Might Some Brain Therapies in Animal Models Fail in Humans due to Fundamental Differences?
Mark Harnett, Ph.D.
Massachusetts Institute of Technology
David Mahoney Neuroimaging Program
September 2017, for 3 years
Might some brain therapies in animal models fail in humans due to fundamental differences?
Investigators will use two-photon imaging in human brain tissue removed during surgical treatment to explore their hypothesis that the human brain, compared to animal model systems, uses uniquely specialized information processing at the level of single neurons and their synapses. These fundamental differences, they suggest, may account for why many promising animal-model studies of experimental therapies for neurological and psychiatric diseases fail when translated into human clinical studies.
As the investigators note, the human brain contains approximately 100 billion neurons, linked together with staggering complexity by approximately 100 trillion connections (synapses). Neurons are thought to represent the elemental units of information processing in the brain. They produce electrical signals that communicate via synapses to other neurons organized into large networks. Much of what science has revealed about neurons and their synapses and circuits has come from studies in laboratory animals and disease-specific animal models. Fundamental to these studies is the assumption that the underlying processes supporting brain function are the same across species.
Recent advances in subcellular imaging are now enabling scientists to explore this assumption at the level of individual neurons and their connections. The biophysicist investigators will use two-photon imaging in human brain tissue that their surgical colleagues at Massachusetts General Hospital have removed from patients undergoing surgical treatment.
Using two-photon imaging in these tissues, the investigators will analyze basic information processing properties of human neurons and their connections in intact neural circuits. Essentially, they will visualize the signaling of individual neurons at each neuron’s single synapse. Through this process, they will investigate whether the computational building blocks of brains are shared across humans and animal models. If not, the study may bring us one step closer to understanding and addressing limitations in identifying therapeutic targets, testing experimental pharmacological interventions aimed at those targets, and translating those findings into human clinical studies.
Significance: Understanding species differences in neuronal circuitry could lead to new innovations in translating treatment results from animal model studies to human clinical testing.
Imaging Synaptic Computations in Human Neural Circuitry
Why humans exhibit such highly advanced cognitive abilities compared to other mammals is unknown. Our brains are not the largest in the animal kingdom, suggesting that absolute number of neurons is not the underlying factor. All mammalian brains exhibit similar gross anatomical principles, with the same basic brain structures found in similar locations, indicating that differences in overall brain design are an unlikely explanation. Functional imaging and electrical recording report broadly comparable signals in analogous regions across species, implying a coarse similarity of brain operation. However, there are intriguing hints from comparative anatomy that fine-grained differences at the level of single neurons may contribute to evolutionary increases in brain processing power across species. This proposal tests the hypothesis that humans exhibit unique computational specializations at the level of the brain’s information processing subunits, single neurons and their synapses. This hypothesis questions a fundamental assumption in modern neuroscience: mammalian brains are generally modular and results from one species are applicable to another. The vast majority of experiments on synaptic and cellular function have been conducted in animal models (mainly rodents but also cats and monkeys). The generalizability of these results to humans has not been rigorously tested. This is increasingly important as animal model systems are being used to guide development of pharmaceutical treatments for human psychiatric and neurological disease. These translational efforts often yield disappointing results, potentially due to species-specific differences in the mechanisms of synaptic transmission and integration. The experiments in this research project directly address these issues by analyzing the computational properties of single human neurons and their synapses in living tissue obtained from neurosurgical patient resection. My laboratory has recently performed our first human brain slice experiments, where we collected 2-photon imaging and simultaneous patch-clamp electrophysiological data from many human cortical neurons. Using this approach, we propose to address the following specific aims: 1) Quantify the fine-scale functional properties of single human synapses; 2) Determine the mechanisms for local integration of multiple synapses in human dendrites; 3) Characterize compartmental interactions across the dendritic tree in human pyramidal neurons. These experiments have the potential for broad and transformative impact: there are no existing analyses of the biophysical properties of human synapses and dendrites, making an exploration of this area truly novel and pioneering. The results from these investigations will substantially expand our knowledge of the processing power of the human brain, directly informing a number of ongoing large-scale neural network modeling endeavors, and may as well lead to the establishment of an entirely new model paradigm for further study of the physiology of the human brain and the dysfunctions associated with neurological and psychiatric disease states.
Mark Harnett, Ph.D.
Mark Harnett joined the Department of Brain and Cognitive Sciences and the McGovern Institute for Brain Research at MIT as an assistant professor in 2015. He received his B.A. in Biology from Reed College in Portland, Oregon and his Ph.D. from the University of Texas at Austin. Prior to joining MIT, he was a postdoctoral researcher at the Howard Hughes Medical Institute’s Janelia Research Campus in Ashburn VA. Professor Harnett’s research at MIT focuses on understanding how the biophysical properties of neurons and their circuits produce the computations that drive behavior. His research program uses an integrative approach to connect synaptic and cellular mechanisms with network dynamics during task performance.