Zebrafish protein unlocks dormant genes for heart repair

> Whatever weird programming language life is made from it seems to be a surprisingly forgiving environment.

Either that, or we truly are more similar than we seem to ourselves. It makes sense that humans would have a greater sense of discernment at the sensual level genetically nearer to us than further from us.

I assume we only hear about the cases when the protein does something interesting, and not the 999/1000 cases when it doesn't.

It's all just chemistry, and the chemistry is extremely conservative all the way back to the first eukaryotes. Once our kind of life figures out a way to do something it's very uncommon that the fundamental pathways change - not by any means impossible.

Cells that have novel things happen to them nearly always die. Others become cancer, which is a career-limiting move for multicellular life at least. So the process of changes to basic chemistry of life is very slow. That's why omnivores can eat almost anything and be fine with it - imagine if different branches of plants and animals had really novel things in them, like a critter with nerve gas for a circulatory fluid.

There's been some discussion of mirror bacteria lately, which have opposite chirality in all their molecules, which would be substantially problematic since we really don't have defenses against that kind of thing. That just shows the degree to which you'd have to rejigger biochemistry to get a truly whole new thing going on.

> I wonder if the leap between a fish and a mouse is less than the distance between a mouse and a human?

No, mice are mammals just like humans. Fish were around long before mammals even existed.

It is a protein that exists in mammals. Presumably they produced the murine version to put in the mouse. Science journalism screwed up the paper's language, calling it a zebrafish protein, when it is a protein that is conserved for all vertebrates. But the function of the protein, and its role in heart regeneration, was identified in zebrafish.

> Directly after LAD ligation during MI surgery, hearts were injected twice with 15 μl of AAV9(CMV:GFP) or AAV9(CMV:HA-Hmga1) (1 × 1012 virus particles per mouse) in opposing regions bordering the area at risk of ischemic injury.

The paper doesn't say whether the viral vector (AAV9) used the murine or zebrafish Hmga1 sequence, but it is more likely that they used the murine.

Babies can heal from heart damage really well, the ability is lost in your life's first few months. So this is just turning on a flag.

one of the earliest and my personal favorite chymeric animals, were bioluminesent jellyfish monkeys, 2001 or before. The only one to glow visibly in normal light didnt live long, but some of the rest, that only glowed under infrared/black lights, did ok. So the zebrafish thing just might work. Also done by some mad scientist experimenting on themselves, was the injection, into there eyes of a chemical produced by deap sea ?squid I think that humans dont produce any apreciable amount of, but we have the receptors for, and gives supper enhanced night vision to humans who are mad enough to give it a go.

Well it's important to note that all organisms (almost certainly) descend from a single common ancestor (https://en.wikipedia.org/wiki/Last_universal_common_ancestor) and the organisms we work with often are quite close on the inheritance tree (really, a graph if you include horizontal gene transfer).

When you collect the same exact protein from a bunch of close species, their underlying sequences often differ by very little and those differences are really just random mutations that occurred and were tolerated. Some proteins got "locked in" as useful and could not really change since so many things depended on them. For example, eukaryotes like humans have a protein called "actin" that is incredibly well consereved across all mammals, and we can even see what appear to be very distant cousins in bacteria- those two parts of the tree of life diverged billions of years ago (you think legacy computing is bad?)

If you start to study biology through the lens of evolution/development (IE, looking at the similarities and differences in close and far related organisms) it starts to become a lot clearer. Many organisms with very different phenotypes actually just contain slight modifications to the parameters of common generator functions- for example, wings and arms seem to derive from the same ancestral body part. Most of the genes that make them are in common, it's just how and when the different genes get expressed that determines the finer details (I'm glossing over a lot here).

So since nearly all the players are the same, it's often not that hard to move a protein from a close relative (mice are pretty close to us) to another without any real loss of function or major change in phenotype. However, there are some fundamental differences that make comparing mice results to humans and sometimes we modify mice to appear more human to make them better model organisms.

(even after working in this field for some 30+ years, I remain constantly amazed any of this works at all. I felt the same way about Google's production environment, which in many ways resembled a very simple organism)

“In mice” is a very important lesson to learn about scientific reporting, but it’s also a very simple and reductive lesson. Most experimental results don’t translate to human studies. That’s it, thats the lesson to learn.

So, maybe someone here doesn’t know that, but it adds so little to the conversation to just repeat this widely known fact over and over again.

Now, let’s talk about the mouse models. Some models are better than others. Many experiments are conducted in mice that don’t have an immune system. Some mouse models have a “disease” that bears very little resemblance to the human condition and was induced through means that are very different from the human natural history. Some mouse models have a genetic defect in a gene that is very well conserved and mimics symptoms of the human condition and are thus excellent models that can tell us a lot about human physiology. All that nuance is lost by “in mice”, which makes it particularly frustrating when that’s used as a blanket reason to dismiss any mouse study.

There's a whole book and movie about it

https://en.wikipedia.org/wiki/The_Secret_of_NIMH

That won't stop Bryan Johnson.

"...in mice, in a lab" as the saying goes

To be fair, the gene also repairs heart damage in Zebrafish which is more than just "in mice". I understand the reasons behind the meme "in mice", but on the other hand dismissing valid progress because humans reflexively insisting on exceptional snowflake status is counterproductive.

Apropos: "Analysis of animal-to-human translation shows that only 5% of animal-tested therapeutic interventions obtain regulatory approval for human applications" - https://journals.plos.org/plosbiology/article?id=10.1371/jou...

Although, even if it doesn't directly work in humans, it might be a stepping stone to something similar that might.

I don't think the image is an artistic representation- I think it's an actual micrograph of fluorescently labelled cells. They may have called it that because some level of postprocessing was applied. In short, when this type of micrograph is taken, it's generated using a sequence of laser pulses, each laser pulse has a different wavelength of light designed to match one of the fluorescent labels, which itself emits a different wavelength of light. The images themselves are collected in grayscale and then a false-color image is created by coloring each grayscale image (mapping gray intensity to luminance) and compositing them. The color mapping does not recapitulate the underlying colors of the images that were taken although usually when these images are rendered, a color very close to the emission wavelength of the fluorescent label is used.

That said it's not impossible to conjure up prompts that would make a generative AI produce images that look a lot like a micrograph of cells. I skimmed the referenced paper and didn't see this image but I suspect the authors provided it in the PR package they must have distributed.

If we were mice this would be an amazing time to be alive.

username checks out haha