Mollusc shells represent excellent systems for the preservation and retrieval of genuine biomolecules from archaeological or palaeontological samples. As a consequence, the post-mortem breakdown of intracrystalline mollusc shell proteins has been extensively investigated, particularly with regard to its potential use as a "molecular clock" for geochronological applications. But despite seventy years of ancient protein research, the fundamental aspects of diagenesis-induced changes to protein structures and sequences remain elusive. In this study we investigate the degradation of intracrystalline proteins by performing artificial degradation experiments on the shell of the thorny oyster, Spondylus gaederopus, which is particularly important for archaeological research. We used immunochemistry and tandem mass tag (TMT) quantitative proteomics, a novel analytical approach that allowed us to simultaneously track patterns of structural loss and of peptide bond hydrolysis. Powdered and bleached shell samples were heated in water at four different temperatures (80, 95, 110, 140 °C) for different time durations. The structural loss of carbohydrate and protein groups was investigated by immunochemical techniques (ELLA and ELISA) and peptide bond hydrolysis was studied by tracking the changes in protein/peptide relative abundances over time using TMT quantitative proteomics. We find that heating does not induce instant organic matrix decay, but first facilitates the uncoiling of cross-linked structures, thus improving matrix detection. We calculated apparent activation energies of structural loss: Ea (carbohydrate groups) = 104.7 kJ/mol, Ea (protein epitopes) = 104.4 kJ/mol, which suggests that secondary matrix structure degradation may proceed simultaneously with protein hydrolysis. While prolonged heating at 110 °C (10 days) results in complete loss of the structural signal, surviving peptide sequences were still observed. Eight hydrolysis-prone peptide bonds were identified in the top scoring shell sequence, the uncharacterised protein LOC117318053 from Pecten maximus. Interestingly, these were not the expected "weak" bonds based on published theoretical stabilities calculated for peptides in solution. This further confirms that intracrystalline protein degradation patterns are complex and that the overall microchemical environment plays an active role in protein stability. Our TMT approach represents a major stepping stone towards developing a model for studying protein diagenesis in biomineralised systems.