), in the wall of small arteries could be responsible for this increase (Fig. 4). In humans, endothelial dysfunction (2,3) was reported in patients with major depression, where impaired forearm flow-mediated vasodilatation was seen (8Y10,33,39,40). In addition, reduced NO-dependent relaxation has been shown in studies of large arteries from rats (30,31) and mice (29,32) exposed to CMS. This inconsistency might be explained by the difference in vessel size (aorta versus resistance arteries). Moreover, the attenuated endothelial relaxation of preconstricted subcutaneous small arteries from elderly patients with depression (41) cannot entirely be explained by reduced NO availability (42). Additional pathogenic mechanisms could be involved. Importantly, the significance of NO-dependent relaxation declines with the reduction of arterial size and plays only a minor role in small arteries (34). Thus, it is likely that endothelial dysfunction is differentially regulated in small and large arteries. The observed increase in endothelium-dependent NO signaling in anhedonia could develop consequently to suppression of the EDH-like response, which is the major relaxing pathway in resistant arteries (43). A counterbalancing interaction between EDH and NO signaling has previously been shown in the aorta from CMS mice (29). The mechanism responsible for the differences in NOdependent relaxation in CMS has not previously been addressed (29Y32). Here we find that the elevation of NO-dependent relaxation in the anhedonic group can be explained by increased 274

expression of eNOS. Consistently, chronic social stress by crowding induces elevated eNOS activity in the rat aorta (44Y46). The highly divergent synthesis pathways for the different eicosanoids suggest a high degree of flexibility for vascular responses to COX activation and inhibition (47). This is further complicated by a spectrum of eicosanoid receptors with limited ligand specificity (48). Thus, changes in both the enzymatic pathways and in receptor expression could lead to a shift between contractile and relaxing responses to COX activation (36). Selective inhibitors of COX-1 and COX-2 revealed a difference in the COX-1/COX-2 balance between anhedonic and resilient rats. No differences in the functional significance and protein expression of COX-1 were observed in the current study, suggesting that the COX-1Ydependent pathway was unaffected by CMS. The increased COX-2Ydependent relaxation in arteries from anhedonic rats cannot be explained by COX2 overexpression. The different functional effects of COX-2 could be due to up-regulation of downstream enzymatic pathways or due to up-regulation/sensitization of prostacyclin receptors (36). It will be important to identify the molecular background for the increased COX-2Ydependent relaxation in arteries from anhedonic rats. A combination of blockers for IKCa and SKCa channels inhibited the eNOS- and COX-independent relaxation, that is, EDH-like relaxation, which is known to play a significant role in relaxation of small arteries (34,49). EDH-like relaxation was attenuated in arteries from anhedonic rats. In contrast, in the aorta from CMS mice, approximately one-third of the AChinduced relaxation was via EDH, whereas no EDH response was seen in the aorta from nonstressed mice (29). This difference may reflect species differences or differences between large and small arteries as discussed above. The significant contribution of EDH-like relaxation to the tone of small arteries suggests that the EDH-like relaxation is important for the control of blood pressure and tissue perfusion. This notion is supported by the observation that the EDH responses are diminished in several cardiovascular pathologies, including some forms of hypertension and diabetes (7). The specific role of IKCa and SKCa varies in different conditions and different experimental models (3,7). We found that mRNA of the IKCa channels was significantly reduced in the wall of arteries from anhedonic rats and suggest that reduced EDH-like relaxation is due to reduced expression of IKCa channels. Similar pathological changes were previously seen in mesenteric small arteries from Type 2 diabetic Zucker diabetic fatty rats (43), where reduced EDH response was also mainly due to reduced expression of IKCa channels. The detailed analysis of endothelial function in small arteries, presented in the current study, demonstrates the complex association between depression symptoms and endothelial dysfunction. We found that the anhedonia-associated suppression of EDH-like relaxation in small resistance arteries is not fully compensated by increased relaxation via eNOS- and COX-2Ydependent pathways. Therefore, it is important in future studies to elucidate the molecular pathways responsible Psychosomatic Medicine 76:268Y276 (2014)

Copyright © 2014 by the American Psychosomatic Society. Unauthorized reproduction of this article is prohibited.

DEPRESSION AND ENDOTHELIAL FUNCTION for the depression-associated endothelial dysfunction, including the effects of antidepressant treatment. We thank Kim Henningsen and Stine Dhiin Heide Hansen for running the CMS model and Jane Holb&k RLnn, Viola Smed Larsen, and Joergen Andreasen for excellent technical assistance during the functional studies. Source of Funding and Conflicts of Interest: The study was supported by the Lundbeck Foundation (R118-A11567). V. Matchkov is currently receiving a grant from the Lundbeck Foundation (No. R118-A11567). For the remaining authors, no conflicts of interest were declared.

22.

23.

24.

25. 26.

REFERENCES 1. Grippo AJ. Mechanisms underlying altered mood and cardiovascular dysfunction: the value of neurobiological and behavioral research with animal models. Neurosci Biobehav Rev 2009;33:171Y80. 2. Joynt KE, Whellan DJ, O’Connor CM. Depression and cardiovascular disease: mechanisms of interaction. Biol Psychiatry 2003;54:248Y61. 3. Vanhoutte PM, Shimokawa H, Tang EH, Feletou M. Endothelial dysfunction and vascular disease. Acta Physiol (Oxf) 2009;196:193Y222. 4. Toda N, Nakanishi-Toda M. How mental stress affects endothelial function. Pflugers Arch 2011;462:779Y94. 5. Johnson AK, Grippo AJ. Sadness and broken hearts: neurohumoral mechanisms and co-morbidity of ischemic heart disease and psychological depression. J Physiol Pharmacol 2006;57(Suppl 11):5Y29. 6. Lett HS, Blumenthal JA, Babyak MA, Sherwood A, Strauman T, Robins C, Newman MF. Depression as a risk factor for coronary artery disease: evidence, mechanisms, and treatment. Psychosom Med 2004;66:305Y15. 7. Grgic I, Kaistha BP, Hoyer J, Kohler R. Endothelial Ca+-activated K+ channels in normal and impaired EDHF-dilator responsesVrelevance to cardiovascular pathologies and drug discovery. Br J Pharmacol 2009;157:509Y26. 8. Wagner JA, Tennen H, Mansoor GA, Abbott G. History of major depressive disorder and endothelial function in postmenopausal women. Psychosom Med 2006;68:80Y6. 9. Sherwood A, Hinderliter AL, Watkins LL, Waugh RA, Blumenthal JA. Impaired endothelial function in coronary heart disease patients with depressive symptomatology. J Am Coll Cardiol 2005;46:656Y9. 10. Broadley AJ, Korszun A, Jones CJ, Frenneaux MP. Arterial endothelial function is impaired in treated depression. Heart 2002;88:521Y3. 11. Rajagopalan S, Brook R, Rubenfire M, Pitt E, Young E, Pitt B. Abnormal brachial artery flow-mediated vasodilation in young adults with major depression. Am J Cardiol 2001;88:196Y8. 12. Katz RJ. Animal model of depression: pharmacological sensitivity of a hedonic deficit. Pharmacol Biochem Behav 1982;16:965Y8. 13. Wiborg O. Chronic mild stress for modeling anhedonia. Cell Tissue Res 2013;354:155Y69. 14. Grippo AJ, Moffitt JA, Johnson AK. Cardiovascular alterations and autonomic imbalance in an experimental model of depression. Am J Physiol 2002;282:R1333Y41. 15. Jayatissa MN, Bisgaard C, Tingstrom A, Papp M, Wiborg O. Hippocampal cytogenesis correlates to escitalopram-mediated recovery in a chronic mild stress rat model of depression. Neuropsychopharmacology 2006;31:2395Y404. 16. Willner P. Chronic mild stress (CMS) revisited: consistency and behaviouralneurobiological concordance in the effects of CMS. Neuropsychobiology 2005;52:90Y110. 17. Willner P. The validity of animal models of depression. Psychopharmacology (Berl) 1984;83:1Y16. 18. Grippo AJ, Francis J, Beltz TG, Felder RB, Johnson AK. Neuroendocrine and cytokine profile of chronic mild stress-induced anhedonia. Physiol Behav 2005;84:697Y706. 19. Garcia LS, Comim CM, Valvassori SS, Reus GZ, Stertz L, Kapczinski F, Gavioli EC, Quevedo J. Ketamine treatment reverses behavioral and physiological alterations induced by chronic mild stress in rats. Prog Neuropsychopharmacol Biol Psychiatry 2009;33:450Y5. 20. Christiansen S, Bouzinova EV, Palme R, Wiborg O. Circadian activity of the hypothalamic pituitary adrenal axis is differentially affected in the rat chronic mild stress model of depression. Stress 2012;15:547Y657. 21. Asnis GM, Halbreich U, Ryan ND, Rabinowicz H, Puig-Antich J, Nelson B, Novacenko H, Friedman JH. The relationship of the dexamethasone

27.

28. 29. 30.

31.

32.

33.

34.

35. 36. 37. 38.

39.

40.

41. 42.

suppression test (1 mg and 2 mg) to basal plasma cortisol levels in endogenous depression. Psychoneuroendocrinology 1987;12:295Y301. Grippo AJ, Moffitt JA, Johnson AK. Evaluation of baroreceptor reflex function in the chronic mild stress rodent model of depression. Psychosom Med 2008;70:435Y43. Henningsen K, Andreasen JT, Bouzinova EV, Jayatissa MN, Jensen MS, Redrobe JP, Wiborg O. Cognitive deficits in the rat chronic mild stress model for depression: relation to anhedonic-like responses. Behav Brain Res 2009;198:136Y41. Bergstrom A, Jayatissa MN, Mork A, Wiborg O. Stress sensitivity and resilience in the chronic mild stress rat model of depression; an in situ hybridization study. Brain Res 2008;1196:41Y52. Grippo AJ, Beltz TG, Weiss RM, Johnson AK. The effects of chronic fluoxetine treatment on chronic mild stressYinduced cardiovascular changes and anhedonia. Biol Psychiatry 2006;59:309Y16. Grippo AJ, Beltz TG, Johnson AK. Behavioral and cardiovascular changes in the chronic mild stress model of depression. Physiol Behav 2003;78: 703Y10. Bouzinova EV, Moller-Nielsen N, Boedtkjer DB, Broegger T, Wiborg O, Aalkjaer C, Matchkov VV. Chronic mild stress-induced depression-like symptoms in rats and abnormalities in catecholamine uptake in small arteries. Psychosom Med 2012;74:278Y87. Mulvany MJ, Baumbach GL, Aalkjaer C, Heagerty AM, Korsgaard N, Schiffrin EL, Heistad DD. Vascular remodeling. Hypertension 1996;28:505Y6. Isingrini E, Belzung C, Freslon JL, Machet MC, Camus V. Fluoxetine effect on aortic nitric oxide-dependent vasorelaxation in the unpredictable chronic mild stress model of depression in mice. Psychosom Med 2012;74:63Y72. Neves VJ, Moura MJ, Almeida BS, Costa R, Sanches A, Ferreira R, Tamascia ML, Romani EA, Novaes PD, Marcondes FK. Chronic stress, but not hypercaloric diet, impairs vascular function in rats. Stress 2012;15: 138Y48. Neves VJ, Moura MJ, Tamascia ML, Ferreira R, Silva NS, Costa R, Montemor PL, Narvaes EA, Bernardes CF, Novaes PD, Marcondes FK. Proatherosclerotic effects of chronic stress in male rats: altered phenylephrine sensitivity and nitric oxide synthase activity of aorta and circulating lipids. Stress 2009;12:320Y7. Isingrini E, Surget A, Belzung C, Freslon JL, Frisbee J, O’Donnell J, Camus V, d’Audiffret A. Altered aortic vascular reactivity in the unpredictable chronic mild stress model of depression in mice: UCMS causes relaxation impairment to ACh. Physiol Behav 2011;103:540Y6. Lavoie KL, Pelletier R, Arsenault A, Dupuis J, Bacon SL. Association between clinical depression and endothelial function measured by forearm hyperemic reactivity. Psychosom Med 2010;72:20Y6. Shimokawa H, Yasutake H, Fujii K, Owada MK, Nakaike R, Fukumoto Y, Takayanagi T, Nagao T, Egashira K, Fujishima M, Takeshita A. The importance of the hyperpolarizing mechanism increases as the vessel size decreases in endothelium-dependent relaxations in rat mesenteric circulation. J Cardivasc Pharmacol 1996;28:703Y11. Mulvany MJ, Halpern W. Contractile properties of small arterial resistance vessels in spontaneously hypertensive and normotensive rats. Circ Res 1977; 41:19Y26. Feletou M, Huang Y, Vanhoutte PM. Endothelium-mediated control of vascular tone: COX-1 and COX-2 products. Br J Pharmacol 2011;164:894Y912. Iuchi T, Akaike M, Mitsui T, Ohshima Y, Shintani Y, Azuma H, Matsumoto T. Glucocorticoid excess induces superoxide production in vascular endothelial cells and elicits vascular endothelial dysfunction. Circ Res 2003;92:81Y7. Leonard B, Maes M. Mechanistic explanations how cell-mediated immune activation, inflammation and oxidative and nitrosative stress pathways and their sequels and concomitants play a role in the pathophysiology of unipolar depression. Neurosci Biobehav Rev 2012;36:764Y85. Broadley AJ, Korszun A, Abdelaal E, Moskvina V, Jones CJ, Nash GB, Ray C, Deanfield J, Frenneaux MP. Inhibition of cortisol production with metyrapone prevents mental stressYinduced endothelial dysfunction and baroreflex impairment. J Am Coll Cardiol 2005;46:344Y50. Broadley AJ, Korszun A, Abdelaal E, Moskvina V, Deanfield J, Jones CJ, Frenneaux MP. Metyrapone improves endothelial dysfunction in patients with treated depression. J Am Coll Cardiol 2006;48:170Y5. Paranthaman R, Greenstein AS, Burns AS, Cruickshank JK, Heagerty AM, Jackson A, Malik RA, Scott ML, Baldwin RC. Vascular function in older adults with depressive disorder. Biol Psychiatry 2010;68:133Y9. Greenstein AS, Paranthaman R, Burns A, Jackson A, Malik RA, Baldwin RC, Heagerty AM. Cerebrovascular damage in late-life depression is associated with structural and functional abnormalities of subcutaneous small arteries. Hypertension 2010;56:734Y40.

Psychosomatic Medicine 76:268Y276 (2014)

Copyright © 2014 by the American Psychosomatic Society. Unauthorized reproduction of this article is prohibited.

275

E. V. BOUZINOVA et al. 43. Brondum E, Kold-Petersen H, Simonsen U, Aalkjaer C. NS309 restores EDHF-type relaxation in mesenteric small arteries from Type 2 diabetic ZDF rats. Br J Pharmacol 2010;159:154Y65. 44. Puzserova A, Csizmadiova Z, Andriantsitohaina R, Bernatova I. Vascular effects of red wine polyphenols in chronic stressYexposed Wistar-Kyoto rats. Physiol Res 2006;55:S39Y47. 45. Puzserova A, Bernatova I. Chronic social stress increases nitric oxideY dependent vasorelaxation in normotensive rats. Interdiscip Toxicol 2010;3: 109Y17.

276

46. Bernatova I, Puzserova A, Navarova J, Csizmadiova Z, Zeman M. Crowding-induced alterations in vascular system of Wistar-Kyoto rats: role of nitric oxide. Physiol Res 2007;56:667Y9. 47. Vane JR, Bakhle YS, Botting RM. Cyclooxygenases 1 and 2. Annu Rev Pharmacol Toxicol 1998;38:97Y120. 48. Tsuboi K, Sugimoto Y, Ichikawa A. Prostanoid receptor subtypes. Prostaglandins Other Lipid Mediat 2002;68-69:535Y56. 49. Busse R, Edwards G, Feletou M, Fleming I, Vanhoutte PM, Weston AH. EDHF: bringing the concepts together. Trends Pharmacol Sci 2002;23:374Y80.

Psychosomatic Medicine 76:268Y276 (2014)

Copyright © 2014 by the American Psychosomatic Society. Unauthorized reproduction of this article is prohibited.

Association between endothelial dysfunction and depression-like symptoms in chronic mild stress model of depression.

Cardiovascular diseases have high comorbidity with major depression. Endothelial dysfunction may explain the adverse cardiovascular outcome in depress...
700KB Sizes 0 Downloads 5 Views