Reframing formalin

[AndyMcKenzie]

I found an interesting article on formalin that I thought readers of this forum might be interested in: "Reframing Formalin: A Molecular Opportunity Enabling Historical Epigenomics and Retrospective Gene Expression Studies". Link: https://pmc.ncbi.nlm.nih.gov/articles/PMC11887604/

One of the quotes from the abstract: "Thus, exposure to formalin essentially corresponds to taking a snapshot of organism‐wide gene expression at the time of death." They noted that fluid-preserved specimens were especially useful because the specimens can be preserved whole.

I think that understanding the history of this topic might help to explain some of the confusion regarding our use of formaldehyde in brain preservation. It turns out that there are old dogmas suggesting that formalin introduces insurmountable damage to molecules, including those in chromatin. But new molecular biology techniques have shown that those are really due to limitations in extraction and sequencing methods, which are already being overcome, and stand to be further overcome in the future. (In my view, all the way towards allowing for revival, assuming that the brain is preserved well enough and that technology continues to improve.)

[PCmorphy72]

This interesting paper made me think: I would really appreciate if there were a similarly exhaustive paper, but focused on glutaraldehyde instead of formaldehyde, so that we could make a direct comparison. As far as I can tell, there is no literature really addressing brain preservation in glutaraldehyde for “revival” in the sense we (ex-)cryonicists mean, with the specific advantages of its protocols.

Since I could not find such a paper, I tried to anticipate the comparison myself.

Various aspects of the brain (with their contribution to infer long-term memory) in different preservation protocols

Code: Select all

Aspect                   | Glutaraldehyde            | Formalin                  | ACP (Formalin → CPA → LN2)
(LT memory contribution) |                           |                           | 
-------------------------+---------------------------+---------------------------+---------------------------
Connectome/architecture  | Excellent inference;      | Very good inference;      | Very good inference at
incl. astrocytic tiling, | requires initial formalin | less architectural        | fixation; osmotic damage; 
synaptic microdomains    | for vascular penetration; | precision compared to     | risk of cracking if ramping/
(88-92%)                 | possible osmotic shrink   | glutaraldehyde.           | gradients not perfect; 
                         | of tissue.                |                           | otherwise most durable.
-------------------------+---------------------------+---------------------------+---------------------------
Synaptic plasticity      | Inference based on        | Richer inference;         | Robust inference; connectome
(5–7%)                   | connectome + residual     | connectome + residual     | plus residual molecular clues,
                         | molecular clues very      | molecular clues readable. | stable long-term.
                         | difficult to read.        |                           |
-------------------------+---------------------------+---------------------------+---------------------------
Cellular DNA/epigenetic  | 10–40% preserved but      | 40–70% preserved;         | 50–80% preserved as in
coverage *see Note       | strongly crosslinked.     | crosslinks moderately     | formalin; LN2 stability
(2–3%)                   |                           | reversible; demonstrated  | maintains quality.
                         |                           | in FFPE protocols.        |
-------------------------+---------------------------+---------------------------+---------------------------
Theoretical inference of | Difficult reading;        | Moderate access;          | Moderate→high access;
DNA/epigenetics per cell | requires aggressive       | practical access today    | similar to formalin,
(2–3%)                   | de-crosslink protocols.   | (FFPE).                   | less degradation over time.
-------------------------+---------------------------+---------------------------+---------------------------
General                  | Partial; integration of   | Moderate; integration of  | More complete; stable
theoretical inference    | connectome (very high) +  | connectome (very high) +  | integration; nanobot use
                         | DNA/epigenetics limited.  | DNA/epigenetics partly    | not plausible in vitrified
                         |                           | demonstrated (FFPE).      | state.
-------------------------+---------------------------+---------------------------+---------------------------
Preservation duration    | Some decades; theoretical | Few decades; theoretical  | Theoretically millions of years
                         | up to ~200 years before   | up to ~100 years before   | if LN2 stable and cracking
                         | significant degradation   | significant degradation   | mitigated; cracking interrupts
                         | (e.g. −10% connectome,    | (same very approximate    | volumetric continuity, severely
                         | −50% DNA/epigenetics;     | parameters).              | compromising connectome
                         | very approximate values). |                           | inference

Note: La The row “Cellular DNA/epigenetic coverage” refers to the number of cells whose contribution to LT memory inferability (2–3%) depends exclusively on the DNA/epigenetics of the cells and is not already inferable from the connectome/architecture. For clarity, here is the list of estimated shares to be considered (parts of the 2–3% total) for each cell type: neurons (55–75%), astrocytes (7–12%), oligodendrocytes (5–10%), microglia (5–10%), endothelial cells (4–7%), progenitor/stem cells (2–5%).