A New Therapeutic Target for MASH: The GPNMB–RYK Signaling Axis Opens New Opportunities

Metabolic-dysfunctional Associated SteatoHepatitis (MASH) is rising globally but effective treatment is limited.  It is characterized by inflammation and variable fibrosis, which can progress to cirrhosis and hepatocellular carcinoma.  Resmetirom (Rezdiffra) got FDA approval in 2024 as the first MASH-specific therapy, a liver-specific thyroid hormone receptor-beta agonist (THR-b) that activates lipid oxidation, LDL receptor expression, and autophagy of lipid-damaged mitochondria.  However, it if not very effective,  is not indicated for the more serious cases, and has significant side effects.  GLP-1/GIP agonists (semaglutide, tirzepatide) are also being heavily studied for MASH via weight loss and improved insulin sensitivity.

A new receptor target for MASH therapy was reported by Xi et al. Nature 652:703, 2026.   First, GlycoProtein Non-metastatic Melanoma Protein (GPNMB) was identified as highly overexpressed in mouse and human MASH and was a key driver of MASH.  Gene knockout in mice had reduced susceptibility to high-fat steatosis.  GPNMB is a transmembrane protein with the extracellular domain (G-ECD) being released following proteolytic cleavage. G-ECD was found to be the immediate driver of murine MASH.  Second, assuming that G-ECD acts by binding to a receptor, a library of 3300 transmembrane proteins was screened for G-ECD binding.  Three with hepatic expression were investigated for steatosis by knockdown, and Receptor tYrosine kinase (RYK) was identified as responsible for MASH.  This was shown by four methods: knockdown described above, antibody to G-ECD, specific siRNA, and a G-ECD vaccine approach.

These results underscore the use of G-ECD as a reliable, convenient non-invasive diagnostic marker for MASH, and the RTK-GPNMB axis as a tractable target for intervention.  An important distinction for drug development, it reframes GPNMB from a membrane target (suited to antibody-drug conjugates or CAR constructs, as in the oncology literature) into a secreted ligand amenable to neutralizing antibodies, decoy receptors, or knockdown strategies aimed at the source protein.  Also, a G-ECD assay for MASH severity stratifies patients and adds a pharmacodynamic readout during therapy.

Mechanism: G-ECD upregulates lipogenic programs

Downstream, G-ECD engagement of RYK activates ERK1/2 signaling, which transcriptionally upregulates two lipogenic/lipid-uptake programs: PPARγ-CD36 (driving free fatty acid uptake) and SREBP1c (driving de novo lipogenesis). This gives the pathway a satisfying mechanistic closure — a receptor-proximal kinase cascade feeding directly into the transcriptional machinery that hepatology has long implicated in steatosis progression to steatohepatitis. It also means the GPNMB-RYK axis sits upstream of, rather than parallel to, several nodes already being targeted pharmacologically.

Open questions

A few things worth watching as this moves forward:

• Redundant RYK ligands. RYK’s established role as a Wnt5a co-receptor, with potential bone/gut toxicity, raises the question of pathway crosstalk.  This would be avoided by therapies that attack  G-ECD  directly.
• Source of G-ECD shedding. Identifying the protease(s) responsible for ectodomain shedding could open a third druggable node and clarify whether shedding itself is a rate-limiting, targetable step.
• Human genetic validation. GPNMB and RYK loci alleles should be investigated in MASH/NAFLD GWAS and biobank datasets to see if human genetics independently supports the axis.  This was a problem for clinical trials with FGF21 and THR-β.

Bottom line

This is a clean, mechanistically complete target discovery paper: overexpression → causal genetic validation → receptor identification → downstream signaling → multi-modality therapeutic proof-of-concept, with a human biomarker correlation woven throughout. The GPNMB-RYK axis is now a legitimate new entry in the MASH target landscape, and the receptor identity — a Wnt co-receptor with no prior metabolic disease association — is the kind of result that should prompt a second look at other “orphan” receptors in the RYK/ROR family for analogous ectodomain-ligand relationships in liver disease.

 
 
Image credit: Portions of the figure in this article were generated using ChatGPT (OpenAI) or Google Gemini.