【作者】 朱颙颙；

【导师】 徐林；

【作者基本信息】 中国科学技术大学 ， 神经生物学， 2012， 博士

【摘要】 神经信息如何编码是神经科学领域核心问题之一。神经元通过动作电位传递信息,对于如何编码内外界的信息,目前的研究主要聚焦在频率编码(rate code)和时间编码(temporal code)。对于下游的神经元如何识别频率和时间信息,研究表明,神经元通过突触传递信息,而突触具有高度可塑性。经验或活动依赖的突触可塑性被普遍认为是学习记忆的细胞机制。传统可塑性为高频刺激(100Hz)诱导突触传递效能的长时程增强(LTP),低频刺激(1-5Hz)诱导突触传递效能的长时程抑制(LTD),并且它们在机制上都依赖于NMDA受体。这些都支持频率编码规律的理论。放电时序依赖的可塑性(STDP)的发现是时间编码的有力证据。除此之外,在动物theta节律的波峰上通过短簇高频电刺激可诱出LTP,而在波谷处给予可诱出LTD。这提示theta串刺激(TBS,100Hz短簇发放,以theta频率重复)也支持时间编码的规律。TBS普遍被认为模拟了海马锥体神经元的短簇发放模式,当给予Schaffer侧枝刺激会同时激活锥体神经元和中间神经元,而中间神经元通常是长簇发放模式。迄今为止,中间神经元的发放模式对于可塑性的影响仍不清楚。我们模拟中间神经元的发放模式,探索了全新的诱导方案。利用在体清醒动物和离体脑片的系统,在海马SC-CA1通路上通过胞外记录场电位(fEPSP)的方式,发现以5Hz低频重复的100Hz的高频刺激(100Hz,每串15次发放,以5Hz重复60串),我们将其称之为长簇高频刺激(long burst high frequency, LHS),可以诱导出长时程抑制现象(LTDLHs)。改变其中的各项参数不能有效的诱导出LTD。有别于经典的海马CA1区LTD, LTDLHS不能被NMDA受体或代谢型谷氨酸受体(mGluR)的拮抗剂阻断,但可以被GABAA受体的拮抗剂picrotoxin或bicuculline阻断并且毒蕈碱型乙酰胆碱(mACh)受体拮抗剂atropine或scopolamine也可以阻断LTDLHS。这说明LTDLHS的诱导依赖于GABAA和mACh受体。进一步的研究发现,LTDLHS也可以被Tat-GluR23Y阻断,说明LTDLHS依赖于AMPA受体的下膜,提示LTDLHS与经典的LTD具备相似的下游机制。本论文的研究在诱导方式和机理上揭示了一种新型的长时程抑制,并且提示在海马CA1区频率和时间编码能够通过特定的高频发放模式共同起作用。本论文的研究对于我们了解海马的神经信息加工方式和以及相应机制都提供了全新的视角,对于更深层次认识海马的功能有着积极深远的意义。更多还原

【Abstract】 Neural coding is one of the key issues in neuroscience. Classic theory assumes that rate code, the instantaneous frequency of a burst of action potentials (APs), is used by neurons to transmit information. In contrast, temporal code, the precise timing of APs on a millisecond time scale is also believed to be used by neurons.Activity-dependent synaptic plasticity, which is mostly studied in hippocampus, is generally believed to be the cellular basis of learning and memory. Vast evidence has demonstrated that a brief high-frequency stimulation (HFS) or low-frequency stimulation (LFS) at Schaffer collaterals can induce LTP or LTD in the hippocampal CA1area respectively. Most forms of LTP and LTD are dependent on the activation of NMD A receptors for induction and the trafficking of AMPA receptors for expression. These developments are consistent with rate code hypothesis. However, spike-timing dependent pLasticity(STDP) is believed to be consistent with temporal code. Nevertheless, a more physiological protocol, theta burst stimulation (TBS),which is short bursts of stimuli at theta frequency to induce LTP is also consistent with temporal code because either LTP or LTD can be induced by TBS at the peak or trough of the hippocampal field theta waves, demonstrating the rule of the timing between TBS-produced bursts and endogenous field oscillations. TBS is always thought to mimic the short burst firing pattern of the CA1pyramidal neurons. Since Schaffer collaterals can activate both pyramidal and GABAergic neurons of the CA1area and GABAergic neurons often show longer burst. However, the effect of GABAergic firing pattern to synaptic plasticity is still unknown until now.Therefore, activity patterns mimicking the CA1GABAergic neurons were used to stimulate Schaffer collaterals and the field excitatory postsynaptic potentials (fEPSPs) were recorded in hippocampal CA1area of freely moving rat and slices. Here, we discovered a particular pattern, long burst high-frequency stimulations (LHS,15pulses at100Hz,60trains at5Hz) but not other similar protocols was able to induce a novel LTD reliably. The LTD was still induced with the present of the NMD A receptor antagonists AP5or MK801, or the mGlu receptor antagonist MCPG, which as expected are able to block LFS-induced LTD. However, the LTD was completely blocked by antagonists to either of the GABAa or mACh receptors. However, it can be blocked by Tat-GluR23Y peptide, indicating it still needed endocytosis of AMPA receptors, the known mechanisms underlying LTD expression.These findings demonstrate that long burst high frequency stimulation mimicking an activity of the CA1GABAergic neurons, induces a novel form of LTD-LTDLHS-LTDLHS depends on GABAA and mACh receptors and its expression relies on endocytosis of AMPA receptors.Our findings is an important supplement of existing form and mechanisms of the classic synaptic plasticity in hippocampus. Furthermore, it suggests a possibility that rate and temporal code could work together through a particular pattern of high-freuqency burst activity in the local circuit of the hippocampal CA1neurons.更多还原