The heart is a gland

For most of the history of physiology the heart had one job: move blood. Then, in 1981, a deceptively simple experiment reassigned it. An extract of atrial muscle, injected into rats, produced a rapid and powerful flush of sodium and water from the kidney (de Bold et al., 1981). Something inside heart tissue was a hormone. The heart, it turned out, is also a gland.

What it secretes are the natriuretic peptides. When the heart wall is stretched, by the extra volume or pressure of a body carrying too much fluid, cardiac muscle releases them into the blood: ANP from the atria, BNP from the ventricles. Their instruction to the rest of the body is, in effect, shed the load. And because the load they relieve is the same stretch that released them, the heart is not just sensing its own strain. It is doing something about it.

Read as a control system, the natriuretic peptides are a negative feedback loop, a thermostat for the circulation. Stretch releases the peptide; the peptide binds a receptor called NPR-A and raises the second messenger cGMP; cGMP drives the kidney to excrete sodium and water and the vessels to relax; blood volume and pressure fall; and the fall relieves the stretch that started it. The loop closes on itself. (The cGMP detail is worth a beat: most hormones in this catalog work through cAMP and a G-protein-coupled receptor, but the natriuretic receptors are themselves enzymes, guanylyl cyclases, a genuinely different signaling family.)

The loop does not run alone. It is the standing opposition to the renin-angiotensin-aldosterone system, the body’s dominant fluid-retaining axis: where angiotensin II and aldosterone constrict vessels and hold on to salt and water, the natriuretic peptides dilate and release them. Most of cardiovascular medicine is a negotiation between these two forces. The heart’s own hormone is the counter-weight.

Because BNP rises in direct proportion to how hard the ventricle is being stretched, its level in the blood carries a clean diagnostic signal: a heart under strain announces itself. That turned a hormone into one of the most useful tests in medicine. In a landmark study, measuring BNP in people arriving at the emergency department short of breath sharply improved the diagnosis of heart failure, separating it from lung disease that looks similar at the bedside (Maisel et al., NEJM, 2002). BNP and its companion fragment NT-proBNP are now routine: a number on a lab report that reads out, quite literally, how stretched the heart is.

Here is the twist that shapes the whole drug story. In heart failure, the natriuretic system is not absent, it is overwhelmed. BNP levels are high precisely because the heart is straining, yet fluid still accumulates: the system is elevated and partly resistant, shouting into a room that has stopped listening. So the obvious move, give more of the peptide, runs into two walls at once. The native peptide is cleared from the blood within minutes, and the target system is already saturated.

The obvious move was tried. Nesiritide, a recombinant form of human BNP, was given for acute decompensated heart failure on exactly this logic. In a large, careful trial it did not meaningfully improve the outcomes that matter (O’Connor et al., NEJM, 2011). Supplying a short-lived hormone against a resistant system is a losing hand. The peptide was right; delivering more of it was not the way.

If you cannot usefully add the hormone, protect the hormone you already make. Natriuretic peptides are destroyed by an enzyme called neprilysin. Block neprilysin, and the body’s own ANP and BNP linger and act for longer, amplifying the brake without infusing anything. That is the mechanism behind sacubitril, and in combination (sacubitril/valsartan) it beat the previous standard of care in a major heart-failure trial (McMurray et al., NEJM, 2014). It is the same principle this catalog keeps meeting from the other direction, in peptide half-life engineering: do not fight the body’s clearance by dosing harder, work with the salvage system instead.

The combination is not an accident, and its history is a caution. Neprilysin does not only clear natriuretic peptides; it also degrades angiotensin II, so blocking neprilysin alone would let that pressure-raising peptide build up. The fix is to pair it with an angiotensin-receptor blocker, valsartan, which is why the drug is a deliberate two-part design. An earlier attempt to combine neprilysin inhibition with an ACE inhibitor, omapatrilat, hit a dangerous rate of angioedema, because that pairing also spared bradykinin, and the program did not survive it (Kostis et al., 2004). Protecting a signal means knowing everything else the same enzyme was quietly clearing.

The family has a third member that barely touches blood pressure at all. CNP is made mostly by the lining of blood vessels and, tellingly, by the growth plate of bones. It signals through a different receptor, NPR-B, and acts locally rather than as a circulating hormone. Its standout job is not fluid balance but endochondral bone growth, the lengthening of long bones at the growth plate.

That sent the family somewhere no one would have predicted from a heart-and-kidney story. Achondroplasia, the most common form of short-limbed dwarfism, is driven by an overactive FGFR3 pathway that brakes growth-plate expansion, and CNP signaling pushes the other way. A CNP analog engineered to last, vosoritide, increased growth velocity in affected children in phase 3 trials (Savarirayan et al., Lancet, 2020). A peptide filed under the cardiovascular system turned out to be a lever on human height, which is the kind of branch a family tree grows only when the underlying biology is older and broader than the label on the drawer.

Put it together and the cardiovascular peptides read as one coherent idea with an unexpected offshoot. The heart is a gland with a thermostat: it senses its own stretch and secretes a hormone that relieves it, in standing opposition to the systems that would retain fluid. That hormone is so faithful a report of cardiac strain that its blood level became a diagnosis, and so short-lived that the way to turn it into a drug was never to add it but to defend it from the enzyme that clears it. And one member of the family slipped the cardiovascular story entirely and became a treatment for how tall a child grows.

• The heart is an endocrine organ. Stretched by volume or pressure, cardiac muscle secretes ANP (atria) and BNP (ventricles), hormones that unload the heart by shedding salt and water and relaxing vessels. A pump that also senses and signals.
• It is a thermostat and a counter-regulator. Through NPR-A and cGMP, the natriuretic peptides oppose the renin-angiotensin-aldosterone system's fluid retention. The feedback loop relieves the stretch that triggered it.
• The hormone is also the test. BNP and NT-proBNP rise with ventricular wall stress, so their blood level is how heart failure is diagnosed and tracked. A signal read straight off a vein.
• Protect the signal, do not replace it. Supplying recombinant BNP (nesiritide) underwhelmed; the peptide is cleared in minutes. Blocking neprilysin, the enzyme that destroys it, made the body's own peptides last and beat the old standard, paired with an ARB because neprilysin also clears angiotensin II.
• One cousin grows bones. CNP acts locally through NPR-B and drives growth-plate bone formation. Its analog, vosoritide, treats achondroplasia. A cardiovascular peptide that became a height drug.