CAREER: Enabling Immunomodulatory Treatment of Influenza Infection using Multiscale Modeling

呼吸道病毒感染是对公共健康的持续威胁。在美国，流感感染每年导致多达70万人住院和5.6万人死亡。免疫反应过度导致的极端肺部炎症是流感病毒和冠状病毒感染(包括非典、中东呼吸综合征和新出现的证据中的新冠肺炎病毒)严重疾病后果的主要因素。免疫系统是一个复杂、互动、动态的系统，必须以最佳方式清除病毒感染，同时将免疫调节过程对肺部和其他器官造成的附带损害降至最低。基于工程的数学建模方法非常适合于创建免疫系统的计算模拟，可以用来估计系统对病毒感染的反应。本研究项目的总体目标是构建现实的、可预测的、生理上准确的流感病毒感染过程中自然免疫反应的计算模型。计算模型将构成基于模拟的研究的基础，以确定允许发生过度免疫反应的免疫学条件。模拟还将确定抑制免疫活动的最佳方法，使用化学抑制来减少组织炎症，同时确保感染迅速清除。由于过度炎症是许多呼吸道病毒感染的共同特征，因此产生的关于免疫调节的见解预计将影响各种呼吸道病毒疾病。在这项研究工作的同时，研究小组将开发虚拟现实(VR)游戏，以更好地让公众参与与呼吸道感染和免疫反应相关的问题。VR游戏将免费分发给用户的智能手机，以确保强大的公众参与度。VR游戏也将在当地高中的研讨会上使用，以促进学生对工程和免疫学的参与。研究人员的长期职业愿景是设计基于计算建模的解决方案，以调节和改善免疫系统反应。为了实现这一愿景，这个职业项目的目标是确定在流感感染期间驱动肺部炎症的分子/细胞机制，并使用多尺度计算模型评估矽肺的免疫调节治疗。研究表明，选择性地抑制免疫系统可以在不影响病毒清除的情况下减少炎症，从而显著改善感染结果。免疫调节也被证明比服用抗病毒药物(如奥司他韦)提供更大的保护，但目前还没有关于免疫调节治疗的全面指南，如抗炎皮质类固醇。这样的免疫调节策略本质上是一个工程优化挑战：修改免疫反应以限制炎症，同时确保病毒清除。所设计的预测模型可以用来评估关于流感病毒感染过程中炎症调控机制的具体假设。该研究计划是在三个协同目标下组织的。第一个目标是构建和使用ODE(常微分方程式)模型来识别流感感染的人肺上皮细胞炎症的分子驱动因素。将开发一种上皮细胞内信号传递模型，用于揭示炎症蛋白产生的分子驱动因素，评估炎症和抑制病毒复制之间的权衡，并在比较高致病性(H5N1)病毒感染和较轻病毒感染时确定可能的病毒特异性炎症调节。这一目标的完成将为一般和特定病毒感染中调节肺上皮炎症的分子机制提供证据。第二个目标是构建和使用ODE模型来识别导致流感感染小鼠肺部炎症增强的免疫学条件。该模型将把肺上皮信号与免疫细胞渗透联系起来。该模型将被用来确定在组织水平上推动炎症增强的机制，并探索治疗方案。不同病毒感染之间的比较以及男性(不太严重)与女性(更严重)感染之间的比较可能揭示病毒和队列特异性炎症调节。这一目标的完成将确定流感引起炎症的关键分子和细胞驱动因素，提供一种新的肺免疫系统计算模型，使重要感染队列之间的比较成为可能，并可能确定驱动差异炎症的病毒特异性或性别特异性免疫调节。第三个目标是构建和使用肺免疫系统的ABM(基于代理的模型)来量化细胞异质性对组织水平炎症的影响。干扰素的产生是随机的，可能会导致不同的感染结果。使用定制的代码，将构建一个ABM，其中ODE定义了代理人(上皮细胞)如何与其环境相互作用，并用于询问在流感感染期间，上皮细胞的异质性如何影响组织水平的炎症。完成这一目标将确定负责维持严格控制的炎症反应的免疫系统组件，提供关于受试者内部炎症如何因免疫调节差异而变化的知识，为模拟社区产生新的代码，并产生流感感染期间肺免疫系统的新ABM。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

Respiratory virus infections are a constant threat to public health. Influenza infection results in up to 700,000 hospitalizations and 56,000 deaths each year in the US. Extreme lung inflammation due to an excessive immune response is a major factor of severe disease outcomes in influenza virus and coronavirus infection (including SARS, MERS, and in emerging evidence, the COVID-19 virus.) The immune system is a complex, interactive, dynamic system that must optimally clear the virus infection while minimizing collateral damage to the lungs and other organs initiated by immune-regulated processes. Engineering-based mathematical modeling approaches are ideally suited to create computational simulations of the immune system that can be used to estimate the system’s response to virus infection. The overall goal of this research project is to construct realistic, predictive, physiologically accurate computational models of the natural immune response during influenza virus infection. The computational models will form the foundation for simulation-based research to identify the immunologic conditions that allow for excessive immune responses to occur. Simulations will also identify the best way to suppress immune activity, using chemical inhibition to reduce tissue inflammation while ensuring rapid clearance of the infection. As excessive inflammation is a common feature of many respiratory virus infections, the insights on immune regulation generated are expected to impact a variety of respiratory virus diseases. In parallel with this research effort, the research team will develop virtual reality (VR) games to better engage the public on matters associated with respiratory infection and immune responses. The VR games will be distributed to user smartphones for free to ensure strong public engagement. The VR games will also be used during workshops at local high schools to promote student engagement in engineering and immunology. The Investigator’s long-term CAREER vision is to engineer computational modeling-based solutions to regulate and improve immune system responses. Toward this vision, the objectives of this CAREER project are to identify the molecular/cellular mechanisms that drive lung inflammation during influenza infection and to evaluate immunomodulatory treatment in silico using multiscale computational modeling. Studies show that selectively inhibiting the immune system can significantly improve infection outcomes by reducing inflammation without compromising virus clearance. Immunomodulation has also been shown to offer greater protection than administration of antiviral medicines (e.g. oseltamivir), but no comprehensive guidelines for administering immunomodulatory treatments, such as anti-inflammatory corticosteroids, exist. Such immunomodulatory strategies are inherently an engineering optimization challenge: modifying immune responses to limit inflammation while ensuring virus clearance. The predictive models designed can be used to evaluate specific hypotheses on the mechanisms regulating inflammation during influenza virus infection. The research program is organized under three synergistic objectives. The FIRST Objective is to construct and use an ODE (Ordinary Differential Equations) model to identify the molecular drivers of inflammation in influenza-infected human lung epithelial cells. A model of epithelial intracellular signaling will be developed and used to uncover the molecular drivers of inflammatory protein production, evaluate trade-offs between inflammation and suppressing virus replication, and to identify possible virus-specific inflammatory regulation when comparing highly pathogenic (H5N1) and milder virus infections. Completion of this objective will provide evidence of the molecular mechanisms regulating lung epithelial inflammation in general and in specific virus infections. The SECOND Objective is to construct and use an ODE model to identify the immunologic conditions that drive enhanced inflammation in influenza-infected mouse lungs. The model will link lung epithelial signaling with immune cell infiltration. The model will be used to identify the mechanisms that drive enhanced inflammation at the tissue level and explore treatment options. Comparisons between infection with different viruses and between male (less severe) vs female (more severe) infection may reveal virus and cohort-specific inflammation regulation. Completion of this objective will identify the key molecular and cellular drivers of influenza-induced inflammation, provide a novel computational model of the lung immune system that enables comparisons between important infection cohorts, and potentially identify virus-specific or sex-specific immune regulation that is driving differential inflammation. The THIRD Objective is to construct and use an ABM (Agent Based Model) of the lung immune system to quantify the impact of cell heterogeneity on tissue-level inflammation. Interferon production is stochastic and may factor into variable infection outcomes. Using customized code, an ABM in which ODEs define how agents (epithelial cells) interact with their environment will be constructed and used to interrogate how epithelial cellular heterogeneity impacts tissue-level inflammation during influenza infection. Completion of this objective will identify the components of the immune system responsible for maintaining a tightly regulated inflammatory response, provide knowledge of how intra-subject inflammation may vary due to differences in immune regulation, produce new code for the simulation community, and produce a novel ABM of the lung immune system during influenza infection.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.