Approaches and Applications of Quantitative LC-MS for Proteomics and Activitomics - LC-MS/MS in Proteomics: Methods and Applications - page 12

of organelle proteins by isotope tagging (LOPIT) ( 50 ) . The main
difference between these two approaches is that PCP uses label-
free LC-MS to quantify the relative enrichment of components in
the subcellular fractions, whereas LOPIT relies on the labelling of
components before quantification. PCP has been used to inves-
tigate the composition and dynamics of nucleolar proteins ( 45 )
and for comprehensive characterization of subcellular proteomes
( 51 ) . LOPIT has been applied to the characterization of mem-
brane proteins ( 52 ) and more recently to identify the protein
composition of several organelles in Drosophila embryos ( 53 ) . A
similar strategy was used to more confidently locate membrane
proteins specific for the basolateral and apical membranes of polar-
ized kidney epithelial cells ( 54 ) . A protocol for PCP is described in
Chapter 15 .

Proteins do not work in isolation but instead have to interact with
other molecules to function. Thus, protein-protein, protein-lipid
and protein-nucleic acid interactions play fundamental roles in
biology. LC-MS is routinely used to detect novel protein-protein
interactions ( 55 - 59 ) . A recent example is the identification of
pyruvate kinase M2 as a phosphotyrosine-binding protein ( 60 ) .
This study used LC-MS to quantify the enrichment of binding
of proteins to tyrosine-phosphorylated peptides relative to bind-
ing of their non-phosphorylated peptide counterparts. In addition
to proteins known to bind phosphotyrosines, the screen identi-
fied pyruvate kinase M2 as a tyrosine-binding protein. This find-
ing inspired further experiments that linked activation of tyro-
sine kinase signalling to the activity of pyruvate kinase, an enzyme
that controls the fluxes of glycolytic pathways often found to be
upregulated in cancer cells ( 61 ) . The Warburg effect (anaerobic
glycolysis in the presence of oxygen, a common occurrence in
cancer cells) may in some instances be the result of the activa-
tion of oncogenic tyrosine kinase signalling pathways, which in
turn affect metabolic pathways through the allosteric activation
of metabolic enzymes ( 61 ) .
As with the case of organelle protein analysis, detecting a pro-
tein in a protein complex by LC-MS/MS does not necessarily
mean that this protein is part of the complex. Protein complexes
are difficult (perhaps impossible) to isolate to homogeneity and
the presence of non-specific contaminants is often impossible to
avoid. An approach to overcome this problem is to over-express a
protein bait with several tags for tandem affinity purification ( 62 ) .
This approach can certainly result in the identification of novel
protein binders but because it needs to over-express the protein
bait, it may result in the introduction of false positives without
biological relevance. An alternative approach is to use quantita-
tive LC-MS to identify dynamic protein complexes that form after
cells are stimulated with specific agents. LC-MS can then be used

2.2.3.Protein
Interactions