A post-translational modifications (PTM) is the covalent modification of a protein after its synthesis, by either addition or removal of functional groups. PTMs significantly increase the complexity of the proteome by allowing each protein to exist in different forms, which can have different activity, stability or binding specificity. Because of their central role in protein regulation, PTMs are often deregulated in many diseases, especially cancer. Since all PTMs introduce a change in the proteins mass, one of the leading techniques to study PTMs is Tandem Mass Spectrometry (MS/MS), which can measure the mass of molecules with high precision and resolution. Here, we apply computational methods to MS data, in order to study PTMs from two points of view. In one project, we conducted a comprehensive analysis of Arginine (R) methylation at a global level. To achieve this, we significantly improved hmSEEKER, our in-house developed computational tool for the analysis of MS data from heavy methyl SILAC (hmSILAC) labelled samples, by implementing a machine learning model to identify methyl-peptides with high confidence. The hmSILAC-validated dataset was then combined with SILAC-based quantitative methyl-proteomics data from a set of SILAC experiments in which we profiled R methylation changes in response to different stimuli (e.g. Cisplatin treatment; inhibition of the major R methyltransferases; PRMT1 expression modulation) to generate the ProMetheus database (ProMetheusDB) of high-confidence methyl-sites. The in-depth analysis of R-methyl-sites inside ProMetheusDB reinforced the notion that protein R methylation modulates protein:RNA interactions but also provided new insights, such as the presence of several R-methyl-proteins involved in metabolism and immune response-related pathways or the fact that R methylation correlates differently with S/T-Y phosphorylation in response to different stimuli. Moreover, we employed computational methods to identify a number of protein:protein interactions that could be affect by this PTM and experimentally validated one of them. Finally, to fully exploit the potential of hmSILAC and hmSEEKER, we explored the application of our pipeline to the annotation of unconventional methyl-sites, which are largely uncharacterized. Since MS has emerged as a powerful tool not only to characterize known PTMs but also to discover new ones, in a second project, we tried to expand the annotation of histone PTMs. Histone lysine (K) acetylation and methylation are routinely used identify cancer subtypes and assess their severity; however, the unbiased nature of MS analysis has revealed the existence of several additional modifications that have yet to be systematically studied. Although theoretically possible, using MS to profile all PTMs that can occur on histones is impractical at the moment, due to limitations in the post-acquisition processing of the MS data. As a matter of fact, the most common computational tools for the identification of modified peptides are not suited to search more than 5-6 PTMs at a time, and this impairs the analysis of the combinatorial nature and cross-talk of histone PTMs. To overcome this limitation, we took advantage of a novel peptide search engine named ionbot, which can perform an “open modification search” to identify an arbitrarily large number of modifications. We reasoned that such a tool would be suitable for the detection of hyper-modified histone peptides. Upon filtering of the results, we were able to annotate not only novel histone PTMs, such as short-chain acylations, but also several amino acid substitutions that could have a biological impact on the interactions between histone and other proteins (i.e. DAXX). The analysis presented here permitted us to better understand the extent, dynamicity and biological role of protein R methylation and identified several modifications and mutation events on histone proteins. We believe further optimization and application of these methods will lead to the discovery of novel regulatory axes and the development of new cancer therapies.
COMPUTATIONAL METHODS TO STUDY KNOWN AND NOVEL PROTEIN POST-TRANSLATIONAL MODIFICATIONS BY MASS SPECTROMETRY / E. Massignani ; added supervisor: G. Pavesi ; internal advisor: M. Pelizzola ; supervisor: T. Bonaldi. - Milano : Università degli studi di Milano. Università degli Studi di Milano, 2021 Dec 13. ((33. ciclo, Anno Accademico 2021.
|Titolo:||COMPUTATIONAL METHODS TO STUDY KNOWN AND NOVEL PROTEIN POST-TRANSLATIONAL MODIFICATIONS BY MASS SPECTROMETRY|
|Tutor esterno:||BONALDI, TIZIANA|
|Supervisori e coordinatori interni:||MINUCCI, SAVERIO|
|Data di pubblicazione:||13-dic-2021|
|Parole Chiave:||Mass spectrometry; Heavy methyl SILAC; Protein methylation; Post-translational modifications; Bioinformatics; Machine Learning;|
|Settore Scientifico Disciplinare:||Settore BIO/10 - Biochimica|
|Citazione:||COMPUTATIONAL METHODS TO STUDY KNOWN AND NOVEL PROTEIN POST-TRANSLATIONAL MODIFICATIONS BY MASS SPECTROMETRY / E. Massignani ; added supervisor: G. Pavesi ; internal advisor: M. Pelizzola ; supervisor: T. Bonaldi. - Milano : Università degli studi di Milano. Università degli Studi di Milano, 2021 Dec 13. ((33. ciclo, Anno Accademico 2021.|
|Digital Object Identifier (DOI):||http://dx.doi.org/10.13130/massignani-enrico_phd2021-12-13|
|Appare nelle tipologie:||Tesi di dottorato|