Data Analysis Pipeline for High Content Screening in Drug Discovery - High-Throughput Image Reconstruction and Analysis - page 210

but nontoxic drug leads. Hence, there is a need for refined approaches to predict
success earlier in the discovery process [5]. HCS, which can simultaneously mea-
sure multiple cellular phenotypes which are relevant to both efficacy and toxicity,
became increasingly popular because it holds the promise of expediting the drug
discovery process [6].
Early HCS applications were mainly focused on secondary assays for the hit-
to-lead process. Evaluating the effects of hit compounds in cellular models is crucial
for making decisions on the fate of the compound, as well as the direction of the
drug discovery program. Hit compounds are usually tested in a variety of assays
reflecting their on-target activities, off-target activities, toxicities, and physico-
chemical properties. Cellular imaging based HCS assays have been developed in all
these areas. Successful cases include nuclear and plasma membrane translocation
of transcription factors or signaling molecules, endocytosis of G-protein coupled
receptor (GPCR) after activation, cell cycle, and neurite morphology assays [7--10].
With the increasing realization of the pitfalls of targeted based drug discovery,
HCS has now been adopted earlier in the hit discovery stage at many major phar-
maceutical companies [11]. Interestingly, phenotypic profiling of drug activities
using high content analysis has been shown to be able to predict drug mechanism
of action (MOA) with remarkable accuracy, comparable to the more expensive
and low throughput transcription profiling technology [12, 13]. The large quantity
of single cell descriptors from high content analysis might hold the key to such a
result. High-throughput prediction of drug MOA and toxicity is one of the most
important challenges facing drug discovery. High content analysis is predicted to
have an increasingly important role in this area as the technology develops.
9.3 Types of HCS Assay
Thanks to the rapid development of knowledge on molecular and cellular functions
in the past two decades, many cellular processes can be adopted into HCS assays
format in rather straightforward ways and usually in a short timeframe. Current
HCS assays can be roughly divided into three main formats: (1) translocation
assays, (2) cytometry assays, and (3) morphology assays. Figure 9.1 illustrates
examples of the three assay formats.
Translocation assays score for concentration shift of a particular protein from
one cellular compartment to another. Immunocytochemistry using specific antibod-
ies and expression of green fluorescence protein (GFP) fusion protein are among
the most commonly used methods for detection. The ratio or difference of the
concentration in different cellular compartments was used as a quantitative score
for such events [14]. Nuclear-cytoplasmic translocation of transcription factors
or signal transduction molecules, such as NFkB, NFAT, FOXO, p38MK2, and
HDAC4/5 are among the most common ones of interest. Because of the simple for-
mat and robust quantitation method, nuclear-cytoplasmic translocation was also
adopted in engineering biosensors to detect binding or cleavage events in cells, such
as p53/MDM2 interaction or caspase3 activation [15, 16].
Cytometry assay treats each cell as a single entity and scores for population
change of cells with certain phenotypic characteristics. Fluorescence activated cell