The Generative Brain

GPNMB as a therapeutic target (Aguzzi, Zurich)

GPNMB is made by a distinctive group of microglia, the brain's immune cells, that appear in prion-infected mouse brains where nerve cells are dying. The Zurich group of Adriano Aguzzi is asking whether pushing GPNMB up could protect neurons, a different strategy from lowering the prion protein. So far the evidence is mouse tissue and cell models plus a small human sample showing the protein fragment is higher in the spinal fluid of people with sporadic CJD. No drug exists and nothing has been tested in a living animal for benefit, so this is not a treatment option. Information only.

Key facts

  • Funding: the CJD Foundation awarded a 2024 research grant to Adriano Aguzzi and Davide Caredio at the University of Zurich to investigate GPNMB as a potential therapeutic target in prion diseases; the foundation page gives no results.
  • Mouse finding: in terminally sick prion-infected mice a microglial state with high Gpnmb, phagocytic gene signatures and raised lysosomal genes appears in regions with pronounced cell death.
  • Trigger: the authors report the state is induced by soluble factors released by dying cells, not by protein aggregates, and that this holds whatever the initiating insult, so it is a general response to cell loss and not prion-specific.
  • Timing and regions: Gpnmb upregulation is detectable at 30 weeks after infection, especially in the thalamus, in mice that start showing symptoms at about 24 weeks and reach terminal disease at 32 to 33 weeks; it was confirmed by western blot in thalamus, hippocampus and cerebellum but not hypothalamus or cortex, and high-Gpnmb regions also had strong PrPSc deposits (preprint stage).
  • Cell source: sorting of nuclei showed that microglia drive the rise in Gpnmb after infection (preprint stage).
  • Human tissue: GPNMB mRNA was higher in prefrontal cortex from five sporadic CJD patients than in five matched controls, a very small sample.
  • Biofluid signal: the shed soluble GPNMB fragment was higher in CSF of infected mice and of sporadic CJD patients but unchanged in serum (group sizes appear in the text as 24 versus 25 and 20 versus 10; effect sizes were not read); the authors say its value as a neurodegeneration biomarker is still debated.
  • Therapeutic idea: slides from the 2025 family conference say apoptotic bodies, not prions, trigger GPNMB upregulation and that recombinant soluble GPNMB prevented neuronal loss without lowering prion levels in cerebellar slice cultures, an ex vivo model.
  • Plan as of July 2025: raise GPNMB expression and shedding using a genome-wide arrayed CRISPR activation screen described as starting soon, then compounds against selected genes.
  • Limits: the peer-reviewed paper frames Gpnmb as a cell-state marker and reports no treatment effect; the neuroprotection claim rests on slice cultures presented at a conference and no animal efficacy or safety data are public.

Timeline

  1. 2024CJD Foundation grant to Aguzzi and Caredio for GPNMB.
  2. 2025Preprint posted on bioRxiv (date from the record identifier).
  3. 2025Zurich group presents the therapeutic idea and screen plan at the foundation family conference.
  4. 2026Nature Communications publishes the mouse paper.
Sources for this page