Action Potentials for February
1: New study publishes the first large-scale reconstruction of 3D electron microscopy data from the CA3 region of the hippocampus. This is probably my favorite brain region, although I recognize that’s like calling blue your favorite color. They found a higher level of recurrent connectivity than previously expected, which means that neurons can send feedback signals to each other, forming loops of interaction:
Notably, the tissue was fixed by perfusing fixatives through the blood vessels of the mouse’s body. They wrote: “This immediate fixation preserves the structure as it was in its last in vivo state, and therefore, gives a representation of the network with as close-to-real likeness as possible.”
2: New study on changes in longish-term gene expression changes after memory recall. They trained mice with a fear conditioning paradigm, followed by a recall of the fear memories days later, and then analyzed changes in gene expression 9 days after recall. Comparing the brains of mice who did and did not experience the fear memory recall, they found some gene expression differences in Gpr88+ neurons in the basolateral amygdala:
Two genes encoding neuropeptides, Tac1 and Penk, have the largest expression differences. I wonder what the functional role of these genes might be in this context. My guess, from having done some single cell RNA sequencing analysis, is that any study finding differences in gene expression in transcripts between conditions is marking for differences in sub-cell type composition until proven otherwise. And in my book, this study didn’t really prove otherwise.
3: Review paper on tissue clearing methods in engram research. Shows some different types of brain-wide cell-level engram mapping:
4: A cost effective method of perfusion fixation in brain banking. Claims that with perfusion, and at postmortem intervals of 10 to 37 hours that are practically achievable in brain banking, the fluorescent immunohistochemistry staining results are similar to the standard results one would find in transcardially perfused mice:
5: A video of cellular structure in electron microscopy data from a cubic millimeter of the mouse brain:
6: Although brain tissue usually liquefies during the dying process, a surprising number of studies have found that it can sometimes form a material that survives for hundreds or thousands of years. One of the most famous examples is the Heslington brain, which was buried in the skull under the soil in 673–482 BC. The cellular architecture is destroyed, but it’s still a fascinating phenomenon.
Previous authors have argued that these brains might have been naturally preserved due to the formation of adipocere, which is caused by the bacterial hydrolysis of fat tissue. However, a new study uses mass spectrometry and finds that the brain material is actually made of a chemical called kerogen, which is formed by a condensation of proteins and lipids, a type of long-lasting biomaterial that has not previously been described.
7: Mapping dysfunctional circuits between the frontal cortex and the subthalamic nucleus using data from 534 deep brain stimulation electrode placements. And then using this knowledge to predict ways to improve the efficacy of brain stimulation therapy in one patient with Parkinson’s disease and one with OCD:
8: The first US trials of CAR-T therapy for multiple sclerosis are starting. The idea is to kill the B cells expressing CD19, thus resetting this aspect of the immune system. The main questions are whether this approach will kill enough of these B cells and whether it will be safe enough.
9: Encouraging preliminary results for GABAergic interneuron cell therapy in drug-resistant mesial temporal lobe epilepsy.
10: Retrospective study of 196,610 visits by veterans to the ED or urgent care with suicidal ideation or suicide attempt. Finds that in patients with suicidal ideation only or a suicide attempt in the past 2 to 7 days, hospitalization was not associated with any significant effect on suicide attempts in the next 12 months. However, for patients with a suicide attempt in the past day prior to the visit, hospitalization was associated with a reduced risk of suicide attempt, from 6.9% to 9.6%.
Considering what a major aspect of psychiatry this is, there are remarkably few studies on it. I agree with the many who have argued that there should be more randomized trials on psychiatric hospitalization, so this aspect of decision-making can be done in a more evidence-based way. The notion that it is too dangerous to perform this type of trial is totally nonsensical — in fact, it is that form of thinking that is too dangerous.
11: New review finds that in treatment resistant depression, 25 mg of psilocybin has a number needed to treat (NNT) of 5 at 21 days, while esketamine has a NNT of 7 at 28 days post-dose. These are almost certainly inflated because they are reliant on new studies and effect sizes drop over time.
12: Study finds that doses of stimulants above the licensed amount are generally not much more effective and are associated with more participant dropout, indicating a higher side effect burden.
13: Awais Aftab with some good points about the possibility of post-SSRI sexual dysfunction.
14: Study finds that elevated levels of different proteins in the blood can predict dementia up to 15 years in advance:
Seems to be an effect of markers for different cell types, as the most predictive protein (NEFL) is expressed in neurons (mostly axons), the second most predictive protein (GFAP) in astrocytes, and the third most predictive protein (GDF15) is a bit more confusing but might be a microglial marker.
15: Interview with Anders Sandberg and Alex Davidson that touches on cryonics. Anders points out that people often link their health practices, such as exercising or not smoking, to moral virtues, believing they deserve a longer life because of their healthy choices. Even though, in reality, these are just mechanical properties of our bodies. Cryopreservation, being more technical and less emotionally fulfilling, doesn't fit this narrative.
16: Video in Spanish about cryonics by Sandra Ortonobes. Her opinion (translated to English): "Let me know in the comments if you would freeze your body waiting for it to thaw in the future (leaving aside the money it costs, only on a hypothetical level). I think so, because I have nothing to lose if I'm already dead, and if I don't like the future in which I wake up, I can always die again. Although surely in the future we celiacs can eat whatever we want, and of course, who would want to die in a future like this?"
17: A friend of the famous Chinese philosopher Li Zehou reveals that Zehou chose to be cryopreserved at Alcor upon his legal death in 2021. His stated motives for this were non-traditional, as he said that he thought revival would be impossible, but that it might contribute to testing his philosophical concepts. This is a complex story and I’d like to learn more about it.
18: "Cryonics, a Comprehensive Approach" by Sebastian Cippitelli of Buenos Aires, Argenita. Performs patient advocacy for people who want to pursue cryopreservation. Also reports results from a 500-person survey on the public street of randomly selected people, who were 48% female, 49% male, and 3% non-binary. When asked “Would you cryopreserve your body under the possibility of being able to bring you back to life in the future?”, 29% answered “Yes”.
19: Nathan Helm-Burger’s Manifold market question on whether we will have synapse-level digitization of a human brain, which requires the whole connectome, approximate synaptic strength, cell type, myelination degree, and >90% synapse fidelity, by 2035. I bet no at 38%.
20: The Less Wrong 2023 survey results are out. As usual, it includes some questions about cryonics. One is about people’s level of interest in cryonics is (not interested, considering, cryocrastinating, signed up, etc.). Another asks about people’s subjective probability of successful restoration to life in the future, conditional on there not being a global catastrophe destroying civilization before then. This is also known as p(success). I thought it might be interesting to plot these (with the subjective probability estimates on a log scale, of course):
It is true that people who are more interested tend to give higher subjective probability estimates of success (median probability estimates: signed up = 17.5%, cryocrastinating = 30%, considering = 10%, not interested = 5%). But the difference is not very large. There must be other factors that are much more important than p(success) estimates in mediating whether someone is interested in signing up for cryonics and/or actually goes through with it.