SESSION IV: I WISH I KNEW… (PRECLINICAL)

The Tau Consortium is a consortium put together by its Scientific Advisory Board (SAB), based on interest of the Principal investigator (you) and publications and the research priorities the Consortium’s SAB has deemed most important to advancing the field of Tauopathy research. The main areas being funded are Basic Discovery Biology (understanding the underlying biology of tauopathies and the creation of tools and models), Drug Discovery and Translation (discovery and development of therapeutics for tauopathies, from target discovery to end stage preclinical development) and Clinical research and trials (natural history studies, patient registries, interventional clinical trials. An important focus for the Tau Consortium is the development of Biomarkers and PET imaging agents. You can learn more about these efforts at https://tauconsortium.org/research/.

Besides the funding opportunities as described in question 1, which is by invitation only, the Rainwater Foundation runs co-funding initiatives

DM: The best way is to go to our website (https://chdifoundation.org/) and review the “Community Resources” section and reach out to the listed contact for your area of interest. You may also contact me at any time.

There is an emerging body of literature of the role of ATM in neurodegenerative diseases; see following reviews and references within.

• Burn et al. P. Cytoplasmic ATM protein kinase: an emerging therapeutic target for diabetes, cancer and neuronal degeneration. Drug Discov Today. 2011;16(7-8):332‐338.
• Fielder, Edward, von Zglinicki, Thomas, and Jurk, Diana. ‘The DNA Damage Response in Neurons: Die by Apoptosis or Survive in a Senescence-Like State?’ 1 Jan. 2017 : S107 –S131.
• Choy KR, Watters DJ. Neurodegeneration in ataxia-telangiectasia: Multiple roles of ATM kinase in cellular homeostasis. Dev Dyn. 2018;247(1):33‐46. doi:10.1002/dvdy.24522

It is a bit of a chicken and an egg problem. One of the reasons we set out to create brain penetrant selective ATM inhibitors was partly to further the understanding of the underlying biology of ATM in DNA repair in the context of Huntington’s disease. While there are other tools, siRNAs, ASOs, antibodies, CRISPR etc. a sharp small molecule chemical tool that can be delivered to the right tissue and site of action is a very useful tool to interrogate biology. Tools
like siRNA, ASO, antibodies and CRISPR have poor brain penetration. We already had indication that ATM is involved in HD biology/pathology from GWAS, SiRNA knockdown, and the cross of the ATM Knockdown mice with the HD mice which rescue several landmark phenotypes of HD in those mice, but being able to dose a small molecule that can inhibit the ATM at the site of action and follow these effects in a controlled fashion was very informative to a program (not all the data was shown given time constrains in the presentation, we have a new manuscript in preparation). We felt this was key, before further resources are deployed to generate a clinical candidate. One of the goals of CHDI is to de-risk targets by creating brain penetrating sharp probes for those targets. Chemical Biology with sharp chemical probes has been very effective in advancing therapeutics.

This is one of the most underappreciated areas of medicinal chemistry and compounds optimization and where chemists often go for big changes in potency (>100fold) and often discard the opportunities that a 10 fold-here and a 10 fold-there tweaks can offer when combined. In this program we encountered different philosophical perspectives, but at the end with a few precedents the team was convinced it was working. I would advise keeping your eyes open for these opportunities, specially, when you have a rational design strategy and the SAR is solid. However, this can only happen if your data is very tight and your assays highly reproducible, to assure that these small effects are real and not just a variability of the assay). I would not say to go only for small effects but be aware of the opportunities they present.

See literature below for a perspective of the challenges comorbidies present in AD.

Kahle-Wrobleski K, Fillit H, Kurlander J, Reed C, Belger M. Methodological challenges in assessing the impact of comorbidities on costs in Alzheimer's disease clinical trials. Eur J Health Econ. 2015;16(9):995‐1004. doi:10.1007/s10198-014-0648-7 and
https://alzheimersnewstoday.com/2018/11/14/comorbidities-common-disabling-in-alzheimers-and-dementia-patients-uk-study-finds/

Yes, I should have defined that better. By proper PK, I meant to dose compounds via a route IV/PO (oral) that allows to understand how much compound got absorbed, distributed, was metabolized and was excreted by determining all pharmacokinetics parameters, (avoid dosing compound in drinking water for example, where it is hard to understand how much compound was up taken by the animal). You should collect sufficient time point to derive a good area under
the curve and understand compounds half-life and elimination. For brain penetration, it is important to understand the free fraction in brain and free compound concentrations to assure you have obtained compound exposures that cover or are above the effective concentrations (EC50) of your compounds at eliciting its effect at the target or mechanism in cells. So if your compound has an EC50 of say 50nM, free brain concentrations of 50nM or above will be needed
to assure any effect observable in vivo can be attributed to the compound. I recommend reading this helpful review and perspective on brain PK ( (Li Di, Demystifying Brain Penetration in Central Nervous System Drug Discovery,J. Med. Chem. 2013, 56, 1, 2–12https://pubs.acs.org/doi/10.1021/jm301297f)

While it is difficult to assert an efficacy without understanding compound exposure at the target site and a pharmacodynamic effect, there are techniques that can help you deconvolute your compound mechanism of action. It is important to assess, however if your compounds is present in sufficient quantities to engage that target of mechanism so you can attribute an efficacy.

There are various chemical biology and proteomics techniques that allow you identify the target of a compound identified in a phenotypic assay:

• Activity-based protein profile (ABPP)
• Cellular thermal shift assay (CETSA)
• Compound-immobilized beads
• Photoaffinity labeling (PAL

Here are some good reviews that describe these techniques:

https://www.nature.com/articles/nrd2410/
https://pubmed.ncbi.nlm.nih.gov/30392561/
https://linkinghub.elsevier.com/retrieve/pii/S1570963918301249

There are several CROs that offer target deconvolution services.

• https://www.hybrigenics-services.com/contents/ychemh-our-target-deconvolution-method
• https://biognosys.com/services/target-deconvolution-limited-proteolysis
• https://www.eurofinsdiscoveryservices.com/cms/cms-content/services/phenotypic-assays/biomap

Directly related to the case study example covered in the talk, I would recommend Roy et al., (2019), Journal of Medicinal Chemistry, 62:5298-5311, and the references cited therein.

Yes. Based on documented biological rationale in the peer reviewed literature, the most proximal alternative clinical indications would be tauopathies (Roy et al., 2019, JMC 62:5298) and certain Autism Spectrum Disorders (Robson et al., 2018, PNAS, 23,115:E10245). The rationale is based on pathophysiology progression mechanisms in the clinical diseases and the evidence of MW150 target engagement and efficacy in relevant preclinical animal models. MW150 is essentially a phase 2 ready clinical therapeutic candidate at this time, so the preclinical GxP data package and phase 1 clinical trial suggests these if funding becomes available. Because the preclinical development was focused on common pathophysiology progression mechanisms, other diseases where those mechanisms are contributors to disease susceptibility or progression are reasonable considerations for the future.

Once the basic research on CNS protein kinases identified how to circumvent the challenges in the field, it took a bit over five years to take a lead compound to a candidate ready for start of phase 1a clinical trial by the spin-out biotech. If we had made better use of complementary foundation and NIH funding to minimize the cycle of submission/peer review/award date time gaps, we most likely would have removed 18 months or more from the timeline.

Even if one has clear go-no-go decision points in advance, the larger challenge is taking a new molecule into disease areas where there are no disease modifying therapeutic precedents. Therefore, a less linear path to the goal is the starting point. A number of alleys must be pursued in try/fail probing – a bit like mapping an unknown territory. The overall Go/NoGo decision is based on a set of activity outcomes from multi-disciplinary studies, not a single test or milestone. Once an experimental protocol in a particular discipline (med chem; pharmacology; animal models) is established and results are being produced readily, the temptation is to continue generating more data in that disciplinary silo in the absence of strategic input from other disciplines. How far down an exploratory alleyway one goes before hitting a dead end, or “wall” as it called in med chem, is the tough decision. For example, medicinal chemistry refinement can hit experimental “sweet spots” where diversification of a precursor allows rapid generation of an entire family of analogs to test in a facile and quantitative primary assay. There is a temptation to fall in love with an emerging beautiful SAR series if one stays only within that disciplinary silo. The key activity is to pick landmark structures at strategic points to test in selected secondary pharmacology assays. The secondary pharmacology outcomes can support a mid-campaign decision to stop or shelve the seductive SAR path that is falling into place. One would then re-visit what alternative chemical space to explore that is a better fit with the campaign’s long term goal. This is where a recursive platform is especially useful as one gains greater knowledge in each round of passage with less expenditure of time and effort. Using the case study presented and examples in the Roy et al. (2019) report, the sole pursuit of higher affinity (lower IC50 values) for the primary target in the absence of selected secondary pharmacology screens would have delivered a different series of drug candidates. One set would have been higher affinity inhibitors with off-target activity in the kinome and non-kinome target space. Another set would have been high affinity inhibitors in an in vitro assay, but the deliverable would have been a ligand with metabolic liabilities or less attractive bioavailability. The adage is to focus on drug candidates, not beautiful ligands. For example, the addition to every compound’s evaluation at a critical point was liver microsome stability and screens for a troublesome cross-over kinase target (Roy et al., 2019). The addition of key secondary pharmacology in vitro screens at strategic end points in the campaign yielded a candidate with higher potential for selectivity that was partially de-risked for pharmacological properties. Addition of focused pilot pharmacodynamic and in vivo pharmacokinetic screens at a later step moved the campaign towards it primary goal, a de-risked drug candidate to take into phase 1 clinical development. The recursive platform using strategic secondary pharmacology inputs decreased the amount of time and effort spent on non-productive synthetic chemistry and later stage in vivo animal studies. The overall Go/NoGo decision points based on well-considered milestones were not changed throughout this process.

There are two aspects: the literature and the intellectual/physical infrastructure of the inter-disciplinary team. For the former, the prevailing view in the literature at the start of our campaign contributed to the decision to bias our primary efforts away from amyloid pathway effects and add increased early testing of safety pharmacology and toxicology de-risking. In this way, we had higher probability to have a drug candidate delivered to first-in-human studies that was both a potential alternative approach as well as a complementary approach amenable to a clinical multi-drug environment for a complex disease. In terms of the intellectual and physical infrastructure of the inter-disciplinary team, the presence of a disease focused structural genomics infrastructure and the availability of qualified assays to probe rapidly neuroinflammation and synaptic dysfunction pathophysiology mechanisms were key infrastructure strengths among team members. The prior experience with azine synthetic chemistry and pharmacology was also leveraged in order to generate initial structures with pharmacological potential for early testing and design of a preclinical drug development campaign.

The most difficult challenge, for me personally, in assisting Neurokine Therapeutics (NKT) get their start was the process of getting the campaign out of academia and into the real world where viable products are generated in a regulated industry. The most rewarding thing so far is the cooperative efforts of NKT members and academic colleagues to get the clinical asset into phase 1 first-in-human safety studies and to generate a commercial scale GMP clinical drug supply for start of first-in-patient trials when funding becomes available. The time line using a virtual spin-out and peer-reviewed funding is remarkable to me.