Criteria for chemical probes to be used in animals

Since the establishment of the Portal (1) we have provided guidelines on criteria for chemical probes that are aimed for use in cellular models and that would apply also for use in organismal models in conditions equivalent to cell culture, for example zebra fish embryos.

But for application in species such as the mouse and rat – the most commonly used in biomedical research – there are further hoops that chemical probes must jump through.

Chemical probes that meet the desired criteria for in vitro cell systems (1) may or may not be suitable for use in animal models, such as rodent species. In particular, it is important that a probe can be administered to animals so as to achieve the appropriate level of exposure that would support modulation of the desired protein target for the required duration. The subsequent phenotypic consequences can then be determined with greater confidence

On the Portal, we provide reviewers’ scores for chemical probes separately in terms of their suitability ‘in cells’ and also ‘in model organisms’ – mouse and rat being by far the most commonly used species in biomedical research. Reviewers may advise on suitable concentration ranges for using recommended chemical probes in in vitro culture models and also suggest dosing schedules for use in rodent models.

In order to aid evaluation of probes in such animal models, particularly rodents, the Chemical Probes Team has put together a set of measured parameters that are considered essential for the in vivo pharmacokinetic (PK) and pharmacodynamic (PD) characterization of chemical probes. These can be applied to guide both the choice of probe and also the rational selection of the dosing schedule of a compound for preclinical in vivo experiments (2,3).

For those not used to these terms, PK parameters describe how an administered compound such as a chemical probe moves through the body – that is, how it is absorbed, distributed, metabolized and excreted (ADME). Whereas PD parameters describe the impact that the compound/probe has on the body – for example a mechanism-based change in a protein biomarker.

It is to be stressed that these in vivo criteria apply in addition to the criteria required for use in cells in vitro. Thus chemical probes to be employed in animal models in vivo should not only be potent and selective for the protein target of interest and show evidence of modulating the target in cells, but also exhibit adequate PK/PD behaviour.

PK data should be provided to inform on the plasma and ideally the target organ exposure achieved with the compound at specified doses and dosing route(s). Protein binding data should be included to enable calculation of the unbound (free) concentration of the compound – the amount of compound not bound by plasma and tissue protein and hence available to elicit PD effects at the site of action.

Together with the measured in vitro potency of the compound, this information will typically enable estimation of the dose(s) of compound required to achieve unbound concentrations sufficient to interact with the compound’s intended molecular target for the required duration (4). The likelihood of interactions with known off-targets can be similarly quantified.

Unfortunately it is common to see publications that show the administration in animal models of small-molecule compounds that have unsuitable or undocumented PK/PD properties. This makes interpretation of the phenotypic consequence difficult or impossible to interpret.

In order to review the suitability of a probe for use in rodents we put together a list of the PK parameters for which information should be provided under the following headings:

  • Specification of the preclinical model(s) in which pharmacokinetic parameters have been measured
  • Data informing on the plasma exposure of the drug and the plasma free fraction
  • Data for compounds targeting the central nervous system

In addition, where available, inclusion of quantitative data on the modulation of specific biomarkers reporting on the activity of the intended molecular target is encouraged, as evidence of target engagement and modulation in vivo – and providing a Pharmacological Audit Trail at the measured exposures (3,5).

For details see

We welcome feedback on these guidelines, please send comments to


  1. Arrowsmith CH et al. Nat Chem Biol 2015 11: 536-541 doi: 10.1038/nchembio.1867
  2. Kleiman RJ, Ehlers MD. Sci Transl Med 2016 8: 320ps1. doi: 10.1126/scitranslmed.aac9888
  3. Rossanese O, Eccles S, Springer C, Swain A, Raynaud FI, Workman P, Kirkin V. Drug Discovery Today: Disease Models 2016 21: 23-32. doi: 10.1016/j.ddmod.2017.07.002
  4. Smith DA, Di L, Kerns EH. Nat Rev Drug Discov 2010 9: 929-39. doi: 10.1038/nrd3287
  5. Banerji U, Workman P. Semin Oncol 2016 43: 436-45. doi: 10.1053/j.seminoncol.2016.06.001

Back to News