A 3D Scaffold Helps Recreate the Human Bone Marrow Niche

The bone marrow serves as the main site of hematopoiesis in humans, giving rise to red and white blood cells. This space is highly organized into distinct regions, or niches, that perform different functions. One of these, the endosteal niche, lies closest to the bone and consists of bone-producing cells called osteoblasts, vascular cells, and hematopoietic stem cells.

In mice, researchers have shown that this niche is often where blood disorders can form or where tumors can metastasize.^1,2 However, the endosteal niche and its functions have not been well studied in humans.

“One of the biggest challenges of modeling the bone marrow in vitro is trying to recreate this cellular complexity that we find in the native tissue,” explained Andrés García-García, a biologist at the University of Basel. Recent bone marrow organoid models that used human induced pluripotent stem cells (hiPSCs) lacked bone scaffolding, and their microscopic size prevented the development of physiological vasculature.^3,4

To overcome these challenges, García-García and his team developed a model of the human endosteal niche that supported hematopoiesis.⁵ These findings, published in Cell Stem Cell, could help researchers explore bone marrow biology and diseases arising from this tissue.

The researchers first produced two groups of organoids from hiPSCs: one of osteoblasts and the second of vascular cells. The team differentiated the cells for the osteoblast organoids on a scaffold made up of hydroxyapatite, a component of bone. After the vascular organoids developed separately, the researchers added these to the scaffold with the osteoblast organoids. They continued to provide vascular and osteogenic growth signals for two weeks, at which point the team assessed their model.

Using fluorescence microscopy and scanning electron microscopy, the team showed that their endosteal niche model contained endothelial cells organized in vascular structures nestled among osteoblasts. This resembled the complex architecture of the true endosteal niche.

“This is joining a pretty small contingent of models that exist to do something like this and adding a really important additional layer onto it, which is having a three-dimensional, bone-like scaffold to direct the differentiation and the self-organization of the different types of organoids,” said Grace Bushnell, a cancer biologist and engineer at the University of Minnesota who was not involved with the study.

Next, to explore how vascularization influenced hematopoiesis, the team cultured hematopoietic stem and progenitor cells from cord blood on their vascularized endosteal niche model or a similar construct that lacked vascular cells. They cultured these two models in vitro or implanted them into a mouse for one week. The researchers saw that, in both settings, models that included vascular cells had a greater representation of myeloid-producing cells compared to those that only had osteoblasts. This demonstrated the vascularization is important to the hematopoietic function of the endosteal niche.

The researchers then developed two additional vascularized endosteal niche models using different hiPSC cell lines. Both models generated vascular structures, although with variability in the distribution of wide vessels and narrow capillaries. The researchers also saw that, when they introduced hematopoietic stem and progenitor cells to these different vascularized models, the niches directed cellular differentiation differently. For example, they saw altered proportions of myeloid-producing cells, monocytes, and platelet-producing cells.

While Bushnell agreed that the experiment demonstrated that the model supports the growth and differentiation of hematopoietic cells, she said, “I wonder how much that really recapitulates the self-organization of where [hematopoietic stem and progenitor cells] are present in a three-dimensional environment in the bone marrow, compared to maybe having them a little bit closer to the endosteal surface.”

To study the cellular composition of these niche models further, the researchers used single-cell RNA sequencing and clustered cells based on their expression profiles. In addition to osteogenic cells and endothelial cells, they also identified epithelial cells and several clusters of mesenchymal cells.

Within these cell clusters, the researchers identified the expression of master regulator genes. They also saw greater proportions of some cell types in their three models, which could explain the variability in hematopoietic differentiation.

“We also got a bit surprised because we saw that, in parallel to these two cell types [osteoblast and endothelial], other very important cell types that were not previously modeled in any other in vitro system, like neural cells, appear in the system associated to the blood vessels,” García-García said.

The emergence of these cell types, including macrophages and neural cells, also stood out to Bushnell. “That's really exciting in terms of pointing to the idea that you put the right things together with the right environment, and you can start to get these really cool emergent properties that obviously couldn't happen if you were taking permanently differentiated cells and trying to recapitulate that niche in the same way,” she said.

Overall, Bushnell said that the study opens many possible lines of future research, including questions about the bone marrow itself, diseases of bone marrow cells, or, relative to her work, studying how tumor cells react in this niche. “I really enjoyed this paper,” she said.