Dear Authors
The review is, of course, highlighting a particular research field which positively believes that there is evidence of more labile biomolecules in Mesozoic fossils, which is entirely based on antibody immunisation experiments and limited perspectives from other approaches to protein detection.
Antibody immunisation was a widely used methodology in archaeological studies in the 1980s and 1990s, but subsequently abandoned due to the inherent issue with false positives and the introduction of more sensitive methods, such as newer generations of mass spectrometers.
iIt is mentioned that even DNA is found in fossil bones, citing an older study. However, it has been shown this spring (Saitta et al. 2019, eLife) that bones are inherently open systems that welcome unique bacterial consortia, perhaps to make use of the inherently biolimiting nutrient that phosphate is, but that dinosaur bones excavated from subsurface environments are more than 80X enriched in DNA (derived from bacteria) relative to surrounding sediments.
Hence, any claims of proteins or DNA need to evaluate the degree of bacterial contamination. Currently, no chemical evidence for any peptide sequence that is older than 3.8 million years old has been positively confirmed. Antibody immunisation necessitates long and basically complete proteins with an intact tertiary structure to positively work. So, if antibodies readily give true positive results, then various suites of peptide sequences should flourish--complete ones and a larger tail of more degraded ones based.
The authors keep stressing the power of the repeatability of their methods in their labs, but true substantiation of a hypothesis is corroboration with a different method with different strengths and weaknesses. This would be the sequencing of peptides with mass spectroscopy in the case of proteins.
Proteins and DNA are some of the least stable organic molecules, while others such as cross-linked and complex molecules such as lignin, melanin and sporopollenin and certain lipids, such as cholesterol and hopanoids are extremely stable. Mass spectroscopic analyses have shown that these molecules preserve routinely in exceptional fossils, while proteins have not found themselves to be detected positively.
it could well be that proteins may preserve as transformed molecules, through cross-linking with oxidised lipids or sugars, but these molecules are not preserving the functionality of the protein and hence would not give rise to positive antibody staining, but their complex structure and potential to have several reactive sites may be a suitable locus for false-positive reactions with antibodies.
Proteinaceous tissues are rarely observed preserved as organic imprints, if ever. Otherwise, why would it be possible to observe colour patterns in fossil feathers? We would preserve all sorts of tendons and in particular notochords in jawless vertebrates. However, they practically shimmer in their absence.
I would invite the authors to test their hypotheses across other platforms, specifically demonstrating an absence of recent contamination and the presence of diagnostic peptide sequences. The recent corroboration of likely proteinaceous material with methods such as TOF SIMS in bones would similarly suffer from the likelihood of bacterial contamination and cannot be cited anymore as support for fossil proteins by other labs until the same bones have been subjected to tests for bacterial contamination and amino acid racemisation tests to evaluate if protein moieties present have an ancient signature.
So, to rephrase your final statement "Why do evidence of pre-Neogene fossil proteins not emerge from conventional and widely approved paleoproteomic protocols?"
