How new drugs are discovered
Drug discovery starts with a biochemical hypothesis for a causative or aggravating effect on a disease state. This hypothesis is mostly derived from comparisons between healthy and diseased collective groups of humans in fundamental medical research or from discoveries of biochemical relationships. The evidence is produced by correlating medical and biological data, e.g. genomic, metabolic, behavioral or proteomic data. Then the hypothesis is experimentally verified in cellular systems and if possible, animal models. Stem cell technology has provided pharma research with the unprecedented possibility to test and screen very early in cellular, human, disease specific models derived from patients.
The basis for drug approval is activity and safety
It is important to know that authorities are ready to accept (and patent) drug candidates with proven activity and safety, even if the exact mode of action is unknown. This means in turn that knowledge of the molecular mode of action is not mandatory for a drug’s approval. The submitter has to demonstrate ‘reasonable effort’ to elucidate the mode of action, and special attention must be given to safety, but if the drug works and is safe, it is very likely to be accepted.
The patent protects the research investment
Without patent protection there would be much less investment in research. A chemical drug substance can not be patented per se, but only in context with its medical application.
Two ways (of many) to find drug candidates
a) the classical target based approach..
.. is the identification and validation of a known molecular target being responsible for the observed hypothesis and the subsequent search for patentable interacting ligands influencing its function.
The screening library of compounds is tested for affinity and activity against the drug target using biochemical and cell based test systems. (High-throughput-screening HTS). The result of the screen should be a confirmed collection of some stuctural families with activity.
This way – supposed to lead to selective drugs with fewer side effects – is called ‘rational’ and has been brought to perfection in the past decades. Structural biology and molecular modeling methods can elucidate all details of the ligand-target interaction and assist chemists in improving the efficiency of their compounds.
Lessons learnt
Off-target side effects could be minimized indeed, but we had to learn on several occasions that the desired medical effect could not be achieved in-vivo even though the biochemical suppression or activation was demonstrated by biomarker analysis. It is therefore suspected that a certain amount of polypharmacology is desirable for a good drug. This approach has been termed ‘Network pharmacology’ [1]
b) the phenotypic approach..
.. is the screening of ligand libraries using the above mentioned hypothesis in a high throughput system, mostly a cell based assay. In recent years, stem cell technology has made disease specific, human cellular models available to drug screening and provided the phenotypic approach with a lot of momentum.
This approach delivers drug candidates in a very short period of time, de-selects poorly penetrating compounds and assesses in a real in-vivo system – but without any information on the mode of action and therefore with a high safety risk. Furthermore, it is difficult to optimize the lead compounds since none of the modern biophysical and structure based methods can be applied and the chemist is rejected to the trial-and-error procedure of earlier times.
In consequence of this inconvenience and the above mentioned ‘reasonable effort’, a new task termed ‘target identification‘ (TID) has emerged. It means the identification of the “interactome” – all biochemical interaction partners – of a given ligand.
Lessons learnt
The target ID effort is very complex to manage. It requires concerted activity of organisationally distributed groups on the same model system, at the same time and with simultaneous data analysis.
The candidate list provided by the TID effort can be quite large and induces big validation efforts: Candidate proteins have to be produced, biophysical assays set up and the obtained ligands tested for binding. It might make sense to wait for the completion of an early safety package: Safety issues can disqualify a compound.
A special issue of Nature Chemical Biology from 2013 gives more background and outlines some of the involved technologies [2].
Sorting out the best candidate in Lead Optimization
In ‘lead optimization’ stage, the discovered lead families (structurally similar hit compounds) are optimized towards maximal potency, optimal duration of action and minimal risk. This is a multifactorial problem and requires a battery of very diverse assays (activity in-vitro and in-vivo, biophysical assessment like Kd, kon, koff, physicochemical properties, safety, metabolism, pharmacokinetics and distribution in-vitro and in-vivo studies).
Each research organization has its own tradition of weighting these results against each other. Often the decisionmaking is influenced by historical failures and successes, and each project presents an unique situation. Therefore the final choice of a clinical candidate is made after discussions among the specialists in the project team and the biology, chemistry and safety management – and yes, it’s not easy.
Bringing it to clinical stage
The decision for entering development – and even more, clinical development – is a very expensive one. The project team is heavily challenged and has to be prepared to questions ranging from market potential, competition situation, medical need and differentiation to established therapy, to name a few.
One of the most difficult tasks in this stage is the translation of the drug candidate’s efficacy from cellular and animal models to man. In this regard it helps a lot if an analytical method for a biochemical target response marker is available and an effect in animal models has been demonstrated at this stage.
Here smv3.ch comes into the game: Liaising project biology and external analytical labs, we develop methods for target response biomarkers and validate them to the required degree. If this is possible, confidence in projects and compounds is increased a lot.
A personal word
The key to a maximal productivity of research is concentration on the essence of drug discovery combined with a meaningful frequency of strategic changes. The essence is to improve the patient’s condition – and therefore it is our duty to deliver new or clearly better medicines with affordable risk, at the earliest possible time and minimal costs. Strategy changes slow this down and waste research resources – Therefore they need to be well reflected. Patient needs rarely change within 5 years..
References
[1] Hopkins, A. L.: Network pharmacology: the next paradigm in drug discovery; Nat Chem Biol 4, 11, (2008); doi:10.1038/nchembio.118, M#5801
[2] Schenone, M., Dancik, V., Wagner, B. K. and Clemons, P. A.: Target identification and mechanism of action in chemical biology and drug discovery; Nat Chem Biol 9, 4, (2013); doi:10.1038/nchembio.1199 , M#5800
[3] Malik, F., Führen Leisten Leben; Campus Ed.; Ist ed. (2014), ISBN 978-3-593-50127-7