Four ways to hear a hormone

The molecules in this reference look unrelated to the point of chaos. A gut peptide that times the stomach, a hormone that builds bone, a signal from fat that reports how much of it there is, a pituitary burst that orders puberty — nothing in the list seems to belong with anything else. But hormones do not act by existing; they act by being heard, and the machinery that hears them is not chaotic at all. Almost the entire catalog collapses into four ways a cell can receive a peptide. Learn the four and the field reorganizes along lines the molecule names hide.

The claim worth holding onto is stronger than “there are four receptors.” It is that the receptor class tells you more about how a hormone behaves than the hormone’s own identity does — its speed, the way it breaks, and whether anyone can build a drug against it. Two peptides that share a receptor class behave like cousins even when they do unrelated jobs; two that do similar jobs through different classes behave like strangers. So the right first question about any peptide on this site is not what it does, but how it is heard.

The overwhelming majority of the catalog is heard the same way — by a G-protein-coupled receptor, a single protein chain that threads back and forth through the membrane seven times. When the peptide docks on its outer face, the coil shifts and, on the inside, switches on a G-protein, which in turn fires a second messenger: usually cAMP (through a Gs protein) or a pulse of calcium (through Gq). The list is enormous — GLP-1 and glucagon, PTH, the melanocortins, oxytocin and vasopressin, the reproductive axis from GnRH down, somatostatin, ghrelin, the gut peptides, amylin’s receptor built from a calcitonin core. If you pick a peptide hormone at random, the smart bet is that a GPCR hears it (Lefkowitz & Kobilka, Nobel Prize, 2012).

The reason this one design dominates is that it is a fast, amplifying, re-usable switch. One bound peptide turns on many G-proteins, each spilling a burst of second messenger, so a handful of molecules outside the cell become a loud signal inside it — in seconds, and reversibly. That suits the moment-to-moment control most hormones are for. It also carries the family’s signature weakness: over-stimulate a GPCR and it desensitizes — pulled off the surface and quieted. That single fact is the engine behind an idea this reference meets again and again, that the body keeps time: a pulsed signal lets the receptor recover between beats, while a continuous maximal one erases the pattern and reads as off. Pulsatility is not a quirk of a few hormones; it is what a desensitizing receptor forces on any system that wants to keep listening.

The first great exception is small in number and huge in consequence: insulin and the two IGFs. Their receptors are not switches that borrow an enzyme — they are the enzyme. A receptor tyrosine kinase sits as a pre-formed pair; the hormone binds across it, the two halves phosphorylate each other, and the receptor’s own newly active tail becomes a docking scaffold that recruits the PI3K/Akt and MAPK pathways (receptor autophosphorylation, 1994). No G-protein, no cAMP — a different machine entirely, built for growth and fuel storage rather than a quick reversible nudge.

This is why insulin and IGF biology always sits a little apart from the hormones around it, and why the family has its own idiosyncrasies: a decoy IGF-2 receptor that signals nothing and exists only to clear the hormone, and engineered analogs like IGF-1 LR3 that win their potency by dodging the binding proteins rather than by hitting the receptor harder. Same receptor class, same logic — a self-activating kinase — running underneath all of it.

The third way is the cytokine receptor, and it is the one that produces the most counter-intuitive behavior. This receptor owns no enzyme of its own. Binding simply clamps two receptor halves together, which brings two tethered JAK kinases into contact; they activate, phosphorylate the receptor, and recruit a STAT protein that — once tagged — detaches, travels into the nucleus, and turns genes on directly (GH receptor activates JAK2, 1993). The signal’s destination is the genome, so its effects are gene-expression effects, and they take hours, not seconds.

Three catalog hormones you would never group by function are siblings here: growth hormone (JAK2/STAT5, driving IGF-1 transcription in the liver), leptin (JAK2/STAT3 in the hypothalamus), and the erythropoietin whose tissue-protective fragment became ARA-290. The shared receptor class explains the shared feel — slow, transcriptional, built to change what a cell is making rather than to flip it on and off. It even clarifies an exception-to-the-exception: adiponectin’s receptor looks like a GPCR’s seven-pass coil but is wired backwards and signals as a membrane enzyme instead (AdipoR1/R2 cloning, 2003) — a reminder that the shape of a receptor is not the same as its class.

The fourth way is the rarest and the tidiest. Here the receptor is a guanylyl cyclase: the peptide binds the outer face and the receptor’s inner face directly manufactures cGMP, collapsing the whole relay into a single protein — no G-protein, no borrowed kinase, no second cast member (natriuretic peptide receptor-A, 1991). In this catalog that design belongs almost entirely to one family: the natriuretic peptides — ANP, BNP and CNP — through which the heart acts as a gland, telling the kidney and vessels to shed salt and relax. cGMP as the vasodilation messenger is the through-line that links these peptides to nitric oxide and to the drugs that work on the same cGMP axis.

There is a line this clarity keeps tempting the page to cross. Knowing exactly how a hormone is heard — the coil, the kinase, the second messenger — feels like knowing what it does. It is not. A receptor and a second messenger are the bottom rung of the evidence ladder: they establish that a molecule can be heard, which is abundant and cheap, and only loosely tied to whether the molecule helps a person, which is rare and expensive. “It binds the GLP-1 receptor” and “it produces the outcome you wanted” are different claims on different rungs, and the receptor story only ever buys you the first.

So the four listeners are a map of the mechanism, not a verdict on the molecules. That is exactly the right use of them, and a genuinely powerful one: the class tells you how a hormone will move, how it will break, and whether it can be drugged — the whole grammar of how the catalog signals — while leaving the question of benefit where it belongs, with the trials. Bullish on the science; the science here is how a cell listens.

• Almost everything is a GPCR. A seven-pass receptor that switches on a G-protein and a second messenger (cAMP or calcium) — GLP-1, PTH, melanocortins, oxytocin/vasopressin, GnRH, somatostatin, ghrelin, the gut peptides, amylin. Fast, amplifying, and the most druggable class there is.
• Insulin and the IGFs are receptor tyrosine kinases. The receptor is its own enzyme: it phosphorylates itself into a scaffold for the PI3K/Akt and MAPK pathways. A growth-and-metabolism machine, not a cAMP switch.
• GH, leptin and EPO are cytokine receptors. No enzyme of their own — they borrow a JAK kinase and send a STAT to the nucleus. Gene-expression effects that take hours, not seconds.
• The natriuretic peptides are guanylyl cyclases. The receptor makes cGMP directly, collapsing the relay into one protein. The heart's salt-and-vessel signal runs on the same messenger as nitric oxide.
• The class predicts behavior; it does not prove benefit. Receptor class sets speed, desensitization and druggability — but a receptor is the bottom rung of the evidence ladder. How a hormone is heard is not whether a molecule works.