Coordinators:
- INSERM U761 : Florence Leroux
- JE2491 : Juergen Siepmann
Mission:
Maximise the output of in vitro compound optimisation and enable the interpretation of in vivo data and optimise the management of early in vivo experiments.
Objectives and Rationale:
Many academic projects fail to deliver acceptable proof of concept for novel modes of actions because they cannot provide solid evidences that compounds modulating the chosen target actually modifies the physiopathology of an animal model in the desired way. Failure to obtain an in vivo proof of concept can have two principal causes. One is pharmacodynamic the other is pharmacokinetic. Either the mode of action (i.e. the target) chosen is not appropriate, or the animal is not appropriately exposed to the compound, which does not reach its target at the required level. If the net present value of any given project is a function of the probability to reach the expected outcome, a leap in industrial value exists between the sequence of a protein whose gene is linked to a disease (“concept”) and an experimental evidence that modulation of the coded protein by a drug-like molecule can actually modify the pathology in the expected way (“proof of concept”). Indeed, this experiment, while not qualifying the compound itself, validates the therapeutic approach using the novel target. A successful experiment in an animal model will be possible if:
- One has access to a recognized animal model of the target pathology,
- If the compound reaches the target concentration at the site of action in a sufficiently sustained manner,
- If it is not toxic at the dose employed to reach the target concentration,
- If the compound triggers the expected physiological response in the animal.
We believe that such a “proof of concept” in the animal would be sufficient to convince a pharma company to license in the drug target and even buy the rights on the compound in the best case. This proof of concept would also be a valuable asset for a biotech company seeking its first venture money. Both scenarios are valuable exit points for an academic institution. Many animal models are available in the academia, which are often recognised as reliable translations of human diseases and as such often in-licensed by pharmaceutical companies. In Lille, for example, biologists have produced animal models of dyslipidemia, metabolic syndrome, or Alzheimer disease. At the other end of the discovery process, academic groups produce large amounts of biological data (from epidemiology, genetic or differential expression studies) that link more or less tightly a gene and its protein product to a disease. However, very few experimental connections are made in a “drug discovery”-relevant way (i.e. using a small molecule) between these potential targets and the animal model. We believe that this link can be created using the tools of chemical biology or medicinal chemistry. Indeed, medium scale screening capacity (10.000 to 100.000 datapoints per campaign) has become affordable to academia. In our lab, we have access to a library of more that 35.000 compounds and we are preparing a joint purchase by CEA, Inserm, and Karolinska to reach a size of 100.000 compounds in a near future. Parallel synthesis has also become a commodity in many academic labs and has increased the productivity of chemistry, making compound optimisation possible. Under a good guidance, i.e. experienced management, and the possibility to determine the right parameters for optimisation, medicinal chemists in academia can reasonable try to produce compounds for animal testing and in the end, add significant value to the initial set of biological data. However the missing link between hit and lead in academia is the capacity to determine the PK profile of the compound and to perform reliable in vivo experiments. Indeed, as stated above, the in vivo experiment is conclusive if one can associate pharmacokinetic data to pharmacodynamic data. Moreover, a compound is worth the lives of animals if its chance to make it to the target in vivo is demonstrated by in vitro experiments predictive of the pharmacokinetic. In our lab, we have built a small PK unit. Along with potency and selectivity measurement, we can measure a number of key ADME parameters, both in vivo and in vitro that are integrated in our dataflow and used as optimisation criteria by our chemists. The main instrument of that unit is a LC-MSMS system from Varian, which we use to determine kinetic aqueous solubility, logDs, microsomal clearance in vitro, and AUC in rodents. We also have access to Caco-2 cells. With this proposal, we wish to increase the capacity of the current platform to support more projects. We also wish to have a continuous access to the animal facility of te Pasteur Institute for preliminary PK studies, in mouse or rats (genetic backgrounds used in the PD studies later on).
