Leveraging the power of HTS in the new and rapidly evolving collaborative drug discovery landscape

With the advent of technologies such as precise genome editing, the ability to screen against specific mutations relevant to disease is in place. Technologies such as high-throughput mass spectrometry are reducing costs and increasing sensitivity in both biochemical and cell screens. Complex imaging assays coupled with machine learning and deep learning are allowing complex multiparametric cell and tissue screens at high throughput. Automation in data analysis in general is enabling increases in productivity and providing insights on data not seen before. Coupled with virtual screening, compound swaps and automated compound synthesis, the range and scope of drug discovery looks set to expand dramatically. Looking to the future, drug hunters must embrace these new technologies and techniques and apply them in early hit discovery.

High-throughput screening (HTS) remains a key method to discover the hits needed to fuel the drug discovery pipeline.^1,2 Despite an upsurge in New Chemical Entity (NCE) registrations over the last few years, success rates are still relatively few in the pharmaceutical industry and HTS must evolve and adapt to generate the hit chemistry needed to sustain growth within the sector. Additionally, the ever-present focus on Speed, Cost and Quality to increase success rates means HTS and the wider pharma industry must adopt different models and technologies.

The rapid evolution that has taken place within the drug discovery industry – which has seen a far more inclusive and collaborative approach³ – has manifested in many different forms.

No single model has yet proven to be the optimal one, although it is our experience that these collaborations are enhancing the early drug discovery process. It is likely that a mix of approaches will remain to maximise these benefits. A long-standing means of achieving collaboration between academia and industry is the joint post-doc proposal or funded PhD studentship, and these remain important channels often exploring fundamental research that underpins subsequent hit finding collaborations. We are also now seeing initiatives such as AstraZeneca’s open innovation portal where a route is provided for academics and other, usually small, companies to pitch ideas. This enables them to form the basis of a joint venture to discover chemical equity using the hit finding power of AstraZeneca’s global HTS unit (UK Centre for Lead Discovery) and compound collection. AstraZeneca’s experience is that this approach enables new ideas to be generated and we have shared several compound series arising from these efforts.⁴ Another way of generating open innovation is to form an alliance with a funding body, as exemplified by AstraZeneca’s collaborations with the Medical Research Council (MRC) and Cancer Research UK (CRUK). The MRC collaboration allows the MRC to offer academic applicants access to screening the AZ compound collection in the AstraZeneca UK Centre for Lead Discovery and the option to share any chemical equity that is found in the screen. AZ retains a first right of refusal on the chemistry if it proves to be of interest, but if the option is not taken up the academic group has rights to develop the chemistry themselves or with other partners.

Automation remains vital when testing large compound sets in HTS and collaboration with vendors has been key to the development of the latest generation of screening robotics within the UK Centre for Lead Discovery. This has seen collaborative robots deployed in screening laboratories for the first time. However, it is easy to overlook the fact that many other technologies such as dispensers, washers, and screening technologies are also rapidly evolving through such collaborations to support efficient screening processes.

Technological gains are being made in recent screening approaches by applying biophysical assays. The widespread adoption of target engagement technologies such as Surface Plasmon Resonance (SPR) and Differential Scanning Fluorimetry (DSF) now allows rapid identification of compounds that have the desired mechanism of action and rapidly remove the artefactual screening hits. Additionally, high-throughput mass spectrometry is an area of great interest in the industry. There are currently two leading technologies: acoustic mist ionisation and high-throughput MALDI-TOF. These techniques are both reducing costs and increasing sensitivity and specificity in both biochemical and cell HTS screens. Our experience is that they are improving the efficiency of post-hit identification as higher quality hits are found more rapidly. These techniques are developing to become the key label-free technologies for HTS and we expect them to become the dominant biochemical screen format in the future.

With the advent of technologies such as precise genome editing, the ability to generate HTS screens against specific mutations relevant to disease is in place and in combination with the use of spheroids and other novel and more physiological cell culture methods the ability to find and/or assess hit compounds in something closer to a physiological setting is becoming possible. While we continue to develop these technologies and reduce the cost of running them, we are now increasingly seeing complex imaging assays being built for HTS cellular screens.^5-9 With the advent of readers such as the Yokogawa CV8000 with its multiple cameras and Intellicyt high-throughput flow cytometry systems, we are also now seeing improved throughput in complex assays.¹⁰ This, in combination with machine learning and deep learning, will allow complex multiparametric cell screens at high throughput. Hence, we are starting to see imaging-based technology realise its full potential in HTS.

While data rich formats are beginning to help define which chemical equity has the desired hit parameters, data storage still presents a problem as data volumes soon become vast with a concomitant increase in costs. However, with data management policies allowing deletion of, for example, primary HTS images after a screen reads out, this is not insurmountable. Automation in data analysis in general is allowing increases in productivity and providing insights on data not seen before – both in imaging and other more main stream areas of screening.

It is unequivocal that the large compound collections held by pharma companies remain an important asset for them in building the collaborations we have described above. The compounds themselves, their annotation^11,12 and scientific know-how collectively lead to greater success. We are also starting to see a revolution where artificial intelligence is being used to select screening data sets for iterative screening approaches enabling the use of more complex assays and assay platforms where throughput is more limited – often due to cell or reagent supply whilst still enabling the coverage of available chemical equity. This, coupled with compound swaps between companies, is increasing the chances of finding hits for previously difficult to drug targets and the collaborative environment in which we now work increases the access to these cutting edge, but costly, techniques and technology. One of the most exciting opportunities that could transform the way in which HTS interacts with the pharmaceutical pipeline is the possibility of combining artificial intelligence with automated chemistry.^9,13 Beyond this, the ability to learn in real time from data being captured from an HTS and to use this to feed live chemistry giving the potential to dramatically increase the speed at which individual hit molecules can be explored, along with rapid project progression offers a paradigm that could be a significant improvement within hit discovery. 

Looking to the future, HTS drug hunters will need to embrace these and other as yet undiscovered new technologies and techniques and apply them in collaborative early hit discovery to ensure the ability to leverage the power of HTS is maintained – and ideally enhanced – to deliver the new medicines that are needed to patients.

Biographies

Dr Mark Wigglesworth leads The UK Centre for lead discovery on behalf of AstraZeneca. He is an active conference supporter having held co-director responsibility for the European Laboratory Research and Innovation Groups (ELRIG), Research and Innovation conference for 4 years. And has delivered high profile presentations at conferences such as SLAS. Mark has contributed to numerous publications in the field of receptor pharmacology, hit identification and compound management. He has led academic projects that support more than 15 Graduate and Postdoctoral researchers and managed over 80 industrial Scientists. In his current role he leads a team of 28 scientists dedicated to identifying novel chemical starting points for drug discovery programmes and maintains a collaborative interest in Synthetic Biology.

Dr David Murray has over 25 years’ experience in the pharmaceutical industry having held roles in high throughput screening, lead generation and lead optimisation organisations. David currently holds the position of associate director in the AstraZeneca UK Centre for Lead Discovery (UKCLD) currently based at Alderley Park, Cheshire. In the UKCLD David leads a group of scientists responsible for high throughput screening and hit identification across both AstraZeneca’s internal portfolio and its external portfolio which comes out of its various open innovation initiatives. David has contributed to numerous publications and book chapters focussed on hit identification in the pharmaceutical industry.

References

1. Leveridge M, Chung CW, Gross JW, Phelps CB, Green D. Integration of lead discovery tactics and the evolution of the lead discovery toolbox. SLAS Discov. 2018;23(9):881-897.
2. Wigglesworth MJ, Murray DC, Blackett CJ, Kossenjans M, Nissink JW. Increasing the delivery of next generation therapeutics from high throughput screening libraries. Curr Opin Chem Biol. 2015;26:104-110.
3. Rees S, Gribbon P, Birmingham K, Janzen WP, Pairaudeau G. Towards a hit for every target. Nat Rev Drug Discov. 2016;15(1):1-2.
4. Jones S, Ahmet J, Ayton K, et al. Discovery and optimization of allosteric inhibitors of mutant isocitrate dehydrogenase 1 (R132H IDH1) displaying activity in human acute myeloid leukemia cells. J Med Chem. 2016;59(24):11120-11137.
5. Joshi P, Lee MY. High content imaging (HCI) on miniaturized three-dimensional (3D) cell cultures. Biosensors (Basel). 2015;5(4):768-790.
6. Ryan SL, Baird AM, Vaz G, et al. Drug discovery approaches utilizing three-dimensional cell culture. Assay Drug Dev Technol. 2016;14(1):19-28.
7. Sant S, Johnston PA. The production of 3D tumor spheroids for cancer drug discovery.Drug Discov Today Technol. 2017;23:27-36.
8. Knight S, Plant H, McWilliams L, Wigglesworth M. Techniques to enable 1536-well phenotypic screening. Methods Mol Biol. 2018;1787:263-278.
9. Knight S, Plant H, McWilliams L, et al. Enabling 1536-well high-throughput cell-based screening through the application of novel centrifugal plate washing. SLAS Discov. 2017;22(6):732-742.
10. Ding M, Clark R, Bardelle C, et al. Application of high-throughput flow cytometry in early drug discovery: An AstraZeneca perspective. SLAS Discov. 2018:2472555218775074.
11. Mervin LH, Cao Q, Barrett IP, et al. Understanding cytotoxicity and cytostaticity in a high-throughput screening collection. ACS Chem Biol. 2016;11(11):3007-3023.
12. Bedard AC, Adamo A, Aroh KC, et al. Reconfigurable system for automated optimization of diverse chemical reactions.Science. 2018;361(6408):1220-1225.
13. Fitzpatrick DE, Maujean T, Evans A, Ley SV. Across-the-world automated optimization and continuous flow synthesis of pharmaceutical agents operating through a cloud-based server. Angew Chem Int Ed Engl. 2018.