Selenium Decipher: Trapping of Native Selenomethionine-Containing Peptides in Selenium-Enriched Milk and Unveiling the Deterioration after Ultrahigh-Temperature Treatment

Selenopeptide identification relies on databases to interpret the selenopeptide spectra. A common database search strategy is to set selenium as a variable modification instead of sulfur on peptides. However, this approach generally detects only a fraction of selenopeptides. An alternative approach, termed Selenium Decipher, is proposed in the present study. It involves identifying collision-induced dissociation-cleavable selenomethionine-containing peptides by iteratively matching the masses of seleno-amino acids in selenopeptide spectra. This approach uses variable-data-independent acquisition (vDIA) for peptide detection, providing a flexible and customizable window for secondary mass spectral fragmentation. The attention mechanism was used to capture global information on peptides and determine selenomethionine-containing peptide backbones. The core structure of selenium on selenomethionine-containing peptides generates a series of fragment ions, namely, C3H7Se+, C4H10NSe+, C5H7OSe+, C5H8NOSe+, and C7H11N2O2Se+, with known mass gaps during higher-energy collisional dissociation (HCD) fragmentation. De-selenium spectra are generated by removing selenium originating from selenium replacement and then reassigning the precursors to peptides. Selenium-enriched milk is obtained by feeding selenium-rich forage fed to cattle, which leads to the formation of native selenium through biotransformation. A novel antihypertensive selenopeptide Thr-Asp-Asp-Ile-SeMet-Cys-Val-Lys TDDI(Se)MCVK was identified from selenium-enriched milk. The selenopeptide (IC50 = 60.71 μM) is bound to four active residues of the angiotensin-converting enzyme (ACE) active pocket (Ala354, Tyr523, His353, and His513) and two active residues of zinc ligand (His387 and Glu411) and exerted a competitive inhibitory effect on the spatial blocking of active sites. The integration of vDIA and the iteratively matched seleno-amino acids was applied for Selenium Decipher, which provides high validity for selenomethionine-containing peptide identification.