I’ve spent some more time reading into the electrophysiological literature, and I’d value your perspective on a remaining concern.
It seems clear that the precise intra- and extracellular ion concentrations around each neuron are important for its electrical behavior. Even small, physiologically plausible changes in the extracellular fluid (for example, switching from native human cerebrospinal fluid to a standard artificial CSF solution) have been shown to alter resting membrane potential, firing threshold, spontaneous firing rate, and overall network dynamics in brain slices.
In that light, I wonder whether it is realistic to expect that simply replacing the original brain fluid with a standardized solution would allow a reconstructed or repaired neuron to fire at precisely the same moments and under the same conditions as before. In a network of billions of interconnected neurons, even modest shifts in excitability could change how electrical activity propagates through the system.
To illustrate my thinking, I keep coming back to an analogy with a precisely measured 1.526472652 V AA battery. If one were to wash away its original electrolyte, preserve the empty shell, confirm perfect structural preservation via electron microscopy, and then refill it with a standard electrolyte mixture that produces, say, 1.500000000 V, the result would still be a fully functional battery of the original size and shape. Yet its electrical properties would no longer be identical to the pre-wash state. In a complex circuit, that small difference could matter.
Neurons, of course, are far more sophisticated than batteries — they are living cells capable of homeostatic adjustment. Still, they achieve their remarkable function by propagating electrical signals through an orchestra-like harmony across vast networks.
Replacing or removing the original brain fluid in a healthy living brain (or during a preservation process) therefore seems likely to alter the network’s original capabilities, at least to some degree.
I’d be very interested to hear how Nectome thinks about this aspect — particularly how future reconstruction or emulation would ensure that the dynamic electrical properties and precise firing behavior of the original network are faithfully recovered from the preserved static structure.
Thank you for the great article. Is it in your opinion irrelevant whether the “ion bags” empty out during chemical preservation, changing the electrochemical properties of the neuron during chemical preservation but leaving the observable structure intact?
Compare preserving an AA battery which has a voltage of -say- 1.516472762 Volt before preservation. All ion chemicals leave the battery during preservation, leaving the structure behind which can be viewed by electron microscopy to be perfectly preserved. However, post-preservation this AA battery will never produce 1.516472762 Volt anymore, the structure having been preserved being irrelevant compared to it’s electrochemical properties.
I like to think of the world in terms of theories and what they predict, and I call the ones that predict a failure of preservation because of loss of dynamic state "blackout theories".
I think blackout theories are sensible a priori. We might indeed live in a world where dynamic state really matters! Our brains could be like the Exhalation aliens. But when I look at evidence like DHCA, ischemia, spreading depolarization, drugs that affect membrane conductance, electroconvulsive therapy, deep anesthesia, and others, I see a consistent pattern: the brain seems to be "restartable" from scrambled or zeroed-out electrical states.
If some variant of blackout theory was correct, I'd expect that DHCA wouldn't work, that it would render someone permanently catatonic or erase their memories or something. I'd also expect synapses to look a lot more uniform and barely change when a long-term memory is formed: why would they, when the information is being stored in the ion distribution? Since I don't see these things, it weighs heavily against blackout theories in my mind.
Thank you Aurelia and Borys, for your extensive answers and describing the situation on a very high level, with strong analogies presented.
We can go a bit deeper into the biological and electrophysiological aspects of the problem I am seeing, yet perhaps you can convince me it’s not an actual problem.
During living circumstances, i.e. a person is alive and conscious or has been shortly unconscious, there’s intra- and extracellular ion concentrations around a neuron which together with neuron physiology determine the mV’s across the cell membrane of the neuron. All of this is constantly in motion since the neuron is operational within the living neural network. K+ and Na+ ions move around all the time, and even if the person has been unconscious for a while with neural damage, the ion concentrations may have shifted intra- and extracellularly, temporarily changing the resting potential in mV’s ( micro voltages ) across the cell membrane. The neuron will not fire the same as before, or it will fire at different moments. Yet, given sufficient recovery time, the ion pumps will bring the intra- and extracellular ion concentrations back into their original balances, and the resting potential across the entire cell membrane slowly recovers back to it’s original state, i.e. the neuron will work again as before, ultimately under the best scenarios. It will recover to a point where it fires and not fires precisely as before, for example as compared to before an accident.
All of this recovery occurs because the extracellular concentrations of ions remains similar as before. The original ions are still there, physically around the neuron, they just moved from inside the neuron to outside the neuron and may have moved around mildy. During recovery, the ion pumps will restore the original concentrations of intra- and extracellular ions of every single neuron, or most of them.
This is the way I understand it works, perhaps this is incorrect to a certain degree.
During a brain preservation procedure, all the original ion concentrations are washed; both the intra- and extracellular ions are gone, meaning the neuron suddenly is preserved in an empty environment. The neuron in principle can be recovered, copied or restored in some fashion, but how can the microvoltages / resting potential across the cell membrane be restored if it’s totally unknown what original concentrations of ions were present around the neuron? The ion pumps need to restore the original state / resting potentials across the cell membrane but from an environment of newly introduced ions. I am concerned the intra- and extracellular ion concentrations can thus not be restored to their original balance, changing the resting potential across the entire cell membrane and the moments when this neuron fires or not fires within the neural network then changes dramatically.
Won’t losing all the original ion concentrations in or around a neuron then not lead to a disfunctional brain?
This may explain my somewhat bad analogy to preserving an AA battery. When ions are just gone through a washout procedure, and their original concentrations unknown, after the preservation experiment, how can the very exact pre-preservation voltage of the battery be restored? A battery cannot recover itself to it’s original state like a neuron can, but still the original concentrations of ions would be needed in the external environment of the neuron to *allow* the neuron to self-restore, or am I perhaps mistaken?
———
AI suggests I rephrase my question above, to you, as follows to make it clearer:
———
Thank you again for your detailed explanations and analogies — they help a lot in understanding the high-level picture.
I’d like to go a bit deeper into one specific concern I still have, and I hope you can convince me it’s not a real issue for your approach.
In normal living (or recovering) brain tissue, the resting membrane potential (~ -70 mV) and the precise moments when a neuron fires or doesn’t fire are determined by the intra- and extracellular ion concentrations (especially Na⁺ and K⁺), maintained by ion pumps. Small shifts in these concentrations temporarily change excitability, but given enough time and a similar extracellular environment, the pumps can largely restore the original gradients and the neuron can behave much like before.
In the ASC procedure, however, the blood is washed out and replaced with fixative (glutaraldehyde) and later cryoprotectants. This removes essentially all original intra- and extracellular ions, so the original concentration gradients and microvoltages across every cell membrane are lost. The neuron is then preserved in a fixed, non-living state.
My question is: when future technology attempts to reconstruct or emulate the person (whether biologically, via molecular nanotechnology, or as whole-brain emulation), how will it restore or replicate the correct resting potentials and firing thresholds for each individual neuron?
If the exact original ion concentrations around each neuron are unknown, and we introduce new (even “healthy”) concentrations, won’t the steady-state resting potential of many neurons shift slightly? And if so, wouldn’t that change their excitability, causing them to fire at different moments or under different input conditions within the network — potentially altering behavior, memory recall, or personality in subtle but important ways?
In short: isn’t losing the original ion milieu analogous to washing out the electrolyte in a battery and then refilling it with a similar but not identical solution? The battery (or neuron) might still “work,” but not with exactly the same voltage and performance characteristics as before. Or am I missing something fundamental about how the information is encoded and will be decoded from the fixed ultrastructure?
I’m genuinely curious how Nectome thinks about bridging this gap between the preserved static structure and the dynamic electrical properties that were present pre-preservation.
There is a risk of drawing false analogies here, but I'll give it a try. I'll start with the caveat that the brain does not work like a digital computer with a van Neumann architecture or like a CPU in the computer I am using now. BUT to care about the dynamical state of the neurons in the electrophysiological sense is a bit like insisting that in order to emulate a CPU or to reverse-engineer it one would need to know for sure what were the values stored in the registers before it was switched of. It is not that they are irrelevant for the operation of the CPU at a given moment, or that the I/O mapping is irrelevant to understanding how the CPU works. What I am saying is that particular values at the time the machine was switched of are irrelevant when one wants to reverse-engineer it or to build an emulation.
Again, using the analogies based on digital computer is very risky here, because it is easy to draw garbage conclusions from garbage analogies. But if we must, it is more productive to compare the neural tissue to an integrated circuit with reconfigurable wiring. An actual reconfigurable circuit, like if the photolithography process could be on-going. So *not* like FPGA even! The circuit performs computations on-the-fly with a fast dynamic, and also self-reconfigures with a slower dynamic. It is how this circuit is configured what matters in terms of long-term memory and other psychological features that it stores or implements. Imagine now you could store such a circuit. With sufficient technology, based on its structure, you can reverse engineer it, or infer how to repair it in situ (if it is not too damaged).
This new (or repaired) circuit will give the same output to the same input as the old circuit (or the one before damage) would. To implement the behavior properly, the circuit needs to be embodied, in the loop with the environment---the output affects the input the machine will get next. I write it because it is easy to forget about it when concentrating only on the computation in the brain. We at Nectome preserve the whole body to make this part of the reconstruction (embodiment) easier.
But going back to the brain/circuit analogy. The fast-dynamic computational process at the time it was stored (or shortly before) does not matter. I mean, all that fast-dynamic that happened over lifetime *does* matter, but "only" in a cumulative sense, because some of it has affected the slow-dynamic reconfiguration process. But for emulation/repair the particular state of the fast dynamic is not important. Back to brains, we argue it is not important what was the fast-dynamical state of individual neurons in electrophysiological sense. Only what was build by the slow-dynamic process matters, the 'circuit properties'. If/when future technology will restore the circuit and makes it work again, it again will be able to have both the fast and slow dynamic. This view is consistent with what is observed after profound hypothermia and rewarming, and after other circumstances that hugely affect the electrophysiological states of individual neurons, across the whole brain.
People (including us) sometimes use the shorthand of 'connectome' for these 'circuit properties', but it is not a perfect term. Perhaps what the future technology will need is more than what is, strictly speaking, defined as neural connectome---more then just the way neurons connect. But our preservation method preservers not only how neurons connect, but also the 3D structure of all cells (neurons and glia included) and their subcellular features, down to the level of macromolecules (proteins, nucleic acids) inside them. So *whatever* the structural basis that underlies the 'circuit properties', aldehydes plus low temperature (with cryoprotectants) are extremely likely to preserve it.
Of course, we cannot be 100% sure about what the future technology will be able (or not) to do. BUT we insist, as Nectome, that what we should be doing now is to preserve the bodies/brains in the best way now possible. By best I mean a way that, according to current knowledge, has the best chance to make the future work of restoration as easy as possible. If we or someone else figures out how to do it better, we will do *THAT*. We want to ensure that we are affecting the structural information as little as possible before the preservation starts, during, and after. We want to be able to demonstrate experimentally, using animal and human brains, that we actually deliver on that, in the setting that is relevant for how humans can be actually preserved. The fast-dynamic state cannot be preserved by any known method---I'd argue it is extremely unlikely that such a method is possible, so I do not plan to work on it. No preservation technique can guarantee that there is absolutely no distortion. What we argue is that our approach introduces as little distortion to structural information as currently possible.
I’ve spent some more time reading into the electrophysiological literature, and I’d value your perspective on a remaining concern.
It seems clear that the precise intra- and extracellular ion concentrations around each neuron are important for its electrical behavior. Even small, physiologically plausible changes in the extracellular fluid (for example, switching from native human cerebrospinal fluid to a standard artificial CSF solution) have been shown to alter resting membrane potential, firing threshold, spontaneous firing rate, and overall network dynamics in brain slices.
In that light, I wonder whether it is realistic to expect that simply replacing the original brain fluid with a standardized solution would allow a reconstructed or repaired neuron to fire at precisely the same moments and under the same conditions as before. In a network of billions of interconnected neurons, even modest shifts in excitability could change how electrical activity propagates through the system.
To illustrate my thinking, I keep coming back to an analogy with a precisely measured 1.526472652 V AA battery. If one were to wash away its original electrolyte, preserve the empty shell, confirm perfect structural preservation via electron microscopy, and then refill it with a standard electrolyte mixture that produces, say, 1.500000000 V, the result would still be a fully functional battery of the original size and shape. Yet its electrical properties would no longer be identical to the pre-wash state. In a complex circuit, that small difference could matter.
Neurons, of course, are far more sophisticated than batteries — they are living cells capable of homeostatic adjustment. Still, they achieve their remarkable function by propagating electrical signals through an orchestra-like harmony across vast networks.
Replacing or removing the original brain fluid in a healthy living brain (or during a preservation process) therefore seems likely to alter the network’s original capabilities, at least to some degree.
I’d be very interested to hear how Nectome thinks about this aspect — particularly how future reconstruction or emulation would ensure that the dynamic electrical properties and precise firing behavior of the original network are faithfully recovered from the preserved static structure.
Looking forward to your thoughts!
Thank you for the great article. Is it in your opinion irrelevant whether the “ion bags” empty out during chemical preservation, changing the electrochemical properties of the neuron during chemical preservation but leaving the observable structure intact?
Compare preserving an AA battery which has a voltage of -say- 1.516472762 Volt before preservation. All ion chemicals leave the battery during preservation, leaving the structure behind which can be viewed by electron microscopy to be perfectly preserved. However, post-preservation this AA battery will never produce 1.516472762 Volt anymore, the structure having been preserved being irrelevant compared to it’s electrochemical properties.
What’s your opinion?
I like to think of the world in terms of theories and what they predict, and I call the ones that predict a failure of preservation because of loss of dynamic state "blackout theories".
I think blackout theories are sensible a priori. We might indeed live in a world where dynamic state really matters! Our brains could be like the Exhalation aliens. But when I look at evidence like DHCA, ischemia, spreading depolarization, drugs that affect membrane conductance, electroconvulsive therapy, deep anesthesia, and others, I see a consistent pattern: the brain seems to be "restartable" from scrambled or zeroed-out electrical states.
If some variant of blackout theory was correct, I'd expect that DHCA wouldn't work, that it would render someone permanently catatonic or erase their memories or something. I'd also expect synapses to look a lot more uniform and barely change when a long-term memory is formed: why would they, when the information is being stored in the ion distribution? Since I don't see these things, it weighs heavily against blackout theories in my mind.
Thank you Aurelia and Borys, for your extensive answers and describing the situation on a very high level, with strong analogies presented.
We can go a bit deeper into the biological and electrophysiological aspects of the problem I am seeing, yet perhaps you can convince me it’s not an actual problem.
During living circumstances, i.e. a person is alive and conscious or has been shortly unconscious, there’s intra- and extracellular ion concentrations around a neuron which together with neuron physiology determine the mV’s across the cell membrane of the neuron. All of this is constantly in motion since the neuron is operational within the living neural network. K+ and Na+ ions move around all the time, and even if the person has been unconscious for a while with neural damage, the ion concentrations may have shifted intra- and extracellularly, temporarily changing the resting potential in mV’s ( micro voltages ) across the cell membrane. The neuron will not fire the same as before, or it will fire at different moments. Yet, given sufficient recovery time, the ion pumps will bring the intra- and extracellular ion concentrations back into their original balances, and the resting potential across the entire cell membrane slowly recovers back to it’s original state, i.e. the neuron will work again as before, ultimately under the best scenarios. It will recover to a point where it fires and not fires precisely as before, for example as compared to before an accident.
All of this recovery occurs because the extracellular concentrations of ions remains similar as before. The original ions are still there, physically around the neuron, they just moved from inside the neuron to outside the neuron and may have moved around mildy. During recovery, the ion pumps will restore the original concentrations of intra- and extracellular ions of every single neuron, or most of them.
This is the way I understand it works, perhaps this is incorrect to a certain degree.
During a brain preservation procedure, all the original ion concentrations are washed; both the intra- and extracellular ions are gone, meaning the neuron suddenly is preserved in an empty environment. The neuron in principle can be recovered, copied or restored in some fashion, but how can the microvoltages / resting potential across the cell membrane be restored if it’s totally unknown what original concentrations of ions were present around the neuron? The ion pumps need to restore the original state / resting potentials across the cell membrane but from an environment of newly introduced ions. I am concerned the intra- and extracellular ion concentrations can thus not be restored to their original balance, changing the resting potential across the entire cell membrane and the moments when this neuron fires or not fires within the neural network then changes dramatically.
Won’t losing all the original ion concentrations in or around a neuron then not lead to a disfunctional brain?
This may explain my somewhat bad analogy to preserving an AA battery. When ions are just gone through a washout procedure, and their original concentrations unknown, after the preservation experiment, how can the very exact pre-preservation voltage of the battery be restored? A battery cannot recover itself to it’s original state like a neuron can, but still the original concentrations of ions would be needed in the external environment of the neuron to *allow* the neuron to self-restore, or am I perhaps mistaken?
———
AI suggests I rephrase my question above, to you, as follows to make it clearer:
———
Thank you again for your detailed explanations and analogies — they help a lot in understanding the high-level picture.
I’d like to go a bit deeper into one specific concern I still have, and I hope you can convince me it’s not a real issue for your approach.
In normal living (or recovering) brain tissue, the resting membrane potential (~ -70 mV) and the precise moments when a neuron fires or doesn’t fire are determined by the intra- and extracellular ion concentrations (especially Na⁺ and K⁺), maintained by ion pumps. Small shifts in these concentrations temporarily change excitability, but given enough time and a similar extracellular environment, the pumps can largely restore the original gradients and the neuron can behave much like before.
In the ASC procedure, however, the blood is washed out and replaced with fixative (glutaraldehyde) and later cryoprotectants. This removes essentially all original intra- and extracellular ions, so the original concentration gradients and microvoltages across every cell membrane are lost. The neuron is then preserved in a fixed, non-living state.
My question is: when future technology attempts to reconstruct or emulate the person (whether biologically, via molecular nanotechnology, or as whole-brain emulation), how will it restore or replicate the correct resting potentials and firing thresholds for each individual neuron?
If the exact original ion concentrations around each neuron are unknown, and we introduce new (even “healthy”) concentrations, won’t the steady-state resting potential of many neurons shift slightly? And if so, wouldn’t that change their excitability, causing them to fire at different moments or under different input conditions within the network — potentially altering behavior, memory recall, or personality in subtle but important ways?
In short: isn’t losing the original ion milieu analogous to washing out the electrolyte in a battery and then refilling it with a similar but not identical solution? The battery (or neuron) might still “work,” but not with exactly the same voltage and performance characteristics as before. Or am I missing something fundamental about how the information is encoded and will be decoded from the fixed ultrastructure?
I’m genuinely curious how Nectome thinks about bridging this gap between the preserved static structure and the dynamic electrical properties that were present pre-preservation.
Looking forward to your thoughts!
Hi Jacob, thank you for your question!
There is a risk of drawing false analogies here, but I'll give it a try. I'll start with the caveat that the brain does not work like a digital computer with a van Neumann architecture or like a CPU in the computer I am using now. BUT to care about the dynamical state of the neurons in the electrophysiological sense is a bit like insisting that in order to emulate a CPU or to reverse-engineer it one would need to know for sure what were the values stored in the registers before it was switched of. It is not that they are irrelevant for the operation of the CPU at a given moment, or that the I/O mapping is irrelevant to understanding how the CPU works. What I am saying is that particular values at the time the machine was switched of are irrelevant when one wants to reverse-engineer it or to build an emulation.
Again, using the analogies based on digital computer is very risky here, because it is easy to draw garbage conclusions from garbage analogies. But if we must, it is more productive to compare the neural tissue to an integrated circuit with reconfigurable wiring. An actual reconfigurable circuit, like if the photolithography process could be on-going. So *not* like FPGA even! The circuit performs computations on-the-fly with a fast dynamic, and also self-reconfigures with a slower dynamic. It is how this circuit is configured what matters in terms of long-term memory and other psychological features that it stores or implements. Imagine now you could store such a circuit. With sufficient technology, based on its structure, you can reverse engineer it, or infer how to repair it in situ (if it is not too damaged).
This new (or repaired) circuit will give the same output to the same input as the old circuit (or the one before damage) would. To implement the behavior properly, the circuit needs to be embodied, in the loop with the environment---the output affects the input the machine will get next. I write it because it is easy to forget about it when concentrating only on the computation in the brain. We at Nectome preserve the whole body to make this part of the reconstruction (embodiment) easier.
But going back to the brain/circuit analogy. The fast-dynamic computational process at the time it was stored (or shortly before) does not matter. I mean, all that fast-dynamic that happened over lifetime *does* matter, but "only" in a cumulative sense, because some of it has affected the slow-dynamic reconfiguration process. But for emulation/repair the particular state of the fast dynamic is not important. Back to brains, we argue it is not important what was the fast-dynamical state of individual neurons in electrophysiological sense. Only what was build by the slow-dynamic process matters, the 'circuit properties'. If/when future technology will restore the circuit and makes it work again, it again will be able to have both the fast and slow dynamic. This view is consistent with what is observed after profound hypothermia and rewarming, and after other circumstances that hugely affect the electrophysiological states of individual neurons, across the whole brain.
People (including us) sometimes use the shorthand of 'connectome' for these 'circuit properties', but it is not a perfect term. Perhaps what the future technology will need is more than what is, strictly speaking, defined as neural connectome---more then just the way neurons connect. But our preservation method preservers not only how neurons connect, but also the 3D structure of all cells (neurons and glia included) and their subcellular features, down to the level of macromolecules (proteins, nucleic acids) inside them. So *whatever* the structural basis that underlies the 'circuit properties', aldehydes plus low temperature (with cryoprotectants) are extremely likely to preserve it.
Of course, we cannot be 100% sure about what the future technology will be able (or not) to do. BUT we insist, as Nectome, that what we should be doing now is to preserve the bodies/brains in the best way now possible. By best I mean a way that, according to current knowledge, has the best chance to make the future work of restoration as easy as possible. If we or someone else figures out how to do it better, we will do *THAT*. We want to ensure that we are affecting the structural information as little as possible before the preservation starts, during, and after. We want to be able to demonstrate experimentally, using animal and human brains, that we actually deliver on that, in the setting that is relevant for how humans can be actually preserved. The fast-dynamic state cannot be preserved by any known method---I'd argue it is extremely unlikely that such a method is possible, so I do not plan to work on it. No preservation technique can guarantee that there is absolutely no distortion. What we argue is that our approach introduces as little distortion to structural information as currently possible.