From FRET Imaging to Practical Methodology for Kinase Activity Sensing in Living Cells - Progress in Molecular Biology and Translational Science - page 187

become an essential step to create and/or optimize a KAR especially when
specific substrates of the kinase of interest are not yet known (for review see
Ref. 105 ).
To this extent, two libraries have been established for substrate identifica-
tion with peptide microarrays. First, knowledge-based libraries contain many
small peptide sequences isolated from known proteins. The second library
contains de novo synthesized peptides using either randomly generated peptides
or a combinatorial approach. The first type of library allows the determination
of phosphorylation sites using overlapping peptide scans from a known pro-
tein sequence. Furthermore, combinatorial and randomly generated peptide
libraries are useful in detecting specific kinase substrates. Combinatorial librar-
ies define one or more amino acids at fixed positions (e.g., isoleucine at

þ

3
position after the unique phospho-acceptor residue), while other amino acids
(except serine, threonine, tyrosine) are placed at randomized positions.
The number of generated peptides is very high compared to that in
knowledge-based libraries. Combinatorial libraries have successfully identified
PKA (protein kinase A) 106,107 and PKG (protein kinase G) substrates. 107,108
In some cases, a phosphorylation site cannot solely determine substrate
specificity of protein kinases. Substrate specificity is also determined by short
sequences named “docking sites,” which are specifically recognized by a ki-
nase. Indeed, many studies report that docking sites dramatically increase the
efficiency with which a substrate is phosphorylated by a kinase in vitro and in
cellulo. 109-111 The case of the kinase MAPK (mitogen-activated protein
kinase)/ERK is a good example because MAPK family members such as
ERK, p38, and JNK have similar phosphorylation sites. Distinct docking
sites for ERK, p38, and JNK have been identified. It ensures substrate
targeting and can be used to design a KAR with high sensitivity and
selectivity.
Erkus was the first genetically encoded FRET biosensor of ERK activity
in different compartments of single living cells. 112 Erkus has a docking site
(D domain) connected to the C-terminal of the sensor. This D domain is a
common docking site contained in most known endogenous substrates of
ERK and increases the probability of substrate phosphorylation by
ERK. 113,114 However, Harvey et al. 42 have developed a new genetically
encoded FRET-based biosensor extracellular regulated kinase activity
reporter (EKAR) (EKARc and EKARn target cytoplasm and nucleus,
respectively). As an improvement, EKAR exhibits a new ERK-specific
docking site (FQFP) adjacent to the substrate, resulting in a FRET signal
three times larger than that of Erkus, while other potential docking sites