9 ms·

I’ve never gotten a good answer as to how an electron microscope can image the synaptic weights. If you look at the pictures it can barely capture the synapses.

return0 9y ago
electron microscopy has the resolution to discriminate individual spines and synapses, figuring out the weight of the synapse though is not straightforward. At least for excitatory synapses however, there is ample evidence that the size of the spine correlates with the maturity and strength of the synapse. https://www.ncbi.nlm.nih.gov/pubmed/20646057 https://www.ncbi.nlm.nih.gov/pubmed/20646057
- mrfusion 9y ago
  I didn’t know that. That’s a good start. We still have, what 20 different neurotransmitters that can be weighted differently at each synapse?
  - kanzure 9y ago
    You can label receptors with aptamers and other technology, and then sequence the aptamer labels (or whatever). Many in situ sequencing techniques can be applied here. Note, however, that this requires more steps than simply scanning with electron microscope multi-array.
    - mrfusion 9y ago
      Interesting. Maybe also doing some kind of rnaseq synapse by synapse could tell you what levels of neurotransmitters and receptors it’s maintaining. I’m not sure if that small of a scale would ever be possible.
      - kanzure 9y ago
        Yeah, it would certainly help to use a method like expansion microscopy, sort of like if Nectome were to subcontract some work to a certain MIT lab that invented the techni--- oh wait..
  - return0 9y ago
    what determines the conductance of the synapse is mostly the number AMPA/NMDA receptors in the synapse for excitatory and GABAa/GABAb for inhibitory. Then there's neuromodulation, g-coupled receptors etc etc. it's clearly not straightforward.
aurellem 9y ago
> If you look at the pictures it can barely capture the synapses Electron micrographs vary quite a bit in quality (good micrographs are an art), but if you can barely see the synapses, you may be thinking of light microscopy, where it’s very difficult to see differences in the sizes of synapses. You can see for yourself what synaptic details electron microscopy allows you to see in the book, "Fine Structure of the Nervous System: Neurons and Their Supporting Cells". You might also want to see the FIB-SEM images from the Brain Preservation Foundation at https://www.youtube.com/watch?v=RYKIePuVENY https://www.youtube.com/watch?v=RYKIePuVENY, which I find quite beautiful. As for how sizes may relate to weights [1]: “Some axons form two or more synapses with the same dendrite, but on different dendritic spines. These synapses should be the same strength because they will have experienced the same history of neural activity…the synaptic areas and volumes of the spine heads were nearly identical. This remarkable similarity can be used to estimate the number of bits of information that a single synapse can store, since the size of dendritic spines and their synapses can be used as proxies for synaptic strength.” I hope that these are informative. I do want to clarify that we are not arguing that electron micrographs of brain tissue are all that’s needed to reconstruct memories. We are interested in building the best brain preservation technology possible, and one of the ways we evaluate brain preservation is with electron microscopy. The connectome is the first step, the “skeleton” upon which models of the mind can be built. [1] "Nanoconnectomic upper bound on the variability of synaptic plasticity" at https://elifesciences.org/articles/10778 https://elifesciences.org/articles/10778

return0 9y ago

electron microscopy has the resolution to discriminate individual spines and synapses, figuring out the weight of the synapse though is not straightforward. At least for excitatory synapses however, there is ample evidence that the size of the spine correlates with the maturity and strength of the synapse. https://www.ncbi.nlm.nih.gov/pubmed/20646057 https://www.ncbi.nlm.nih.gov/pubmed/20646057

mrfusion 9y ago
I didn’t know that. That’s a good start. We still have, what 20 different neurotransmitters that can be weighted differently at each synapse?
- kanzure 9y ago
  You can label receptors with aptamers and other technology, and then sequence the aptamer labels (or whatever). Many in situ sequencing techniques can be applied here. Note, however, that this requires more steps than simply scanning with electron microscope multi-array.
  - mrfusion 9y ago
    Interesting. Maybe also doing some kind of rnaseq synapse by synapse could tell you what levels of neurotransmitters and receptors it’s maintaining. I’m not sure if that small of a scale would ever be possible.
    - kanzure 9y ago
      Yeah, it would certainly help to use a method like expansion microscopy, sort of like if Nectome were to subcontract some work to a certain MIT lab that invented the techni--- oh wait..
- return0 9y ago
  what determines the conductance of the synapse is mostly the number AMPA/NMDA receptors in the synapse for excitatory and GABAa/GABAb for inhibitory. Then there's neuromodulation, g-coupled receptors etc etc. it's clearly not straightforward.

aurellem 9y ago

> If you look at the pictures it can barely capture the synapses Electron micrographs vary quite a bit in quality (good micrographs are an art), but if you can barely see the synapses, you may be thinking of light microscopy, where it’s very difficult to see differences in the sizes of synapses. You can see for yourself what synaptic details electron microscopy allows you to see in the book, "Fine Structure of the Nervous System: Neurons and Their Supporting Cells". You might also want to see the FIB-SEM images from the Brain Preservation Foundation at https://www.youtube.com/watch?v=RYKIePuVENY https://www.youtube.com/watch?v=RYKIePuVENY, which I find quite beautiful. As for how sizes may relate to weights [1]: “Some axons form two or more synapses with the same dendrite, but on different dendritic spines. These synapses should be the same strength because they will have experienced the same history of neural activity…the synaptic areas and volumes of the spine heads were nearly identical. This remarkable similarity can be used to estimate the number of bits of information that a single synapse can store, since the size of dendritic spines and their synapses can be used as proxies for synaptic strength.” I hope that these are informative. I do want to clarify that we are not arguing that electron micrographs of brain tissue are all that’s needed to reconstruct memories. We are interested in building the best brain preservation technology possible, and one of the ways we evaluate brain preservation is with electron microscopy. The connectome is the first step, the “skeleton” upon which models of the mind can be built. [1] "Nanoconnectomic upper bound on the variability of synaptic plasticity" at https://elifesciences.org/articles/10778 https://elifesciences.org/articles/10778