Protein Protocols Protein Protocols

614 Bonenfant, Mini, and Jenö
included) by the sum of all intensities of the phosphorylated and its non-phosphorylated
forms (see Note 5). For example, comparison of the integrated peak areas of the
various charge forms for the two phosphopeptides K 2 and K 26 of the Drosophila lamin
indicates an approximate extent of phosphorylation of 23–29%, and 13–19%,
respectively (Fig. 2A,C). This assumes that the ionization efficiencies of the phosphoand
the nonphosphopeptide are equal. Our observations indicate that the relative areas
between phosphorylated and unphosphorylated peptide varies with the charge state:
the higher charge forms tend to overrepresent the phosphoform, whereas the lower
charge forms tend to underrepresent the phosphoform. Accordingly, the stoichiometry
of phosphorylation can be determined only as an average from a number of charge
states with a certain variation. Nevertheless, large changes in phosphorylation that
occur at a given site can still be followed that peak areas from spectra of the
corresponding phospho-nonphospho- forms are compared.
For example, the extent of phosphorylation of the K 2 and K 26 fragment is changed
considerably when a soluble isoform of the protein (the so-called mitotic form) is isolated
from young embryos. The extent of phosphorylation of the K 2 fragment increases
from 13–19% to 66–71%, whereas phosphorylation of the K 26 fragment decreases to
levels too low to be quantified with certainty (Fig. 2B,D).
These examples clearly demonstrate that quantitation of the extent of phosphorylation
of a given phosphopeptide is possible within certain limits. However, it must be
stressed that this becomes more difficult when multiple phosphorylation is clustered
within a short stretch of a protein. In our studies to determine the insulin-induced phosphorylation
sites of ribosomal protein S6 (DS6A) in the fruit fly Drosophila
melanogaster, the sites were found to be located in a short stretch at the carboxy-(C)terminal
end of the protein (18). Two-dimensional polyacrylamide gel electrophoresis
of ribosomal proteins indicated that up to five phosphates are incorporated into DS6A
upon insulin/cycloheximide stimulation of Kc167 cells. Although LC/MS analysis of
an endoproteinase LysC digest of DS6A demonstrated the presence of singly, doubly,
and triply phosphorylated forms of the C-terminal fragment, the quadruply, and fivefold
phosphorylated forms could not be detected. Because polyacrylamide gel electrophoresis
clearly demonstrated that the quadruply, or fivefold phosphorylated form of
the protein was predominant, it was concluded that ionization efficiency decreases with
increasing degree of phosphorylation. Therefore, an endoproteinase LysC digest of
DS6A from insulin- and cycloheximide-stimulated Kc167 cells was treated with
Ba(OH) 2 to induce β-elimination of the phosphate groups (10,19). Subsequent capillary
LC/MS analysis revealed strong signals with mass-to-charge ratios of 692.1, 698.4,
and 704.5 Da, corresponding to the triply charged ions of the RRRSASIRESKSSVSSDKK
(K 35–37) peptide containing three, four, and five dehydroalanines (Fig. 3B). The intensity
of the corresponding ions closely resembled the staining intensity of the individual
phosphoderivatives of DS6A, with the tetradehydroalanine derivative being the most
abundant species (Fig. 3B). Furthermore, when DS6A was isolated from cells pretreated
with okadaic acid, a type 2A phosphatase inhibitor, prior to stimulation, substantially
higher amounts of the penta-dehydroalanine derivative could be observed in the
spectrum of the K 35–37 peptide (Fig. 3C).