Circuit Elements of the Nervous System Ed Boyden Synthetic Neurobiology Laboratory Massachusetts Institute of Technology Invited lecture at Stanford in CS379C on April 8, 2013
Today’s outline PARTS • Cells as engineerable circuit elements • Molecules in the context of cells • Brain regions in the context of cells THEMES • Implications for neuroengineering • What properties of cells make them observable? • What properties of cells make them amenable to being controlled?
Why Neurotechnology? PERSISTENT, ENORMOUS, • Economic Costs of UNMET NEEDS Neuro-disorders: Orange font = top 7 $1.3 trillion/yr in the U.S (greater than 200 million alone people worldwide) • People: 2 billion people Alzheimer's Insomnia worldwide, >100 million Addiction Migraine Americans suffer from a ALS Multiple Anxiety sclerosis brain or nervous system Blindness Obesity illness Chronic Pain Paralysis Depression Parkinson's Epilepsy Schizophrenia Source: Neurotechnology Industry Organization, Oct 2008 Headache Sleep disorders http://www.neurotechindustry.org/ Hearing loss Spinal Cord Huntington’s Injury Stroke
The Numbers Cost (US alone, 2008) People (global, 2008) • Addiction: 790m • Addiction: $366b • Anxiety: 400m • Alzheimer's: $146b • Obesity: 300m • Obesity: $123b • Chronic pain: 290m • Chronic pain: $95b • Migraine: 240m • Depression: $83b • Depression: 240m • Attention deficits: $77b • Sleep: 238m • Sleep: $75b • Hearing loss: 140m • Stroke: $57b • Attention deficits: 120m • TBI: $56b • Alzheimer's: 90m • Vision loss: $52b • Stroke: 60m • Epilepsy: 50m • Hearing loss: $50b • Anxiety: $47b Source: Neurotechnology Industry Organization, Oct 2008 http://www.neurotechindustry.org/
What is the right abstraction layer for engineering the brain? Lewis et al. Denk and Horstmann 2005 2004 Brain regions 1700s-present 1 µ m ‘Precise’ microcircuitry? 1990s-present Ramon y Cajal 1899 ‘Population’ microcircuitry in tissue/slices 1890-present
Goals for a good technology • Make the most meaningful measurements, and interpret them accurately • Make the most precise perturbations of specific substrates, and test the necessity and sufficiency of those substrates • Understand the mechanisms behind symptoms in the most precise way possible • Control a neural circuit in the most powerful way, with the fewest side effects
Today’s outline PARTS • Cells as engineerable circuit elements • Molecules in the context of cells • Brain regions in the context of cells THEMES • Implications for neuroengineering • What properties of cells make them observable? • What properties of cells make them amenable to being controlled?
Cells as engineerable circuit elements • Cell types are defined by geometry, genome state/gene expression, physiology, connectivity, and protein state • Self-contained computational units – Maintain stable resting potential; contain genome; survive autonomously – Go smaller: vastly more incomplete picture; go larger: heterogeneity • Molecules only make sense in the context of the specific cell type they’re in • Brain regions only make sense in the context of the cell types found therein
The Neuron
Molecular components of the neuron • Postsynapse and dendrites – Receptors for neurotransmitters (excitatory - glutamate, inhibitory - GABA, modulatory – DA, ACh, 5-HT, NE; slow vs. fast forms of each) – Integrate signals passively on the membrane, as well as with active conductances (potassium, calcium, sodium) – Prominent plastic elements (kinases, receptors, mRNA) • Cell bodies – Nonlinearity towards a spike • Axons – Sodium and potassium channels: conduction of spike • Presynapse – Calcium channels admit calcium; then, release of neurotransmitter (excitatory, inhibitory, modulatory; short-range vs. long-range) – Prominent plastic elements (release machinery, kinases, cytoskeleton) • Throughout the neuron – Pumps and exchangers to maintain resting potential (Na+ out, K+ in, Cl- out) – Receptors for neurotransmitters – Ion channels – Gap junctions
Cell classes: shapes, location • Unlike many other fields of biology, in neuroscience, “Anatomy is function” • Shapes (Markram et al., 2004)
Molecular markers that correlate with shape/location • Shapes correlate with markers, but not perfectly (Markram et al., 2004)
Cell types: spiking • Physiology (Markram et al., 2004)
Shape, molecules, physiology • Not perfect correspondence (Markram et al., 2004)
Whole cell patch clamp: enables simultaneous measurement of electrophysiology, morphology, and gene expression in single cells in living brain Kodandaramaiah et al. (2012) Nature Methods 9:585–587.
A robot that can automatically patch clamp neurons in living brain Kodandaramaiah et al. (2012) Nature Methods 9:585–587. Commercialized by Neuromatic Devices, Inc. (ESB has no financial affiliation)
The patch algorithm: robotic assessment of sequences of pipette resistances, followed by fast action Kodandaramaiah et al. (2012) Nature Methods 9:585–587.
In vivo robotics: converting an art form to software Kodandaramaiah et al. (2012) Nature Methods 9:585–587.
Derived an algorithm: high-performance recording, with high yields Kodandaramaiah et al. (2012) Nature Methods 9:585–587.
Derived for the cortex, the algorithm works in the hippocampus as well Kodandaramaiah et al. (2012) Nature Methods 9:585–587.
Good access resistances, holding currents, resting potentials, and holding times – independent of depth Kodandaramaiah et al. (2012) Nature Methods 9:585–587.
Integrative analysis of cell types of the brain: molecule to morphology to physiology + Ragan et al., Nature Methods 2012 + gene expression Suhasa Kodandaramaiah, Ian Wickersham, Craig Forest, Hongkui Zeng and Allen Institute for Brain Science
How do the properties of a cell give us a handle on controlling or reading it? • Given a cell, can isolate a promoter – DNA in front of the gene – to target a gene to it – Small promoter – can put in virus – BAC transgenics • Big promoter – knock-ins • Insert your gene into that locus
Labeling something by activity? • Fos-driven feedback loop – Fos = immediate early gene (IEG), turned on by activity (and perhaps plasticity) – Only on for a short time, though. – Solution: turn on a nonlinear switch (Reijmers et al., 2007)
Glial cells • Long thought to be “metabolic supporting” cells, but found to release • Calcium waves can transmitters and spread throughout glial communicate networks, via gap junctions and ATP (Guthrie et • >90% of cells in human al., 1999) brain • Connected by gap junctions, have similar receptors and release machinery to neurons • Don’t fire spikes
Glial cells • Adenosine: • Glial cells release ATP and glutamate (Innocenti et al., – If you block release of ATP, then animals do not develop 2000) sleep debt (normally you sleep more on night 2, if deprived on night 1) – Also, cognitive consequences of sleep loss are ameliorated – Glial release of adenosine to A1A receptors important for effects of DBS? (Bekar et al., • Block release: impairs 2008) synaptic function and • Glutamate release implicated plasticity (Pascual et al., 2005) in neural activity-blood flow coupling (Schummers et al., 2008)
Astrocytes and blood flow • Neuronal activity releases glutamate à astrocyte activation (calcium waves) à release vasoactive substances (prostaglandins) à vasodilation (Zonta et al., 2003)
Astrocytes and blood flow • Neuronal activity releases glutamate à astrocyte activation (calcium waves) à release vasoactive substances (prostaglandins) à vasodilation (Zonta et al., 2003) • Astrocytes are the ‘invisible cell’ in between
Today’s outline PARTS • Cells as engineerable circuit elements • Molecules in the context of cells • Brain regions in the context of cells THEMES • Implications for neuroengineering • What properties of cells make them observable? • What properties of cells make them amenable to being controlled?
Molecules make sense in the context of the specific cell type they’re in: leptin • Leptin released by fat cells, regulates energy homeostasis by action in brain – Potential target for obesity and diabetes • But, leptin receptor expressed in multiple cell classes in the hypothalamus – which is the site? • Delete leptin receptor in one subclass, the steroidogenic factor-1 neurons – neurons are not activated by leptin • Potential brain stimulation target? Clustered in ventromedial hypothalamus …
Channelopathies • Rare, but instructive – Can link molecule to phenotype – Mutations in a single channel can result in many different results
Synapsopathies • A model of obsessive compulsive disorder: molecular deletion of Sapap3, scaffolding protein important for glutamate receptor presence, and synaptic strength (Welch et al., 2007) • Leads to excessive grooming • Striatal intervention reduces phenotype
Recommend
More recommend
