A targeted protein degrader shows promise against arthritis in mice
Researchers used AI-guided single-cell analysis to find an enzyme in rheumatoid-arthritis joint-lining cells, then built a targeted degrader that eased arthritis in mice — still preclinical, not a human treatment.
The 60-second version
AI-guided single-cell analysis identified OGT in rheumatoid-arthritis joint-lining cells, and an FLS-targeted PROTAC reduced arthritis severity in mice — a preclinical proof of concept, not a human treatment.
Key points
- OGT was found enriched in aggressive fibroblast-like synoviocytes from RA synovium using AI-based single-cell transcriptomics.
- In vitro, OGT drove FLS invasiveness; FLS-specific OGT knockout in mice reduced joint swelling and erosion.
- The team built a nucleolin-targeted PROTAC that degrades OGT selectively in those cells and attenuated arthritis in mice.
- All results are from cells and mice; human efficacy, safety and dosing are unproven.
Verdict. Promising research pathway with a clear mechanism, but it does not change rheumatoid arthritis treatment today.
What the study didAI found an enzyme inside joint-lining cells
Researchers applied AI-based single-cell analysis (scTransMIL plus SCimilarity) to synovial tissue data and identified OGT — an enzyme that adds O-GlcNAc sugar modifications to proteins — as highly enriched in the fibroblast-like synoviocytes (FLS) that drive rheumatoid arthritis.
The biology checkTurning OGT off reduced the aggressive cell state
In lab experiments, activating OGT made FLS more invasive; silencing it reversed that. Deleting OGT specifically in FLS of arthritic mice reduced joint swelling, synovial overgrowth, and bone and cartilage erosion. Mechanistically, OGT stabilizes the protein SAP130, which leads to silencing of the growth-suppressor gene BTG2.
The therapeutic ideaA targeted degrader instead of a blunt inhibitor
The team built a PROTAC that destroys OGT: an OGT-inhibitor piece (OSMI-1) joined to an aptamer (AS1411) that homes to nucleolin on the surface of pathogenic FLS. In mice, systemic delivery eased arthritis, and adding the TNF inhibitor etanercept showed added benefit.
| Study design | AI-guided target discovery, in vitro gain/loss-of-function, FLS-specific knockout, and PROTAC treatment in mouse arthritis models. |
|---|---|
| Evidence level | Preclinical only: human cells and mice; no human trials. |
| Main findings | OGT enriched in RA-FLS; OGT-SAP130-BTG2 axis drives FLS aggressiveness; FLS-targeted PROTAC attenuates arthritis in mice. |
| Combination | PROTAC plus etanercept showed additive benefit in mice. |
The notable part is the pipeline: AI-driven discovery of a stromal enzyme, mechanism first, then a cell-type-specific degradation strategy.
What it does not meanNot a treatment yet
There is no human efficacy, safety, or dosing data. The molecule is experimental, and the study does not change current care for rheumatoid arthritis. The honest reading is a proof of concept for targeting the stromal side of the disease.