[6] The amino acid sequence for bovine BPTI is RPDFC LEPPY TGPCK ARIIR YFYNA KAGLC QTFVY GGCRA KRNNF KSAED CMRTC GGA.

[8] The long, basic lysine 15 side chain on the exposed loop (at top left in the image) binds very tightly in the specificity pocket at the active site of trypsin and inhibits its enzymatic action.

Its physiological functions include the protective inhibition of the major digestive enzyme trypsin when small amounts are produced, by cleavage of the trypsinogen precursor during storage in the pancreas.

[13] In September 2006, Bayer A.G. was faulted by the FDA for not revealing during testimony the existence of a commissioned retrospective study of 67,000 patients, 30,000 of whom received aprotinin and the rest other anti-fibrinolytics.

[14] In a Public Health Advisory Update dated October 3, 2006, the FDA recommended that "physicians consider limiting Trasylol use to those situations in which the clinical benefit of reduced blood loss is necessary to medical management and outweighs the potential risks" and carefully monitor patients.

The producer, Bayer, reported to the FDA that additional observation studies showed that it may increase the chance for death, serious kidney damage, congestive heart failure and strokes.

FDA warned clinicians to consider limiting use to those situations where the clinical benefit of reduced blood loss is essential to medical management and outweighs the potential risks.

[17] On November 5, 2007, Bayer announced that it was withdrawing Aprotinin because of a Canadian study that showed it increased the risk of death when used to prevent bleeding during heart surgery.

[5] Small amounts of aprotinin can be added to tubes of drawn blood to enable laboratory measurement of certain rapidly degraded proteins such as glucagon.

The conjugate retains its antiproteolytic and carbohydrate-binding properties[21] and has been used as a fluorescent histochemical reagent for staining glycoconjugates (mucosubstances) that are rich in uronic or sialic acids.

It was one of the earliest protein crystal structures solved, in 1970 in the laboratory of Robert Huber,[28] and it's substrate-like interaction mode deciphered in the context of the bovine trypsin complex in 1974.

[30][31] Because it is a small, stable protein whose structure had been determined at high resolution by 1975,[32] it was the first macromolecule of scientific interest to be simulated using molecular dynamics computation, in 1977 by J. Andrew McCammon and Bruce Gelin, in the Karplus group at Harvard.

[33] That study confirmed the then-surprising fact found in the NMR work[34] that even well-packed aromatic sidechains in the interior of a stable protein can flip over rather rapidly (microsecond to millisecond time scale).