Looking Closely

March 6, 2021 5:01 PM

From a Single Cell to a Salamander "In this six-minute time lapse video, you can watch a single cell grow into an alpine newt salamander. I got this via Craig Mod’s post about looking closely, in which he asks: when precisely does this collection of cells become a salamander?"
posted by kathrynm (43 comments total)

I've been idly thinking about the "when precisely?" question for a long time. I've come to the conclusion that "it's a salamander" is merely a rough heuristic. It's an approximation of reality. A thinking hack. It's often a useful hack for helping us deal with the world, but we shouldn't be surprised when it doesn't work very well. We shouldn't put too much effort into the question of "when precisely?" because that's precisely when our rough heuristics break down and stop being useful.

How many species are in a ring species? Not a useful question. We're asking "where precisely do these species divide?" when there is no precisely where.

What's the exact definition of biological sex which allows us to precisely divide everyone into male and female? Not a useful question. We're asking "what precisely defines a sex?" when there is no precisely what.

Etc.

Oh, and... great video! :-)
posted by clawsoon at 5:20 PM on March 6, 2021 [14 favorites]

The under-the-surface crawlies around 2:28... are those something like, I dunno, fibroblasts, or neurons, or...?
posted by clawsoon at 5:25 PM on March 6, 2021 [2 favorites]

This blew my mind.

I can't get over the way it folds around itself - forming tubes for the GI tract, etc.? - while remaining roughly spherical... and then just stretches out into a noodle. Neat trick!
posted by colin.jaquiery at 6:02 PM on March 6, 2021 [1 favorite]

The under-the-surface crawlies around 2:28...

I was curious about this too. Following some links through to the nat geo article (yucky sign-up-wall):

"Starting at around two minutes and 20 seconds into the time-lapse video, or around five days of development, Hurney points to the passage of individual cells migrating across the salamander’s surface. Each is taking cues from the genetic blueprints within it as well as signals from surrounding cells to determine what sort of tissue it will become."
posted by colin.jaquiery at 6:14 PM on March 6, 2021 [8 favorites]

a) cell division is INSANE

b) we are all just gigantic bodies with very elaborate food-catchers wrapped around simple GI tracts. Brains included.

c) It remains astonishing how much simple fish/amphibitans look like human fetuses, ontogeny recapitulating philogeny etc.
posted by Eyebrows McGee at 6:20 PM on March 6, 2021 [3 favorites]

It's funny how life, something that has never concerned itself with hard delineations, has given rise to human intelligence, something that desperately wants everything to be precisely categorized.
posted by Mr.Encyclopedia at 6:28 PM on March 6, 2021 [6 favorites]

when precisely does this collection of cells become a salamander?

Always has been.
posted by malphigian at 6:31 PM on March 6, 2021 [4 favorites]

Which is the argument abortion foes make.
posted by PhineasGage at 6:51 PM on March 6, 2021 [5 favorites]

A newt?!?
posted by Greg_Ace at 7:18 PM on March 6, 2021 [3 favorites]

I liked the part where it has multiplied in to a lot of cells, makes a sphere, then a torus, to become its GI tract. There is an elegance in all of this and an urging outward to fill that sac. What a beautiful film.
posted by Oyéah at 7:25 PM on March 6, 2021 [3 favorites]

In order to reason about any thing, the very first step has to be to make a distinction between the thing and all that the thing is not.

Too often, I think, people forget that these distinctions are a part of what we do in order to think, rather than some inherent property of whatever it is we're thinking about.
posted by flabdablet at 7:28 PM on March 6, 2021 [15 favorites]

Metafilter: an urging outward to fill that sac
posted by Greg_Ace at 7:32 PM on March 6, 2021 [5 favorites]

Nature Shows, it is time to quit with the foley already.
posted by TheCoug at 7:47 PM on March 6, 2021 [6 favorites]

Yes, so we start out as a single cell, then that cell divides, those cells divide, and some magic formulas called DNA makes these cells move and change to create different kinds of cells, and these different kinds of cells form larger structures, and eventually a whole extremely complex living thing is there with many many levels of magnitude greater numbers of cells than the starting cell.

I learned about how DNA makes chains of amino acids to create different proteins but do we know and understand how DNA orchestrates what we are watching in that video? Why does this happen now and that happen next? Where is the chronological script that governs when and what things happen?
posted by njohnson23 at 8:03 PM on March 6, 2021

In the case of living things, I'm not at all convinced that positing a "script" that is in any way distinct from the processes that any such script would putatively "govern" is a good use of reasoning resources. It's been my experience that analogies commonly drawn between the behaviours of living things and those of engineered things all break down on close examination, and that they way they all break down is pretty similar: somebody works out that there are huge amounts of causal connection between what have previously been thought of as conceptually separable organizational layers - connections that no engineer would ever design in, purely because doing so would make reasoning about system behaviour uselessly difficult.
posted by flabdablet at 8:14 PM on March 6, 2021 [5 favorites]

> Nature Shows, it is time to quit with the foley already.

For real. Also, documentaries. I used to love watching historical doccos, but now all I end up doing is focusing on how terrible the dubbed-in sounds are on the old war footage or whatever.
posted by glonous keming at 8:20 PM on March 6, 2021 [2 favorites]

Fixed that for you
posted by flabdablet at 9:34 PM on March 6, 2021 [3 favorites]

Previously, but without the little National Geographic laurels, whatever those mean.

This video is one of the most amazing things I have ever seen in my life. I've probably watched it twenty times in the last two years. I'm glad it's here again.
posted by fantabulous timewaster at 9:45 PM on March 6, 2021 [3 favorites]

The under-the-surface crawlies around 2:28... are those something like, I dunno, fibroblasts, or neurons, or...?

I'm not a developmental biologist so I'm not sure exactly what they are, but cell migration is an important motif throughout development. Developmental biologists speak of a cell's "fate," which is sort of the set of possible things it can become, locked in by its history so far. The initial fertilized egg, and all the cells for the first several cell divisions, have a fate that encompasses every possible cell type, and throughout development as each cell responds to signalling molecules, its fate starts to narrow. One of the most common ways of establishing a cell's fate is by its position in space relative to some organizing point: the embryo can take advantage of the diffusion of molecules produced there to serve as a signal telling each cell what its fate should be. For example, the first pore that we see in that video helps to establish three different types of tissue, governed by their locations relative to the pore: endoderm, mesoderm, and ectoderm. The possible fate of ectodermal cells includes skin cells and neurons, but an ectodermal cell can never become a muscle cell; only mesodermal cells can do that. But it's frequently useful to have some cells of a certain type present in an area where there aren't any precursor cells with the appropriate fate. So a relatively small number of cells will migrate, sometimes huge distances (relatively speaking) from the region where they can form to the region where they're needed. So anyway, I don't know exactly what types of cells we see migrating in this video, but I think there's a lot of possibilities, and it's a hugely important part of development even if it only seems like a comparatively small number of cells doing it. I think it's amazing we can watch it happening here.

I can't get over the way it folds around itself - forming tubes for the GI tract, etc.? - while remaining roughly spherical... and then just stretches out into a noodle. Neat trick!

The one big fold that happens at about 2 minutes in is actually the formation of the neural tube, which becomes the brain and spinal cord. This area produces a huge number of signalling molecules that help orchestrate the organization of the entire body, but especially the head. It was really beautiful to watch!

In the case of living things, I'm not at all convinced that positing a "script" that is in any way distinct from the processes that any such script would putatively "govern" is a good use of reasoning resources. It's been my experience that analogies commonly drawn between the behaviours of living things and those of engineered things all break down on close examination

I agree completely, but I do think these analogies are still useful for helping us develop our understanding, even if we need to treat them carefully and be ready to discard them as our understanding improves. The idea that DNA provides a "script" or a "blueprint" for development isn't all that accurate, but it does capture the gist of the idea. A better metaphor is that of "source code": DNA specifies various conditional checks and responses that each cell performs during development. No timetable is required to orchestrate things, because each cell just responds to the signals produced by its neighbors by undergoing some change and emitting its own signals in turn. Perhaps a better metaphor still is that of feedback control systems, but that's less familiar to most people. And of course each metaphor fails to capture how development actually proceeds in some important ways.
posted by biogeo at 12:09 AM on March 7, 2021 [12 favorites]

Oh, and the very first feature we see on the embryo, that first pore that forms at about 1 minute in?

That is what will become its butthole.
posted by biogeo at 12:19 AM on March 7, 2021 [8 favorites]

I do think these analogies are still useful for helping us develop our understanding, even if we need to treat them carefully and be ready to discard them as our understanding improves.

Agree completely. Just pushing back against taking them too seriously. For example, how long did the central dogma of molecular biology impede understanding that epigenetics and prion diseases are both actually things?

Life's karma just keeps on running over our dogma and I've never seen any indication it's about to stop doing that. I'm quite fond of the principle that if a young biologist tells you that something is possible they're quite probably right, and if an old biologist tells you that something is impossible they're quite probably wrong.

the very first feature ... is what will become its butthole

Thanks for that; I will tuck that little factoid away and whip it out next time some "spiritually advanced" yoga type has the temerity to ask me why I'm so attached to my colon.
posted by flabdablet at 12:54 AM on March 7, 2021 [2 favorites]

No timetable is required to orchestrate things, because each cell just responds to the signals produced by its neighbors by undergoing some change and emitting its own signals in turn.

I know diddly squat about biological processes, but this sounds a lot like event-driven asynchronous programming. This is very popular right now with nodejs for developing network software and websites, but has been around for a long time.

The way it works is event-driven, i.e. something happens (an event), and there is code listening for that event, so it does stuff as a result. When it's finished, it emits its own event(s), and some other pieces of code sees THOSE events, and acts in turn. The beauty is that it's not doing step a, step b, step c in order or to a timetable, but each piece of code acts only when there's a need for it to do so, and at its own speed, then triggers what needs to happen next as soon as its done. So it doesn't really matter what overall order things happen in, as long as each small step has what it needs when it starts. It all runs in parallel, so you can scale it up very easily - multiple events trigger multiple responses, and it cascades through. If a step fails, it triggers an error event to clean up the mess, and the subsequent steps just don't happen because they're not triggered in the first place.

Of course, any such analogy falls down in the real world, but it sounds like it might be a closer model than a blueprint, or synchronous step-by-step code that people are more familiar with.
posted by Absolutely No You-Know-What at 1:28 AM on March 7, 2021 [3 favorites]

something happens (an event), and there is code listening for that event, so it does stuff as a result. When it's finished, it emits its own event(s), and some other pieces of code sees THOSE events, and acts in turn. The beauty is that it's not doing step a, step b, step c in order or to a timetable, but each piece of code acts only when there's a need for it to do so, and at its own speed, then triggers what needs to happen next as soon as its done.

Biology-inspired hardware that also works this way is a thing now.
posted by flabdablet at 2:00 AM on March 7, 2021 [1 favorite]

when precisely does this collection of cells become a salamander?

This is called the “Paradox of the Heap,” and it was coined some 2300 years ago. My favorite answer is the Stoics’, which amounts to “I have better things to do than deal with your nonsense.”
posted by GenjiandProust at 3:01 AM on March 7, 2021 [10 favorites]

I'm not a developmental biologis--

Shuddup-and-answer-my-questions:
(srs tho, thanks for the great comments)

When we see the blastopor-- butthole-pore form, the material appears to be getting 'sucked' into the center of the zygote. And when the neural tube is formed, it looks like a wave of cells straight up crawling around the equator.

So how does all that relative movement happen? Is it mostly driven by different rates of cell replication (slowly growing/dividing areas tending to 'fall' to the center of the zygote while fast growing tissues spread outward as all those squished-together cells push against their neighbours)? ...or are the blighters growing legs?

For the first nine divisions the cells seem to divide in lockstep. At 0:52 we jump cut and have entirely lost synchrony. Is this just boring stochastic drift, or is there something else going on?
posted by colin.jaquiery at 3:57 AM on March 7, 2021 [1 favorite]

Oh, and the very first feature we see on the embryo, that first pore that forms at about 1 minute in?

That is what will become its butthole.

It is always fascinating to reflect on the earliest feature we all develop and how some of us never really get met much past there.
posted by ricochet biscuit at 3:58 AM on March 7, 2021 [2 favorites]

I wonder where it got those cute Chinese tattoos.
posted by MtDewd at 4:42 AM on March 7, 2021 [1 favorite]

TFA: when precisely does this collection of cells become a salamander?

malphigian: Always has been.

PhineasGage: Which is the argument abortion foes make.

I'd say that it became a salamander the moment G-d installed a salamander's soul into it. And She's no fool, so She's not wasting time and materials installing souls into salamander babies She KNOWS aren't going to go full term.

There's never been a moral issue with salamander abortion, because not a single aborted salamander baby ever had a soul. Because G-d isn't a fool.
posted by mikelieman at 4:48 AM on March 7, 2021 [2 favorites]

Thanks for that; I will tuck that little factoid away and whip it out next time some "spiritually advanced" yoga type has the temerity to ask me why I'm so attached to my colon.

You...are asked this often?
posted by tiny frying pan at 5:46 AM on March 7, 2021 [3 favorites]

colin.jaquiery: ...or are the blighters growing legs?

As I understand it (and, as an interested amateur, I could be wrong), the cells are moving around something like amoeba, those distant relatives of ours, do. Some of our cells, including a few types in our immune system, continue to crawl around that way our whole lives.
posted by clawsoon at 5:52 AM on March 7, 2021 [2 favorites]

GenjiandProust: This is called the “Paradox of the Heap,”

Thanks for finally giving me a name for it! I've been thinking about it for 30 years, ever since I started wondering when a rotting orange stops been an orange while vacuuming the church floor on some Saturday morning during my teens. Now I can finally learn what other, smarter people have been thinking about it for the past couple of thousand years. Thanks!
posted by clawsoon at 5:58 AM on March 7, 2021 [4 favorites]

tiny frying pan: You...are asked this often?

I imagine you only have to be asked once for the question to stay with you.
posted by clawsoon at 5:59 AM on March 7, 2021 [4 favorites]

Absolutely No You-Know-What: I know diddly squat about biological processes, but this sounds a lot like event-driven asynchronous programming.

Not a bad analogy, but I'd add a caution: If there are multiple ways to do something, there's a good chance that nature will do all of them. So some of it is event-driven, and some of it is timer-driven, and somehow they all line up at exactly the right time (most of the time) to successfully build a body. (Computer science-wise, sometimes the way that development works reminds me of the Story of Mel.)

I'm not sure how much this chapter has changed since 2002, but you can read about some of the technical details of the interactions between gene expression and development in the free online 4th edition of Alberts' Molecular Biology of the Cell, Cell Movements and the Shaping of the Vertebrate Body:

Most cells of the animal body are motile, and in the developing embryo their movements are often extensive, dramatic, and surprising. Controlled changes of gene expression create ordered arrays of cells in different states; cell movements rearrange these cellular building blocks and put them in their proper places. The genes that the cells express determine how they move; in this sense, the control of gene expression is the primary phenomenon. But the cell movements are also crucial, and no less in need of explanation if we want to understand how the architecture of the body is created. In this section, we examine this topic in the context of vertebrate development.

Down in the section "A Gene-Expression Oscillator Controls Segmentation of the Mesoderm Into Somites" you'll see an example of a timer mixed into all the asynchronicity and somehow successfully coordinating.

I haven't read them, but I'm seeing these two books recommended as easier introductions to developmental biology with a computing-y approach:

Wetware: A Computer in Every Living Cell
Life's Ratchet: How Molecular Machines Extract Order from Chaos
posted by clawsoon at 6:29 AM on March 7, 2021 [6 favorites]

This is called the “Paradox of the Heap”

Also called a “sorites series” or “sorites sequence.” (“soros” means heap, per that article) That’s one of those jargon words that I think should enter mainstream vocab because it comes up all the time!
posted by yarrow at 6:59 AM on March 7, 2021 [4 favorites]

So how does all that relative movement happen? Is it mostly driven by different rates of cell replication (slowly growing/dividing areas tending to 'fall' to the center of the zygote while fast growing tissues spread outward as all those squished-together cells push against their neighbours)? ...or are the blighters growing legs?

Great question, and this is definitely starting to get above my pay grade as a mere neurobiologist. I think the answer is that for the kind of large-scale coordinated "flow"-like movement you're pointing to here, it's differential rates of cell replication/expansion that push/pull the tissues around, whereas when you see individual cells flitting around through the larger matrix of tissue, that's the ameboid motion that clawsoon described above. But I'm not actually totally sure about that!

For the first nine divisions the cells seem to divide in lockstep. At 0:52 we jump cut and have entirely lost synchrony. Is this just boring stochastic drift, or is there something else going on?

This is another one I'm not really sure about; it seems like something that is probably well-known in developmental biology, but just not by me. I suspect that there's probably some sort of signalling molecule that the cells use to remain synchronized during the first few divisions, but I don't really know! It could even be that as the number of cells increases and synchronization becomes harder to maintain, this itself becomes a signal that tells the cells that it's time to start differentiating from the initial totipotent stem cell type, but that's just speculation on my part.

If there are multiple ways to do something, there's a good chance that nature will do all of them.

Quoted for truth.
posted by biogeo at 7:02 AM on March 7, 2021 [3 favorites]

Oh, and the very first feature we see on the embryo, that first pore that forms at about 1 minute in?

That is what will become its butthole.

I saw that and thought how fun it would have been to have seen this video when I learned about deuterostome (iirc, the sea stars and chordates, who make their butts first) vs protostome (worms and molluscs and bugs, who make their mouths first). One of the things I remember from those classes, how very useful of my brain.
posted by jeather at 7:50 AM on March 7, 2021 [3 favorites]

I was momentarily kind of confused as to why the movie was such a revelation, but then it occurred to me that not everyone here has spent years teaching developmental biology. It’s funny, but the process was just so obvious to me after all that time helping students learn to differentiate between things like body plan development via involution (salamander) vs. ingression (bird/mammal) that it was hard to take a step back and remember that this is not at all intuitive to people who have not spent time learning it.

People SHOULD learn and see this though. The miracle to me is not that it happens, not that there are so many intricate dependencies that must be fulfilled for it to work. The miracle is that it goes right so often and wrong so infrequently. The odds of all this just working appear to be just so infinitesimal, but every single day it’s happening like clockwork in so many embryos. You take something like a roundworm, where every cell has a predetermined fate, where removal of one cell will leave a defective organism, and contrast with a mammal, where you can straight break the zygote in half or in quarters even and rather than going wrong you end up with duplicates or quadruplicates of your developing organism. You meet twins and it’s cool that they shared an egg that split by happenstance, and it seems fantastic until you meet a 9-banded armadillo and learn that they always have identical quadruplets, that this 4-way splitting of the embryo is the default for an armadillo, and suddenly the accidental creation of twins doesn’t seem quite as astounding.

This process though, and the question of when it is a salamander, and the misunderstanding of how development works. It’s one of the reasons that the pro-life across America billboards offend me so deeply. Not the fact that they are pushing for elimination of abortion - they have the right to voice their opinions - but the incredibly bad science they use to push this! It makes me angry. The billboard will feature a chubby blue-eyed infant with a phrase like “I had eyes 14 days after conception!” Look at the video, around the 3:49 mark. That barely-defined protrusion from the neural tube? That’s a bulge that will eventually cause the epithelium overlying it to cup inwards and pinch off, to form the eye. The bit of brain in the back will become the retina. It looks nothing like an eye, it does not yet function, but the pro-life crowd knows that most people don’t know or can’t understand the nuance here. That will become the eye, so calling it an eye means the general public will think that at 14 days there is a wholly-formed microscopic human in utero, when really it’s a barely defined loose collection of cells just beginning to organize and assemble into what may eventually be an independent organism. Hundreds of years ago science banished the notion of the homunculus, only to now have the anti-abortion crowd doing their damndest to resurrect the idea to serve their own political purposes. You an argue for or against abortion all you want, but you should not be using shitty science to do so.
posted by caution live frogs at 8:21 AM on March 7, 2021 [14 favorites]

What an amazing video-- life is ridiculous :)
posted by Static Vagabond at 10:57 AM on March 7, 2021

A newt?!?

An eft.
posted by ovvl at 11:09 AM on March 7, 2021 [3 favorites]

What's the scale of this? It almost seems like the first cell is the size of the final salamander, but that can't be right, or?
posted by ymgve at 2:02 PM on March 7, 2021

The initial diameter of the newt's egg is about 2 mm, and the final larva isn't much bigger. Here are some photographs of the stages of development of the axolotl for comparison, which are all at the same scale and have a handy 1 mm scale bar. Keep in mind that during early development, the embryo isn't feeding or taking in nutrients, so it can't really grow much: it has to work with what it was provisioned with in the egg. Early development is characterized by division and differentiation, rather than growth. Growth comes later once the animal has developed enough to be able to feed (or otherwise acquire nutrients) so it can add to its total mass.
posted by biogeo at 3:35 PM on March 7, 2021 [6 favorites]

A newt?!?

An eft.

So are you a crossword aficionado or herpetologist? (I found a red eft in my yard a few days ago; a nifty little harbinger of spring.)

FYI it becomes a salamander at 05:11. Let me know if you have any further questions.
posted by TedW at 7:09 PM on March 7, 2021 [1 favorite]

TedW, that's hopelessly imprecise. You're going to need to provide screenshots of the frame immediately before the becomes-a-salamander instant and the one immediately after it to be at all convincing. And if those frames straddle an editing jump cut, all bets are off.
posted by flabdablet at 8:37 PM on March 7, 2021 [3 favorites]