Abstract
It is believed that protein analysis by LC-MS will be an important diagnostic tool for different kind of diseases in the future. Proteins can not always be directly analyzed by LC-MS, several sample preparation steps can be necessary. One of these steps is protein deglycosylation. In order to study different deglycosylation techniques, bovine fetuin was used as a model protein in this thesis. Additionally other sample preparation steps have been performed in conjunction with deglycosylation. This includes tryptic digestion, denaturation, reduction, alkylation and sample cleanup. For all results obtained by LC-MS, identification of peptide sequences was attempted.
Bovine fetuin was first digested by the proteolytic enzyme trypsin, in aqueous conditions, before analysing the sample with LC-MS. Peptide products were identified by comparing data from LC-MS with in-silico digest results produced by ProteinProspector. Optimizing the sample preparation with denaturation, reduction and alkylation was then attempted before new analysis and peptide identification was performed.
Two different deglycosylation techniques were tested on the glycoprotein fetuin; Chemical and enzymatic.
Chemical deglycosylation was perfomed by using TFMS which has the ability to deglycosylate all types of glycan bonds. Results obtained here were however, inconclusive. For sample cleanup after chemical deglycosylation, a new in-house SPE technique was developed. It is suspected that because this is a new sample cleanup technique, this step could have contributed to the inconclusive results. Initial testing of this technique was also perfomed, giving a starting point for further optimizing of the technique.
Enzymatic deglycosylation was carried out with the glycopeptidase PNGase F. This enzyme deglycosylates N-glycans. Experiments with enzymatic deglycosylation of fetuin showed new peptide products compared with trypsin digest only. This indicates that performing deglycosylation gives better possibilities for protein identification.