Hi Max, sorry for the buzzwords. Here are some more specifics. Please keep confidential. The Fission virtual screening method uses VirtualFlow 2.0 to efficiently screen a custom subset of the 68B Enamine and 48M Mcule Ultimate libraries. We leverage a 1000-node GPU cluster for accelerated HTVS using a series of progressively more stringent methods. The workflow includes QuickVina → Smina → GlideXP → openfe (free energy calculation). The limitations of docking partially depends on the specific target and whether the crystal structure accurately models the binding site. In our case, we have already de-risked the project significantly since we have demonstrated a successful in silico screen on the target (see the 2023 NatComms paper below describing the discovery of SC9). The original method was extremely limited, relying on a pharmacophore filter on 7M compounds and docking on ~600 compounds. Despite its limitation, 3/3 of the tested compounds were nM potent at the target site in recombinant protein and 1/3 was active in cells and in vivo. This provides proof of concept that the crystal structure model is predictive, and supports of the funding of this dramatically improved method (termed the Fission method). We have already conducted a 50K diverse compound pilot study, increasing the number of docked compounds by ~100x and showing promising results. Please note that the method itself is not intended to be part of the evaluation of the company since we have no plans of commercializing this method.
References:
@Max_Unfried Additional responses from the founder, Luis.
Response : Since WT mice do not die from neurodegeneration or even display significant mitochondrial dysfunction in age, they are not a suitable model to test for lifespan extension for this mechanism of action. Therefore, we rely on genetic models that introduce elements of human aging such as protein aggregation and mitochondrial damage to demonstrate a proof of concept that there could be benefit for humans in extending brain healthspan, a major limiting factor in human lifespan. The sepsis study was simply a fast model to show P110 mimetic activity in vivo and demonstrate that it does not work by simply inhibiting cytokine release. This model was not intended as a proof of concept for lifespan extension in humans. However, SC9 has shown lifespan extension in the ALS SOD1 mouse (induces ROS and mitochondrial damage), and P110 has shown this in zQ175 (HD, protein aggregation), R6/2 (HD, protein aggregation), SOD1 ALS model (mitochondrial damage), 5XFAD (AD model, protein aggregation), scrapie (prion protein aggregation), MS model (autoimmunity), anesthetic neurotoxicity (chemical damage), stroke (ischemia), and human iPCS models of HD, PD, ALS. Indeed, a single model can be contrived by inducing a specific stress (such as A-beta overexpression) and treating with an agent that blocks that specific stress (such as an A-beta antibody). In this case we have not used any contrived models and rely on the totality of evidence that P110 and SC9 display highly generalized anti-degenerative activity to a diverse range of stresses.
-
Response: See above. Additionally, yes, we are open to testing by Ora. A lifespan study in WT mice would require far more funding and is a very poor model of human brain aging. A positive result would be interesting but is unfortunately irrelevant for the intended human use. I would recommend chatting with Christin Glorioso from NeuroAge about this conundrum. The best we can do is a dog study or collect aging biomarkers once it is in human clinical trials. (edited)
-
[12:25 PM]
- Response : There are no agents that specifically protect mitochondria from inflammatory damange. The closest agent is NLRP3 inhibitors, however these work primarily by inhibiting cytokine release from innate immune cells. We are not reliant on this being fundamentally novel on all levels, however it is a novel mechanism of action that is distinct from all other strategies. (edited)
Thanks for providing details Luis. This sounds reasonable.