#SfN13: Day 3

Ed Boyden gave a fascinating lecture today about a variety of methods that he's worked on with a swath of other scientists and engineers. Each method was powerful on its own, and to hear about them all in one talk was quite amazing. Here's an overview of them:

  • Automated patch-clamping. Whole-cell patch-clamp already exists, enabling simultaneous electrical, chemical and gene analysis of single cells. However, it takes a long time for trainees to learn and is very difficult to do in vivo. One of Ed's grad students automated this method. They apply square waves to look at changes in impedance continually as a pipette is lowered into the brain region of interest, detecting nearby neurons. Then, by looking at changes in impedance over a series of lowering steps they can tell that a neuron is approaching and they can reliably and automatically patch-clamp neurons. The equipment for automatic patching is available online at http://www.neuromatic-devices.com.
    • Using this method, their group has looked at how different anesthetics affect single neurons in living animals.
    • Another application is to look at how a synaptic connection changes during a learning task. They've been using a setup that uses 4 individually controlled pipettes at once to record from pre- and post-synaptic neurons simultaneously.
    • He's also teamed up with Allen Institute for Brain Science to do integrative analysis of the different cell types in the brain, including their morphology, electrophysiology and molecular features.
  • Automated animal surgeries. This works by recording the impedance of the drill. Since the skull has a high impedance, it is easy to tell when the drill has reached the inner edge of skull during drilling, and the drill stops. Automating this method has helped them avoid bleeding during surgery. Also, by having a rapid way to make a grid of holes in the skull with this method, one might be able to do high-throughput in vivo pharamcological testing on many patches of cortex.
  • Advanced electrode arrays. They are developing electrode arrays with 120 recording sites to listen to neurons, using algorithms to sort out which neurons each electrode is listening to. He's not sure how many neurons their electrode will be able to listen to, as they are still testing this.
    • The arrays are tiny, with each electrode being only 10s of microns wide, reducing bleeding, morbidity, and mortality.
  • Spike recording with DNA. This was probably my favorite. When spikes occur, calcium flows into the cell. This causes a DNA polymerase that's copying strands of DNA during the recording time to mutate its shape and make mistakes. By looking for mistakes in the copied DNA and comparing it with the rate of replication, they can have a record of spiking behavior. At the moment, this method is too slow to record neural data, but they are working on improving it.
  • Of course he's also worked on optogenetic molecules, of which there are 3 classes: archaerhodopsins and bacteriorhodopsins (pass protons), halorhodopsins (pass chloride ions), and channelrhodopsins (pass protons and sodium, potassium, and calcium ions).
    • Now they're working on making a red light sensitive rhodopsin. Because red light can penetrate to deeper tissue than other wavelengths, it is better suited for deep neural stimulation.
    • Using two new types of rhodopsins, chronos and crimson, they can activate neurons with blue and red light with no measurable overlap, allowing them to stimulate two different populations of cells individually.
  • He also is working on ways to deposit cells in 3-D space to make 3-D cell cultures, in order to analyze, for instance, how connections form between neurons.

I was amazed at all the tools that Ed had to offer, and eagerly anticipate the ones still in development.

Which method do you think is the most promising or exciting? Which method do you think we should prepare for?

2013-11-11 22.56.39-1

Posted November 12th, 2013 in Career, Neuroscience, Science.