De novo origins of multicellularity in response to predation | Hacker News Reader

De novo origins of multicellularity in response to predation

nature.com

92 pointsmazsa46 comments

This has big implications for the Fermi Paradox: https://en.wikipedia.org/wiki/Fermi_paradox

In particular, the jump from unicellular to multicellular life is (was) one of the top leading candidates for the Great Filter. https://en.wikipedia.org/wiki/Great_Filter

Importantly, it was one of the filter candidates "behind" us. It will be exciting to see if other potential filter steps can be so conclusively eliminated over time.

Exciting, but also slightly terrifying. I know this isn’t universal, but I, for one, very much hope that the filter is behind us and not in front of us.

That being said I’m not too shocked. The transition from unicellular to multicellular has always seemed like less of a hurdle to me compared to the start of unicellular life itself.

The jump to the start of unicellular life seems like the big stumbling block. It might be so incredibly rare that it's lucky for it to even happen once in the universe.

You need heredity, differential reproduction and selection. We really just need to start with a single case of a self-replicating system. Hard yes, but with countless trials running in parallel for millions of years it seems intuitively possible

If the universe exists for eternity, and in approximately the same form, then even very small odds may approach near certainty.

Evolution is a process rooted in randomness. The more faces a dice has, the longer you need to throw it to get the highest number. Or, reversely: the longer you have to throw your dice to get the highest number, the more faces it generally has.

It took "only" < 400 million years for life to develop, meaning unicellular life, but it took 3 billion years for multicellular life to develop. If we apply this basic rule, multicellular life seems the hard thing. Of course though, statistics wise this is pretty crappy ground to stand on because we only have one sample :).

The hard part is (probably) the transition from Prokaryotes to Eukaryotes. The transition from unicelular Eukaryotes to multicelular Eukaryotes is (probably) easy [1] .

Edit: [1] From https://en.wikipedia.org/wiki/Multicellular_organism#Occurre...

> Multicellularity has evolved independently at least 46 times in eukaryotes, and also in some prokaryotes, [...] . However, complex multicellular organisms evolved only in six eukaryotic groups [...]

"It took "only" < 400 million years for life to develop"

Yes, that's what people assume, but the universe existed for billions of years before that, so what if life might have used all that time to evolve before it landed on Earth deep inside some rocks?

One sample is hard to read because even low odds can happen. Picture the following statement: "The chance of life forming anywhere in the universe over a 100 billion year period is equal to 1 in Graham's number".

If this statement is true then it is incredibly unlikely life forms in our universe at our current stage exist. Low odds can occur and the only reason anyone is here to observe them is because they did occur. Since we only know of life forming once we cannot make assumptions about the odds of it happening.

Unicellular in the form of bacterium may in fact not be that much of a jump. It may even have happened twice in the form of archaea and bacterium.

The jump from prokaryotic to eukaryotic is far more intimidating.

That's actually happened a bunch of times, too. I'm just going to link to a comment I made a while back: https://news.ycombinator.com/item?id=18456049

"Diatom-derived plastids" blew my mind.

Not to be too pessimistic, but there's no reason that there should only be one "Great Filter". There could easily be multiple astronomically improbable events barring interstellar civilizations from bearing fruition. It might be just as improbable for unicellular life to form as it is for unicellular life to beget multicellular life, so on and so forth.

The great filter exists at all times as cited by the fact that the longevity of a species does not guarantee their future survival. All life may die off at any time.

Or even reason to assume that exobiotic life is cellular, and needed to jump from singular to multi-cellular structures like ours did on this planet.

I mean, in principle, but that's probably like assuming that exobiology is carbon based. Carbon and cells are both really super useful.

Is it really that much of a paradox?

- "Metallicity": We probably need a recent / "Population I" star, i.e. it' s probably unlikely that there can be civilizations much older than us (in astronomical dimensions): https://en.wikipedia.org/wiki/Stellar_population#Population_...

- In 1.1 billion years from now, the Sun will be 10% brighter than it is today, and this increase in luminosity will also mean an increase in heat energy, which Earth’s atmosphere will absorb. This will trigger a moist greenhouse effect here on Earth that is similar to the runaway warming that turned Venus into the hellish environment we see there today. So we have appeared only after 80% of the "usable" life span of our star has already been over -- a window that might be easy to miss (an larger stars burn even faster)

You don't need an astronomically old civilization for it to have galaxy-wide impact. Already a "few" million years should be enough to colonize a dead-except-for-you galaxy.

One of the interesting ones I've heard talk about lately is the "interstellar travel" step - Earth is on the lower end of the size range for rocky planets that can support an atmosphere and magnetic field to shield surface life from radiation. And because of the tyranny of the rocket equation, life developing on larger rocky planets may have a much harder time reaching orbit.

A good demo video: https://www.youtube.com/watch?v=amjuJJwI3iM

On a planet with twice the Earth's mass, it would take a larger-than-Saturn-sized rocket to put a Mercury-sized payload into orbit.

EDIT: And said Mercury-sized payload would have to spend more of its mass on heat shielding than the human Mercury capsule if it's intended to come back down.

Yes but the rocket equation wouldn't stop intelligent species from going "loud" with radio signals.

The rocket equation itself, no, but the planet mass would also have drastic effects on the evolution of species. On our planet, high density fibrous plants like trees have an advantage when it comes to access to the sun but what if on a bigger planet, that advantage goes to plants than can emit a spidersilk-like material to float on the wind in colonies? There goes a practically infinitely renewable building material that can be used for almost anything until the industrial age. How can a civilization get to the electronic age when they spend all their time between extinction events hauling rocks and mining ores to build 2 story dwellings?

There's just so many unknowns.

Not only dwellings but fuel as well. I mean, if you don't have trees how are you going to invent fire...

Some people consider cooking a critical step in our evolution (cooking enables harvesting more calories from food).

With no good fuel available cooking could also be ... off the menu.

Dry grass catches fire quite easily and high gravity shouldn't hinder the growth of low grass.

If not grass, dried escrements are also good fuel. People use them in areas of this world with little or no trees.

This wouldn't stop von neumann probes https://en.wikipedia.org/wiki/Self-replicating_spacecraft

The upmass requirement for bootstrapping that kind of system would be quite large.

Just to clarify for anyone unfamiliar with the US space program in the 50s-70s, or those just briefly confused by the discussion about planet sizes using planet names for things that aren't planets...

Mercury was a small space capsule containing 1 person designed to last a few hours in space. The Saturn V rocket was until recently the most powerful rocket ever launched, and took 3 crew, a crew module, a lander module, and payloads, to the moon.

Someone evolved a multicellular yeast with a similar approach back in 2011 or so. They used centrifuging to select for larger 'clumps' of yeast, rather than a predator to eat the smaller ones, but still pretty similar.

https://www.newscientist.com/article/mg21028184-300-lab-yeas...

Yes, although keep in mind we already knew of three independent evolutions of multicellularity — animals, plants, and fungi — so it arguably couldn’t be an improbable enough transition to qualify for great filter status.

Don't forget the humble Volvox!

https://en.m.wikipedia.org/wiki/Volvox

Do these cells already have the blueprints for multicellular organization, latent until selection exerts pressure for its manifestation? Is it much easier to reproduce the phenomenona since it's already happened before? Do cells ever go back to the single life?

The authors indicate the blueprints were already present but the multicellular expression stopped being optional in their evolved population:

The ability of wild-type C. reinhardtii to form palmelloids suggests that the founding population in our experiment already possessed a toolkit for producing multicellular structures. However, while the palmelloid condition is expressed facultatively in wild-type C. reinhardtii, the strains that evolved in our experiment are obligately multicellular.

Maybe I'm not familiar with how this term is used in biology but given the above it seems the "de novo" claim in the paper title is a little sensationalist.

I agree, but IANAB. From Wikipedia, this specie is of the genus Chlamydomonas of the family Chlamydomonadaceae https://en.wikipedia.org/wiki/Chlamydomonadales that include a lot of multicelular species. So, they have many multicelular relatives, we should ask a biologist if they have a multicelular ancestor.

That makes it way less interesting. The first time multicellular life happened it took billions of years and simple single cell organisms had to figure it out from scratch.

...or, it's not that different from dividing completely into two, and they just didn't do it for a long time because there weren't enough predation-based species to make it worth doing?

All of these questions are answered in the article.

But the answers are all less interesting than the headline, making me not want to take the time for the article.

I actually wrote a program which showed behavior that appeared to be "social" after predictors were added to the simulation.

The postulation being organisms have a pressure (from predators) to form groups and eventually societies. Essentially, they'd either survive by evolving to be social, evolve defense mechanisms or die out. Social evolution may actually be the shortest path for non-aggressive species because they simply have to bare one another, as opposed to evolve long claws or something.

Very hard to prove, but our model showed given the options social interactions appeared more likely with basic reward circuitry.

So the paper says that filter-feeding predators might have been the reason, if you are bigger than a certain size you become irrelevant to them. But how does the filter predator become so big without multicellularity? I mean: filter predators always have to be bigger than their food, no? So isn't it rather coevolution?

Cells can become quite large, the nerve cells in your back run the length of your spine for example, so it's certainly possible for a single cell to trawl through the ocean like a filter feeder and pick on smaller targets. Maybe tendril feeder would be a better description for a single cell though.

There's also caulerpa (https://en.wikipedia.org/wiki/Caulerpa) can be large but only have a single cell, googling turns up several other surprisingly large single cells.

Cells can become large, but can't a single big cell also escape a filter predator that's also a single big cell? There needs to be some reason for why in most cases, both are multicellular.

From a mechanics of materials perspective, having a larger size also means being easier to break. Multicellularity seems to be a way to have preferred breaking points that are cheaper and easier for the organism to repair.

> but can't a single big cell also escape a filter predator that's also a single big cell

Maybe if it's big enough, but there's a huge cost to being big and the cot is harder to bear for more herbivorous cells. For most organisms the better survival strategy is probably to be more numerous and accept a certain level of attrition, hence ants outnumbering humans.

Perhaps I'm missing something, but it seems they artificially selected for single-celled organisms that are sticky and clump together (thus making them -- as a group -- too large to be eaten.)

As the 'stickiness' doesn't really pose a disadvantage to the single-celled organism, the trait persists even after the predator is removed.

In short, can a collection of 'stickier-than-normal' single-celled organisms truly be referred to as a multicellular organism? Aren't they stretching the definition of multicellular? Each of the units, after all, reproduces on its own and there is no differentiation.

Not biologist, but I wonder whether these model organisms are valid. One might expect that today's single cell organisms may have had multicellular ancestors, but have kept the--I struggle for the word--the required genetic machinery for multicellular form in dormant/unexpressed fashion, but otherwise relatively intact and ready for expression given certain evolutiinary presures. This would be in contrast to the original evolution of multicellular life. Edit: I see others here have expressed this idea already, and more elegantly.

Great that I could just read all that info without jumping through hoops or just getting the abstract.

Interesting that the changes were then stable over many generations once the predator was removed.

As far as I understand this paper, the experiment did not generate a multicellular species. Perhaps the experiment actually generated an algae colony?

I guess this means that thousands of new multicellular “species” are born in the wild each day. Fascinating!

Reply on news.ycombinator.com