Macrophage Polarization in Response to Infections and Inflammation

摘要 巨噬细胞在希腊语中的意思是“食量大的人”，是先天免疫的强大细胞成分。它们扮演着 通过‘吃掉’病原体、死亡细胞或癌细胞，在免疫防御中发挥关键作用。它们也有助于组织的形成 动态平衡、发育和修复。当巨噬细胞工作时，巨噬细胞对周围环境做出反应并触发 急性发炎来解决问题。他们通过假设两个州中的一个来实现这一点 被识别的，即免疫反应性(促炎)和免疫耐受(分别为M1和M2)。而当 在生理学上，有限的反应性和耐受性是可取的，超过任何一种状态都是不可取的，并且 总是与疾病的发病机制有关(例如，金发姑娘难题)。例如，高反应性 被认为是一系列疾病(结肠炎、败血症、NASH)中组织损伤的根本原因 过度耐受性是一个共同的决定因素，它导致了大多数，如果不是所有的慢性疾病都是无法治愈的，例如， 癌症。关于这些生理性和病理性巨噬细胞状态的定义还没有达成共识 可能是因为4个主要挑战：异质性、生物稳健性、时间进化 和伪影(巨噬细胞在分离时迅速漂移，具有巨大的可塑性 纸巾)。我们使用了一种新的计算方法，布尔蕴涵网络[Saho2008]，来 分析汇集的人类巨噬细胞基因表达数据集。这种识别不对称基因的方法 表达模式模糊了噪声(异质性/伪像)，但揭示了事件的时间模型 在所有数据集上都可以看到。分析揭示了到目前为止未知的反应之间的连续过渡态 到五条路径的容忍状态；机器学习确定其中一条是主要路径，随后 经受住了跨物种(小鼠)对多个公开可用的转录数据集的严格测试/验证 和人类)、巨噬细胞亚型和疾病状态。最重要的是，与其他常用基因不同 通过集群签名，布尔路径可以预测不同疾病的结果。初步验证 对遗传模型的研究证实，这条途径可以被用来调节巨噬细胞的极化 改变内毒素/TLR4应答。我们现在将使用迭代来询问这些发现的影响 方法，即模型驱动的实验和实验驱动的模型精化，通过三个目标： 揭示新分子驱动因素在新发现的巨噬细胞基因特征中的重要性 半HTP化学/遗传屏蔽对小鼠和人单核细胞来源巨噬细胞的极化作用 (目标1)，在高反应性和高耐受性的小鼠疾病模型中(目标2)，以及在“类人”，即人类 以有机物质为基础的微生物/免疫细胞共培养模型(“肠道在盘”；目标3)。尽管我们的重点是 胃肠道感染和炎症，这一发现将定义巨噬细胞在多个 器官/疾病背景，因此，影响许多领域。我们希望确定高价值的治疗靶点 这可以限制和/或重置巨噬细胞对“金发地带”内感染和炎症的反应。

Abstract Macrophages in Greek means “big eaters" are powerful cellular components of innate immunity. They play a pivotal role in immune defense by ‘eating’ pathogens, dead or cancerous cells. They also contribute to tissue homeostasis, development and repair. When doing their job, macrophages react to their surroundings and trigger acute inflammation to resolve the problems. They do so by assuming one of the two states that have been widely recognized, i.e., immunoreactive (proinflammatory) and immunotolerant (a.k.a, M1 and M2, respectively). While finite degrees of reactivity and tolerance are desirable in physiology, excess of either state is undesirable and invariably associated with disease pathogenesis (i.e., the Goldilocks conundrum). For example, hyperreactivity is recognized as the root cause of tissue injury in a wide array of diseases (colitis, sepsis, NASH) and hypertolerance is a common determinant that drives most, if not all chronic diseases that are incurable, e.g., cancers. Consensus on the definition of these physiologic and pathologic macrophage states has not been reached, perhaps because of 4 major challenges: heterogeneity, biological robustness, the temporal evolution of the network, and artifacts (tremendous plasticity of macrophages as they drift rapidly when isolated from tissues). We have used a novel computational methodology, Boolean Implication Network [Sahoo 2008], to analyze pooled human macrophage gene expression datasets. This method, which identifies asymmetric gene expression patterns, blurs noise (heterogeneity/artifacts) but reveals a temporal model of events that is invariably seen across all datasets. The analysis revealed hitherto unknown continuum transition states between reactive to tolerant states along five paths; machine-learning identified one of them as the major path which subsequently stood the rigorous test/validation on multiple publicly available transcriptomic datasets, across species (mouse and human), macrophage subtypes and disease states. Most importantly, unlike other commonly used gene cluster signatures, the Boolean path can prognosticate outcomes across diverse diseases. Preliminary validation studies on a genetic model confirm that the path could be exploited for modulating macrophage polarization by altering LPS/TLR4 responses. We will now interrogate the impact of these discoveries using an iterative approach, i.e., model-driven experimentation and experiment-driven model refinement, through three aims: Unravel the importance of novel molecular drivers in the newly identified gene signatures of macrophage polarization using semi-HTP chemical/genetic screens on murine and human monocyte-derived macrophages (Aim 1), in murine disease models of hyperreactivity and hypertolerance (Aim 2) and in “Humanoids”, i.e., human organoid-based microbe/immune cells co-culture models (“gut-in-a-dish”; Aim 3). Although our focus is gastrointestinal infection and inflammation, the findings will define macrophage transition states in multiple organs/disease contexts and therefore, impact many fields. We expect to identify high-value therapeutic targets that can restrict and/or reset macrophage responses to infections and inflammation within the “Goldilocks zone.”