EHD1 and RUSC2 Control Basal Epidermal Growth Factor Receptor Cell Surface Expression and Recycling.

Epidermal growth factor receptor (EGFR) is a prototype receptor tyrosine kinase and an oncoprotein in many solid tumors. Cell surface display of EGFR is essential for cellular responses to its ligands. While postactivation endocytic trafficking of EGFR has been well elucidated, little is known about mechanisms of basal/preactivation surface display of EGFR. Here, we identify a novel role of the endocytic regulator EHD1 and a potential EHD1 partner, RUSC2, in cell surface display of EGFR. EHD1 and RUSC2 colocalize with EGFR in vesicular/tubular structures and at the Golgi compartment. Inducible EHD1 knockdown reduced the cell surface EGFR expression with accumulation at the Golgi compartment, a phenotype rescued by exogenous EHD1. RUSC2 knockdown phenocopied the EHD1 depletion effects. EHD1 or RUSC2 depletion impaired the EGF-induced cell proliferation, demonstrating that the novel, EHD1- and RUSC2-dependent transport of unstimulated EGFR from the Golgi compartment to the cell surface that we describe is functionally important, with implications for physiologic and oncogenic roles of EGFR and targeted cancer therapies.

Global gene expression profiling of healthy human brain and its application in studying neurological disorders.

Brain function is governed by precise regulation of gene expression across its anatomically distinct structures; however, the expression patterns of genes across hundreds of brain structures are not clearly understood. Here, we describe a gene expression model, which is representative of the healthy human brain transcriptome by using data from the Allen Brain Atlas. Our in-depth gene expression profiling revealed that 84% of genes are expressed in at least one of the 190 brain structures studied. Hierarchical clustering based on gene expression profiles delineated brain regions into structurally tiered spatial groups and we observed striking enrichment for region-specific processes. Further, weighted co-expression network analysis identified 19 robust modules of highly correlated genes enriched with functional associations for neurogenesis, dopamine signaling, immune regulation and behavior. Also, structural distribution maps of major neurotransmission systems in the brain were generated. Finally, we developed a supervised classification model, which achieved 84% and 81% accuracies for predicting autism- and Parkinson's-implicated genes, respectively, using our expression model as a baseline. This study represents the first use of global gene expression profiling from healthy human brain to develop a disease gene prediction model and this generic methodology can be applied to study any neurological disorder.