攻击性、交配与唤起的生物学机制

摘要

Caltech生物学教授、霍华德·休斯医学研究所研究员David Anderson博士探讨了情绪、攻击性与性行为的神经科学。他阐释了情绪最好被理解为由特定神经回路支配的内部状态,并分享了关于控制攻击行为、交配行为及其令人惊讶的重叠之处的突破性研究。这场对话揭示,关于激素、行为与大脑功能的许多常见假设,比大众认知所认为的复杂得多——也往往是错误的。


核心要点

  • 情绪是内部状态的一种亚类型,并非纯粹的心理体验——它们是持续存在、能够泛化并影响行为的神经生物学过程。
  • **腹内侧下丘脑(VMH)**既是攻击、恐惧、摄食和交配的信号接收中枢,也是广播中枢——“四F”(摄食feeding、冻结freezing、战斗fighting、交配mating)紧密交织在一起。
  • 进攻性攻击对雄性小鼠具有奖励效应——它们会主动寻求战斗机会,提示攻击行为在某些神经回路中具有正向效价。
  • 恐惧优先于进攻性攻击:激活VMH中的恐惧神经元会立即终止打斗并触发冻结反应。
  • 雌激素而非睾酮才是VMH神经元层面驱动攻击行为的主要激素——雄性的攻击行为依赖于睾酮经芳香化酶转化为雌激素。
  • 相同的行为(如爬跨)可以反映完全不同的内部状态——性爬跨与支配性爬跨涉及不同的脑回路,可通过是否存在超声波发声来加以区分。
  • 在雌性中,VMH内存在两个不同的神经元亚群——一个负责战斗,一个负责交配——从未产雌鼠到哺乳母鼠,两者之间的平衡发生了显著转变。
  • **内侧视前区(MPOA)**包含专门对应交配不同阶段的神经元(嗅探、爬跨、抽插、射精),激活雌性的这些神经元可以引发雄性式爬跨行为。
  • 唤起并非单一的统一状态——存在由不同神经回路支配的、针对特定行为的唤起形式,即便某些形式需要相同的神经化学物质(如dopamine)。

详细笔记

情绪与内部状态

  • Anderson博士将情绪定义为内部状态的一类,与唤起、动机和睡眠并列。
  • 状态改变了大脑的输入-输出转换——相同的刺激在不同的当前状态下会产生不同的反应。
  • 状态的关键维度:
    • 唤起:状态的强度
    • 效价:积极或消极的性质
    • 持续性:情绪的持续时间超过其触发刺激(不同于反射)
    • 泛化性:在某一情境中触发的状态可以影响无关情境下的行为(例如,工作中糟糕的一天会改变对哭闹孩子的反应)
  • 动机状态(饥饿、口渴)具有稳态性和特异性;情绪状态更为广泛,可跨情境泛化。

唤起:并非单一现象

  • 唤起存在于从昏迷到恐慌的连续谱上,但在相同唤起水平下效价可以不同(例如,性唤起与恐惧性唤起)。
  • 对果蝇的研究发现了两个可分离的唤起回路——睡眠-觉醒唤起和惊吓性唤起——两者均需要dopamine,但使用完全不同的神经通路。
  • 这支持了行为特异性唤起形式的观点,而非单一的泛化唤起系统。
  • 决定唤起类型和质量的是回路,而不仅仅是神经化学物质。

攻击行为:回路与类型

  • “攻击”描述的是一种行为,而非单一的内部状态——它可以反映愤怒、恐惧或饥饿(掠食性攻击)。
  • 关键脑区:腹内侧下丘脑(VMH)
    • VMH下部(“梨形的宽大部分”):攻击神经元
    • VMH上部:恐惧神经元
    • 还包含:代谢/葡萄糖感应神经元和体温调节神经元
  • 大鼠VMH受损会导致肥胖——历史上被视为”抗肥胖中枢”,但同样是社会行为的核心区域。
  • Dayu Lin的里程碑式研究(Anderson实验室):使用光遗传学激活小鼠VMH神经元,可靠地触发攻击行为——而电刺激在过去50年来从未实现这一点,原因是电流扩散会同时激活恐惧神经元。

攻击行为的类型

  • 进攻性攻击:具有奖励效应;雄性小鼠会主动争取机会攻击从属个体;与VMH激活相关;大鼠表现为侧腹部咬伤。
  • 防御性攻击:与被攻击或被欺骗相关;主观上不具有奖励感;大鼠表现为咽喉/颈部咬伤;由恐惧增强(与进攻性攻击相反)。
  • 掠食性攻击:涉及与VMH不同的回路;用于捕捉猎物(例如蟋蟀)。
  • 一个称为**无名质(substantia innominata)**的区域可能是三种攻击类型共同的最终通路。
  • VMH被驱动得越强,触发攻击所需的阈值就越低——但在没有目标的情况下,不会发生明显的攻击行为。驱动力需要一个释放因素。

恐惧与攻击:毗邻的回路

  • 恐惧神经元位于VMH中攻击神经元的正上方——激活恐惧神经元会立即终止正在进行的攻击并触发冻结反应。
  • 可能的进化解释:防御性恐惧回路最先进化;进攻性攻击回路可能后来通过复制和修改与恐惧相关的前体细胞而发展形成。
  • 在发育上,两个神经元群体早期共享共同的前体细胞;基因表达随后发生分化。
  • 功能上的邻近性可能使恐惧能够在成本-收益分析发生改变时(例如在打斗中落败)迅速抑制攻击

激素与攻击:雌激素,而不仅仅是睾酮

  • 雄性VMH中的攻击神经元以雌激素受体为标志,而非雄激素受体。
  • 去势雄性小鼠失去战斗能力;仅通过雌激素植入即可恢复战斗能力,完全绕过睾酮。
  • 睾酮对攻击行为的影响主要通过芳香化酶将其转化为雌激素来介导。
  • 芳香化酶抑制剂会同时抑制动物的攻击行为和性行为。
  • 孕激素受体也在攻击神经元上表达——雌激素和孕激素(传统意义上的”女性激素”)在雄性攻击行为中均发挥关键作用。

雌性攻击行为与从未产到哺乳母鼠的转变

  • 雌性小鼠作为未产鼠时没有攻击性,但在产下一窝幼鼠后会变得极具攻击性,无论闯入者性别如何都会发起攻击。
  • 雄性小鼠的攻击神经元仅对雄性闯入者有反应;雌性攻击神经元对雄性和雌性闯入者同等反应
  • 雌性VMH内存在两个不同的雌激素受体神经元亚群:
    • 战斗神经元:未产鼠中活性低,哺乳母鼠中活性高
    • 交配神经元:在未产鼠中占主导
    • 两者之间的平衡——如同跷跷板——在分娩后从以交配为主转变为以战斗为主。
  • 仅激活战斗特异性亚群即可在未产雌鼠中触发攻击行为。
  • 这种转变是否需要幼仔的存在、哺乳或妊娠期间的激素变化,目前仍基本未知

交配回路:内侧视前区(MPOA)

  • MPOA是传统意义上雄性性行为的核心区域,包含在不同交配阶段活跃的神经元:
    • 嗅探、爬跨、抽插、射精——每个阶段都有可区分的神经元群体(通过成像而非刺激识别)。
    • 其中也存在在攻击行为中活跃的神经元——可能用于在打斗时抑制交配行为。
  • 在打斗中途激活MPOA交配神经元,会使雄性停止打斗、开始鸣叫(超声波发声),并试图爬跨雄性对手
  • MPOA =“以爱代战”神经元;VMH =“以战代爱”神经元。
  • MPOA与VMH之间存在密集的相互抑制性连接,阻止攻击时发生交配以及交配时发生攻击——但两者之间也可能存在协作性互动,这或许可以解释某些交配行为中的攻击性成分。
  • 激活雌性的MPOA等效神经元可引发对任意性别个体的雄性式爬跨行为。

爬跨:相同行为,不同状态

  • 雄性间爬跨可以是支配行为,而非性行为
  • 区分特征:


English Original 英文原文

The Biology of Aggression, Mating, & Arousal

Summary

Dr. David Anderson, a professor of biology at Caltech and Howard Hughes Medical Institute investigator, discusses the neuroscience of emotions, aggression, and sexual behavior. He explains how emotions are best understood as internal states governed by specific neural circuits, and shares groundbreaking research on the brain regions and mechanisms controlling aggression, mating, and their surprising overlap. The conversation reveals that many common assumptions about hormones, behavior, and brain function are far more complex — and often wrong — than popular understanding suggests.


Key Takeaways

  • Emotions are a subtype of internal states, not purely psychological experiences — they are neurobiological processes that persist, generalize, and bias behavior.
  • The ventromedial hypothalamus (VMH) acts as both an antenna and broadcasting center for aggression, fear, feeding, and mating — the “four Fs” (feeding, freezing, fighting, mating) are all tightly intermingled.
  • Offensive aggression is rewarding in male mice — they will actively seek out opportunities to fight, suggesting aggression has a positive valence in certain neural circuits.
  • Fear dominates over offensive aggression: stimulating fear neurons in the VMH instantly stops a fight and triggers freezing.
  • Estrogen, not testosterone, is the primary hormonal driver of aggression at the level of VMH neurons — male aggression depends on testosterone being converted to estrogen via aromatization.
  • The same behavior (e.g., mounting) can reflect completely different internal states — sexual mounting vs. dominance mounting involve different brain circuits and can be distinguished by the presence or absence of ultrasonic vocalizations.
  • In females, there are two distinct subsets of VMH neurons — one for fighting and one for mating — and the balance between them shifts dramatically from virgin to nursing mother.
  • The medial preoptic area (MPOA) contains neurons specific to different phases of mating (sniffing, mounting, thrusting, ejaculation), and stimulating them in females can elicit male-type mounting behavior.
  • Arousal is not a single, unified state — there are behavior-specific forms of arousal governed by distinct neural circuits, even if some require the same neurochemicals (like dopamine).

Detailed Notes

Emotions vs. Internal States

  • Dr. Anderson defines emotions as a class of internal state alongside arousal, motivation, and sleep.
  • States change the input-to-output transformation of the brain — the same stimulus produces different responses depending on current state.
  • Key dimensions of states:
    • Arousal: intensity of the state
    • Valence: positive or negative quality
    • Persistence: emotions outlast their triggering stimulus (unlike reflexes)
    • Generalization: a state triggered in one context can influence behavior in an unrelated context (e.g., a bad day at work altering response to a crying child)
  • Motivation states (hunger, thirst) are homeostatic and specific; emotion states are broader and generalize across contexts.

Arousal: Not One Thing

  • Arousal exists on a spectrum from coma to panic, but valence can differ at the same arousal level (e.g., sexual arousal vs. fearful arousal).
  • Research in fruit flies identified two separable arousal circuits — sleep-wake arousal and startle-based arousal — both requiring dopamine but using completely different neural pathways.
  • This supports the idea of behavior-specific forms of arousal rather than a single generalized arousal system.
  • The circuit, not just the neurochemical, determines the type and quality of arousal.

Aggression: Circuits and Types

  • “Aggression” describes a behavior, not a single internal state — it can reflect anger, fear, or hunger (predatory aggression).
  • Key brain region: ventromedial hypothalamus (VMH)
    • Lower VMH (“fat part of the pear”): aggression neurons
    • Upper VMH: fear neurons
    • Also contains: metabolic/glucose-sensing neurons and thermoregulatory neurons
  • VMH destruction in rats produces obesity — historically viewed as an “anti-obesity center,” but also central to social behavior.
  • Dayu Lin’s landmark work (Anderson lab): used optogenetics to activate VMH neurons in mice and reliably trigger aggression — something electrical stimulation had failed to do for 50 years due to current spread activating fear neurons simultaneously.

Types of Aggression

  • Offensive aggression: rewarding; male mice will work to access a subordinate to attack; associated with VMH stimulation; flank-directed biting in rats.
  • Defensive aggression: associated with being attacked or cheated; does not feel rewarding subjectively; throat/neck-directed biting in rats; enhanced by fear (opposite of offensive aggression).
  • Predatory aggression: involves different circuits from VMH; used for catching prey (e.g., crickets).
  • A region called the substantia innominata may be a final common pathway for all three types of aggression.
  • The more strongly VMH is driven, the lower the threshold needed to trigger attack — but with no target present, overt attack does not occur. Drive requires a releaser.

Fear and Aggression: Neighboring Circuits

  • Fear neurons sit directly above aggression neurons in VMH — stimulating fear neurons immediately stops ongoing aggression and triggers freezing.
  • Possible evolutionary explanation: defensive fear circuits evolved first; offensive aggression circuits may have developed later through duplication and modification of fear-related precursor cells.
  • Developmentally, both populations share early common precursors; gene expression later diverges.
  • Functional proximity may allow fear to inhibit aggression rapidly when the cost-benefit analysis shifts (e.g., losing a fight).

Hormones and Aggression: Estrogen, Not Just Testosterone

  • Aggression neurons in male VMH are marked by the estrogen receptor, not androgen receptor.
  • Castrated male mice lose the ability to fight; fighting can be restored by estrogen implants alone, bypassing testosterone entirely.
  • Testosterone’s effects on aggression are largely mediated through conversion to estrogen via aromatase.
  • Aromatase inhibitors suppress both aggression and sexual behavior in male animals.
  • Progesterone receptor is also expressed on aggression neurons — both estrogen and progesterone (classically “female hormones”) play key roles in male aggression.

Female Aggression and the Virgin-to-Mother Transition

  • Female mice are non-aggressive as virgins but become highly aggressive after delivering a litter, attacking any intruder regardless of sex.
  • Male mice’s aggression neurons respond only to male intruders; female aggression neurons respond to both male and female intruders equally.
  • Within female VMH, two distinct subsets of estrogen receptor neurons exist:
    • Fighting neurons: low activity in virgins, high in nursing mothers
    • Mating neurons: dominant in virgins
    • The balance shifts — like a seesaw — from mating-dominant to fighting-dominant after parturition.
  • Stimulating the fighting-specific subset alone can trigger aggression even in virgin females.
  • Whether the transition requires pup presence, lactation, or hormonal changes from pregnancy remains largely unknown.

Mating Circuits: The Medial Preoptic Area (MPOA)

  • MPOA is the traditional center for male sexual behavior and contains neurons active during different phases:
    • Sniffing, mounting, thrusting, ejaculation — each phase has distinguishable neuron populations (identified by imaging, not stimulation).
    • Some neurons active during aggression are also present — possibly to inhibit mating during fighting.
  • Stimulating MPOA mating neurons mid-fight causes a male to stop fighting, begin singing (ultrasonic vocalizations), and attempt to mount the male opponent.
  • MPOA = “make love not war” neurons; VMH = “make war not love” neurons.
  • Dense, mutually inhibitory connections between MPOA and VMH prevent mating during attack and vice versa — but cooperative interactions may also exist, potentially explaining the aggressive component in some mating behaviors.
  • Stimulating MPOA-equivalent neurons in females elicits male-type mounting behavior toward either sex.

Mounting: Same Behavior, Different States

  • Male-male mounting can be dominance behavior, not sexual behavior.
  • Distinguishing features: