The European Lead Factory

In drug discovery, the search for new chemical equity to be developed into valuable pre-clinical and clinical candidates is very time and cost-intensive. This is a particular difficulty for academics or small and medium enterprises (SMEs) who are likely to have limited resources. Additionally, small companies are often created around a single compound, or chemically related series of compounds spun out from academic discoveries. These limited asset business models present a significant risk when new data can bring into question the likelihood of success in progressing the compounds into clinical development. Large pharma mitigates against such risk by employing strategies aimed at identifying chemically distinct back-up or complimentary compounds. One hit finding strategy is the use of high throughput screening (HTS), whereby libraries of hundreds of thousands to millions of compounds are screened against the molecular target of interest. Whilst this strategy proves very fruitful for large pharma it is hardly cost effective for academics or SMEs due to the requirement for expensive infrastructure and specific expertise. In addition to large compound libraries, the infrastructure requirements include facilities and automation for storing and handling the compounds, assay development, protein production, screening automation, large scale data storage and analytics, biochemical and biophysical technologies, structural biology and computational and medicinal chemistry. Low cost access to a large drug discovery platform would, therefore, be a highly attractive proposition for SMEs, particularly if cost can be uncoupled from the potential for failure in finding tractable hit compounds, which is an inherent risk to any hit discovery approach.

The European Lead Factory
The Innovative Medicines Initiative (IMI), Europe's largest public-private initiative in healthcare, identified the difficulty in accessing such infrastructure as an impediment to translating academic discoveries into new therapeutics, and growing and creating innovative new SMEs in Europe. To address this difficulty IMI funded the formation of the European Lead Factory (ELF) a public-private consortium of 30 different partners consisting of industrial infrastructure and expertise from academia, SMEs and large pharma.ELF provides all European academics or SMEs the opportunity to access a large-scale drug discovery platform at no up-front cost. In particular, they gain access to the Joint European Compound Library (JECL). The JECL is a unique collection of up to 500,000 drug-like, mostly proprietary small molecules, more than 300,000 of which are supplied by seven large pharma companies and a further 200,000 synthesised bespoke for the project from novel chemistry ideas sourced throughout Europe. Access is gained through a competitive call whereby molecular target proposals are submitted to ELF to be reviewed by an independent committee of experts who gauge the scientific quality of the proposal taking into consideration the target rationale, availability of a screening assay and plans for the follow up and exploitation of any hits identified from a screen. Once a target has been approved a project team is formed, consisting of ELF scientists and the submitter of the target, known as the Programme Owner (PO). Together, they develop and agree on a plan to screen the target against the JECL. The aim is to design a plan that will best meet the PO’s goals in identifying the most promising hits and generate an associated package of data that will help leverage funding to further develop the compounds along the value chain outside of the ELF. A key deliverable of this process is a list of up to 50 hit structures, called the Qualified Hit List (QHL), for which the PO gains the exclusive rights to exploit. In return there are a series of cash payments that the PO must make to the ELF if certain commercialisation milestones are reached, namely patent filing, first IND approval and the beginning of phase II trials and the beginning of phase III trials. If the output does not successfully meet these commercialisation milestones then no costs are borne by the PO. When a QHL is nominated for a programme, subject to review of the data and the availability of resources within the portfolio, ELF can also provide follow up services including hit characterisation, medicinal chemistry and computational chemistry and crystallisation. The aim of this phase of work is to agree upon a prioritisation of compounds between ELF scientists and the PO. New samples are then synthesized from a couple of series to be used in further studies characterising target engagement, mechanism of action and cross target selectivity. The discussion between ELF scientists and the PO involves developing a new plan with realistic goals, assuming six to nine months of focussed effort by a team of three to six ELF scientists and whatever resources the PO can bring to the programme. If initial data on the freshly synthesised samples is promising then synthesis of analogues aims to explore structure activity relationships and improve the compound’s physicochemical properties, selectivity and potency. In vitro drug metabolism and pharmacokinetic (DMPK)data on the most developed compounds typically finalises the effort to deliver what is termed the Improved Hit List (IHL).

Compound Sharing, Confidentiality and Data Usage
A unique feature of this initiative is the sharing of compounds between pharma companies and the public. The chemical structures of the JECL are held in a secure database called the Honest Data Broker (HDB), which has been designed specifically to protect the confidentiality of compounds, as they remain the property of the company contributors. The HDB administers a series of strict business rules, rights and permissions with the compound structures blinded to all but a small group of individuals within the ELF called the Programme Clearance Team (PCT). Even then, the PCT are only able to see a limited number of hit structures during the screen which become un-blinded when the data facilitates a selection of 0.1per cent of the total number of compounds screened (≤500). While this is principally a data-driven process, there is a considerable array of data annotation associated with the compounds and cheminformatic tools for aiding selection, including physicochemical properties, similarity scoring, clustering information, scaffold-based clustering, multi-parameter desirability scoring, SMARTS filters for undesirable features from all company partners and the ability to exclude/include compounds based on user-defined substructural features. All of this enables a structure blinded but ‘chemically aware’ triage process. The HDB also holds all of the screening data, providing a substantial repository for cross-target/cross-compound analysis to identify and eliminate target or technology specific frequent hitters or non-selective compounds. In addition to providing the computational tools to enable effective triage, the HDB includes comprehensive auditing functionality. It enables the tracking of compound progression through the triage process and records the locations and testing dates of hit compounds selected by the triage team in anticipation of future IP generation and due diligence requirements.

Portfolio and Progress to Date
So far the ELF has built a portfolio of 85 distinct targets and delivered 42 QHL reports. Now operating at a capacity of 15 to 20 HTS per year this is equivalent to the scale of screening in large pharma. The portfolio of targets is diverse with all major target classes represented.  A benefit of ELF being a publicly funded initiative is that a lack of profit motive allows for the inclusion of more challenging targets or under-represented disease indications. Indeed 26 per cent of the portfolio consists of protein:protein (20 per cent), protein:DNA (4 per cent) and protein:RNA (2 per cent) interactions and 20 per cent of the portfolio are targets for infectious diseases, half of which are classed as neglected tropical disease and the other half predominantly classed as gram negative bacteria. To date,1517 compounds on 45 QHLs have been distributed to 10 SMEs and 18 different academics throughout Europe. This has so far led to the creation of two new SMEs; a Swedish company ScandiCureAB spun out from the University of Gothenburg, which is developing ELF compounds as treatments for type 2 diabetes and Keapstone Therapeutics a UK virtual biotech created as a partnership between the University of Sheffield and Parkinson’s UK, which is progressing novel KEAP1:Nrf2 targeting compounds discovered within the ELF. One of the earliest and most developed ELF programmes was from Chris Schofield at the University of Oxford aimed at discovering and developing inhibitors of New Deli Metallo-β-lactamase-1 (NDM-1), a critically important gram negative bacterial multi-drug resistance target. The promising nature of these compounds, including highly potent broad spectrum metallo-β-lactamase activity, good DMPK properties and the ability to rescue carbapenem efficacy in clinically relevant bacterial strains secured the entry of the programme into another IMI funded programme, ENABLE, which is now progressing the molecules towards the clinic. These initial successes are all programmes accepted in the first couple of years of the ELF when operations were being established and so it is likely that many more should emerge in the coming years. Obviously, the true realisation of the benefits will only be forthcoming in the fullness of time, given the long lead times in drug development, but it is a clear example of a functional and ambitious public-private collaboration designed to promote innovation and tackle unmet medical needs whilst protecting IP and promoting valorisation of the outputs.