Incorrectly annotated keratin derived peptide sequences lead to misleading MS/MS data interpretation

In the course of the nanoLC-nanoESI-MS/MS analysis of Chelidonium majus proteins we detected an extremely abundant 12 aa peptide (sequence: TNAENEFVTIKK). The Mascot search against NCBInr database with Viridiplantae taxonomic restriction revealed its complete identity to Unknown protein 18 from Pseudotsuga menziesii (P85925). The same sequence has also been submitted as Unknown protein 1 (P86104) from Vitis rotundifolia and as a part of the oxygen-evolving enhancer protein 1 (OEE1) from Pinus strobus (P84718). After careful inspection the record P84718 turned out to comprise a set of peptides of unconfirmed origin rather than complete protein sequence. In this paper we present extensive data indicating that the peptide in question may originate from type II cytoskeletal keratin - a common contaminant in protein samples. We found empirical evidence that it can be detected in several types of keratin-contaminated samples and its sequence is identical to one of the proteotypic peptides commonly observed for keratins. Nevertheless, the peptide has been annotated as plant protein and thus leads to data misinterpretation. We advise extreme caution when dealing with sequences of Unknown protein 18 (P. menziesii), Unknown protein 1 (V. rotundifolia) and OEE1 (P. strobus) since they are at best poorly annotated, if not artifactual. Biological significance To our knowledge, this is the first report indicating that the peptide TNAENEFVTIKK, which is identical to Unknown protein 18 from P. menziesii (P85925), Unknown protein 1 (P86104) from V. rotundifolia and a part of the oxygen-evolving enhancer protein 1 (OEE1) from P. strobus (P84718), may originate from type II cytoskeletal keratin - a common contaminant in protein samples. We found empirical evidence that it can be detected in several types of keratin-contaminated samples and its sequence is identical to one of the proteotypic peptides commonly observed for keratins. Nevertheless, the peptide has been annotated as plant protein and thus, without proper quality assessment, leads to biological misinterpretation of proteomic data.