SESSION II. DRUG DISCOVERY: FROM TEST TUBE TO ANIMAL STUDIESCHAIR: Nadine Tatton, PhD—The Association for Frontotemporal Degeneration (AFTD)
Compound Optimization after Target Identification: From HTS to LeadKurt Brunden, PhD—University of Pennsylvania
What Can We Still Learn from Studies in Animal Models of Neurodegenerative Disease?Joseph Araujo—InterVivo Solutions
Translational PK/PD in NeurodegenerationNatasha Penner, PhD—Biogen



Session Overview

Nadine Tatton, PhD - Chair, Session II

Dr. Nadine Tatton joined AFTD as Scientific Director in June 2013. She has more than 20 years’ experience as a neuroscientist in basic science and translational research, combined with technology transfer and business development expertise.

She has a personal connection to ALS, having been a family caregiver and is deeply committed to the translation of scientific discoveries into treatment opportunities that will benefit patients and caregivers alike in FTD and other neurodegenerative disorders.


Does AFTD offer scholarships or grants for PhD and postdocs?

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, for more details.

Compound Optimization after Target Identification: From HTS to Lead

Kurt Brunden, PhD—University of Pennsylvania

Kurt Brunden, PhD

Dr. Kurt Brunden is Director of Drug Discovery and Research Professor in the Center for Neurodegenerative Disease Research (CNDR) within the Perelman School of Medicine at the University of Pennsylvania, where he oversees drug discovery and translational research programs in the areas of Alzheimer’s disease (AD), frontotemporal lobar degeneration and Parkinson’s disease.

Dr. Brunden’s work is funded by NIH grants and through a number of collaborative research programs with pharmaceutical and biotechnology companies.

Prior to joining CNDR in 2007, Dr. Brunden was an executive in the biotechnology sector, where he served as VP of Research at Gliatech, Inc. and later as Sr. VP of Drug Discovery at Athersys, Inc. In these positions, he initiated and managed drug discovery and development programs in AD, cognitive enhancement, schizophrenia, inflammation, metabolic disease and cancer.

Prior to joining industry, Dr. Brunden was an NIH-funded faculty member within the Biochemistry Department at the University of Mississippi Medical Center, with a research focus on the regulation of myelination. He obtained his B.S. degree (magna cum laude) from Western Michigan University, with dual majors of Biology and Health Chemistry, and his Ph.D. in Biochemistry from Purdue University, with a post-doctoral fellowship at the Mayo Clinic.

Dr. Brunden has over 100 scientific publications, 8 issued U.S. patents with foreign equivalents, and a number of pending patent applications.



I have a question regarding the first BBB study that you recommend prior to performing the full PK. Is that done using PO or IV? Do you have a recommendation?

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.).

Can you recommend CROs for an academic lab that wants to do PK/BBB penetrance of a small molecule? Can you recommend CROs for an academic lab that wants to PK/BBB penetrance of a biologic?

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.

How should we calculate the Pka of non-ionization compounds in CNS MPO scores, please? It looks like that it wasn't reported.

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.

For non-ionization compounds, how should we calculate their Pka in CNS MPO score , please? Many thanks in advance?

KB: See 3. above.

We have developed a very potent, soluble compound. Excellent PK, great properties in vivo, where it rescue cognition in animal models (3x). We also showed reduction in the amount of plaques. However, we don’t have a direct way to measure target engagement, since we don’t know our target. What do you think is the next critical experiment? Should we test our drug in a different animal model?

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.

At which point should I do the hERG binding? I thought that it should be done during IND-enabling studies and not earlier.

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.

What do you think are the minimum, basic toxicology tests that I should on hit compounds?

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.

I’m a computational chemist at my 2nd year of PhD, and I have identified and characterized the binding site of BDNF with TrkB receptor. Based on my work, my lab has synthesized a small molecule with pretty good in vitro properties. I’d like to test this molecule in vivo as soon as possible. Being very potent in vitro, do you think it is worth to try? Or should we do some optimization first?

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.

In your experience, what are the best in vitro model to measure BBB penetration? Are they reliable and predictive of good in vivo results?

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.

What Can We Still Learn from Studies in Animal Models of Neurodegenerative Disease?

Joseph Araujo—InterVivo Solutions

Joseph Araujo

Joseph Araujo is President and CEO of Intervivo Solutions. He is focused on optimizing the translational value of InterVivo’s services to facilitate the development and approval of novel drugs. Joseph continues to be closely involved in projects utilizing, characterizing and developing the aged dog as a natural model of age-related human diseases. Joseph is a strong proponent for the use of validated natural and/or translational animal models in drug discovery.

His scientific background includes graduate training in pharmacology at the University of Toronto, more than 25 refereed publications and several invited presentations, which exemplifies his continued passion for scientific innovation and expertise in natural aged canine models of human disease. Joseph directly supports developing of local biotech talent and companies and has co-founded and held executive level positions in several Ontario-based Contract Research and Life Science companies helping to build their global presence and success.


Dear Joseph, thank you for the excellent presentation. We are now working on a neuroscience drug pipeline and have just started to think about what animal models to use so this is very timely, thanks. I have a couple of questions. What is the current FDA mandate/requirement for a high powered in vivo model? As you mentioned in this talk, there is no POC model - what can we use? In other words, can we skip the full in vivo efficacy model, and just assess the safety tox (dogs?) and then go straight to FIH studies. Another question, what do you think of the SAMP-8 model? Thank you, Mo
JA: Thank you for the question, Mo. To the best of my knowledge, there is no specific requirements from the FDA for high powered in vivo models. A case could be made for skipping in vivo efficacy models depending on the rationale of the drug program; for example, if the target does not exist in any animal model. My opinion is that non-targeted models that naturally develop age-related parallels to AD, like the aged dog, are potential POC models The SAMP-8 model is similar in many respects. I do not have any direct experience with the SAMP-8 mouse model, but there are several features of the model that parallel changes seen in sporadic AD. For a review of the model see Pallas et al., Journal of Alzheimer’s Disease, 15 (2008), 615-624.
How do you view on the Senescence Accelerated Mouse-Prone 8 (SAMP8) as a model of AD and aging? What are the pros and cons for this model?
JA: I do not have any direct experience with the SAMP-8 mouse model, but there are several features of the model that parallel changes seen in sporadic AD. For a good review of the model see Pallas et al., Journal of Alzheimer’s Disease, 15 (2008), 615-624. One positive feature is that the model is not genetically targeted and there is a cascade of neuropathological events that model early neurodegeneration. This suggests the model may be useful for investigating early stages of disease progression. One of the negative features, like all models, is that the predictive validity of the model is not well established.
Hi! Joseph, learnt lots from you, thanks! Just two small questions: 1. Aged dogs look like having less Abeta 1-42 in the CSF. Is this kind of Abeta a monomer or oligomer? Is this phenomenon in the dogs same as that in the AD patients? 2. Your data show that both Tau and phospho-Tau will increase in the CSF of aged dogs. Could you please tell me which phospho site of Tau you detected? You know, there are many p-sites in Tau protein. That slide is blurred. Thank you so much! Roy
JA: Thank you for the questions, Roy.

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.

Thank you Joseph, great presentation. We have been trying to use aged rats, but most of the times these animals were either dying or having issues and ultimately we couldn’t use them. Any thoughts? Have you heard or had similar issues? Thank you in advance.
JA: Our group has used aged rats up to 24 months, although not all rats are suitable due to health-related changes and not all survive. Some of the health related issues can be reduced or managed through changes in housing and a larger cohort is required to obtain suitable subjects.
Hi Dr. Araujo, thank you very much for your presentation. Do you think that is useful to develop new mice model for AD or other neurodegenerative diseases? In your opinion, what is the animal model that is missing?
JA: I think there is value in developing new models as a means of better understanding the disease and finding new potential targets. In my opinion, it would be great to better understand the predictive validity of current models and use gaps in that analysis to address what animal model is missing.
Hi Joseph, thanks for sharing your presentation, very informative. To save money, can we run behavior and then biochemistry studies in the same animals? Or do you think that behavioral tests can affect the level of some biomarkers? Somehow related, would you recommend to expose the same animal through a variety of behavioral and cognitive tests? Or different cohort should be used?
JA: Excellent questions. I think behavior and biochemistry studies can be done in the same animal, but it will depend on the model. For example, in dogs, we are able to look at multiple cognitive domains in the same animal and collect biofluids for biomarker analysis. The size of the model allows us to collect relatively large amounts of blood and CSF compared to rodent and murine models. We also try to time fluid collection so it doesn’t impact behavior. While behavioral tests may impact some biomarkers, there are many factors, such as time of day and vigilance state that can also affect biomarkers. Therefore, these are all important considerations in designing studies. For the last question, I think it also depends on the model and tests used. For example, if a cognitive test requires extensive training, then it may be more efficient to use different cohorts.
Thank you for your talk, Dr. Araujo. Before starting IND-enabling studies, how many different animal models should I test?
JA: Thank you for your question – I don’t think there is a right answer to this question. Ideally, one needs to balance comfort that the compound is sufficiently efficacious with cost and time considerations.
Quick question: If you could test a drug in only one AD mouse model, which one would that be? Why?
JA: Interesting question. I think it would depend on the drug and the purpose for using the model. If I knew nothing about the drug and could only evaluate in one model, then I would likely select a non-targeted mouse model such as the SAMP-8 model. My rationale is that a positive signal in a model that is not linked to a specific pathology and that develops age-related behavioral/cognitive and histopathological changes would increase likelihood of success.

Translational PK/PD in Neurodegeneration

Natasha Penner, PhD—Biogen

Natasha Penner, PhD

Dr. Natasha Penner is Director in the Clinical Pharmacology and Pharmacometrics group at Biogen, Cambridge.

Dr. Penner is a chemist by training (MS in Chemistry and PhD in Analytical Chemistry from Moscow State University, Russia). She also has MS degree in Pharmacometrics from University of Maryland. She spent 2 years as a post-doc at Purdue University then worked in drug metabolism characterization group at Schering-Plough (now Merck) and Biogen.


To be published soon.