The formation and radical-directed dissociation (RDD) of multiple hydrogen abstracted peptide cations [M+H-mH]^•+ have been examined using MALDI In-source decay (ISD) with four different dinitro-substituted matrices, i.e., 3,5-dinitrosalicylic acid (3,5-DNSA), 3,4-dinitrobenzoic acid (3,4-DNBA), 1,5-dinitronaphthalene (1,5-DNN) and 2,4-dinitro-1-naphthol (2,4-DNNL). The MALDI-ISD of peptides using 3,5-DNSA and 3,4-DNBA as matrices resulted in multiple hydrogen abstraction from protonated peptide [M+H]⁺ and production of fragment [a]⁺ ions, i.e., [M+H-mH]^•+ and [a-mH]^•+ (m=1–6). Fragmentation pattern of the intensity of [a_n]⁺ ions with 4-nitro-1-naphthol (4,1-NNL) depended on amino acid residue, while that of the [a_n]⁺ ions with 3,5-DNSA did not.

The diynene natural products are effective drugs for cancer or leukemia owing to their strong DNA cleavage activity. Bioreduction triggers Masamune–Bergman cyclization of their diynene portion producing a 1,4-dehydrobenzene diradical species, and it abstracts two hydrogens from deoxyribose. MALDI mass spectrum of Esperamicin-A₁, one of the first isolated diynene antibiotics, with thiol containing matrix (2,5-diamino-1,4-benzenedithiol) gave peaks of the cleavage of the trisulfide with thiols (m/z 1248) and cyclized analyte (m/z 1250, Esperamicin-Z).However, three matrices including 2,3-diaminonaphthol gave no peaks of Esperamicin-Z, while it gave peaks of m/z 1248. The results suggested that thiol group of matrix has influence on cyclization mechanism of the diynene portion.

Chemically modified trypsin by reductive methylation is a standard reagent for identifying proteins by mass spectrometry. The reductive methylation of trypsin is assumed to protect the enzyme against autolysis. However, it has been reported that peptides with mono- or di-methylated Lys, or Asn residues at their C-termini have been observed in autolysis products of chemically modified trypsin. In most cases, the chemical modification of trypsin is not taken into consideration in database searches for protein identification. Spectral data derived from autolysis peptides of chemically modified trypsin might be incorrectly assigned to peptides derived from other proteins, thus creating false-positive identifications. In this paper, we propose a simple strategy for searching databases by using spectral libraries to identify autolysis peptides from heavily chemically modified trypsin, which cannot be identified by a normal search strategy.