【作者】 胡华；

【导师】 王立伟； 施慎逊； 诸索宇；

【作者基本信息】 复旦大学 ， 精神病与精神卫生学， 2010， 博士

【摘要】 第一部分CUMS抑郁症模型大鼠行为学及脑葡萄糖代谢[目的]通过运用慢性不可预见性轻度应激(CUMS)方法建立抑郁症大鼠模型,研究CUMS致抑郁症大鼠行为学特征,探讨抑郁症与脑葡萄糖代谢之间的关系。[方法]正常对照组(CON)、抑郁症模型组(CUMS) Sprague-Dawley雄性大鼠各10只。采用国际公认的“慢性不可预见性轻度应激”(Chronic Unpredicatable Mild Stress, CUMS)方法建立抑郁症大鼠模型：实验大鼠持续4周接受不同的应激因子,包括群居饲养、通宵照明、鼠笼倾斜、潮湿垫料、禁水后空瓶放置、频闪光照及白噪音。对所有大鼠(共20只),CUMS实验前、应激开始后的每一周末,在同等条件下共进行5次体重测量和糖水实验。糖水实验的指标包括总液体消耗量、糖水消耗量、纯水消耗量,并计算糖水相对消耗量、纯水相对消耗量及糖水偏好程度(糖水偏好程度=糖水消耗量/总液体消耗量×100%)。糖水偏好常用于评价快感缺乏程度。CUMS建模前、后进行敞箱实验(open-field test),以观察大鼠的行为学特征,根据数码摄像机回放记录,评估行为学指标：①5min内运动总格子数目；②5min内运动的周边格子数目；③5min内运动的中间格子数目；④后肢直立次数；⑤第一分钟内运动格子的总数目;⑥修饰次数；⑦大便粒数。对CUMS组大鼠(共10只)在应激前、后以18氟标记的脱氧葡萄糖(18F-FDG)为示踪剂进行18F-FDGmicro PET扫描。统计方法：糖水实验和体重采用重复测量的方差分析；open-field实验采用non-parametric Friedman ANOVA;运用SPM 5软件对18F-FDG PET图像进行基于体素的配对t检验；相关分析采用Spearman秩相关。[结果]①体重：慢性应激第3周、第4周时CUMS组大鼠体重增加量较CON组大鼠轻度减少(p<0.05)。②糖水实验：应激期间CUMS组大鼠逐渐减少糖水相对消耗量,以应激时间(第1-4周)为主效应,两组间糖水相对消耗量存在差异[F(1,18)=40.72,p<0.001],应激时间与组别之间存在明显交互作用[F(4,72)=7.49,p<0.001]。基线、CUMS建模期间两组对纯水相对消耗量并无显著性差异。应激4周期间,特别是第3周、第4周,CUMS组大鼠糖水偏好明显缺乏(allp<0.001)。③Open-field实验：与CON组相比,CUMS组大鼠第一分钟内运动格子的总数目增多[χ2(10,1)=6.4,p=0.011]; 5min内运动的中间格子数目、后肢直立次数显著减少[χ2(10,1)=9.0,p=0.003;χ2(10,1)=5.4,p=0.02, respectively],修饰次数及大便粒数明显增多[χ2(10,1)=6.4,p=0.011；χ2(10,1)=5.4,p=0.02, respectively]。④18F-FDG micro PET:慢性应激后大鼠听觉皮层葡萄糖代谢显著增高,下丘、中脑导水管周围灰质、梨状皮层、中隔核代谢明显减低(all p<0.01);差异性脑区均出现在左侧,并且听觉皮层葡萄糖代谢增高值与下丘葡萄糖代谢降低值之间呈显著负相关,相关系数为(R=-0.785,p=0.007)。[结论]①运用CUMS方法可以成功建立抑郁症大鼠模型,抑郁症大鼠除了具有抑郁症核心症状,即对糖水偏好明显缺乏外,还具有焦虑、恐惧样行为；②抑郁症与脑功能异常变化有关,以梨状皮层、中脑导水管周围灰质、中隔核、下丘、听觉皮层葡萄糖代谢变化较为显著,且具有明显侧化现象(左侧)。第二部分高脂饮食致脑血管因素模型大鼠行为学及脑葡萄糖代谢[目的]通过运用高脂饮食方法建立脑血管因素模型,研究该模型大鼠行为学特征及脑葡萄糖代谢情况,探讨高脂饮食致高脂血症与抑郁症之间的关系。[方法]正常对照组(CON*)、脑血管因素组(CVF) Sprague-Dawley雄性大鼠各10只。以高脂饲料(饲料配方：蛋黄粉10%、猪油20%、胆固醇2.5%、胆酸钠0.2%、甲巯咪唑0.1%、蔗糖2%、普通饲料65.2%)喂养CVF组大鼠9周。高脂饮食前、实验第9周时,在同等条件下共进行2次糖水实验、体重测量及open-field实验,各实验观察指标同第一部分。实验第9周时对20只大鼠经尾静脉采血5ml,用酶标法测定血清甘油三酯(TG)、血清总胆固醇(TC),用磷钨酸-镁沉淀法测定血清高密度脂蛋白胆固醇(HDLC)、血清低密度脂蛋白胆固醇(LDLC)。另外分别对两组大鼠以18氟标记的脱氧葡萄糖(18F-FDG)为示踪剂进行18F-FDG micro PET扫描。统计方法：糖水实验和体重采用重复测量的方差分析；open-field实验采用non-parametric Friedman ANOVA;两组间血脂指标比较采用student’s t-test;运用SPM5对18F-FDG PET图像进行基于体素的组间t检验；相关分析采用Spearman秩相关。[结果]①体重与血脂：实验第9周时,与CON*组大鼠相比,CVF组大鼠体重显著增高(t=18.824,p<0.001); TC (t=11.76,p<0.001)、LDL (t=6.96,p<0.001)均显著增高,HDL (t=-3.43,p<0.001)显著降低。②糖水实验：两组大鼠在基线时、9周时糖水相对消耗量无差异[F(1,18)=0.896,p=0.356]；与基线纯水相对消耗量相比,CVF组大鼠在9周时纯水相对消耗量明显增多(t=5.747,p<0.001),对糖水的偏好程度明显缺乏(t=-5.565,p<0.002).③Open-field实验：两组大鼠在open-field各项行为学指标上均无显著性差异(all p>0.05)。④18F-FDG microPET：与CON*组大鼠相比,CVF组大鼠丘脑、纹状体脑葡萄糖代谢显著减低(allp<0.01)；且对糖水偏好缺乏程度与丘脑、纹状体局部葡萄糖代谢降低呈显著正相关(R=0.782,p<0.01; R=0.806,p<0.01, respectively)。[结论]①高脂血症大鼠具有抑郁症核心症状；该症状严重程度与丘脑、纹状体脑葡萄糖代谢异常呈正相关；②高脂血症通过影响“边缘系统-纹状体-丘脑”环路功能参与抑郁症发生,且具有侧化现象(右侧)；③高脂血症不仅是脑血管危险因素,也是抑郁症重要危险因素之一第三部分高脂饮食致脑血管因素叠加CUMS抑郁症模型大鼠行为学及脑葡萄糖代谢[目的]成功建立高脂饮食致脑血管因素叠加CUMS抑郁症动物模型,研究叠加模型大鼠行为学特征及脑葡萄糖代谢情况,探讨高脂血症、CUMS在导致抑郁症中所起的作用。[方法]正常对照组(CON)、叠加模型组(DMY) Sprague-Dawley雄性大鼠各10只。高脂饲料(同第二部分)喂养DMY组大鼠9周,继续高脂饮食的同时采用CUMS方法(同第一部分)叠加抑郁症模型4周。对所有大鼠(共20只),实验分组时、高脂饮食第9周时、CUMS开始后的每一周(即实验的第10、11、12、13周)在同等条件下共进行6次糖水实验和体重测量；实验分组时、高脂饮食9周时、CUMS第4周后(即实验的第13周)进行3次open-field实验；在实验第9周时对20只大鼠经尾静脉采血5ml,进行血脂指标测定；各实验观察指标均同第二部分。对DMY组大鼠(共10只)在应激前、后进行以18氟标记的脱氧葡萄糖(18F-FDG)为示踪剂进行18F-FDG micro PET扫描。统计方法：同第一、第二部分。[结果]①血脂：与CON组大鼠相比,第9周时DMY组大鼠TC(t=7.548,p<0.001)、LDL(t=6.175,p<0.001)均显著增高,HDL显著降低(t=-5.300,p<0.001)。②体重：9周时,DMF组体重增加量较CON组显著增多(p<0.01); CUMS第4周(即13周)时DMF组体重增加量较CON组轻度减少(p<0.05)。③糖水实验：与CON组大鼠相比,应激期间各个时点DMY组大鼠对糖水相对消耗量均有显著减少[F(1,18)=104.267, F(1,18)=55.986, F(1,18)=46.084,F(1,18)=61.080, respectively;allp<0.001]；在9周、11周、13周时DMY组大鼠纯水相对消耗量有所增加[F(1,18)=27.72, F(1,18)=16.13, F(1,18)=12.26, respectively; all p<0.001];第9周及4周应激期间,DMY组大鼠逐渐降低对糖水的偏好程度[F(1,18)=28.891,F(1,18)=51.630, F(1,18)=103.794, F(1,18)=103.645, F(1,18)=119.104, respectively; all p<0.001]。④Open-field实验：与CUMS前相比,DMY组大鼠在CUMS后5min内运动总格子数目、中间格子数目、周边格子数目均减少[χ2(10,1)=3.60,p=0.048；χ2(10,1)=10,p=0.002；χ2(10,1)=3.60,p=0.048, respectively];后肢直立次数减少[χ2(10,1)=9.0,p=0.003]；第一分钟内运动格子总数目增多[χ2(10,1)=6.4,p=0.011]；修饰次数及大便粒数增多[χ2(10,1)=9.0,p=0.003;χ2(10,1)=9.0,p=0.03, respectively]。⑤18F-FDG micro PET:叠加模型后大鼠下丘脑、岛叶皮层葡萄糖代谢显著增高,而海马、内嗅皮层葡萄糖代谢显著降低(all p<0.01);并且下丘脑葡萄糖代谢增高值与海马葡萄糖代谢减低值之间呈负相关,相关系数为(R=-0.736,p=0.015)。[结论]①高脂饮食致高脂血症叠加CUMS动物模型可较好模拟高脂血症与抑郁症共病症状,是建立共病动物模型的方法之一；②高脂血症未加控制基础之上一旦接触慢性应激因子会较早出现严重抑郁症状,并伴明显自主神经功能紊乱症状；③高脂血症与抑郁症叠加对与情绪、神经内分泌功能相关中枢葡萄糖代谢影响显著,其中海马功能降低最为明显；另外内嗅皮层、岛叶葡萄糖代谢改变也参与其中,且具有明显侧化现象(左侧为主)；④CUMS与高脂血症所致抑郁症在对行为学、大脑功能的影响中起着协同作用,大脑神经元葡萄糖代谢水平发生改变是二者对机体影响的共同通路之一。更多还原

【Abstract】 PartⅠBehavioral and brain glucose metabolism study in rat depression model induced by Chronic Unpredicatable Mild Stress[Objectives] To investigate behavioral features of depression rats and explore the relationships between depression and brain glucose metabolism by establishing rat depression model induced by Chronic Unpredicatable Mild Stress.[Methods] The study was carried out using twenty healthy male Sprague-Dawley (SD) rats:normal control group (CON, n=10), depression model group (CUMS, n=10). To establish rat depression model, it was designed to meet the procedures of Chronic Unpredictable Mild Stress, which were internationally recognized and were carried out continuously for a total of 4 weeks, including water deprivation, empty water bottle, continuous lighting, cage tilt, paired housing, damp bedding, white noise and strobe light. For all of rats, sucrose test and body weight were measured before beginning the CUMS procedures, and then conducted weekly throughout the CUMS periods under similar conditions. All parameters in sucrose test were recorded, including total fluid intake, sucrose intake and water intake, and then sucrose relative intake, water relative intake and sucrose preference=[sucrose solution intake (ml)/ total fluid intake (ml)]×100 were calculated. Sucrose preference tests were employed to operationally define anhedonia. Besides these, all rats were tested one by one in an open field test at baseline and week 4 to investigate behavioral characters. All rats’ behaviors were recorded using a video camera and were scored manually from a TV screen. Behavioral parameters were as follows:the number of total activity, peripheral activity, central activity in five minutes; the number of first minute activity, the number of grooming and defecation. Ten rats (CUMS, n=10) were given 18F-FDG micro PET scan at baseline and week 4, respectively. Statistical analysis:sucrose test and body weight were analyzed using a repeated measurement ANOVA. Open-field test was analyzed using a non-parametric Friedman ANOVA. We used voxel-based statistical analyses by SPM5 software and used paired t-tests for the comparison between pre-CUMS and CUMS groups. Spearman rank correlation coefficients were calculated.[Results]①Body weight:the body weight gain of the CUMS rats was slightly lower than the CON rats during the last two weeks of the CUMS periods (p<0.05).②Sucrose test:CUMS gradually reduced the relative intake of the sucrose solution. Tests of the main effects for the periods of CUMS (weeks 1-4) showed a difference between the two groups of sucrose relative intake [F(1,18)=40.72, p<0.001]. There was a strong tendency for interaction between group and week [F (4,72)=7.49, p<0.001]. However, no significant differences in water relative intake were found at any point at baseline and during the CUMS periods between two groups. Compared with the CON rats, CUMS rats show no difference at the baseline, but a significantly reduced sucrose preference, especially at week 3 and 4 (all p<0.001).③Open-field test:The CUMS rats exhibited enhanced locomotor activity during the first minute [χ2 (10, 1)=6.4,p=0.011] compared with CON animals. Furthermore, there was a tendency to lower activity in the center squares and reduced rearing for the CUMS rats than the CON rats at week 4[χ2 (10, 1)=9.0,p=0.003;χ2 (10,1)=5.4,p=0.02, respectively]. In comparison to the CON rats, a significant increase in grooming and defecation was observed in CUMS rats[χ2(10, 1)=6.4,p=0.011;χ2(10, 1)=5.4,p=0.02, respectively).④18F-FDG micro PET:using 18F-FDG micro PET, we found that piriform cortex, septal nuclei, inferior colliculus and periaqueductal gray were deactivated while only auditory cortex was activated after 4 weeks of Chronic Unpredictable Mild Stress (all p<0.01). These significant brain region changes were predominantly found in the left hemisphere, and the change of glucose metabolism in auditory cortex has negative correlation with the change value of inferior colliculus (R=-0.785,p=0.007).[Conclusions]①CUMS rats can mimic the core symptoms of depression, an anhedonia to sucrose preference, and anxiety-, fear-like behaviours, which support the face validity of CUMS as an animal model for human depression.②The changes of brain activity are associated with depression and left hemisphere is dominated, especially in many brain regions, such as piriform cortex, septal nuclei, inferior colliculus, periaqueductal gray and auditory cortex, whose glucose metabolism changes are most notable.Part II Behavioral and brain glucose metabolism study in rat cerebrovascular factor model induced by high lipid feed[Objectives] To investigate behavioral features of hyperlipidemia rats and explore the relationships between depression and hyperlipidemia by establishing rat cerebrovascular factor model induced by high lipid feed.[Methods] The study was carried out using twenty healthy male Sprague-Dawley (SD) rats:normal control* group (CON*, n=10), cerebrovascular factor model group (CVF, n=10). To establish rat cerebrovascular factor model, it was designed to high lipid feed for a total of 9 weeks (feed formulation:custard powder 10%, lard 20%, cholesterol 2.5%, sodium cholate 0.2%, thiamazole 0.1%, sucrose 2% and common feed 65.2%). For all of rats, sucrose test, body weight and open-field test were measured before beginning high lipid feed, and then conducted at week 9 under similar conditions. All observed parameters in above tests have been given detailed descriptions in the partⅠ. Besides these, all rats were taken blood 5ml via caudal vein at week 9. Triglycerides (TG) and total cholesterol (TC) were measured using the method of enzyme linked immunosorbent assay, and high density lipoprotein cholesferol (HDLC) and low density lipoprotein cholesferol (LDLC) were measured using the method of phosphotungstic acid-magnesium sedimentation. Twenty rats (CON*, n=10; CVF, n=10) were given 18F-FDG micro PET scan at week 9. Statistical analysis:sucrose test and body weight were analyzed using a repeated measurement ANOVA. Open-field test was analyzed using a non-parametric Friedman ANOVA. Comparison of serum lipids was analyzed using student’s t-test. We used voxel-based statistical analyses by SPM5 software and used student’s t-test for the comparison between CON* and CVF groups. Spearman rank correlation coefficients were calculated.[Results]①Body weight and serum lipids:the body weight gain of the CVF rats was significantly higher than the CON* rats (t=18.824,p<0.001); TC (t=11.76, p<0.001) and LDL (t=6.96, p<0.001) were both significantly increased, however, HDL (t=-3.43,p<0.001) was significantly decreased at week 9.②Sucrose test:no significant differences in sucrose relative intake were found at baseline and week 9 between two groups [F(1,18)=0.896,p=0.356]. Compared with water relative intake at baseline, CVF rats increased water relative intake at week 9 (t=5.747,p<0.001) and significantly decreased sucrose preference (t=-5.565,p<0.001).③Open-field test:no significant differences were found at baseline and week 9 between two groups (all p>0.05).④18F-FDG micro PET:using 18F-FDG micro PET, we found that thalamus and striatum were deactivated of CVF rats, compared with CON* rats at week 9 (all p<0.01). On the other hand, the correlation between the changes of sucrose preference and the changes of brain glucose metabolism in thalamus and striatum was positive (R=0.782,p<0.01, R=0.806, p<0.01, respectively).[Conclusions]①CVF rats present the core symptom of depression, and severity of this symptom has positive correlation with the changes of brain glucose metabolism in thalamus and striatum.②The changes of brain activity in "limbic system-striatum-thalamus" circuit induced by hyperlipidemia are associated with depression and right hemisphere is dominated.③Hyperlipidemia is one of cerebrovascular risk factors, but it is more important to find that hyperlipidemia is also a risk factor in depression.PartⅢBehavioral and brain glucose metabolism study in rat cerebrovascular factor model combined with Chronic Unpredictable Mild Stress model of depression[Objectives] To establish successfully an animal model of cerebrovascular factor model induced by high lipid feed combined with Chronic Unpredictable Mild Stress (CUMS) model of depression, to investigate behavioral features and brain glucose metabolism changes of this model rats, and to explore what roles hyperlipidemia and CUMS played in depression.[Methods] The study was carried out using twenty healthy male Sprague-Dawley (SD) rats:normal control group (CON,n=10), double model factors group (DMY, n=10). To establish rat DMY model, it was designed to high lipid feed for a total of 9 weeks (see the PartⅡ), and then followed by CUMS for 4 weeks (see the PartⅠ). For all of rats, sucrose test and body weight were measured before beginning high lipid feed, at week 9, and then conducted weekly throughout the CUMS periods (week 10,11,12,13) under similar conditions. All rats were tested one by one in an open field test at baseline, week 9 and week 13. Besides these, all rats were taken blood 5ml via caudal vein at week 9 and then serum lipids were measured. All observed parameters in above tests have been given detailed descriptions in the partⅡ. Ten rats (DMY, n=10) were given 18F-FDG micro PET scan at week 9 and week13, respectively. Statistical analysis:for details, see the partⅠand partⅡ.[Results]①Serum lipids:TC (t=7.548,p<0.001) and LDL (t=6.175,p<0.001) were both significantly increased, however, HDL (t=-5.300,p<0.001) was significantly decreased at week 9 compared with the CON rats.②Body weight:the body weight gain of the DMY rats was significantly higher at week 9 (p<0.01) and slightly lower at week 13 (p<0.05) compared with CON rats.③Sucrose test:CUMS gradually reduced the relative intake of the sucrose solution at 4 time points during CUMS periods in DMY rats [F(1,18)=104.267, F(1,18)=55.986, F(1,18)=46.084, F(1, 18)=61.080, respectively; all p<0.001]. DMY rats increased water relative intake at week 9,11 and 13[F(1,18)=27.72,F(1,18)=16.13,F(1,18)=12.26, respectively; all p<0.001]. DMY rats gradually decreased sucrose preference at week 9 and 4 time points during CUMS periods [F(1,18)=28.891, F(1,18)=51.630, F(1,18)=103.794, F(1,18)=103.645, F(1,18)=119.104, respectively; all p<0.001].④Open-field test: compared with pre-CUMS, significant decrease activities in total squares in 5-min, central squares and peripheral squares were observed in DMY rats[χ2 (10, 1)=3.60, p=0.048;χ2 (10, 1)=10,p=0.002;χ2 (10, 1)=3.60,p=0.048, respectively]. Besides these, the number of rearing was reduced[χ2 (10, 1)=9.0, p=0.003], however, the number of grooming and defecation was significantly increased[χ2 (10, 1)=9.0, p=0.003;χ2 (10, 1)=9.0,p=0.03, respectively].⑤18F-FDG micro PET:using 18F-FDG micro PET, we found that hypothalamus and insular cortex were activated while hippocampus and entorhinal cortex were deactivated after 4 weeks of Chronic Unpredictable Mild Stress in DMY rats (all p<0.01). These significant brain region changes were predominantly found in the left hemisphere and the change of glucose metabolism in hypothalamus was negatively correlative with the change of hippocampus(R=-0.736, p=0.015).[Conclusions]①It is a method to mimic hyperlipidemia combined depression by establishing cerebrovascular factor model induced by high lipid feed for a total of 9 weeks, followed by depression model induced by Chronic Unpredictable Mild Stress (CUMS) for 4 weeks.②Under uncontrolled hyperlipidemia, severe depressive symptoms will present more early once exposure to a series of chronic stressors followed by significant autonomic nervous dysfunctional symptoms.③When hyperlipidemia combined with depression, it will significantly change in brain glucose metabolism of many brain regions related with emotion and neuroendocrine, especially in hippocampus. Furthermore, there are significant changes in entorhinal cortex and insular cortex, and left hemisphere is dominated.④CUMS and hyperlipidemia play a synergistic action in the changes of behavior and brain activity of depression.更多还原