This interdisciplinary approach has the potential to develop first- or best-in-class medicines across a number of disease areas. While we have initially focused on neurodegenerative diseases, our science has generated attractive opportunities in neuro-inflammation and immunology.
Creating better medicines through AI and genetics
Human genetic evidence is far superior to animal data in establishing the relationship between a drug target and its relevance in disease. Incorporating genetic data into a drug discovery engine doubles the likelihood that therapeutic modulation of protein targets will lead to success in the clinic and in the marketplace. By leveraging data from a well-characterized subset of patients with known mutations, we can achieve a deeper understanding of the biology of neurodegeneration across a wide range of patients.
Targets in the Neuron23 pipeline represent potential therapeutic opportunities to treat both mutation-carrying patients as well as a subset of the broader population. The mutations of interest may point to a target’s broader role in the biology of disease, laying the groundwork for efficacy in a larger cross-section of patients. In some cases, the genetics may point to targets that are relevant in patients who carry a single mutation, or a combination of mutations, and might offer a large treatment effect in those patients.
With deep expertise in functional genomics and advised by some of the field’s most accomplished genetic experts, we are mapping the impact that genetic mutations have on disease biology. Our scientists, using proprietary machine learning algorithms, are adept at connecting the dots between genes of interest and the molecular pathways that drive neurodegeneration.
Experience, speed and quality are critical success factors in drug discovery, and one of the most effective new tools for achieving these success factors is artificial intelligence. In addition to Neuron23’s deep and broad knowledge and experience in drug discovery, our Origenis-built AI platform is a multi-faceted toolkit based on an abundance of experimental data that is constantly refined by new data to enhance its predictive power. These tools allow us to:
- Design new drug molecules in silico and predict their key attributes, including activity, selectivity, drug-like properties, and optimal synthesis routes.
- Rapidly transition molecules from AI-driven design to automated synthesis, allowing us to quickly produce and screen the most promising structures for further evaluation.
- Perform in silico and early in vitro assessment of a compound’s ability to pass through the blood-brain barrier and confirm brain penetrance through rapid and efficient early-stage assays.
- Continually scan worldwide patent filings to ensure the novelty of the molecules we advance and our freedom to operate.
While AI can dramatically accelerate the early phases of small molecule drug discovery, world-class chemistry and biology are required to ensure the quality of the data and models used in the AI system and to perform the essential work of compound optimization.
The Neuron23 and Origenis scientists behind our pipeline have unique experience leveraging computational capabilities to explore the vast universe of chemical space, allowing them to design and synthesize novel and highly optimized chemical matter. As a result, our lead compounds have excellent drug-like properties but bear little similarity to competitor compounds directed against the same targets. This enhances the prospects for delivering unique molecules that have differentiable efficacy and safety and a strong intellectual property position.
Our partner in Germany, Origenis, is a world leader in the use of AI to create the drugs of the future. Their synergistic small molecule drug pipeline is target-centric, disease-centric or organ-centric, with a focus on discovering drugs with high potency and outstanding selectivity in difficult-to-reach tissues, such as the brain.
In collaboration with external specialists, we are developing novel PET (positron emission tomography) tracers to support our preclinical and clinical development programs. We are applying these tools to assess target engagement, target occupancy, and the biological impact of our therapies, both in animal models and clinical trial subjects. The aim is identify and advance molecules that provide well-balanced exposure levels in the brain and plasma. Insights from imaging studies will enable us to calculate doses that achieve optimal target saturation in the brain while limiting exposure elsewhere in the body to maximize efficacy and mitigate the risk of peripheral side effects.
A positron emission tomography (PET) scan showing decreased dopamine transporter activity in a Parkinson’s patient (right) compared to a healthy control (left).
Integrating pharmacodynamic biomarkers of target engagement and downstream pathway engagement, pharmacokinetic analysis of our drugs in a variety of biofluids (e.g. blood and CSF), and novel PET tracer data from preclinical and clinical studies will allow us to better understand our drug and optimize dose selection for increased clinical success.
Neuron23’s research is focused on the ability to distinguish molecular subtypes of Parkinson’s (as well as other diseases of interest), and we will use these biomarkers to guide patient recruitment in our clinical programs. The analyses of clinical data, biomarker data, and genetics data from patients and the integration of this information using AI should allow us to better understand the genotype-phenotype relationship. This precision medicine approach enables our clinical program in 2 ways: (1) Identifies a molecular fingerprint of disease, along with a subset of patients most likely to respond to our drug, and (2) Identifies patients in advance of our clinical studies who are candidates for our drug, accelerating our development timelines.
LRRK2 (leucine-rich repeat kinase 2) is a complex, multidomain protein found in neurons and many other tissues and cell types throughout the body. Variants in the LRRK2 gene are the most common mutations found in inherited Parkinson’s disease, and there is also emerging evidence that LRRK2 activity may play some role in the larger population of patients with non-familial Parkinson’s. Neuron23 has a highly potent and selective LRRK2 inhibitor in the late stages of preclinical development. In addition to Parkinson’s disease, LRRK2 inhibition may have therapeutic potential in inflammatory disorders such as Crohn’s disease.
Neuron23 is also developing a companion PET tracer to be used in clinical studies of both patients with mutant and wild type LRRK2.
TYK2 (tyrosine kinase 2) is an immune signaling protein with a genetically validated role in inflammatory disorders. Studies have shown that certain mutations in the TYK2 gene are protective against autoimmune disease but do not limit immune function to the point of causing immunodeficiency, which makes it an attractive therapeutic target for a range of inflammatory disorders.
Neuron23 has potent and selective TYK2 inhibitors in preclinical development. Unlike many competitor compounds, Neuron23’s TYK2 inhibitors can cross the brain-blood barrier, giving the program potential in CNS disorders where neuroinflammation contributes to disease progression, such as MS.
Additionally, Neuron23 has highly selective peripherally acting inhibitors, for the treatment of a range of immune disorders.
In our pipeline
Neuron23 has a pipeline of five preclinical, small molecule candidates, licensed as part of a drug discovery partnership with Origenis. Our lead candidate targets LRRK2 as a potential treatment for Parkinson’s Disease.
|Target||Therapeutic Area||Discovery||Preclinical||Phase 1|
|TYK2||Systemic inflammation, Neuroinflammation|