【作者】 杨晶；

【导师】 胡三觉； 吴南屏；

【作者基本信息】 第四军医大学 ， 神经生物学， 2007， 博士

【摘要】 兴奋性是神经元最重要的基本性质之一。长时期以来尽管人们对兴奋性的表现方式,通道电流机制及其变化规律有相当深入的研究,然而对于神经元在刺激条件下为什么会产生不同的反应,在相同离子电流参于下何以出现多种多样的放电模式,以及噪声对兴奋性的作用等基本问题仍然了解得很少。本项研究通过电生理实验与非线性理论的结合,对几个涉及兴奋性的基本问题展开了探索性研究,以揭示兴奋性的主要非线性特征,深入认识兴奋性的本质过程。第一部分三叉神经中脑核(Mes V)神经元兴奋性的分类与转型兴奋性是指可兴奋细胞受到刺激时产生动作电位(action potential,AP)的能力。传统观念以阈值高低或产生AP的多少衡量兴奋性的高低。1948年Hodgkin在研究甲壳类动物轴突的时候,首次按放电频率与刺激电流强度之间的关系将兴奋性分为三种类型,但未引起足够的注意。80年代以来随着非线性理论与计算技术的发展,理论学家们运用数学模型论证了兴奋性的分类与其动力学机制。然而在实际的神经元,特别是哺乳动物的神经元,是否存在兴奋性分类以及分类形成的机制和内在联系尚不清楚。本项研究对这一问题进行了探试,不仅在哺乳动物的神经元上证实了兴奋性的分类,还发现了兴奋性转型现象。显然随着对兴奋性分类的深入研究,必将超越传统观念的局限性,为进一步探索兴奋性的内在本质过程奠定基础。实验选择幼鼠三叉神经中脑核(Mes V)神经元为标本,在利用红外直视显微镜结合电生理膜片钳(电流钳和电压钳)和药理学技术的基础上展开。主要结果:1、对140个Mes V神经元施加去极化方波刺激,观察到三种兴奋性类型。1类兴奋性神经元(9/140):阈值电流较低,放电前存在潜伏期,放电频率随着刺激电流强度的增强而逐渐增加。2类神经元(55/140):阈值电流较高,首个AP之前没有潜伏期,达到阈值后,AP成串出现,放电频率对刺激电流的强度相对不敏感。3类神经元(76/140):不仅阈值电流较高,通常只产生1个AP,高强度刺激( > 1000 pA)才产生3-5个AP。2、去极化斜波刺激:1类神经元的放电频率随着膜电位的逐渐去极化由低到高,呈线性增加,动作电位之前没有出现阈下膜电位振荡(SMPO)。2类神经元在膜电位去极化的过程中出现了SMPO,放电频率相对恒定。3类神经元没有出现AP。3、50μM 4-AP(低剂量选择性阻断I4-AP)分别使2类和3类兴奋性神经元转变为1类神经元。4、2μM riluzole (一种持续Na+通道(INaP)非特异性阻断剂)能使2类神经元转变为3类神经元,对3类神经元却无明显影响。5、当4-AP阻断I4-AP使2类兴奋性神经元转型变为1类时,随后再施加riluzole,此时阻断INaP不是恢复I4-AP,却可使1类兴奋性返回到2类兴奋性。当riluzole使2类兴奋性神经元转变为3类,继之再施加4-AP,此时阻断I4-AP不是恢复INaP,又可使其返回2类。我们把上述兴奋性分类转变的现象称之为兴奋性转型(transformation of excitability)。6、在上述神经元兴奋性转型的过程中,可出现具有不同类型混合特征的复杂放电模式,我们称之为中间态(intermittent phenomiral)。7、钾通道阻断剂TEA(20 mM),Ih电流特异阻断剂ZD 7288(10μM)以及钙通道阻断剂Cd2+(300μM)对兴奋性的类型没有产生显著影响。8、2类和3类神经元的I4-AP在激活电压位置、V1/2和k三方面存在显著性差异。当去极化达到并且超过-56 mV以后,I4-AP在2类兴奋性神经元中的平均幅值明显地小于3类神经元中的平均幅值。2类和3类神经元的INaP在上述三方面以及平均幅值上没有显著性差异。9、Mes V神经元数学模型近似地模拟了实验中所观察到的兴奋性分类和转型现象,揭示了兴奋性分类与动力学分岔之间的关系(数学模型工作由刘一辉博士完成,详见其博士论文)。主要结论:1、首次证实Mes V神经元存在三种兴奋性类型,并且用去极化斜波方式能简便地检测兴奋性类型。2、发现了神经元兴奋性转型现象,揭示了介导Mes V神经元兴奋性转型的主要离子通道机制是I4-AP and INaP相对比例的动态变化。3、在兴奋性转型过程中可出现“中间态”放电模式。4、兴奋性分类与转型的实验结果得到了神经元数学模型的理论验证。第二部分噪声提高Mes V神经元兴奋性及其与共振性质的关系近年神经科学与非线性科学交叉研究的一项重大进展,是发现适当强度的噪声可以增强感受器对弱信号的检测能力。理论模型研究进一步提示神经元可以通过随机共振或自共振方式接受噪声的影响。实际上,神经元生存在噪声环境,如离子通道的随机开闭,递质的随机释放等。然而在脑内,神经元的活动是否以及如何接受噪声的影响却了解很少。部分神经元存在对输入频率选择性放大的性质,即频率共振(frequency resonance)性质,阈下膜电位振荡被看作是频率共振放大的表现。本研究通过观察与比较噪声对上述三类兴奋性神经元的作用,揭示噪声对兴奋性的作用及其与共振性质的关系,进而为深入阐明噪声在神经系统的作用机制提供了新的线索。实验材料与方法除与第一部分相同外,增加了高斯白噪声和ZAP(一种幅值恒定而频率随时间连续线性增加的正弦电流)两种细胞内注入刺激方式。主要结果:1、2类兴奋性神经元出现了双重的电压依赖性共振,即存在高低两个共振主频:其中去极化达-50mV时共振主频为75.4±2.11 Hz,超极化达-70mV时的共振主频为5.46±0.31 Hz。2、50μM 4-AP可使高频共振主峰消除,但不影响低频共振主峰。10μM ZD 7288可使低频共振主峰消除,但不影响高频共振主峰。3、1类神经元上没有观察到明显的共振现象。4、部分(13/21)3类神经元上也存在高低两个共振主频,主频位置和共振电流与2类神经元近似。5、在2类兴奋性神经元,一定强度范围内的噪声,能够降低出现SMPO的膜电位水平,其降低程度随噪声强度而增大。6、在2类兴奋性神经元,一定强度范围内的噪声降低产生首个AP的膜电位水平。当噪声强度增大时不仅AP阈值降低程度增大,而且放电数目增加,放电节律性也相应增强。当达到最佳噪声强度时(range: 150-250,n = 10),放电节律性与噪声强度之间的关联性(β值)最佳;随后,噪声继续增大,放电数目虽然增多,放电序列却杂乱无序,关联性下降。7、riluzole抑制2类神经元SMPO的同时也消除噪声对神经元的兴奋性作用。8、噪声对1类(n = 12)和3类(n = 20)神经元没有产生明显的影响。主要结论:1、2类兴奋性神经元分别在去极化与超极化水平存在两个共振主频,分别由I4-AP和Ih介导。2、噪声能够通过降低出现SMPO和放电阈值的膜电位水平来提高2类神经元的兴奋性,并改善放电的节律性。3、噪声对2类神经元兴奋的调变作用依赖于SMPO的存在,其通道电流机制尚需进一步探讨。第三部分坐骨神经“起步点”反应性与动力学状态的关系长期以来,人们对处于静息背景状态下的神经元对刺激的反应规律有了相当充分的了解,然而,对处于活动(放电)状态下的神经元的刺激反应规律却了解得较少。本研究室的前期工作以及相关研究表明神经元的反应依赖于放电节律的动力学状态。加周期分岔过程是可兴奋细胞中的一个非线性现象。本研究力图判定当神经元活动处于分岔不同时相的动力学状态时,其反应性(也称兴奋性)可能发生的变化。实验采用单纤维记录技术,引导大鼠受损坐骨神经“起步点”的放电活动,改变细胞外液中Ca2+离子浓度,使“起步点”放电活动分别处于不同的动力学时相。局部施加方波电场刺激。主要结果:1、在受损坐骨神经的A类纤维上成功复制了加周期分岔的全过程,并可区分临近分岔点与远离分岔点等不同时相。2、在远离分岔点时相对神经纤维施加去极化方波电场刺激,随着刺激强度的增加,放电频率近似线性增加,放电模式基本不变;在临近分岔点时相,施加与上述相同的去极化刺激,当达到一定强度时放电频率显著增加,伴有放电模式的转化。临近分岔点的刺激-反应曲线较远离分岔点的曲线显著上移。3、在临近分岔点时相,施加相同强度的超极化刺激仅轻度降低放电频率,也没有出现放电模式的转化。主要结论:实验结果提示当神经纤维异位放电“起步点”处于邻近分岔点时相,对电场刺激的反应较远离分岔点时相的反应敏感,并具有方向选择性,我们称之为“临界敏感”现象。该现象进一步支持本研究室关于神经元的反应性依赖于放电节律动力学状态的假说。更多还原

【Abstract】 The most significant feature of neurons is their excitability. For a long time, the expressional manner, mechanism of ion currents, and its changing rule of excitability have been detected quite well, while little is known about the following questions: the reason that different responsiveness of neuron under the same stimulus; the reason that neuron could exhibit various firing patterns within the same ion currents; the effect of noise to excitability, et al. We investigate some basic problems about excitability through combining electrophysiological experiments with nonlinear theory. The aim is try to expose some main nonlinear characteristics of excitability, in order to well realize the essential process of excitability.First part: Classification and Transformation of Excitability in Mesencephalic V NeuronsExcitability is the ability that excitable cells are able to produce the action potential (AP) when stimulated. Traditionally, we estimated it just through the intensity of stimulus threshold and the number of AP. It is Hodgkin (1948) who first classified the excitability into three classes according to the relationship of firing frequency and applied current intensity, in a study of crustacean axon, but didn’t bring enough attention. In 1980s, with the development of nonlinear theory and computation technology, theory scientist demonstrated the excitability classification and its dynamic mechanism by using some mathematical models. In real neurons, especially mammal’s neuron, however, it is still not clear about the excitability classification and the relation among them, and the mechanism of them. Our research not only finds that there existed the excitability classification in mammal’s neuron, but also testifies the transformation among them. Obviously, with the further investigation to excitability classification, we will transcend the limitation of traditional conception and could establish the basis to advance explore the intrinsic essence of excitability.In our study, the mesencephalic V (Mes V) neurons slices from neonatal rats were adopted as specimen. Whole-cell recording by infrared visual patch clamp was combined with pharmacological technique.Main results:1. 140 Mes V neurons in our experiments were divided into three classes when giving the constant-current depolarize stimulus. Class 1 excitability neurons (9/140): low injected current produce spike discharges with a latency and the firing frequency increased continuously with the injected current intensity increase. Class 2 excitability neurons (55/140): relatively high injected current produce high frequency spike train without latency. The firing frequency was less sensitive to the injected current intensity. Class 3 excitability neurons (76/140): only fired a single AP at the relatively high injected current. Very high stimulation intensity (over 1000 pA) could evoke 3 to 5 action potentials.2. Ramp depolarize stimulus: the firing frequency of Class 1 neurons increased lineally from low to high with the depolarizing membrane potential. And before the AP, there is no subthreshold membrane potential oscillation (SMPO). Class 2 excitability neurons exhibited SMPO and the firing frequency remained relatively constant even though the magnitude of the injected current continually increased. Class 3 neurons didn’t show any spike under the ramp injected current.3. 50μM 4-AP (low concentration could selectivity block I4-AP) transformed originally Class 2 and Class 3 excitability neurons into Class 1 excitability behavior.4. 2μM riluzole (a kind of non- specific blocker of persistent sodium current (INaP)) abolished the spike discharges and transformed originally Class 2 into Class 3 excitability neurons. But did not made obvious change to the originally Class 3 excitability neurons.5. First, blocked I4-AP with 4-AP could transform originally Class 2 into Class 1, then after additional riluzole application (means blocking INaP instead of resuming I4-AP), the spike discharges of this neuron transformed from Class 1 back into Class 2. Riluzole transformed the originally Class 2 excitability neuron into Class 3 type and additional 4-AP (means blocking I4-AP instead of resuming INaP) could restore the spike discharges like Class 2. We named these changes of the excitability classification as transformation of excitability. 6. During the process of the transformation of excitability, the cell exhibited mixed characteristics of different classes. We termed it as“intermediate phenomenon”.7. 20 mM tetraethylammonium (TEA) (blocker of K+ current), 10μM ZD 7288 (specific blocker of Ih), and 300μM Cd2+ (blocker of Ca2+ current) did not transform neuron excitability class.8. The difference of activation threshold, V1/2 and k value were statistically significant between the Class 2 and Class 3 neuron. The mean amplitude of I4-AP was significantly smaller in Class 2 than in Class 3 type neurons when depolarized to and above -56 mV. There was no significant difference of the above kinetic characteristics and mean amplitude of current between two class neurons.9. The mathematics model simulations of Mes V neuron replicate the classification and the transformation of excitability that observed in experiments, indicate the relationship between excitability classification and the dynamic bifurcation (the work of mathematics model was finished by Doc. Liu Yihui).Main conclusions:1. For the first time testify that there exist three excitability classes in Mes V neuron, and could use ramp function to check it conveniently.2. Detect the transformation of excitability. The dynamic change of the relative amplitude proportion of I4-AP and INaP is the crucial mechanism in deciding which class of excitability behavior a Mes V neuron exhibits. 3. The neuron could exhibits“intermediate phenomenon”during the process of the transformation of excitability.4. The classification and the transformation of excitability were testified by using neuronal mathematics model.Second part: Noise increase the excitability and the relationship with its resonant character in Mes V neuronRecently, one of the important developments of overlapping research in the non-linear science and neuroscience is that some degree of noise can play a constructive role in the detection of weak signals. Moreover, theoretical model research indicates that neuron could take in the effect of noise through stochastic resonance (SR) or autonomous SR (ASR). Actually, neuron lives in an environment full of noise, such as the stochastic open and close of the ion channels, and the stochastic release of the neurotransimitters. In brain, however, little is known about whether and how the neuron accepts the effect of noise.Part of neurons exist the character that could selective amplify the input frequency, i.e., character of frequency resonance, and SMPOs were regarded as the exhibition of amplificatory frequency resonance. Through observing and comparing effect of noise to the three classes’excitability neuron, this research indicates the relationship between the effect of noise and the character of resonance, and could provide a new clue for clarifying the effect of noise in nerve system.Besides the same material and methods as the first part, we also add two kinds of stimulus way - Gauss white noise and ZAP (some kind of sine wave with constant amplitude and linear-increased frequency)Main results:1. All of the Class 2 neuron showed a dual voltage dependence electrical resonance, that means there have two forms of resonance frequency: the high resonance frequency were 75.4±2.11 Hz when the membrane potential was depolarized to -50 mV; the low resonance frequency were 5.46±0.31 Hz when the membrane potential was hyperpolarized to -70 mV.2. 50μM 4-AP could abolish the high frequency resonant peak, but caused little or no change of the low frequency resonant peak. 10μM ZD7288 could abolish the low frequency resonant peak, but caused little or no change of the high frequency resonant peak.3. There is no voltage dependence of resonance behavior in a Class 1 neuron.4. Parts of Class 3 neuron (13/21) also have two forms of resonance frequency, which are similar with Class 2 both in resonance frequency and resonance currents.5. In Class 2 neuron, noise with a certain range of intensity could reduce the membrane potential level of SMPO, the degree of such reduction was increase with the increment of noise intensity.6. In Class 2 neuron, noise with a certain range of intensity could reduce the membrane potential level of the first AP. With the increase of noise intensity, not only the degree of such reduction, but also the number and rhythm of spike were increase. The connection (βvalue) between spike rhythm and noise intensity is best when the intensity of noise reach the best(range: 150-250,n = 10). Then with the noise continuously increased, though the spike number increased, the rhythm of spike disordered and the connection decreased.7. Riluzole could inhibit the SMPO of Class 2, and eliminate the excitability effect that noise made.8. Class 1 (n = 12) and Class 3 (n = 20) neurons do not show any obvious responses to noise.Main conclusions:1. In the level of depolarized and hyperpolarized membrane potential, Class 2 neuron have two forms of resonance frequency, and respectively mediated by I4-AP and Ih.2. Noise could increase the excitability of Class 2 neuron by means of reducing both the membrane potential levels of SMPO and spike threshold.3. The effect that noise increasing the excitability of Class 2 neuron depends on the existence of SMPO, but the ion channels mechanism of it still needs more investigate.Third part: Relationship between responsiveness and dynamic state on sciatic pacemakerFor some time, the responsiveness of neurons has been detected under a resting background condition, while little is known about the response rule when neurons are stimulated under an active background in which neurons may display many kinds of firing patterns. Our previous works and other relative researches indicate that the responsiveness of neuron may depend on the dynamic states of its firing pattern. Period-adding bifurcation was a nonlinear phenomenon in excitable cells. Our research tries to determine the possible change of its responsiveness (also called excitability) when the dynamic states of neuron are belong to different time physics during bifurcation.We guide the discharge from the pacemaker of CCI model (Chronic Compression of Sciatic model) in rat using the single fiber recording technology, then change the calcium concentration of the solution perfusing the pacemaker, and make it possible for artificially maintained different dynamic states within period-adding bifurcation. Finally, we give the square-wave electrical field stimulus.Main results:1. Successfully recorded the whole process of period-adding bifurcation form A-type fiber in CCI model, and could distinguish different time physics during the bifurcation, such as near or far from the bifurcation point.2. In the time physic far from the bifurcation point, with the increase of the intensity of excitatory stimulus, the firing rate increased in an approximately linear manner and no firing pattern transition was observed. While in the time physic near the bifurcation point, the firing rate increased markedly higher accompanied with the transition of firing pattern when the intensity of excitatory stimulus remained the same. The stimulus-response of the time physic near the bifurcation point shifted upward significantly compared to that of the time physic far from the bifurcation point.3. Inhibitory stimulus with the same intensity, however, decreased the firing rate slightly without the transition of firing pattern in the region near the bifurcation point. Main conclusions:These results suggest that the responsiveness in the time physic near the bifurcation point is more sensitive than that in the time physic far from the bifurcation point, which we named“critical sensitivity”, and this has directional selectivity. This phenomenon further supports our hypothesis that the responsiveness of neuron may depend on the dynamic states of its firing pattern.更多还原