Thank you for attending the 2021 Showcase Symposium - Day 1 (December 7) or viewing the recording on Youtube .
Below are questions from our audience that were answered with written responses by our speakers. Additional Q&A is at the end of each talk in the event. We invite you to use this thread to ask additional questions or follow up on the conversation more generally.
Question for Tiny Blue Dot team: 2 questions for Irene’s group: 1) why isoflurane? we know that TIVA can be easily titrated in humans when stimulating the brain so might it be a better drug to use and 2) how can you ensure you are not stimulating subcortical pathways (white matter) of the corticospinal tract? that could explain why responses to TMS are more robust (?lower anesthesia effect)
Leslie Claar: 1) we chose isoflurane as a starting point with these experiments because it is widely used in surgical procedures with mice. Though we would like to continue this work by showing that the decrease in complexity generalizes to many anesthetics (as has been shown in humans). 2) At this time, I would say we can’t ensure we are not stimulating subcortical pathways. We have performed some experiments stimulating the cortex in superficial vs. deep layers. In both cases the anesthesia strongly suppresses evoked responses.
Christina Kim, NGL, question for Tiny Blue Dot team: For Ethan: if you modulate the amount of excitation used (less or greater than 500ms stim), do you expect to see a difference in percentages of neurons activated/suppressed? Was 500ms chosen for a particular reason? Is there any difference in timing onset of excitatory or inhibitory responses across cells in cortex?
Ethan McBride: Thanks for the questions! We did use different stimulation parameters, we just focused on 500ms in the talk for time purposes. For example we used a 5ms pulse, which is more similar to what previous studies have used. With 5ms stimulation we saw strong recruitment of FS neurons, but many fewer RS neurons activated. We chose 500ms because in another study from the institute where clasutrum cells were imaged with a miniscope in the context of behavior, some claustrum neurons had long lasting responses up to 500ms or so. And yes, the inhibitory neurons tend to be activated with less latency than the excitatory cells.
Question for the Tiny Blue Dot team: Hi Ethan, great talk - it seemed like you saw a wave of activity expanding with some temporality after your claustrum stimulation. Do you think this was all monosynaptic from the claustrum, or was this some cascade of intercortical projections?
Ethan McBride: Thanks, great question. With our data I don’t think we can know for sure. I do think that the differences in latency are probably mostly driven by the differences in when APs arrive from the claustrum axon, but there could definitely be a contribution of intercortical projections.
Question for Emilia Favuzzi: Hi Emilia, really interesting discussion there re: PV-PV and VIP synapses being distinct from other GABAergic synapses. Do you think this implies the “fundamental unit” of microglial specificity might be “synaptic partners” rather than “neurotransmitter type”?
Emilia Favuzzi: I think the evidence says that GABA is a recruitment signal that starts the process in those specific microglia and activates several downstream molecules needed for them to interact with inhibitory synapses in general but my current hypothesis is that synapse-specific molecules are the ones that ultimately allow the fine discrimination. So yes, although not proven yet, the fundamental unit may be made of a repertoire of e.g. cell adhesion molecules that will have binding partners at specific synapses.
Question for Rusty and RA team: Rusty: for mapping, what is the threshold for high quality vs low quality? For example, would 15% of another type be considered low or high quality?
Rusty Mann: Our highest quality patch-seq data should map to a specific type at least 90% of the time, but we also consider data that maps to a specific type at least 70% of the time to be good quality as well, but like I said before, there are also other QC factors that play a role in this as well.
Mike Lodato, NGL, question for RA team: Great talks! Regarding patch-seq RNA-seq data. Rusty said that patch-seq cells are not clustered alone but mapped to FACS data, because the patch-seq data are noisier. Do I understand correctly that the FACS data are from nuclei, while the patch-seq data are from soma? Could this be driving some of the differences between clustering between the experiments? How do you deal with that?
Rusty Mann: Mike, the main reason for the differences comes the technique for how we collect the patchseq data. We have to navigate a pipette through the brain slice and the pipette can pick up contamination from neighboring cells during that process. Similar when we extract the nucleus and then have to drag it out through the tissue.