Introduction

Osteoprotegerin (OPG), encoded by TNFRSF11B , is a secreted glycoprotein that inhibits bone resorbing activity of osteoclasts [1]. By preventing binding of RANKL to RANK on osteoclast progenitors, OPG inhibits osteoclast differentiation and subsequent bone resorption [2,3]. Based on genetic perturbation studies in animal models, loss of OPG produces severe, early-onset osteopenia/osteoporosis with exuberant osteoclastogenesis and high bone turnover, whereas OPG gain-offunction or therapeutic RANKL blockade suppresses osteoclast formation and preserves bone mass [3-6]. In humans, biallelic loss-of-function variants of TNFRSF11B cause juvenile Paget disease, a high bone-turnover disorder characterized by bone pain, deformity, and fracture susceptibility, firmly linking OPG deficiency to pathological remodeling [7,8]. Thus, OPG is currently recognized as a central negative regulator of osteoclastogenesis in vitro and in vivo . As a key source for secreting OPG, mesenchymal progenitor cells (MPCs) give rise to osteoblasts (OB), chondrocytes, and regulate bone microenvironments through extracellular matrix (ECM) production, growthfactor presentation, and immunomodulation [9,10]. While the anti-bone resorbing role of OPG secreted from MPCs is well studied, little is known regarding autocrine or cell-intrinsic roles of OPG within MPCs themselves. To address this knowledge gap, we previously established OPG knock-out human induced pluripotent stem cells (OPG-KO hiPSCs) [11]. Using this genetic ablation cellular model, we tested the hypothesis that OPG is not merely an external inhibitor of osteoclastogenesis but also an intrinsic regulator of MPC gene networks through comparative RNA sequencing of wild-type (WT) versus OPGdeficient (OPG-KO) mesenchymal progenitors.

Materials and Methods

1. Maintenance of human induced pluripotent stem cells

The human iPSC line (hFmiPS2) was obtained from the National Stem Cell Bank of Korea and as used in this study as the WT reference line [12]. In parallel, our previously established OPG-KO-hiPSCs were maintained and expanded following the reported culture method [11]. In brief, undifferentiated hiPSCs were propagated under feeder-free conditions in mTeSR1 medium (STEMCELL Technologies) on Matrigel-coated culture dish and maintained at 37℃ in a humidified atmosphere of 5% CO₂. Cultures were passaged every 5–6 days by gentle enzymatic dissociation with Accutase (STEMCELL Technologies) or ReLeSR^TM (STEMCELL Technologies) and routinely reseeded at approximately a 1:20 split ratio.

2. Differentiation of hiPSCs into mesenchymal progenitors

For MPC differentiation, hiPSCs were enzymatically dissociated to a single-cell suspension with Accutase and seeded onto Matrigel-coated substrates in mTeSR1 supplemented with 10 μM Y-27632 (ROCK inhibitor) for 2 days. The culture medium was then changed to the STEMdiff^TM Mesenchymal Progenitor Kit (STEMCELL Technologies) medium and differentiated to MPCs according to the manufacturer’s protocol. In brief, on Day 0, cultures were transferred to STEMdiff^TM ACF Mesenchymal Induction Medium with daily medium exchanges through Day 3. On Day 4, medium was replaced with complete MesenCult^TM-ACF Plus and subsequently maintained with routine changes. Passaging was performed by treating Trypsin/ EDTA for 5 minutes at 37℃, immediately neutralized with DMEM medium with 10% fetal bovine serum (FBS), and cells were re-seeded onto ACF Cell Attachment Substrate in complete MesenCult-ACF Plus. Three weeks post-differentiation (Day 21), MPC identity was verified by flow cytometry (Navios EX, Beckman Coulter) based on characteristic morphology and expression of canonical surface markers (CD73 and CD105), with data processed in FlowJo v10.8.1.

3. RNA-sequencing

Total RNA (1 μg per sample) was extracted from WT and OPG-KO hiPSC-derived mesenchymal progenitors (WTMPC vs. OPG-KO-MPC) using TRIzol^TM Reagent (Thermo Fisher Scientific) and treated with DNase I to remove residual genomic DNA. Libraries were prepared by Macrogen using TruSeq Stranded Total RNA Library Prep Gold with Ribo-Zero rRNA depletion, controlled fragmentation, random-hexamer priming for first-strand synthesis, and subsequent amplification/ size selection. Sequencing was performed on an Illumina HiSeq 4000 in paired-end mode (2 × 100 bp). Although the study was designed for n = 3 per group, one WT-MPC RNA failed prespecified provider QC (integrity/quantity below acceptance criteria) prior to library construction and was excluded, yielding WT-MPC n = 2 and OPG-KO-MPC n = 3 for all downstream analyses. After sequencing, adapters and lowquality bases were removed with Trimmomatic v0.38 (including ILLUMINACLIP and sliding-window filtering; MINLEN = 36). Cleaned reads were aligned to GRCh38 (NCBI annotation NCBI_109.20200522) using HISAT2 v2.1.0. Alignments were processed with StringTie to obtain gene-level count matrices as well as fragments per kilobase of transcript per million mapped reads/transcripts per millions estimates. Differential expression was assessed in DESeq2 (negative-binomial Wald framework). Unless stated otherwise, genes were considered differentially expressed at |fold change (FC)| ≥ 2 with raw p ＜ 0.05. Unequal group sizes (WT-MPC n = 2; OPG-KO-MPC n = 3) were accommodated by DESeq2’s generalized linear model framework. Data structure and sample relatedness were examined by distance-matrix clustering and hierarchical heatmaps (Euclidean metric; complete linkage). For functional analysis of the differentially expressed genes (DEGs), Gene Ontology (GO) analysis was performed using g:Profiler (https:// biit.cs.ut.ee/gprofiler/orth). In addition, molecular interaction structure and pathway connectivity of the DEGs were analyzed within the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway database (http://www.genome.jp/kegg/pathway. html).

4. Quantitative real-time polymerase chain reaction

Total RNA was isolated using the easy-BLUE^TM Total RNA Extraction Kit (Intronbio, Cat. 17061). First-strand cDNA was synthesized with the PrimeScript^TM RT Reagent Kit (Takara, Cat. RR037A). Quantitative real-time polymerase chain reaction (qRT-PCR) was performed using TOPreal^TM SYBR Green qPCR PreMIX (Enzynomics, Cat. RT500M) on a StepOnePlus^TM Real-Time PCR System (Applied Biosystems). Expression levels were normalized to β-actin, and relative quantification was calculated using the 2 ^–ΔΔCt method. Primer sequences are provided in Supplementary Table 1.

5. Differentiation of hiPSC-MPCs into osteoblasts

To differentiate hiPSC-derived MPCs into OB, cells were cultured in osteogenic medium (alpha-MEM with 10% FBS supplemented with ascorbic acid, β-glycerophosphate, and dexamethasone). On 14 days after OB differentiation, the resulting cells were analyzed by qRT-PCR for osteoblast marker genes and Alizarin Red S staining.

6. Alizarin Red S staining

WT-MPC or OPG-KO-MPC-derived OBs were first rinsed with PBS and fixed in 4% (w/v) paraformaldehyde for 15 minutes at room temperature. Then, samples were incubated with Alizarin Red S solution (EMD Millipore, Cat. 2003999) for 20 minutes at room temperature with gentle agitation. Excess dye was removed, and cultures were washed repeatedly with distilled water until the background was clear.

7. Statistical analysis

Quantitative RT-PCR data are presented as mean ± standard deviation. Comparisons between two groups were performed using an unpaired, two-tailed Student’s t-test. Each condition included three independent biological replicates. Graphs and statistical analyses were generated in GraphPad Prism (v10.6.1; GraphPad), and p ＜ 0.05 was considered statistically significant.

Results

1. Differentiation of OPG-KO hiPSCs into mesenchymal progenitor cells

To investigate the molecular mechanisms and possible roles of OPG within the mesenchymal lineage, we used our previously generated CRISPR/Cas9-engineered TNFRSF11B (OPG) knockout human iPSC line (KSCBi002-B-2) [11]. This OPG-KO hiPSC line was previously confirmed to exhibit complete ablation of OPG expression and maintain pluripotency [11]. Thus, this established line serves as a validated isogenic model to explore the transcriptional mechanisms regulated by OPG deficiency. From these characterized OPG-KO hiPSCs, we successfully derived MPCs (Fig. 1A). Specifically, the differentiation process commenced with an initial four-day phase (Day 0–4) involving mesoderm induction using STEMdiff-ACF on a Matrigel-coated dish. This was followed by MPC induction and expansion (Day 4–6) using MesenCult-ACF media, culminating in MPC maturation through continued culture until Day 21 on an animal-component-free substrate. The resultant cells exhibited the typical spindle-shaped morphology (Fig. 1B). Furthermore, flow cytometric analysis confirmed high expression of canonical MPC markers CD73 and CD105 in OPG-KO MPCs at levels comparable to WT, indicating successful mesenchymal specification. Following comprehensive characterization, WT and OPG-KO MPCs were harvested and subjected to total RNA sequencing analysis to profile the transcriptional changes induced by OPG deficiency.

2. Global transcriptomic comparison of WT- and OPG-KO MPCs

Following differentiation of WT- and OPG-KO hiPSCs into MPCs, we performed total RNA sequencing to determine the transcriptional changes driven by OPG deficiency (TNFRSF11B knockout) (Fig. 2). Global transcriptome profiling revealed significant differences between WT-MPCs and OPG-KO-MPCs (Fig. 2A). Unsupervised hierarchical clustering analysis clearly separated the samples into two distinct groups based on their transcriptional profiles, indicating a robust biological difference caused by the OPG deficiency (Fig. 2A). Applying a threshold of |FC| ≥ 2 and raw p-value ≤ 0.05, we identified a total of 1,758 DEGs. Of 1,758 DEGs, 878 genes were significantly upregulated and 880 genes were significantly downregulated in OPG-KO MPCs compared to WT controls (Fig. 2B). Furthermore, a scatter plot visualizing the expression levels of all genes confirmed the overall distribution of transcripts and highlighted the significant downregulation of the knock-out gene, TNFRSF11B (OPG), in the OPG-KO-MPCs (Fig. 2C). However, several canonical mesenchymal stem cell markers (e.g., CD44 , NT5E , THY1 , and ENG ) maintained comparable expression levels between the two groups, confirming that the OPG deficiency did not compromise the fundamental identity of the MPCs (Fig. 2C).

3. Osteoprotegerin regulates extracellular matrix remodeling and cell adhesion in mesenchymal progenitor cells

To further dissect the transcriptional signature induced by OPG deletion, we performed functional enrichment analysis using GO and KEGG pathway databases (Fig. 3). The analysis of the DEGs between WT- and OPG-KO-MPCs revealed a strong and consistent enrichment across multiple categories, predominantly focusing on the ECM and cellular interaction. In the Biological Process (GO:BP) category, the top enriched terms were highly concentrated in functions related to structural organization and cell dynamics, including ‘extracellular matrix and structure organization’ and ‘cell migration’ (Fig. 3A). Similarly, the Cellular Component (GO:CC) analysis highlighted components essential for extracellular environment, with ‘collagen-containing extracellular matrix’ and ‘collagen trimers’ being the most significantly enriched terms (Fig. 3B). Furthermore. the GO Molecular Function analysis (GO:MF) provided molecular reinforcement to this focus on the ECM, where binding activities related to the ECM, such as ‘integrin binding,’ ‘heparin binding,’ and ‘collagen binding,’ dominated the top results (Fig. 3C). Finally, KEGG pathway analysis confirmed these findings by showing significant enrichment in pathways critical for cell-ECM communication, including ‘Focal adhesion,’ ‘ECM-receptor interaction,’ and ‘Cell adhesion molecules’ (Fig. 3D). Furthermore, the concurrent enrichment of downstream signaling pathways such as PI3K-Akt signaling and MAPK signaling highlights that OPG deficiency impacts major regulatory cascades governing cell behavior. To validate the functional signature derived from GO and KEGG pathway enrichment analysis, we examined the expression profile of the corresponding key genes (Fig. 4). Heatmap analysis of the top GO terms, including ‘Extracellular matrix structural constituent’ (Fig. 4A), ‘Extracellular matrix organization’ (Fig. 4B), and ‘Collagen-containing ECM’ (Fig. 4C), demonstrated a consistent and broad downregulation of key ECM components in OPG-KO-MPCs compared to WT-MPCs. Collectively, these analyses suggest that OPG deficiency in MPCs primarily impacts the regulation of the ECM and associated structural and signaling pathways, with potential implications for their ECM remodeling and cell adhesion.

4. Impairment of osteoblast differentiation and mineralization by OPG deficiency

Given the known roles of OPG in bone homeostasis and our finding regarding the regulatory roles of ECM remodeling, we next investigated whether OPG deficiency also directly impacts the osteogenic differentiation capacity of MPCs. Expression of key osteogenic differentiation marker genes (ALPL, COL1A1, and RUNX2) was significantly down regulated in OPG-KO MPCs compared to WT MPCs (Fig. 5A). We further confirmed that the reduced expression of key early OB differentiation factors in OPG-KO MPCs negatively impacted OB differentiation, as evidenced by Alizarin Red S staining. (Fig. 5B). Collectively, these findings suggest that OPG deficiency in MPCs severely compromises their ability to undergo osteoblast differentiation and subsequent matrix mineralization.

Discussion

The traditional role of OPG (encoded by TNFRSF11B ) is reported as a soluble decoy receptor that inhibits osteoclastogenesis by binding to RANKL. However, accumulating evidence underscores versatile role of OPG extending beyond bone-specific modulation, encompassing diverse physiological and pathological processes, including immunoregulation, vascular function, and cancer progression [13]. Nevertheless, the direct, cell-autonomous role of OPG within MPCs and its influence on their differentiation potential remains underexplored. In this study, we utilized hiPSC-derived OPG-knockout (OPGKO) MPCs and RNA-seq to dissect the intrinsic functions of OPG in the mesenchymal lineage. Our findings suggested that OPG deficiency alters the functional signature of MPCs. GO and KEGG pathway enrichment analyses (Fig. 3) consistently highlighted significant perturbations in terms related to the ECM, cell adhesion, and associated structural signaling pathways (e.g., focal adhesion, PI3K-Akt signaling, and MAPK signaling). In addition, the subsequent visualization of differential expressions showed a broad and consistent downregulation of numerous key ECM structural components and remodeling enzymes in OPG-KO MPCs (Fig. 4). This suggests that OPG exerts a cell-intrinsic role in maintaining the integrity and organizational status of the ECM niche. In addition, we investigated the osteogenic capacity and found that OPG deficiency severely compromised the ability of MPCs to differentiate into OB. This defect was evidenced at both the transcriptional level, through the marked attenuation of essential osteogenic markers (RUNX2, ALPL, COL1A1), and at the functional level, showing significantly impaired matrix mineralization (reduced Alizarin Red staining) (Fig. 5). Despite these insights, our study has several limitations. First, the RNA-seq analysis was performed with a relatively small sample size, and will require further validation by qRT-PCR, proteomics, and additional biological replicates. Second, all experiments were conducted in vitro using a single hiPSC background. Therefore, potential line-to-line variability remains to be clarified. Future studies using rescue experiments by OPG overexpression and in vivo implantation experiments will be important to determine how OPG-dependent ECM programs in MPCs influence osteoblast differentiation and bone formation. Collectively, our findings demonstrate a cell-intrinsic role for OPG that extends beyond its anti-resorptive role against osteoclasts, suggesting that OPG is a crucial regulator of the cellular microenvironment and osteogenic fate of MPCs.