INTRODUCTION

Interleukin-1β (IL-1β) is one of the major proinflammatorycytokines involved in the regulation of hypothalamic-pituitary-adrenal (HPA) axis and stress responses in the brain (1-5).

Cytokines can access a cytokine network in the brain andinfluence neurotransmission within stress regulatory brain circuitsand induce hormonal changes similar to those observed followingstress exposure (6). Cytokines may also markedly affectneurotransmission within regulatory brain circuits for emotion,

JOURNALOF PHYSIOLOGYAND PHARMACOLOGY 2015, 66, 3, 427-440

www.jpp.krakow.pl

A. GADEK-MICHALSKA, J. TADEUSZ, P. RACHWALSKA, J. BUGAJSKI

CHRONIC STRESS ADAPTATION OF THE NITRIC OXIDE SYNTHASES AND IL-1βLEVELS IN BRAIN STRUCTURES AND HYPOTHALAMIC-PITUITARY-ADRENAL

AXIS ACTIVITY INDUCED BY HOMOTYPIC STRESS

Department of Physiology, Institute of Pharmacology, Polish Academy of Sciences, Cracow, Poland

The aim of this study was to determine the effect of repeated restraint stress (RS) on a single restraint (homotypic) stress-induced expression of interleukin-1β (IL-1β), neuronal nitric oxide synthase (nNOS) and inducible nitric oxide synthase(iNOS) in the prefrontal cortex (PFC), hippocampus and hypothalamus and their adaptational changes in chronic stress.Also the involvement of plasma IL-1β in adrenocorticotropic hormone (ACTH) and corticosterone secretion duringchronic stress was investigated. Rats were subjected to a single restraint for 10 minutes or restrained twice a day for 3, 7and 14 consecutive days and 24 hours after the last stress period rats were restrained for 10 min and rapidly decapitated0, 1, 2 and 3 hours later. The IL-1β, nNOS and iNOS protein levels in brain structures samples were analyzed by Westernblot procedure and IL-1β, ACTH and corticosterone levels were determined in plasma. Single restraint induced a strongestdecrease of iNOS protein levels (1-3 h) in the PFC and a weaker decline in the hippocampus and hypothalamus (3 h) afterstress cessation. Single restraint markedly increased IL-1β protein level in PFC and hippocampus. In the PFC repeatedrestraint for 3 days significantly increased the homotypic stress induced iNOS and IL-1β protein levels and this increasegradually declined after 7 and 14 days of repeated restraint. Much weaker yet a parallel changes appeared with neuronalNOS level. In the hippocampus prior stress for 3, 7 and 14 days significantly increased the homotypic stress induced iNOSprotein level parallel with IL-1β level which gradually declined with prolonged period of repeated restraint. In thehippocampus a longer restraint period, 7 and 14 days markedly decreased nNOS protein level evoked by homotypic stress.In the hypothalamus prior stress for 3 days strongly enhanced the homotypic stress-induced iNOS level and repeatedstress for 7 and 14 days blunted this effect. Repeated stress increased IL-1β level in response to homotypic stress after 3days and after 14 days. The present results indicate time-related similarities in the potent alterations in IL-1β and iNOSprotein levels in brain structures. Single restraint induced a significant increase of plasma IL-1β level which was abolishedby pretreatment with IL-1 receptor antagonist (IL-1Ra). A parallel strong increase of plasma ACTH and corticosteronelevels were significantly impaired by IL-1Ra suggesting a marked involvement of stress-induced stimulation of ACTHand corticosterone by IL-1β in single restraint. In repeatedly restrained rats IL-1Ra significantly blunted plasma IL-1βlevel induced by homotypic stress. A parallel strong increase in plasma ACTH level by homotypic stress was notsubstantially altered by pretreatment with IL-1Ra in repeatedly stressed rats. Plasma a corticosterone level increased byhomotypic stress in rats restrained for 3 and 14 days was not affected by pretreatment with IL-1Ra, but after for 7 daysits level was significantly enhanced. These results suggest that repeated stress desensitizes IL-1β-induced stimulatorycomponent in a single restraint stress-induced hypothalamic-pituitary-adrenal (HPA) axis stimulation. A sensitization byhomotypic stress of corticosterone response after restraint for 7 days may depend on other stimulatory systems actingwithin adrenal glands during prolonged stress. Comparative data from the same model of rather mild psychological stressallows for the comparison of functional adaptive changes of NO synthases and IL-1β in brain structures involved in stressregulation. In general, the iNOS system is strongly sensitized by repeated stress for 3 days in prefrontal cortex andhippocampus. Increased plasma IL-1β level by a single restraint stress is significantly involved in ACTH andcorticosterone secretion. Repeated stress for 3–14 days reduces this participation of IL-1β in pituitary-adrenal stimulation.

K e y w o r d s :nitric oxide synthases, interleukin-1β, interleukin-1 receptor antagonist, stress, brain structures, hypothalamic-pituitary-adrenal axis, adaptation

mainly in the prefrontal cortex (PFC) (7). Pro-inflammatorycytokines are able to induce the release of a corticotropin-releasing hormone (CRH) and arginine vasopressin (AVP) fromthe hypothalamic paraventricular nucleus (PVN) and augmentneurochemical stimulation of the pituitary gland (8). Stressorssuch as footshock and immobilization (IMO) have been shown toinduce hypothalamic interleukin-1 (IL-1) production, while otherstressors such as restraint, maternal separation or social isolationhave no effect on hypothalamic IL-1 levels (9). Stress-increasedIL-1β levels in the circulation directly stimulates pituitary cells tosecrete adrenocorticotropic hormone (ACTH). Exposure to somepredominantly emotional and systemic stressors has been found toinduce long-term sensitization of the HPA responsiveness tofurther superimposed stressors (10). It is possible that in an intactanimal the major or initial effect of IL-1 is mediated through thehypothalamus. Chronic psychosocial stress decreased IL-1βmRNA levels in the hippocampus, the IL-1 receptor in pituitaryand increased the plasma corticosterone level (11). During stresscombined networks from brainstem nuclei to specific limbicsystem structures exercise their regulatory function on HPA axisactivity and glucocorticoid regulation. The key target of thesevarious direct and indirect pathways is the PVN of thehypothalamus (12, 13). The specific role of the hippocampus,PFC, and brainstem nuclei in regulating the HPA axis under stresscondition depends on the type of stressor (14). Different types ofstressors, such as reactive versus anticipatory stressors inducestimulation of the HPA axis through activation changes in distinctbrain regions involved in HPA regulation (2, 9). Stressful stimuli,whether physical, metabolic or psychological activate all levels ofthe HPA axis as indicated by the increased release of brain CRH,anterior pituitary ACTH and consequent increase of plasmacorticosterone or cortisol concentration.

Accumulated evidence indicates that nitric oxide (NO) iswidely distributed in the central nervous system (CNS) andsubserves the functions via its synthetic and signal transductionmechanisms. The special proteins NO receptors are coupled tocyclic guanosine monophosphate (cGMP) formation (15, 16)and cytokines modulate of NO production.

A large number of reviews on the role of NO in CNSregulatory function in neuroendocrinology of stress system havebeen published in the last years (15, 17-20). NO is synthesizedby two constitutive enzyme isoforms: neuronal nitric oxidesynthase (nNOS) and endothelial NO synthase (eNOS) byconversion of L-arginine to citrulline. The third type inducibleiNOS, although generally undetectable under basal conditions,can be constitutively expressed in different brain regions of ratsmainly by astrocytes in normal, undisturbed mammalian brainsand in microglia during immunological challenge (21). Inneurones nNOS is located mainly at the post-synaptic terminal.NO functions in the mammalian CNS like a conventionalneurotransmitter, because of its ability to freely cross a cellmembrane. NO can act in an autocrine and paracrine manneralso on distant targets (22). Brain nNOS exists in particulate andsoluble forms and the differential subcellular localization ofnNOS may contribute to its diverse functions. nNOS is involvedin a variety of synaptic events, modulating physiologicalfunctions and in a number of human diseases (23). NOparticipates in signal transduction pathways in physiologicalprocesses such as corticosterone release from the adrenal gland(24-26). NO is widely involved in interaction betweenneuroendocrine and neuroimmune systems in physiological andpathological processes. nNOS is implicated in modulatinglearning, memory and neurogenesis and is involved in a numberof human diseases, including depression. nNOS is highlyexpressed by cells of the hypothalamic PVN, the convergencepoint for the sympatho-adrenal system, the HPA axis and thehypothalamo-hypophyseal regulatory systems (27). These

systems are key regulators of the neuroendocrine stressresponses. Constitutive iNOS activity would interfere withneuronal processes that favor stress adaptation (28).

We have recently found that repeated restraint for 3 daysmarkedly altered the homotypic stress induced IL-1β level andiNOS level in brain structures and influenced HPA axis activity(29, 30). These changes were brain structure-specific anddifferent in magnitude. Despite numerous studies on theinvolvement of central IL-1β and NO systems in stress reactions,their role and functional adaptation during prolonged stress havenot yet been elucidated. The aim of the present study was toevaluate the expression of nNOS, iNOS and IL-1β in brainstructures involved in stress responses and to determineadaptational changes of IL-1β and NO-synthases induced bychronic restraint for the homotypic stress. We also investigatedthe involvement of plasma IL-1β in the functional adaptation ofpituitary-adrenocortical axis during prolonged stress.

MATERIALS AND METHODS

Animals

All the experiments were performed on unanesthetized (n =480) male Wistar rats (6 weeks old, 190 – 220 g) housed ingroups of 5 per cage (52 × 32 × 20 cm) under standardconditions with an artificial 12-h light/dark cycle (lights on from7 a.m. to 7 p.m.), constant temperature 22 ± 2°C, food and tapwater available ad libitum. The animals were given 1 week ofhabituation period before the experimental session.

All the procedures were approved by Local BioethicsCommission for Animal Experiments at the Institute ofPharmacology, Polish Academy of Sciences in Cracow and metthe requirements of the European Council Guide for the Careand Use of Laboratory Animals (86/609/EEC).

Experimental procedures

To avoid circadian variability, acute stress protocols werecarried out at light cycle between 10 – 12 a.m. All the decapitationswere performed rapidly by a skillful, always the same person,within few seconds after taking the animal from its cage. Thisprocedure did not affect the basal plasma ACTH and corticosteronelevels. After decapitation, the brains were removed from the skullsand three whole structures prefrontal cortex, hippocampus andhypothalamus were excised on an ice-cold glass plate, immediatelyfrozen on dry ice and stored at –70°C until assayed.

The animals were divided into four groups. Animals from thecontrol group were not subjected to any restraint (the first group).Rats from the second group were restrained for 10 min only once,on the day of the experiment, and decapitated 0, 1, 2 and 3 hoursafterwards. Rats from the third group were subjected to 10 minrestraint stress sessions twice a day for 3, 7 and 14 consecutivedays. The break between stress sessions was at least 8 hours. Theuse of a single time point in the present investigation might notallow for the determination of peak variations in some of the NOsynthases and IL-1β. Therefore, the changes in these componentslevels were measured during 3 hours in the post-stress period.

Restraint stress was carried out using metal tubes with ampleholes for ventilation. After stress the animals were returned tothe same cages. Some of the rats which were subjected to singlerestraint stress for 10 min, were injected i.p. with a single doseof IL-1 receptor antagonist (IL-1Ra) (100 µg/kg, Sigma-Aldrich) 15 min before the stress session and decapitated 0, 1, 2and 3 hours afterwards (the fourth group). The dose of 100 µg/kgof IL-1Ra was selected as effective antagonist of IL-1β-inducedeffect in our earlier study (31).

428

Plasma hormones and IL-1β measurement

Trunk blood for plasma determination was collected in thepresence of EDTA (10% w/v; Merck; 25 µl/ml of blood) and forACTH immunoassay in the presence of EDTA and aprotinin (0.6TIU/ml of blood; Sigma Aldrich). The total corticosterone,ACTH and IL-1β levels were measured in plasma usingcommercially available kits: the Rat/Mouse CorticoteroneEnzyme Immuno Assay (EIA) kit (Immunodiagnostic Systems),the ACTH Rat/Mouse EIAkit (Phoenix Pharmaceuticals) andthe Rat IL-1β enzyme-linked immunosorbent assay (ELISA)(BioVendor) kit.

Sodium dodecyl sulfate polyacrylamide gel electophoresis andWestern blot

Protein extracts were prepared according to Gadek-Michalska (29). Equal amounts of denaturated proteins (10 µgper lane) were separated using a 7.5% (for nNOS, iNOS) or 15%(for IL-1β) Sodium dodecyl sulfate (SDS) polyacrylamide gelelectophoresis (PAGE) (Mini PROTEAN® Tetra Cell, Bio-Rad)under constant voltage (90 V in stacking gel; 140 V in resolvinggel) and were transferred electrophoretically (90 V, 1 hour) ontoa nitrocellulose membrane (0.45 µm, Bio-Rad). All themembranes were blocked with 5% non-fat dry milk in TBS with0.05% Tween20 at room temperature. Then they were incubatedovernight at 4°C in 1% non-fat dry milk in TBS with 0.05%Tween 20 with the following primary antibodies provided bySanta Cruz Biotechnology: rabbit anti-nNOS and anti-iNOS(1:400) polyclonal antibodies and mouse anti-β-actinmonoclonal antibody (1:10 000). After incubation, themembranes were washed four times for 10 min with TBST thenfinally incubated in 1% non-fat dry milk in TBS with 0.05%Tween20 with goat anti-rabbit horseradish peroxidase-conjugated secondary antibody (1:10 000, Santa CruzBiotechnology).

For IL-1β determination, the membranes were incubated inSignal Boost Immunoreaction Enhancer Kit (Merck) instead of1% non-fat dry milk, with anti-IL-1β polyclonal primaryantibody (1:5000, AbD Serotec) and with goat anti-rabbithorseradish peroxidase-conjugated secondary antibody (1:20000, Santa Cruz Biotechnology) respectively. Afterwards, themembranes were washed four times with TBST, the proteinswere detected using ECL(Lumi-LightPlus Western Blotting Kit,Roche) and evaluated with a luminescent image analyzer(Fujifilm LAS-4000). The optical density of appropriate bandswas quantified by densitometric analysis of blots using ImageGauge V4.0 Software (Fujifilm) and normalized to β-actinlevels. All the values are expressed as a percentage of controls.

Statistical analysis

All results are expressed as the means ± standard error of themean (S.E.M.) where n = 10–12 rats per group. Statisticalanalyses were performed using GraphPad Prism 5 (GraphPadSoftware Inc.). Data were analyzed by one-way analysis ofvariance (ANOVA) followed by Tukey's multiple range test.Groups that are significantly different from the non-stressedcontrol group are indicated in the figures as: + P < 0.05, ++ P <0.01, +++ P < 0.001. For comparison of the influence of repeatedrestraint stress for 3, 7 and 14 days on homotypic stress-inducednNOS, iNOS and IL-1β levels, data were analyzed by two-wayanalysis of variance (ANOVA) followed by Bonferroni'smultiple comparison test. Groups subjected to repeated restraintthat are significantly different from the respective groupsubjected to single 10-min restraint stress and are shown as: * P< 0.05, ** P< 0.01, *** P< 0.001.

RESULTS

Repeated restraint-induced nNOS, iNOS and IL-1β responsesto homotypic stress in brain structures

In the prefrontal cortex, a single restraint for 10 min initiallyinduced a modest increase of nNOS and iNOS protein levels atthe time of the cessation of restraint. Then a progressive,significant decrease of the iNOS level appeared 1 – 2 hours laterand a subsequent, non-complete return occurred 3 hours later(Anova F(4, 46) = 6.707; P< 0.0002]. The changes of the nNOSlevel were parallel to those of iNOS but not significant (AnovaF(4, 46) = 1.993; P= 0.1083].

In PFC, a single restraint immediately after its terminationmoderately increased the IL-1β protein level, which significantlyincreased 2 and 3 hours later (Anova F (4, 44) = 12.61; P< 0.0001).The level of IL-1β in PFC changed parallel to the alterations ofnNOS and iNOS, but appeared at significantly higher than controllevel, whereas the nNOS and iNOS levels decreased below thecontrol values in non-stresses rats. (Fig. 1A). In PFC, repeatedrestraint for 3 days markedly enhanced the homotypic stress-induced nNOS level 1 hour after stress termination, whencompared to the level in the non-stressed group (Anova F(4,46)=5.287; P< 0.0002). Concomitantly, the iNOS level induced bysubsequent homotypic stress increased considerably, immediatelyafter restraint cessation (Anova F(4, 42) = 11.92; P< 0.0001) andslowly decreased 1 – 3 hours later (Fig. 1B). In PFC, repeatedrestraint for 3 days significantly enhanced the IL-1β level, inducedby homotypic stress, during the first 2 hours to the maximum value(Anova F(4, 46) = 6.456; P< 0.0001) and later it declined to the thesame level as immediately after stress termination (Fig. 1B). InPFC restraint for 7 days blunted the increase of the homotypicstress induced nNOS level after prior restraint for 3 days (AnovaF(4, 42) = 1.086; P= 0.3744) and markedly impaired the increaseof iNOS level, thought it remained at significantly higher levelwhen compared with control level in previously non stressed rats(Anova F(4, 42) = 4.350; P< 0.0011). Restraint for 7 days alsoblunted the significant increase of the IL-1β level in rats restrainedfor 3 days but did not appreciably alter the IL-1β level induced byhomotypic stress in rats stressed for 3 days (Anova F(4, 46) =1.154; P= 0.3361) (Fig. 1C). Repeated restraint for 14 days furtherdiminished but did not abolish the homotypic stress-inducedincrease of iNOS after 7 days of restraint (Anova F(4, 42) = 5.386;P< 0.0002). Homotypic stress after repeated restraint for 14 days,significantly lowered the nNOS level when compared to the levelafter 3 and 7 days of prior restraint, as well as after single restraint(Anova F(4, 48) = 3.945; P< 0.0023). After restraint for 14 daysthe IL-1β level in PFC did not significantly differ from its controllevel in previously non-restrained rats (Anova F(4, 46) = 1.345; P= 0.2590) (Fig. 1D).

In the hippocampus a single restraint induced transient andsignificant elevation of the nNOS level 2 hours after stresstermination (Anova F(4, 44) = 3.467; P< 0.0103) and a significantdecrease of the iNOS level 1 hour later (Anova F(4, 46) = 2.860; P< 0.0275). Single restraint gradually increased IL-1β level whichreached the highest value 3 hours after stress cessation (Anova F(4,46) = 3.741; P< 0.0084) (Fig. 2A). In the hippocampus repeatedrestraint for 3 days markedly enhanced the homotypic stress-induced increase of the nNOS level 1 hour later (Anova F(4, 46) =3.678; P< 0.0038) and evoked significant increase of the iNOSlevel in the whole post-stress period, when compared to the level incontrol non-stressed rats (Anova F(4, 42) = 8.762; P< 0.0001)(Fig. 2B). Repeated restraint for 3 days also significantly enhancedthe homotypic stress-induced increase of the IL-1β level (AnovaF(4, 44) = 3.704; P< 0.0045) (Fig. 2B). After longer restraint, for7 days, homotypic stress induced a moderate decrease of nNOSlevel immediately after stress termination (Anova F(4, 46) = 2.925;

429

P < 0.0187) and a significant increase of iNOS throughout thewhole post stress period (Anova F(4, 46) = 3.760; P< 0.0039).Repeated stress for 7 days transiently increased the restraint stress-induced IL-1β level 1 hour after stress termination (Anova F(4, 42)= 0.8667; P= 0.5073) (Fig. 2C). In the hippocampus 14 days ofrestraint significantly decreased the nNOS level induced by

homotypic stress (Anova F(4, 46) = 4.667; P< 0.0008) andsubstantially impaired the increase of iNOS level during 2 hoursafter stress termination (Anova F(4, 44) = 6.260; P< 0.0001).Prolonged restraint also markedly increased the homotypic stress-induced IL-1β level during the post stress period (Anova F(4, 42)= 4.693; P< 0.0014) (Fig. 2D).

430

Fig. 1. Neuronal (nNOS) and inducible(iNOS) nitric oxide synthase andinterleukin 1β (IL-1β) content inprefrontal cortex in rats exposed tosingle restraint stress (1 × RS) for 10min. Rats were restrained in metaltubes (for 10 min) and decapitated atthe termination of restraint and 1, 2 and3 hours later (Fig. 1A). Effect of priorrestraint for 10 min two times per dayfor 3 (Fig. 1B), 7 (Fig. 1C) and 14 (Fig.1D) days on 10 min RS-induced nNOS,iNOS and IL-1β content in prefrontalcortex. Data was assessed by one-wayANOVA followed by Tukey's multiplerange test. Each point represents themean ± S.E.M. of 10 – 12 rats pergroup +P< 0.05, ++P< 0.01, +++P< 0.001vs. non-stressed control group. Rightpanels show the representativeimmunoblots showing the expressionof nNOS, iNOS and IL-1β in prefrontalcortex.

In the hypothalamus a single restraint markedly enhancednNOS level after stress cessation (Anova F(4, 55) = 4.795; P<0.0022) and it significantly decreased iNOS level 3 hours afterrestraint (Anova F(4, 46) = 3.9; P< 0.0057). Single restraintstress failed to markedly affect IL-1β level in this structure(Anova F(4, 53) = 0.2750; P= 0.9257), moderate changes were

parallel to these in iNOS level (Fig. 3A). In hypothalamus,repeated stress for 3 days slightly increased nNOS level inducedby homotypic stress, compared to corresponding control levels innon-stressed rats (Anova F(4, 46) = 2.041; P= 0.0787) (Fig. 3B).The response of iNOS to the following homotypic stress wasconsiderably enhanced and most pronounced immediately after

431

Fig. 2. Neuronal (nNOS) and inducible(iNOS) nitric oxide synthase andinterleukin 1β (IL-1β) content inhippocampus in rats exposed to singlerestraint stress (1 × RS) for 10 min.Rats were restrained for 10 min anddecapitated at the termination ofrestraint and 1, 2 and 3 hours later (Fig.2A). Effect of prior restraint for 10 mintwo times per day for 3 (Fig. 2B), 7(Fig. 2C) and 14 (Fig. 2D) days on 10min RS-induced nNOS, iNOS and IL-1β content in hippocampus. Data wasassessed by one-way ANOVA followedby Tukey's multiple range test. Eachpoint represents the mean ± S.E.M. of10 – 12 rats per group +P< 0.05, ++P <0.01, +++P < 0.001 vs. non-stressedcontrol group. Right panels show therepresentative immunoblots showingthe expression of nNOS, iNOS and IL-1β in hippocampus.

stress cessation (Anova F(4, 44) = 9.189; P<0.0001), thengradually declined but persisted at significantly higher thancontrol level during the whole post-stress period. Repeatedrestraint for 3 days also significantly increased the homotypicstress-induced IL-1β level after 1 hour at the maximum (F(4, 46)

= 6.130; P< 0.0001) and then slightly declined (Fig. 3B). Alonger period of repeated restraint, for 7 days, significantlydiminished nNOS level (Anova F(4, 42) = 7.076; P< 0.0001),evoked by subsequent homotypic stress (Fig. 3C). The parallelincrease of iNOS level (Anova F(4, 44) = 3.790; P< 0.0046) was

432

Fig. 3. Neuronal (nNOS) andinducible (iNOS) nitric oxidesynthase and interleukin 1β (IL-1β)content in hypothalamus in ratsexposed to single restraint stress (1 ×RS) for 10 min. Rats were restrainedin metal tubes (for 10 min) anddecapitated at the termination ofrestraint and 1, 2 and 3 hours later(Fig. 2A). Effect of prior restraint for10 min two times per day for 3 (Fig.2B), 7 (Fig. 2C) and 14 (Fig. 2D) dayson 10 min RS-induced nNOS, iNOSand IL-1β content in hypothalamus.Data was assessed by one-wayANOVA followed by Tukey's multiplerange test. Each point represents themean ± S.E.M. of 10 – 12 rats pergroup +P < 0.05, ++P < 0.01, +++P <0.001 vs. non-stressed control group.Right panels show the representativeimmunoblots showing the expressionof nNOS, iNOS and IL-1β inhypothalamus.

433

Fig. 4. Comparison of the influence of restraint stress for 10 mintwo times per day for 3 days and 1 × RS, restraint for 10 min twotimes per day for 7 days and 1 × RS, restraint for 10 min two timesper day for 14 days and 1 × RS on nNOS (A), iNOS (B) and IL-1β (C) content in prefrontal cortex. Rats were restrained in metaltubes (for 10 min) and decapitated at the termination of restraintand 1, 2, and 3 hours later. Data were analyzed by two-wayanalysis of variance (ANOVA) followed by Bonferroni's multiplecomparison test Each point represents the mean ± S.E.M. of 10 –12 rats per group *P< 0.05, ** P< 0.01, *** P< 0.001 vs. grouprestrained for 10 min at 0, 1, 2 and 3 hour respectively.

Fig. 5. Comparison of the influence of restraint stress for 10min two times per day for 3 days and 1 × RS, restraint for 10min two times per day for 7 days and 1 × RS, restraint for 10min two times per day for 14 days and 1 × RS on nNOS (A),iNOS (B) and IL-1β (C) content in hippocampus. Data wereanalyzed by two-way analysis of variance (ANOVA) followedby Bonferroni's multiple comparison test Each point representsthe mean ± S.E.M. of 10 – 12 rats per group *P< 0.05, ***P <0.001 vs. group restrained for 10 min at 0, 1, 2 and 3 hourrespectively.

considerably less marked compared to the level in rats stressedfor 3 days. Prior restraint for 7 days did not substantially alter thehomotypic stress-induced IL-1β level (Anova F(4, 46) = 0.2750;P = 0.9257) compared to the control level in non -stressed rats.(Fig. 3C). After prolonged restraint for 14 days, the hypothalamicnNOS (Anova F(4, 44) = 1.849; P= 0.1402) and iNOS (AnovaF(4, 48) = 2.041; P= 0.0787) level responses to homotypic stressdid not significantly differ from their control values in non-repeatedly stressed rats (Fig. 3D). However, prolonged restraintfor 14 days strongly enhanced the homotypic stress-inducedincrease of IL-1β level in the hypothalamus during 3 hours poststress-period (Anova F(4, 44) = 7.547; P< 0.0001) (Fig. 3D).

Prior restraint period-dependent homotypic stress-induced nNOS,iNOS and IL-1β levels in brain structures

In the prefrontal cortex, prior restraint for 3 days induced amarked, transient increase of nNOS level 1 hour after homotypicstress termination (Anova F(3, 56) = 3.774; P< 0.0156). Repetedrestraint for 7 days transiently diminished the homotypic stress-induced nNOS level (Anova F(3, 56) = 5.898; P< 0.0014) andafter 14 days of restraint nNOS level significantly decreased(Anova F(3, 56) = 10.62; P< 0.0001) when compared to the levelsin control non-stressed rats (Fig. 4A). Repeated restraint for 3 daysconsiderably enhanced the homotypic stress-induced iNOSprotein level (Anova F(3, 56) = 6.005; P< 0.0013), during 3 hoursof a post-stress period. Prior restraint for 7 days evoked a lesspronounced, though significant, increase of the homotypic stress-induced iNOS response (Anova F(3, 56) = 4.288; P< 0.0086) andafter 14 days of repeated restraint, iNOS level further decreasedalthough it remained at a significantly higher level than aftersingle restraint (Anova F(3, 56) = 16.99; P< 0.0001) (Fig. 4B).Prior repeated restraint for 3 days also significantly enhanced thehomotypic stress-induced IL-1β level in the prefrontal cortex(Anova F(3, 56) = 16.59; P< 0.0001) to the highest level 1 and 2hours after stress termination. A longer restraint period, for 7 days,did not alter the homotypic stress-induced IL-1β response (AnovaF(3, 56) = 2.340; P= 0.083) and prior restraint for 14 dayssignificantly diminished this response (Anova F(3, 56) = 9.239; P< 0.0001), 3 hours after stress termination when compared to theresponse in control non-stressed rats (Fig. 4C).

In the hippocampus, repeated restraint for 3 days moderatelyincreased the homotypic stress-induced nNOS level immediatelyafter cessation of stress (Anova F(3, 56) = 4.563; P= 0.0063),which gradually vanished in the post-stress period. A longerrepeated restraint, for 7 days, moderately lowered the homotypicstress-induced nNOS level (Anova F(3, 56) =3,534; P= 0.0204),compared to the level in previously non stressed rats. After 14 daysof prior restraint the homotypic stress evoked a marked decrease ofnNOS level during 2 hours after cessation of restraint and later itsignificantly increased nNOS response (Anova F(3, 56) = 10.49; P< 0.0001), over the control level in prior non stressed rats (Fig. 5A).In the hippocampus, repeated restraint for 3 days induced a strongsignificant increase of the homotypic stress-induced iNOSresponse, during 3 h of post-stress period (Anova F(3, 56) = 4.290;P < 0.0085). A longer period of restraint, for 7 days, evoked aweaker, though significant increase of the homotypic stress-induced iNOS level (Anova F(3, 56) = 9.993; P< 0.0001) andrestraint for 14 days induced a still weaker enhancement of iNOSlevel (Anova F(3, 56) = 10.39; P< 0.0001) when compared to theincrease effected by 3 days of repeated restraint (Fig. 5B). Previousrestraint for 3 days also induced a marked, more pronouncedincrease of the homotypic stress-induced IL-1β level (Anova F(3,56) = 3.424; P< 0.0027) in the post-stress period and a less,moderately increased response appeared after 7 days (Anova F(3,56) = 4.711; P< 0.0053) and 14 days of prior restraint (Anova F(3,56) =1.050; P= 0.3779), (Fig. 5C).

434

Fig. 6. Comparison of the influence of restraint stress for 10 mintwo times per day for 3 days and 1 × RS, restraint for 10 min twotimes per day for 7 days and 1 × RS, restraint for 10 min two timesper day for 14 days and 1 × RS on nNOS (A), iNOS (B) and IL-1β (C) content in hypothalamus. Data were analyzed by two-wayanalysis of variance (ANOVA) followed by Bonferroni's multiplecomparison test Each point represents the mean ± S.E.M. of 10 –12 rats per group *P< 0.05, **P< 0.01, ***P < 0.001 vs. grouprestrained for 10 min at 0, 1, 2 and 3 hours respectively.

In the hypothalamus, restraint for 3 days did not markedlyalter the homotypic stress-induced increase of nNOS protein level(Anova F(3, 56) = 0.6308; P= 0.5982),during the post-stressperiod. Previous stress for 7 days markedly decreased nNOS level(Anova F(3, 56) = 3.596; P< 0.0190) and after restraint for 14days, nNOS level induced by homotypic stress altered from aninitial fall 1 hour after stress termination to a significant increase(Anova F(3, 56) = 8.876; P< 0.0001) 1 hour later (Fig. 6A).Repeated restraint for 3 days strongly intensified the homotypicstress- induced increase of iNOS level (Anova F(3, 56) = 6.157; P< 0.0011), which slightly declined but remained at a significantlyhigher level throughout 3 hours post-stress period. Prior restraintfor 7 days induced a significant increase of iNOS levels (AnovaF(3, 56) = 3.205; P< 0.0300), over the control level. Repeatedrestraint for 14 days markedly increased the homotypic stress-induced iNOS level when compared to the control level in non-stressed rats (Anova F(3, 56) = 0.7850; P= 0.05073), (Fig. 6B).The homotypic stress-induced increase in the hypothalamic IL-1βlevel was significantly enhanced (Anova F(3, 56) = 4.854; P<

0.0045), by previous restraint for 3 days. Restraint for 7 days didnot substantially alter IL-1β level (Anova F(3, 56) = 2.374; P=0.0798), compared to that in non-stressed rats, and prolongedstress for 14 days evoked the most pronouced enhancement of thehomotypic stress-induced increase in IL-1β levels (Anova F(3, 56)= 6.977; P< 0.005), above the control levels in non- stressed rats(Fig. 6C).

Effect of IL-1Ra on homotypic stress-induced plasma IL-1β,ACTH and corticosterone levels after repeated restraint

A single restraint for 10 min significantly increased plasma IL-1β level immediately after stress termination (Anova F(4, 44) =55.22; P< 0.0001) and to a maximum level 1 hour later. IL-1Ra(100 µg/kg i.p.) given 15 min before restraint abolished thisincrease in the whole 3 hours post-stress period (Anova F(3, 88) =32.5; P < 0.0001) (Fig. 7A). Repeated restraint for 3 dayssignificantly increased the homotypic stress-induced plasma IL-1βlevel 1 hour later than a single restraint (Anova F(4, 57) = 5.592; P

435

Fig. 7. Influence of IL-1 receptor antagonist (IL-1Ra), given i.p. 15 min before single restraint stress for 10 min, on IL-1β plasma levels(Fig. 7A). Influence of IL-1 receptor antagonist (IL-1Ra), given i.p. 15 min before restraint stress for 10 min, on IL-1β plasma levels inducedby restraint stress for 3 days (Fig. 7B), 7 days (Fig. 7C) and 14 days (Fig. 7D). On the day of the experiment, rats were exposed to singlerestraint for 10 min and decapitated at the termination of restraint and 1, 2 and 3 hours later. Data were analyzed by one-way ANOVAfollowed by Tukey's multiple range test (each point represents the mean ± S.E.M. of 10 – 12 rats per group +P< 0.05, ++P< 0.01, +++P <0.001 vs. non-stressed control group) and two-way ANOVA followed by Bonferroni's multiple comparison test (each point represents themean ± S.E.M. of 10 – 12 rats per group *P< 0.05, ** P< 0.01 vs. group restrained for 10 min at 0, 1, 2 and 3 hour respectively).

< 0.0005). IL-1Ra, given i.p. 15 min before homotypic stressblunted this increase to the control level in rats exposed to repeatedrestraint for 3 days (Anova F(3, 56) = 5.437; P< 0.00241) (Fig.7B). Previous restraint for 7 days significantly enhanced homotypicstress-induced plasma IL-1β level 0 – 3 h after restraint termination(Anova F(4, 54) = 5.861; P< 0.0003) and IL-1Ra abolished thisincrease in the whole post stress period (Anova F(3, 56) = 5.019; P< 0.0038) (Fig. 7C). Similarly, chronic restraint for 14 dayssignificantly elevated the homotypic stress-induced plasma IL-1βlevel (Anova F(4, 60) = 3.006; P< 0.0181) and IL-1Raadministered before restraint totally reduced this increase 1 and 2hours after stress termination (Anova F(3, 56) = 5.467; P< 0.0023)(Fig. 7D).

Single restraint rapidly and strongly increased the plasmaACTH level which decreased in a fast manner to resting levelduring 2 hours after stress termination (Anova F(4, 55) = 46.05;P < 0.0001). IL-1Ra given i.p. significantly inhibited therestraint-induced increase of plasma ACTH level immediately

and 1 h after stress termination (Anova F(3, 84) = 14.73; P<0.0001) (Fig. 8A). In rats previously restrained for 3 days IL-1Raadministered 15 min before homotypic stress moderatelyweakened the increase of plasma ACTH response immediatelyafter termination of stress (Anova F(3, 56) = 2.434; P= 0.0743),(Fig. 8B), modestly diminished this response after 7 days ofprior restraint (Anova F(3, 56) = 0.1766; P= 0.09118), (Fig. 8C),and did not alter this response after repeated restraint for 14 days(Anova F(3, 56) = 0.1766; P= 0.9987), (Fig. 8D).

Single restraint immediately after termination rapidly andstrongly increased plasma corticosterone level which returned in asimilar manner to control level 2 hours later (Anova F(4, 55) =56.46; P < 0.0001). IL-1Ra administered i.p. before stresssignificantly diminished the restraint-induced the increase and thetime of returning of plasma corticosterone to control level (AnovaF(3, 88) = 5.76; P< 0.0001) (Fig. 9A). Plasma corticosterone levelsignificantly increased by homotypic stress in repeatedly restrainedrats for 3 days was not altered by pretreatment with IL-1Ra (Anova

436

Fig. 8. Influence of IL-1 receptor antagonist (IL-1Ra), given i.p. 15 min before single restraint stress for 10 min, on ACTH plasmalevels (Fig. 8A) Influence of IL-1 receptor antagonist (IL-1Ra), given i.p. 15 min before restraint stress for 10 min, on ACTH plasmalevels induced by restraint stress for for 3 days (Fig. 8B), 7 days (Fig. 8C) and 14 days (Fig. 8D). On the day of the experiment, ratswere exposed to single restraint for 10 min and decapitated at the termination of restraint and 1, 2 and 3 hours later. Data were analyzedby one-way ANOVA followed by Tukey's multiple range test (each point represents the mean ± S.E.M. of 10 – 12 rats per group +P <0.05, ++P< 0.01, +++P< 0.001 vs. non-stressed control group) and two-way ANOVA followed by Bonferroni's multiple comparison test(each point represents the mean ± S.E.M. of 10 – 12 rats per group *P< 0.05, ** P< 0.01 vs. group restrained for 10 min at 0, 1, 2and 3 hour respectively).

F(3, 56) = 0.4969; P= 0.6859) (Fig. 9B). After repeated restraintfor 7 days this antagonist markedly enhanced the homotypic stress-induced rise of plasma corticosterone at the cessation of restraint(Anova F(3, 56) = 3.797; P< 0.0150) (Fig. 9C). In rats previouslyrestrained for 14 days IL-1Ra moderately enhanced the fall ofplasma corticosterone level 1 and 2 hour after homotypic stresstermination (Anova F(3, 56) = 1.833; P= 0.1517), (Fig. 9D).

DISCUSSION

Effect of restraint stress on nNOS, iNOS and IL-1β levels in brainstructures

In the present study, a single restraint which is a milderexperimental type of stressors, affected nitric oxide synthasesand IL-1 systems in brain structures regulating stressresponses.

In the prefrontal cortex restraint for 10 min induced parallelalterations of nNOS and iNOS protein levels. Immediately afterrestraint cessation both izoenzyme levels transiently increased and1 – 2 hours later the iNOS level significantly decreased andreturned to the resting level 3 hours after stress termination. Asignificant fall of iNOS level may suggest either excessive use of alimited amount of constitutive iNOS or/and its insufficientgeneration during increased demand in the initial phase of acutestress reaction in PFC. Although it is generally accepted that iNOSis undetectable under basal conditions, some evidence indicatesthat iNOS can be constitutively expressed in different brain regionsof rats (21). Constitutive levels of nNOS and iNOS in PFC may notbe sufficient to compensate for an increased demand in the initialphase of acute stress reaction. Neuronal NOS is constitutivelyexpressed in neurons, microglial cells and astroglial cells (32). Thegeneration of NO by the nNOS isoform appears very rapidly sinceit does not require mRNAsynthesis, and can act as a synapticneurotransmitter. During stress the generation of NO by iNOS is

437

Fig. 9. Influence of IL-1 receptor antagonist (IL-1Ra), given i.p. 15 min before single restraint stress for 10 min, on corticosteroneplasma levels (Fig. 9A). Influence of IL-1 receptor antagonist (IL-1Ra), given i.p. 15 min before restraint stress for 10 min, oncorticosterone plasma levels induced by restraint stress for for 3 days (Fig. 9B), 7 days (Fig. 9C) and 14 days (Fig. 9D). On the dayof the experiment, rats were exposed to single restraint for 10 min and decapitated at the termination of restraint and 1, 2 and 3 h later.Data were analyzed by one-way ANOVA followed by Tukey's multiple range test (each point represents the mean ± S.E.M. of 10 – 12rats per group +P < 0.05, ++P < 0.01, +++P < 0.001 vs. non-stressed control group) and two-way ANOVA followed by Bonferroni'smultiple comparison test (each point represents the mean ± S.E.M. of 10 – 12 rats per group *P< 0.05, **P< 0.01 vs. group restrainedfor 10 min at 0, 1, 2 and 3 hour respectively).

long-lasting but in far higher local concentration than by nNOS oreNOS. In our experiment after single restraint both nNOS andiNOS protein levels in PFC initial declined during 2 hours and thenapproached their control levels which suggests the induction ofNOS gene expression by acute restraint stress (33). Single restraintstress caused acute changes in NO producing neurons in theamygdala complex and delayed modifications in the hippocampalformation (34). Our data show that in PFC acute restraint stressinduced a moderate, transient decrease of the IL-1β level 1 hourafter restraint cessation which significantly increased 2 and 3 hourslater probably via a genomic mechanism. IL-1 is regarded as anessential mediator of adaptive stress responses. This cytokine andstressors are able to impair neuronal plasticity andneurotransmission contributing to depression illness (1, 35, 36). Inour experiment the changes of the IL-1β levels in PFC wereparallel in time to the alterations of nNOS and iNOS, but appearedat a significantly higher level. Cytokines increased by exposure tostress can induce iNOS expression (37) and NO synthesis, which isinvolved in the modulation of stress induced neuroendocrineaction. In the present experiment a single restraint increased the IL-1β level in PFC, whereas the nNOS and iNOS levels weremarkedly decreased below the control level in non- stressed rats.Our results show that neuromodulatory changes during acute stressrapidly disrupt PFC network connection of IL-1 and NO systemsand markedly impair function of the most sensitive structure todetrimental effects of stress (7).

In the hippocampus a single restraint induced transientelevation of the nNOS level and a significant decrease of the iNOSlevel 2 and 3 hours, after restraint termination. This may suggest amarked involvement of nNOS and iNOS in an acute stress-induced reaction before further compensation by a genomicmechanism. A single restraint significantly increased the IL-1βlevel 3 hours after stress termination in the hippocampus, similarto the increase in PFC but not in the hypothalamus. A singlerestraint after 3 hours induced reciprocal changes in hippocampusiNOS level (a significant decrease) and IL-1β levels (a significantincrease), but repeated restraint for 3 days evoked a strongincrease of both iNOS and IL-1β levels in that structuresuggesting the involvement of NO generated by iNOS and IL-1βin adaptational compensatory response during repeated stress.

In the hypothalamus a single restraint markedly enhancedthe nNOS protein level in the whole post stress period, perhapsdue to a larger presence of nNOS in the neuronal processes ofdifferent neurons involved in the central regulation of stressreactions in that structure.

These results suggest that single stress activates the nNOSprotein system in brain structures containing a neuronal net, likethe hippocampus and hypothalamus. By contrast single stresssignificantly decreased the iNOS protein level 3 hours afterstress termination below the control level in non- stressed rats.Acute restraint stress also significantly declined the IL-1β levelparallel to similar decrease of the iNOS level. NO may regulatebrain IL-1 gene expression in the hypothalamic PVN whichreleases CRH (38), the first stimulator of the HPA axis. Thesechanges indicate that in the hypothalamus single restraint stressmoderately enhanced nNOS protein level and markedlydecreased iNOS and IL-1β levels (and functions).

Repeated stress-induced-adaptational changes of the nNOS, iNOSand IL-1β levels in brain structures induced by homotypic stress

The present data show that in PFC repeated restraint for 3days consistently induced the strongest increase of the iNOSprotein level in response to followed homotypic stress. Thisstrong sensitized increase of iNOS response gradually declinedafter prior prolonged exposures to restraint for 7 and 14 days, butremained at a markedly higher level when compared to the

parallel decreased iNOS response in control prior non- stressedrats. Also the responses of nNOS and IL-1β protein levels to thehomotypic stressor in PFC are sensitized to a maximum extentafter 3 days of repeated restraint. Our results suggest that thecumulative effects of repeated stress for 3 days may be morevulnerable for animals and humans subjected to a subsequenthomotypic stressor.

In the hippocampus, repeated stress for 3 days moderatelyincreased the homotypic stress induced nNOS level. Thehomotypic stress induced iNOS protein level was significantlyand most strongly enhanced after 3 days and to a gradually lesserextent after prolonged exposure to stress for 7 and 14 days. Thisindicates that also in the hippocampus like in PFC, the highestincrease of iNOS level appears after 3 days of prior stress.Repeated restraint for 3 days also significantly enhanced thehomotypic stress-induced increase of the IL-1β level but alonger prior restraint did not alter the IL-1β protein level whencompared with a single restraint-induced level. Our resultsindicate that restraint, a prevailing psychological stressor,engages more strongly the hippocampus for adapting the stress-induced NO and IL-1β regulatory functions (39). Our resultssuggest that in the hippocampus homotypic stress-inducednNOS, iNOS and IL-1β levels are significantly sensitized byrepeated restraint, and that the iNOS and IL-1β levels areparticularly enhanced after 3 days of prior stress.

Restraint stress repeated for 3 days selectively enhancedexcitatory synaptic inputs to parvocellular neurons of the PVNleading to increased CRH release and HPA axis stimulation (40).In the hypothalamus repeated stress for 3 – 14 days slightlyaltered the nNOS protein level induced by subsequenthomotypic stress as compared with the level after singlerestraint. Repeated restraint for 7 days significantly diminishedthis level and after 14 days of restraint a marked increase of thenNOS level in response to successive homotypic stressappeared. On the contrary, prior restraint for 3 days considerablyincreased the homotypic stress-induced iNOS level during 3hours after stress cessation. A vanishing increase of iNOS levelin response to homotypic stress appeared after 7 and 14 days,after repeated stress. In the hypothalamus repeated restraint for3 days and more strongly after 14 days enhanced the homotypicstress-induced IL-1β level 1 hour after stress termination. Thesebiphasic adaptational changes of homotypic stress-induced IL-1β in the hypothalamus suggest strong sensitization of the IL-1βresponse after 3 and 14 days of prior repeated stress treatment.Single administration of IL-1β induced long lasting sensitizationof the HPA axis to novelty stress 3 – 22 days later, parallel fastand delayed with biphasic increases of CRH and HPA axissensitization (41). Prolonged, biphasic increase of IL-1β level inthe hypothalamus after 3 and 14 days in our experiment may bedependent in part on HPA system adaptation to chronicemotional stress. Emotional stress in restrained rats can induce along-lasting sensitization of IL-1β-evoked stimulation ofhypothalamic PVN (41, 42). Proinflammatory cytokines inplasma may be involved over a long period in human metabolicliver disease (43). Multiple pathogenic factors can be involved inthe regulation of autonomic nervous system. Prolongedemotional stress might disturb neuromodulation in the brain-gutinteractions during gastrointestinal disorders (44).

A single restraint, significantly increased the plasma IL-1βlevel immediately and 1 hour after stress cessation. This increasewas abolished by IL-1Ra administered i.p. 15 min before restraintsuggestion the selectivity of IL-1β system in this stress-inducedreaction. Parallel to the changes of the plasma IL-1β level, theACTH level also significantly increased immediately aftercessation of restraint and declined to around control level 2 hourlater. Our data indicate that this initial increase of ACTH level ina significant part depends on the elevated IL-1β level, since IL-

438

1Ra reduced the stress-induced increase in plasma ACTH levelafter restraint termination. After 3 days of repeated restrainthomotypic stress-induced a rapid increase in plasma ACTH levelwhich was moderately diminished by pretreatment with IL-1Raand after longer repeated restraint, for 7 and 14 days, IL-1Ra didnot affect the homotypic stress induced increase of ACTH level.These results suggest that prior repeated stress inhibitedcompletely the IL-1β stimulatory component observed aftersingle restraint stress. Single restraint stress-induced considerableincrease of plasma corticosterone level that was significantlydiminished by prior i.p. IL-1Ra administration. Repeated stressfor 3 and 14 days abolished the inhibitory action of IL-1Ra onhomotypic stress-induced corticosterone secretion. However, theenhancement by IL-1Ra the homotypic stress inducedcorticosterone response after prior repeated stress for 7 days isdifficult to explain. Corticosterone release during adaptation toprolonged stress is regulated by a multiplicity of localadrenocortical signaling factors like cytokines, NO,prostaglandins. During stress adaptation the activity of thesefactors may be suppressed or sensitized in not a parallel time anddirection manner. It is known that IL-1β has a marked stimulatoryeffect in adrenal gland whereas NO may stimulate or inhibitcorticosterone synthesis and release (42). At present the action ofthe temporal dependent factors, connected with a localstereoidogenesis in vivo, in regulation of corticosterone secretionduring stress adaptation is not explained. The long and commonlyaccepted view that the functions of the adrenal cortex weremainly, or exclusively, regulated by circulating ACTH wassuccesively completed by pointing out that manyneurotransmitters and neuropeptides (45), as well as cytokinesand NO also participate in the local paracrine regulation ofadrenocortical steroidogenesis (3, 24, 25). Recent evidencesuggests interaction between cytokine pathways and HPAhormones directly at the level of the pituitary or adrenal glands tostimulate the secretion of ACTH and corticosterone (46, 47). IL-1β dose-dependently enhanced plasma corticosteroneconcentrations. Different stressors induce a rapid and highincrease of IL-1β expression within the adrenal glands. NOsignificantly decreased corticosterone production both inunstimulated and in corticotropin stimulated zona fasciculataadrenal cells. Both nNOS and eNOS were detected in rat adrenalzona fasciculata and several NO donors significantly decreasedboth basal and ACTH-stimulated corticosterone generation in rats(24). Inflammatory messengers like IL-1β and prostaglandin(PG) E2 influence (in vitro) the release of ACTH andcorticosterone by induction production of IL-1β as well asinducible cyclooxygenase (COX-2) in the adrenal immune cellpopulation (47). The adrenal gland adapts to various forms ofacute and chronic stress by producing cytokines and upregulationof cyclooxygenase and nitric oxide synthase enzymes that canmodulate corticosterone release. Our results indicate that plasmaIL-1β is markedly involved in a single restraint stress-inducedstimulation of ACTH and corticosterone secretion.

We conclude that chronic stress may both sensitize ordecrease specific control circuits to stimulate or inhibit stress-sensitive systems. In PFC repeated restraint for 3 days enhancedmost strongly the homotypic stress-induced increase of iNOSand to a lesser extent IL-1β levels. The sensitization of iNOS andIL-1β responses vanished after 14 days. In the hippocampus, thestrongest increase of homotypic stress-induced iNOS, nNOS andIL-1β levels appeared after 3 days of prior stress and decreased7 and 14 days later. In the hypothalamus the iNOS level inducedby homotypic stress was strongly enhanced by a prior 3 days ofstress and returned to the control level after 7 days. Repeatedstress induced a biphasic increase of the IL-1β level in responseto homotypic stress after 3 days and 14 days. This suggests thathomotypic stress significantly increases the IL-1β response after

3 days, which had vanished after 7 days, and increased morestrongly after a further 7 days of stress. Our results present adistinct structure specificity and intensity of cytokine and nitricoxide in brain structures required to preserve or adapthomeostatic and protective cooperation mechanisms.

Acknowledgements: This research was supported by grant:POIG 01.01.02-12-004/09-00 financed by European RegionalDevelopment Fund as well as by the statutory funds from theInstitute of Pharmacology, Polish Academy of Sciences.

Conflict of interests: None declared.

REFERENCES

1. Goshen I, Yirmiya R. Interleukin-1(IL-1): a central regulatorof stress responses. Front Neuroendocrinol 2009; 30: 30-45.

2. Gibb J, Hayley S, Poulter MO, Anisman H. Effects ofstressors and immune activating agents on peripheral andcentral cytokines in mouse strains that differ in stressorresponsivity. Brain Behav Immun 2011; 25: 468-482.

3. Gadek-Michalska A, Tadeusz J, Rachwalska P, Bugajski J.Cytokines, prostaglandins and nitric oxide in the regulation ofstress-response systems. Pharmacol Rep 2013; 65: 1655-1662.

4. Nguyen KT, Deak T, Will MJ, et al. Timecourse andcorticosterone sensitivity of the brain, pituitary, and seruminterleukin-1β protein response to acute stress. Brain Res2000; 859: 193-201.

5. O'Connor KA, Johnson JD, Hansen MK, et al. Peripheraland central proinflammatory cytokine response to a severeacute stressor. Brain Res 2003; 991: 123-132.

6. Capuron L, Miller AH. Immune system to brain signaling:neuropsychopharmacological implications. Pharmacol Ther2011; 130: 226-238.

7. Arnsten AF. Stress signalling pathways that impairprefrontal cortex structure and function. Nat Rev Neurosci2009; 10: 410-420.

8. Silverman MN, Pearce BD, Biron CA, Miller AH. Immunemodulation of the hypothalamic-pituitary-adrenal (HPA)axis during viral infection. Viral Immunol 2005; 18: 41-78.

9. Deak T, Bordner KA, McElderry NK, et al. Stress-inducedincreases in hypothalamic IL-1: a systematic analysis ofmultiple stressor paradigms. Brain Res Bull 2005; 64: 541-556.

10. Belda X, Rotllant D, Fuentes S, Delgado R, Nadal R,Armario A. Exposure to severe stressors causes long-lastingdysregulation of resting and stress-induced activation of thehypothalamic-pituitary adrenal axis. Ann NY Acad Sci 2008;1148: 165-173.

11. Bartolomucci A. Palanta P, Parmigiani S, et al. Chronicpsychosocial stress down-regulates central cytokine mRNA.Brain Res Bull 2003; 62: 173-178.

12. Busnardo C, Tavares RF, Resstel LB, Elias LL, Correa FM.Paraventricular nucleus modulates autonomic andneuroendocrine responses to acute restraint stress in rats.Auton Neurosci 2010; 158: 51-57.

13. Herman, J.P. Seroogy, K. Hypothalamic-pituitary-adrenalaxis, glucocorticoids, and neurologic disease. Neurol Clin2006; 24: 461-481.

14. Dedovic K, Duchesne A, Andrews J, Engert V, PruessnerJC. The brain and the stress axis: the neural correlates ofcortisol regulation in response to stress. Neuroimage 2009;47: 864-871.

15. Garthwaite J. Concepts of neural nitric oxide-mediatedtransmission. Eur J Neurosci 2008; 27: 2783-2802.

16. Zidek Z. Role of cytokines in the modulation of nitric oxideproduction by cyclic AMP. Eur Cytokine Netw 2001; 12: 22-32.

439

17. Lopez-Figueroa MO, Day HE, Akil H, Watson SJ. Nitric oxidein the stress axis. Histol Histopathol 1998; 13: 1243-1252.

18. Bugajski J, Gadek-Michalska A, Bugajski AJ. Nitric oxideand prostaglandin systems in the stimulation ofhypothalamic-pituitary-adrenal axis by neurotransmitters andneurohormones. J Physiol Pharmacol 2004; 55: 679-703.

19. Nelson RJ, Trainor BC, Chiavegatto S, Demas GE.Pleiotropic contributions of nitric oxide to aggressivebehavior. Neurosci Biobehav Rev 2006; 30: 346-355.

20. Zhou QG, Hu Y, Hua Y, et al. Neuronal nitric oxide synthasecontributes to chronic stress-induced depression bysuppressing hippocampal neurogenesis. J Neurochem 2007;103: 1843-1854.

21. Amitai Y. Physiologic role for "inducible" nitric oxidesynthase: a new form of astrocytic-neuronal interface. Glia2010; 58: 1775-1781.

22. Orlando GF, Wolf G, Engelmann M. Role of neuronal nitricoxide synthase in the regulation of the neuroendocrine stressresponse in rodents: insights from mutant mice. Amino Acids2008; 35: 17-27.

23. Zhou L, Zhu DY. Neuronal nitric oxide synthase: structure,subcellular localization, regulation and clinical implications.Nitric Oxide 2009; 20: 223-230.

24. Cymeryng CB, Lotito SP, Colonna C, et al. Expression ofnitric oxide synthases in rat adrenal zona fasciculata cells.Endocrinology 2002; 143: 1235-1242.

25. Djikic D, Budec M, Vranjes-Djuric S, et al. Ethanol andnitric oxide modulate expression of glucocorticoid receptorin the rat adrenal cortex. Pharmacol Rep 2012; 64: 896-901.

26. Mohn CE, Fernandez-Solari J, De Laurentiis A, BornsteinSR, Ehrhart-Bornstein M, Rettori V. Adrenal glandresponses to lipopolysaccharide after stress and ethanoladministration in male rats. Stress 2011; 14: 216-226.

27. Yamaguchi N, Ogawa S, Okada S. Cyclooxygenase andnitric oxide synthase in the presympathetic neurons in theparaventricular hypothalamic nucleus are involved inrestraint stress-induced sympathetic activation in rats.Neuroscience 2010; 170: 773-781.

28. Montezuma K, Biojone C, Lisboa SF, Cunha FQ, GuimaraesFS, Joca SR. Inhibition of iNOS induces antidepressant-likeeffects in mice: pharmacological and genetic evidence.Neuropharmacology 2012; 62: 485-491.

29. Gadek-Michalska A, Tadeusz J, Rachwalska P, Spyrka J,Bugajski J. Effect of prior stress on interleukin-1β and HPAaxis responses to acute stress. Pharmacol Rep 2011; 63:1393-1403.

30. Gadek-Michalska A, Tadeusz J, Rachwalska P, Spyrka J,Bugajski J. Effect of repeated restraint on homotypic stress-induced nitric oxide synthases expression in brain structuresregulating HPA axis. Pharmacol Rep 2012; 64: 1381-1390.

31. Gadek-Michalska A, Tadeusz J, Rachwalska P, Spyrka J,Bugajski J. Brain nitric oxide synthases in the interleukin-1β-induced activation of hypothalamic-pituitary-adrenalaxis Pharmacol Rep 2012; 64: 1455-1465.

32. Abbott NJ, Ronnback L, Hansson E. Astrocyte-endothelialinteractions at the blood-brain barrier. Nat Rev Neurosci2006; 7: 41-53.

33. Hueston CM, Deak T. The inflamed axis: the interactionbetween stress, hormones, and the expression ofinflammatory-related genes within key structurescomprising the hypothalamic-pituitary-adrenal axis. PhysiolBehav 2014; 30: 124, 77-91.

34. Echeverry MB, Guimaraes, Del Bel E. Acute and delayedrestraint stress-induced changes in nitric oxide producingneurons in limbic regions. Neuroscience 2004; 125: 981-993.

35. Hayley S, Poulter MO, Merali Z, Anisman H. Thepathogenesis of clinical depression: stressor- and cytokine-induced alterations of neuroplasticity. Neuroscience 2005;135: 659-678.

36. Hayley S. Toward an anti-inflammatory strategy fordepression. Front BehavNeurosci 2011; 13: 5-19.

37. Garcia-Bueno B, Caso JR, Leza JC. Stress as aneuroinflammatory condition in brain: damaging andprotective mechanisms. Neurosci Biobehav Rev 2008; 32:1136-1151.

38. Yang WW, Krukoff TL. Nitric oxide regulates bodytemperature, neuronal activation and interleukin-1 beta geneexpression in the hypothalamic paraventricular nucleus inresponse to immune stress. Neuropharmacology 2000; 39:2075-2089.

39. Pruessner JC, Dedovic K, Pruessner M, et al. Stressregulation in the central nervous system: evidence fromstructural and functional neuroimaging studies in humanpopulations - 2008 Curt Richter Award Winner.Psychoneuroendocrinology 2010; 35: 179-191.

40. Kusek M, Tokarski K, Hess G. Repeated restraint stressenhances glutamatergic transmission in the paraventricularnucleus of the rat hypothalamus. J Physiol Pharmacol 2013;64: 565-570.

41. Schmidt ED, Aguilera G, Binnekade R, Tilders FJ. Singleadministration of interleukin-1 increased corticotropinreleasing hormone and corticotropin releasing hormone-receptor mRNAin the hypothalamic paraventricular nucleuswhich paralleled long-lasting (weeks) sensitization toemotional stressors. Neuroscience 2003; 116: 275-283.

42. Johnson JD, O'Connor KA, Watkins LR, Maier SF. The roleof IL-1β in stress-induced sensitization of proinflammatorycytokine and corticosterone responses. Neuroscience 2004;127: 569-577.

43. Celinski K, Konturek PC, Slomka M, et al. Effects oftreatment with melatonin and tryptophan on liver enzymes,parameters of fat metabolism and plasma levels of cytokinesin patients with non-alcoholic fatty liver disease-14 monthsfollow up. J Physiol Pharmacol 2014; 65: 75-82.

44. Przybylska-Felus M, Furgala A, Zwolinska-Wcislo M, et al.Disturbances of autonomic nervous system activity anddiminished response to stress in patients with celiac disease.J Physiol Pharmacol 2014; 65: 833-841.

45. Delarue C, Contesse V, Lenglet S, et al. Role ofneurotransmitters and neuropeptides in the regulation of theadrenal cortex. Rev Endocr Metab Disord 2001; 2: 253-267.

46. Blandino P, Hueston, CM, Barnum CJ, Bishop C, Deak T.The impact of ventral noradrenergic bundle lesions onincreased IL-1 in the PVN and hormonal responses to stressin male sprague dawley rats. Endocrinology 2013; 154:2489-2500.

47. Engstrom L, Rosen K, Angel A, et al. Systemic immunechallenge activates an intrinsically regulated localinflammatory circuit in the adrenal gland. Endocrinology2008; 149: 1436-1450.

R e c e i v e d :October 20, 2014A c c e p t e d :April 8, 2015

Author's address: Dr. Anna Gadek-Michalska, Institute ofPharmacology Polish Academy of Sciences, 12 Smetna Street,31-343 Cracow, Poland.E-mail: gadek@if-pan.krakow.pl

440