Bioactivity-driven discovery of repurposable antivirals as OSCAR inhibitors that promote cartilage protection via transcriptomic reprogramming

This study identifies the repurposed antivirals adefovir and brivudine as effective OSCAR inhibitors that promote cartilage protection and reverse pathological gene expression in osteoarthritis through a bioactivity-driven virtual screening approach.

The Gist

The Big Problem: The "Broken Joint" Mystery

Imagine your knee is like a high-end car engine. Inside, there's a special cushion (cartilage) that keeps the metal parts from grinding against each other. In Osteoarthritis (OA), this cushion starts to wear away, rust, and crumble.

Currently, doctors can only give you painkillers (like oiling the squeaky hinge) or suggest surgery (replacing the whole engine). There is no "magic spray" that actually stops the rust or rebuilds the cushion. The main reason? The cushion is isolated. It's like a fortress with no roads leading to it, so medicines taken as pills can't get inside to fix the damage.

The Villain: A "Double Agent" Protein

The researchers found a specific protein called OSCAR.

• In healthy joints: OSCAR is like a quiet librarian who barely does anything.
• In arthritic joints: OSCAR gets angry and goes on a rampage. It starts acting like a "double agent," telling the body to break down the cartilage cushion instead of protecting it. It's like a security guard who suddenly decides to start smashing the windows instead of locking the doors.

The goal? Find a way to shut this angry OSCAR down.

The Challenge: Locking a Door Without a Keyhole

Usually, to stop a protein, scientists look for a specific "keyhole" (a deep pocket in the protein's shape) and design a key (a drug) to fit inside it.

• The Problem: OSCAR doesn't have a deep keyhole. It has a flat, wide surface, like a smooth table. Trying to stick a key to a table is hard because there's nothing to grab onto. Traditional drug design failed here.

The Solution: The "Digital Detective" (sBEAR)

Instead of trying to design a key from scratch, the researchers used a clever computer program called sBEAR.

• The Analogy: Imagine you are looking for a thief, but you don't know what they look like. Instead of drawing a sketch, you look at a database of all the crimes committed in the city and ask: "Who else has committed crimes that look similar to this one?"
• How it worked: The computer scanned millions of existing drugs (mostly for viruses) to see which ones had a "chemical fingerprint" similar to things that might jam up OSCAR. It didn't need to know the shape of the protein; it just looked for patterns in how drugs behave.

The Discovery: Old Drugs, New Tricks

The computer flagged two old antiviral drugs:

1. Adefovir (ADV): Used for Hepatitis B.
2. Brivudine (BRV): Used for Shingles.

Think of these as veteran actors who used to play "virus fighters." The researchers asked, "Can these actors play a new role as 'cartilage protectors'?"

The Experiment: Putting the Drugs to the Test

The team tested these drugs in three ways:

1. The Lock-and-Check (Lab Test): They confirmed that these drugs could physically stick to the "flat table" of the OSCAR protein, blocking it from grabbing onto the cartilage. It was like putting a sticky note over the security guard's eyes so he couldn't see the windows to smash.
2. The Mouse Model (The Crash Test): They gave mice with damaged knees an injection directly into the joint (bypassing the "fortress" problem).
   • Result: The mice treated with these drugs had much healthier knees. The cartilage didn't wear away as fast, the bone didn't get deformed, and the inflammation went down. It was like pouring a "repair gel" into the engine that stopped the rust.
3. The Cellular Level (The Factory Floor): They looked at the cells under a microscope.
   • Normally, inflammation tells the cells to become "demolition crews" (breaking down cartilage).
   • With the drugs, the cells stopped being demolition crews and started acting like "construction crews" again, rebuilding the cushion.

The Star Player: Brivudine (BRV)

While both drugs worked, Brivudine was the superstar.

• It didn't just stop the destruction; it actively reprogrammed the cells.
• The Analogy: If the cell's DNA is a recipe book, inflammation had written a recipe for "Destruction Soup." Brivudine erased that page and rewrote the recipe for "Rebuilding Stew." It turned on the genes that make the cartilage strong and turned off the genes that cause pain and swelling.

The Bottom Line

This paper is a huge win for two reasons:

1. New Target: It proves that the "flat table" protein (OSCAR) can actually be targeted by drugs, opening the door for future treatments.
2. Drug Repurposing: It shows that we don't always need to invent new drugs from scratch. Sometimes, the cure for a broken knee is hiding in a box of old virus medicine, waiting to be rediscovered.

In short: The researchers used a smart computer search to find old virus drugs that can stop a specific "bad guy" protein in your joints, effectively turning your body's repair crew back on and stopping the arthritis in its tracks.

Technical Summary

1. Problem Statement

Osteoarthritis (OA) is a progressive degenerative joint disorder characterized by cartilage degradation, chronic pain, and loss of function. Currently, no clinically effective disease-modifying osteoarthritis drugs (DMOADs) exist; treatments are largely palliative.

• Therapeutic Challenge: The avascular nature of cartilage limits systemic drug delivery, necessitating chondrocyte-specific targets.
• Target Challenge: The Osteoclast-Associated Receptor (OSCAR) is upregulated in osteoarthritic chondrocytes and drives OA pathogenesis by interacting with Type II collagen. However, targeting OSCAR is difficult because it binds collagen via an extended, surface-exposed interface (in the D2 domain) rather than a deep, well-defined binding pocket. This makes traditional structure-based drug discovery (SBDD) and ligand-based approaches ineffective due to the lack of known small-molecule ligands.

2. Methodology

The authors employed a bioactivity-driven, structure-independent virtual screening approach to overcome the lack of structural data for OSCAR.

• Screening Framework (sBEAR): They utilized sBEAR (Structurally similar Bioactive compound Enrichment by Assay Repositioning). Unlike standard BEAR, sBEAR integrates chemical similarity-based enrichment. It mines large-scale bioactivity profiles (PubChem BioAssay: ~1.23 million compounds) to identify compounds that share structural features with known OSCAR ligands, even if they are not identical hits.
• Virtual Screening:
  • Query set: Known OSCAR ligands from humans and six mammalian species.
  • Process: Identified "Hit-Enriched Assays" (HEAs) where structurally similar compounds showed activity.
  • Prioritization: Ranked compounds based on aggregated enrichment scores.
• Experimental Validation Pipeline:
  1. ELISA Competition: Tested 45 purchasable candidates for their ability to inhibit OSCAR–collagen binding.
  2. Cellular Assays: Evaluated top candidates in IL-1β-stimulated chondrocytes for suppression of catabolic markers (MMP3).
  3. Overexpression Assay: Tested ability to suppress OSCAR receptor overexpression (Ad-OSCAR).
• Structural Analysis: Performed molecular docking (AutoDock Vina) and pose refinement using Protenix (AlphaFold3-based) to visualize how candidates bind to the OSCAR D2 domain.
• In Vivo Models: Used a DMM (Destabilization of the Medial Meniscus) mouse model of post-traumatic OA. Mice received intra-articular (IA) injections of lead compounds (0.5 or 2 mg/kg) weekly for 8 weeks.
• Transcriptomics: RNA-seq analysis of chondrocytes treated with lead compounds under inflammatory (IL-1β) conditions to map gene regulatory changes.

3. Key Contributions

• Novel Discovery Paradigm: Demonstrated that sBEAR can successfully identify inhibitors for "undruggable" targets lacking deep binding pockets by leveraging large-scale bioactivity data and structural similarity.
• Target Validation: Confirmed OSCAR as a therapeutically actionable target for OA, specifically showing that disrupting the OSCAR–collagen interaction halts disease progression.
• Drug Repurposing: Identified two clinically approved antiviral drugs, Adefovir (ADV) and Brivudine (BRV), as potent OSCAR inhibitors.
• Mechanistic Insight: Revealed that OSCAR inhibition not only blocks ligand binding but also induces transcriptomic reprogramming, shifting chondrocytes from a catabolic/inflammatory state to an anabolic/homeostatic state.

4. Key Results

A. Identification and Binding

• Screening Outcome: From ~1.2 million compounds, sBEAR prioritized candidates with high predicted affinity.
• Lead Compounds: Adefovir (ADV) and Brivudine (BRV) were selected as top hits.
  • They inhibited OSCAR–collagen binding in ELISA.
  • They suppressed IL-1β-induced MMP3 expression.
  • They downregulated OSCAR overexpression in chondrocytes.
• Structural Mechanism: Docking simulations showed both compounds bind to the OSCAR D2 domain, specifically occupying the shallow, surface-exposed groove where collagen normally binds.
  • ADV: Forms hydrogen bonds with His171 and Tyr177; electrostatic interactions via its phosphate moiety.
  • BRV: Engages aromatic residues via $\pi$–$\pi$ stacking and hydrogen bonding; the brominated nucleobase facilitates polarizable interactions.
  • Both act as competitive inhibitors by sterically blocking the collagen-recognition site.

B. In Vivo Efficacy (DMM Mouse Model)

• Treatment: Intra-articular administration of ADV or BRV (2 mg/kg) starting 1 week post-surgery.
• Outcomes:
  • Cartilage Protection: Significant reduction in cartilage erosion and lower OARSI scores compared to controls.
  • Bone & Synovium: Reduced osteophyte maturation, decreased subchondral bone plate (SBP) thickness, and reduced synovial inflammation.
  • Regeneration: Both compounds increased Sox9 expression (a master regulator of chondrogenesis).

C. Transcriptomic Reprogramming (Focus on BRV)

• Gene Expression: BRV effectively reversed IL-1β-induced transcriptional changes:
  • Downregulated: Matrix-degrading enzymes (MMP3, MMP9, MMP13, ADAMTS5) and inflammatory pathways (complement cascade, chemokine signaling).
  • Upregulated: Anabolic markers (ACAN, COL2A1) and chondrogenic pathways (Hedgehog, FGFR3).
• Pathway Analysis: BRV restored pathways related to cell viability, amino acid transport, and metabolic balance while suppressing stress responses (oxidative stress, NF-κB, MAPK).
• Transcription Factors: BRV increased activity of SOX9, GLI2, and E2F (pro-survival/chondrogenic) while reducing HIF1A, EPAS1, and ARNT (hypoxia/stress).
• Comparison: While ADV showed similar in vivo protection, its transcriptomic reprogramming was less pronounced than BRV, suggesting ADV may act via partially transcription-independent mechanisms.

5. Significance

• Proof of Concept for "Undruggable" Targets: This study proves that targets with extended surface interfaces (like OSCAR) can be targeted by small molecules using data-driven, bioactivity-centric screening rather than relying on crystal structures.
• New Therapeutic Class: It identifies ADV and BRV as potential DMOADs. Since these are already FDA-approved (for Hepatitis B and Herpes Zoster, respectively), they offer a rapid pathway to clinical translation via drug repurposing.
• Mechanism of Action: The work elucidates that protecting cartilage involves not just blocking a receptor but actively reprogramming the chondrocyte transcriptome to resist inflammation and promote regeneration.
• Clinical Implication: Intra-articular delivery of these repurposed antivirals could provide a viable strategy to halt or reverse osteoarthritis progression, addressing a major unmet medical need.