Computational Identification of Phospho-Tyrosine Sub-Networks Related to Acanthocyte Generation in Neuroacanthocytosis

Acanthocytes, abnormal thorny red blood cells (RBC), are one of the biological hallmarks of neuroacanthocytosis syndromes (NA), a group of rare hereditary neurodegenerative disorders. Since RBCs are easily accessible, the study of acanthocytes in NA may provide insights into potential mechanisms of neurodegeneration. Previous studies have shown that changes in RBC membrane protein phosphorylation state affect RBC membrane mechanical stability and morphology. Here, we coupled tyrosine-phosphoproteomic analysis to topological network analysis. We aimed to predict signaling sub-networks possibly involved in the generation of acanthocytes in patients affected by the two core NA disorders, namely McLeod syndrome (MLS, XK-related, Xk protein) and chorea-acanthocytosis (ChAc, VPS13A-related, chorein protein). The experimentally determined phosphoproteomic data-sets allowed us to relate the subsequent network analysis to the pathogenetic background. To reduce the network complexity, we combined several algorithms of topological network analysis including cluster determination by shortest path analysis, protein categorization based on centrality indexes, along with annotation-based node filtering. We first identified XK- and VPS13A-related protein-protein interaction networks by identifying all the interactomic shortest paths linking Xk and chorein to the corresponding set of proteins whose tyrosine phosphorylation was altered in patients. These networks include the most likely paths of functional influence of Xk and chorein on phosphorylated proteins. We further refined the analysis by extracting restricted sets of highly interacting signaling proteins representing a common molecular background bridging the generation of acanthocytes in MLS and ChAc. The final analysis pointed to a novel, very restricted, signaling module of 14 highly interconnected kinases, whose alteration is possibly involved in generation of acanthocytes in MLS and ChAc.

棘形红细胞（acanthocytes）是形态异常的带刺红细胞（red blood cells, RBC），属于神经棘红细胞增多症综合征（neuroacanthocytosis syndromes, NA）的生物学标志之一，该类疾病为一组罕见的遗传性神经退行性疾病。由于红细胞易于获取，对神经棘红细胞增多症综合征患者体内棘形红细胞的研究或可为神经退行性病变的潜在发病机制提供新的研究线索。既往研究表明，红细胞膜蛋白的磷酸化状态改变会影响红细胞膜的机械稳定性与细胞形态。本研究将酪氨酸磷酸化蛋白质组学分析（tyrosine-phosphoproteomic analysis）与拓扑网络分析（topological network analysis）相结合，旨在预测可能参与两种核心神经棘红细胞增多症综合征——麦克劳德综合征（McLeod syndrome, MLS，XK相关，Xk蛋白）和棘红细胞增多症舞蹈病（chorea-acanthocytosis, ChAc，VPS13A相关，chorein蛋白）——患者体内棘形红细胞生成的信号子网络。通过实验获得的磷酸化蛋白质组数据集，使我们能够将后续的网络分析与疾病的致病背景相结合。为降低网络复杂度，我们整合了多种拓扑网络分析算法，包括基于最短路径分析的聚类识别、基于中心性指数的蛋白质分类，以及基于注释的节点筛选。我们首先通过识别连接Xk蛋白与chorein蛋白，以及在患者体内酪氨酸磷酸化水平发生改变的对应蛋白质组的所有相互作用组最短路径，确定了XK相关与VPS13A相关的蛋白质-蛋白质相互作用网络。这些网络涵盖了Xk蛋白与chorein蛋白对磷酸化蛋白质发挥功能影响的最可能通路。我们进一步通过提取高度相互作用的信号蛋白限制性集合来优化分析，该集合代表了连接麦克劳德综合征与棘红细胞增多症舞蹈病患者体内棘形红细胞生成的共同分子背景。最终分析指向了一个全新的、高度受限的信号模块，该模块由14个高度相互连接的激酶组成，其异常或参与了麦克劳德综合征与棘红细胞增多症舞蹈病患者体内棘形红细胞的生成过程。