JAX-NYSCF and GSK Partner on Human Stem Cell Models for Neurodegeneration

Drug development for Alzheimer's disease has a well-documented problem: most compounds that show promise in mouse models fail in human clinical trials. The disconnect stems partly from the models themselves - rodent neurons do not replicate the full complexity of how neurodegeneration unfolds in human brain tissue. A new five-year collaboration between The Jackson Laboratory-New York Stem Cell Foundation Collaborative (JAX-NYSCF) and pharmaceutical company GSK is built around addressing that gap directly.

The partnership, announced February 17, 2026, combines JAX-NYSCF's patient-derived induced pluripotent stem cell (iPSC) technology and large-scale automation platform with GSK's deep drug discovery and translational science infrastructure. The stated aim is to generate human cellular models that more faithfully capture the biology of neurodegenerative conditions, with a particular focus on Alzheimer's disease.

From mouse models to human neurons at scale

iPSC technology allows scientists to take cells from human patients - typically skin or blood cells - and reprogram them into a pluripotent state, then coax them to develop into specific cell types including neurons. The resulting cells carry the patient's full genetic background, including any disease-associated variants. That makes them substantially more relevant to human neurodegeneration than cells derived from genetically engineered mice.

JAX-NYSCF acquired the New York Stem Cell Foundation's research arm, which had spent years developing automated, high-throughput methods for generating and characterizing iPSC lines at scale. That automation capacity - the ability to produce large, standardized panels of patient-derived cells - is central to what the new collaboration offers GSK.

"Now we have the opportunity to extend those discoveries and accelerate therapies to patients by bridging traditional disease models with next-generation patient-derived iPSC models that more accurately capture the complexity of human biology at scale," said Lon Cardon, president and CEO of JAX.

What each partner contributes

GSK brings established expertise in neurodegeneration biology, particularly through its work in immunology and inflammation - two pathways increasingly linked to Alzheimer's pathology. The company also brings its translational science infrastructure, which includes the tools and assays needed to evaluate whether a given cellular model actually predicts drug behavior in humans.

JAX-NYSCF contributes access to relevant patient cohorts, its stem cell derivation and quality-control platform, and data science capabilities to analyze the resulting cellular phenotypes at scale.

"By working together with JAX-NYSCF, we have an opportunity to develop more predictive models and ultimately, to use that better understanding to develop potential new medicines and determine which patients are most likely to benefit from new therapies," said Chris Austin, senior vice president and global head of Research Technologies at GSK.

The underlying problem this targets

Alzheimer's disease affects approximately 55 million people globally, and the rate of clinical trial failure for Alzheimer's drugs has historically exceeded 99%. Many researchers attribute part of that failure rate to the use of animal models that do not adequately replicate the disease mechanism in humans. Late-stage amyloid-targeting therapies began showing some clinical signal only after decades of development - and even approved drugs like lecanemab and donanemab show modest cognitive benefit in early-stage patients, not broader populations.

Human iPSC-derived neurons allow researchers to test whether a drug candidate actually modifies relevant disease pathways in cells that genetically and phenotypically resemble those of the patients who would receive treatment. The models do not replicate every aspect of the disease - they cannot capture aging, systemic factors, or the full three-dimensional architecture of the brain - but they represent a more direct window into human neuronal biology than rodent models provide.

Scope and what remains to be demonstrated

The collaboration covers neurodegenerative diseases broadly, not Alzheimer's alone. Details about specific research targets, budgets, or intellectual property arrangements were not disclosed in the announcement. The five-year timeframe suggests the partners expect this to be a sustained research effort rather than a single project, but no specific clinical milestones or drug candidates were mentioned.

Whether improved cellular models actually translate into better drug candidates - and eventually approved therapies - remains to be demonstrated. The collaboration is positioned as a platform-building effort, not a late-stage drug program.