Oncodevelopmental plasticity of the skeleton in myeloid neoplasms

This study challenges the traditional view of myelofibrosis by revealing that a single clonal driver mutation triggers a unified osteochondral injury program involving THBS1+ stromal cells, which drives both pathological bone formation and resorption across distinct skeletal lineages, thereby establishing THBS1 inhibition combined with JAK blockade as a promising therapeutic strategy to halt disease progression and restore bone homeostasis.

The Gist

The Big Picture: A Broken Construction Site

Imagine your body's bones are like a massive construction site. Normally, there is a perfect balance: Demolition Crews (cells that break down old bone) and Construction Crews (cells that build new bone) work together to keep the building strong and healthy.

In a group of blood cancers called Myeloproliferative Neoplasms (MPN), something goes wrong with the "foreman" (a mutated gene called JAK2). This bad foreman sends mixed signals to the construction site.

For a long time, doctors thought this cancer only caused the site to get too crowded with new, hard, useless bone (a condition called osteosclerosis). They thought the demolition crews had stopped working.

This paper says: "Actually, the demolition crews are still working, and they are destroying the building in specific spots, even while the construction crews are piling up useless bricks elsewhere."

It's like a construction site where the workers are frantically building a wall in the living room (making it hard and solid), while simultaneously tearing down the foundation in the basement. The result is a house that looks solid from the outside but is crumbling inside.

---

Key Discoveries Explained

1. The "Two-Face" Skeleton

The researchers looked at patients and mice and found that the cancer doesn't treat all bones the same way.

• The Long Bones (Legs/Arms): These are made from "mesoderm" (a specific embryonic origin). Here, the cancer causes Osteosclerosis. It's like pouring too much concrete. The bone gets thick and hard, but it's brittle and full of tiny cracks.
• The Jawbones (Teeth): These are made from "neural crest" (a different origin). Here, the cancer causes Osteoporosis (bone loss). The bone actually disappears, leading to loose teeth and gum disease.

The Analogy: Imagine a house where the upstairs is being filled with concrete (hard but useless), while the downstairs is being eaten away by termites. The cancer is doing both at the same time, depending on which "room" of the body it is in.

2. The "Chameleon" Workers

The most surprising discovery is about the Stromal Cells. Think of these as the "landscapers" or "architects" living inside the bone marrow.

• Normal Job: In the jaw, these architects are supposed to build bone directly (like laying bricks).
• The Cancer Effect: When the cancer strikes, these architects get confused. They stop building bone and start acting like cartilage builders (the kind of soft tissue found in your nose or ears).
• The Result: Instead of building a strong jawbone, they start building soft, useless cartilage in the wrong place. This weakens the bone and allows the "demolition crew" (osteoclasts) to eat it away.

The Analogy: It's like hiring a bricklayer to build a wall, but the bad foreman tricks them into thinking they are supposed to be making a jelly sculpture instead. The wall collapses because no bricks are being laid.

3. The "Traitor" Signal (THBS1)

The researchers found the specific signal causing this chaos. It's a protein called THBS1.

• Think of THBS1 as a loudspeaker on the construction site.
• The cancer cells turn this loudspeaker up to maximum volume.
• The loudspeaker tells the "architects" (stromal cells) to stop building bone and start building cartilage.
• It also tells the "demolition crew" (osteoclasts) to work overtime and destroy the bone.

The Analogy: The cancer is hijacking the PA system and screaming, "Stop building! Start destroying! And by the way, build jelly instead of bricks!"

4. The Solution: Turning Off the Loudspeaker

The team tested a new treatment strategy. They used a drug (called LSKL) that acts like earplugs for the THBS1 loudspeaker.

• They combined these earplugs with a standard cancer drug (a JAK inhibitor).
• The Result: When they silenced the loudspeaker:
  1. The cancer cells slowed down.
  2. The "architects" stopped making jelly and started laying bricks again.
  3. The "demolition crew" stopped destroying the bone.
  4. The bone density returned to normal in both the legs and the jaw.

The Analogy: By putting earplugs on the construction site, the workers finally heard the real instructions again. The concrete pouring stopped, the termite destruction stopped, and the house started getting repaired.

---

Why This Matters

• Old View: Doctors thought MPN was just about "too much bone." They didn't realize patients were also losing bone (fractures, loose teeth) because the "hard" bone masked the "soft" destruction happening underneath.
• New View: This paper proves that the cancer causes both hardening and softening simultaneously, but in different places.
• The Future: Targeting the THBS1 signal could be a "magic bullet" to fix the skeleton in MPN patients, stopping the fibrosis (scarring) and restoring healthy bone, rather than just treating the blood cancer symptoms.

Summary in One Sentence

This study reveals that a single blood cancer mutation tricks bone-building cells into becoming cartilage-makers and demolition crews, causing bones to harden in some places and crumble in others, but silencing a specific "traitor signal" (THBS1) can stop this chaos and heal the skeleton.

Technical Summary

1. Problem Statement

Myeloproliferative neoplasms (MPNs), particularly those driven by the JAK2V617F mutation, are traditionally characterized by bone marrow fibrosis and osteosclerosis (excessive bone formation). The prevailing paradigm suggests that osteosclerosis results from impaired osteoclast-mediated bone resorption. However, clinical observations present a paradox: while patients exhibit osteosclerosis, they also suffer from high rates of osteoporosis, pathological fractures, and periodontal disease. Current models fail to explain how a single clonal driver mutation can simultaneously induce opposing skeletal phenotypes (bone formation vs. bone destruction) and how these processes coexist within the same anatomical sites. Furthermore, the role of ontogeny (developmental origin) in skeletal susceptibility to MPN remains unexplored.

2. Methodology

The study employed a multi-modal, translational approach integrating clinical data, advanced imaging, single-cell genomics, murine models, and human organoid systems:

• Clinical Imaging & Epidemiology:
  • PET/CT & CT Analysis: Analyzed whole-body 18F-FDG PET/CT and retrospective CT scans of MPN patients to quantify Bone Mineral Density (BMD) and metabolic activity in ontogenetically distinct sites (neural crest-derived maxilla/mandible vs. mesoderm-derived long bones).
  • UK Biobank PheWAS: Conducted a phenome-wide association study (n=2,548 MPN patients) to identify associations between MPN and musculoskeletal/dental phenotypes.
• Single-Cell & Spatial Transcriptomics:
  • scRNA-seq: Profiled paired bone marrow and dental (tooth/PDL) tissues from JAK2V617F-mutant MPN patients and healthy controls.
  • Mouse Models: Used inducible JAK2V617F knock-in mice and Thrombopoietin (ThPO)-driven myelofibrosis models. Performed scRNA-seq on murine maxillae/mandibles and long bones.
  • Spatial Transcriptomics: Utilized Visium on murine tibias to map gene expression (specifically Thbs1) to anatomical regions (growth plate, metaphysis).
• In Vitro & Ex Vivo Models:
  • Human BM Organoids: Generated iPSC-derived bone marrow organoids engrafted with isogenic JAK2V617F or WT hematopoietic stem and progenitor cells (HSPCs).
  • Organ-on-a-Chip: Developed a microfluidic multi-organ system connecting BM organoids with Periodontal Ligament (PDL) fibroblast spheroids to mimic systemic circulation and niche crosstalk.
• Therapeutic Intervention:
  • Tested pharmacological inhibition of Thrombospondin-1 (THBS1) using the peptide LSKL (a TGFβ activation antagonist) alone and in combination with the JAK inhibitor Fedratinib in both organoid and murine models.

3. Key Contributions & Findings

A. Ontogeny-Dependent Skeletal Remodeling

The study reveals that MPN pathology is not uniform but depends on the developmental origin of the bone:

• Mesoderm-derived bones (Long bones): Exhibit classic osteosclerosis in the marrow cavity (proximal femur) but concurrent bone loss (trabecular thinning) in the growth plate and distal regions.
• Neural crest-derived bones (Jaws): Show disproportionate involvement. The maxilla (highly trabeculated) suffers severe periodontal bone loss and expansion of the periodontal ligament (PDL), whereas the mandible is less affected.
• Clinical Correlation: Imaging confirmed that MPN patients have increased BMD in long bones but significantly reduced BMD in the maxilla, alongside a high prevalence of periodontal disease and osteoporosis in population data.

B. The "Osteochondral Injury Program"

The authors identified a shared pathological mechanism across distinct skeletal sites:

• Lineage Plasticity: In the maxilla, neural crest-derived stromal cells (normally committed to intramembranous ossification) undergo ectopic chondrogenesis (switching to a cartilage-like state) in response to mutant hematopoietic stress.
• Convergence: This chondrogenic shift mirrors the pathological remodeling seen in the growth plates of long bones. Both sites show a suppression of osteogenic programs and an activation of chondrocyte-associated genes (e.g., Acan, Cytl1).
• Mechanism: This plasticity leads to a failure in bone formation and creates a microenvironment permissive for excessive osteoclastogenesis.

C. THBS1 as the Unifying Mediator

• Discovery: A conserved population of THBS1+ (Thrombospondin-1) stromal cells was identified as the central driver linking fibrosis to bone loss.
• Localization: THBS1 is highly expressed in ectopic chondrocytes (maxilla), fibrotic niches (long bones), and tooth fibroblasts in MPN patients.
• Signaling: THBS1 acts as a ligand for CD47 on myeloid cells, driving a pro-fibrotic and pro-osteoclastogenic signaling axis. It is upregulated by TGF-β/SMAD signaling and RUNX3.
• Systemic Activation: Circulating levels of THBS1 and CD47 are significantly elevated in MPN patients, indicating a systemic activation of this pathway.

D. Therapeutic Reversal

• Dual Targeting: While JAK inhibition (Fedratinib) alone reduced myeloproliferation, it failed to reverse fibrosis or bone defects.
• Combination Therapy: The combination of Fedratinib + LSKL (THBS1 inhibitor) was uniquely effective. It:
  1. Suppressed mutant myeloid expansion.
  2. Halted fibrosis progression.
  3. Restored bone homeostasis: It corrected both the osteosclerosis in long bones and the osteoporosis/periodontal loss in the maxilla.
  4. Reversed the ectopic chondrogenic program in stromal cells.

4. Significance

• Paradigm Shift: The paper challenges the view that osteosclerosis and osteoporosis are mutually exclusive or stage-dependent in MPN. Instead, it proposes they are spatially segregated, co-existing outcomes of a single underlying stromal injury program.
• Developmental Insight: It highlights the critical role of ontogeny (neural crest vs. mesoderm) in determining how different bones respond to hematologic malignancy, explaining the specific vulnerability of the maxilla.
• Novel Mechanism: The identification of ectopic chondrogenesis in neural crest-derived stromal cells as a driver of bone loss provides a new mechanistic link between fibrosis and skeletal degeneration.
• Therapeutic Breakthrough: The study identifies THBS1 as a "disease-modifying" target. By targeting the stromal injury response (THBS1) alongside the hematopoietic clone (JAK), the therapy restores the balance of bone remodeling, offering a potential cure for the skeletal complications of MPN that current JAK inhibitors cannot address.

In summary, this work unifies fibrosis, aberrant skeletogenesis, and bone loss into a single mechanistic framework driven by THBS1-mediated stromal plasticity, providing a robust rationale for combination therapies to restore skeletal health in myeloid neoplasms.