09. The First Animals

In the Mackenzie Mountains of northwestern Canada, a band of ancient reef limestone called the Little Dal Group holds branching, fibrous networks entombed inside 890-million-year-old microbial rock — the Tonian period, the opening act of the Neoproterozoic Era. Elizabeth C. Turner, of Laurentian University, published an analysis of these structures in Nature in 2021, interpreting them as fossilized tissue from a demosponge, one of the simplest kinds of sponge. If the interpretation holds, these would be animal fossils roughly 350 million years older than the next-oldest undisputed sponge body fossils, and they would predate the rise of atmospheric oxygen long thought necessary for animal life by about 90 million years. Turner herself, and outside commentators, are careful to call the finding tentative and dependent on later verification: no unambiguous spicules, the mineral or protein skeletal needles that support a sponge's body, were found alongside the fibrous networks, and the structures could instead be abiotic mineral growths or ordinary microbial-mat artifacts. Best treated as a disputed maximum, not an accepted date.

Firmer footing comes from chemistry rather than shape. A steroid molecule called 26-methylstigmastane is made, among living organisms, only by demosponges; its presence in ancient rock, reported by Love and colleagues in 2009, places definite demosponges at least 635 million years old. Demospongiae remains the most species-rich sponge class alive today, and one of its living members, Amphimedon queenslandica, a sponge native to the Great Barrier Reef, has become the standard genomic model for the whole basal sponge lineage — the organism whose DNA keeps answering questions no fossil can.

For a date on the whole animal kingdom rather than sponges alone, the reference point is a 2015 study by Mario dos Reis, Yuttapong Thawornwattana, Konstantinos Angelis, Maximilian J. Telford, Philip C. J. Donoghue and Ziheng Yang. Their method, a Bayesian molecular clock, treats the accumulated differences between living species' genes as a kind of stopwatch, converted into calendar time by anchoring select branches of the tree to fossils of known age. Using 203 nuclear proteins across 71 species and 38,577 amino-acid sites, and testing four different fossil-calibration strategies against one another, they placed the origin of crown Metazoa — the group containing every animal alive today plus its most recent common ancestor — in a 95 percent credibility interval of 649.8 to 833.5 million years ago, spanning the boundary between the Tonian and Cryogenian periods, the era of Snowball Earth glaciations.

The same study dated crown Eumetazoa (cnidarians plus bilaterians, effectively every animal with true tissue layers) to 625.9–746.2 million years ago, crown Cnidaria (corals, jellyfish, sea anemones) to 531.5–641.8 million years ago, and crown Bilateria (everything else, including every vertebrate) to 595.7–688.3 million years ago. The paper's real headline, though, is a warning rather than a number: its own authors conclude that building firm evolutionary narratives on molecular-clock timescales this deep remains premature, because unavoidable violations of clock-like behavior over such vast stretches of time, combined with genuine uncertainty in fossil calibration, keep the true dates from narrowing further. Before this sensitivity analysis, competing estimates for crown Metazoa in the literature had ranged as widely as 1,298 million years to 615 million years — a spread the dos Reis team's wide, honest ranges were built specifically to explain.

Among the animals whose age these methods try to pin down is one of the strangest: Trichoplax adhaerens, representative of the group called Placozoa. It is flat, about a millimeter across, brainless, gutless and muscle-less, built from only a handful of cell types — among the simplest known free-living animals. Yet its genome, sequenced and described in 2008 by Mansi Srivastava, Emina Begovic, Jarrod Chapman, Daniel S. Rokhsar and colleagues at the DOE Joint Genome Institute and UC Berkeley, runs to roughly 98 million base pairs and encodes a surprisingly rich set of transcription factors and signaling-pathway genes, the kind otherwise associated with the complex development of animals with true tissues. Molecular phylogenetics places the placozoan lineage's divergence from other animals at more than 600 million years ago; the genome paper's own analysis placed Placozoa inside a broader animal clade alongside cnidarians and bilaterians, with sponges branching off earliest of all. Several additional, genetically distinct placozoan lineages, hiding under what had looked like one species, have since been identified in comparative genomic work published in 2018. Exactly where Placozoa sits relative to sponges, ctenophores and the rest remains one of the least settled nodes in the whole animal tree.

A demosponge without a single neuron offers the clearest hint of what came before nervous systems existed. Amphimedon queenslandica, the same sponge used as the genomic model above, was shown in 2007 by Onur Sakarya, Kenneth W. Armstrong, Kevin J. Peterson and colleagues to carry, in its genome, a nearly complete postsynaptic protein scaffold — the molecular anchoring structure that, in an animal with a nervous system, holds signal-receiving proteins in place on the receiving side of a synapse. This includes the scaffolding gene dlg, known in humans as PSD-95, with the exact residues that contact its binding partner 100 percent identical between sponge and human, and co-expressed together during both larval and adult stages. Missing entirely from the sponge genome, however, are the actual ionotropic glutamate receptors that would dock onto that scaffold inside a real neuron. The reading is that the molecular hardware for a synapse assembled before neurons themselves existed, with a key receptor component added later, elsewhere in the tree. Follow-up work in 2019 found the same genes co-expressed in adult sponge choanocytes, the hair-ringed cells sponges use to filter food from water, reinforcing a pre-neural, non-neuronal function.

Two Candidates for First

Ctenophores, the comb jellies, complicate the picture further. Leonid L. Moroz, of the University of Florida, and independently Joseph F. Ryan and colleagues, sequenced ctenophore genomes — Pleurobrachia bachei by Moroz's team, and Mnemiopsis leidyi, at roughly 150 million bases, by Ryan's — both published in 2013 and 2014. Ctenophores plainly have a working nervous system. But it lacks every classic neurotransmitter found in other animals with nerves: no acetylcholine, serotonin, histamine, dopamine, adrenaline, noradrenaline or octopamine signaling turned up in either genome. In their place, an unprecedented diversification of ionotropic glutamate receptors — the same receptor family conspicuously missing from the sponge genome — appears to carry the signaling load instead. The implication argued by both teams is that ctenophore nervous systems were built from a different molecular parts list entirely, meaning they may have arisen independently of, rather than by inheritance from, the nervous systems of every other animal.

Whether that independence is real depends on a question still unresolved: which lineage, sponges or ctenophores, branched off earliest from every other living animal. The answer decides whether nervous systems evolved exactly once, later lost in sponges and placozoans, or at least twice, independently, once in ctenophores and again in the shared ancestor of cnidarians and bilaterians. Roberto Feuda, Davide Pisani and colleagues argue for sponges as the earliest branch, having reanalyzed multi-gene phylogenomic matrices in 2017 using improved models of amino-acid compositional differences between species, and concluding that the rival ctenophore-first result is a "tree-reconstruction artifact" produced by inadequate modeling. Joseph F. Ryan, Leonid Moroz, Casey Dunn and colleagues argue the opposite. In 2023, Darrin T. Schultz, Stephen A. Smith, Ryan and colleagues mustered a different kind of evidence for the ctenophore-first position: patterns of conserved gene neighborhood, called synteny, in which particular genes stay next to the same neighboring genes across enormous stretches of evolutionary time. Sponges, cnidarians and bilaterians share certain rare synteny patterns that ctenophores lack, a marker the Schultz team argues resists the modeling problems that plague amino-acid-based methods. A 2023 exchange in Nature Communications shows neither camp treats the question as closed.

The fossil record, when it finally arrives for ctenophores, arrives much later than any of these molecular estimates. The oldest confirmed ctenophore body fossil is an embryo from the Meishucun assemblage in Shaanxi, China, confirmed as organic rather than purely mineral preservation by Raman spectroscopy in 2007, and dated to the Early Cambrian, roughly 540 million years ago. Adult forms from the Chengjiang biota in Yunnan, China — among them Maotianoascus octonarius, built with skeletal support and as many as 24 comb rows for swimming, compared with just 8 in living species — date to about 518 million years ago. Both dates fall more than 100 million years after the Cryogenian-period origin the molecular clock implies for crown Metazoa, a textbook illustration of the gap that regularly opens between when a lineage's genes say it must already have existed and when the fossil record actually catches it in stone.

That fossil record, once it starts, is thin and easy to misread. A separate Chengjiang fossil, Stromatoveris psygmoglena, once proposed as a ctenophore relative, is now argued by some researchers to belong instead to an unrelated, frond-shaped group left over from the earlier Ediacaran period, called the Petalonamae — a reminder of how easily soft, frond-like Cambrian fossils get miscategorized. A further stem-ctenophore fossil, roughly 358 million years old, found at Miguasha in Canada and dated to the Late Devonian, shows the lineage persisting, sparsely documented, across the quarter-billion years between the Cambrian and the more familiar comb jellies alive today: some 100 to 150 living species, among them Mnemiopsis leidyi and Pleurobrachia bachei, all still running on the glutamatergic signaling pattern their Cambrian-and-older ancestors apparently invented.

None of these animals, brainless sponge, cell-poor placozoan, or nerve-net-bearing ctenophore, had anything resembling a mind. What they had, whichever of them came first, was ancient signaling chemistry, already billions of years old, finally being wired for the first time into cells whose only job was passing a signal on. The molecules that follow next are not simple amines mopping up radicals or nudging a cell's metabolism. They are peptides, small chains of amino acids, built for a purpose none of these early animals yet needed: holding one animal to another.

Overturned if: a phylogenomic method accepted by both the sponge-first camp of Feuda and Pisani and the ctenophore-first camp of Ryan, Moroz and Schultz converges on a single, mutually agreed sister lineage to all other animals, settling whether nervous systems arose once or independently at least twice.

Sources

• 2021 · Turner, "Possible poriferan body fossils in early Neoproterozoic microbial reefs" — Nature 596, 87–91; biomarker evidence — Love et al., Nature 457, 718–721 (2009)
• 2015 · dos Reis, Thawornwattana, Angelis, Telford, Donoghue & Yang, "Uncertainty in the Timing of Origin of Animals and the Limits of Precision in Molecular Timescales" — Current Biology 25(22), 2939–2950 / PMC4651906
• 2008 · Srivastava, Begovic, Chapman, Putnam, Hellsten, Kawashima, Rokhsar et al., "The Trichoplax genome and the nature of placozoans" — Nature 454, 955–960
• 2007 · Sakarya et al., "A Post-Synaptic Scaffold at the Origin of the Animal Kingdom" — PLOS ONE 2(6), e506 / PMC1876816; follow-up — Najle et al., Scientific Reports (2019)
• 2014 · Moroz et al., "The ctenophore genome and the evolutionary origins of neural systems" — Nature 510, 109–114; 2013 · Ryan et al., "The genome of the ctenophore Mnemiopsis leidyi" — Science 342(6164), 1242592
• 2017 · Feuda et al., "Improved Modeling of Compositional Heterogeneity Supports Sponges as Sister to All Other Animals" — Current Biology 27(24), 3864–387031453-7); countered by — Schultz, Haddock, Bredeson, Green, Simakov & Rokhsar, "Ancient gene linkages support ctenophores as sister to other animals," Nature 618, 110–117 (2023)