Neural multiomics Q&A

Dr. Lee: I don't know the exact percentage of what the overlap is, but we are able to take the framework from our RNA-seq platform and map our cells onto that. What we have found is that when the nucleus is included, we get a nice alignment with particular types, subclasses, and transcriptomic types to the RNA-seq platform.

Dr. Lee: Epigenetic information is possible, since we have the nucleus. We've recently talked about this, but the effort has not yet started. However, we acknowledge that there are epigenetic mechanisms there that we can use to further classify cell types.

Dr. Lee: Yes, we do. What I showed initially was the three standardized stimulus sets that we run. We also have additional stimulus sets that we run on particular types of neurons; we're always adjusting our stimulus sets too.

Dr. Lee: We do have RNase inhibitors and that was quite critical. When we initially started it, we had quite a bit of degradation. We stepped it up to what was currently used in our RNA-seq pipeline of one unit per microliter for the RNase inhibitor.

Due to the components of the RNase inhibitor, that skyrocketed the osmolarity of our internal solution. When we patched onto a cell, the cells would blow up and swell to 10 times the size. We had to methodically lower the RNase inhibitor and settled down to a 0.5 units per microliter. To do that, we had to offset the potassium gluconate concentration in our solution to maintain the proper osmolarity.

Dr. Lee: I think it depends on what type of neuron you're looking at, the diversity of those neurons, and then what modality that you really want to try to incorporate. For example, inter neuron types within the visual cortex, we have six or seven different subclasses and as you saw, 10 15, 20 different met types.

So really to make those, I think that you need hundreds, if not thousands of cells. On the flip side, if we were to go to a subcortical region and we start to see some initial results of the diversity, and the diversity between those types is not as great, then maybe we can make that work with a hundred cells. I think that's something that you have to start evaluating as you collect the data.

Dr. Lee: The Cell paper that we published focused on the visual cortex; this year we're expanding to subcortical regions. As anyone may hypothesize or know from personal experience, the neurons at different parts of the brain are different.

What we've started to realize is that is, at least with a couple of the subcortical regions that we're targeting now (one of them being the nucleus accumbens), the membrane properties are actually different. We have to adjust the parameters such that we're able to extract the nucleus while preserving the somatic membrane. We had a standard set that we use for the cortex, but we've found that that is too strong for some types or maybe too weak for other types.

Dr. Lee: The biggest thing that we're working on right now is the rate of morphological recovery. In our paper, figure one shows that this is our lowest rate of being able to retain high-quality fills, so our goal is to hone our technique to improve that. We are good at being able to classify the cell types and obtaining good transcriptomics. We're good at obtaining the appropriate amount of RNA and thanks to the Takara Bio SMART-Seq kiton it.

Dr. Lee: That's one thing we really had to spend a bit of time to do. We invited the folks from molecular biology to look over our rigs to see how suitable they were for obtaining molecular biology information, and they were actually astonished to see how dirty our rigs are compared to a molecular biology wet lab.

They gave us some tips and techniques, and we just tried to be as clean as possible. We wear gloves, we wipe down all surfaces. Anything that we touch, we change gloves multiple times a day. We use ethanol to make sure our hands are clean. We have brought hoods into the ePhys lab. So we have hoods just adjacent to the ePhys rigs. As we collect the data, we're able to quickly bring it to the hood and deposit it and reduce any type of air contamination or RNase contamination.

Dr. Lee: We're open to collaborations and have many of them going on right now. I think we're open to multiple types, grant related or non-grant related. If someone has a specific disease model they want to look at, but they don't have the ability to do thousands of recordings, that's one collaboration we can assist with: identifying the cell types that they're currently looking at in a disease model.

At the same time, we also have collaborations going where we're simultaneously collecting data. We have a grant with a collaboration in Amsterdam and Hungary where they are performing the same experiments and sending us the samples to process. It is definitely a wide range of collaborations that we currently have going in.