To distinguish protein splicing isoform seems to be quite difficult by using mass spectrometry, because of the nearly identical sequence of them. And after fragmentation in mass spectrometry, often we cannot tell, for example, whether the detected peptide sequence is from one splicing isoform, or a combination of two alternative splicing isoforms.
But a NATURE article A draft map of the human proteome (the link below) offers detialed information of isoform and their distribution across tissues, which makes me really confused. Do we also need information other than result of MS to eventually determine splicing isoform? Frankly speaking I'm not familiar with proteomics at all but is there anything wrong of my reasoning?
Many thanks for any relevant information.
http://www.nature.com/nature/journal/v509/n7502/full/nature13302.html
The way to do this, and no doubt how it was done in the Nature paper, is to use isoform specific peptides. You need to find peptides that are unique to the isoform, or splice variant, and search for those in the MS/MS data. This obviously raises a number of issues about how well the unique peptides are detected in the MS. ie. are they a good size, do they ionise well etc. Expression levels will also affect detection limits.
Maybe you can use Ion Mobility-MS. I used for differentiate peptides and reverse peptides. It is a 3rd separation at mass spec. It is considered as a electrophoresis in gas phase.
The way to do this, and no doubt how it was done in the Nature paper, is to use isoform specific peptides. You need to find peptides that are unique to the isoform, or splice variant, and search for those in the MS/MS data. This obviously raises a number of issues about how well the unique peptides are detected in the MS. ie. are they a good size, do they ionise well etc. Expression levels will also affect detection limits.
The chance to identify / quantify protein isoforms increases with the depth of the analysis. The more peptides you identify the higher the chance that you come across isoform specific peptides. With a list of around 1000 proteins you can expect to find only the isoforms of the most abundant proteins.
Once you have a long list of identified proteins, you can extract the isoforms by finding proteins with redundant / same gene names (gene names with count > 1 in the whole list). This is because isoforms are different protein species which are nevertheless encoded by the same protein.
Gradient chromato focusing is a popular technique to separate isoforms of proteins based on the gradient in pH. We separate various basic and acidic proteins and its isoforms on ion exchange columns with a pH gradient around the iso-electric point of protein.
You may find interesting these two references about protein inference:
Nesvizhskii A, Aebersold R: Interpretation of shotgun proteomic data. Mol Cell Proteomics 2005, 4(10):1419-1440.
Prieto G, Aloria K, Osinalde N, Fullaondo A, Arizmendi JM, Matthiesen R. PAnalyzer: a software tool for protein inference in shotgun proteomics. BMC Bioinformatics. 2012 Nov 5;13:288. doi: 10.1186/1471-2105-13-288.
Just to add in the previous replies, you can search your data using a reference database containing canonical and isofroms sequences. these databases are available through Uniprot (uniprot.org)
Just to add a few "field" terms, what Peter is talking about is "unique peptides". These are peptides which a sequence characteristic to only one protein (not family, single protein).
It is normally quite hard to find a unique peptide for your protein. A good example is PSA (prostate specific antigen)
Here a good paper about it
http://www.ncbi.nlm.nih.gov/pubmed/20102753
- Badges
- Science method
- Similar topics
- Chemistry
- General Biochemistry
- Biochemical Techniques
- Analytical Chemistry