Bottom-up and top-down computational modeling approaches to study CMV retinitis

项目摘要 人巨细胞病毒（HCMV）是一种广泛存在的病毒，感染后可导致从无症状到严重的综合征。 由于免疫功能低下个体中潜伏感染的再激活，潜伏至危及生命的脓毒症。一 免疫抑制患者中HCMV感染的主要但经常被忽视的后果是CMV视网膜炎 （CMVR）。鉴于其普遍存在的性质，包括失明在内的视觉并发症的风险以及与 诊断和治疗，HCMV，CMVR的病原体，是一个重要的研究目标，以提高 知识和制定有效的治疗策略。视网膜色素上皮（RPE）是一种重要的色素上皮细胞。 它被认为是病毒（包括HCMV）进入视网膜神经层的入口。一步HCMV 在成纤维细胞中，裂解性复制周期在体外持续约96小时， 感染性病毒体和被感染细胞的破坏。已知成人RPE（ARPE 19）的HCMV感染 细胞导致更慢，破坏性更小的持续感染和多步复制。计算建模 是一种方法上的进步，最近才被应用于研究的动态概况， 一些病毒蛋白在单步复制中，并且存在很少的多步复制模型，可能 由于缺乏一个易于测试的体外持久性系统，从而产生了两个相关的知识空白。的 该奖学金的目的是提供培训，整合湿实验室和计算策略，以研究 使用HCMV裂解复制动力学和CMVR作为模型系统的复杂视觉病理。的目标 该研究金的目的是：（1）确定和计算模型的影响，不同的多重感染（MOI） 和细胞类型（即，成纤维细胞vs. ARPE 19）对HCMV复制动力学的影响;（2）发展 类似于在ARPE 19细胞中采用多步复制的持续HCMV感染的体外模型系统; 实验研究和计算机模拟抗HCMV药物更昔洛韦的作用 (GCV)使用ARPE 19细胞的新型持久性样体外系统。在目标1中，我将使用自下而上的生物- 基于化学的，MOI依赖的计算模型来描述蛋白质相互作用和调节， 单步裂解HCMV复制。我将通过比较模型参数来研究细胞类型依赖的变异性- 使用MRC 5成纤维细胞来源的数据和ARPE 19细胞来源的数据进行了量化。在目标2中，我将使用自顶向下 一种方法来计算模型的影响，多步HCMV复制在ARPE 19细胞在细胞 水平使用一种新的，在体外持久性样系统和靶细胞有限的模型。我会在体外 持久性样系统使用以低MOI感染的ARPE 19细胞，允许多步复制，以实现 病毒DNA（vDNA）的稳态水平。我将把这个系统对抗HCMV药物GCV的反应与 从现有的目标细胞有限的模型，使用患者数据参数化的预测。最后要 使用靶细胞限制模型预测最佳体外GCV治疗持续时间，导致vDNA水平 衰减到qPCR检测限以下，然后使用新的持久性样系统对其进行测试。

PROJECT SUMMARY Human cytomegalovirus (HCMV) is widespread and infection results in syndromes ranging from asymptomatic latency to life-threatening sepsis due to reactivation of a latent infection in immunocompromised individuals. One major, yet often overlooked, consequence of HCMV infection in immunosuppressed patients is CMV retinitis (CMVR). Given its ubiquitous nature, risk of visual complications including blindness, and cost associated with diagnosis and therapy, HCMV, the causative agent of CMVR, is a crucial target for investigation to advance knowledge and develop effective therapeutic strategies. The retinal pigment epithelium (RPE) has been impli- cated as a point of entry for viruses, including HCMV, into the neural layer of the retina. The single-step HCMV lytic replication cycle has an approximately 96 h duration in vitro in fibroblasts that culminates in production of infectious virions and destruction of the infected cell. It is known that HCMV infection of adult RPE (ARPE19) cells leads to a slower, less destructive persistent infection and multi-step replication. Computational modeling is a methodological advancement that has only been recently applied to the study of the dynamic profiling of some viral proteins in single-step replication, and there are very few multi-step replication models that exist, likely due to a lack of a readily testable in vitro persistence-like system, thus creating two related knowledge gaps. The objective of this fellowship is to provide training that integrates wet-lab and computational strategies to study complex visual pathologies using HCMV lytic replication kinetics and CMVR as a model system. The goals of this fellowship are to: (1) identify and computationally model the effects of varying multiplicity of infection (MOI) and cell type (i.e., fibroblast vs. ARPE19) on HCMV replication kinetics during single-step replication; (2) develop an in vitro model system similar to a persistent HCMV infection employing multi-step replication in ARPE19 cells; and (3) experimentally investigate and computationally simulate the effects of the anti-HCMV drug ganciclovir (GCV) on the novel persistence-like in vitro system using ARPE19 cells. In Aim 1, I will use a bottom-up, bio- chemically-based, MOI-dependent computational model to describe the protein interactions and regulation in single-step lytic HCMV replication. I will investigate cell type-dependent variability by comparing models param- eterized using MRC5 fibroblast-derived data and ARPE19 cell-derived data. In Aim 2, I will use a top-down approach to computationally model the effects of multi-step HCMV replication in ARPE19 cells at the cellular level using a novel, in vitro persistence-like system and a target-cell limited model. I will generate an in vitro persistence-like system using ARPE19 cells infected at a low MOI, allowing for multi-step replication, to achieve a steady-state level of viral DNA (vDNA). I will correlate this system’s response to the anti-HCMV drug GCV with the predictions from extant target-cell limited models that were parameterized using patient data. Finally, I will use the target-cell limited model to predict the optimal in vitro GCV treatment duration causing vDNA levels to decay below the qPCR limit of detection and then test this using the novel persistence-like system.