Electrospray mass spectrometry of peptides and lens proteins

Mammalian lenses, including human and bovine, contain three major classes of soluble lens proteins α-, β-, and γ-crystallins. Each respective class of crystallins is comprised several closely-related subunits ranging in molecular weight from 20-30 kDa. A variety of gel filtration chromatography, fast protein liquid chromatography, and ion exchange chromatography methods have been used to isolate the individual protein subunits from bovine lenses which then have been examined by electrospray mass spectrometry (ESI-MS). This study shows the considerable advantages (and some short-comings) ESI-MS has for the examination of the bovine lens crystallins. ESI-MS of α-crystallins showed the two major subunits (and their phosphorylated forms). The presence of a number of C-terminal degradation products (i.e. αA1-172, αAp1-154, αA1-151 αA1-101) appear to be age-dependent. Quantitation studies on the α-crystallins show that the phosphorylation of αβ crystallin does not significantly affect ionisation efficiency of the protein as long as the skimmer potential is carefully controlled. Further, the response of αβ and αβp are equivalent over a limited concentration range. However, the ESI response of alpha;A and αB in mixtures differed significantly so these proteins cannot be quantitatively compared by ESI. ESI-MS and MS/MS studies on glycated acrystallin and glycated C-terminal peptide extensions of αA and αB subunits show that galactose reacts specifically at the lysine residues within the sequence. Further, the degree glycation at these residues was shown to be greater than expected from NMR and other data. The β subunit separation was considerably more complex because of the high degree of homology between these proteins. The largest and most hydrophobic β subunit, i.e. βB1, exhibited unusual behaviour in that it could only be detected in the βH aggregate and not following FPLC isolation. An age-dependent cleavage of the C-terminal serine residue of βB2 was characterised. The published sequence for βB3 was shown to be incorrect. Of the remaining β subunits, only the mass determined by ESI-MS for βA2 corresponded to that calculated from the published sequence. ESI-MS of the γ-crystallins showed the observed masses of γS, γll, and γlllb corresponded to those calculated from their respective sequences. The sequences of γllla and γIVb were shown to be in error. The sequence errors in γlVb were examined in detail by a combination of enzymatic and chemical digestion, gas-phase sequencing, ESI-MS, and MS/MS and up to nine changes have been identified. As a prelude to the studies on lens crystallins, the effect of ESI conditions on lowenergy MS/MS spectra of peptides (Mr 400-1200) was also examined in the initial stage of this work. It was found that for singly-charged peptide ions, the source skimmer potential can strongly influence the extent of fragmentation observed in MS/MS spectra, especially at low collision energies. Low-energy MS/MS spectra of peptides are also shown to be equally as sensitive to peptide structure and the type of adduct studied as MS IMS obtained following older ionisation methods such as fast atom bombardment. Finally, the major proteins contained within the bovine lens also display a considerable degree of sequence homology to the major human lens proteins. Therefore, the techniques developed for the separation and/or characterisation of bovine lens crystallins should be able to be applied with little or no modification to the study of human lens crystallins