DISEÑO CONVERSOR REDUCTOR

The methods developed in Chapter 3 were here applied to the analysis o f three proteins in which occurrence o f phosphorylation was either known or strongly indicated to play an essential role in their functional behaviour. Peptides with m/z

values corresponding to both known and novel phosphorylation sites were detected by mass spectrometry from IMAC-separated fractions, and in some cases could be subjected to further tandem mass spectrometric analysis to determine the precise location of the phosphate moiety upon the peptide chain. In several cases, the phosphopeptides which were observed following IMAC separation were completely absent in spectra prior to IMAC. The ease with which effective and definitive phosphorylation site mapping could be achieved using IMAC and mass spectrometry did, however, vary considerably, depending upon the substrate and the kinase which was believed to have carried out the phosphorylating activity. There are a number o f potential causes for this variation, which I shall address here.

In the analysis o f GSK-Sp, a novel site of modification was identified conclusively using IMAC in combination with mass spectrometric methods. Thrggg was present in untreated purified recombinant GSK-3p. The expected site of modification of GSK- 3p following in vitro kinase assay with the tyrosine kinase ZAKl was not observed following IMAC separation o f the protein. This experiment illustrated that even in cases where the potential phosphorylation sites upon a protein are known, the current IMAC protocol cannot be guaranteed to identify them. Since the IMAC sample preparation method has not been observed to specifically select for a particular

Chapter 4: Analysis of Signal Transduction Proteins

phosphorylation site of interest, the observation of novel modifications is not unexpected.

In the case of 4E-BP1, multiple peptides corresponding to phosphopeptides containing known sites o f phosphorylation were detected by mass spectrometry, IMAC allowed removal of some of the non-phosphorylated peptides to increase the relative signal-to-noise for the phosphorylated peptides. In both the GSK-3P and 4E- BPl case o f phosphorylation site identification using IMAC prior to mass spectrometry, a number o f non-phosphorylated peptides containing multiple (two or more) acidic residues were detected in the eluted fraction following IMAC. Such non-specific binding via carboxylate anions to the immobilised metal ion has been observed in a number o f previous studies (Ficarro, 2002b; Muszynska, 1992; Raska, 2002; Zhou, 2000), as well as within the method development work presented in this thesis (Chapter 3). In all o f the experiments upon 4E-BP1, prominent acidic peptide ion signals were observed in fractions which had been separated using IMAC (see Figure 4.2.4), representing a significant obstacle to effective phosphopeptide separation.

In the case o f phosphorylation site analysis of RhoE, very poor data were obtained following IMAC separation of digested RhoE. Although the tentatively identified phosphopeptides were in regions which prior studies had indicated to be important for ROCK-mediated phosphorylation of RhoE, these data are insufficiently strong to be able to draw any conclusions regarding the phosphorylation status of these peptides. Since parallel radiolabelling experiments had strongly suggested that phosphorylation of RhoE by ROCK occurs under the sample preparation conditions used, there are three likely explanations regarding the lack o f phosphorylated material in the IMAC- separated fractions.

Firstly, the stoichiometry o f modification of the protein by ROCK phosphorylation has not been conclusively determined. Simple detection of phosphorylation by monitoring electrophoretic shifting o f silver-stained ID bands is not an ideal method for the determination o f the stoichiometry of modification, as not all phosphorylation events induce significant band shifting, and silver staining methods are knovm to give

non-linear detection o f phosphorylated proteins (Patton, 2000). A better method for the determination of stoichiometry might be to use 2D gel electrophoresis, which should be more capable o f resolving the phosphorylated forms o f proteins, with a linear staining method such as the fluorescent dye SYPRO Ruby™, to be able to determine the relative intensity of the differently phosphorylated protein spots.

A second problem which may give rise to the lack o f phosphorylated signal in eluted fractions is a lack o f binding of phosphorylated peptide to the IMAC resin. Discussions with colleagues at conferences who have tried IMAC for phosphorylation site identification has indicated that there is considerable variation in the binding of phosphorylated peptides to IMAC resins, and that a number of proteins show poor retention to immobilised metal ions. A significant observation from these experiments is that phosphorylation site identification was relatively simplistic in the case o f 4E-BP1, which has a pi o f 5.9. In the case o f GSK-3/3, although the protein has a pi of 8.3, the phosphorylation site identified occurred in an acidic region of the protein, the acidic non-phosphorylated counterpart to the identified phosphopeptide also being found in IMAC-eluted fractions. In the case of RhoE, which also has a pi of 8.3, the phosphorylated peptides which were tentatively indicated by IMAC binding were in regions where there were fewer acidic residues than basic residues, the basic residues may be involved in some sort of secondary interactions within the peptide which prevent high-affinity binding of the phosphorylated peptides to IMAC resins, such as electrostatic interaction between basic residues upon the side-chain and phosphate moieties.

A third possibility for the lack of phosphopeptide observation in eluted fractions is a lack of elution under the conditions used. The studies using ^^P-labelled casein digests (Chapter 3, Figure 3.2.2) indicated that under the current elution conditions some phosphorylated material may not be recovered from IMAC beads. If poor recovery o f phosphorylated peptides from IMAC beads was occurring, this could also explain the low yield of phosphopeptide in eluted fractions. The latter two of these points indicate that the IMAC binding and elution process had not yet been fully understood or optimised, and that there may be some sequence specificity of IMAC binding and elution which has hitherto remained unreported, probably since a lack of

Chapter 4: Analysis of Signal Transduction Proteins

phosphorylation site analysis. In order to establish IMAC as a routine, robust and comprehensive method for phosphorylation isolation and analysis, more detailed investigation is merited.

Chapter 5: Investigation into the Binding Properties of