Action Potentials for January
1: Have you ever wondered why you can’t remember things backwards? Probably not. It’s not a very intuitive idea. But sometimes weird Gedankenexperiments (thought experiments) like this can help us to better understand how the world works.
A new study helps address this question. The authors recorded from around 90 neurons in the motor cortex of rhesus monkeys using multielectrode arrays. They also used a brain computer interface (BCI) that mapped neural activity directly to the position of a cursor on a screen, allowing the monkeys to control the cursor based on their own neural population activity in real time.
They first observed that during basic cursor control, there were natural patterns in the neural activity — structured trajectories in neural space that followed different paths depending on which direction the cursor was moved.
They then challenged the monkeys to produce these same trajectories in the reverse order. However, even with very strong incentives and multiple training approaches, the monkeys were not able to violate these natural patterns.
Instead of directly following the instructed path, they found that the neural trajectories “insisted” on following specific paths through the population activity space. The monkeys could not get their brains to route around these constraints, at least over the course of around 500 trials in a single day of training.
If these kinds of constraints are common across the brain, it could mean there are relatively strict “rules” about how neural activity patterns can change from moment to moment, due to the underlying network architecture. This theoretically limits the total number of distinct patterns that the brain can generate, potentially making the inference of those patterns from incomplete data an easier problem.
Of course, this would need to be validated across different types of neural activity before we generalize the principle too far. Most importantly, to see if similar constraints shape how we encode and recall memories.
2: A new expansion microscopy technique using potassium acrylate-based hydrogels allows for the 40x expansion of whole Drosophila (fly) brains while maintaining structural integrity. Using this method and a lightsheet microscope, they performed nanoscale resolution imaging of entire fly brains. They could visualize fine neuronal structures like mitochondria and they could also map neural circuits such as those of Tm5a visual neurons (below). Somewhat surprisingly to me, they also found that using the stronger fixative glutaraldehyde actually helped in retaining fluorescent proteins in the expanded samples.
3: A new method called SHARD allows for improved 3D visualization in archived human brain tissue. It builds on the SHIELD tissue clearing technique. The brains were initially preserved via immersion fixation, with post-mortem intervals of between 6-72 hours. They reported excellent visualization of multiple cell types in the dorsolateral prefrontal cortex:
4: Many have long thought that Huntington’s disease causes neurons to slowly degenerate over decades due to constant exposure to toxic mutant huntingtin protein. A new study finds instead that the disease-causing CAG repeats in the huntingtin gene must expand beyond 150 repeats to become toxic. They find that neurons die quickly once reaching this threshold, due to losing features of neuronal identity and de-repressing senescence and apoptosis genes. This suggests that at any given time, most neurons in the brains of people with Huntington’s disease are healthy, spending about 95% of their life with a non-toxic gene that is gradually expanding. This obviously suggests that interventions to prevent DNA expansion could be a key to treatment.
5: A new study of changes in the postmortem mouse brain finds that the spatial distribution of several types of RNA is reliably detectable for up to 24 hours at room temperature. The signal weakens and becomes less reliably detectable at 48 hours. Storing the body at 4°C extends the time window for reliable spatial RNA detection for up to 48 hours postmortem (the longest time period they tested).
6: How does one’s genetic predisposition (polygenic scores) for various psychiatric diagnoses or other traits affect what profession one will end up in? A new study evaluates this. They studied the correlations among all individuals in their cohort and also those without any neuropsychiatric diagnosis (“controls”). They report a few statistically significant findings, including an association of schizophrenia polygenic scores with arts and design and an association of autism spectrum disorder polygenic scores with computers and math professions. However, while the effects are statistically strong, they are not very practically strong, with most odds ratios being less than 1.1 per polygenic score standard deviation.
7: New-to-me study from 2023 analyzes Swedish military conscription records of more than 400,000 men born between 1971-1983. They leveraged the random assignment of 18-year-old men to different doctors during mandatory military screening. Some doctors were more likely to diagnose mental illness than others, allowing the researchers to measure the causal effects of receiving a diagnosis. They found that the group of people who had a “marginal” diagnosis of mental illness (i.e. cases where diagnosis depended on which doctor a person saw) experienced higher mortality rates, more hospitalizations, and reduced employment.
There are some problems with the study. The effects might have been due to discrimination within the military or otherwise, and the stigma of a diagnosis has lessened over time. It is obviously different than settings where people are actively seeking a diagnosis. Most concerning to me as a confound is that there may have been an incentive to do poorly on the mental health screen to decrease or avoid military service, so these “marginal” cases may have been a proxy for something. However, this study is still a compelling datapoint about potential downsides of mental health diagnosis in certain contexts.
8: A new study looks at compulsive behaviors in pets by analyzing their specific motor patterns. They found that dogs and cats with suspected compulsive disorders showed distinct behavioral signatures, particularly “non-functional acts” like licking the air, as opposed to functional behaviors like grooming. It is challenging to determine the extent to which there are parallels with human OCD. However, I find this an interesting line of research because it would be surprising to me if OCD did not also exist in animals.
9: A meta-analysis of 27 randomized clinical trials involving 1,369 participants found that glutamatergic medications (including N-acetylcysteine, memantine, lamotrigine, and riluzole) show some promise in treating OCD and related disorders. They suggest that these glutamatergic medications could provide a valuable treatment option for the approximately 60% of patients with OCD who do not respond adequately to first-line SSRI therapy.
However, there seems to be a lot of publication bias in this literature. Their funnel plot shows an asymmetrical distribution of study outcomes (blue dots) around the estimated effect size (vertical red line), suggesting that smaller studies with less favorable or negative results may be missing from the literature. So it’s hard to be too confident about this data.
10: A company is trying to make one of the best antipsychotics, amisulpride, available in the US. This is great. Unfortunately, it seems like it is going to be a slightly different version — once daily dosing rather than twice daily — and correspondingly be quite expensive. It would be such a better idea for the FDA to simply approve amisulpride based on the data used for approving it in Europe and other areas. The FDA being so restrictive on approving drugs that are approved in other peer countries is a major error.
11: New study with data on step counts in different US states:
12: Seth Herd on why so many people who apparently believe that a singularity is happening somewhat soon still care so much about the culture wars: Because we have primate brains and are irrational. Seems basically accurate.
13: A French YouTube video about cryonics from 3 months ago now has 2 million views. While the coverage itself was fairly positive, the comments expressed a variety of criticisms: it perpetuates inequity by being only for the rich, it damages the environment, it goes against nature, it falsely assumes future societies would want to revive preserved people, it won’t work, and even if it did work, it would be traumatizing to wake up alone in a radically different world. It’s nice to see that people from different cultures can come together in their reasons for loathing cryonics.
14: User JoeStrout on Reddit with an interesting comment: “I work in connectomics — i.e., developing the technology to read out the complete wiring diagram of a brain, synapse for synapse. This is a field that is advancing exponentially. We’ve gone from nematodes (302 neurons) to fly central hemibrain (25,000 neurons) to complete fly brain (140,000 neurons), and are now seriously eyeing a complete mouse brain within 10 years (quite likely less). The technology, both imaging and AI-related (needed to automate terabytes of image processing), are advancing so fast it’s hard to keep up, even for those of us in the field.
This technology will allow us to upload to digital form, very much like the TV show Upload (which is a drama/comedy and obviously fiction, but has a surprisingly sensible take on the core tech). This is how I expect cryonics patients will be revived, and I’d guess it will be in less than 50 years — possibly much less.”
15: Personal and emotional reflections of the husband of a woman recently cryopreserved at the Cryonics Institute. The author discusses grappling with grief while maintaining hope for a future reunion via technological advancement. Notably includes insights from an anonymous cryobiologist (“Dr. Yates”) who has faced significant professional censure for his work in cryonics, who said “My whole life I’ve had to live in the shadows because the world was not ready for it.”
16: Our review of the effect of cryopreservation on brain cell structure is now available on PubMed. We searched the literature for all of the studies we could find that evaluate how cryopreservation affects the structure of cells as seen under the microscope — either light microscopy or electron microscopy.
We found that artifacts due to cryopreservation — including but not limited to ice crystal formation — clearly alter microscopic structure. The extent to which that means it is no longer possible to trace neural structures with AI-assisted methods is unclear. Evidence also clearly shows that this damage can be ameliorated with cryoprotectants.
For more data, we also provide an example comparison from one human brain, one hemisphere of which was immersion fixed and the other hemisphere of which was cryopreserved (without cryoprotectants). The pictures on the left (a, c, e, and g) are from the fixed tissue while those on the right (b, d, f, and h) are from the cryopreserved tissue.
You can see numerous clear, white spaces in the cryopreserved and rewarmed brain tissue. This is thought to be areas where ice crystals formed. The remaining parts of the tissue are compressed together and distorted.
Overall, the review explains why aldehyde fixation has been better validated to result in high quality structural preservation than cryopreservation alone.
re: #7 Based on my clinical experience, I think this is way more about the types of people who are likely to receive those sorts of equivocal diagnoses. I'm willing to bet that those patients have some sort of personality pathology mixed with anxious/depressive symptoms (i.e. borderlines/Cluster B/cPTSD/whatever you want to call them) and there is much disagreement between docs about whether or not we should put a diagnostic label on them. It is also very unclear what to do to help them, as medications rarely work, so a diagnostic label is unlikely to help. These patients often engage in self-harm of varying levels of risk (which ends up with them psychiatrically hospitalized, or dead), and almost uniformly have pretty bad interpersonal situations, so (assuming my hypothesis is correct) I'm not surprised to hear that they have poor outcomes.
From the fly paper: "Freshly dissected Drosophila brains were fixed with 4% PFA for 30 min on ice, before undergoing 0.2% GA fixation on ice under vacuum for another 20 min."
That is impressive! I wonder if the implications for human brain perfusion preservation will hold up...