eliminating false-negative hits in ATP-luminescence


eliminating false-negative hits in ATP‑luminescence viability screens by use of an alternative phenotypic approach with acumen® Cellista


Cell viability and proliferation assays are a fundamental tool in the drug discovery process. They are used to evaluate both the potency of compounds, as well as their toxicity profiles for drug safety assessments. Many commonly used plate reader assays (e.g. ATP-luminescence measurements) assume a linear relationship between the assay signal, the number of cells and their viability. However, as we illustrate here, this assumption is not always justified: ATP-luminescence measurements can significantly overestimate toxicity and underestimate potency, leading to false-negative viability/ efficacy hits.

Here we show the development and implementation of an alternative no-wash high content (HCA) assay using acumen, which eliminates common problems associated with ATP-luminescence measurements. We compare the drug-responses of several well-characterised anti-cancer drugs on HeLa cells, measured by acumen and a commercial ATP-luminescence detection system, and highlight key differences between the measurements.

key benefits

TTP Labtech’s acumen facilitates replacement of ATP-luminescence based cell health assays, resulting in better quality data with cheaper reagent costs:

  • multiplexed readouts  including cell number, % live cells, cell cycle phase
  • ready distinction between cytostatic and cytotoxic mechanism of action for higher quality hits
  • simple, homogenous assay protocol @ 2.28 pence/well (acumen) versus 14.6 pence/ well (CellTiter-Glo)
  • compatible with high density 384-well plates, scanned in 8 minutes
  • applicable to 1536-well plates for even higher throughputs

materials and methods

cell culture and drug addition
All plates are prepared as duplicates, one for imaging on the acumen Cellista and the other for ATP-luminescence determination.

day 1
Seed 500 HeLa cells in 25 μL of medium into the wells of a 384-well plate and culture overnight.

day 2
Prepare 2-fold serial dilutions of drugs in medium (2x concentrated), add 25 μL to each well and incubate for 48 hours.

day 4
HCA assay: add 5 μL of stain “master mix” in PBS to each well, to give final dye concentrations of 10 μM Hoechst 33342, 1 μM calcein-AM and 1.5 μM propidium iodide. Incubate the plate for 90 minutes at room temperature and then image on the acumen Cellista to determine the cell number, % live cells and cell cycle phase distribution for each well.

ATP-luminescence assay: add 40 μL of CellTiter-Glo reagent (Promega) to each well, agitate the plate for 60 minutes at room temperature and then read the luminescence signal on a multimode plate reader.

data analysis
Average and normalise the well readouts for each drug concentration to the control wells. Plot drug-response curves of the normalised values and fit to sigmoidal curves to derive EC50 values.

reagent unit price unit size & stock concentration cost per well
calcein-AM £233 1 mL, 1 mM 1.12p
proidium iodide £85 10 mL, 1.5 mM 1.12p
Hoechst33342 £82 5 mL, 20 mM 0.04p
CellTiter-Glo £56 10 mL (sufficient for 1 x 384-well plate) 14.58p


phenotypic assay validation % live determination
Control or doxorubicin-treated cells were classified as either live or dead based on the ratio of the green (calcein-AM, FL2) to the red (propidium iodide, FL3) fluorescence intensity per cell. The percentage of live cells was calculated from the total number of live and dead cells.

cell cycle determination
acumen cell cycle classification gates were derived from the total FL1 fluorescence intensity histograms of untreated cells, and of cells treated with 3 μM etoposide for 48h (Figure 1). The predominant cell cycle response for each drug is summarised in Table 1.

comparison of drug‑responses across different assay formats

The drug-response curves (Figure 3) and EC50 values (Table 1) show that the agreement between the acumen Cellista cell number and % live cell readings, and the ATP luminescence readings varies widely between the compounds.

For doxorubicin, a classic cytotoxic agent, there is very close agreement between all three readouts.

For etoposide and cycloheximide there is less agreement, with the EC50 values for the acumen % live cell readout differing from the cell count by more than 1 log unit. The discrepancy between the two acumen readouts immediately highlights the cytostatic potency of these compounds and the cell cycle profiles confirm that etoposide is indeed a cell cycle inhibitor, whereas cycloheximide (a protein synthesis inhibitor) by elimination must work via a different mechanism of action. Furthermore, the acumen Cellista Tiff images (Figure 4) show an unusual enlarged morphological phenotype conclusions cannot be drawn from the ATP-luminescence measurements: The cycloheximide drug-response curves mirror the acumen cell count, which taken in isolation would falsely suggest a highly toxic compound. The etoposide drug-response curve falls in between the acumen cell count and % live readouts, probably because the antiproliferative action of the drug is accompanied by an increase in average cell size (and mitochondrial mass). With etoposide, the ATP luminescence measurements would also overestimate the toxicity of the compound.

With staurosporine, the acumen readouts show clear monotonic drug-response curves, with a cytostatic mechanism of action at lower drug concentrations, which becomes cytotoxic only at higher drug concentrations. The ATP measurements are more complicated, with a bi-phasic dose response showing the composite effect of cell health and number. In this example, the bi-phasic nature of the dose response curve hints at different mechanisms of action operating at different drug concentrations, however, the effect is quite subtle and could easily be missed. As with etoposide and cycloheximide, staurosporine would appear much more toxic at lower concentrations than it actually is.

drug log EC50 cell cycle response
cell count % live cells CTG
staurospine -9.1 -6.7 -8.7 (monotonic)-6.8 (bi-phasic) G2/M, >4N
doxorubicin -6.4 -6.3 -6.8 G1
etoposide -5.9 -4 -4.8 G2/M
cycloheximide -6.6 N.D. -6.5 G1

Table 1. Summary of Log EC50 values


Read our high-throughput phenotypic screening workflow for cytotoxicity determination poster here






  • acumen Cellista provides content rich cell viability data in a simple and homogenous assay format that utilises cost-effective reagents.
  • Whereas ATP measurements alone can significantly overestimate the toxicity of cytostatic drugs, acumen Cellista viability assays can readily distinguish between cytostatic and cytotoxic mechanisms of action.
  • Cellular DNA content measurements by acumen Cellista can easily identify drugs affecting the cell cycle, whilst the Tiff export function can highlight morphological changes of the cells.
  • acumen Cellista assays are amenable to miniaturisation and automation, offering improved process efficiencies.

about acumen

TTP Labtech’s acumen is a laser scanning imaging cytometer designed to provide single-shot, whole well, content-rich cytometric and image-based analysis. Its F-theta lens gives a uniform illumination across the field of view with a large focused depth of field, which enables high throughput, whole well image acquisition across a range of plate types. acumen enables a wide range of fluorescent reagents to be combined in
multicolour, multiplexed assays. Its easy-to-use, template-driven software offers an industry-proven route for quick adoption across a wide range of applications.

about cellista

Windows based Cellista software is powerful yet approachable. Cellista gives you the best of both worlds: application fl exibility combined with simplicity of operation. The software has been refined over many years and allows even the most novice users to rapidly become confi dent with the instrument.

acumen Cellista uses the principle of cytometry, rather than image analysis. This intelligent cytometric approach is quick and easy to learn, eliminating the requirement for specialised operators and endless ongoing training. Data output files are small, enabling acumen Cellista to be integrated into a screening workflow with little change to your existing infrastructure.


app notes

acumen: multiplex measurement of toxicity indicators
TTP Labtech’s acumen Cellista provides a way to incorporate toxicity screening during the earliest phases of the drug discovery pipeline

acumen: high throughput quantification of hepG2 cell colony formation in soft agar
The acumen can rapidly and accurately quantify cell colony formation in soft agar.

acumen: detecting kinase phosphorylation
acumen provides an extremely fast, sensitive and robust method to detect the effects of small molecules or RNAis on signalling pathways.

acumen: detecting protein translocation
acumen provides a fast, simple method to study protein translocation, without the requirement for off-line complex image analysis algorithms.

acumen: rapid analysis of cell cycle phases
acumen hci allows rapid high-content screening of the effect of compounds on the cell cycle.

acumen: speeding up angiogenesis assays
The acumen hci offers rapid whole well image acquisition, with no need for image stitching prior to data analysis.