[proteomics] Proteomic Analsysis of Glycoproteins

  • From: Mavi Gozler <mavigozler@xxxxxxxxx>
  • To: proteomics@xxxxxxxxxxxxx
  • Date: Thu, 9 Nov 2006 22:32:21 -0800 (PST)

Glycoproteins can present special challenges to those trying to identify them 
by examining peptides of them using MALDI-MS of gel spots or LC-ESI-MS/MS of 
purified proteins.

In a paper by Valmu et al, they were working with recombinant human transferrin 
(TRF) purified from mammalian cell hosts in which it was expressed.  TRF 
circulates in serum and its function is to transport the essential element iron 
(Fe).  It is normally N-glycosylated at two sites; N-glycosylation occurs at 
the Asn-X-Ser/Thr (where X is any amino acid but Pro), and they had a 
recombinant in which Asn-413 and -611 were changed to Asp (TRF-NG or 
nonglycosylated form).  Human TRF exists as a bilobed globular protein, with a 
C- and N-side lobes connected by a linking sequence.

The group treated TRF and TRF-NG with chymotrypsin, which like trypsin, is 
often used to generate peptides for doing peptide mass fingerprinting with the 
MALDI to identify proteins.  Chymotrypsin prefers to cut on the carboxyl side 
of aromatic amino acid residues (Phe, Tyr, and Trp) except when the C-side 
amino acid is Pro;  trypsin, also a member of the serine protease family, works 
in the same way, but prefers basic amino acids (Lys and Arg).  They found that 
the TRF-NG generated fewer peptides than TRF, and that is because the glycans 
attached to TRF protected the protein from digestion.  Some target sites for 
chymotrypsin are located near the glycosylation sites of TRF.  They observed 
changes in the HPLC chromatogram (using OD214) as well as MALDI.

In using the MALDI, a higher energy setting on the laser was necessary to 
generate a spectrum for the TRF but not TRF-NG peptides, which they attribute 
to the difficulty of generating a good yield of peptides.

They found some lower intensity peaks in the TRF-NG spectrum and tried to 
confirm their identity (sequence) using LC-MS/MS.  They particularly looked at 
the chymotryptic peptides that contained the Asn->Asp substitutions, peptides 
413-427 (containing Asn/Asp-413, which is C-side of a chymotrypsin target), and 
peptides 609-622, containing Asn/Asp-611.  These peptides were seen in MALDI 
with about a 3-6 signal-to-nose ratio (and they were using purified proteins 
NOT from a gel spot cut!!!). They had a difficult time getting fragmentation 
spectra on the  609-622 peptide; the fragmentation spectrum is used to get the 
sequence.  But they had much better luck getting fragmentation spectra on the 
413-472 peptide.  They say it was because the peptide contained a Lys, adding 
more positive charge to the peptide during the electrospray, increasing its 
ionization efficiency...the additional charges also help fragment the peptides 
more easily.  The mass of the 413-472 peptide is seen at
 m/z 1690.472 in the MALDI (the [M+H]+ ion), but in ESI, it is seen at m/z 
845.46 (the [M+2H]2+ ion).  When the collision energy is turned up (to 32 V), 
you generate the fragmentation.

So keep in mind that glycans (oligosaccharide modifications) on proteins can 
affect their digestion and the peptide masses that you might expect!  Many of 
you might see this in the mass spectral data on your projects, and you need to 
think about possibly different masses.

L Valmu, N Kalkkinen, A Husa, PD Rye (2005) Biochemistry 44: 16007-16013.  
Differential Susceptibility of Transferrin Glycoforms to Chymotrypsin:  A 
Proteomics Approach To The Detection of Carbohydrate-Deficient Transferrin

Want to start your own business? Learn how on Yahoo! Small Business.

Other related posts:

  • » [proteomics] Proteomic Analsysis of Glycoproteins