The research, led by Selwin G. Samuel, Hsu-Wen Tseng, Bastien Rival, and colleagues including senior author Kylie A. Alexander, builds on a growing body of evidence that inflammation is a central engine of NHO development. Clinicians have long observed retrospective associations between NHO and conditions that inflame the body—smoking, pressure ulcers, pneumonia, urinary tract infections, and polytrauma. Even more intriguingly, heterotopic ossification has been documented in patients with severe COVID-19 who had no central nervous system injury at all, hinting that a sufficiently intense systemic infection might be capable of driving ectopic bone formation on its own. The new work provides the mechanistic backbone for these clinical observations, showing in precise molecular detail how the molecular signatures of pathogens can fan the flames of abnormal bone growth.

The team’s experimental platform is a mouse model in which the spinal cord is surgically transected between the eleventh and thirteenth thoracic vertebrae, followed by a controlled injury to the hamstring muscle induced by cardiotoxin purified from snake venom. In this setting, the injured muscle, deprived of normal neural regulation and flooded with inflammatory signals, becomes fertile ground for ectopic ossification. Previous work by the same group had established a critical piece of the puzzle: lipopolysaccharide, or LPS, a molecular component of the outer membrane of gram-negative bacteria such as E. coli, worsened NHO in mice in a dose-dependent manner. That effect depended on Toll-like receptor 4, or TLR4, and its signaling adaptor TRIF, revealing for the first time that a single bacterial molecule could amplify pathological bone formation through a defined immune receptor pathway.

But LPS is just one molecular signature among many. Pathogens carry an arsenal of conserved molecular structures, collectively called pathogen-associated molecular patterns, or PAMPs, which are detected by pattern recognition receptors, or PRRs, displayed on the surfaces of and within immune and non-immune cells. When a PRR binds its ligand, it triggers downstream inflammatory cascades, notably through the NF-κB and interferon signaling pathways, releasing cytokines that orchestrate the immune response. The research team asked a deceptively simple question: if one bacterial PAMP can worsen NHO, what about the rest of the pathogen world?

The first step was to map which pattern recognition receptors are actually present in the cells that matter. Using flow cytometry to sort individual cell populations from mouse muscle—satellite cells, which regenerate muscle fibers; fibro-adipogenic progenitors, or FAPs, which are the mesenchymal cells that turn into bone in NHO; endothelial cells; and monocytes/macrophages—the researchers quantified messenger RNA for a broad panel of PRRs. The results were revealing. Macrophages expressed essentially the full complement of these sensors. More surprisingly, several receptors, including TLR3, RIG-I, STING, MDA-5, PKR, NOD1, and NOD2, were detectable in every cell type isolated from muscle. In human samples, FAPs isolated from muscles surrounding surgically excised NHO biopsies at Raymond Poincaré Hospital in France expressed TLR4, TLR6, NOD1, and several viral RNA sensors, while CD14-positive blood monocytes carried a complementary set including TLR7, TLR8, TLR9, and the fungal-sensing lectins Dectin-1, Dectin-2, and Mincle. Muscle tissue, in other words, is thoroughly wired to detect infection.

With that receptor map in hand, the team conducted a systematic in vivo screen, injecting purified PAMPs representing bacterial, viral, and fungal signatures into mice undergoing the SCI-plus-muscle-injury procedure, then measuring ectopic bone volumes by micro-computed tomography at seven and twenty-one days after surgery. The outcomes split into two camps. Agonists that activate TLR2-containing receptor complexes—the synthetic lipopeptides Pam3CSK4 and Pam2CSK4, which mimic bacterial lipoproteins, and lipoteichoic acid from gram-positive bacteria—all significantly exacerbated NHO in a dose-dependent fashion, with effects visible as early as day seven. CpG oligodeoxynucleotide ODN1668, which mimics bacterial DNA by activating TLR9, also increased NHO volumes, though its effect emerged more slowly, becoming apparent by day twenty-one.

The exceptions proved equally informative. Flagellin, the protein that forms bacterial tails and activates TLR5 as well as cytosolic NLRC4 and NAIP5 receptors, had no effect on NHO even at the highest dose mice could tolerate. The study’s title captures this nuance precisely: many, but not all, pathogen-associated molecular patterns aggravate neurogenic heterotopic ossification. This selectivity suggests that the pro-ossification signal is not simply generic inflammation, but a specific receptor-driven program that certain microbial signatures engage and others do not.

To understand the cellular machinery connecting infection to bone, the researchers turned to in vitro systems. When PAMPs were applied directly to purified FAPs, most had little effect on calcium mineralization, the laboratory readout of bone formation. But when macrophages were stimulated with PAMPs and their conditioned medium was transferred to FAP cultures, mineralization surged. The message was clear: macrophages act as intermediaries, translating the molecular language of pathogens into signals that directly reprogram muscle progenitor cells toward an osteoblastic fate. The team pinpointed two key messengers in that translation: the inflammatory cytokines interleukin-1 and oncostatin M. When human FAPs were treated with conditioned medium from PAMP-stimulated human CD14-positive monocytes, blocking the interleukin-1 receptor with an antagonist, or neutralizing oncostatin M with a specific monoclonal antibody, markedly reduced the calcium mineralization and expression of RUNX2, the master transcription factor of osteogenic differentiation. Both IL-1 and OSM had previously been implicated in NHO in the group’s earlier work; the new study cements them as convergent, druggable endpoints of PAMP-driven signaling.

The clinical implications are substantial. SCI patients are disproportionately vulnerable to infections—pneumonia and urinary tract infections are among the most common complications of paralysis, and the study notes that the animal model itself requires prophylactic antibiotics and manual bladder expression to prevent urosepsis. The findings suggest that a bacterial lung infection or a fungal urinary colonization during the critical early weeks after injury could biochemically amplify the deposition of ectopic bone in the very muscles surrounding immobilized joints. They may also explain why retrospective cohorts consistently report higher NHO prevalence in patients with infections, pressure ulcers, and tracheostomies, and why heterotopic bone appeared in intubated, severely inflamed COVID-19 patients without any central nervous system injury.

Beyond explaining risk, the work opens potential therapeutic avenues. If TLR2 signaling is a major driver, then TLR2 inhibitors—such as the compound C29 used in the mouse experiments—might blunt NHO exacerbation during infections. More immediately, the confirmation that IL-1 and oncostatin M sit at the convergence point of multiple PAMP pathways strengthens the case for targeting these cytokines. Interleukin-1 receptor antagonists are already approved drugs for autoinflammatory diseases, and neutralizing antibodies against oncostatin M are in clinical development for other inflammatory conditions, meaning repurposing trials for NHO prevention are at least conceivable. Because the only reliable preventive measure currently available is cyclooxygenase inhibition, based on retrospective evidence alone, the field is clearly in need of mechanism-based interventions.

The study also carries a caveat worth noting: the mouse experiments used female animals, in part because manual bladder expression is easier and safer in females, and infection-associated NHO risk in humans will need direct clinical validation. Still, by systematically surveying the pathogen landscape rather than a single molecule, the researchers have transformed NHO from an inexplicable complication into a process with identifiable immune triggers, intermediate cell types, and downstream cytokine effectors. For the hundreds of thousands of people worldwide living with spinal cord injury, that molecular clarity is the first step toward keeping their joints—and their independence—from being locked away by bone that should never have formed.