Biology Reference
In-Depth Information
assays also suffer from a number of limitations that must be
taken into careful consideration when interpreting the results
from a given screen. For instance,
on key components of the signaling pathway itself or on
proteins that form a part of the cell's response to activity
through the pathway. For instance, genome-wide screens
using phospho-specific antibodies have identified regula-
tors of dually phosphorylated MAP kinase/ERK and
phosphorylated Akt downstream of receptor tyrosine kinase
(RTK) signaling
[84
the use of a single
pathway-responsive promoter
reporter construct assumes
that all signaling through the pathway converges on the
single readout being assayed, and as a result components that
do not converge on this readout will be missed. Furthermore,
synthetic promoter constructs that are composed of a string
of binding sites for a single downstream transcription factor
do not reflect the endogenous context, as they lack sites for
co-regulators or sequences that may be important for
epigenetic regulation. These limitations contribute to the
false negative rate (the number of true regulators missed)
associated with the screen. Although it is not possible to
completely eliminate false negatives in a genome-wide
screen, one can estimate the false negative rate from
benchmarking the data obtained with known pathway
regulators. Perhaps a more serious problem is experimental
variations due to non-specific factors affecting assay
readout, leading to the accumulation of false positive hits in
the screen. These include factors that affect the level of the
reporter indirectly by having an effect on cell viability/
proliferation, global transcription, protein translation and
stability. Experimental variations are also introduced owing
to differences in transfection efficiency in the case of screens
where the reporter and/or RNAi reagents are transiently
transfected into cells. Thus, appropriate normalization
methods are required to account for false positives due to
such non-specific factors. A commonly used normalization
procedure is co-transfection of a control reporter (for
example Renilla Luciferase) along with the experimental
reporter (for example Firefly Luciferase). A good control
reporter for assay normalization should include a constitu-
tively active promoter that is inert to the pathway under
study and achieves reporter expression significantly higher
than background
[80]
.
Transcriptional reporter assays have also been em-
ployed to identify transcription factor/signaling pathway
regulators in a number of in vivo RNAi screens (i.e., in
whole organisms rather than tissue culture cells). For
instance, large collections of transgenic RNAi lines have
been generated to conduct spatially and temporally defined
in vivo screens in Drosophila
[81,82]
. Such resources can
be used to systematically screen gene functions in specific
tissues for phenotypes or effects on gene expression. For
example, a LacZ reporter for Suppressor of Hairless
(Su(H)) expression in the wing imaginal disc has been used
as a readout in a screen for Notch pathway regulators
[83]
.
In another case, serine proteases that are involved in the
activation of the Toll pathway upon infection have been
identified using a drosomycin-LacZ reporter assay
[80]
.
Another quantitative cell-based assay makes use of
specific antibodies that recognize protein modifications
such as phosphoserine/tyrosine or methyl-lysine residues
e
86]
. The success of antibody-based
assays is critically dependent on the availability of specific
antibodies. Such assays can be performed using a simple
plate reader to measure fluorescence emitted by the fluo-
rescently coupled secondary antibody. Antibody selection
and validation are critical for the development of any high-
quality assay. The specificity of the antibody should be
precisely evaluated by both Western blotting and immu-
nocytochemistry (cell staining). An antibody that generates
strong non-specific bands in a Western blot is not suitable
for plate-based assays. Importantly, it must be determined
beforehand that the antibody truly detects pathway activity
in response to known stimuli and perturbations. A major
difference between Western blots and plate-based assays is
the context in which proteins are analyzed. In plate-based
assays cultured cells are fixed to the bottom of a microplate,
and therefore the immobilized antigens present a slightly
different conformation than those that have been processed
by SDS-PAGE (see
Box 5.1
for definition) prior to Western
blotting. Thus, it should be noted that although some
primary antibodies perform well for Western blotting, they
might exhibit poor binding characteristics on fixed anti-
gens, resulting in low fluorescence signal in the plate-based
format. As in the case for transcriptional reporter-based
assays, the signal from the phospho-specific antibody must
be normalized to account for variations in cell number
across different wells in the plate. A wide array of fluo-
rescent molecules, including DNA-binding dyes (DAPI,
TO-PRO 3), actin-binding dyes (Phalloidin), antibodies to
total protein and non-specific cytoplasmic protein stains,
can be used for normalization. However, it must be first
determined that the stain of choice does indeed provide
a linear measure of cell number. These plate-based
approaches provide an attractive alternative to high-
throughput microscopy (also known as high-content
screening, HCS) for assay readouts that are based on
immunofluorescence detection. If high-resolution infor-
mation is not central to the results of the screen, then plate
reader assays
[84
e
86]
provide a significant advantage in
terms of ease and speed of detection, as well as simplifying
the downstream analysis to a single intensity measurement
per well to report effect on pathway activity.
An alternative to the plate-based assays described above
is the transfected cell microarrays that allow the miniaturi-
zation and simplification of high-throughput assays
[87]
.
RNAi reagents are spotted on the surface of a standard
glass microarray slide and are used to transfect cells. This
generates a living cell microarray comprising locally
e
