Towards a cognitive framework for understanding cellular behavior

摘要 自由生活的细胞对个体环境的变化表现出强烈的全基因组转录反应 氧气和温度等参数。这种转录动力学被认为是 稳态反应，试图逆转由 具体的环境变化。我提出了另一种解释，其中转录反应 反映针对预期的全球环境变化的多方面行为计划 跟随扰动。这是因为原生微生物栖息地是高度结构化的，这使得 环境参数之间存在很强的相关性。在地质时间尺度上，这种相关性 可以通过“联想学习”过程内化，该过程塑造了事物的连通性和动态性 监管网络。这种内部模型应该允许微生物预测未来的轨迹 基于即时感官信息的环境。我们已经看到了这种预期行为的证据 大肠杆菌对温度和氧气变化的反应，对应于 外部环境和哺乳动物胃肠道之间的过渡。这些内部 表征似乎反映了真正的联想学习范式，因为它们在接触后表现出可塑性 到新奇的环境。这些现象越来越要求我们从一个角度来解释微生物的行为： 认知视角，就像我们理解动物行为一样。我建议进行多方面的研究 计划旨在 1）确定这些现象确实代表了微生物的认知模型 栖息地，2）揭示潜在的网络机制，3）探索小说的联想学习 通过实验室实验进化的环境。拟议的工作建立了深厚的联系 微生物生态学、调控网络进化和行为等不同领域之间的关系。这样做，它 挑战了具有百年历史的体内平衡概念的主导地位，对我们如何维持生命具有根本性影响 了解和控制微生物行为，尤其是在疾病背景下。

Abstract Free-living cells display strong genome-wide transcriptional responses to changes in individual environmental parameters such as oxygen and temperature. Such transcriptional dynamics are thought to be the basis of a homeostatic response that attempts to reverse the immediate intracellular consequences resulting from the specific change in the environment. I present an alternative interpretation where transcriptional responses reflect a multifaceted behavioral program in response to global changes in the environment that are anticipated to follow the perturbation. This results from the fact that native microbial habitats are highly structured, giving rise to strong correlations between environmental parameters. Over geological timescales, such correlations can be internalized through an "associative learning" process that shapes the connectivity and dynamics of regulatory networks. Such internal models should allow microbes to predict the future trajectory of the environment based on immediate sensory information. We have seen evidence of this anticipatory behavior in responses of the bacterium Escherichia coli to changes in temperature and oxygen that correspond to transitions between the outside environment and the mammalian gastrointestinal tract. These internal representations seem to reflect a true associative learning paradigm, since they show plasticity upon exposure to novel environments. These phenomena increasingly demand that we interpret microbial behaviors from a cognitive perspective, much as we do for understanding animal behaviors. I propose a multi-faceted research program aimed at 1) establishing that these phenomena, indeed, represent cognitive modeling of microbial habitats, 2) revealing the underlying network mechanisms, and 3) exploring associative learning of novel environments through laboratory experimental evolution. The proposed work establishes deep connections between the disparate fields of microbial ecology, regulatory network evolution, and behavior. In so doing, it challenges the dominance of the century-old notion of homeostasis, with fundamental implications for how we understand and control microbial behavior, especially in the context of disease.