Porous hydrogel under perfusion: modeling and optimization of an in vitro model of bone defect reconstruction – HydrOs

To rebuild a damaged bone, it's not enough to create a simple scaffold. We need to recreate an environment capable of hosting living cells, nourishing them, guiding them… and convincing them to become real bone tissue. That’s the core of our approach: combining a porous material, similar to natural bone, with stem cells and optimized culture conditions.

To do this, we use a matrix made of a polysaccharide hydrogel enriched with hydroxyapatite (HAp). This mineral is naturally found in our bones and teeth. In our system, it plays a dual and crucial role: on one hand, it mimics the mineral phase of bone, giving the material strength and structure; on the other, it sends biochemical signals to the cells, encouraging them to transform into osteoblasts—the cells responsible for bone formation.

One of the keys to success lies in the porous structure of the matrix: the pores must be large enough for cells to settle, multiply, and receive nutrients—but not so large that they compromise mechanical strength. To precisely control this structure, we adjust various parameters during fabrication, such as the concentration of hydroxyapatite or the thickness of the material. The process used, called freeze-drying, is carefully monitored in real time using X-ray microtomography to optimize pore size and distribution. We then analyze the hydrated matrix using optical coherence tomography (OCT).

At the same time, we study the mechanical properties of the material to ensure it can withstand the stresses of the human body while remaining compatible with the cells. We also measure how well oxygen and glucose—two vital elements for cell survival—can diffuse through the matrix, to ensure a favorable environment for cell growth and differentiation.

These matrices are then seeded with mesenchymal stem cells (which can become bone) and endothelial cells (which form blood vessels). The entire system is cultured in a perfusion bioreactor—a device that mimics the body’s natural dynamics. Oxygen, nutrients, and mechanical signals are continuously supplied to stimulate both bone and blood vessel formation.

Finally, to test how well these bone substitutes work, we implant them in animal models with significant bone loss. These in vivo experiments allow us to assess the system’s ability to regenerate well-vascularized bone tissue and integrate with the surrounding bone.

This combined approach—smart material + living cells + controlled environment—opens the door to producing real "replacement parts" for damaged or missing bones, with strong potential for clinical translation in human medicine.

The work carried out marks an important step toward the creation of living bone “replacement parts.” In our system, stem cells and vascular cells were cultured within a porous material designed to mimic the natural environment of bone. The result: the cells spontaneously organized into small spherical clusters called spheroids, a highly promising structure for tissue regeneration.

This behavior was observed using advanced imaging techniques (micro-computed tomography, confocal microscopy, and optical coherence tomography), which showed that the matrix’s interconnected porous structure allows cells to infiltrate, group together, and form these spheroids. The shape and architecture of the pores play a key role—they directly determine the size of the cellular clusters. By tuning this architecture, we can therefore influence how the cells organize and behave.

Adding hydroxyapatite—a natural component of bone—to the matrix proved especially useful. It strengthens the material’s structure while sending a chemical signal to the cells, encouraging them to turn into bone-producing cells. It also improves the material’s mechanical strength without compromising porosity. Using a combination of experimentation and computational modeling, we simulated how oxygen and nutrients move through the material. The result: the conditions surrounding the cells remain favorable to their survival and activity, even in the deeper areas of the matrix.

We then developed a dynamic culture system using a perfusion bioreactor, where the cell-laden material is continuously supplied with nutrients and oxygen. This technique, inspired by how the body naturally works, significantly improved cell viability, proliferation, and differentiation—especially in spheroids composed of both bone stem cells and vascular cells. These co-cultured spheroids showed strong potential to produce both bone tissue and blood vessels.

We compared two types of human stem cells: BMSCs (from bone marrow) and SHEDs (from baby teeth). Both cell types survived well and contributed to vascular network formation, but only the BMSCs led to significant bone formation.

Finally, to assess the therapeutic potential of our approach, these cell-based constructs were implanted into damaged bones in rats. The results showed progressive bone regeneration and improved vascularization—especially when the cells had been preconditioned in dynamic culture. These findings confirm how physical environments influence cell behavior and demonstrate that our method can successfully create living tissue capable of repairing large bone defects.

This project has laid the groundwork for creating living “replacement parts” capable of repairing damaged bone. By combining lab-based experiments with advanced computer modeling, researchers gained new insights into how cells behave and interact within a bio-inspired material, cultured in a system that mimics the human body. This innovative, cross-disciplinary approach—merging biology, physics, and engineering—has the potential to revolutionize how we develop solutions for bone regeneration.

But the impact goes far beyond bone. The method developed here is highly adaptable: it can be applied to different types of tissues (such as cartilage, liver, or muscle), various cell types, and a wide range of biomaterials. By precisely tuning material properties (like porosity and stiffness) and culture conditions (such as oxygen levels or flow rates), it becomes possible to guide cells to regenerate specific tissues. This represents a major step toward personalized regenerative medicine.

Looking ahead, one of the main challenges will be to monitor how these materials degrade and are replaced by living tissue once implanted in the body. Advanced imaging techniques—like micro-computed tomography (µCT) or fluorescent labeling—can be used to track this process in real time and help refine material design accordingly.

Beyond medical applications, this technology holds significant promise in other sectors. In pharmaceutical development, for instance, it could enable more realistic 3D tissue models for drug screening, reducing reliance on animal testing and improving safety and efficiency. In the food industry, similar bioengineering strategies could support cultured meat production or improve bioprocessing techniques.

In short, the results of this project go well beyond bone repair. They pave the way for a new generation of cell culture technologies that are more lifelike, precise, and effective. By integrating cutting-edge modeling and imaging tools, this interdisciplinary approach could reshape biomedical research, regenerative medicine, and various industrial applications in the years to come.

Bone is a living tissue that is subjected to various stimuli. Cells receive mechanical, chemical or biophysical signals through their environment. The information translates into a capacity to adapt to any environmental modification. In the context of bone tissue engineering (BTE), this property can be exploited to produce osteogenic substitutes from Mesenchymal Stem Cells (MSCs) seeded within porous scaffolds cultured under dynamic conditions. Porous scaffold is required to accommodate cells and guide their growth and tissue regeneration in a 3D environment. Design variables for producing optimum scaffold architecture include the provision of adequate space for growth and the development of sufficient transport pathways within the porous material. To overcome nutrients transport limitations and to subject cells to optimal mechanical stresses, bioreactors are used. The production of osteogenic substitutes in a bioreactor lie on a rudimentary approach compared to those developed in process engineering, which rely on modeling and integrate the multi-scales and multi-physics aspects. It appears necessary to rationalize the running of bioreactors for bone tissue engineering. The originality of our approach lies in the development of methods to control and quantify the mass and momentum transfers within a perfusion bioreactor. Numerical simulations providing the stress and the nutrients concentration field experienced by the cells will systematically extend experiments.

At the interface of physics, mechanics and biology, HydrOs aims to produce vascularized osteogenic bone substitutes from a coculture of MSCs and endothelial cells (ECs) seeded in hydrogel scaffolds of polysaccharides supplemented, or not, with hydroxyapatite particles (HAp). They will be cultured in our experimental and numerical characterized perfusion bioreactor. Current knowledge regarding the fabrication of osteogenic porous scaffolds, the influence of its internal structuring on the formation of physiologically relevant biological structures, understanding of the influence of species transport and momentum mechanisms do not currently allow to envisage a realistic translation of the process to the clinic. Our ambition is to bring knowledge and methods in the field by carrying out the various tasks proposed in this project.

A key parameter is the understanding and control of the structuration of the scaffold porosity during hydrogel processing. In WP1, the morphological, mechanical and diffusional properties of hydrogel will be characterized. The mechanical environment and nutrients gradients will influence MSCs fate. In WP2, we will develop a digital twin of the bioreactor that will provide the nutrients concentration field and the stress field within the bioreactor in order to know the local conditions seen by the cells. For that purpose, we need to model the mechanical behavior of the swollen scaffolds, describe the interactions between the deformable scaffolds and the hydrodynamics within the bioreactor and model the transport of nutrients within the bioreactor. In WP3, a bioreactor will produce bone constructs from hydrogel of controlled porosity and mechanical properties (develop in WP1) seeded with a co-culture of MSCs and ECs then cultivated under optimized operating conditions given by the digital twin developed in WP2. Osteoblastic differentiation markers will be quantified and thanks to numerical simulation, correlate with hydrodynamics, stress field and mass transfers within the bioreactor. The potential of the bioreactor to improve a vascular network formation within spheroids will be assessed. In WP4, the potential of osteogenic cellularized bone grafts produced in the bioreactor and the evaluation of the added value of the supplementation of hydrogels with HAp for promoting osteoinduction will be assessed in vivo.