【作者】 彭云丽；

【导师】 蒋春雷； 王云霞；

【作者基本信息】 第二军医大学 ， 航空、航天与航海医学， 2013， 博士

【摘要】 抑郁症是复杂的、多因素相互作用的结果，包括环境因素、遗传因素以及共患疾病间的相互影响等。应激是抑郁症的重要诱发因素，但其作用机制尚未明确。对于抑郁症的治疗，现有抗抑郁药和治疗手段尚不能满足临床需要。一方面，现有抗抑郁药存在严重副作用，临床起效慢；另一方面，相当一部分抑郁患者对现有抗抑郁药无疗效反应，且近半数抑郁患者治疗停止后出现复发。因此，多角度阐明应激诱发抑郁症的机制，寻找潜在抗抑郁靶点，开发新的、更有效、更特异的抗抑郁药物，具有重要的理论价值和现实意义。随着对抑郁症研究的不断深入，学者们提出了一系列关于抑郁症发病机制的假说或学说，包括单胺假说、神经退行性假说、细胞因子学说、下丘脑-垂体-肾上腺皮质(hypothalamic-pituitary-adrenal, HPA)轴功能障碍假说等。其中，细胞因子学说自1991年提出以来即引起广泛关注，并在过去二十年的时间里被众多学者不断论证补充。该学说认为抑郁症可能与免疫激活导致细胞因子的过度分泌有关。细胞因子是由激活的免疫细胞分泌的生物活性蛋白。研究发现，通过急性注射LPS或持续卡介苗免疫，均可激活动物免疫系统，并诱发抑郁行为。各种生理性、心理性应激源可引起免疫系统的激活和促炎细胞因子的大量释放。急、慢性应激暴露，特别是慢性应激，同样也能诱发人及实验动物抑郁行为的出现。免疫激活和应激具有相似的行为学效应。因此，免疫炎性反应可能是应激调控抑郁进程的作用机制之一。细胞因子间相互调节，构成一个复杂的网络结构，每种细胞因子除诱导自身合成外，还可以通过级联反应（cascade）的方式诱导其他细胞因子的合成，包括TNFα、IL-1β、IL-6等。中枢内，神经元、星形胶质细胞、小胶质细胞也构成细胞因子网络，它们不但能合成细胞因子，还可表达细胞因子受体，从而放大细胞因子信号。免疫信号传递的这种神经性基础为外周促炎细胞因子参与神经精神性障碍的病理进程提供了可能。外周活化的促炎细胞因子信号可以通过体液和神经两种途径进入中枢。细胞因子参与调节抑郁进程的可能机制包括调节神经递质代谢、影响神经内分泌功能等。近年来，随着抑郁症细胞因子学说的发展，犬尿氨酸通路（kynurenine pathway,KP）在这一学说中的作用日益凸显。在哺乳动物，与抑郁密切相关的神经递质5-羟色胺（5-HT）的合成原料色氨酸有两条不同的代谢通路，分别是KP和5-HT通路，其中90%的色氨酸通过KP代谢。吲哚胺2,3双加氧酶（indoleamine2,3-dioxygenase,IDO）是KP通路的第一个限速酶，将底物色氨酸催化生成犬尿氨酸。IDO是炎症诱导型酶，可被多种炎性因子所诱导，且细胞因子间具有协同效应。其中IFNγ和TNFα是其强诱导剂。机体免疫系统激活引起细胞因子的过度分泌，诱导激活IDO，使色氨酸沿KP通路代谢增加，从而竞争性拮抗5-HT的生物合成过程，影响中枢神经传递。因此，IDO在抑郁症细胞因子学说中的地位至关重要。但IDO是否介导慢性应激诱导动物抑郁行为的发生尚不明确。本研究将立足于抑郁症的细胞因子学说，从炎症的角度探讨慢性应激诱发动物抑郁行为的可能机制。首先建立慢性应激性小鼠抑郁模型，在此基础上应用药物拮抗应激诱导的炎症反应，观察阻断炎症对应激所致小鼠抑郁行为的改善作用，评价炎症在介导应激诱发抑郁行为中的作用。其次，进一步观察慢性应激对炎症诱导型酶IDO的影响，同时观察IDO特异性拮抗剂1-MT对应激所致小鼠抑郁行为的改善作用，评价IDO在介导应激诱发抑郁行为中的作用地位。再次，在上述研究基础上，寻找与抑郁发病密切相关的特异性细胞因子，并从IDO的角度初步阐明其介导抑郁发病的可能机制。具体内容包括：1.慢性应激性抑郁动物模型的建立与鉴定。采用慢性不可预见轻度应激性抑郁小鼠模型，将心理性、生理性应激方式交替联合应用，最大程度地模拟人类所处的复杂环境。通过糖水偏爱测试、强迫游泳实验、悬尾实验检测小鼠抑郁行为，并在此基础上给予小鼠临床抗抑郁药氟西汀，观察小鼠抑郁行为的改善效果，综合评价建模是否成功。实验结果显示，慢性轻度应激使小鼠体重增长减缓；使小鼠糖水偏爱分数时间依赖性降低。应激作用1、2周后，与对照组相比，糖水偏爱分数无明显改变，但第2周时出现下降趋势；应激作用3、4周后，糖水偏爱分数显著降低。氟西汀可以明显阻断慢性应激引起的小鼠糖水偏爱降低。应激作用4周后，小鼠强迫游泳实验和悬尾实验中的不动时间也明显延长。结果表明，应激性抑郁模型建立成功。2.炎症在慢性应激诱发抑郁行为中的作用。基于应激与免疫激活、免疫激活与抑郁症的关系，我们推测免疫激活引起的炎症反应可能参与调节慢性应激诱发抑郁行为的过程。应激的同时，给予小鼠广谱抗炎药米诺环素，观察用药不同时间对动物抑郁行为的改善效果。结果发现，慢性应激作用2周后，小鼠糖水偏爱、不动时间均无明显改变，米诺环素对小鼠上述行为也无明显影响；慢性应激作用3周后，与对照组相比，小鼠糖水偏爱分数明显降低，强迫游泳实验中的不动时间明显延长，且米诺环素可以明显阻断应激引起的上述抑郁行为；用药4周后，结果与第3周时相似，米诺环素可以明显阻断应激引起的小鼠抑郁行为，包括糖水偏爱分数的降低和不动时间的延长。为了进一步确证炎症在介导慢性应激诱发动物抑郁行为中的作用，我们对给药组动物停止用药，而应激刺激持续，观察动物抑郁行为是否会出现。结果发现，停药2周后，在应激的持续作用下小鼠出现糖水偏爱降低和不动时间延长的抑郁行为表现。同时，我们还观察了慢性应激及米诺环素对小鼠外周和中枢炎症反应的影响。结果显示，慢性应激使小鼠外周促炎细胞因子的蛋白表达明显升高，包括TNFα、IL-1β、IL-18；使中枢皮层内TNFα蛋白表达水平明显升高，海马内TNFα蛋白表达下降，但对中脑内TNFα无明显影响；对IL-1β、IL-18、IL-6、IFNγ等促炎细胞因子在皮层、海马、中脑内的蛋白表达均无明显影响。米诺环素可以明显拮抗应激对中枢内TNFα蛋白表达的调节作用。慢性应激暴露可明显激活中枢皮层内的小胶质细胞，而米诺环素可以明显阻断慢性应激引起的促炎细胞因子升高和小胶质细胞活化。另外，通过尼氏染色还发现，慢性应激对皮层神经元具有明显的损伤作用，使神经元尼氏小体溶解丢失，而米诺环素可以阻断应激效应，对皮层神经元具有明显的保护作用。3. IDO在慢性应激诱发抑郁行为中的作用。前面实验结果发现，慢性应激使小鼠外周及中枢免疫系统激活，TNFα等促炎细胞因子表达上调。IDO是炎症诱导型酶。我们推测慢性应激可能通过触发免疫炎性反应激活IDO，影响神经递质的代谢，从而诱发动物抑郁行为。由此，我们观察了应激对IDO表达及其活性的影响。结果发现，慢性应激使IDO mRNA表达及活性明显增强。在此基础上运用IDO特异性拮抗剂1-MT，观察其对应激诱导的动物抑郁行为的改善效应。结果发现，IDO可以明显阻断慢性应激引起的IDO表达、活性增强及小鼠抑郁行为，包括糖水偏爱降低和强迫游泳实验中不动时间的延长。结果提示，IDO在介导慢性应激诱发抑郁行为中起重要作用。4. TNFα参与介导慢性应激诱发抑郁行为。前面实验发现，慢性应激暴露后只有TNFα在外周及中枢蛋白表达均上调，TNFα是IDO的强诱导剂，由此我们推测TNFα在介导应激所致IDO活化及动物抑郁行为中起关键作用。应用TNFα单克隆抗体英夫利昔阻断其生物学效应，观察应激诱导的动物抑郁行为的改善效果。结果发现，英夫利昔可以完全阻断慢性应激引起的TNFα蛋白表达升高，并且可以明显拮抗慢性应激引起的小鼠抑郁行为，包括糖水偏爱分数降低和强迫游泳实验中不动时间的延长。检测中枢IDO表达变化发现，英夫利昔可以明显阻断慢性应激引起的IDO mRNA表达上调。另外，尼氏染色发现，英夫利昔明显抑制慢性应激引起的皮层神经元损伤。结果提示，TNFα可能是介导慢性应激所致动物抑郁行为的核心细胞因子之一，其机制是通过诱导激活IDO。更多还原

【Abstract】 Depression is one of most common psychiatric disorders, which has a complex andmultifactorial aetiology originating from the interaction between environmental andgenetic factors and presents frequent co-morbidity. Stress is well known as one of the mostpathogenic factors of depressive disorder; however, the underlying mechanisms are stillunknown. There is a pressing need for improvements in current depression treatment.Existing antidepressants have many negative effects, and their efficacy is slow and requireschronic treatment. Despite an increasing number of available antidepressants nowadays,there are still many depressed patients do not show improved mood after advanced therapy,or even do not response to any antidepressant therapies. Therefore, to find new potentialtargets and subsequently explore novel antidepressants, it is necessary to clarify themechanisms underlying stress-induced depression.Over the last few decades, researches have made a great progress on the pathogenesis ofdepression. Several theories have been proposed, including the monoamine hypothesis, theneurogenesis hypothesis, the cytokines hypothesis, the hypothalamic-pituitary-adrenal axisdysfunction theory, and so on. The cytokines hypothesis of depression has attractedconsiderable attention ever since it was proposed in1991. It has been and justifieddeveloped by many researchers during the past twenty years. This hypothesis emphases therole of cytokines released by activated immune system in the pathological processunderlying depression.Cytokines are bioactive proteins secreted by activated immune cells. It has beendemonstrated that mice, administrated with LPS acutely or inoculated with BacillusCalmette-Guerin, display depressive behaviors accompanied with the immune systemactivation. Various physiological and psychological stressors can induce the activation ofimmune system and the release of pro-inflammatory cytokines. Exposure to stressedconditions, acute or especially chronic, can induce depressive behaviors in human andanimals. Immune activation and stress exposure have similar ethology effects. Therefore,cytokines-mediated inflammatory reactions may be involved in stress-induced depression.Cytokines functions as part of an integrated network, each of them inducing their ownsynthesis while also inducing the synthesis of other pro-inflammatory cytokines, includingTNFα, IL-1and IL-6, in a “cascade” fashion. In the brain, there is a cytokine networkconsisting of cells, e.g., neurons, microglia and astrocytes, which are able to produce cytokines, express cytokine receptors and amplify cytokine signals. This neural substrateof immune signals under-lies the potent effects of peripheral pro-inflammatory cytokineson pathways involved in the pathophysiology of neuropsychiatric disorders. Peripheralcytokine signals are able to reach the brain through humoral and neural pathways.A number of pathways by which cytokines regulate behaviors have been identified.These pathways include neurotransmitter function, neuroendocrine activity, neuralplasticity and so on. Accompanying with the development of cytokines hypothesis ofdepression, the kynurenine pathway attached much acctention. In mammals, essentialamino acids tryptophan, the synthetic material of neurotransmitter5-HT related todepression, has two metabolic pathways, i.e. kynurenine pathway (KP) and5-HT pathway.The KP is responsible for about90%of the tryptophan metabolism. Indoleamine2,3-dioxygenase (IDO) is a rate-limiting enzyme of KP, breaks down tryptophan intokynurenine. IDO is inflammation inducible, can be stimulated by a number of cytokinesalone or incombination. IDO is activated robustly by IFNγ and TNFα. The activation ofIDO induced by pro-inflammatory cytokines released from activated immune system,enhances the degradation of tryptophan along KP. The breakdown of tryptophan results ina relative decrease in tryptophan bioavailabity, thereby setting the stage for a decrease inserotonin synthesis and serotoninergic neurotransmission. However, it is unclear whetherIDO is involved in stress-induced depression.In this present research, the inflammatory mechanisms underlying chronicstress-induced depressive behaviors were explored based on the cytokines hypothesis ofdepression. Firstly, based on the establishment of the animal model of chronicstress-induced depression, investigated the improvement in stress-induced depressivebehaviors after blocking the inflammation induced by stress exposure withantiinflammatory agents. Secondly, the effects of chronic stress on IDO were evaluated,including its expression and activity. Applying the IDO specific inhibitor1-MT, to estimatethe roles of IDO in stress-induced depression, by investigating the improvement indepressive behaviors induced by stress after the administration of1-MT. Thirdly, found outthe specific cytokines related to depression and the underlying mechanisms regarding toIDO. The details are as follow:1. The establishment and assessment of the animal model of chronic stress-induceddepression. In our present research, the chronic unpredictable mild stress model ofdepression was chosen, combining the physiological and psychological stressors to mimic the complex environments around the human. Applying sucrose preference test, forcedswim test and tail suspension test to evaluate the depressive behaviors induced by chronicstress. Furthermore, the responses of depressed animals to antidepressant fluoxetine wereinvestigated, to further estabilish the animal model. The results showed that chronic stressslows down the increas of weight, and reduces the sucrose preference in a time-dependentmanner. After exposed to stress for one or two weeks, the sucrose preference had nochange compared with control. But the sucrose preference of stress-exposed mice began todecrease at the end of the second week although there was no significance compared withthat of control group. After exposed to stress for three or four weeks, the sucrosepreference decreased remarkablely. Antidepressant fluoxetine blocked the decrease ofsucrose preference induced by chronic stress. The immobility time of mice increasedsignificantly in the forced swim test and tail suspension test, after exposed to stress for fourweeks. These results indicated the establishment of the animal model of stress-induceddepression.2. The roles of inflammation in chronic stress-induced depressive behaviors. Based onthe relations between stress and immune activation, besides the relations between immuneactivation and depression, we proposed that inflammation initiated by immune activationmay be involved in the process of induction of depressive behaviors by chronic stress.Administrated mice with minocycline when exposed them to various stressors, theimprovement in depressive behaviors were observed by blocking inflammatory reactionswith minocycline. The results showed that chronic stress and minocycline had no effect ondepressive behaviors after exposed to stress for two weeks, including sucrose preferenceand immobility time in the forced swim test. After exposed to stress for three weeks, thesucrose preference of stress-exposed mice decreased significantly compared with that ofcontrol. Stress exposure also increased the immobility time remarkablely in the forcedswim test. Minocycline abrogated the depressive behaviors induced by stress exposurecompared with that of the stress group, including surose preference and immobility time.After exposed to stress for four weeks, the results were similar as those of three weeks.Exposure to stress for four weeks induced depressive behaviors obviously, including thedecrease of sucrose preference and increase of immobility time. These above depressivebehaviors were abrogated by minocycline compared with that of the stess group. There wasalmost no chang behviors of the mice treated with minocycline alone. Moreover, to furtherdetermine the roles of inflammation in stress-induced depression, the depressive behaviors of mice treated with minocycline and stress were investigated after the removal ofminocycline. The results showed that after removal of minocycline for two weeks,depressive behaviors were induced by chronic stess compared with that of the controlgroup, including the decrease of sucrose preference and increase of immobility time in theforced swim test and tail suspension test. The effects of chronic stress on peripheral andcentral inflammatory mediators were also determined. As shown, the levels of circulatingproinflammatory cytokines were significantly increased by stress exposure, includingTNFα、IL-1β and IL-18. The protein expression of TNFα of stress-exposed mice increasedremarkablely in the cortex, rather than the hippocampus or midbrain. Stress had no effecton the expression of proinflammatory cytokines in the cortex, hippocampus and midbrainof stress-exposed mice, including IL-1β、IL-18、IL-6、IFNγ. Microglias were activated inthe cortex of stress-exposed mice, indicating by the increased expression of CD11b.Minocycline abrogated completely the increase of proinflammatory cytokines and theactivation of microglia induced by chronic stress. The results of Nissl staining showed thatneurons in the cortex of stress-exposed mice were damaged obviously, indicating by thereduction and dissolution of nissl bodies. Minocycline attenuated the reducdion of nisslbodies compared with that of the stress group, protected neuron from damage induced bystress.3. The roles of IDO in chronic stress-induced depressive behaviors. Our present researchshowed that the peripheral and central immune systems were activated by stress exposure,accompanying by increased expression of proinflammatory cytokines. IDO is aninflammation inducible enzyme. So it was proposed that the activation of IDO, induced byinflammation intiated from stress exposure, account for the dysbolism of neutransmittersand subsequent depressive behaviors. Therefore, the effects of chronic stress on theexpression and activity of IDO were detected. The results showed that both of the mRNAexpression and activity of IDO are enhanced by stress exposure compare with that of thecontrol group. Furthermore, administrated mice with the IDO specific inhibitor1-MTwhen they were subjected to various stressors, and the improvement in depressivebehaviors were detected. As shown,1-MT blocked the enhancement of IDO transcript andactivity, and accompanying depressive behaviors induced by stress, including thedecreased sucrose preference and increased immobility time. These results indicated thatIDO is involved in the pathogenesis of chronic stress-induced depression.4. The roles of TNFα in chronic stress-induced depressive behaviors. The above data demonstrated that TNFα was significantly upregulated in periphery and cortex after stressexposure for four weeks. TNFα can activate IDO robustly. Therefore, we hypothesized thatTNFα might be one of the key biomarkers in this chronic stress-induced depression model.To confirm this hypothesis, we measured the expression of TNFα and depressive behaviorin the presence or absence of TNFα monoclonal antibody infliximab. As shown, the levelof TNFα in stressed mice was back to the control level with infliximab pretreatment.Along with the downregulation of TNFα, infliximab pretreatment abrogated the decreaseof sucrose preference and the increase of immobility time in FST induced by stress. Therewas no significant change in behaviors of the mice treated with infliximab alone. So,blockage of TNFα with infliximab abrogated the depressive behavior induced by stress.The cortex was selected for morphological assessment to investigate the effects ofinfliximab pretreatment on cortex neurons. As shown, normal neurons were observed incontrol group and stress plus infliximab group, while damaged neurons obviouslyincreased in stress group. Furthermore, infliximab pretreatment abrogated this decrease ofundamaged neurons.更多还原