SESSION II: DRUG DISCOVERY: FROM TEST TUBE TO ANIMAL STUDIES

NT: AFTD has been offering a Basic Science and a Translational Research postdoctoral fellowship every other year, these are 2 year programs. Unfortunately we don’t have a doctoral scholarship program yet. AFTD will be hosting an FTD Mini-symposium at SfN this October led by Dr. Fen-Biao Gao and we hope to hold our FTD Mixer pre-SfN – both are great opportunities to meet early career scientists and postdocs, many of whom are active researchers in FTD and FTD/ALS. Please contact Dr. Debra Niehoff, AFTD’s Research Manager dniehoff@theaftd.org, for more details.

KB: Our usual preliminary brain-to-plasma determination done 1 hour after dosing at 2-5 mg/kg is done via intraperitoneal injection for convenience. However, it could also be done i.v., which while more technically demanding ensures 100% compound exposure. If dosed orally, there is a possibility that low plasma and brain compound levels at 1 hour are not due to rapid metabolism, but instead to poor oral bioavailability, so we don’t usually do oral analysis until we have some sense of plasma and brain levels by i.p. (or i.v.).

KB: I would refer you to the ADDF website, where they have a Science Exchange that lists a number of vetted CRO’s for research needs, including PK.

KB: The CNS MPO as originally introduced by Wagner et al. (ACS Chem Neurosci. 2010, 1, 435-49) considers among other things, the pKa value of the most basic center in the molecule (e.g., basic nitrogens). Compounds that are largely positively charged at pH 7.4 (i.e., pKa of the associated acid being >8) are assigned a lower pKa MPO score (between 0 and 1) compared to compounds with pKa of 8 or less (these are scored as 1 for the pKa MPO score). If the compound is non-ionizable, then I think you could score the pKa MPO score as 1, as these can be equated to compounds having a pKa <<8.

KB: See 3. above.

KB: It is somewhat difficult for me to answer your questions without having greater knowledge on how this compound was identified. I presume it was first shown to be active in some in vitro screen, and if so might you be able to use the activity endpoint from the in vitro assay as a target engagement readout in vivo? In general, if the longer term desire is to license this or a related compound to a biotech or pharma for further clinical development and commercialization, there will be a strong desire by the company to have a short-term (i.e., measurable in days, not months) in vivo biomarker that can be used to help establish dosing in humans. It sounds like you already have established efficacy in an AD mouse model, and of course you could verify this by testing in another model,  but establishing a pharmacodynamic marker would certainly be useful.

KB: Formal (i.e., GLP level) hERG testing is done at the time of pre-IND studies. However, there are relatively inexpensive laboratory assays that allow an initial assessment of hERG liability in a potential clinical candidate compound prior to initiating GLP testing to gain confidence that the compound won’t encounter problems in GLP hERG testing.

KB: If by hit compounds you mean those coming directly from HTS, then it probably premature to worry about toxicology testing until more refined candidates emerge from med chem efforts. Ultimately, I would refer you to the first slide of my presentation, where I suggest that as candidates advance they undergo pharmacokinetic analysis, followed by tolerability testing in mice if brain exposure and PK looks adequate. If the ultimate objective is to try to pursue an IND filing, then more sophisticated analyses including CYP450 metabolism and inhibition profiling, hERG  testing and analysis in a commercial selectivity panel may make sense.

KB: It’s great that you were able to identify a highly potent molecule based on your molecular modeling. However, keep in mind that drug optimization is a multi-dimensional process where activity is just one dimension. Other things that need to be considered are solubility, metabolic stability, and safety/tolerance. Thus, I would suggest that you first determine if your compound has suitable PK (assuming that it has reasonable solubility), and then progress through the schema I outlined in my presentation.

As noted in my presentation, we do not use in vitro models to predict BBB penetration, but instead go directly to mice in our “probe” 1 hour PK studies. Thus, I do not have firsthand experience regarding whether, for example, PAMPA is better than MDCK assays in their predictability. However, there are likely literature reports comparing these in vitro methods.

1. The decrease in Abeta 1-42 seen in aged dogs is likely the monomer, although oligomeric amyloid has also been reported in the aged dog. Head et al., Journal of Alzheimer’s Disease 20 (2010) 637-646 indicates the decline in %Abeta 1-42 is inversely proportional to brain amyloid load which is similar to what is reported in humans.
2. The antibodies in the slide included AT8 and one targeting phospho ser 199/202.