Action Potentials for June
1: A new whole animal connectome dropped this month, this time for a three-day-old larva of the annelid worm Platynereis dumerilii. It contains 9162 cells, 1642 neurons, and 28,717 presynaptic sites. This is the third whole animal synaptic connectome mapped after C. elegans (first published in 1986, although incomplete at the time) and the tadpole larva of Ciona intestinalis (2016). Here is what the three-day-old larva organism looks like on SEM:
Here is an example of what the underlying electron microscopy looks like, with some segmented axons highlighted in color:
Here are 3d reconstructions of the cells, the neuronal morphologies, and the effector structures (e.g., muscles, glands):
Here are 3d reconstructions of different modules of the nervous system, such as those responsible for vision or posture control:
The organization of the nervous system is mostly feed-forward, with flow from sensors through interneurons to effectors. But they also identified some recurrent networks, which are suspected to play a role in internal processing. For example, they identified a rhythmic pacemaker circuit thought to drive the ciliomotor system.
As technology improves, we will be able to generate whole body connectomes for more species, and more individual animals within each species, just as happened in genomics.
2: Deep learning method to generate synthetic 3D segmentations from sparsely annotated 2D electron microscopy images:
The models trained with their synthetic data approach have similar accuracy while requiring 3 orders of magnitude less training data. This is important because the need for human annotation is the main current bottleneck to scaling up connectomics.
3: Accelerating scanning electron microscopy imaging by re-scanning selected error-prone areas. Claims that it leads to a 7-fold speedup compared to traditional methods while still maintaining segmentation accuracy.
4: Whole brain light sheet microscopy in mice, showing just a few of the reconstructed neurons. It’s always fascinating to me how long axons can be, reaching across macroscopic areas of the brain. Scale bar = 1 millimeter.
5: More whole brain light sheet microscopy data (preprint), this time from a human cerebral hemisphere fixed with formaldehyde. Different antibodies stain for different cellular features:
6: Sometimes I’ve heard people argue that brain regions aren’t really distinct things. So I was interested to see a comprehensive counter-argument in defense of distinct cortical brain regions.
First, it’s pretty straightforward how such regions could form during neurodevelopment, with gradients of gene expression and distinct anatomical connectivity:
Second, neuroimaging provides good evidence for functionally distinct cortical brain regions. For example, here is a parcellation of the human cerebral cortex into distinct areas based on the correlations of their functional neural activity with other areas. Different studies have apparently found similar parcellations, despite having pretty different methodologies.
I’m not at all an expert in the area, but I at least find this counter-argument pretty compelling.
7: A new neural network model shows how synaptic plasticity during memory formation can create neuronal assemblies that can then be reactivated in order to recall the memory.
8: Big steps forward for gene therapy in the brain due to developments in AAV engineering.
9: Sleep deprivation is associated with decreased hippocampal replay of the memories that the organism formed while awake. And replay of procedural experiences in the striatum also occurs during sleep, independent of the hippocampus.
10: Five more years of data shows that naked mole rats simply do not age. They defy Gompertz. Compare the age-related mortality risk of naked mole rats to horses, humans, or mice. Less clear: why they don’t age, and how we can use that knowledge to prevent us from aging.
11: Steve Horvath’s team with another major finding, this time in the form of an epigenetic predictor for maximum lifespan. They use a multivariate regression model of the methylation of cytosines that are highly conserved across 348 mammalian species.
Notably, females have an epigenetic longevity advantage in 17 of the 348 species, including humans, while males have a longevity advantage in only 1 (harbor seals, what is up with them?).
Partial reprogramming of cells did not significantly affect the methylation-based maximum lifespan predictions. The authors suggest that it is a fundamental property of each species that is difficult to change.
12: One piece of rarely-discussed good news in recent years is that age-adjusted dementia rates have fallen by about 25 percent since the 1990s. My personal read of the literature is that this is primarily due to better cerebrovascular health, in turn due to better treatment of certain medical conditions such as hypertension. I don’t think it is just due to a change in diagnosis thresholds, although I do agree that should be high on the differential when considering this type of trend (for example: teen mental health).
A new analysis of Framingham study data finds that people born in later decades have larger brains on age-adjusted structural MRI scans. They especially have larger cortical surface areas. This may help to explain the decline in dementia diagnosis rates over time.
13: New commentary highlights the major annual impact of several psychiatric conditions on quality-adjusted life-years (QALYs), including opioid use disorder (−0.16/year), generalized anxiety disorder (−0.10/year), and major depression (−0.09/year). These yearly QALY estimates are higher than some common non-psychiatric conditions, like diabetes (−0.06/year) and obesity (−0.04/year).
14: Suicide is most common at 12 pm, while homicide is most common at 10-11 pm. But most people are awake during the day. A new study finds that, adjusting for how likely people are to be awake, the risk of suicide for someone awake at 3 am is 5x higher. This effect was stronger in young people, but not among those with a history of suicidal ideation or suicide attempt. It is evidence that there is a higher risk of impulsive behavior at night.
15: Retrospective study of 83,000 patients from U.S. electronic health records finds a 50-56% decrease in alcohol use disorder incidence and recurrence over a 1 year period with the use of semaglutide compared to other non-GLP1 agonist obesity medications.
16: I continue to be confused that news reports about research on ketamine, such as this one, don’t mention that its mechanism of action seems to be related to the opioid system. It doesn’t make sense to me. It seems like this should be a bigger part of the discussion around ketamine research. Probably I am missing something. Maybe the argument is that it is stimulating the endogenous opioid system rather than directly binding to opioid receptors, which makes it way different from opioids, and I just don’t know enough psychopharmacology to understand that.
17: Twin study on the heritability of OCD finds that genetic factors explain 50% of the variance in OCD, with environmental factors and measurement error accounting for the rest.
18: We know that people who discontinue antidepressants often have discontinuation symptoms. But how often do people who take placebo medications in trials also experience discontinuation-like symptoms when they are stopped?
New study finds that approximately 1 in 6 people (17%) in the placebo group whose placebo medication are stopped experience discontinuation-like symptoms. This is compared to approximately 1 in 3 people (31%) in the actual antidepressant group. They also found that approximately 1 in 33 people (3%) experience severe symptoms after discontinuing antidepressants.
19: Excellent article on the long-term trajectories of people who experience psychosis and how antipsychotics can affect this.
20: A speculative argument that psychedelics reduce fear of death by inducing comforting religious or mystical beliefs about the possibility of postmortem experience. I.e. by relaxing people’s beliefs and making them think that a purely naturalistic belief of the world is false. If true, is it sort of like taking the blue pill?
21: Completely agree that randomized control trials (RCTs) are often unethical and we should move away from them. Perhaps, in order for a RCT to be required for drug approval — and in turn for people to not be able to access them — there would need to be good evidence of actual clinical equipoise via prediction markets, rather than this just being assumed. But more importantly, the current system is wrong because it restricts our pursuit of life. Â
22: Study of brain tissue after death in rats finds that the severity of microscopic changes after 48 hours at room temperature is greater than after 7 days under refrigeration and similar to or slightly lower than after 14  days under refrigeration.
23: An excellent new study on electrophysiology in brain slices from Alexander German and Enes Yağız Akdaş at the University of Erlangen–Nuremberg.
They preserved 400 μm-thick slices of the mouse hippocampus using vitrification, which — when it is successful — turns tissue into a glass-like state without forming ice crystals. After rewarming these brain slices, they measured field excitatory postsynaptic potentials (fEPSPs) to test whether the neurons could still communicate with each other.
Their results showed partial preservation of neural function. Vitrified slices exhibited synaptic transmission, although with reduced amplitude compared to control slices. They also observed high-frequency stimulation (HFS)-induced potentiation of synaptic transmission in two out of three samples.
This partial success shows that while the vitrification process preserved some brain function, it didn't maintain all the electrophysiologic properties seen in control brain slices.
This is admirable research because they cite previous work, clearly explain their methods, and mention the percentage of times that the procedure failed due to inadvertent ice formation despite their attempt at vitrification (40%).
What is especially great about this study is that they also performed microscopy on the tissue. On the left is an electron micrograph of control brain tissue and on the right is tissue after their vitrification procedure:
They identify areas of vacuolization (asterisk) and cell membrane rupture (double asterisks) in the hippocampus after vitrification, suggestive of some degree of necrosis. They conclude that their abnormal electrophysiology measurements can be explained by their imperfect structural preservation results. The converse of this is that if a study could show consistently excellent ultrastructure after vitrification of brain slices, then the tissue would also be expected to have excellent electrophysiology if tested.
Unsurprisingly, structure predicts function. This has been the story of biology since Aristotle, Galen, and van Leeuwenhoek.
How do you interpret data shown in the figure for 3d reconstructions of different nervous system modules? It is a beautiful connection of interspersed and colorful lines, but what can be drawn just from looking at the figure?