Cancer-associated Osteoblasts Exhibit Inhibited Mineral Deposition and Altered Gene Expression in an Engineered Bone Metastatic Niche

Introduction: : Bone is the most common site of breast cancer metastasis, with cancer cell dissemination to bone occurring even before primary tumor diagnosis [1]. Metastatic tumors alter normal bone physiology and can cause uncontrolled bone resorption (osteolysis) leading to fracture risk along with altered hematopoiesis and immune response [2]. Biochemical factors (IL-6, PTHrP, TNFα) produced by cancer cells trigger bone resorption by osteoclasts, releasing cytokines and chemokines that drive further metastatic progression in a vicious cycle. Mechanical stimulation has been reported to inhibit tumor progression and prevent bone destruction [3,4,5] and decrease tumor burden in a dose dependent manner [6]. In engineered model systems, mechanical stimulus provided protective effects against tumor-induced osteolysis and promoted osteogenic activity [7]. However, cancer cells induce an osteoblast inflammatory response [8] and disrupt osteoblast arrangement when in physical contact [9], suggesting that phenotypically modified osteoblasts – i.e. cancer-associated osteoblasts (CAO) – may disrupt bone physiology during metastasis, though this is not well understood.
The goals of this study were to identify an altered osteoblast phenotype induced by cancer paracrine signaling by 1) measuring changes in osteoblast differentiation and mineralization in the presence of cancer cells; 2) determining whether the phenotype depends on mechanical stimulation or osteoclast crosstalk; and 3) identifying transcriptomic changes in both cancer cells and bone cells that are associated with the altered phenotype.

Materials and

Methods: : We employed a segregated 3D co-culture hydrogel model to isolate the paracrine effects of cancer cells on osteoblasts. MC3T3-E1 murine osteoblastic cells (MC3) were expanded in standard conditions and embedded at 2×106 cells/mL in gelatin, nano-hydroxyapatite (nHA) hydrogels cross-linked with microbial transglutaminase. Triple negative murine mammary cancer cells (4T1) were expanded and embedded in a separate hydrogel with the same formulation. MC3 hydrogels were cultured alone (CTL; N=12) or in the same well as a 4T1 hydrogel as segregated co-cultures (N=12). In 24 additional constructs , RAW264.7 osteoclast precursors (RAW) were layered atop the MC3 hydrogels on day 21 of the experiment (N=12 RAW-CTRL, N=12 RAW+4T1). Six constructs from each group remained in static culture, while the remaining six were subjected to 0.5% strain for 30 min for 7 d starting on day 21 (CellScale Mechanoculture TX). Live-dead staining was performed at day 21 and 28 to quantify cell viability and apoptosis in all groups. FITC, DAPI, DMP1, and Ki67 staining were performed at days 7, 14, 21, and 28 to analyze cell morphology, differentiation, and proliferation. DNA, calcium, and ALP assays were performed at days 7, 21, and 28 to analyze cell number and osteogenic differentiation. Six additional MC3 and MC3+4T1 constructs were cultured for 7 d for RNA isolation and bulk RNA sequencing. Sequencing results were analyzed using STAR alignment and DeSeq2 programs.

Results, Conclusions, and Discussions:: Osteoblast mineral deposition was inhibited in the presence of cancer cells at all timepoints, regardless of mechanical loading and osteoclast presence (Fig. 1B,C). Osteoblast viability was lower when cultured with cancer cells alone (Fig. 1D). DMP1 expression was similar between groups, indicating a lack of change in osteogenic differentiation (Fig. 1I). Osteoblasts were smaller and more circular in cancer co-culture compared to monoculture (Fig. 1A,E, H). ALP expression increased on day 28 in cancer co-culture in unloaded and loaded conditions (Fig.1F,G).
RNA sequencing of osteoblasts showed clear clustering of control (osteoblast only) and cancer associated osteoblasts after 7 days (Fig. 2A). Over 6500 differentially expressed genes were found between these groups (Fig. 2C). Pathway analysis showed downregulation of several bone formation associated pathways in CAO’s, including collagen synthesis and bone mineralization signaling pathways (Fig. 1E). Gene expression analysis of cancer cells also showed clustering of cancer monoculture compared to co-culture with osteoblasts, with 807 differentially expressed genes found in cancer cells cultured with osteoblasts (Fig. 2B,D).
We identified a cancer associated phenotype in osteoblasts that is induced by soluble factors from triple negative mammary cancer cells. Notably, these cancer associated osteoblasts (CAOs) did not lay down mineral, an effect not previously demonstrated in metastatic models. Mechanical stimulation, which otherwise increased calcium deposition, did not rescue this phenotype. CAOs also exhibited a less spread morphology, compared to monoculture, though it is not clear whether this is a consequence of the inhibited mineralization or an effect of cancer secreted factors. ALP expression, which normally precedes mineralization, increased at later timepoints, suggesting that CAO’s do not secrete the appropriate matrix to support nucleation and crystallization. Transcriptomic analysis uncovered thousands of differentially regulated genes in the CAO’s, including downregulation of osteogenic pathways. The transcriptome of the cancer cells was also in the presence of osteoblasts, indicating a potential feedback or feedforward mechanism. Taken together, these results demonstrate that osteoblasts take on a cancer associated phenotype in the presence of triple negative mammary cancer cells, characterized by inhibited mineralization, even under mechanical stimulation. This phenotype may contribute to altered bone physiology in metastasis.

Acknowledgements and/or References (Optional):: [1] Braun, S et al., N Engl J Med, 342:525-533, 2000. [2] Breuer, EK et al., Cell Death & Dise, 10:1-15, 2019. [3] Yang, S et al., The FASEB Journal, 33:10742-10752, 2019. [4] Lynch, ME et al., JBMR, 28:2357-2367, 2013. [5] Wang, S et al., Bone, 153:116100-116112, 2021. [6] Fan, Y et al., Bone Research, 8:1-11, 2020. [7] Kumar Vet al., Cell Rep, 43(5):114043, 2024. [8] Kinder, M et al., Experimental Cell Research, 314:173–183, 2008. [9] Kimura Y et al. Scientific Reports, 7: 1–11, 2017.