Action Potentials for March
1: Big news this month as Johann Danzl’s group at the Institute of Science and Technology Austria released their work on light-microscopy based connectomics, a new technology that enables dense reconstruction of brain circuitry with light microscopy at synaptic resolution. It is an advance in hydrogel embedding and expansion microscopy. It allows them to directly incorporate molecular information into synapse-level brain tissue reconstructions.
In one demonstration, they were able to reconstruct a ~1 million cubic micrometer volume of mouse somatosensory cortex. To validate the reliability of their tracing, they compared human consensus skeletons to ground truth obtained from Thy1-eGFP mice that have cytosolic expression of enhanced green fluorescent protein (eGFP) in a subset of neurons. They also performed reconstructions in the CA1 subregion of the hippocampus:
What is most groundbreaking about the use of expansion microscopy is that it allows them to use immunostaining to identify the distribution of biomolecules. They trained models to predict the locations of pre-synaptic (Bassoon) and post-synaptic (Shank2) proteins from the structure alone, achieving high accuracy when compared to ground truth immunolabeling data (F1 scores > 0.9). They then mapped the excitatory synaptic input and output fields of individual neurons:
Their method also allows them to draws insights from aspects of the brain’s microstructure that are challenging to address in other ways. They used immunolabeling to identify specific cell types, such as interneurons and astrocytes, and to clarify the identity of subcellular structures like axon initial segments, myelinated axons, and primary cilia. They also demonstrated the ability to map electrical connectivity via gap junctions in addition to chemical synapses:
Ten years ago, there were debates about whether whole brain emulation could ever work because electron microscopy could only perform profile a subset of biomolecules, and ultrastructure alone might not be sufficient. With the advances in expansion microscopy, culminating in this study so far but sure to advance much further in the future, it’s hard to make that argument today.
2: Study uses focused ion beam-scanning electron microscopy to generate 3d images from the prefrontal cortex of a human brain donated 6 hours after death. Ultrastructural features reflecting synaptic function and activity, like active zone and post-synaptic density sizes, were well-preserved in the postmortem tissue. This finding is consistent with the idea that brain ultrastructure is preserved in the early postmortem period. It is the first study I’m aware of that has used volume electron microscopy to demonstrate this.
3: Study finds that light and sound stimulation at the gamma rhythm frequency of 40 Hz promotes the clearance of amyloid from the brains of Alzheimer's disease model mice via the glymphatic system. The clearance was associated with increased efflux of interstitial fluid and was mediated by vasoactive intestinal peptide interneurons that regulate arterial pulsatility.
I don’t really understand how to evaluate this paper. There was a non-replication of the main finding last year, covered here at NN. One of the corresponding authors’ arguments at the time was that there is a distinction between stimulation-evoked vs native gamma oscillations, but they didn’t discuss that in their new paper.
In a recent clinical trial of gamma stimulation as a treatment for mild to moderate Alzheimer's disease, there was no significant effect on their primary outcome metric. But that trial was likely underpowered, so we will need to await Phase 3 results.
4: Maybe formaldehyde is a signaling molecule? If true, this would be consistent with the idea that formaldehyde is widespread in biological systems and more evidence that the reversal of the crosslinks it forms is physically possible.
5: Study finds that there is an association between common variant and rare variant predispositions for neurodevelopmental conditions, which the authors attribute to assortative mating.
6: Study finds that antipsychotics at relatively higher doses (greater than 100 mg chlorpromazine equivalents) are associated with higher mortality in people 18-24 (but there is no association with mortality at lower doses, and not with either dose in younger people). The association is due to higher non-overdose unintentional injury deaths. They speculate this might be due to sedation or because it is a marker for more severe psychiatric illness, indicating decreased injury prevention skills.
7: Swedish cohort study finds that a diagnosis of hydrochondriasis is associated with a 4-fold higher risk of death of suicide. This is not surprising because it is a debilitating condition. The study also finds that people with this diagnosis have an increased risk of death due to diseases of nervous system, circulatory system, respiratory system, and death due to unknown reasons. The authors speculate that this might be due to chronic stress leading to a wide variety of bad effects on the body. Surprisingly, the only major cause of death that hydrochondriasis was not associated with was cancer.
8: A randomized trial (n = 73) of Magnetic Seizure Therapy (MST) compared to Electroconvulsive Therapy (ECT) finds that MST is as effective with fewer subjective side effects.
9: A new study in mice suggests that the antidepressant effects of ketamine requires the endogenous opioid system, particularly mu opioid receptor signaling in the central amygdala (CeA). However, it finds that ketamine does not directly mimic opioids. Instead, the mu opioid system appears to play a permissive role that gates the effects of ketamine. My guess is that the role of ketamine in enhancing endogenous opioid signaling might also help explain its analgesic and sometimes addictive properties.
10: The FDA grants a breakthrough designation to MindMed's lysergide d-tartrate — a salt of LSD — for generalized anxiety disorder, after a phase 2b study of a single dose shows rapid, robust, and sustained efficacy through 12 weeks.
11: Awais Aftab on a new study that performed unbiased dimensional clustering of psychiatric symptoms and found that many classic DSM diagnoses do not emerge as distinct syndromes.
12: Chatsirisupachai and de Magalhães make a principled argument for something that has seemed intuitively likely to me for many years: the accumulation of somatic mutations with age is unlikely to drive most aging phenotypes.
13: Eric Minikel on Ionis's new trial of an antisense oligonucleotide for lowering prion protein levels in patients with prion disease. He discusses why this is such an important trial, because it is targeting a plausible biological mechanism. Even though it could fail for many reasons, perhaps because it is unable to sufficiently diffuse into deep brain regions, that would still yield valuable lessons.
14: X-Therma, a company that has part of its focus on cryobiology, raises 22M.
15: Mark Woodward on using machine learning to find better cryoprotectants. He proposes training a model on existing cryoprotectant data to predict properties based on molecular structure, exploring different chemicals and combinations, and iteratively experimenting on the most promising candidates. One challenge here is that the existing data on cryoprotectants is not very robust. So it might be necessary to integrate some other sources of data, like molecular dynamics simulations.
16: Anti-cryonics articles this month by two prominent intellectuals, Arthur Caplan and Venki Ramakrishnan. I call them “anti-cryonics” because they both make a wide range of arguments against it and focus on some of the more sensational aspects. There appear to be some factual misunderstandings in both. No, there is no annual fee once someone is preserved at any organization I’m aware of. No, nobody is proposing “pickling” the brain. Ramakrishnan also uses the phrase “no evidence”, which is a red flag indicating imprecise communication. Most importantly, neither engages deeply with key technical considerations in cryonics. This is unfortunate, because there are some very important technical discussions to be had, and I wish more smart people like these two would give their input on them.
17: Philosophy paper by Daniel Story on the possibility of worse-than-death outcomes in cryonics. As usual, academic articles are about 10 years behind Less Wrong, Reddit, and random mailing lists, as this topic has been discussed endlessly online. What I don’t understand is how anyone who would avoid cryonics because of the presumably small but obviously non-zero possibility of a worse-than-death outcome could possibly want to have children or encourage others to. And if everyone stops having children, then what happens to humanity? The frequency with which I’ve seen this argument made against cryonics is one reason that I think some forms of anti-natalism are more popular than many realize.
18: Nice segment from 60 Minutes on deciphering the ancient scrolls of Herculaneum that were preserved with lava following the eruption of Mount Vesuvius. I love this quote from the archaeologist: "The paradox is that catastrophic destruction is also exceptionally good preservation." This reminds me of brain preservation with aldehydes and at very low subzero temperatures.
19: Tomorrow Bio has posted their four patient case reports from 2023. They report uniform cryoprotectant distribution throughout the brain tissue in two out of the three cases in which cryoprotectant perfusion was attempted, an outstanding result for their organization and the field in general. In both of these cases, there was a relatively short periods of brain ischemia prior to perfusion, just a few hours. In the fourth case, the brain was immersion fixed by hospital staff prior to transfer to their facility.
20: Excellent episode with John Smart on the Cryonics Underground podcast. Includes a lot of material related to brain preservation using aldehydes. John has been a mentor to me and source of deep encouragement for many years now.
21: Nice video by Colin Kakama titled "My optimistic view of cryonics":
Includes the subsection “Reasons cryonics isn't yet mainstream and why they won't last”. Argues that the main reason that people are not interested in cryonics is that it costs too much. And that if it were cheaper than burials, everyone would do it. Also argues that cryopreserving people is an intrinsically good thing, because people have a lot of knowledge, which is beneficial to humanity.
Can you please explain thinking behind #17 comment: "...worse-than-death outcome could possibly want to have children or encourage others to. And if everyone stops having children, then what happens to humanity?" Just confused why this was added! Thanks.
Arthur Caplan is better than many bioethicists but, on this topic, he's no more able to think without blinkers than the rest.