High-throughput screens are a recognised hit-finding process but the use of an integrated discipline approach is key to maximising the output from such a screening methodology. Ian Linney and Scott Maidment, Charles River, describe the value of integration of medicinal chemistry in this method.
A review in the Journal of Medicinal Chemistry1identified the lead generation approach of 66 clinical development candidates. It showed that random high-throughput screening (HTS) was used as the starting point in 29% of compounds and 13% were identified by either directed- or fragment-based screening approaches. These findings highlight that screening approaches of compound collections are still a significant method for the identification of chemical hit matter.
An integrated HTS team: medicinal chemists and screening biologists
The use of an integrated discipline approach is key to maximising the output from such a screening methodology. The combination of medicinal chemistry expertise alongside highly-skilled screening biologists allows for the identification of high-quality chemical starting points that will make the passage through multiple Design, Make, Test and Analyse (DMTA) optimisation cycles to a clinical candidate more successful and hopefully faster.
Prior to running an HTS, three fundamental questions need to be addressed:
i) What compounds will be screened2?
ii) How will the compounds be screened? (Essentially: what assay format will be used?)
iii) Consideration of the target.
Currently, there is an emphasis on the quality of compounds to be screened as opposed to sheer numbers. This has led to the generation of the ‘lead-like’ libraries containing compounds, devoid of structural alerts, with molecular weights in the region of 175 to 400 Daltons, logP less than 4, and a tPSA of less than 100. The initial physicochemical properties of these potential hits give the medicinal chemist significant leeway when applying the DMTA optimisation process.
However, we are increasingly screening targets that are more challenging. Whether that be screening for inhibitors of protein-protein interactions, enzymes such as phosphatases or helicases or seeking allosteric inhibitors of proteins. In these cases, the need to survey much wider chemical space to find hits, gives rise to the Diversity-type collections, where the physicochemical descriptors are less rigidly applied and the library tends to be significantly larger in number to cover this increased chemical space.
A third library type is the focussed target-based variety. These are specifically designed to target gene family members such as kinases, GPCRs or ion channels. They tend to be smaller in nature but can offer a rapid and cost-effective means for generating hit matter against those targets. The screening collections are also ‘living things’ in that they are consistently evolving to incorporate new thinking in the medicinal chemistry field, such as the ‘escape from flatland’3 by incorporating more 3-dimensional character or new modalities such as covalent inhibitors. It is at this first stage that the incorporation of the medicinal chemist into the HTS team can impact on the quality of the output from an HTS. A demanding selection criterion should be applied to all new compounds going into the library:
i) synthetic tractability of the scaffold and ease of introduction of outlying substituents
ii) calculated physical properties
iii) 3-dimensional character
The fundamental question is whether the hit obtained from an HTS can be considered as a developable lead. From the screening biology perspective, a key point will be around any liabilities associated with the particular screening methodology. When deciding the format for the primary assay the researcher needs to consider the target biology and assess whether compound interaction with the target can be measured using a functional assay, such as inhibition of an enzyme or cellular signalling pathway, or whether a binding assay can be employed.
The choice of assay format may be dictated by relevance to the therapeutic area, reagent availability, throughput, cost or a combination of these. In all cases the inclusion of an appropriate counter screening assay will be an essential component of the cascade for the elimination of false positives. The counter screen for a target screened in an over-expressed cell line could, for example, utilise the non-transfected host cell line, or one where the target of interest has been knocked out using CRISPR. Likewise, in a biochemical assay set-up the counter screen may be run with a mutated form of the target protein, rendering it inactive, to establish whether compounds interfere with the detection system. Further confidence in the target specificity of the hit compounds may be obtained by the addition of an orthogonal assay with an alternative read out to the screening cascade.
The understanding of the target is also key in setting the expectations for the chemical output from the HTS campaign. Leveraging the experience of the integrated HTS team, who will have a deep understanding of assay formats and targets, in the identification of targets that have the potential for delivering false positives is crucial. An example of this is the well documented role of metal impurities as an issue in HTS. The combined team can mitigate any potential issues by appropriate early use of LCMS purity confirmation and understanding of chemical functionalities that could potentially act as metal chelators4,5.
The typical workflow for an HTS is outlined in Figure 1 and the first impact of the combined team can be felt at target disclosure. Literature analysis around the target can identify potential tool compounds to aid in the assay development and provide understanding of known pharmacophores. This information can aid in the early understanding of the hit matter from the HTS. One possible example of this enhanced understanding could be a screen that is seeking allosteric modulators of a protein with known orthosteric ligands. Understanding the pharmacophore of the orthosteric ligand can allow for the de-prioritisation of any hits that match this motif. In the assay development step the use of annotated screening libraries, such as the SelleckChem FDA-approved Drug library or Prestwick Chemical Library, is commonplace. The imbedded medicinal chemist can confirm that any annotated compound behaves as expected but also help identify any potential assay interference compounds.
An example of this could be if the assay contains ssDNA do DNA intercalators such as doxorubicin register as positive hits, which could be considered a false positive if the screen is attempting to identify a protein /ss DNA interaction. Within the primary HTS the medicinal chemist will add a structural component to the data generated by the screening team. The team will look for the biological data to translate into clusters of similar compounds and for there to be a range of biological effect within these clusters. This structural analysis alongside the assay statistics from the assay itself, such as the robustness (Z’) and signal to background (S:B) of the assay, provide confidence that the HTS has worked to identify potential chemical hit matter.
The integrated team will then look to confirm the activity of these clusters, alongside potent and structurally diverse singletons in a hit confirmation step. To increase confidence in the hit matter identified at this stage a frequent hitter analysis can be performed6. This would determine how often the individual compounds confirmed in the hit confirmation have been annotated as a hit in previous HTS campaigns. If a compound is flagged as a frequent hitter, the integrated team will need to decide the risks involved in carrying the compound forward into the XC50 determination step. The risk could be that the compound is either associated with assay interference or is too promiscuous to be consider a hit for the screened target.
After the hit confirmation step the XC50 or potency determination step is undertaken. The joint team will ensure that sufficient hits from the hit confirmation step are taken into this phase, because it is at this stage that detailed structure-activity relationships (SAR) are determined. The use of R-group analysis on the individual clusters can aid determination of the optimal substituents and the identification of obvious gaps in the SAR. This SAR analysis also allows for the early transposition of SAR from cluster to cluster. The output from the HTS with a complete structural breakdown and an early understanding of structure activity relationship against the target is now available.
Beyond the HTS: the role of continued integration
By the time the screen is transitioning from the hit confirmation into the potency/XC50 determination phase the integrated medicinal chemist will have a good understanding of the structural features required for interaction with the target protein. They will know the scaffold from the most populated clusters and may have some idea of substituents that the protein tolerates. With this information they can start to design the next set of novel compounds to be made – the Hit Explosion. Synthetic routes to the scaffold can be identified and routes to rapidly install outlying substituents can be devised. This would allow rapid array chemistry to further generate confidence in the hit matter from the HTS. A complimentary computational approach would be undertaken by the computer-aided drug design (CADD) group of the integrated team to identify further commercially available analogues using Hit Expansion techniques – the so-called ‘SAR by catalogue’ approach.
From the workflow shown in Figure 1 the finish of the XC50 determination step would align with the synthesis of new compounds from the Hit Explosion. The compounds sourced from the computational Hit Expansion would be available to screen alongside the compounds from synthesis. These combined, parallel activities would aim to build significant confidence in the identified chemical matter
To further bridge towards a hit-to-lead programme, between the hit confirmation and potency determination the integrated team can start to identify an appropriate hit-to-lead screening cascade. Crucially, that first step in building the screening cascade is already available and running – the primary assay used within the HTS itself. The integrated team would then work towards the identification of a suitable cell assay aiming to generate a cell assay that could have multiple biomarker read-outs to allow for translation of a biomarker into the in vivo setting.
The integrated team would also confirm that the compounds identified in the HTS are interacting with the target. This may involve the addition of a structural biology member to the integrated team. They would work to determine: is it possible to confirm target binding through the generation of an x-ray crystal structure? If this is not possible what other orthogonal approaches are available such as surface plasma resonance or isothermal calorimetry to confirm target binding.
The integrated team would also consider what selectivity assays are required within the cascade. Close gene family member selectivity or are there counter targets suggested by the team’s understanding of the fundamental biology of the target.
There is also a need to ensure that the responses observed within the routine screening cell assay are replicated in primary cells is satisfied by the generation of a suitable phenotypic primary cell assay – the team would be looking for this in both mouse and human – again with one eye firmly on developing a screening cascade that will translate to humans.
Alongside this, the integrated team would access the wider drug discovery infrastructure to ensure that the relevant ADME and DMPK disciplines are engaged so that the project team can be aiming for putting compounds into PK studies with the ultimate goal of replicating the cellular phenotypes within an animal setting. The screening cascade would be designed to allow for translation from in vitro to in vivo and from rodent to human.
The advantages of having an experienced medicinal chemist embedded within the HTS process and then expanding that integration into the wider drug discovery group allows for hit validation activities, with the aim of building confidence in the quality of the hits identified in the HTS, to run in parallel. The planning of the screening cascade in conjunction with these hit validation assays will shorten the time between the HTS and hit-to-lead activities.
The integrated team decision making in this combined HTS and early hit-to-lead phase, the identification of quality chemical hit matter with early ADME characterisation coupled with the clear translational screening cascade, will save time in any transition milestones, such as entry into lead optimisation. The time saved in this enhanced decision making is key both when working within a competitive target area and in terms of the costs of a full project.
Volume 22, Issue 2 – Spring 2021
About the authors
Ian Linney is a Research Leader within the Medicinal Chemistry group at Charles River, where he is involved in running integrated drug discovery projects from target identification through to clinical candidate nomination. Linney is also heavily involved in providing medicinal chemistry support to the many High Throughput Screening campaigns prosecuted for our clients at Charles River.
Scott Maidment is a Research Leader at Charles River in the High Throughput Screening Biology group, working on hit identification projects to deliver small molecule hits for progression into further drug development.
- Brown DG and Boström J. Where Do Recent Small Molecule Clinical Development Candidates Come From? J. Med. Chem. 61(21), 9442-9468, 2018.
- Volochnyuk DM, Ryabukhin SV, Moroz YS et al. Evolution of commercially available compounds for HTS. Drug Discovery Today, 24(2), 390-402, 2019.
- Lovering F, Bikker J, Humblet C. Escape from flatland: increasing saturation as an approach to improving clinical success. J. Med. Chem. 52(21), 6752-6, 2009.
- Hermann JC, Chen Y, Wartchow C, et al. Metal impurities cause false positives in high-throughput screening campaigns. ACS Med. Chem. Lett. 4(2), 197-200, 2012.
- Morreale FE, Testa A, Chaugule VK, et al. Mind the Metal: A Fragment Library-Derived Zinc Impurity Binds the E2 Ubiquitin-Conjugating Enzyme Ube2T and Induces Structural Rearrangements. J. Med. Chem. 60 (19), 8183-8191, 2017.
- Nissink, JWM and Blackburn S. Quantification of frequent-hitter behavior based on historical high-throughput screening data. Future Med Chem. 2014 June;6(10):1113-26