内容来源：https://www.quantamagazine.org/how-the-brain-balances-excitation-and-inhibition-20250929/

内容总结：

【本报专稿】人类大脑如何维持兴奋与抑制的精密平衡，一直是神经科学领域的核心议题。最新研究表明，这种动态平衡不仅关乎大脑正常功能，更与癫痫、自闭症等神经系统疾病密切相关。

人脑拥有近千亿个神经元，其中绝大多数可归为两类：兴奋性神经元通过释放谷氨酸促使其他神经元放电，如同电路中的"启动开关"；抑制性神经元则通过释放γ-氨基丁酸（GABA）阻止放电，相当于"安全断路器"。二者协同工作，确保神经信号在正确的时间抵达正确的位置，使人能够完成抓取物品、回忆信息等日常行为。

值得注意的是，尽管大脑皮层中兴奋性神经元数量占优，但在哺乳动物进化过程中，抑制性神经元的多样性和数量持续增加。最新《自然》杂志研究证实，抑制性神经元通过选择性降低放电频率，能增强小鼠对食物位置的记忆，这表明其在学习记忆过程中扮演着比既往认知更积极的角色。

研究还发现，大脑中存在着作用更持久的第三类神经元——神经调节神经元。这类细胞通过释放多巴胺、血清素等神经调节素，以"慢波信号"方式大面积影响脑区活动。例如去甲肾上腺素能强化情绪相关记忆的神经连接，这解释了为何情感体验往往令人记忆深刻。

科学家指出，当兴奋与抑制的平衡被打破，可能导致灾难性后果：兴奋过度会引发癫痫发作，而兴奋不足则与自闭症等疾病相关。随着脑细胞图谱绘制的推进，学界正深入探索神经网络失衡机制，相关突破将为神经系统疾病治疗开辟新途径。

"这项研究不仅关乎个体生活质量，更对整体社会发展具有重要意义。"不列颠哥伦比亚大学神经科学家马克·塞姆布劳斯基强调。目前，全球科研团队正在小鼠视觉皮层测绘等领域持续攻关，试图揭示数万次神经元放电如何汇聚成人类思想与意识的奥秘。

中文翻译：

大脑如何平衡兴奋与抑制

19世纪末至20世纪初，神经解剖学家圣地亚哥·拉蒙-卡哈尔用画笔勾勒出枝杈与螺纹、棘突与网状结构——这些如今享誉学界的绘图首次揭示了哺乳动物大脑基本构件「神经元」的独特性与多样性。此后百余年，后继者孜孜不倦地对这些细胞进行计数、追踪、鉴定、标记与分类。如今神经元分类体系已繁复得令人目眩，常以色彩斑斓的复杂脑细胞图谱呈现。通过这些目录，我们可以按功能将神经元归类：区分支配运动的运动神经元、感知视觉的感觉神经元，或是估算数量的计数神经元；也可以根据轴突长短、位于海马体或嗅球等解剖特征进行划分。但绝大多数神经元无论功能、形态或位置如何，都归属于两大基本类型：促使其他神经元放电的兴奋性神经元，以及阻止其他神经元放电的抑制性神经元。

维持恰当的兴奋抑制比例对大脑健康协调至关重要。英属哥伦比亚大学神经科学家马克·塞姆布拉夫斯基指出：「任何方向的失衡都可能引发严重后果」，导致神经系统疾病。过度兴奋可能诱发癫痫发作，而兴奋不足则与自闭症等病症相关。神经科学家正致力于揭示这两类细胞的工作机制，特别是它们如何与影响其行为的第三类稀有细胞相互作用。这些发现最终或能帮助重建因正常衰老等原因失衡的神经环路。

平衡之道
兴奋性与抑制性神经元以相似方式工作。它们大多释放称为神经递质的化学信使，这些物质穿越名为突触的微小间隙，与下游神经元上杯状受体蛋白结合。区分二者的关键在于所释放的神经递质类型。

大脑中的兴奋性神经元在激活时几乎专一性释放谷氨酸。谷氨酸会引发大量正离子涌入神经元，升高其内部电压，促使动作电位产生——这种强烈的电脉冲沿神经纤维传导，促使神经元释放自身分子群继续传递信号。相反，抑制性神经元激活时释放名为γ-氨基丁酸（GABA）的神经递质，引导带负电离子涌入或带正电离子逸出相邻神经元。内部电压降低使得下游神经元无法放电。加州大学旧金山分校神经科学家托马什·诺瓦科夫斯基解释：「抑制性神经元犹如电路断路器」。

这种「停止-通行」机制构成了大脑的高速公路系统，确保信号在正确时间抵达正确位置，使你能抓起桌上的苹果、哼唱最爱的旋律或记起手机存放处。在哺乳动物大脑皮层中，兴奋性神经元数量远超抑制性神经元。但在哺乳动物脑演化过程中，抑制性神经元不仅数量增加，更呈现功能分化，暗示它们在高级功能中扮演关键角色。

佐治亚理工学院与埃默里大学神经科学家安纳贝尔·辛格指出，抑制性神经元「常被归为辅助角色」，这或许源于兴奋性神经元更易研究。例如海马体的兴奋性位置细胞会在动物处于特定方位时激活，这种激活对其他细胞的兴奋作用清晰可辨。但抑制性神经元「在所有区域都频繁放电，难以界定其响应对象」，我们既不知它抑制何种信号，与之连接的细胞也不会产生自身放电反应。

然而研究正逐步揭示抑制性神经元的激活机制与时机。辛格团队在《自然》期刊的最新研究发现，当小鼠接近食源位置时，抑制性神经元通过选择性降低放电频率来帮助它们快速学习并记忆觅食点。辛格解释：「随着小鼠靠近目标位置，抑制性神经元放电频率下降会增强目标信号，从而促进对关键位置的学习」。这表明它们在记忆过程中发挥着比既往认知更主动的作用。

艾伦研究所神经科学家努诺·马萨里科·达科斯塔补充道，既往观点认为抑制性神经元活动更具普适性，实施「地毯式抑制，阻断其轴突周围所有信号」。但达科斯塔团队通过「微米计划」（旨在精细绘制小鼠视觉皮层1立方毫米区域的大型项目）发现，抑制性神经元对抑制对象具有高度特异性。

大脑环路由以不同方式对话的抑制性与兴奋性细胞混合构建。例如某些抑制性细胞倾向将信号传递至树突细枝，另一些则靶向神经元胞体，还有些协同抑制特定细胞群。这些动态组件通过尚未完全阐明的机制交织融合，最终形成我们的反应、思维、记忆与意识。

但神经元通信速度远超其产生的认知效应，能在数十毫秒内传递信号。塞姆布拉夫斯基指出：「神经递质作用极快，但行为与认知过程却相对缓慢」，这种明显悖离是「大脑的核心谜题之一」。

第三维度
另一类细胞或许能解决这个时序难题。

神经调节性神经元在大脑中更为稀有，它们以较慢节奏工作，但作用范围更广、持续时间更长。这类细胞不局限于通过突触向单个神经元传递分子，而是将被称为神经调质的分子群释放至整个区域，与众多突触相互作用。它们释放的多巴胺、血清素等物质会引起兴奋性或抑制性神经元内部变化，调节其放电倾向。塞姆布拉夫斯基描述：「它们创造了一种缓慢的潜流信号，对大脑快速动态施加重要改变」。例如神经调质去甲肾上腺素在情绪记忆形成中作用显著，其释放能增强神经元间连接，巩固记忆痕迹，从而「引导特定情感体验进入记忆轨道」。

尽管兴奋性、抑制性、神经调节性这三大基本属性为我们理解神经元运作提供了框架，但其角色存在模糊地带。例如某些兴奋性和抑制性神经元兼具神经调节功能；少数与情绪相关的神经元能同时包裹释放GABA与谷氨酸，兼具兴奋与抑制特性；在慢性压力等条件下，部分神经元还能实现兴奋与抑制属性的转换。

随着脑细胞图谱持续更新，研究证实广大神经元类别内部存在巨大多样性，但所有细胞都参与着兴奋与抑制的节律律动。塞姆布拉夫斯基表示，神经科学家对网络失衡后果的探索才刚起步，但这可能催生更多修复疗法。「这不仅关乎个体生活质量，更将对整个社会产生深远影响」。

英文来源：

How the Brain Balances Excitation and Inhibition
Introduction
From Santiago Ramón y Cajal’s hand came branches and whorls, spines and webs. Now-famous drawings by the neuroanatomist in the late 19th and early 20th centuries showed, for the first time, the distinctiveness and diversity of the fundamental building blocks of the mammalian brain that we call neurons.
In the century or so since, his successors have painstakingly worked to count, track, identify, label and categorize these cells. There is now a dizzying number of ways to put neurons in buckets, often presented in colorful, complex brain cell atlases. With such catalogs, you might organize neurons based on function by separating motor neurons that help you move from sensory neurons that help you see or number neurons that help you estimate quantities. You might distinguish them based on whether they have long axons or short ones, or whether they’re located in the hippocampus or the olfactory bulb. But the vast majority of neurons, regardless of function, form or location, fall into one of two fundamental categories: excitatory neurons that trigger other neurons to fire and inhibitory neurons that stop others from firing.
Maintaining the correct proportion of excitation to inhibition is critical for keeping the brain healthy and harmonious. “Imbalances in either direction can be really catastrophic,” said Mark Cembrowski, a neuroscientist at the University of British Columbia, or lead to neurological conditions. Too much excitation and the brain can produce epileptic seizures. Too little excitation can be associated with conditions such as autism.
Neuroscientists are working to uncover how these two classes of cells work — and specifically, how they interact with a rarer third category of cells that influence their behavior. These insights could eventually help reveal how to restabilize networks that get out of balance, which can even occur as a result of normal aging.
Balance Is Key
Excitatory and inhibitory neurons work in similar ways. Most release chemical messengers known as neurotransmitters, which travel across the tiny gaps known as synapses and dock onto cuplike proteins called receptors on the next neuron. What distinguishes excitatory and inhibitory neurons is the type of neurotransmitters they release.
Excitatory neurons in the brain almost exclusively release glutamate when they activate, or fire. Glutamate triggers a bunch of positive ions to flood into a neuron, increasing its internal voltage and spurring it to fire an action potential, a strong burst of electricity that travels down a nerve fiber and makes the neuron release its own set of molecules to communicate with others, and so on.
In contrast, when inhibitory neurons fire, they release a neurotransmitter known as GABA that triggers negatively charged ions to flood into the neighboring neuron or positively charged ions to flood out. With a lower internal voltage, the next neuron won’t fire. Inhibitory neurons “function as sort of a breaker,” said Tomasz Nowakowski, a neuroscientist at the University of California, San Francisco.
These stops and gos enable a highway system in the brain, ensuring that the signals end up in the correct places at the correct times, so that you can grab the apple on your desk, hum your favorite tune or remember where you left your phone.
In the mammalian cortex, excitatory neurons vastly outnumber inhibitory ones. But throughout mammalian brain evolution, inhibitory neurons have diversified and increased in quantity, suggesting that they play critical roles in higher-order functioning.
Inhibitory neurons have “often been ascribed support roles,” said Annabelle Singer, a neuroscientist and neuroengineer at the Georgia Institute of Technology and Emory University. That’s likely because it’s simply easier to study excitatory neurons. For example, an excitatory place cell in the hippocampus can fire when an animal is in a particular location. When this happens, its excitation of other cells can be observed. “It’s very clear-cut,” she said. But an inhibitory neuron “fires a lot everywhere, and it’s much harder to say what is it responding to,” she said. We don’t know what signal it is inhibiting, and the cells connected to it don’t respond with firing of their own.
Still, studies are starting to illuminate how and when inhibitory neurons fire. In a recent study published in Nature, Singer and her colleagues found that inhibitory neurons help mice learn rapidly and remember where to find food by selectively decreasing how much they fire when the animal is near a location where food can be found. By firing less frequently as the mouse approaches the location, inhibitory neurons enhance the desired signals, thereby “enabling this learning about the important location,” Singer said. This suggests that they play a much more active role in memory than previously thought.
What’s more, the prevalent view of inhibitory neurons once cast them as more generalist in their activity, doing this kind of “blanket-y inhibition, inhibiting everything that is around their axons,” said Nuno Maçarico da Costa, a neuroscientist at the Allen Institute. But da Costa and his team, as part of the Microns project, a large-scale effort to fully map out a 1-cubic-millimeter portion of a mouse’s visual cortex, discovered that inhibitory neurons are very specific in choosing what cells to inhibit.
The brain’s circuits are all built from a mixture of inhibitory and excitatory cells conversing in diverse ways. For example, some inhibitory cells prefer to send signals to another neuron’s little branches called dendrites, while others send signals to a neuron’s cell body. Others tag team to inhibit certain other cells. These different moving parts weave together, through mechanisms not entirely understood, to create our reactions, thoughts, memories and consciousness.
But neurons communicate thousands of times faster than the cognitive effects they generate, transmitting signals in tens of milliseconds or less. “Neurotransmitters work really fast, but a lot of the behavioral and cognitive components that we need are really slow,” Cembrowski said. This apparent mismatch is “one of the central and great mysteries of the brain.”
A Third Category
Another category of cells might help to resolve this timing issue.
Neuromodulatory neurons, which are much rarer in the brain, work on slower timescales, but their effects last much longer and are much more widespread. Rather than sending molecules across a synapse exclusively to the next neuron, they can spill their molecules — a subset of neurotransmitters called neuromodulators — into an entire area, where they interact with many different synapses. The molecules they release, such as dopamine or serotonin, lead to changes within excitatory or inhibitory neurons, making them more or less likely to fire. They create “a slow undercurrent of signaling that imparts important changes in the fast dynamics of the brain,” Cembrowski said.
For example, the neuromodulator norepinephrine plays a strong role in emotionally charged memory. When released, it helps strengthen connections between neurons that form and reinforce memory, so that they fire more often and thus “guide particularly emotional experiences into memory,” he said.
These basic identities — excitatory, inhibitory, neuromodulatory — bring some structure to the way that our various types of neurons operate, but their roles can blur. For example, some excitatory and inhibitory neurons also seem to have a neuromodulatory function built into them. A small number of neurons, especially ones related to emotion, can fire GABA and glutamate packaged together, giving them both excitatory and inhibitory properties. Some neurons can switch identities, say, from an excitatory to an inhibitory neuron, under chronic stress and other conditions.
Though much diversity exists within broad categories of neurons — as one brain cell atlas after another is showing — they all enable the rhythm of excitation and inhibition. Neuroscientists are only scratching the surface of what happens when the networks are thrown off balance, but the work could lead to more treatments to fix them, Cembrowski said. “This can make a huge difference, both in individuals’ quality of life and society as a whole.”