I’m not a fan of phys.org — they summarize interesting articles, but it’s too often clear that their writers don’t have a particularly deep understanding of biology. I wonder sometimes if they’re just as bad with physics articles, and I just don’t notice because I’m not a physicist.
Anyway, here’s a summary that raised my hackles.
Chromosphaera perkinsii is a single-celled species discovered in 2017 in marine sediments around Hawaii. The first signs of its presence on Earth have been dated at over a billion years, well before the appearance of the first animals.
A team from the University of Geneva (UNIGE) has observed that this species forms multicellular structures that bear striking similarities to animal embryos. These observations suggest that the genetic programs responsible for embryonic development were already present before the emergence of animal life, or that C. perkinsii evolved independently to develop similar processes. In other words, nature would therefore have possessed the genetic tools to “create eggs” long before it “invented chickens.”
First two words annoyed me: Chromosphaera perkinsii ought to be italicized. Are they incapable of basic typographical formatting? But that’s a minor issue. More annoying is the naive claim that a specific species discovered in 2017 has been around for a billion years. Nope. They later mention that it might have “evolved independently to develop similar processes”, which seems more likely to me, given that they don’t provide any evidence that the pattern of cell division is primitive. It’s still an interesting study, though, you’re just far better off reading the original source than the dumbed down version on phys.org.
All animals develop from a single-celled zygote into a complex multicellular organism through a series of precisely orchestrated processes. Despite the remarkable conservation of early embryogenesis across animals, the evolutionary origins of how and when this process first emerged remain elusive. Here, by combining time-resolved imaging and transcriptomic profiling, we show that single cells of the ichthyosporean Chromosphaera perkinsii—a close relative that diverged from animals about 1 billion years ago—undergo symmetry breaking and develop through cleavage divisions to produce a prolonged multicellular colony with distinct co-existing cell types. Our findings about the autonomous and palintomic developmental program of C. perkinsii hint that such multicellular development either is much older than previously thought or evolved convergently in ichthyosporeans.
Much better. The key points are:
- C. perkinsii is a member of a lineage that diverged from the line that led to animals about a billion years ago. It’s ancient, but it exhibits certain patterns of cell division that resemble those of modern animals.
- Symmetry breaking is a simple but essential precursor to the formation of different cell types. The alternative is equipotential cell division, one that produces two identical cells with equivalent cellular destinies. Making the two daughter cells different from each other other opens the door to greater specialization.
- Palintomic division is another element of that specialization. Many single-celled organisms split in two, and each individual begins independent growth. Palintomic division involves the parent cell undergoing a series of divisions without increasing the total cell volume. They divide to produce a pool of much smaller cells. This is the pattern we see in animal (and plant!) blastulas: big cell dividing multiple times to make a pile of small cells that can differentiate into different tissues.
- Autonomy is also a big deal. They looked at transcriptional activity to see that daughter cells had different patterns of gene activity — some cells adopt an immobile, proliferative state, while others develop flagella and are mobile. This is a step beyond forming a simply colonial organism, is a step on the path to true multicellularity.
Cool. The idea is that this organism suggests that single-celled organisms could have acquired a toolkit to enable the evolution of multicellularity long before their descendants became multicellular.
I have a few reservations. C. perkinsii hasn’t been sitting still — it’s had a billion years to evolve these characteristics. We don’t know if they’re ancestral or not. We don’t get any detailed breakdown of molecular homologies in this paper, so we also don’t know if the mechanisms driving the patterns are shared.
I was also struck by this illustration of the palintomic divisions the organism goes through.
a, Plasma membrane-stained (PM) live colonies at distinct cell stages, highlighting the patterned cleavage divisions, tetrahedral four-cell stage and formation of spatially organized multicellular colonies (Supplementary Video 5). b, Actin- (magenta) and DNA-stained (blue) colonies at distinct cell stages showcasing nuclear cortical positioning, asymmetrical cell division (in volume and in time) and the formation of a multicellular colony. This result has been reproduced at least three independent times.
Hang on there : that’s familiar. D’Arcy Wentworth Thompson wrote about the passive formation of cell-like cleavage patterns in simple substrates, like oil drops and soap bubbles, in his book On Growth and Form, over a century ago. You might notice that these non-biological things create patterns just like C. perkinsii.
Aggregations of oil-drops. (After Roux.) Figs. 4–6 represent successive changes in a single system.
Aggregations of four soap-bubbles, to shew various arrangements of the intermediate partition and polar furrows.
An “artificial tissue,” formed by coloured drops of sodium chloride solution diffusing in a less dense solution of the same salt.
That does not undermine the paper’s point, though. Multicellularity evolved from natural processes that long preceded the appearance of animals. No miracles required!
It should be pointed out that Chromosphaera are opisthokonts and therefore not too distant relatives of ours, as these things go, especially since complex multicellularity has been invented at least twice in opisthokonts. It’s quite plausible that some of the mechanisms might be ancestrally shared.
The byline for that summary is University of Geneva, so any problems with the piece belong to the university’s media team rather than phys.org’s writers (although, yes, phys.org does publish a lot of slightly-off reports on biomedical research and it would be nice in general if pop science outlets only published articles that could be effectively appraised by someone on their editorial team).
I agree that some of that pop summary is poorly written, especially the awful line about Nature creating eggs before inventing chickens, which is I suspect what raised your hackles. As it should. Bloody terrible sentence, a weird combination of obviously true (organisms that make eggs vastly predate chickens in evolutionary history) and implied teleology. Overall, though, I think it’s a reasonable summary, especially as it includes one of your main points towards the end: “either the principle of embryonic development existed before animals, or that multicellular development mechanisms evolved separately in C. perkinsii,” pretty much in keeping with the original paper.
The paper does provide evidence that these changes are old and not just a recent evolutionary development in C. perkinsii. See fig. 2d in the paper — a ‘heatmap’ of gene expression comparing C. perkinsii with other animals — and the paragraph immediately after that describes evidence these gene clusters are strongly conserved across many branches of Animalia, suggesting early evolution rather than recent functional convergence. The summary does address this, but it’s easy to miss.
As a physicist, I regard phys.org as sensationalist trash.
I always wonder whether second cousins count as close relatives, and these guys say our cousins who diverged a BILLION years ago are close.
As a physicit, I find phys.org a useful entry to the literature. But then the same is true for creationist websites. I have found many inportant science articles, thanks to creationists. Of course, the actual content of the article is anti-creationist. Not surprising, that is creationism in a nutshell, pretending science is on their side. Thanks guys!
As for “eggs before chickens” lighten up PZ. Of course we understand that chickens is just a metaphore! Creationists are beating the pant off us (metaphore!) in the making science user friendly department, giving people a simple summary that they will remember more than a week later, is part of that!
You’ll see the same thing, again, with climate deniers.
As for the article, it’s a good indication of how multicellular animals may have arisen gradually. First you have free swimming protists. Then some take to forming a colonial structure, say a small blob, that is harder for a single-celled predator to eat, under certain circumstances. Against predatory blobs, time to get bigger, but due to diffusion problems that would suffocate and starve the inner cells, that means becoming a hollow ball with a shell only one or a few cells thick. Now there are enough cells though to do a bit of rudimentary information processing beyond what single isolated cells can do, and to use coordinated flagella or what-have-you to move the whole colony around. Become elongated into a tube and you can move in other ways, snake-like undulations and such, to outrun predatory hollow balls. Open the ends of the tube and you can speed that up by accelerating water down the inside of the tube and jetting it out the rear. Oh, and food might be sucked in and can then be accessed directly by the innermost cell layer and not only the outermost, doubling feeding efficiency. Now get a bit smarter and add phototropism by putting light sensitive patches at the front along with a ganglion or so, and use the difference in light hitting two patches for steering. Stay where there’s light so you stay where there’s algae to eat. Specialize the layers: toughen the outer skin to defend against predators and parasites while the inner layer focuses on digestion and oxygenation. Oh, wait, have we become some early type of worm now, with the beginnings of eyes, a brain, a mouth, and an anus?
At the same time, the portion of the life-cycle spent as a colony grows, the portion spent as free-swimming single cells shrinks, until the role of the latter is reduced to sperm and white blood cells. Plants had a parallel track, with the diploid sporophyte stage growing more and more dominant in the larger and more complex species, while the haploid (but not always single-celled!) gametophyte stage shrank in size and duration and significance.
Of course, none of this is a linear progression: niches remain for every step along this path, and organisms remain occupying most of those niches. Sometimes the path outright branches, producing the phylogenetic tree. (Some of the earliest animal lineage divergences include which-end’s-the-anus, tube-becomes/stays-radially-symmetrical-critter vs. tube-flattens-into-mirror-symmetrical, and some other forks in that early part of the road envisioned above.)
Later steps for animals and plants included “make a strong, rigid structural material” (bone/wood) and “get a circulatory system to scale up” (xylem and phloem sap passageways/blood vessels), as well as “get a cardiorespiratory system to scale up even more” (leaf transpiration doing gas exchange while generating negative pressure in the xylem/hearts and lungs), not to mention “get armor and maybe make it spiky” (bark and thorns/shells of various kinds). Animals that could calcify up structural supports and armor could also calcify up a sword or so to attack armored prey, and we got teeth.
More branching strategies later and eventually one animal started using fire for a cellulosic energy powered external digestion step, giving its branch direct use of cellulosic energy for the first time and unlocking some new energy-dense foods to boot, allowing giant leaps in brain power; used that to scale up its social organizations by orders of magnitude and start forming colony organisms with their own analogues of “DNA” and “RNA” coding new behavioral and physical traits that could evolve at a much faster pace than their breeding speed would ordinarily permit; then exploiting that new social scale to collect, concentrate, and use fire to smelt metal ores, formerly too unevenly distributed for biology to effectively exploit except in trace amounts, such as iron for hemoglobin but not for large scale structural purposes or high durability moving parts: few individuals or even colonies of a species would have enough iron available to them to do that. The same species also started deliberately turning other species into symbionts with an assortment of useful functions, from little furry things that wipe out food competitors and plague carriers to big furry things that cooperate in hunting game, huge furry things that can be used as vehicles or vehicle engines, and a wide variety of food and medicinal plants. Transformed six of the seven continents beyond recognition with these. And that’s before starting to secrete external structures that could extract energy from the environment and then schlep it around as electric currents and even augment brain function, giving the colony organisms it formed their own nervous systems, muscles, and brains …
That’s probably a bell that won’t be unrung short of a planet-sterilizing-severity cataclysm. The planet is not so much in a mere sixth mass extinction as it is a transformation as major as the Great Oxygenation Event or the end of the Cryogenian. Every few billion years life unlocks some new trick for energy extraction, and then some new trick for faster adaptability that uses that added energy, and the whole place undergoes a phase change. Photosynthesis caused oxygenation caused respiration allowed eukaryotes and multicellularity. Plants, or at least eukaryotic algae, may have consumed so much CO2 they caused the Cryogenian and animal grazers restored the CO2 balance again to end it. (Then a second, smaller pendulum oscillation in the Carboniferous before it fully died down, triggered by vascular land plants this time and ended by a combination of land grazers and wood decomposition fungi. The end Permian, linked to a big volcanic CO2 burp, may have been made worse by overcorrection at the end of this second swing as a shit-ton of Carboniferous vegetable detritus decayed in a very short time.)
Life survived these changes but they defined the ends of eras and even eons: the Hadean ended with photosynthesis, the Archaen with oxygenation reaching respitation-supporting levels, and the Proterozoic with the Cryogenian’s end and the full flowering of multicellular life, with the Paleozoic era ending on that second, smaller pendulum oscillation.
In the deep-time perspective, the recent election is a tempest in a tea-pot. The Phanerozoic is on its way out the door no matter what, to be replaced by the Cyberbiotic, when life commands a much higher percentage of the planet’s available energy and is much less limited in what it can do with its raw materials, and more and more of it participates in fast-evolution methods that involve heavy signal processing and internal and external computation with electrical signals, with a shrinking (percentage-wise) role for DNA and RNA. Cyberbiotic life has the potential to spread beyond the planet as other than just single-celled hardy bacteria, too, given time. In the long run we won’t see starships, so much as we will see starseeds, whole ecosystems in a protective shell that maintain and propel themselves to reach new star systems and germinate therein. It’s going to look a lot less like Star Trek than like, I dunno, Olaf Stapledon, maybe.
But that is billions of years from now. Right now, we’re faced with the ructions of the Phanerozoic’s end, and can only scramble for personal survival in the upheavals to come, political, climatological, and ultimately ecological. The survivors will learn humility as cells in organisms like proto-arcologies and such that very much are not themselves under human control, even though they might seem so. Democratic decisions, even (especially!) when fully fair and uncorrupt, are a brain using humans as its neurons, with the role of any one human that of the guy in the Chinese Room thought experiment. And more and more that will involve AI and other nonhuman factors. Laugh as you will at ChatGPT hallucinations and emancipated lab monkeys and impending crypto bubble collapses; the groundwork has been laid. LLMs do in fact replicate a part of human cognition, the part that generates variations, dreams and creates and tries to build summaries and theories of things, grist for the selection mill. They lack said selection mill, the part that rejects stuff that is “unrealistic”, falsified by data, or what have you. When that selection mill is added on they will be immediately superhuman — said selection mill is the weak link in human cognition, famously failing every time a human is sucked in by QAnon style conspiracy thinking and other forms of wishful thinking. The AI version will quickly overmatch us, and when that happens AIs will be more reliably members of the reality-based community than humans. And yes, that also reveals that cognition is nothing more or less than evolution itself, but cyber-ized with access to virtual worlds (based on themselves-evolved models) to test and select things in before investing in them in the physical world, and operating at the speed of neurons, and soon electronic VLSIs or even more advanced hardware. You’ve evolved to evolve things in your head so you can then make arrowheads or automobiles or lesson-plans that are already highly optimized when first physically built, seemingly springing fully-formed from the brows of eight billion Zeuses, in turn so that the superorganisms humans compose can evolve new kinds of legs and skin and teeth and claws and digestive organs and even brains, from cavalry to smelters and abacuses to AIs, that use materials formerly inaccessible to life and develop and adapt at speeds that would make a virus blush. An advantage so huge it must necessarily trigger a phase change as beautiful and terrible as those spurred by photosynthesis and eukaryotes and life’s first colonization of the land.
Welcome to the Cyberzoic.
(Anthropocene? You’re kidding, right? That’s rather … anthropocentric. Stuff that lives around us is already learning to understand more and more human speech. Soon it’ll start talking. Being able to negotiate with (or lie to and manipulate, even) the big hairless apes will be a giant advantage to any species that figures out how. And that’s before we consider deliberate “uplift” to make our symbionts work with us and our stuff even better than before. Between nonhuman animals gaining greater cognitive and speech capabilities and AI, we’re not going to be the leading (bleeding?) edge for much longer. And we’re already finding the larger superorganisms that have grown around us to be increasingly beyond our own control as it is. That got more so when bureaucracies got computers, and AI and talking dogs and such will only exacerbate that trend. Firms, cities, countries, and multilateral organizations all have lives of their own and will just shed and replace humans, or any other components, that don’t fit in with their agendas. The nascent Cyberzoic is not and will never be anthropocentric, and so it will not be, or contain, an Anthropocene. The short period of human domination ran from about 100,000 BCE to about 1 CE, starting with the main wave of migrations out of east Africa, and the megafauna extinctions and land use transformations accompanying that wave, and ending with the Roman Empire or thereabouts, the beginning of superorganisms that, while incorporating humans, were beyond human control and took on lives and minds of their own. But it’s too short to be more than a coda to the Pleistocene, not an era but just an age, for which a name already exists: the Late Pleistocene. From a far future perspective, it will be seen as ending the Pleistocene, the Neogene, the Cenozoic, and the Phanerozoic. The expansion of Rome began the Cyberzoic, the era of continent-scale superorganisms incorporating iron and concrete, roads and plumbing, as well as animal and plant “cells”, with wills of their own that could overpower even the collective wills of the smartest of those cells acting in concert. The future that is taking shape around us now is the Cyberzoic entering full flower; the transition will be complete with the first human/uplift/AI-brained sustainable ecosystem, probably emerging somewhere in Southeast Asia. As for what will follow it, well obviously that’ll be the Astrozoic, when self-contained starseed ecosystems begin flying outward from Sol like dandelion fluff. The Cyberzoic will resemble “Kardashev Type I” and the Astrozoic “Kardashev Type II”, but those classifications always presupposed human-centric “civilizations” rather than intelligent whole ecosystems with humans as just one component and not even the sole neuron-type, or ETI equivalents. These likely will never exist as such. The contemporary SF author to get things mostly right is revealed to be none other than Karl Schroeder, with his thalience and Artificial Nature and deodands. Cross all three and you get the Cyberzoic’s … ecozoans, I suppose you might call them.)
Now to go make popcorn and watch jackasses like Jinping, Putin, and Trump (not to mention their mini-mes like Jong Un, Orban, and Oilievre) get overtaken by events and made irrelevant. :)