Defining a transcriptional periodic table of the human brain using reprogramming

点击翻译按钮获取中文摘要

The human brain contains diverse neural cell types that are differentially responsible for distinct aspects of human behavior, cognition and neurologic disease. Advances in DNA sequencing are providing new insight into human specific neurobiology by providing lists of genes that increase risk for disease or correlate with behaviors and cognitive properties that differ between individuals. However, the lack of available human neuronal subtypes and our limited understanding of which genes are expressed in different human neurons are major barriers to exploiting these growing genomic resources. One way to overcome this barrier is to use direct reprogramming to produce induced neuronal cells in vitro, by transiently expressing transcription factors (TFs) in fibroblasts. Direct reprogramming produces induced neurons that share many features with endogenous neurons, including characteristic morphologies, ligand-evoked synaptic responses and characteristic patterns of gene expression. Reprogramming therefore offers a new tool to identify transcriptional circuits that establish distinct features of neuronal identity. We, and others, have used candidate gene approaches to engineer induced neurons that functionally mimic well-characterized neuronal subtypes, such as the peripheral sensory neurons that detect pain and itch produced recently by my laboratory. These studies led us to hypothesize that direct reprogramming engages conserved transcriptional circuits similar to those that actively maintain neuronal subtype identity in endogenous neurons. This hypothesis predicts that it should be possible to identify multiple TF combinations that induce distinct features of different neuronal subtypes. To test this hypothesis and establish a systematic method to produce and classify human neuronal subtypes, we will conduct an unbiased screen for new TF combinations that can induce human neuronal subtypes in vitro. We will then characterize the induced neurons transcriptionally, morphologically and functionally. We are well suited to perform this study because we recently conducted a pilot screen of ~600 TF pairs and identified more than 70 new pairs that produce candidate induced neurons from mouse fibroblasts. Gene expression profiling and functional analyses of these cells confirm that they exhibit extensive subtype diversity. Therefore, by using unbiased screens to define sets of human TFs that can induce neuronal identity in fibroblasts, we will identify new methods to produce human neuronal cell types with defined functional properties in vitro and establish a database of transcriptional programs and cellular properties that emerge from transient expression of different sets of TFs. These studies will have impact on our understanding of the basic biology of human neuronal diversity and will provide conceptual and practical tools to enable neuroscience researchers to produce diverse subtypes of human induced neurons for research and translational applications.