to set the record straight, you do not need saturation phosphorylation to
detect phosphorylation sites by mass spectrometry, you just need an
appropriately sensitive approach. That said however, since you have the
luxury of having radiolabelled protein, there is a particular analytical
route that we take as a routine in this situation.
The following description may seem a little long winded and "old fashioned"
but I believe it to be the most thorough. The literature is replete with
"I've found another phosphorylation site in a previously well characterized
phosphoprotein" stories, this being the sad result of phosphopeptides being
very easy to miss.
Before I digested my radiolabelled protein I would do two things, first try
to get a mass of the intact protein at sufficient resolution to see
phosphorylation, we use nano-spray ESI MS for this purpose. This will give
you an indication of how many sites you are dealing with. Secondly, I would
run the protein on a 2D gel, again the number of spots will give you an
idea of how many species you are dealing with. Hopefully, you will see the
expected mass and a +80 partner on the mass spectrum and a single spot on
the 2D gel (I use the term "hopefully" a lot when discussing
phosphorylation site analysis)
Then get yourself access to a "Tony Hunter" style high voltage
electrophoresis apparatus and learn how to run traditional two dimensional
phosphopeptide maps. This is the only proven way to completely account for
a protein's phosphorylation. The catchword here is "completely", those folk
who rely on recovery of digestion products from HPLC live dangerously, I
can show you long lists of phosphorylation sites in peptides that never
would stick to a reversed-phase column and others that would never elute. A
2D map on the crude digest shows you everything you have got in the mixture
and it is important to account for every spot. Hopefully you will see a
single spot. Even if you see more than one spot, it doesn't mean that you
have more than one site, phosphorylations inevitably affect the efficiency
of protease digestion and multiple fragments of a single site are very
common.
>From your analytical map, you can scrape the spot(s) and do a phospho-amino
acid analysis, again using the high voltage electrophoresis/TLC approach.
>From the same map you can scrape the spot(s) and do phosphate release
sequencing on your Edman sequencer. I would not recommend traditional Edman
sequencing but solid phase phosphate release sequencing. You will have to
attach each peptide to a membrane like the Perseptive Sequelon AA product
and use a special extraction solvent like 90% Methanol/0.05% phosphoric
acid since the inorganic (32P)phosphate liberated during the sequencing
reaction is poorly soluble in the regular extraction solvents. It is
possible to do this style of sequencing on a few hundred counts which,
depending on your stoichiometry, can be a few femtomoles of peptide.
Therefore, there is no use directing any of the sequencer stream to the LC,
just collect it in a fraction collector and count the fractions.
Alternatively, you might try separating/purifying the phosphopeptides by
HPLC, at least with the 2D map of the crude digest in hand you will know
how many peaks of radioactivity to expect (and see how many you loose!) ,
phospho-AAA and phosphate release sequencing can be done direct from the LC
fractions. Expect poor recoveries from your HPLC and you won't be
disappointed.
You will then be faced with knowing how many peptides you have, which type
of amino-acids are involved and how many residues each phospho-amino acid
is from the N-terminus of the peptide. With this data, one can usually
deduce the phosphorylation site(s).
Before the advent of mass spectrometry, this, combined with showing the
synthetic phosphopeptide migrated in the same place on a 2D map, was
regarded as sufficient data for site identification.
If you haven't used all your digest up by this stage, I would then
partially separate the digest on a ZipTip (say five or six different
acetonitrile concentrations) and use MALDI mass spectrometry to identify
the predicted phosphopeptide by mass. Don't forget to look in the fraction
that doesn't stick to the ZipTip. You can treat some of each fraction with
phosphatase and look for the loss of 80 in the MALDI spectrum.
This will be sufficient to nail your phosphorylation site.
There are other very elegant ways of doing phosphopeptide analysis but they
generally rely on desalting (by tip or full chromatography). Unless you do
2D mapping on the crude, untreated digest, there is a serious chance you
will miss something, especially if your kinase consensus site involves
charged and/or hydrophilic amino-acids.
There is a very nice description and proposed solution to the sample
handling issue in a recent paper out of Matthias Mann's lab which I would
recommend you read (Neubauer G, Mann M; Anal Chem 1999 Jan 1;71(1):235-42
"Mapping of phosphorylation sites of gel-isolated proteins by
nanoelectrospray tandem mass spectrometry: potentials and limitations.), I
especially like the use of the word "limitations" in the title ;-)
Hope this is useful.
Regards....Ken
>Dear colleagues,
>
>I would like to determine phosphorylation sites on a protein that has been
>phosphorylated in vitro with P32 (ATP) using a kinase. I was thinking of
>digesting the radioactively labeled protein with proteinase followed by
>HPLC purification of the labeled peptide (i.e. phosphorylated) and using an
>automated protein sequencer (e.g. Procise) determine which amino acid is
>phosphorylated. How reliably can I determine the phosphorylation site
>using this strategy and are there better methods? Using Mass Spec is
>another strategy but it seems that one needs almost saturation
>phosphorylation at the amino acid to detect it.
>I would appreciate any comments from colleagues who have experience in this
>matter. Jay.
>
>
>
>Address: Jay H. Chung MD,PhD
> NIH
> Bldg. 10-7D13
> 10 Center Dr.
> Bethesda, MD 20892-1654
>
>Tel: (301)496-3075
>FAX: (301)496-9985
********************************
Ken I. Mitchelhill
The John Holt Protein Structure Laboratory
St. Vincent's Institute of Medical Research
41 Victoria Parade
Fitzroy 3065 Victoria
AUSTRALIA
Telephone: 61-3-9288 2480
Facsimile: 61-3-9416 2676
Email: k.mitchelhill@medicine.unimelb.edu.au
Laboratory: http://www.medstv.unimelb.edu.au/WWWDOCS/SVIMRdocs/JHPSL.html
ABRF: http://www.abrf.org
***********************************