At the Cosmos Club, Washington, DC
October 21, 2016
President Larry Millstein called the 2367th meeting of the Society to order at 8:05 p.m. He announced the order of business and welcomed new members. The minutes of the previous meeting were read and approved. President Millstein presented a summary of the 29th meeting of the Society, held in 1872. President Millstein then introduced the speaker for the evening, Anthony Zador, the Harrison Professor of Biology and Chair of Neuroscience at The Cold Spring Harbor Laboratory. His lecture was titled “Understanding the Brain's Network Using DNA Sequencing”.
Dr. Zador began by explaining that our consciousness is an emergent consequence of the physical circuitry of the brain. Improving our understanding of the brain’s “wiring diagram”, he said, would help us to heal or augment its processes, or maybe even emulate them entirely within a computer.
Dr. Zador’s investigation into the brain’s wiring began with studying how we make decisions about sensory stimuli. Through training he was able to induce rats and mice to respond consistently to stimuli, demonstrating a change in their brain’s wiring, but Zador and his team had no insight into the actual writing diagram that would allow them to correlate changes in neural circuitry with changes in behavior.
This holds true for benign changes, such as learning, as well as psychiatric disorders, such as autism, schizophrenia, depression, addiction, and bipolar disorder, which we now suspect may arise in part or in whole from disruptions of neural circuity.
Dr. Zador pointed out that traditional psychiatric treatments lack circuit specificity. Most pharmacological interventions treat the brain as “brain soup” and assume that bathing the brain in a particular chemical will produce the desired result. He described this as akin to inferring that someone with a headache suffered from an “aspirin deficiency”. To the extent pharmacological treatments work, he said, they have likely tapped into circuit specificity.
Treatment approaches that target specific neural circuits, such as deep brain stimulation, have been very effective at addressing conditions once considered intractable, such as Parkinson’s, chronic pain, epilepsy, depression, Tourette’s, and OCD.
Thus, even a rudimentary understanding of the brain’s wiring diagram is powerful, and we need to achieve better resolution in order to know what to manipulate in order to precisely affect behavior. The challenge of getting a high resolution neural map is that the human brain has 100 billon neurons interconnected by 100 trillion dendrites in the cortex alone, making it a tangled mess. Techniques such as injecting a bulk tracer can show where a cluster of neurons connects to, but not the exact route or cross connections. Single neuron tracing can provide an accurate map, but is labor-intensive and painstakingly slow.
Dr. Zador’s breakthrough was to recast a difficult connectomics problem as problem of high-throughput DNA sequencing, which we have already solved. By embedding individual neurons with random unique sequences of RNA, he would have an RNA barcode for each neuron. His technique, called MAPSeq, works by creating a library of reengineered viruses that are identical except for a 30 nucleotide sequence. The viruses are injected near the neurons of interest, whereupon each virus infects a nearby neuron and travels down the axon to its connection areas. The brain is then dissected and each location is sequenced to identify the RNA barcodes present, thereby correlating the origination and destination points of the viruses, and thus of the neurons.
An early practical test of MAPseq showed that the wiring of the locus coeruleus—a neuromodulatory center that affects brain state—was more complicated and idiosyncratic than predicted, helping discard several hypotheses and suggest new lines of research.
Future refinements in the MAQseq process will provide better spatial resolution by allowing the sequencing in-situ. This in turn will permit scientists to create a very high resolution “baseline” brain that can be compared to the brains of mice that have disorders such as autism, in order to characterize any systematic miswiring.
Dr. Zador concluded that MAQseq will provide substantial insight into brain function and may point the way to more effective therapeutic techniques, but is unlikely to provide the key to digitally recreating a specific person’s brain in a computer.
After the conclusion of the talk, President Millstein invited questions from the audience.
One questioner asked whether discoveries gained from MAQseq will be useful to building better neural networks. Dr. Zador said that the sheer number of evolutionary tricks innately wired into human brains makes this far too complex to expect results in the short term. For very discrete tasks, however, it is certainly possible that the wiring of a specific function may indeed be able to be mapped well enough to serve as an improved starting point for neural networks.
Another questioner whether MAPseq could assess the relative strength of connections in the brain. Dr. Zador confirmed that future versions of MAQseq will not only reveal connection strength, but also whether connections have recently gotten stronger.
After the question and answer period, President Millstein thanked the speaker, made the usual housekeeping announcements, and invited guests to join the Society. At 10:06 p.m., President Millstein adjourned the 2367 meeting of the Society to the social hour.
|The weather:||Mostly cloudy and unseasonably humid|
External Communications Director
Abstract & Speaker Biography
Lecture Series Index - Home