Ben-Ari Yehezkel
INMED/INSERM, Parc Scientifique de Luminy, Marseille, France.
Epileptic Disord. 2006 Jun;8(2):91-102.
The construction of the human brain with its 10(15) synapses follows a set of complex developmentally and environmentally regulated steps. A series of sequences have been described that are instrumental, in the sense that a failure of any one of them leads to dramatic, life-long consequences. Hence the importance of determining the sequential maturation of neurons, synapses and cortical maps. It is also important to determine how network-driven events become installed, as neuronal activity intervenes in all of these steps and modulates, for better or worse, the outcome. A fundamental consequence of these sequential events is that any disruption will have very different consequences depending on when it occurs, indeed, "when is as important as what". An obvious aspect of these general features is related to seizures. In fact, the developing brain has both a higher incidence of seizures in human and animal models, and experiences seizures that can produce long-lasting consequences that are also stage-dependent. This seminar and the series of slides presented are an introduction to these issues, summing up several studies made notably by INMED researchers during the last two decades (http://www.inmednet.com). It concentrates on four basic developmental rules: i) the generation by very immature neurons, of very large currents mediated by the activation of receptors in neurons that bear no synapses. This is due to the release of GABA that diffuses to distal sites and acts as a paracrine factor; ii) the excitatory/inhibitory shift of the actions of GABA during development because of a progressive reduction in the intracellular chloride concentration; iii) the sequential formation of GABAergic synapses and networks before glutamatergic ones, implying that, at an early stage, all the excitatory drive will be GABAergic; iv) the presence, at an early stage, of a unique, primitive pattern in all developing structures, this pattern disappears when most GABAergic synapses have shifted to their adult configuration. Several consequences of these sequences are described including: i) a control of neuronal migration by GABA-acting drugs, and the possibility that migration disorders are also generated by environmental factors that include the effects of GABA-acting agents; ii) If GABA excites immature neurons and inhibits adult one, then GABA-acting agents will also produce different effects on the mother and the embryo; iii) early brain oscillations are generated by the periphery and propagate centrally - notably to the sensory-motor cortex, suggesting that peripherally-generated movements may provide an important signal for the formation of cortical maps, in keeping with the importance of embryonic movements; iv) "seizures beget seizures" in the developing brain. This has now been shown in a triple chamber with the two intact hippocampi that we developed, and with which it has been possible to show that only recurrent seizures that include high frequency oscillations can transform the naïve, contralateral hippocampus to an epileptic one that seizes spontaneously. Most interestingly, at an early developmental stage, when GABA excites many neurons and the density of glutamatergic synapses is not sufficiently high, purely glutamatergic seizures cannot lead to long-term consequences, the additional excitatory drive provided by GABAergic synapses is needed. In other words, at that stage, blocking GABA synapses generates seizures, as in adults, but these do not lead to long-term consequences. The mechanisms that underlie these differences is due to the need for high frequency oscillations (> 80 Hz or so), and these can only be generated when GABA synapses are operative in the developing brain: GABA receptor antagonists are ictogenic, but not epileptogenic. To facilitate teaching purposes the paper is published together with supplemental data (as a PowerPoint presentation included in the accompanying DVD), thus allowing an overview of important developmental steps and their implications.
人类大脑拥有10¹⁵个突触,其构建遵循一系列复杂的、受发育和环境调控的步骤。已经描述了一系列有序过程,从某种意义上说,其中任何一个过程的失败都会导致严重的、终身的后果。因此,确定神经元、突触和皮质图谱的顺序成熟至关重要。确定网络驱动事件是如何形成的也很重要,因为神经元活动参与了所有这些步骤,并或好或坏地调节最终结果。这些连续事件的一个基本结果是,任何干扰根据其发生时间的不同都会产生非常不同的后果,实际上,“何时与何事同样重要”。这些一般特征的一个明显方面与癫痫发作有关。事实上,在人类和动物模型中,发育中的大脑癫痫发作的发生率更高,并且经历的癫痫发作会产生长期后果,这些后果也与发育阶段有关。本次研讨会及展示的一系列幻灯片将介绍这些问题,总结了INMED研究人员在过去二十年中所做的几项研究(http://www.inmednet.com)。它集中在四个基本的发育规则上:i)非常不成熟的神经元通过激活无突触神经元中的受体产生非常大的电流。这是由于γ-氨基丁酸(GABA)的释放,它扩散到远端部位并作为旁分泌因子起作用;ii)由于细胞内氯离子浓度的逐渐降低,GABA在发育过程中的作用发生兴奋性/抑制性转变;iii)GABA能突触和网络在谷氨酸能突触之前依次形成,这意味着在早期阶段,所有的兴奋性驱动都将是GABA能的;iv)在早期阶段,所有发育中的结构都存在一种独特的原始模式,当大多数GABA能突触转变为成熟构型时,这种模式就会消失。描述了这些顺序的几个后果,包括:i)GABA作用药物对神经元迁移的控制,以及环境因素(包括GABA作用剂的影响)也可能导致迁移障碍;ii)如果GABA使未成熟神经元兴奋而使成熟神经元抑制,那么GABA作用剂也会对母体和胚胎产生不同的影响;iii)早期脑振荡由外周产生并向中枢传播,特别是传播到感觉运动皮层,这表明外周产生的运动可能为皮质图谱的形成提供重要信号,这与胚胎运动的重要性一致;iv)在发育中的大脑中“癫痫发作引发癫痫发作”。这一点现在已在我们开发的双完整海马体的三室模型中得到证实,利用该模型已能够表明,只有包含高频振荡的反复癫痫发作才能将未发作的对侧海马体转变为自发发作的癫痫海马体。最有趣的是,在发育早期,当GABA使许多神经元兴奋且谷氨酸能突触密度不够高时,纯粹的谷氨酸能癫痫发作不会导致长期后果,需要GABA能突触提供额外的兴奋性驱动。换句话说,在那个阶段,阻断GABA突触会像在成年人中一样引发癫痫发作,但这些不会导致长期后果。这些差异背后的机制是由于需要高频振荡(约>80Hz),而只有当GABA突触在发育中的大脑中起作用时才能产生这种振荡:GABA受体拮抗剂是致痫的,但不是致癫痫的。为便于教学,本文与补充数据(作为随附DVD中的PowerPoint演示文稿)一起发表,从而可以概述重要的发育步骤及其影响。