Sparks Brain Preservation

The recent Memories thread prompted me to draft a conceptual proposal, which I began to see as being based on a possible complement to the Future Technologies page.
At the end of that page, I read: “Regrowing a new body would be by far the easiest of any of the technologies listed.”
My proposal is based on something similar, but my aim here is not to repeat what is already well explained there, nor to prescribe methods, but just to open a community discussion about a rather different approach than simply “growing an entire new body around an existing repaired brain”.

This needs to be preceded by some background.

• Timing and target for any hypothetical memory transfer: if technologies for transferring memory or identity related information were to become available, it may be preferable for such transfer to occur mostly during the growth of an embryo/fetus rather than into an adult substrate. The embryo in question would be a genetic clone of the patient, since such a clone would plausibly be more predisposed to host the patient’s biological characteristics and memory fragments. Only a portion of personal traits — mainly character and biological predispositions — would be implanted early; the bulk of long term memories would be added gradually in multiple sessions during growth. During the gestational phase, it would be possible to envision an easier version of the rebuilding described in the section Molecular Scan and Rebuild, limited to the most important areas and/or morphological patterns of the brain that are truly identitary, and to those predisposed for future memory implants. This process would require the individual to remain in an extended virtual dormancy until maturity (in a way reminiscent of the “pods” in the film The Matrix: while continuing to be judged “in favor of pragmatism” … “it is inevitable”). As a personal hypothesis, I believe short term memory is lost after death due to the shutdown of neuronal electrical activity, comparable to memory loss after trauma (much like RAM in a computer, which is erased when power is cut): therefore, it should be preserved — whether intact or through partial revival and reconstruction from related information — on a medium distinct from a dead brain…
• Ethical and legal considerations: these topics raise profound ethical, legal and social issues. Being the patient recognized as the donor of the clone, the embryo would be not a child of an unrelated couple, so the proposal does not rely on or disturb existing families. To avoid involving a third party uterus, this discussion assumes the use of an artificial gestational environment. Decisions about preserving additional tissues should be governed by clear informed consent processes, transparent policies on future use, and oversight by appropriate ethics committees.
• Central importance of extracerebral somatic cells: the preservation of somatic cells as fibroblasts, in small tissue samples under cryogenic conditions, is a central component of any long‑term strategy for cloning, because they are easy to obtain, retain intact nuclear DNA with the complete genetic information of the patient, are well suited for nuclear transfer as demonstrated in animal cloning (e.g., Dolly the sheep), are already used in research (e.g., reprogramming to iPSCs), and their nuclei are more resistant to hypoxia and cryogenic stress than fragile neural cells.

PROPOSAL

1. Offer an optional add‑on for members who wish to preserve small additional tissue samples (explicitly including fibroblasts), with clear limits and documented consent. In the past, Oregon Cryonics employed classical cryogenic preservation of whole brains. The practical and scientific difficulties of that approach convinced me to embrace the more realistic and scientifically grounded method of chemical fixation, which Sparks Brain Preservation now fully adopts. However, I believe there is value in re-using cryogenic principles for very small tissue samples. The cryostats once bought for whole brains could, if repurposed, accommodate thousands of patients per unit, since these tissue samples occupy minimal space. This would make cryogenic preservation logistically feasible and cost effective, while complementing the chemical fixation protocol for brains.
2. Provide an option to store durable digital archives co-located with the biological deposit (few GB, curated). This would be a straightforward example pointing in the direction of the abovementioned “partial revival and reconstruction” of short-term memory. Beyond the fact that everyone would be happy, and proud, to re-explore their own works, it would be meaningful to revisit their “work notes” (sometimes proper “work in progress”) to give more depth to their lives. They would help reconstruct short-term memories, especially when combined with the help of images and videos, whether personal ones or, with a different impact, public ones recalled by reading, among their own works or specially saved biographical notes, which were their favorites. Therefore, in addition to biological samples, members could deposit durable digital archives (such as a single M-DISC) stored adjacent to the biological deposit under the same stewardship. While external services exist for indefinite digital preservation (such as SecurSafe), co-location ensures unified governance and redundancy. We are not talking about terabytes, but only a few gigabytes of curated data — enough, at least for now, to preserve one’s works, favorite memories, and essential identity traces.

Thanks for your suggestions and perspective. I will offer my personal thoughts on the topic:

1. Regarding preserving fibroblasts, I don't really see what the benefit is. DNA is also preserved by fixatives, it's just difficult to access the DNA for sequencing with current technology. There are some chemical changes as well, but averaged across millions of cells it seems very easy to figure out what the original chemical states were. Perhaps having access to non-fixed tissue would allow for regrowing a new body to be done more rapidly, but given the scale and complexity of the problems involved in revival, I think extracting usable DNA from fixed cells would be a relatively minor challenge compared to everything else that needs to be solved. And adding on the storage of a whole different type of tissue and modality of preservation adds a lot of challenges and complexity to our operations.

2. The durable digital archives idea is a good one. I agree with your points about this. I personally would find it valuable. Of course, I'm not sure how exactly this would be done in practice.

Thank you for your thoughtful feedback. I agree that, in principle, DNA preserved in fixed tissue can be recovered and sequenced, and that the technical hurdles of extraction are relatively minor compared to the broader challenges of revival.

That said, fibroblasts are not just “DNA containers.” They are living, cultivable cells, and this makes them a fundamentally different resource. Having viable fibroblasts means you can actually grow biological material in vitro, not merely read DNA from fixed tissue. This opens pathways toward cloning or cellular reconstruction that sequencing alone cannot provide. And unlike the idea of simply producing “a new body more rapidly”, cloning from fibroblasts offers the possibility of generating an identical body (already predisposed to a very similar brain architecture, which makes artificial implantation of memory and modification of brain features more feasible), preserving continuity in a way that fixed DNA alone cannot guarantee.

I acknowledge that maintaining a separate preservation modality adds logistical complexity. For this reason, fibroblast preservation should be framed as a paid premium option. In this way, the operational burden is compensated, and those who want to maximize future possibilities can choose it, while the standard service remains streamlined.

In short: fibroblasts are valuable precisely because they are cells you can grow, not just DNA you can read.

If we could grow a new body around an existing brain, we would only be about half-way to the technology we would need for revival. So there would be no point in growing a new body yet because it would be too early. By the time we had the technology to repair a brain, growing the new body would be very old technology. At that point, there's just no conceivable way that preserved fibroblasts would be at all useful because they could just whip up a brand new living fibroblast on a whim. Fibroblasts would only be useful in the near future when we would still have low technology rather than for revival.

Yes, I love the idea that a clone would have very similar brain architecture. I think that the DNA that determines brain architecture could form a really strong foundation that connectome inference could build on. In other words, if there is ambiguity when tracing neurons, the DNA places certain limits on which paths are correct. Revival is going to be all about inferring the original structure and DNA will play a huge role in that. But I think all that inference will be done in software. I don't think it will involve a clone brain. That seems a lot harder than just fixing or duplicating the original. But if that does turn out to be the easier path like you envision, they will easily be able to do it.

I hope you are right when you say “they could just whip up a brand new living fibroblast on a whim.” If that turns out to be the case, then it would simply be a matter of sharing opinions, and I could easily agree with what you “think” and what it “seems” (e.g. “all that inference will be done in software, not involving a clone brain, that seems a path a lot harder than just fixing or duplicating the original”).

In fact, my point was only to suggest that future brain rebuilding might benefit from an option that reduces, by one or two orders of magnitude, “the controlled movement of about 10,000,000,000,000,000,000 molecules” (to quote your Future Technology page). In other words, fibroblasts could serve as a shortcut to ease some of the overwhelming complexity. This would not require changing your organization’s protocol, but could be possible only if your initial assumption about fibroblasts proves correct.

But if what you envision proves to be wrong, the responsibility would be similar to what Robert Ettinger faced when he proposed that a functional brain (and even its original memory) could be reconstructed by nanorobots from the information stored in a vitrified brain. His vision was inspiring, and the sheer amount of information still preserved in a vitrified brain made his assumption plausible, but it underestimated how fragile the connectome really is, and how easily its fine structure could be lost. (Instead, you were great in understanding that chemical fixation preserves it a lot better, although you were supported by more updated scientific knowledge).

As far as I know, the nuclei in fixed neural cells can be so damaged by chemical fixation that even millions of them may not suffice. That is why I hope you or your colleagues can convincingly explain — to someone like me, without the necessary technical expertise — why it should be guaranteed that damaged DNA in millions of fixed neural cells is still enough to “whip up a brand new living fibroblast” with the complete original genome in good “health”, despite the fragility of nuclei even to seemingly harmless radiation.

Note:
Maybe, I have to specify why I used “health” between quotation marks. Here, “health” flags uncertainty: nuclei reconstructed from chemically fixed cells may be unstable, because fixation induces cross-linking, fragmentation, and chemical modifications of DNA that compromise sequencing fidelity and nuclear integrity. Multiple studies on FFPE samples show that such damage can hinder the reconstruction of a nucleus equivalent to the original, despite advances in next-generation sequencing and correction protocols.

Supporting references

• Nucleic Acids Research (2023) – Formalin fixation introduces fragmentation, cross-linking, and artefacts that reduce sequencing fidelity, limiting the reliability of genome reconstruction from FFPE samples.
• Nature Research Intelligence – Reviews how degradation and artefacts remain a known limitation in FFPE DNA, only partially mitigated by advanced sequencing protocols.
• Radiation Research (2004) – Shows that neural precursor cells are highly sensitive to ionizing radiation, with effects on cell cycle checkpoints, apoptosis, and oxidative stress, highlighting nuclear vulnerability in neural tissue.
• UNSCEAR Report (2000/2020) – Documents natural background radiation (radon, terrestrial isotopes, cosmic rays) as a constant low-level exposure, relevant even in storage environments such as basements or laboratories.

I'm glad you're thinking hard about this. Most people take the easy path and just don't even bother. I'm not sure if you've seen this yet:
https://www.sparksbrain.org/images/crosslinks400.jpg
I created that to help visualize how crosslinking works. The red molecules are formaldehyde. That image shows 4 crosslinks between some proteins and a DNA strand on the right. That is the approximate density of crosslinks. Some proteins would have no crosslinks; most would have one or two. Crosslinks can only happen where two molecules are already nearly touching. So looking at that image, does the DNA still look readable to you if you had some higher technology? I would say yes. I would say it's not damaged at all. So what we're saying is that DNA is easily readable to high tech after crosslinking, but you just can't sequence that DNA with current technology because it has that protein stuck to the side of it.

I now see that I underestimated the probability of what future technologies might achieve, but I was proposing just a “ready for use” abundant, high redundancy archive of the genome, that simultaneously acts as a duplicator of the original body, with predisposed and already very similar brain architecture. This avoids the need to “whip up a brand new living fibroblast” (hoping “on a whim” for them, or better for that future AI) and also avoids any repairing from crosslinks (or better, a reconstruction from a still “readable” biologic source — “on a whim” too?): this work saving is not just a request for thanks from future restorers, but, for example, if preserving fibroblasts will become a normal practice for people in a few decades, older preserved brains could then be prioritized in a different way by the future restorers (perhaps to be done much later, or, in the worst-case scenario, never).

Also, although your image makes clear that future methods can read through aldehyde crosslinks, the cryopreservation of fibroblasts halts enzymatic activity and preserves much of their regulatory architecture which chemical fixation distorts (haplotype phasing, epigenetic marks such as methylation profiles, and chromatin topology which importantly influences which genes are accessible or silenced), in a way that cannot be faithfully interpolated or inferred even from billions of fixed cells.

Anyway, I appreciate your keeping the focus on brain chemical fixation, so let me connect my thread to the broader issue of memory preservation in the chemically fixed brain. I see that in pages like https://sparksbrain.org/scientificBasis.html there is written much about “memory reconstruction” but only vague or insufficient arguments about “memory preservation”, which I only realized yesterday is a completely different issue from “Structural Preservation”. I realized I could add to your sentence “reconstructing memories based on tracing the pathways, otherwise known as the connectome” other one of yours: “… as long as memory is not encoded in smaller structures than is currently believed”. Now I think that these “smaller structures” exist on a scientific basis… To be clearer, I had assumed that long-term memory would be completely available from the connectome of the chemically fixed brain, but I was disappointed to realize that this assumption was too optimistic: the molecular states that encode synaptic weights are altered or lost, so the probability of reconstructing memory is much lower than I had believed.

Perhaps I should prepare a more detailed thread entirely on memory preservation, or a post in reply to threads like “Cryosphere Discord server, Chemical Fixation vs. Vitrification”, but perhaps I will anticipate the main points here tomorrow.

About my last post, I now recognize that I misrepresented certain scientific aspects. I wrongly attributed fibroblast‑specific features — such as transcriptional phasis, nuclear topology, and epigenetic regulation — to the brain, whereas they actually pertain only to fibroblasts themselves and would not contribute to similarity in the cloned brain. In this sense, the current protocol used by SBP is sufficient if one relies on a maximized contribution from possible future technologies, but it should be understood within its proper limits.

To repair that contamination, here is a corrected and carefully structured comparative synthesis in English, focusing strictly on the scientific aspects of cloning strategies.

Comparative Synthesis: Fibroblast Nuclei vs. Reconstructed Neuron‑Derived Nucleus

Interesting discussion. I just wanted to respond to a couple of the points.

> To be clearer, I had assumed that long-term memory would be completely available from the connectome of the chemically fixed brain, but I was disappointed to realize that this assumption was too optimistic: the molecular states that encode synaptic weights are altered or lost, so the probability of reconstructing memory is much lower than I had believed

Yes, I agree that biomolecular information will most likely be necessary (or at least, this should be our assumption), not just the geometric aspects of the connectome. However, it is not the case that fixation destroys all of these molecular states. In fact, for preserving the spatial location of biomolecules, chemical fixation is the gold standard method -- see for example, immuno-electron microscopy studies.

> the cryopreservation of fibroblasts halts enzymatic activity and preserves much of their regulatory architecture which chemical fixation distorts (haplotype phasing, epigenetic marks such as methylation profiles, and chromatin topology which importantly influences which genes are accessible or silenced), in a way that cannot be faithfully interpolated or inferred even from billions of fixed cells.

It is also not the case that chemical fixation destroys epigenetic information. This is why chemically fixed tissues can be used for epigenetic profiling. To take one example of very many, see here: https://pmc.ncbi.nlm.nih.gov/articles/PMC4065180. For a broader review of the molecular effects of one fixative (glutaraldehyde), here is a review: https://osf.io/preprints/osf/8zd4e_v1

I've replied on the Memories thread.

I have thought of a possible future update to my proposal. It is based on the extraction of a few neuronal colonies, with a total number of neurons below one million. This could be possible in the not-too-distant future through equipment similar (if not identical but adapted to be multifunctional) to those Jordan Sparks describes in the thread Brain Implants (itself based on a thread from a few months earlier: Kurzweil uploads). Of course, one would need to find a way to select those neurons that do not compromise brain functionality, by preferentially identifying neurons that are naturally approaching apoptosis, while still in good health, and especially those with the most complete and representative nuclear DNA, together with representative non-genetic nuclear features (such as transcriptional phasis, nuclear topology, and epigenetic regulation). This would avoid functional damage beyond what occurs physiologically in every healthy brain (through devices that might select neurons differently but would still cause less damage than today’s Neuralink devices, as Jordan Sparks rightly points out).

The point is that if the technology were not yet mature enough for a complete brain upload, an ideal solution could be to use such devices to extract some living neurons before performing the protocol of chemical brain fixation. Using technologies that might even be very close to those available today, starting from a nucleus of these neurons — already selected but perhaps also modified (with techniques we do not yet possess) to adapt it further to the “average nucleus” through information available statistically from other neurons (transcriptional phasis, nuclear topology, and epigenetics) — one could then implant it directly into an enucleated oocyte (in the header line of the scheme in my previous post I wrote “Reconstructed fibroblast nucleus” as an example “in a whim”: now I’ve corrected it by changing to “Reconstructed oocyte nucleus”). This embryo would then produce a brain as similar as possible to the original, starting from the genetic and non-genetic information of individual original neurons.

The next step, once full brain upload becomes available, would be to control the embryo’s growth process through colonies of microrobots (note: not yet nanorobots for molecular level interventions) that constantly adjust mechanically the cerebral morphologies as they form, making them more similar to the original. After all, mechanical alterations are the main morphological differentiators between monozygotic twins even during human gestation. Perhaps it should also be considered that the simple three dimensional molecular fabrication of an entire brain — as Jordan Sparks imagines (“the controlled movement of about 10,000,000,000,000,000,000 molecules”), perhaps via a kind of 3D printer — might face mechanical stability problems (even in a theoretical absence of gravity), such as structures too soft, if not liquid, that cannot “stay put” while construction is underway. Instead, having a scaffold already in place — one that progressively adapts to growth and does not need to be removed, as a predisposed brain can be considered — could reduce by up to four orders of magnitude (being optimistic) the immense work required to correctly move all those molecules. I believe that in the future we might realize that unjustifiably postponing these considerations was simply due to laziness.

Postponing these considerations will be required because of low tech. If all we have is microrobots and not nanobots, we won't be anywhere close to being able to upload. It would be too soon to worry about revival. Most of this post seems to be about how to revive earlier. Our little ape brains are not nearly smart enough to start trying to solve that problem. Some heavy AI lifting will be required. Growing new cells is way down at the bottom of the list of problems that we will need to solve. It's super easy compared to the repairs.

Thanks Jordan, I see your point that some of these technologies are still far in the future. My intention was not to suggest premature protocols, but simply to outline a possible complementary safeguard that could be considered alongside fixation.