Are cell-based assays the next big thing in laboratory automation?
Cell-based assays are labor-intensive, time consuming, and tricky. They are also becoming increasingly popular in the early stages of the drug development process, making them a potential road block during a point when a researcher might be hoping to keep costs down. Could automation be the answer?
We spoke with Roger Clark, Group Leader within the HTS Biology department at Charles River, to find out. Roger is also Head of the Global Compound Management group at Charles River – leading multiple teams supporting compound plate provision to all HTS and iterative SAR projects, along with management of client proprietary libraries and the provision of Assay Ready Plates for screening at client premises.
Over his pharmaceutical R&D career Roger has worked across many fields of early discovery including HTS, SAR Screening, High Content Biology and Laboratory Automation. Here are his edited responses to our questions about the promises and pitfalls of cell-based assay automation.
What makes cell-based assay automation an ideal technology for drug discovery?
RC: Cellular assays are an incredibly important part of every drug discovery cascade, but can be some of the most complex to run due to multiple reagent addition steps (often coupled with long incubation times). This makes cellular assays labor-intensive, and therefore some of the best candidates for automation.
What are the most common applications of cell-based assay automation in drug discovery today?
RC: As cellular assays are moved earlier in the drug discovery process, either as primary screening platforms, to confirm cellular target engagement, or to gain early insights into off-target cell-tox effects, the requisite throughput of these assays inevitably increases. Automating these more complex multi-step assays enables the best use of time and resources not only in early Hit identification, but also in later stage iterative hit-lead or lead optimization phases.
How has cell-based assay automation changed over the past 5-10 years?
RC: The key changes are linked to the reliability and accuracy of the automation deployed. Vendors are now producing integrated platforms employing more accurate multi-axis robots, enabling far better teaching of plate-positions, and therefore lower likelihood of failed assay runs through avoidable system crashes. Reliability of peripheral devices has also improved immensely – particularly with devices central to cellular assays such as incubated plate carousels and automated plate-washers.
What are the key benefits of using cell-based assay automation in drug discovery?
RC: As mentioned previously – automating cellular assays enables the throughput to be increased for these complex assays. However, automation not only allows more samples to be processed, it also enables more consistency across assays day-to-day, which is hugely important when the key drive of any assay is to measure differences in response to a given variable. Enabling cellular assays to be lifted higher up the drug-discovery cascade means that targets which would be unavailable to prosecute with isolated biochemical approaches can be considered, and robust screening campaigns put together. This also allows molecules to be tested on more physiologically relevant assay systems much earlier, and therefore efforts can be employed on prioritized chemical series which drive the desired phenotype in those relevant cell backgrounds.
What are the biggest challenges to implementing this technology?
RC: Cost and space requirements for automation can be a barrier for smaller organizations. The current trend towards smaller, more modular automation platforms helps address the space concern. Furthermore, for more complex automation (including High Content Imaging plate-readers or High Throughput flow cytometers) it is important to maintain specialist skills within an organization. You need skilled personnel who can train robotic platforms or recover platforms from crashes, along with those who can configure the more complex cellular readers.
Within Charles River we are fortunate enough to have dedicated experts in all of these areas, allowing us to use the more complex assays to support our clients’ projects from Hit ID right through the preclinical phases of drug-discovery.
Are you seeing any new trends emerging for cell-based assay automation?
RC: We are seeing a resurgence in phenotypic screening approaches from our client base, even though molecular target-based approaches continue to dominate. This is a trend replicated across the drug-discovery industry that looks like it is now translating to smaller clients. With this trend, the demand for automation to facilitate higher throughput cellar assays obviously builds, and we see vendors responding to this. High Content Imaging generates a wealth of information from cellular assays, and lends itself extremely well to this phenotypic resurgence. This enables automated assays to not only report out data on key target responses, but also on localization of target(s) along with initial cellular-tox data. The partner to High Content Imaging is High Throughput flow cytometry – enabling similar levels of population response data to be acquired for suspension rather than adherent cells.
Where do you see the technology taking drug discovery in the next 5-10 years?
RC: We believe the resurgence of phenotypic drug discovery will continue as researchers explore more complex disease types and want to screen against the interplay of different cell types in a system more closely replicating the disease state. This complex co-culture system will be employed in the lower throughput lead optimization or safety profiling phase to a greater extent initially, and certain cellular assays will require specialized plate-types (so called ‘micro-physiological systems’ or ‘organs on a chip’). These assays again lend themselves to High Content Imaging as a detection modality, and as such even faster plate-based confocal imaging systems will emerge. In order to exploit the vast amounts of information available within confocal cellular images, the analysis software used will continue to improve. The application of Deep Learning to automate the interpretation and clustering of high content imaging output will bring a step-change in the way this technology is employed in drug-discovery cascades.