Most entries in this library describe something sold in a vial. This one does not. BNP is a hormone the human heart makes, and it appears here for the same reason C-peptide does: the word peptide is in its name, it turns up on ordinary blood test reports, and a reader who arrives that way deserves to find out in one sentence that it is unrelated to the compounds most of these pages describe.
It also happens to be the clearest illustration in the library of something worth understanding: one molecule can produce two different laboratory numbers, and those numbers can move in opposite directions.
The arithmetic of one hormone becoming two tests
The gene NPPB encodes a 134-residue preprohormone (UniProt P16860). The first 26 residues are a signal peptide, removed as the molecule is processed. What remains is proBNP, 108 residues long.
ProBNP is then cut once. The cut produces:
- NT-proBNP — residues 27 to 102 of the precursor, 76 residues. The N-terminal leftover. No established hormonal activity of its own.
- BNP-32 — residues 103 to 134, 32 residues. The active hormone.
The arithmetic closes exactly: 76 + 32 = 108. Anyone can check it against the UniProt record; that is the whole relationship between the two tests.
The heart's ventricular muscle makes this in response to stretch of the chamber wall. Both fragments reach the bloodstream, and both can be measured — which is why a single hormone ended up with two assays rather than one.
"BNP" is a family, not a molecule
The UniProt record for the same protein lists nineteen distinct named BNP-derived peptide forms circulating in human blood. Alongside full-length BNP-32 there are BNP(1-30), BNP(1-29), BNP(1-28), BNP(2-31), BNP(3-32), BNP(3-30), BNP(3-29), BNP(4-31), BNP(4-30), BNP(4-29), BNP(4-27), BNP(5-32), BNP(5-31), BNP(5-29) and BNP-29, among others — the same chain shortened from one end or the other.
This is not a curiosity. What a given assay reports depends on which of those forms its antibodies bind, which is a large part of why results produced by different manufacturers' assays are not interchangeable. A number on a report is a measurement made by a particular instrument, not a count of molecules.
The finding: a drug label that states the two tests disagree
The most useful documented example of why the distinction matters is printed on a prescribing information sheet.
Entresto combines sacubitril, a neprilysin inhibitor, with valsartan. Neprilysin is the enzyme that degrades natriuretic peptides; inhibiting it is the point of the drug. BNP is one of that enzyme's substrates. NT-proBNP is not.
The label puts the consequence in a single sentence. In the PARADIGM-HF trial, it records, the drug decreased plasma NT-proBNP (not a neprilysin substrate) and increased plasma BNP (a neprilysin substrate) compared with enalapril.
One drug. One hormone. Two tests that move in opposite directions — and the manufacturer says so in its own clinical pharmacology section. The same label reports NT-proBNP falling by 17% in PARAMOUNT and by 24% at week 16 in PARAGON-HF.
The scale of it
Two figures put this entry in proportion against the rest of the library. On 2026-09-19, PubMed held 32,314 records for B-type natriuretic peptide and 15,865 for NT-proBNP. The FDA's 510(k) database returned 14 clearances for devices named for natriuretic peptide and 12 for NT-proBNP.
For comparison, the largest literature attached to any compound sold as a research peptide in this library is BPC-157 at 230 PubMed records, none of them a clinical trial. The contrast is the point: this is what a molecule with an established clinical role looks like in the public record, and it does not resemble the compounds most of these entries describe.
Where it stands
Nothing here is sold and nothing is taken. There is no research-chemical market for BNP or NT-proBNP, no vial, and no dose — only assays, and a large clinical literature about how to read them.
This entry deliberately carries no reference ranges and no interpretation of any result. Those numbers depend on the assay used and on the person, and they are read alongside symptoms, examination and imaging by the clinician who ordered the test. If you are holding a result, that clinician is the person to ask.
Related entries: C-peptide, the other clinical lab marker in this library that people reach through the word peptide; what a peptide actually is; and the peptide bond, if the 108-into-76-plus-32 arithmetic above is the part you want to check yourself.