Control of Fear/Defensive Behavior by Brain Derived Neurotrophic Factor

多项研究表明，一侧的皮质和海马体等高级大脑结构与另一侧的所谓防御回路(包括杏仁核、下丘脑和中脑导水管周围灰质)之间的相互作用决定了恐惧/防御行为。此外，这种互动的变化可能会导致精神疾病。 我们假设，脑源性神经营养因子(BDNF)通过调节高级脑结构和低级防御回路之间突触的可塑性来调节这种相互作用。为了验证这一假说并确定调节恐惧/防御反应的机制，我们研究了条件性敲除BDNF的小鼠的行为和相关神经元的变化。 这项研究的目标包括： 1)确定BDNF KO小鼠恐惧/防御行为的可能变化 2)确定这种变化背后的分子和神经机制。 1)确定BDNF KO小鼠恐惧/防御行为的可能变化 在前些年，我们发现在海马区CA3区缺乏BDNF的小鼠对它们的笼子伴侣具有异常的攻击性。在上个财政年度，我们通过记录相同或不同基因型子之间发生攻击的频率和年龄来描述这种家庭笼子攻击。到60日龄时，90%的基因敲除对，60%的野生型和敲除对之间，只有10%的野生型对表现出攻击性行为。在基因敲除小鼠和野生型小鼠之间的所有配对中，基因敲除动物是侵略者。 我们正在测试几个可能解释家庭笼子打架表型的行为假设：1)领土攻击性增加，2)社会识别记忆受损，3)未能建立社会等级。 为了测试领土侵略，我们使用了入侵者测试。令人惊讶的是，对隔离动物的测试显示，不同基因型之间的领土攻击没有差异。这一结果表明，BDNF基因敲除中的攻击行为是一种社会起源，因为动物之间可能存在相互作用的异常。它可能源于社会记忆的丧失，导致另一种动物被认为是入侵者。我们通过确认去卵巢雌性小鼠的正常社会认知来排除这种可能性。为了确定来自另一种动物的感官信号是否在BDNF基因敲除中触发了更强的攻击性，我们将动物关在带有穿孔隔板的笼子里，这种隔板不允许任何打斗，但允许相邻动物之间进行嗅觉、视觉和一些触觉接触。与完全隔离的条件相比，这种住所并没有改变对照动物的攻击性，但在淘汰赛中增加了攻击性，表明它们的过度攻击性行为是由社会经验引发的。我们通过观察主导/顺从行为的发展，继续研究触发攻击的特定线索以及BDNF基因敲除对社会等级形成的可能影响。 2)发现BDNF基因敲除小鼠攻击性增强的神经机制 我们研究的主要目标是确定导致攻击性增加的神经元回路。我们正在测试两种可能解释攻击性增强的假说：外侧隔假说和单胺假说。 由于BDNF的敲除主要局限于海马区CA3，我们正在寻找CA3和其他已知控制攻击性的大脑区域之间联系的可能变化。唯一与CA3直接相连的区域是外侧隔区，它被认为可以抑制攻击行为，并接受CA3的大量投射。因此，我们的工作假设是，CA3细胞缺乏BDNF分泌可能会影响外侧隔区神经元的功能，而后者通常会抑制攻击。在过去的一年里，我们发现BDNFCA3限制性基因敲除小鼠的外侧隔区BDNF水平降低。这一发现表明，BDNF从CA3传递到外侧隔，在那里它可能在局部发挥作用，并抑制攻击。我们正在开发测试这一想法的工具。我们在加州大学旧金山分校与Reichardt博士的实验室开始合作，生产允许有条件地敲除BDNF受体TrkB的小鼠。为了定位负责控制攻击行为的BDNF靶点，我们将使用条件基因敲除的方法去除外侧隔区和海马区CA3区这两个候选区域的BDNF受体TrkB。为了从外侧隔中删除TrkB，我们开发了表达Cre重组酶的高滴度慢病毒，该慢病毒将被注射到Reichardts实验室培育的小鼠的外侧隔中。为了删除海马区CA3区的TrkB，我们使用了Nakazawa博士产生的CA3特异性Cre-line。 单胺假说是基于之前的研究，研究表明BDNF可能会影响血清素的水平，当血清素水平降低时，可能会增强攻击行为。为了确定BDNF基因敲除小鼠的5-羟色胺水平是否发生了变化，在上一财年，我们建立了一个微透析/高效液相系统来测量自由活动动物的单胺水平。目前，测量5-羟色胺水平的实验正在进行中。

Multiple studies suggest that interaction between the higher brain structures like cortex and hippocampus on one side, and a so called defensive circuitry (which includes amygdala, hypothalamus and periaqueductal grey), on the other side, determine fear/defensive behavior. Moreover, changes in such interaction may lead to mental illness. We hypothesized that Brain Derived Neurotrophic Factor (BDNF) modulates such interaction by regulating plasticity in synapses between the higher brain structures and lower defensive circuit. To test this hypothesis and to identify mechanisms regulating fear/defensive responses, we study behavioral and related neuronal changes in mice with the conditional knockout of BDNF. The goals of this study include: 1) Identification of possible changes in fear/defensive behaviors in BDNF KO mice 2) Determining molecular and neuronal mechanisms underlying such changes. 1) Identification of possible changes in fear/defensive behaviors in BDNF KO mice In the previous years we found that mice lacking BDNF in the CA3 area of hippocampus are abnormally aggressive towards their cage partners. During the last fiscal year, we characterized this home cage aggression by documenting frequency and age of its onset between paired littermates of either same or different genotypes. By the age 60 days, 90 % of paired knockouts, 60 % of pairs between wild type and knockout and only 10 % of the wild type pairs show aggressive behavior. In all pairs between knockout and wild type mice, knockout animal was the aggressor. We are testing several behavioral hypotheses which might explain the home cage fighting phenotype: 1) increase in territorial aggression, 2) impaired social recognition memory leading, 3) failure to establish social hierarchy. To test territorial aggression we used intruder test. Surprisingly, testing of isolated animals revealed no differences in territorial aggression between genotypes. This result suggests that aggressive behavior in BDNF knockouts is of a social origin due to possible abnormality in interactions between animals. It might originate from loss of social memory leading to a perception of another animal as an intruder. We excluded this possibility by confirming normal social recognition of ovariectomized female mice. To determine whether sensory cues from another animal were triggering heightened aggression in BDNF knockouts, we housed animals in cages with perforated partition that does not allow any fighting, but permits olfactory, visual and some tactile contact between neighboring animals. Such housing, in comparison with completely isolated conditions, did not alter aggressiveness of the control animals, but increased it in the knockouts, indicating that their excessive aggressive behavior is triggered by social experience. We continue investigation of specific cues triggering aggression and possible effects of BDNF knockout on formation of social hierarchy by looking at development of dominant/submissive behaviors. 2) Finding neuronal mechanisms of increased aggressiveness in BDNF knockout mice The major goal of our study is identification of neuronal circuits responsible for increase aggression. We are testing two potential hypotheses that may explain increased aggression: the lateral septum hypothesis and monoamine hypothesis. Since the knockout of BDNF is mostly restricted to hippocampal area CA3, we are looking at possible changes in connections between CA3 and other brain areas known to control aggression. The only such area that is directly connected to CA3 is lateral septum, which is known to suppress aggressive behaviors and which receives abundant projections from CA3. Thus our working hypothesis is that lack of BDNF secretion from CA3 cells might influence function of neurons in lateral septum that normally suppresses aggression. In the past year we found a reduction in BDNF level in the lateral septum of BDNF CA3restricted knockout mice. This finding indicates that BDNF is delivered form CA3 to lateral septum where it may act locally and suppress aggression. We were developing tool for testing this idea. We initiated collaboration with Dr. Reichardt lab in UCSF on production of mice allowing conditional knockout of BDNF receptor TrkB. In order to locate BDNF target responsible for control of aggressive behavior, we will ablate BDNF receptor TrkB in two candidate areas, lateral septum and hippocampal area CA3, using conditional knockout. To delete TrkB from lateral septum, we developed high titer lentivirus expressing Cre-recombinase that will be injected in the lateral septum of mice generated in Reichardts lab. To delete TrkB in hippocampal area CA3, we are using CA3-specific Cre-line generated by Dr. Nakazawa. The monoamine hypothesis is based on previous studies showing that BDNF may influence levels of serotonin that, when decreased, may enhance aggressive behavior. To determine whether serotonin levels are altered in BDNF knockout mice, during the last fiscal year, we established a microdialysis/HPLC system for measuring levels of mono-amines in freely moving animals. Currently experiments on measuring levels of serotonin are on the way.