PHGDH: an upstream driver of sporadic Alzheimer's disease

In Alzheimer's disease, pathology begins with rising levels of amyloid-beta (Aβ), long before the hallmark plaques, tau tangles, and neuronal loss appear. Phosphoglycerate dehydrogenase (PHGDH) is a pivotal mediator that integrates the environmental stresses associated with sporadic Alzheimer's disease and, through a newly discovered RNA-binding function, initiates Aβ production at its source.

In mouse and human brain-organoid models, raising PHGDH induces amyloid pathology and suppressing it reverses that pathology. This mechanism was discovered by our founder and is documented in three peer-reviewed publications.

  • Cell Metabolism, 2022. PHGDH expression increases with progression of Alzheimer's disease pathology and symptoms. Read the paper.
  • Cell, 2025. Transcriptional regulation by PHGDH drives amyloid pathology in Alzheimer's disease. Read the paper.
  • PNAS, 2026. RNA-binding activity of PHGDH drives amyloid-beta production in a human brain organoid model of sporadic Alzheimer's disease. Read the paper.

Where LEY0214I acts: upstream, at the disease driver

PHGDH Disease driver LEY0214I acts here
→
p-eIF2α Phosphorylation reduced
→
BACE1 Normalized
→
Amyloid-β Lowered Where antibodies act
→
Cognitive decline Slowed

LEY0214I acts at the top of the cascade: it inhibits PHGDH's RNA-binding function, which reduces EIF2AK1-driven eIF2-alpha phosphorylation and, in turn, BACE1 translation. This lowers amyloid-beta at its source, rather than clearing plaque near the end of the cascade, the step where the approved antibodies act.

An oral therapy upstream of the antibodies

Every approved disease-modifying therapy today is an injectable anti-amyloid antibody, infused every two to four weeks, acting on plaque only after amyloid has already formed. Neurospan is taking a different path.

The gap

Approved therapies clear plaque downstream, carry ARIA risk with mandatory MRI monitoring, and require repeated infusions. Nothing on the market acts on the driver upstream of amyloid.

The target

PHGDH drives sporadic Alzheimer's disease, roughly 95% of all cases. Raising it induces amyloid pathology; suppressing it reverses it, in both mouse and human brain-organoid models.

Our position

LEY0214I is an oral small molecule with novel composition-of-matter IP and a strong freedom-to-operate position. No approved drug addresses this target.

LEY0214I: a clean, brain-penetrant drug candidate

A novel upstream target paired with a drug-like oral small molecule.

Oral, once-daily

A small-molecule tablet, designed for daily use and low-cost, scalable supply.

Brain-penetrant

Natively crosses the blood-brain barrier (brain:plasma ~0.67); not a P-gp substrate. No delivery shuttle needed.

Highly selective

Extremely selective for PHGDH, engaging its RNA-binding function rather than a shared enzymatic site.

Wide safety margin

NOAEL more than 700x the efficacious exposure in preclinical studies; negative genotoxicity (Ames and micronucleus).

What the preclinical package shows

Across 5XFAD and APP knock-in models, LEY0214I engaged the target and moved the outcomes that matter: Aβ plaque area fell 53 to 62% across the dentate gyrus, CA1 and cortex, plaque number fell about half, and spatial memory and nesting behavior improved. Markers of neuroinflammation (IBA1, GFAP) decreased, while neurons (NeuN) and synaptic markers (SYN1, PSD95) were preserved or increased. An exposure-based CSF amyloid-beta readout has been nominated as the clinical pharmacodynamic biomarker carried into Phase 1.

Findings are from discovery pharmacology (non-GLP) and CRO reports. Definitive safety margins are set by the GLP toxicology studies now in review. These are preclinical results and mechanism-based hypotheses, to be confirmed in human studies.

How LEY0214I compares to approved antibodies

Target product profile LEY0214I (preclinical) Approved anti-amyloid antibodies
Route and dosing Oral, once-daily tablet IV infusion, every 2 to 4 weeks
Point of action Upstream, before amyloid forms Clears plaque after it forms
ARIA risk and MRI monitoring None expected, no Fc clearance Elevated risk, MRI mandatory
APOE ε4 dosing restriction None expected Genotype-based cautions
Neurons and synapses Preserved in preclinical models Not the mechanism
Cost of goods and supply Low, scalable small molecule Biologic manufacturing
Combines with existing therapy Yes, distinct mechanism Not applicable
Clinical efficacy proven in humans Not yet, preclinical stage Yes, 25 to 35% slowing

AI-powered drug development

Artificial intelligence accelerated the discovery and optimization of our lead compound, from target biology to a clean, brain-penetrant candidate.

Our work was featured on the NVIDIA Technical Blog, 2025.

The team, and the network behind it

Scientific discovery paired with decades of Alzheimer's clinical experience.

Sheng Zhong, PhD

Founder and Principal Investigator

Discovered the PHGDH-amyloid mechanism. Former Professor and Director of Bioengineering and the Center for AI in Biomedicine at UC San Diego. Published in Cell Metabolism 2022, Cell 2025 and PNAS 2026.

Douglas Galasko, MD

Clinical Advisor

Board-certified neurologist and Alzheimer's clinical expert, with more than 20 completed Alzheimer's trials.

Gabriel Leger, MD

Clinical Advisor

Board-certified neurologist supporting Phase 1 design and clinical strategy.

Yuan Ann Liu, PhD

Biomarker Lead

Leads biomarker strategy for the program, including the CSF amyloid-beta readout carried into Phase 1.

Investors and partners

Neurospan is advancing LEY0214I toward IND filing and the first human study.
We would be glad to share more, and we welcome your questions and interest.

Sheng Zhong, PhD, Founder and Principal Investigator
szhong@neurospanllc.com