Impaired mechanosensitivity and metabolism hinder diabetic bone healing, guiding scaffold design

Patients with type 2 diabetes have long faced a frustrating clinical reality: even when surgeons use the patient’s own healthy bone to fill a defect, the fracture site heals slowly, poorly, or sometimes not at all. For decades, the biological reason behind this stubborn healing failure remained murky, and no treatment specifically designed for diabetic bone repair existed. Now, a team of researchers in China has pieced together the mechanism in unprecedented detail and, in the same study, engineered a solution: a custom-built scaffold that corrects two fundamental cellular defects in diabetic bone. The work, published in the Journal of Translational Medicine, combines multiomics profiling, live-cell biophysics and biomaterials engineering into a single translational pipeline.

The research, led by Xintao Wang of the Department of Orthopedics at the Second Affiliated Hospital of Harbin Medical University, with co-first authors Yongqiang Mo, Linqin Tang and Jingwei Zhang, began with a deceptively simple question: what actually goes wrong inside diabetic bone tissue when it tries to regenerate? To find out, the team created a femoral condyle defect model in diabetic rats and grafted autologous bone into the injury, recreating the clinical scenario in which a patient’s own bone should, in principle, heal reliably. They then interrogated the healing tissue with proteomic sequencing and targeted energy metabolomics, two complementary techniques that together capture both the protein machinery of the cells and the small-molecule fuel economy that powers it.

The multiomics results revealed a tissue in metabolic crisis. Proteomic analysis pointed to an imbalance between two competing cell-fate programs: osteogenic differentiation, in which bone marrow stromal cells become bone-forming osteoblasts, and adipogenic differentiation, in which the same cells drift into fat-producing adipocytes. In diabetic bone, that balance had tipped away from bone. At the same time, targeted energy metabolomics uncovered glycolytic dysfunction — a breakdown in the glucose-burning pathway that aerobic, energy-hungry cells rely on. In other words, the diabetic bone environment was not merely inflamed or vascularly compromised; it was reprogrammed at the level of cellular energetics, pushing resident stem cells toward the wrong fate with the wrong fuel supply.

To pin down how this metabolic shift translated into altered cell behavior, the researchers turned their attention to bone marrow stromal cells, or BMSCs, the multipotent progenitors that supply new osteoblasts during repair. RNA sequencing of BMSCs isolated from diabetic rats revealed a striking pattern: genes and pathways governing mechanosensitivity — the ability of a cell to sense and respond to mechanical forces — were markedly downregulated. This suggested that the core defect was not simply a shortage of stem cells or growth factors, but a kind of sensory numbness: the cells could no longer “feel” the mechanical cues that normally tell them to build bone.

That hypothesis was put to a direct biophysical test. Using flow cytometry with a calcium ion probe, the team measured calcium signaling, a rapid and readout-friendly indicator of mechanical sensing, because mechanically stimulated cells typically respond with calcium influx. BMSCs from diabetic rats showed blunted calcium responses compared with healthy controls. In parallel, the researchers examined YAP nuclear localization. YAP, the key downstream effector of the Hippo pathway, functions as a mechanical transducer: when a cell experiences mechanical loading, YAP translocates into the nucleus and drives expression of osteogenic genes. In diabetic BMSCs, YAP failed to accumulate in the nucleus efficiently, confirming that the mechanosensory apparatus — from membrane-level calcium signaling to nuclear mechanotransduction — was functionally impaired. The picture that emerged was coherent: glycolytic dysfunction and reduced mechanosensitivity were not separate problems but intertwined features of the same reprogrammed state, jointly steering BMSCs away from bone formation.

Having identified the mechanism, the team then asked the translational question: could a biomaterial be designed to reverse it? Their answer came in the form of a scaffold with two engineered features, each targeted at one of the two cellular defects. The first was chemical. The researchers modified the scaffold surface with silicon ions, creating a silicon-doped calcium phosphate material, or Si-CaP. Silicon ion release, the study indicates, acts on cellular metabolism, increasing glycolytic flux in the BMSCs and thereby restoring the energy profile these cells need to commit to the osteogenic lineage. The second feature was architectural. The scaffold’s internal structure was designed using a triply periodic minimal surface, or TPMS, geometry — a mathematically defined, smoothly curved porous architecture that has become popular in bone tissue engineering because of its high surface area, controllable pore size and mechanical properties that mimic trabecular bone. The team’s rationale was that the TPMS microstructure would subject cells to favorable mechanical microenvironments, promoting YAP nuclear translocation and thereby compensating for the lost mechanosensitivity of diabetic BMSCs.

The logic of the design is notable because it treats the scaffold not as a passive filler but as an active therapeutic device. Where conventional bone graft substitutes aim to provide structural support and an osteoconductive surface, the Si-CaP TPMS scaffold is engineered to intervene in two specific intracellular programs — metabolism and mechanotransduction — that the study identified as the root causes of diabetic healing failure. The chemical component targets the fuel supply; the physical component targets the sensory apparatus. Together, they were intended to redirect BMSCs toward osteogenic differentiation, effectively re-teaching diabetic cells to behave like healthy ones in the context of a healing defect.

The final validation came in vivo. The researchers implanted the structured Si-CaP TPMS scaffold into the diabetic rat femoral condyle defect model — the same model in which autologous grafting had performed poorly — and evaluated bone regeneration over time. The results showed that the scaffold significantly promoted bone regeneration in diabetic animals, outperforming what the impaired healing environment would otherwise allow. While the study was conducted in rats, the demonstration that a rationally designed biomaterial can rescue healing in a diabetic skeletal environment represents a meaningful proof of concept for a clinical problem that currently has no targeted therapy.

The clinical significance of the work is considerable. Diabetes affects more than half a billion people worldwide, and a substantial fraction of them will require orthopedic interventions — fracture fixation, joint reconstruction, spinal fusion, or revision surgery — at some point. Delayed union and nonunion of fractures are well-documented complications in diabetic patients, and the risk extends to the implants themselves: bone ingrowth into prosthetic surfaces is compromised, raising failure rates in joint replacements. Because the mechanism identified here — metabolic reprogramming coupled with reduced mechanosensitivity in BMSCs — operates at the level of fundamental cell biology, it offers an explanation that spans many of these scenarios. It also suggests that other diabetic complications involving impaired tissue regeneration might share similar roots, since mechanosensing and metabolic state are central to the behavior of progenitor cells in many tissues.

The study’s methodological breadth deserves attention as well. Rather than starting from a single candidate pathway, the team allowed unbiased multiomics data to point toward the core mechanism, then confirmed it with independent functional assays — calcium flow cytometry and YAP localization imaging — and finally built a materials-based intervention that directly addresses the confirmed mechanism. This chain of evidence, from omics discovery to mechanistic confirmation to engineered therapeutic device, validated in the same animal model that revealed the problem in the first place, is a model of translational design. Each step tightens the causal link: the omics data generated hypotheses, the biophysical assays verified them, and the scaffold experiments tested whether correcting the mechanism corrects the outcome.

Several questions remain open before such scaffolds reach the clinic. The dose and release kinetics of silicon ions, the long-term degradation behavior of the Si-CaP material, and the optimal TPMS pore architecture for human-scale defects will all require further engineering refinement. Whether the same dual-targeting strategy — metabolic enhancement combined with mechanostimulatory geometry — works in larger animal models with diabetes-like physiology, and whether it can be combined with existing osteosynthesis hardware, remains to be shown. The authors also note that the published version is being shared early as a citable accepted manuscript, with a final Version of Record to follow, so some details may be refined in the definitive edition.

Even so, the conceptual advance is clear and likely to resonate widely in both the diabetes and biomaterials communities. Bone regeneration, the study shows, fails in type 2 diabetes not because of a single missing factor but because progenitor cells have been reprogrammed at the levels of energy metabolism and mechanical sensing simultaneously. By designing a scaffold that attacks both defects at once — silicon ion chemistry to revive glycolysis, and triply periodic minimal surface geometry to force YAP back into the nucleus — the researchers demonstrated that understanding a disease mechanism precisely enough makes it possible to engineer around it. For the millions of diabetic patients whose bones heal too slowly or not at all, that principle may ultimately matter more than any single scaffold: it establishes that diabetic bone repair is a solvable engineering problem, and it supplies the biological blueprint for solving it.