目的:针对富含亮氨酸的神经胶质瘤灭活蛋白1(LGI1)的自身抗体引起最常见的自身免疫性脑炎亚型,主要累及边缘系统,与癫痫发作和记忆缺陷有关。LGI1及其受体ADAM22是突触蛋白复合物的一部分,该复合物包括几种参与突触前神经递质释放和突触后谷氨酸感知的蛋白质。针对LGI1的自身抗体增加兴奋性突触强度,但是基因破坏LGI1-ADAM22复合物的研究报告了突触后谷氨酸受体介导的反应减少。因此,LGI1自身抗体诱导的突触强度增加的潜在机制仍然难以捉摸,突触前分子对LGI1-突触复合物的贡献仍不清楚。因此,我们研究了介导自身抗体诱导的突触增强的突触前机制。
方法:我们研究了患者来源的纯化多克隆LGI1自身抗体对突触结构和功能的影响,方法是将突触前神经元的膜片钳直接记录与海马培养物和脑切片的超分辨率光镜和电子显微镜相结合。我们还使用结构域特异性患者来源的单克隆抗体鉴定了介导突触前效应的蛋白质结构域。
结果:LGI1自身抗体在高频传播过程中剂量依赖性地增加短期抑郁,与释放概率增加一致。神经传递的增加与突触前钙通道无关,因为突触前Cav2.1通道密度,钙电流幅值,和钙通道门控不受LGI1自身抗体的影响。相比之下,LGI1自身抗体的应用均匀地降低了突触前bouton表面的Kv1.1和Kv1.2通道密度。直接突触前膜片钳记录显示,LGI1自身抗体会导致突触前动作电位明显变宽。在神经元体细胞处分析LGI1自身抗体的结构域特异性作用。通过多克隆LGI1自身抗体和靶向表位结构域的患者来源的单克隆自身抗体诱导体细胞动作电位增宽,但不是富含亮氨酸的重复结构域.
结论:我们的结果表明,LGI1自身抗体降低了突触前束上Kv1.1和Kv1.2的密度,对钙通道密度或功能没有作用,从而扩大突触前动作电位和增加神经递质的释放。这项研究为LGI1自身抗体患者中观察到的神经元过度活跃提供了分子解释。
OBJECTIVE: Autoantibodies against the protein leucine-rich glioma inactivated 1 (LGI1) cause the most common subtype of autoimmune encephalitis with predominant involvement of the limbic system, associated with seizures and memory deficits. LGI1 and its receptor ADAM22 are part of a transsynaptic protein complex that includes several proteins involved in presynaptic neurotransmitter release and postsynaptic glutamate sensing. Autoantibodies against LGI1 increase excitatory synaptic strength, but studies that genetically disrupt the LGI1-ADAM22 complex report a reduction in postsynaptic glutamate receptor-mediated responses. Thus, the mechanisms underlying the increased synaptic strength induced by LGI1 autoantibodies remain elusive, and the contributions of presynaptic molecules to the LGI1-transsynaptic complex remain unclear. We therefore investigated the presynaptic mechanisms that mediate autoantibody-induced synaptic strengthening.
METHODS: We studied the effects of patient-derived purified polyclonal LGI1 autoantibodies on synaptic structure and function by combining direct patch-clamp recordings from presynaptic boutons and somata of hippocampal neurons with super-resolution light and electron microscopy of hippocampal cultures and brain slices. We also identified the protein domain mediating the presynaptic effect using domain-specific patient-derived monoclonal antibodies.
RESULTS: LGI1 autoantibodies dose-dependently increased short-term depression during high-frequency transmission, consistent with increased release probability. The increased neurotransmission was not related to presynaptic calcium channels because presynaptic Cav2.1 channel density, calcium current amplitude, and calcium channel gating were unaffected by LGI1 autoantibodies. By contrast, application of LGI1 autoantibodies homogeneously reduced Kv1.1 and Kv1.2 channel density on the surface of presynaptic boutons. Direct presynaptic patch-clamp recordings revealed that LGI1 autoantibodies cause a pronounced broadening of the presynaptic action potential. Domain-specific effects of LGI1 autoantibodies were analyzed at the neuronal soma. Somatic action potential broadening was induced by polyclonal LGI1 autoantibodies and patient-derived monoclonal autoantibodies targeting the epitempin domain, but not the leucin-rich repeat domain.
CONCLUSIONS: Our results indicate that LGI1 autoantibodies reduce the density of both Kv1.1 and Kv1.2 on presynaptic boutons, without actions on calcium channel density or function, thereby broadening the presynaptic action potential and increasing neurotransmitter release. This study provides a molecular explanation for the neuronal hyperactivity observed in patients with LGI1 autoantibodies.