Col2a1 signaling disrupts blood and lymph vessels in jaw osteonecrosis

Scientists have uncovered a previously unrecognized mechanism that helps explain why a small but devastating fraction of patients who take bisphosphonate medications develop osteonecrosis of the jaw, a condition in which the jawbone dies and becomes exposed through the gum. The new study, published in Nature Communications, reveals that in bisphosphonate-related osteonecrosis of the jaw, or BRONJ, the failure of the bone is driven not primarily by the death of bone-forming cells, as long suspected, but by a profound dysregulation of the blood vessels and lymphatic vessels that normally sustain and drain the jaw tissue. Crucially, the researchers traced this vascular and lymphatic breakdown to signaling from cells expressing the gene Col2a1, a collagen gene better known for its role in cartilage, and showed that this process unfolds without the endothelial-to-mesenchymal transition, or EndoMT, that has often been invoked in vascular disease.

Bisphosphonates are among the most widely prescribed drug classes in the world. They are the mainstay of treatment for osteoporosis in postmenopausal women, for Paget’s disease of bone, and for the bone complications of certain cancers, including multiple myeloma and metastatic breast and prostate cancer. Their therapeutic effect comes from their ability to suppress osteoclasts, the specialized cells that resorb bone, thereby slowing bone turnover and reducing fracture risk. For the overwhelming majority of patients, these drugs are safe and effective. Yet since the early 2000s, oral and maxillofacial surgeons have documented a troubling complication: exposed, necrotic jawbone that fails to heal, most often after a tooth extraction or other dental procedure, in patients exposed to antiresorptive therapy. The jaw, with its exceptionally high bone turnover, rich blood supply, and constant microbial challenge through the oral cavity, appears uniquely vulnerable.

For two decades, the dominant explanation for BRONJ centered on the suppression of osteoclast activity and the resulting accumulation of microdamaged, poorly remodeled bone, along with direct toxicity of bisphosphonates to oral epithelial cells and the inhibition of angiogenesis, the formation of new blood vessels. While these mechanisms almost certainly contribute, they did not fully account for the clinical picture. Patients with BRONJ frequently present not merely with dead bone but with swollen, inflamed, poorly draining soft tissue, persistent infection, and a failure of the mucosa to close over the wound. Those features hinted that something was wrong not just with bone remodeling but with the entire vascular and lymphatic infrastructure of the affected region. The new study set out to test that idea directly, using patient tissue samples together with experimental models to characterize, cell by cell, what happens to the endothelial cells lining the blood and lymphatic vessels in necrotic jaw tissue.

The investigative approach combined single-cell RNA sequencing, which profiles gene expression in thousands of individual cells, with histological and molecular analyses of affected tissue. By isolating the endothelial compartment of the diseased jaw, the team could ask which endothelial subpopulations were present, which genes they were activating, and whether they showed signs of the transformation processes that vascular biologists have come to associate with vascular dysfunction. This resolution matters because bulk tissue analysis averages signals across many cell types and can easily miss or misattribute the critical changes. The single-cell strategy allowed the researchers to pinpoint a distinctive population of Col2a1-expressing endothelial cells that accumulated in lesions of bisphosphonate-related osteonecrosis and to characterize their behavior in detail.

The findings point to a fundamental rethinking of the disease mechanism. Endothelial-to-mesenchymal transition, in which endothelial cells lose their identity, detach from their vascular walls, and adopt migratory, matrix-producing, mesenchymal characteristics, has been implicated in fibrosis, vascular calcification, and impaired healing in numerous organs. A plausible hypothesis going into this work was that bisphosphonate exposure drives EndoMT in the jaw, stripping away the vessel lining and starving the tissue of perfusion. Instead, the researchers found that the Col2a1-positive endothelial cells in BRONJ lesions do not undergo EndoMT. They retain their endothelial identity and remain part of the vessel wall, but they are functionally deranged. The vessels they line are structurally and molecularly abnormal, and the surrounding tissue shows the consequences: inadequate blood perfusion, impaired oxygen delivery, and, critically, defective lymphatic function.

The lymphatic component of the discovery is particularly striking. Lymphatic vessels are often overlooked in studies of bone disease, yet they perform the essential tasks of draining interstitial fluid, clearing tissue debris and inflammatory mediators, and orchestrating immune cell traffic. In the confined anatomy of the jaw, where infection pressure from the oral microbiome is relentless, effective lymphatic drainage may be the difference between resolution of inflammation and chronic, escalating tissue destruction. The study found that Col2a1 signaling in the endothelial compartment is associated with lymphatic dysregulation, with drainage failure that would trap fluid, inflammatory cytokines, and bacteria-derived products in the lesion, perpetuating a vicious cycle of inflammation, ischemia, and necrosis. This offers a mechanistic account for a clinical observation that has puzzled surgeons for years: BRONJ lesions are not simply ischemic wounds, they are congested, inflamed wounds that cannot drain.

The Col2a1 connection adds another layer of biological intrigue. Type II collagen, the product of the Col2a1 gene, is the signature structural protein of hyaline cartilage. Its expression in endothelial cells within diseased jaw tissue suggests that these vessels adopt an aberrant, cartilage-like differentiation program under the conditions created by bisphosphonate exposure, altered mechanical loading, chronic inflammation, and suppressed bone turnover. The researchers demonstrated that this non-EndoMT Col2a1 signaling pathway actively drives the vascular and lymphatic dysfunction, rather than being a passive byproduct of tissue death. In experimental models, manipulating this pathway influenced the severity of the vascular phenotype, establishing it as a causal driver and not merely a biomarker.

Translating these findings into clinical practice, several implications emerge. First, the identification of a specific molecular pathway in the endothelium opens the door to targeted interventions. If Col2a1-driven signaling in endothelial cells can be modulated pharmacologically, before or after dental procedures in at-risk patients, it may become possible to prevent the vascular collapse that precipitates necrosis. Second, the emphasis on lymphatic dysfunction suggests that therapies aimed at promoting lymphangiogenesis or restoring lymphatic drainage could be valuable adjuncts to current BRONJ management, which today relies on antimicrobial rinses, antibiotics, conservative debridement, and in advanced cases, surgical resection with reconstruction. Third, the study underscores the importance of vascular status in predicting which patients are at risk, potentially motivating the development of biomarkers, perhaps detectable in oral fluids or blood, that reflect endothelial and lymphatic health in the jaw.

The research also carries a broader message for vascular biology. The finding that endothelial cells can become dysfunctional and drive disease without undergoing EndoMT challenges the tendency to view endothelial-to-mesenchymal transition as the default explanation for vascular pathology in chronic inflammatory settings. Endothelial cells, the study suggests, have multiple modes of maladaptation, and distinguishing among them is essential for choosing the right therapeutic target. Inhibiting EndoMT, for example, would do nothing for a patient whose problem is a non-EndoMT Col2a1 program in lymphatic and blood endothelium. The single-cell approach that made this distinction possible is likely to be applied to other forms of medication-related osteonecrosis, including those associated with denosumab and antiresorptive agents used in oncology, and to other jaw and periodontal diseases with a vascular component.

For patients, the practical guidance in the near term remains what it has been: maintain meticulous oral hygiene, complete any necessary invasive dental work before starting bisphosphonate therapy when feasible, and inform both physicians and dentists of all antiresorptive medication use. But the horizon is now brighter than it has been since BRONJ was first recognized. By identifying the vascular and lymphatic dysregulation mediated by non-EndoMT Col2a1 signaling as a central event in the disease, the researchers have converted a confusing clinical syndrome into a tractable molecular problem. The jaw, once thought to fail in these patients simply because its bone could not remodel, is now understood to fail because its vessels cannot sustain it, and that insight is precisely the kind that can be acted upon.

The study represents a collaborative effort spanning clinical oral surgery, vascular biology, and genomics, and its publication in Nature Communications reflects the growing recognition that medication-related osteonecrosis is a systemic vascular disease with local consequences. As the population receiving bisphosphonates continues to grow with global aging, and as these agents see expanding use in oncology, understanding and preventing BRONJ becomes an increasingly urgent public health matter. The Col2a1 pathway identified in this work now stands as both a warning sign of the disease process and, potentially, the first specific molecular target in a field that has long managed symptoms without a root cause to aim at. Follow-up studies will need to validate the pathway in larger patient cohorts and test whether intercepting Col2a1 signaling in animal models prevents lesion development, the essential next steps on the road from mechanism to medicine.