10. The Bonding Peptides

On a warm night beside a rain pool in Panama, a male túngara frog inflates his throat and calls. The call is the whole of his advertisement: a whine sliding into a chuck, aimed at any female within earshot. Researchers who inject him with a single molecule called vasotocin do not change his desire to call. They change the call itself — its fine structure, and how readily other males move toward it. A single messenger molecule, hardly more than a handful of atoms strung into a chain of nine amino acids, is enough to retune a piece of mating behavior that evolution spent hundreds of millions of years building.

Nine amino acids joined end to end is called a nonapeptide — a name that describes only the length of the chain, not its job. In mammals, two members of this short family carry familiar names: oxytocin, tied to bonding, birth, and milk letdown, and vasopressin, tied to holding onto water in the kidney and to some of the same social wiring. Popular science writing sometimes calls oxytocin the love hormone, which oversells a molecule that also does plumbing work in the bladder and blood vessels. The underlying chemistry, though, is not oversold at all. The same nine-link peptide, wearing different names, still runs through worms, snails, fish, frogs, and birds — a thread of attachment chemistry older than bone.

The mammalian pair did not appear from nothing. Daniel Ocampo Daza, Christina Bergqvist, and Dan Larhammar, working across Uppsala University and UC Merced, traced the six-gene family of oxytocin and vasotocin receptors found in living jawed vertebrates back to one ancestral gene pair, sitting side by side on one ancient chromosome. That pair became six through two separate rounds of whole-genome duplication — an event in which an organism's entire genetic library is copied all at once, not gene by gene, doubling every chromosome in a single generation. Most duplicate genes made this way are later lost, deleted as redundant; a few survive by finding new work. The oxytocin/vasotocin receptor family is a case where several duplicates survived, each specializing.

The evidence for two duplications, not one, is written into the chromosomes themselves as synteny — the tendency of genes to keep the same neighbors and the same order along a chromosome even after millions of years of separate evolution. The surviving genes still sit in matching arrangements on four different chromosomes, a fourfold pattern that is very hard to produce by any process except two doublings of the whole set. The study appeared online on 22 November 2021 and in print on 3 February 2022, in Frontiers in Endocrinology, volume 12, article 792644, drawing on freshly sequenced genomes — the reedfish Erpetoichthys calabaricus, the thorny skate Amblyraja radiata, the Western clawed frog Xenopus tropicalis — to confirm the fourfold symmetry across a wider slice of the vertebrate tree than earlier work had reached.

On the question of age, the authors are candid about the limits of their own method. A naive molecular clock — reading the number of accumulated mutations as if it were a stopwatch — applied to this gene family, in their own assessment, is often "overestimated," sometimes reaching back beyond the protostome-deuterostome split: the boundary where one great branch of the animal tree, the protostomes (worms, snails, insects), separated from the other, the deuterostomes (sea stars, fish, and eventually us), conventionally placed just over 600 million years ago. Rather than trust that clock on its own, the team bounded the duplications between two independently dated events instead — no older than the split separating vertebrates from tunicates, no younger than the point where bony and cartilaginous fish parted ways. That places the ancestral nonapeptide gene system somewhere between 600 and 450 million years old: a wide window, honestly reported as a window rather than rounded to one reassuring number.

Not every specialist agrees on how it happened. A rival reconstruction, built by Theofanopoulou and colleagues from lamprey gene sequences, argues for only one whole-genome duplication, with the remaining gene copies produced afterward by smaller chromosomal duplications and translocations. Ocampo Daza's team rebuts this in the same paper, arguing that their two-duplication, fourfold-symmetry model is the more parsimonious account — parsimonious meaning it explains the same pattern using fewer separate assumed events, which is usually, though not always, the way to bet. The disagreement has not been settled by new data. It remains a live dispute about mechanism, layered under an already wide age estimate.

Whatever the precise mechanism, the family diversified, and every living jawed vertebrate carries some descendant of it. Ray-finned fish kept a version called isotocin. Birds, reptiles, and the mammals that never evolved a placenta carry mesotocin. Amphibians and most reptiles carry arginine vasotocin, differing from mammalian vasopressin by exactly one amino acid — a single spelling change in a nine-letter word, conserved since before mammals existed. Each name marks the same ancestral gene, inherited by a different branch and put to slightly different work.

Vasotocin is the best-mapped of the non-mammalian versions. A review led by Mario Penna at the Universidad de Chile, synthesizing decades of work by researchers including Boyd, Moore, Hollis, González, Smeets, and Vandesande and Dierickx, traces vasotocin-producing cells running from forebrain to hindbrain in newts, salamanders, frogs, and lizards — a wiring pattern conserved across the whole group. Newts named Pleurodeles waltl and Taricha granulosa carry it. So does the salamander Plethodon shermani. So do frogs across several genera: the European common frog Rana temporaria, the edible frog and marsh frog of the Pelophylax group, the American bullfrog, the common toad Bufo bufo, the Japanese toad Bufo japonicus, the African clawed frog Xenopus laevis, the cricket frog Acris crepitans, and the túngara frog from the opening scene. Two lizards, Lacerta muralis and Acanthodactylus pardalis, carry it as well. Across this list, vasotocin drives calling, clasping, and other pieces of courtship — but not as a fixed switch. In newts, the hormone only induces courtship behavior once the animal has already been exposed to the right level of a separate hormone, androgen; block that exposure and vasotocin does much less. The molecule sets a possibility, not an inevitability.

Ray-finned fish took the ancestral gene in an oddly contested direction. Babak Ataei Mehr, Shawn Garner, and Bryan Neff at the University of Western Ontario injected isotocin — the fish paralogue of oxytocin, paralogue meaning a gene copy descended from the same ancestral duplication rather than borrowed from a different species — into ninety-two adult guppies, forty-six of each sex. Guppies given isotocin spent twenty-nine percent more time near other guppies than guppies given a chemical blocker of the same receptor, with no difference between the sexes. The same molecule turns up in zebrafish, in goldfish, and in the cooperatively breeding cichlid Neolamprologus pulcher. What makes the 2020 guppy result worth flagging honestly is that it cuts against most of what came before it: the authors note plainly that earlier fish studies had generally found no social effect from isotocin, or even an antisocial one. This is one of the first clean demonstrations of a pro-social effect, not a settled rule for fish in general.

Birds run the same gene as mesotocin, and in gregarious species it does something distinctive: it holds flocks together, not just pairs. James Goodson and colleagues at Indiana University found that blocking central mesotocin receptors in estrildid finches cut the time these birds spent with large groups and with familiar social partners. In the zebra finch, mesotocin acting on midbrain receptors is necessary for forming a pair bond in the first place; in the pinyon jay, it is implicated in broader social bonding; the common waxbill carries it too. The same molecule that binds a mammalian pair together here holds together an entire flock — an old tool reused at a different social scale, not a contradiction so much as a demonstration of how readily one signaling pathway gets redeployed.

Beyond the Backbone

The nonapeptide family is older than backbones. In Caenorhabditis elegans, a roundworm about a millimeter long with no brain to speak of, a peptide called nematocin acts through a receptor named NTR-1 to let the worm learn an association between salt concentration and past experience — a capacity called gustatory associative learning, meaning the worm can change how it responds to a taste based on what happened the last time it met that taste. Isabel Beets and colleagues at KU Leuven, with a near-simultaneous and independent report from a second team, showed that worms missing nematocin or its receptor lose this learned response. Nematodes and vertebrates last shared a common ancestor before the protostome-deuterostome split, conventionally placed above 600 million years ago — pushing the functional origin of this signaling system deeper than any vertebrate fossil, into an ancestor whose nervous system had barely begun to specialize.

In the earthworm Eisenia foetida, the same family goes by annetocin, and it does something entirely different: injected in doses above five nanomoles, it triggers egg-laying — rotating body movements, a change in body shape, and the mucous secretion from the worm's clitellum that builds the cocoon holding its eggs. Chizu Oumi and colleagues at the Suntory Institute for Bioorganic Research described this in 1996. Earthworms and vertebrates, like nematodes and vertebrates, diverged before that same protostome-deuterostome split — a second, independent invertebrate lineage pointing at the same deep age from a completely different branch of the tree.

Snails carry a version called conopressin, and mollusks diverged from vertebrates at the same boundary, over 600 million years ago — a third independent phylum converging on the same depth. The great pond snail Lymnaea stagnalis shows something the vertebrate lineage no longer does: one single receptor, cloned by van Kesteren and colleagues at the Vrije Universiteit Amsterdam and named LSCPR, handles both the oxytocin-type reproductive signaling and the vasopressin-type metabolic signaling that vertebrates eventually split across two separate receptor families. The same peptide, chemically identified in the marine cone snail Conus geographus as well, still does both ancestral jobs at once in the pond snail — a plausible glimpse of what the system looked like before it specialized, and itself indirect evidence that the ancestral system had not yet split into separate paralogues this far back. Later work on the nudibranch Berghia stephanieae, published in 2024, extended this picture rather than challenging it.

Laid end to end, the family tree runs from a millimeter-long worm tasting salt, through an earthworm laying eggs, through a snail running two jobs on one receptor, up through fish that shoal, frogs that call, finches that flock, to the mammalian pair that gave the whole family its popular name. Six branches, one ancestral trunk, and the same short peptide doing recognizably related work in every one of them: some version of drawing together, of committing reproductive or metabolic machinery to another individual's presence.

None of this makes the peptide a single, simple lever. Its job changes with the wiring it lands in — a pair bond in one species, a flock in another, an egg-laying reflex in a third. What survives across six hundred million years and half a dozen phyla is not a fixed behavior. It is a chemical vocabulary word, reused wherever a nervous system needs to say: stay close to this one. Not every ancient messenger speaks about closeness. Some speak about rank instead — and one of them is about to raise its claws.

Overturned if: a confirmed nonapeptide-family gene — a genuine ortholog of nematocin or vasotocin — turns up in a cnidarian or sponge genome, pushing the family's origin back past the last common bilaterian ancestor this chapter treats as its deepest confirmed boundary.

Sources

• 2021 · Ocampo Daza, Bergqvist, Larhammar, "The Evolution of Oxytocin and Vasotocin Receptor Genes in Jawed Vertebrates: A Clear Case for Gene Duplications Through Ancestral Whole-Genome Duplications" — Frontiers in Endocrinology
• 2017 · Penna et al., "Arginine Vasotocin, the Social Neuropeptide of Amphibians and Reptiles" — Frontiers in Endocrinology
• 2020 · Ataei Mehr, Garner, Neff, "Effect of isotocin on shoaling behaviour of the Guppy (Poecilia reticulata)" — Animal Cognition, SpringerLink
• 2009 · Goodson, Schrock, Klatt, Kabelik, Kingsbury, "Mesotocin and Nonapeptide Receptors Promote Estrildid Flocking Behavior" — PubMed
• 2012 · Beets et al., "Vasopressin/Oxytocin-Related Signaling Regulates Gustatory Associative Learning in C. elegans" — Science
• 1996 / 1995 · Oumi et al., "Annetocin, an annelid oxytocin-related peptide, induces egg-laying behavior in the earthworm, Eisenia foetida" — PubMed; van Kesteren et al., "A Novel G Protein-Coupled Receptor Mediating Both Vasopressin- and Oxytocin-Like Functions of Lys-Conopressin in Lymnaea stagnalis" — PubMed