Assay and Drug Development Technologies - 12

GIULIANO ET AL.
subsequent expert assessment. Cellomics has employed
modern bioinformatics, proprietary text mining algorithms,
and a biological vocabulary (or ontology) to provide an automated
knowledge extraction system for the scientific literature.
Such a system is embodied within the CellSpace
Knowledge Miner, from which end users can rapidly search
for relationships in the scientific literature (Medline) and use
this associative knowledge to
qualify, annotate, and classify the
results from HCS experiments.
For example, if an HCS experiment
identifies a set of compounds
with key target activities,
the known relationships between
these compounds and target
molecules or pathways can be
rapidly qualified. In this manner,
the novelty of the result can be
easily assessed, and decisions
made as to whether to proceed
with that compound set, or target.
The field of HCS has expanded
over the last 5 years from essentially
nothing to a state where
many pharmaceutical and biotechnology
companies report
increasing use of these technologies
in their drug discovery efforts.
Maximal discovery benefit
from HCS is gained when the
acquisition, management, analysis,
and reporting ofHCS data are
optimized. Achieving this requires
a thorough understanding
of the discovery workflow, HCS,
informatics, and bioinformatics.
The image generated by instrumentation,
based on the biology,
is just the start of an entire
workflow and user experience.
Furthermore, for the data and
information from HCS to be
turned into knowledge about
targeted cellular function, the
context of that data and information
needs to be grounded
with respect to other disciplines
such as genomics, proteomics,
and chemistry.
EXAMPLE OF AN INTEGRATED SOLUTION: HIGH
CONTENT PROFILING OF A SMALL COMPOUND
LIBRARY FOR CELLULAR TOXICITY
Cellular toxicity is a complex response involving myriad
cellular and molecular processes at multiple levels of organization
from ions and metabolites to macromolecules and organelles.
By integrating an automated liquid handling system
FIG. 4. A combinatorial cell biology approach to toxicity screening: High-content profiling of a 160
compound toxin library (Killer Plates 1 and 4; MicroSource Discovery Systems, Gaylordsville, CT)
against 15 molecular, organellar, and cellular targets in multiple cell types (HeLa cervical carcinoma
and HepG2 heptocellular carcinoma). The targets included transcription factor activation (e.g.,
nuclear factor kB [NF-kB], c-jun, STAT1, etc.), organelle physiology (e.g., lysosomal [Lyso] and
mitochondrial [Mito] mass), stress kinase activation (e.g., extracellular signal-regulated kinase
[ERK], p38 MAP kinase [MAPK], stress-activated protein kinase [SAPK], etc.), and whole- cell
measurements (e.g., morphology [Morph], membrane permeability [Perm], etc.). An automated
liquid handling robot, the Beckman (Fullerton, CA) Biomek FX, was used with Cellomics HitKit
HCS reagent kits to produce the data mosaic containing the name and number of HCS assays that
produced hits for each of 59 compounds. Inset: An example data summary used to produce hit
results from one HCS assay, p38 MAPK kinase activation. Each datum point represents the p38
activation level that one compound induced, but only half the number of compounds tested are
displayed here. Compounds that induced p38 activation at or above the 99% confidence limit set by
untreated control values were included as a hit in the large mosaic graph. Thus, an entire profiling
campaign with an aggregate of over hundreds of thousands of single-cell measurements can be
made in a matter of days.
12 ASSAY and Drug Development Technologies
ª 2022 MARY ANN LIEBERT, INC.

Assay and Drug Development Technologies

Table of Contents for the Digital Edition of Assay and Drug Development Technologies