09. The Opioid Peptides: The Pain Ledger

In Hans Kosterlitz's laboratory at the University of Aberdeen, John Hughes is working through pig brain. Speed is the whole problem. What he is hunting are peptides — short chains of amino acids — and a brain left standing digests its own peptides with its own enzymes faster than a chemist can pull them out, so the tissue has to arrive fresh and go straight into the grind. Brain in, homogenise, extract, fractionate, test, repeat, batch after batch, chasing an activity that behaves like morphine and that naloxone, the drug that does nothing but sit in opiate receptors, switches off. In 1975 Hughes and Kosterlitz, with Smith, Fothergill, Morgan and Morris, published the answer in Nature. Two peptides, five amino acids each, read out by dansyl-Edman sequencing and confirmed by mass spectrometry: tyrosine-glycine-glycine-phenylalanine-methionine, and the same chain ending in leucine instead. They called them the enkephalins. The brain had been making its own morphine all along.

That sentence reads backwards from how the science actually ran. Two years earlier, Pert and Snyder had found the lock before anyone had the key — stereospecific, saturable opiate binding in rat brain tissue, tagged with radioactive naloxone, with a dissociation constant of 20 nanomolar and a receptor density around 30 picomoles per gram of tissue. Their 1973 Science paper proved that a vertebrate nervous system contains a site built to hold an opiate. It could not say why. A receptor that binds a poppy alkaloid is either an extraordinary coincidence or evidence that the animal already makes something the alkaloid resembles. Hughes and Kosterlitz settled it: the receptor came first, and the poppy is an imitator.

The family kept growing. In 1979 Avram Goldstein's group, working from pig pituitary, isolated a third endogenous opioid — dynorphin-(1-13), roughly seven hundred times more potent than leucine-enkephalin in the guinea pig ileum assay, and fully reversed by naloxone. Sixteen years later a fourth arrived, discovered twice in the same year by two teams working independently: Meunier's group in Nature and Reinscheid's in Science, both reporting the natural ligand of a receptor that resembled the opioid receptors but refused to bind any known opioid. It is called nociceptin. Three peptides in six years, a fourth two decades later, and — as it turned out — four separate genes behind them, not one gene read four ways.

The sequences give away the relationship immediately. Met-enkephalin and Leu-enkephalin come from the gene PENK; the dynorphins from PDYN; β-endorphin from POMC, a gene already famous as a pituitary stress hormone before anyone knew it was also an opioid. All three products open with the identical four residues, Tyr-Gly-Gly-Phe, before diverging. Dynorphin A does not merely resemble Leu-enkephalin — its first five residues are Leu-enkephalin itself, with the chain simply allowed to run on. β-endorphin opens with Met-enkephalin in exactly the same way. This is not three designs converging on one solution. It is one ancestral sentence, copied, and then finished differently three times.

It also explains the poppy. The receptor's binding pocket recognises a specific physical arrangement — a positively charged nitrogen held at a fixed distance and angle from an aromatic ring — and every one of these peptides presents exactly that arrangement in its first residue, tyrosine. Morphine's fused ring system happens to reproduce the same geometry closely enough to be read as the same key. The plant did not evolve to speak to vertebrates; the vertebrate receptor evolved for a message the plant's molecule imitates by accident. Nociceptin is the exception that confirms the rule. Its chain opens Phe-Gly-Gly-Phe — phenylalanine where the others carry tyrosine, a single missing hydroxyl group at position one. That one substitution is very likely why nociceptin does not activate the classical receptors, and why naloxone, a molecule built to compete with a tyrosine-anchored ligand, cannot touch the fourth receptor at all. One family, one shared opening line, and one member that lost the exact letter the others kept.

Underneath the chemistry sits an unusually clean piece of genome history. There are four peptide precursor genes — POMC, PENK, PDYN, PNOC — and four receptors: mu, delta, kappa and the nociceptin receptor, MOR, DOR, KOR and NOP. Four and four is not a coincidence. Dan Larhammar's group at Uppsala showed in 2008 that the four receptor genes sit inside a chromosomal region carrying at least fourteen other gene families, every one of them showing the same fourfold duplication pattern — synteny, the tendency of genes to keep their neighbours across hundreds of millions of years, doing the work that sequence similarity alone cannot. That signature is the fingerprint of two rounds of whole-genome duplication at the base of the vertebrates, the same 2R event Vol. I's chapter on the bonding peptides used to date the oxytocin family. The peptide side is messier and the same group said so: their 2010 analysis puts three of the four precursor lineages on the whole-genome duplications and the POMC/PNOC separation on a smaller local duplication, on roughly the same timescale but by a different mechanism.

POMC is the strangest gene in the set. It is transcribed once and then cut apart differently in different tissues, so that a single messenger yields adrenocorticotropic hormone, which drives the stress axis; the melanocyte-stimulating hormones, which set skin and coat colour; and β-endorphin, which relieves pain. Dores and Baron trace this architecture to an ancestral opioid-coding gene that acquired its melanocortin cassette around the earlier single chordate duplication, with the second round adding a further segment on top. Stress, pigmentation and analgesia are not three genes that happen to sit near each other. They are three phrases in one continuously read sentence, separated only by which enzyme does the cutting. That is very likely why POMC survives essentially unchanged in every vertebrate lineage examined: losing it does not cost one function, it costs three at once.

The System That Arrived Whole

How far back does any of it go? The cleanest answer arrived only recently. In work published online in 2020 and carried in the journal's 2022 volume, Huang and colleagues sequenced a functional mu-opioid receptor and two enkephalin-like precursor genes directly out of the Pacific hagfish, Eptatretus stoutii — a jawless fish, one of the two surviving cyclostome lineages — and found the lamprey, the other, carrying a clearly related receptor sequence. That places a working opioid system at the root of Vertebrata itself, in the neighbourhood of 550 million years ago: not a molecular-clock estimate but a bounded one, sitting below the split between jawless and jawed vertebrates that Vol. I placed at roughly 560 million years, and cross-braced by synteny and phylogeny from the other direction. The authors describe it as earlier than any previously authenticated account. What is remarkable is not the date but the shape. The system does not appear as one gene that later diversified. It appears already quadruple, already differentiated, already the four-and-four architecture a human carries — pain relief shipped as a complete package, in a single commit.

Whether anything opioid predates that root is the chapter's genuinely open question, and it is where the evidence grades earn their keep. From 1989 onward, George Stefano, with Cadet and Scharrer and a run of later collaborators, pursued a long, peer-reviewed program reporting opioid peptides and morphine-like activity in invertebrate tissue — mussel, leech, and immune cells — using immunoreactivity, chromatography and pharmacology; Laurent and colleagues reported related findings from leech ganglia in 2000. Against that stands one much newer result: the same 2020 hagfish paper ran a genome-wide search across invertebrate genomes and found no opioid receptor sequences and no opioid precursor analogues anywhere, despite deliberately looking. This does not prove the earlier data were artefacts on their own terms — an antibody that binds and a tissue that responds are real observations. It means that whatever those assays detected was very likely not a vertebrate-homologous opioid gene product. And the alternative readings are not hypothetical: morphine itself has been detected in hay, lettuce, human and cow milk, and commercial rodent feed, which is why the modern standard for claiming an animal makes its own opiate is isotope-labelled precursor tracing — watching the cell build the molecule — rather than finding the molecule present. Grade C, not D. This is a named dispute with data on both sides, not a bare hypothesis, and it is decided by a genome or not at all.

The human end of the system is less tidy than its genetics. Henning Boecker's team PET-scanned ten athletes before and after a two-hour run using a tracer that competes with the body's own opioids for receptor space; binding fell across prefrontal and orbitofrontal cortex, anterior cingulate, insula and temporoparietal regions, and the size of the fall tracked how euphoric each runner said he felt. That was the first direct image of runner's high in a living brain, and it mattered because the field's previous proxy — β-endorphin measured in blood — cannot demonstrate anything central, since peptides do not cross the blood-brain barrier efficiently. Then, in 2015, Fuss and colleagues found that in mice, blocking opioid receptors left running's anxiety- and pain-reducing effects essentially intact, while blocking cannabinoid receptors abolished them. The two results are not a direct contradiction — different species, and different halves of the phenomenon — but the popular single-molecule story does not survive meeting both. Runner's high is partly opioid at most.

The one place the receptor's role in a human is unambiguous is stranger still. In 1978, Jon Levine, Newton Gordon and Howard Fields gave patients recovering from third-molar extraction an injection of saline described as possible pain relief. Those who reported it working were then given naloxone without being told, and their pain came back. Reported in The Lancet, it remains the cleanest demonstration that a person can manufacture analgesia from expectation alone, running it through the same receptors morphine occupies. That fits what the endogenous opioid system appears to be for. As Richard Bodnar's long-running review series frames it, analgesia is not a standing discount on pain — pain is useful, and an animal that could not feel it would not last — but a conditional override, triggered by stress, that lets an animal finish a fight or an escape and settle the bill afterwards.

The same pocket has a darker salience. Fentanyl is a synthetic molecule that clinical pharmacology places at roughly fifty to a hundred times morphine's potency at that receptor, precisely because of how efficiently it fits. The most recent finalised United States figures record 79,384 overdose deaths in 2024 against 105,007 the year before — an age-adjusted rate down 26.2 percent, the steepest single-year fall of the decade — with opioid-involved deaths dropping from 79,358 to 54,045 and the synthetic-opioid rate, fentanyl's category, down 35.6 percent: a real reversal in a crisis that is not over, and a figure worth re-checking against newer data before it is read as a trend. Naloxone is the thread running through the whole century: the radioactive probe that first tagged the receptor in 1973, the drug that took back a placebo's borrowed relief in 1978, and — under other names, at higher doses, carried now by paramedics and bystanders — the thing that can still displace a fentanyl molecule from a 550-million-year-old binding pocket fast enough to restart breathing.

The honest standing is lopsided in an unusual way. The genetics are as solid as anything in this book — two independent kinds of evidence, chromosomal neighbourhoods and gene trees, agreeing with a functional receptor pulled from a living jawless fish, all pointing at one moment near the vertebrate root when four peptides and four receptors came into existence together. The behaviour is softer: analgesia is confirmed in mammals and inferred in everything older, reward has a floor but not a bottom, and the one claim that would break the whole frame — a real opioid gene in a snail or a worm — has a research program behind it and no genome to show for it. That is the shape of the ledger below, and it is the eighth and last of them. Every family in this book now carries its own dates, its own grades, and its own particular way of being wrong; what is left is to lay all eight side by side and read what the log itself says.

Commit Ledger

Overturned if: a confirmed opioid precursor or receptor gene turns up in any protostome genome — mollusc, annelid or arthropod — verified by chromosomal synteny or robust phylogeny rather than by immunoreactivity, which would move this family's origin off the vertebrate root and rewrite every date above it.

Sources

• 1973 · Pert & Snyder, "Opiate Receptor: Demonstration in Nervous Tissue" — Science
• 1975 · Hughes, Smith, Kosterlitz, Fothergill, Morgan, Morris, "Identification of two related pentapeptides from the brain with potent opiate agonist activity" — Nature
• 1978 · Levine, Gordon & Fields, "The Mechanism of Placebo Analgesia" — The Lancet — year, venue, authors and finding confirmed; volume and page not independently verified for this chapter
• 1979 · Goldstein, Tachibana, Lowney, Hunkapiller, Hood, "Dynorphin-(1-13), an extraordinarily potent opioid peptide" — PNAS
• 1989 · Stefano, Cadet & Scharrer, "Stimulatory effects of opioid neuropeptides on locomotory activity and conformational changes in invertebrate and human immunocytes" — PNAS; title, venue and year as indexed, volume, page and DOI not independently verified for this chapter
• 1995 · Meunier et al., "Isolation and structure of the endogenous agonist of opioid receptor-like ORL1 receptor" — Nature
• 1995 · Reinscheid et al., independent report of the same peptide — Science; exact title not independently verified for this chapter
• 2000 · Laurent et al., opioid-peptide findings in leech ganglia — European Journal of Biochemistry; exact title not independently verified for this chapter
• 2008 · Boecker et al., "The Runner's High: Opioidergic Mechanisms in the Human Brain" — Cerebral Cortex
• 2008 · Dreborg, Sundström, Larsson & Larhammar, "Evolution of vertebrate opioid receptors" — PNAS
• 2010 · Sundström, Dreborg & Larhammar, "Concomitant Duplications of Opioid Peptide and Receptor Genes before the Origin of Jawed Vertebrates" — PLOS ONE
• 2011 · Dores & Baron, "Evolution of POMC: origin, phylogeny, posttranslational processing, and the melanocortins" — Annals of the New York Academy of Sciences
• 2013 · Laux-Biehlmann, Mouheiche, Vériépe & Goumon, "Endogenous morphine and its metabolites in mammals: History, synthesis, localization and perspectives" — Neuroscience
• 2015 · Fuss et al., on the cannabinoid dependence of a runner's high in mice — PNAS; exact title not independently verified for this chapter
• 2018 · Bodnar, "Endogenous opiates and behavior: 2016" — Peptides
• 2020 · Huang, Taylor, Ghogha, Pribadi, Wang, Kim, Cahill, Coppola & Evans, "Genetic and functional analysis of a Pacific hagfish opioid system" — Journal of Neuroscience Research; published online 2020, carried in the journal's 2022 volume
• 2026 · Garnett & Miniño, "Drug Overdose Deaths in the United States, 2023–2024" — NCHS Data Brief 549, January 2026