Ось miR4261/PAMM в множестве миэломы стимулирует ресорбцию костей.

Introduction

Multiple myeloma (MM) is the second most common hematological malignancy worldwide, characterized by the abnormal proliferation of clonal plasma cells, with approximately 35,000 new cases and 12,000 deaths annually 1 .Most patients are diagnosed through symptoms of hypercalcemia, renal insufficiency, anemia, and bone disease (CRAB). Notably, MBD affects up to 80% of MM patients 2 , 3 , with its pathological features manifesting as enhanced osteoclast activity and suppressed osteoblast function, collectively resulting in systemic osteolytic lesions 4 . Current anti-resorptive therapies, such as bisphosphonates (e.g., zoledronic acid) and RANKL inhibitors (e.g., denosumab), reduce skeletal-related events but have significant limitations 5 – 7 . More importantly, these therapies fail to restore bone homeostasis, leaving patients at persistent risk of fractures.

Recent studies have revealed that tumor-derived exosomes play a crucial role in intercellular communication within the MM bone microenvironment 8 , 9 . Previous findings demonstrated that miR-4261 was present in exsomes and its expression was markedly increased in the peripheral blood circulation of patients with MM, and confirmed that MM-secreted exosomes carry the myeloma cell marker CD138 10 .This study focuses on the mechanistic role of exosome-derived miR-4261 in MM patients, particularly those with myeloma bone disease (MBD). We found that exosomes secreted by myeloma cells deliver miR-4261 to osteoclast precursor cells, where it targets and inhibits PAMM, a protein activated by M-CSF 11 , 13 .Additionally, a 2010 study demonstrated that in stably transfected RAW264.7 cells, PAMM inhibits RANKL-induced activation of p100 NF-κB and c-Jun, as well as osteoclastogenesis, revealing PAMM’s regulatory role in osteoclast differentiation in vitro 11 . A 2025 publication indicated that enhanced oxidative stress in patients with steroid-induced femoral head osteonecrosis reduces PRXL2A (PAMM) expression, which may promote osteoclast differentiation and accelerate local femoral head bone destruction 12 . These studies underscore the importance of PAMM in bone destruction, supporting the feasibility and scientific rationale of our research.Mechanistically, the suppression of genes that negatively regulate osteoclast differentiation ultimately leads to pathological overactivation of osteoclasts and the onset of MBD. These findings suggest that dysregulation of the miR-4261/PAMM axis may contribute to the pathogenesis of MBD.

In the present study, we found that exosomes secreted by myeloma cells delivered miR-4261 to osteoclast precursor cells, and miR-4261 promoted osteoclast differentiation and bone resorption. To gain molecular insights into miR-4261 induced osteoclast differentiation, systematic studies were performed and PAMM, a protein activated by M-CSF (11) was identified as a bona fide target of miR-4261. The clinical relevance of the novel miR-4261/PAMM axis was explored as well. Overall, this study reveals a new role of exsomes-derived miR-4261 in MM patients and provides a new clue to improve the management of MM patients.

Results

miR-4261 secreted by myeloma cells promotes THP-1 cell differentiation into osteoclasts and enhances bone resorption activity

Previous studies by our research group demonstrated that MM patients exhibited significantly higher expression of miR-4261 in peripheral blood exosomes compared to healthy controls ( P < 0.05). Exosomes were isolated by ultracentrifugation, and transmission electron microscopy observations revealed an average exosome diameter of 132.30 ± 16.52 nm. Furthermore, Western blot analysis confirmed the expression of exosomal membrane markers CD9, CD63, and CD81. Notably, exosomes isolated from MM peripheral blood also expressed the clonal plasma cell marker CD138, indicating that exosomes are secreted from MM cell membranes 10 . Based on this, we hypothesized that exosomes secreted by myeloma cells in MM patients may deliver miR-4261 to osteoclasts. First, we evaluated miR-4261 expression in THP-1 cells with co-culture systems. The results showed a significantly increased miR-4261 expression in THP-1 cells as the co-culture duration extended from day 0 to day 7 (Fig. 1 , A-C). To verify whether miR-4261 is transferred from MM cells to THP-1 cells via exosomes, we treated the Transwell co-culture system with the exosome biogenesis inhibitor GW4869 (4 µM) for 7 days. qPCR analysis showed no statistically significant difference in miR-4261 expression between the GW4869-treated co-culture group and THP-1 cells cultured alone. However, miR-4261 levels were markedly decreased in the GW4869-treated group compared with the untreated co-culture group, with a statistically significant difference. These results indicate that the transfer of miR-4261 to THP-1 cells is mediated by exosomes (Fig. 1 ,D–F). Next, we collected bone marrow mononuclear cells from 27 MBD patients and 11 MM patients with non-myeloma bone disease (N-MBD) at the Second Hospital of Shanxi Medical University between January and July 2024. RT–qPCR analysis showed that miR-4261 expression was significantly higher in the MBD group than in the N-MBD group, as determined by the Mann–Whitney U test (Fig. 1 , G). Then miR-4261 mimics were delivered to THP-1 cells and the differentiation of these cells to osteoclasts upon M-CSF and RANKL induction was assessed. After seven-day differentiation induction, TRAP staining showed that the miR-4261 mimics group exhibited multinucleated giant cell-like morphology earlier (Fig. 1 , H), with a significantly higher osteoclast count compared with the control group (Fig. 1 , I). Additionally, bone resorption induced by miR-4261 mimics transduced cells and their control was assessed using micro-CT on bovine bone slices. In micro-CT scans of bovine bone slices, Ctrl samples showed a smooth, dense surface with uniform intensity and no resorption lacunae, while mimics samples exhibited a rough surface with dense lacunae, lower intensity, and large low-density areas—indicating markedly enhanced osteoclastic resorption (Fig. 1 , J). BV/TV and Tb.Th (avg) showed a decreasing trend in the mimics group, but did not reach statistical significance compared to the control group (Fig. 1 , K and L). These results indicate that the miR-4261 mimics overexpression group did not significantly alter bone volume fraction and trabecular thickness at the 21-day detection. At 21 days, the data showed that BMC and TMC were significantly reduced in the miR-4261 mimics group compared with control group (Fig. 1 , M and N). This reduction in BMC and TMC reflects bone destruction, mineral loss, and enhanced bone resorption. The reduction in bone slice mineral content is likely due to calcium ion release. Clinically, this aligns with the frequent occurrence of hypercalcemia in MBD patients, which is widely considered a consequence of bone destruction-induced calcium mobilization into the circulation. Previous study by our research group demonstrated that MM.

Fig. 1 miR-4261 delivered by exsomes from MM cells promotes osteoclasts differentiation and causes bone resorption. ( A – C ) In co-culture systems of OPM2/THP-1, RPMI 8226/THP-1, and NCI-H929/THP-1, THP-1 cells were subjected to RNA extraction, and miR-4261 expression over the duration of co-culture (up to 7 days) was quantified by RT‒qPCR. ( D – F ) Relative expression of miR-4261 in THP-1 cells under various culture conditions, compared with untreated THP-1 cells. Conditions include: THP-1 cells cultured alone, co-culture of myeloma cells with THP-1 cells, and co-culture of myeloma cells with THP-1 cells in the presence of 4 µM exosome inhibitor GW4869. ( G ) The expression of miR-4261 in myeloma bone disease (MBD) roup ( n = 27) compared with the non-myeloma bone disease(N-MBD)group ( n = 11) was measured by RT‒qPCR. ( H ) miR-4261 mimics and control vectors were delivered to THP-1 cells in the presence of both M-CSF and RANKL (100 ng/ml), seven days later, TRAP staining was performed. The typical photos are displayed. Multinucleated cells (≥ 3 nuclei per cell) with wine-red cytoplasm (TRAP-positive) are visible in the images. Cells are indicated by black circles, and nuclei are marked by white arrows (40× objective magnification). ( I )The number of osteoclasts in the miR-4261 mimics-overexpressed and control groups was quantified and compared. ( J ) Representative Micro-CT reconstruction images of the miR-4261 mimics-overexpressing group and the control group. (K-N)The miR-4261 mimics-overexpressed and control groups were subjected to Micro-CT analysis. Original micro-CT images; BV/TV, bone volume/total volume; Tb.Th, trabecular thickness; BMC (g/cm³), bone mineral content; TMC (g/cm³), tissue mineral Content. Data are presented as the mean ± SD from more than 3 biological replicates. Student’s t test was used to estimate the statistical significance. miR-4261 expression in MBD ( n = 27) vs. non-MBD ( n = 11) groups. Significant difference was detected by Mann-Whitney U test ( P < 0.001) despite unbalanced sample sizes.* p < 0.05, ** p < 0.01, *** p < 0.001,**** p < 0.0001.

miR-4261 directly binds to and regulates PAMM gene expression

Identifying the direct target is critical to understand how miR-4261 regulates osteoclast differentiation and increases bone resorption. Using the TargetScan database, 5,336 transcripts were predicted as potential targets of miR-4261. Concurrently, the Gene Ontology (GO) database identified 221 genes linked to osteoclast differentiation regulation (GO:45670). The intersection of these datasets yielded 25 candidate genes, which were further refined using a cumulative weighted context + + score threshold of <-0.35, resulting in two candidates: PRXL2A/PAMM (score: -0.37) and PPARGC1B (score: -0.53) (Fig. 2 ,A). Peroxiredoxin Activated in M-CSF stimulated Monocytes (PAMM) 13 is a novel member of the peroxiredoxin-like protein 2 family, which acts as a redox-regulatory protein to modulate osteoclast differentiation in vitro. Therefore, we speculated that PAMM might be a direct target of miR-4261 secreted by MM cells. Whereas PPARGC1B is a gene associated with diabetes and obesity, we did not consider it further, and subsequent experiments verified the regulatory relationship between miR-4261 and PAMM. To test this, PAMM expression levels in THP-1 cells were measured through day 1 to 7 in OPM2/THP-1, RPMI 8226/THP-1, and NCI-H929/THP-1 co-culture systems. Across all three systems, PAMM expression significantly decreased over time (Fig. 2 ,B - D). A dual-luciferase reporter assay was performed to show that miR-4261 mimics significantly reduced luciferase activity of the wild-type (WT) PAMM 3’UTR ( P < 0.01) but not the mutant construct (Fig. 2 ,E and F). To confirm the regulation of PAMM by miR-4261, both 293T and THP-1 cells were transduced by miR-4261 mimics and the control. 293T cells were not used as a disease-relevant model but served as standard tool cells for high-efficiency transfection, lentiviral packaging, and construct validation. THP‑1 cells were used as the biologically relevant monocytic model for functional experiments related to MM-induced bone mineral changes. PAMM was quantified by RT‒qPCR, and protein expression was assessed by Western Blot. In 293T cells, the miR-4261 mimics significantly decreased the transcript and protein expression of PAMM (Fig. 2 ,G and H). Similar results were obtained in THP-1 cells (Fig. 2 ,I and J).

Fig. 2 miR-4261 directly regulates PAMM expression in THP-1 cells. ( A ) A cartoon illustrates the screening process for miR-4261 targets in this study. The TargetScan database predicted 5,336 genes as putative targets of miR-4261. The gene ontology database identified 221 genes associated with regulated osteoclast differentiation (GO:45670). The intersection of these two datasets yielded 25 candidate genes, with 2 candidates selected under more stringent criteria. ( B – D ). In co-culture systems of OPM2/THP-1, RPMI 8226/THP-1, and NCI-H929/THP-1, THP-1 cells were subjected to RNA extraction, and PAMM expression was quantified by RT‒qPCR. ( E ) Schematic diagram depicting the predicted miR-4261 binding site in the 3’UTR of PAMM mRNA. ( F ). Dual luciferase reporter assay. The 3’UTR fragments (wild-type and mutant sequences) of PAMM were cloned into the psiCHECK2 vector and co-transfected with miR-4261 mimics or negative control (NC). Relative firefly/Renilla luciferase activity was compared among the wild-type (WT), mutant (MUT) and negative control (NC) groups. ( G , H ) PAMM expression in miR-4261-overexpressed and control 293T cells was analyzed by RT‒qPCR and Western blot. ( I , J )PAMM expression in miR-4261-overexpressed and control THP-1 cells was analyzed by RT‒qPCR and Western blot.

Negative correlation between miR-4261 and PAMM expression in MM patients, with low PAMM expression linked to hypercalcemia

Bone marrow mononuclear cells from 27 newly diagnosed MM patients with MBD and 10 healthy controls for the measurement of miR-4261 and PAMM by RT‒qPCR. The results showed that MM group showed significantly higher miR-4261 expression ( P < 0.05) (Fig. 3 , A) and lower PAMM expression ( P < 0.0001) (Fig. 3 , B) compared with normal bone marrow (NBM) group. Correlation analysis indicated a negative relationship between miR-4261 and PAMM (R²=0.3160, P = 0.0065) (Fig. 3 , C). Given the association of the PAMM gene with osteoclast differentiation and bone resorption, we examined the relationship between PAMM expression and some clinical features, including calcium levels. Patients were stratified into high- and low-PAMM expression groups according to the mean relative expression value across all samples. Samples with expression levels above the mean were defined as PAMM-high group ( n = 11), whereas those below the mean were defined as PAMM-low group ( n = 16). The results showed that no significant difference was observed in plasma cell percentage, hemoglobin, or albumin levels, while Patients with low PAMM expression displayed significantly higher serum calcium levels ( P = 0.0032) for hypercalcaemia compared to the PAMM-high group (Fig. 3 , D). To further evaluate the diagnostic potential of miR-4261 and PAMM as biomarkers in MM patients with MBD, we performed receiver operating characteristic (ROC) curve analysis using expression data from 27 MM patients and 10 healthy controls. The area under the ROC curve (AUC) for miR-4261 expression was 0.785 (95% CI: 0.58–0.99, P = 0.0085), indicating moderate discriminatory ability in distinguishing MM with MBD from the control group. The AUC for PAMM expression was 0.9185 (95% CI: 0.82-1, P = 0.001), suggesting a stronger ability to differentiate between the two groups (Fig. 3 , E). The optimal cutoff value for PAMM was < 0.861 (sensitivity: 81.48%, specificity: 90.00%) (Fig. 3 , E). These findings support the potential clinical relevance of miR-4261 and PAMM in identifying MM-associated MBD. Taken together, our data indicate a potential association of the miR-4261/PAMM axis with MBD pathogenesis.

Fig. 3 PAMM is negatively correlated with miR-4261 and PAMM low expression is associated with Hypercalcemia in MM patients at diagnosis.( A , B ) The bone marrow mononuclear cells from healthy donors (NBM, n = 10) and multiple myeloma patients (MM, n = 27) were collected and subjected to RT–qPCR analyses of miR-4261 and PAMM expression. ( C ) The correlation between miR-4261 and PAMM was also analyzed. ( D ) Clinical characteristics of MM patients with high PAMM expression ( n = 11) and low PAMM expression ( n = 16) were analyzed. Calcium ion levels in PAMM-high and PAMM-low groups were compared. ( E ) Receiver operating characteristic (ROC) curve analysis was performed on the miR-4261 and PAMM expression data from 27 MM patients and 10 healthy controls.

Molecular mechanism of PAMM silencing enhancing THP-1 cell differentiation into osteoclasts and bone resorption

In MM patients, low PAMM expression correlates with calcium ion levels. To assess the impact of PAMM silencing on osteoclast function, THP-1 cells were transduced with shPAMM#1, shPAMM#2, and control vectors. RT-qPCR screening results showed that both shPAMM1 and shPAMM2 significantly reduced the mRNA expression levels of PAMM (Fig. 4 A). 由于 shPAMM1 exhibited higher knockdown efficiency (~ 90%), we preferentially selected this sequence for subsequent protein-level validation and functional experiments to ensure maximal gene silencing in the cell model. Western Blot analyses showed that shPAMM vectors effectively reduced the protein expressions of PAMM (Fig. 4 , B).Similar to miR-4261 mimics, TRAP staining showed that shPAMM vectors increased multinucleated giant cell-like morphology upon M-CSF and RANKL induction (Fig. 4 , C), with a significantly higher osteoclast counts compared with controls (Fig. 4 , D). Similarly, micro-CT scans showed that bone slices in the scrambled group had a smooth, dense surface with uniform intensity, whereas bone slices in the shPAMM group exhibited a rough surface with dense resorption lacunae, reduced intensity, and large low-density areas (Fig. 4 ,E). Micro-CT analysis showed that BV/TV and Tb.Th (avg) exhibited a decreasing trend in the shPAMM group, but the differences did not reach statistical significance compared to the scrambled group (Fig. 4 , F and G). At 21 days, the data revealed that BMC and TMC were significantly reduced in the shPAMM group compared with the scrambled group (Fig. 4 , H and I).To obtain molecular insights into how PAMM silencing promotes osteoclast differentiation. RNA-seq data comparing ShPAMM and scrambled THP-1 cells were generated. Gene Set Enrichment Analysis (GSEA) was performed and the gene set negatively regulating osteoclast differentiation (GO:0045671) showed a normalized enrichment score (NES= -1.71, P = 0.009, FDR q = 0.010). Genes including CCL3, TLR3, LILRB3, PIK3R1, LILRB4, UBASH3B, GPR137B, and TNFAIP6 were significantly downregulated in the shPAMM group (Fig. 5 , H and I). Compared with the scrambled group, the shPAMM group exhibited reduced expression of CCL3, TLR3, LILRB3, PIK3R1, LILRB4, UBASH3B, GPR137B, and TNFAIP6 (Fig. 5 , L).These data suggested that PAMM silencing enhances THP-1 cell differentiation into osteoclasts and bone resorption, likely by attenuating the the expression of negative regulatory genes.

Fig. 4 PAMM silencing enhances THP-1 Cell differentiation to osteoclast and enhances osteoclast activity. ( A ) THP-1 cells were transduced by shPAMM#1, shPAMM#2, and control (scrambled) vectors, and PAMM expression was evaluated by RT‒qPCR. ( B ) PAMM protein expression in shPAMM and scrambled THP-1 cells was assessed by Western blot. ( C ) Both shPAMM and scrambled cells were analyzed by TRAP staining after 1 week of differentiation, and typical photos are displayed. ( D ) The osteoclasts were quantified in shPAMM and control THP-1 cells. ( E ) Representative examples of Micro-CT reconstructed images from the shPAMM group and scrambled group. ( F – I ) Micro-CT analysis demonstrated that the shPAMM group exhibited significantly lower BMC, and TMC compared to the control group, with enhanced bone resorption. ( J , K ) Eukaryotic reference-based transcriptome analysis and GO-GSEA revealed significant downregulation of genes associated with negative regulation of osteoclast differentiation (e.g., CCL3, TLR3, LILRB3, PIK3R1, LILRB4, UBASH3B, GPR137B, and TNFAIP6) in shPAMM-treated THP-1 cells, with TNFAIP6 exhibiting the most pronounced downregulation (NES = -1.71, P = 0.009, FDR = 0.010). ( L ) Heatmap of differentially expressed genes in ShPAMM THP-1 cells. RNA-seq data comparing shPAMM with scrambled controls in THP-1 cells ( n = 3). Color scale represents expression changes ranging from downregulation (blue) to upregulation (red).

Proposed model: exosomal miR-4261 drives myeloma bone disease via PAMM-mediated disinhibition of osteoclastogenesis

Overall, the present study uncovers that miR-4261 delivered by exsomes form MM cells enhances osteoclastic bone resorption and hypercalcemia partially via suppression of PAMM expression in osteoclasts (Fig. 5 ). Therefore, effective targeting of miR-4261/PAMM axis provides a new clue to improve the management of MBD patients.

Fig. 5 Targeting miR-4261 or PAMM represents an optional strategy to alleviate clinical symptoms in MBD. PAMM has been identified as a bona fide target of miR-4261. Downregulation of PAMM protein expression relieves its negative regulation on osteoclast differentiation-related genes (e.g., TNFAIP6, CCL3, TLR3), thereby enhancing osteoclast differentiation and activity. This leads to increased bone resorption, manifesting as bone matrix dissolution and resorption pit formation. The imbalance where bone resorption exceeds bone formation ultimately promotes the development of MBD and hypercalcemia.

Discussion

In the present study, we identified exosomal miR-4261 derived from MM cells as a key secretory factor contributing to MBD. MBD is characterized by osteolytic lesions resulting from enhanced osteoclast activity and suppressed osteoblast function, primarily driven by imbalance of the RANKL/OPG axis and complex interactions within the bone marrow microenvironment 14 , 15 . Exosomes play a critical role in mediating communication between MM cells and the bone microenvironment. The bioactive molecules they carry, including miRNAs, are capable of regulating bone remodeling processes 16 , 17 . The upregulation of miR-4261 promotes osteoclastogenesis, suggesting that exosomal miR-4261 may drive osteolytic bone destruction. Our group previously reported significant overexpression of miR-4261 in plasma, exosomes, and mononuclear cells from MM patients, confirming that exosomes are secreted from MM cell membranes and contain the MM-specific marker CD138 10 . In the present study, GW4869-mediated inhibition of exosome secretion from MM cells prevented the elevation of miR-4261 in co-cultured THP-1 cells, indicating that MM cells transfer miR-4261 to THP-1 cells via exosomes (Fig. 1 D-F). Analysis of bone marrow mononuclear cells revealed higher miR-4261 expression levels in MM patients with bone disease compared to those without. In vitro co-culture experiments demonstrated that MM cell lines (OPM2, RPMI 8226, and NCI-H929) deliver miR-4261 to THP-1 cells through exosomes, resulting in significant upregulation of miR-4261 expression (Fig. 1 A-C) and promotion of osteoclast differentiation and bone resorption (Fig. 1 M and N).Regarding bone resorption, the miR-4261 mimics overexpression group and shPAMM group exhibited a decreasing trend in BV/TV and Tb.Th (mean values) (Figs. 1 K and L and 4 F and G), although statistical significance was not achieved due to limitations in differentiation time. However, by day 21 of THP-1 induction, significant differences in BMC and TMC were observed between the miR-4261 mimics overexpression group and shPAMM group compared to controls (Figs. 1 M and N and 4 H and I), suggesting that bone resorption initially manifests as mineral loss, potentially followed by macroscopic changes in BV/TV and Tb.Th (mean values). These findings imply that elevated miR-4261 expression may be associated with MM-related symptoms. Our observations align with reports from other bone-metastatic cancers where microRNAs promote osteoclastogenesis. For instance, Wu et al. reported significant upregulation of exosomal miR-19a and integrin-binding sialoprotein secreted by bone-tropic ER+ breast cancer cells, wherein exosomal miR-19a is internalized by osteoclasts to induce their formation 18 . Recent studies have also shown that exosomal miR-503-3p regulates osteoclast differentiation via the Hpse axis, further confirming the critical role of exosomal miRNAs in bone microenvironment remodeling 19 . Notably, MM cells interact with bone marrow stromal cells through exosomes to promote disease progression, and the miRNA cargo in these exosomes may serve as predictive biomarkers for MM relapse and drug resistance 20 .

The role of miR-4261 varies across different cancers. For example, MEG3 interacts with miR-4261 to upregulate DKK2 expression, blocking the Wnt/β-catenin signaling pathway and thereby inhibiting tumorigenesis and progression in esophageal squamous cell carcinoma (ESCC) 21 . Conversely, in colorectal cancer, the mutated in colorectal cancer (MCC) gene is a direct target of miR-4261. Overexpression of miR-4261 enhances proliferation, promotes G1/S phase transition, and increases migration of HCT116 and HT29 cells, indicating an oncogenic role 22 . These studies indicate the context-dependent functions of miR-4261. In MM, our data show that elevated miR-4261 expression promotes osteoclast differentiation and bone resorption, contributing to MBD development and impacting the quality of life of MM patients. Numerous miRNAs play significant roles in osteoclast differentiation, function, survival, and communication between osteoblasts and osteoclasts 23 , 24 . For example, a recent review indicates that miRNAs regulate osteoclast-mediated bone resorption by targeting key signaling pathways such as the RANKL/OPG axis, offering novel therapeutic avenues for MBD 25 . Elevated miR-4261 expression may enhance osteoclast activity, leading to bone matrix degradation and calcium ion release. These data suggest that serum exosomal miR-4261 could serve as a liquid biomarker for identifying MM patients at high risk for myeloma bone disease.

Mechanistic studies revealed that miR-4261 directly suppresses PAMM (alias: FAM213A, PRXL2A) expression by targeting its 3’-UTR (Fig. 2 E and F). PAMM belongs to the peroxiredoxin-like 2 family, featuring a conserved cysteine active site that enables it to function as an antioxidant enzyme, maintaining cellular redox homeostasis by scavenging reactive oxygen species (ROS) 13 , 26 . During osteoclast differentiation, ROS serve as second messengers to amplify RANKL signaling, influencing pathways such as NF-κB, MAPK, c-Fos, and NFATc1 27 , 28 . Consequently, PAMM’s regulation of ROS levels indirectly modulates these pathways; high PAMM expression attenuates osteoclastogenesis by suppressing oxidative stress-induced activation 13 . PRXL2A, as an antioxidant gene, influences various diseases and can predict prognosis while aiding in the selection of immunotherapy candidates 29 , 30 . In hematological disorders, FAM213A regulates oxidative stress and hematopoiesis, with its high expression in acute myeloid leukemia (AML) associated with poor prognosis 31 . Our data from THP-1 cells overexpressing miR-4261 or with PAMM knockdown demonstrate enhanced osteoclast differentiation, activity, and bone resorption (Figs. 1 M and N and 4 H and I). High PAMM expression inhibits osteoclast formation by suppressing NF-κB and c-Jun, thereby reducing RANKL-induced activation 11 . This aligns with our current findings, wherein elevated PAMM expression inhibits osteoclast formation and differentiation, whereas PAMM knockdown significantly enhances these processes. This study expands the understanding of miR-4261, confirming PAMM as its key target and establishing a logical chain: upregulation of miR-4261 leads to downregulation of PAMM, ultimately promoting osteoclastogenesis. Mechanistic investigations involved eukaryotic transcriptome sequencing data from scrambled control and shPAMM experimental groups to identify differentially expressed genes (DEGs) and perform enrichment analysis. In the shPAMM group, most DEGs were downregulated compared to the control. Enrichment analysis showed significant enrichment in GO:0045671 (negative regulation of osteoclast differentiation), with genes involved in this process exhibiting negative enrichment in both early and late stages. Downregulation of PAMM enhances bone resorption by attenuating the expression of genes that negatively regulate osteoclast differentiation (e.g., TNFAIP6, CCL3, and TLR3) (Fig. 4 J-L). Although traditional studies on bone destruction focus on the RANKL pathway 32 , 33 , this study emphasizes PAMM-mediated negative regulation of osteoclast-related gene sets, broadening the understanding of the molecular network in MBD. Our transcriptomic data further elucidates the role of the miR-4261/PAMM axis as an upstream regulator.

In clinical samples, miR-4261 expression was significantly elevated in MBD patients compared to MM patients without bone disease. In newly diagnosed MM patients, miR-4261 expression exhibited a negative correlation with PAMM expression. Patients with low PAMM expression displayed higher serum calcium levels, with significant differences in calcium ion levels between high- and low-PAMM expression groups (Fig. 3 D).Our study found that PAMM knockdown reduced the mineral content of bone slices, indicating enhanced in vitro osteoclast-mediated bone resorption (Fig. 4 H and I).Given the frequent occurrence of hypercalcemia in MBD patients, we hypothesize that miR-4261-mediated regulation of PAMM may influence systemic calcium levels by modulating osteoclast activity. However, it should be noted that this study did not directly measure calcium release; thus, the association between the miR-4261/PAMM signaling axis and hypercalcemia remains speculative. According to our ROC curve analysis, miR-4261 and PAMM show diagnostic potential in distinguishing MM patients with MBD from healthy controls. The AUC for miR-4261 was 0.785 (95% CI: 0.58–0.99, P = 0.0085), indicating moderate discriminatory ability. Although the confidence interval is relatively wide, possibly due to limited sample size, its statistical significance suggests that miR-4261 may hold important biological relevance in the pathogenesis and progression of MM-associated bone disease. In contrast, PAMM achieved an AUC of 0.9185, exhibiting superior diagnostic performance. As shown in Fig. 3 E, the ROC curve for PAMM lies above that of miR-4261, further supporting its stronger discriminatory capacity between groups. Additionally, at the optimal PAMM cutoff value of < 0.861, the sensitivity was 81.48% and specificity was 90.00%, indicating good sensitivity while maintaining high specificity, with potential clinical utility. Nevertheless, this study has certain limitations, including a relatively small sample size (27 MM patients and 10 healthy controls), which may affect the stability of AUC estimates and the range of confidence intervals. Overall, this study provides preliminary evidence for the clinical application of miR-4261 and PAMM in MM with MBD. This suggests that PAMM knockdown may lead to osteoclast-mediated bone matrix dissolution and calcium ion release, thereby potentially contributing to MBD and hypercalcemia.

Although our study focuses on the paracrine effects of miR-4261 on osteoclasts, its potential autocrine roles in MM cells themselves warrant consideration. In other cancers, miR-4261 exhibits oncogenic properties by promoting proliferation and migration 22 .Although our study primarily focuses on the paracrine effects of miR-4261 on osteoclasts, its potential autocrine role in myeloma cells also warrants consideration. In other cancers, miR-4261 exhibits oncogenic properties by promoting proliferation and migration 22 .Similarly, PAMM participates in oxidative stress regulation in hematological malignancies such as acute myeloid leukemia 31 , it may influence myeloma cell survival or drug resistance. Therefore, we cannot exclude the possibility that the miR-4261/PAMM axis functions within myeloma cells 34 . Future studies should explore whether modulating miR-4261/PAMM in myeloma cells impacts tumor progression, for example, by evaluating cell proliferation, apoptosis, or chemoresistance through knockdown experiments.

This study has several limitations that should be acknowledged. First, the sample size was relatively small (27 patients with MBD, 11 MM patients without bone disease, and 10 healthy controls), which limits the generalizability of the findings and necessitates validation in larger-scale, multicenter cohorts. Second, reliance on in vitro models failed to fully capture the complexity of the MM bone microenvironment, warranting future in vivo investigations of the miR-4261/PAMM axis. Third, the clinical samples consisted of bone marrow mononuclear cells, which represent a heterogeneous population including but not limited to osteoclasts and various other cell types. Consequently, the observed expression patterns of miR-4261 and PAMM may reflect contributions from multiple cellular lineages rather than being specific to osteoclasts. Nevertheless, given the focus of this study on bone damage in MBD, our investigations primarily targeted osteoclasts as the cell type of interest. Additionally, our dependence on the THP-1 cell line as an osteoclast model presents specific limitations.THP-1 cells are derived from acute monocytic leukemia and possess a cancerous origin, which may result in gene expression profiles that differ from those of primary CD14 + monocytes from peripheral blood 35 .Furthermore, THP-1-derived osteoclasts have limited resorptive capacity, often failing to form fully mature multinucleated cells with robust bone resorption activity, which is inferior to that observed in peripheral blood mononuclear cell (PBMC)-derived osteoclasts 36 . To overcome these limitations, future validations should employ primary human CD14 + monocytes or murine bone marrow-derived precursors in both in vitro and in vivo models 37 .

In summary, this study identifies the miR-4261/PAMM axis as a novel therapeutic target for MBD. Recent investigations have further demonstrated that targeting miRNAs, such as miR-29b, can simultaneously suppress tumor growth and promote bone remodeling, offering new avenues for comprehensive MBD therapy 38 .First, this study represents the initial identification of miR-4261 as an exosomal miRNA derived from MM cells that promotes osteoclastogenesis. Second, it confirms the functional role of PAMM in osteoclast differentiation. Third, it establishes the miR-4261/PAMM axis as a regulatory model in MBD.MM cells deliver exosomal miR-4261 to osteoclast precursor cells, where it directly suppresses PAMM, thereby alleviating the inhibition of negative regulators of osteoclast differentiation, including TNFAIP6, CCL3, and TLR3. Compared to scrambled controls, shPAMM treatment resulted in modest gene downregulation, as indicated by the blue shading in the heatmap. This led to a significant enhancement in osteoclastogenesis and bone resorption, potentially driving the osteolytic lesions and hypercalcemia observed in MBD. Through analyses of patient samples and cellular-level functional experiments, PAMM emerges as a promising non-invasive biomarker. Combined assessment of PAMM and miR-4261 may further improve diagnostic accuracy. Overall, these molecules exhibit dual utility: serving as liquid biopsy biomarkers for early MBD detection and risk stratification, while also representing potential therapeutic targets that warrant further exploration to mitigate myeloma-related bone destruction and enhance clinical outcomes.

Methods

Cells and human samples

THP-1 cells (catalog number: SCSP-567) and 293T cells (catalog number: SCSP-502) were purchased from the Chinese Academy of Sciences Cell Bank ( www.cellbank.org.cn ). The NCI-H929, RPMI 8226, and OPM-2 cell lines were generously provided by Dr. Su-Ning Chen from the First Affiliated Hospital of Soochow University. All cell lines were confirmed to be free of mycoplasma contamination, and their identities were authenticated by short tandem repeat (STR) profiling.THP-1 cells were cultured in Roswell Park Memorial Institute (RPMI) 1640 medium supplemented with 10% fetal bovine serum (FBS), while 293T cells were maintained in Dulbecco’s Modified Eagle Medium (DMEM) containing 10% FBS. The NCI-H929, RPMI 8226, and OPM-2 cell lines were cultured in RPMI-1640 medium with 20% FBS.

Bone marrow cells (BMCs) were obtained from patients with multiple myeloma (MM) and healthy donors at the Second Hospital of Shanxi Medical University. These BMCs were residual samples from clinical testing. All research involving human participants was performed in accordance with the Declaration of Helsinki and relevant guidelines/regulations. The study was approved by the Ethics Committee of the Second Hospital of Shanxi Medical University (approval number: 2022 YX (244), Taiyuan, China), and informed consent was obtained from all participants. The clinical characteristics of the recruited MM patients are summarized in Table 1 .

Table 1 The clinical characteristics of multiple myeloma patients recruited in this study. Characteristic MBD ( n = 27) N -MBD ( n = 11) Median age (range) - years 50–76(62) 53–75(65) Male sex - number 20 6 Female sex - number 7 5 Types of multiple myeloma IgG-λ 9 3 IgG-κ 6 1 IgA-λ 4 IgA-κ 2 IgD-λ 1 1 IgD-κ λ 1 2 κ 2 3 Non- secretory 2 1 D-S system Ⅰ 2 1 Ⅱ 4 2 Ⅲ 21 8 ISS system Ⅰ 2 1 Ⅱ 7 2 Ⅲ 18 8 Median C-BM-PCs (range) - % 15–92(42) 10–45(24) Median Hb (range) - g/L 63–165(103) 86–143(120) Median ALB (range) - g/L 17–46(33) 18–44(36.2) Median sβ-2MG (range) - mg/L 1.2–24.8(9.09) 1.4–21.4(6.4) Median sCa (range) – mmol/L 1.86–3.64(2.54) 1.91–2.5(2.25) MMBD: myeloma bone disease, N-MMBD: on-myeloma bone disease, C-BM-PCs: Clonal Bone Marrow Plasma Cells, Hb: Hemoglobin, ALB: Albumin, sCa: Serum Calcium, sβ-2MG: Serum β-2 Microglobulin, D-S system: Durie-Salmon System, ISS system: International Staging System.

RNA extraction and RT-qPCR

Total RNA were extracted using TRIzol (Thermo Scientific, Waltham, MA, USA). For each sample, 1000 ng of total RNA was used as the template for mRNA reverse transcription using random primers and the Transcriptor First Strand cDNA Synthesis Kit (Roche (China) Holding Ltd., Shanghai, China). Separately, another 1000 ng of total RNA was allocated for miRNA reverse transcription, where miR-4261-specific stem-loop primers(GenePharma Co., Ltd., Shanghai, China) were employed to ensure accurate synthesis of miRNA cDNA. The resulting cDNA products were then used for Real-time quantitative polymerase chain reaction (RT-qPCR) to determine the expression levels of target mRNA and miR-4261, respectively.RT-qPCR was conducted with 0.5 µL of cDNA, 400 nM primers, and 12.5 µL of 2× Power SYBR Green PCR Master Mix (ThermoFisher, Applied Biosystems, Foster City, CA, USA) on a 7500 Real-Time PCR System (Applied Biosystems). β-ACTIN or U6 was used as an endogenous control to normalize mRNA and miRNA input levels, respectively. Details of gene-specific primers are provided in Table 2 .

Table 2 The gene-specific primers used in this study. Name sequence(5’-3’) For RT-qPCR Human actin F CACCATTGGCAATGAGCGGTTCC R GTAGTTTCGTGGATGCCACAGG Human PAMM F ATAGACCTGAAAACACTGGA R GCAGCTTCCTCTCGACAGAG hsa-miR-4261 F TGCTGGACGAGGAAACAGGG R TATGGTTGTTCACGAGTCCTTGTC u6 snRNA F CGCTTCGGCAGCACATATAC R TTCACGAATTTGCGTGTCATC For 3’-UTR of PAMM PAMM(WT) F TCGAGTTATTGTTTCCAAGGGCATGC R CAATAACAAAGGTTCCCGTACGCCGG PAMM (MT) F TCGAGTTATCTCAGATAAGGGCATGC R CAATAGAGTCTATTCCCGTACGCCGG The mutant sequences are shown in the boxes. MT, mutant.

Western blot analysis

Total protein was extracted using RIPA lysis buffer (Beyotime, Shanghai, China) containing 50 mM Tris (pH 7.4), 150 mM NaCl, 1% Triton X-100, 1% sodium deoxycholate, and 0.1% SDS, supplemented with 1 mM phenylmethanesulfonyl fluoride (PMSF) and a protease inhibitor cocktail (Sigma-Aldrich, St. Louis, MO, USA). Protein concentrations were determined using a BCA protein assay kit (Cat. No. P0010, Beyotime, Shanghai, China). For each sample, 30 µg of protein lysate was separated by SDS-PAGE and transferred onto PVDF membranes (Millipore, Billerica, MA, USA). After blocking with 5% non-fat milk for 1 h at room temperature, the membranes were incubated overnight at 4 °C with primary antibodies targeting PAMM (1:500; Cat. No. 20446-1-AP, Proteintech, Wuhan, China), ACTIN (1:1000; Cat. No. 66009-1-Ig, Proteintech), and Tubulin (1:1000; Cat. No. 10094-1-AP, Proteintech). The membranes were then incubated with HRP-conjugated secondary antibodies (1:5000; Cat. Nos. RGAM001 and RGAR002, Proteintech) for 1 h at room temperature. Protein bands were visualized using an ECL detection system (New Cell & Molecular Biotech Co., Ltd., Suzhou, China) and developed with a Kodak Medical X-ray Processor 102. ACTIN and Tubulin were used as loading controls.

Co-culture

THP-1 cells were seeded at the bottom of 6-well plates (1 × 10⁵ cells per well), while NCI-H929 cells were seeded in the upper inserts of 6-well Transwell plates (2 × 10⁴ cells per insert; 0.4 μm pore size). The culture medium consisted of RPMI-1640 supplemented with 20% fetal bovine serum (FBS). During co-culture, the bottom cells were collected at seeding (Day 0), 24 h (Day 1), 72 h (Day 3), 120 h (Day 5), and 168 h (Day 7) for subsequent analysis. The same methodology was applied to experiments involving RPMI 8226 and OPM-2 cells. To inhibit exosome generation, GW4869 (Beyotime, S1971, Shanghai, China) was dissolved in DMSO to prepare a 5 mM stock solution and stored at − 80 °C. Preliminary experiments using the CCK-8 assay confirmed that treatment with 4 µM GW4869 for 7 days had no significant effect on the viability of THP-1 cells or multiple myeloma (MM) cell lines (OPM2, H929, and RPMI-8226). For experiments, the stock was diluted in complete culture medium to achieve the desired working concentration (with DMSO ≤ 0.1%). In the Transwell co-culture system, MM cells were seeded in the upper chamber and THP-1 cells in the lower chamber. The medium in both chambers was replaced every 24 h (half-volume change) with fresh complete medium containing GW4869 to maintain a final concentration of 4 µM throughout the 7-day co-culture period. At the end of the co-culture, THP-1 cells in the lower chamber were collected, total RNA was extracted, and qPCR was performed to determine the expression levels of miR-4261.

Plasmid construction and dual-luciferase reporter assay

Wild-type (WT) and mutant (MUT) double-stranded DNA fragments of PAMM, along with their antisense strands, were designed based on the miR-4261 recognition sequence in the 3’-untranslated region (3’-UTR) of PAMM. These fragments were ligated into the reporter plasmid psiCHECK-2 (Promega, USA) at matching sticky ends. Using LipoD293™ DNA transfection reagent (SL100668, SignaGen, USA), WT or MUT psiCHECK-2 plasmids were co-transfected with miRNA mimics or negative controls (NC) into HEK293T cells. After 48 h, cells were harvested and lysed, and luciferase and Renilla luciferase activities were measured using the Dual-Luciferase Reporter Assay Kit (E1910, Promega, USA). Results were expressed as the Firefly/Renilla ratio. The relative reporter activity in miR-4261-overexpressing cells was normalized to that in control cells. Synthesis of miRNA mimics and control nucleic acid fragments was outsourced to GenePharma ( https://www.genepharma.com/ ).

Lentivirus production and transduction

Lentiviral expression vectors encoding miR-4261 and control sequences were synthesized by GenePharma Co., Ltd. (Suzhou, China). Short hairpin RNA (shRNA) constructs targeting PAMM (two independent sequences) and scrambled control shRNA were obtained from GeneCopoeia™ (Rockville, MD, USA).

For lentivirus production, human embryonic kidney (HEK) 293T cells were seeded at a density of \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:6\times\:{10}^{6}$$\end{document} cells per 10 cm dish in Dulbecco‘s Modified Eagle‘s Medium (DMEM) supplemented with 10% fetal bovine serum (FBS). After 16–24 h, the culture medium was replaced with 4.5 mL of fresh DMEM 2–4 h prior to transfection. For each construct, a DNA mixture was prepared in 250 µL of Opti-MEM containing 6 µg of the lentiviral vector plasmid, along with the packaging plasmids: 3.9 µg of ΔR, 1.5 µg of REV, and 2.1 µg of vesicular stomatitis virus glycoprotein (VSV-G) envelope plasmid. Separately, 40 µg of polyethyleneimine (PEI, 1 mg/mL) was diluted in 210 µL of Opti-MEM. The DNA mixture was added dropwise to the PEI solution, incubated for 20 min at room temperature, and then added dropwise to the 293T cells. The cells were incubated at 37 °C with 5% CO₂ for 48 h. Viral supernatants were collected at 48 and 72 h post-transfection, filtered through a 0.45 μm low protein-binding filter (MilliporeSigma), and concentrated by ultracentrifugation at 25,000 rpm for 90 min at 4 °C. The viral pellet was resuspended in serum-free Iscove‘s Modified Dulbecco‘s Medium (IMDM) containing 5% DNase under gentle agitation for at least 1 h at room temperature. Aliquots were stored at − 80 °C until use.

For lentiviral transduction, THP-1 cells were plated at \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:2\times\:{10}^{5}$$\end{document} cells per well in complete culture medium and incubated for 16 h. The medium was then supplemented with 20 µL of the lentiviral suspension per well, and cells were incubated for an additional 16 h at 37 °C with 5% CO₂. Following transduction, cells were washed three times with Hanks‘ Balanced Salt Solution (HBSS) containing 2% FBS and replated in 1 mL of fresh complete medium. After 48 h of recovery, GFP-positive cells were selected based on the expression of the GFP marker encoded in the lentiviral backbone. All sequences are listed in Table 3 .

Table 3 The sequences of shRNA, mimics in this study. Name Sequence (5’-3’) scrambled GCTTCGCGCCGTAGTCTTA shPAMM#1 GGAGACAAAGTAAACCTACTT shPAMM#2 GGAGTGAGAAACCCATTTATA For has-miR-61 mimics overexpression has-miR-61 mimics Sense AGGAAACAGGGACCCA Antisense GGUCCCUGUUUCCUUU Negative control Sense UUCUCCGAACGUGUCACGUTT Antisense ACGUGACACGUUCGGAGAATT

Differentiation of THP-1 cells into osteoclasts

THP-1 cells (3 × 10⁵ cells per dish) were cultured in 3.5 cm dishes with RPMI 1640 medium supplemented with 10% FBS, 100 IU/mL penicillin, and 100 µg/mL streptomycin at 37 °C in a 5% CO₂ atmosphere. Cells were treated with 100 ng/mL phorbol 12-myristate 13-acetate (PMA) for 0, 1, or 2 days, followed by incubation with 100 ng/mL receptor activator of nuclear factor kappa-B ligand (RANKL) and macrophage colony-stimulating factor (M-CSF) for 7 days 39 . The medium was replaced every 2 days, and cell morphology was monitored.

Tartrate-resistant acid phosphatase (TRAP) staining

Osteoclasts were identified via TRAP staining. Cells were fixed with TRAP fixative (Beijing Solarbio Science & Technology Co., Ltd.) at 4 °C for 60 s, washed with water, and air-dried slightly. Samples were then incubated in TRAP staining solution at 37 °C in the dark for 60 min, followed by washing with water. Cells were counterstained with hematoxylin for 5 min at 37 °C, washed again, dried, and examined under a light microscope at 40× magnification.

Microscopic quantification of osteoclasts

After TRAP staining, cells were observed in culture dishes under an inverted optical microscope. TRAP-positive multinucleated cells were defined as those exhibiting a wine-red cytoplasm and containing three or more nuclei; such cells were identified as osteoclast-like cells.For quantitative analysis, a standardized field-selection protocol was applied. Samples were first rapidly screened under a 40× objective to confirm the presence and multinucleated morphology of TRAP-positive cells. Formal counting was then performed under a 20× objective. For each well, 5–10 non-overlapping fields were randomly selected. The exact number of fields was adjusted according to cell density: when more than 10 TRAP-positive multinucleated cells were observed in most fields, five fields per well were randomly selected for counting; when the number was ≤ 10 in most fields, ten fields were counted.This design aimed to balance counting accuracy and experimental efficiency. In high-density conditions, fewer fields were analyzed to reduce workload, whereas in low-density conditions, additional fields were included to minimize sampling error. This approach ensured that the calculated mean values reliably reflected the actual number of osteoclast-like cells in each group.Results were expressed as the average number of TRAP-positive multinucleated cells per field. Data for each group were obtained from at least three independent experiments, and the final results are presented as mean ± SD.

Micro-computed tomography (micro-CT) analysis

Bovine cortical bone slices (Boneslices, Product Catalog Number: 12827; Lot Number: 815, Date:2022/09/15) were used to assess osteoclast-mediated bone resorption. THP-1 cells were seeded at 2 × 10⁵ cells per well in 24-well plates, with three bone slices per well. Osteoclast differentiation was induced in various experimental groups by continuous stimulation with M-CSF and RANKL for 21 days. Bone resorption and osteoclast formation were analyzed using micro-CT with a Micro CT system (NMC-200, NEMO). Imaging parameters included a maximum axial scan range of 250 mm, a radial field of view (FOV) of 100 mm, a scan speed of up to 4 s per bed position, a minimum reconstruction pixel size of 2 μm, and a spatial resolution of < 7.5 μm at 10% modulation transfer function (MTF). Micro-CT data were reconstructed into TIFF images and quantified as bone volume/total volume (BV/TV), trabecular thickness (Tb.Th, mean, mm), bone mineral Content (BMC, g/cm³), and tissue mineral Content (TMC, g/cm³) 40 , 41 .

RNA sequencing and analysis

THP-1 cells were transduced with shPAMM or scrambled control. Total RNA was isolated using TRIzol reagent (Invitrogen), and 1 µg per sample was used for library construction. Poly(A) + RNA libraries were prepared using the NEBNext Ultra RNA Library Prep Kit for Illumina (New England Biolabs) according to the manufacturer’s protocol and sequenced on the Illumina platform. Raw reads were trimmed for adapters and quality filtered using fastp (v0.23.2), and clean reads were aligned to the GRCh38.p14 reference genome using HISAT2 (v2.2.1, --dta). Transcript assembly was performed using StringTie (v2.2.1), with merged annotations generated via stringtie --merge. Novel transcripts were identified using gffcompare (v0.12.6), retaining unannotated transcripts (class_code “u”) with ≥ 100 bp and ≥ 2 exons. Gene expression was quantified as FPKM. For differential expression gene (DEG) identification, read counts were obtained using Subread (v2.0.3) featureCounts, and analysis was conducted in R (v4.2.1) using DESeq2 (v1.34.0), with thresholds of |log2 fold change| >1 and adjusted P-value < 0.05; most DEGs in the shPAMM group were downregulated. Gene set enrichment analysis (GSEA) was performed using GSEA software (v4.3.2, Broad Institute), based on ranked lists of early and late DEGs, employing MSigDB GO gene sets; GO:0045671 (negative regulation of osteoclast differentiation) exhibited significant negative enrichment.

Statistical analysis

Differences among multiple groups were analyzed using one-way analysis of variance. The relationship between the independent variable \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:X$$\end{document} and the dependent variable \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:Y$$\end{document} was evaluated using simple linear regression analysis, with the regression coefficient and its 95% confidence interval (CI) used to assess the strength of the association. The Mann–Whitney U test was applied for small sample sizes, non-normally distributed data, or datasets containing extreme values. Data are presented as mean ±Standard Deviation (SD) from three independent biological replicates. Statistical significance was determined using Student’s t-test, with a threshold of p < 0.05 considered significant. P-values were calculated using the log-rank test (* p < 0.05, ** p < 0.01, *** p < 0.001,**** p < 0.0001).

Supplementary Information

Below is the link to the electronic supplementary material.

Supplementary Material 1

Supplementary Material 2

Supplementary Material 3

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

These authors contributed equally: Jiangxia Cui, Sicheng Bian and Zhuanghui Hao.

Acknowledgements

We express our sincere gratitude to Dr. Su-Ning Chen from the First Affiliated Hospital of Soochow University for generously providing the NCI-H929, RPMI 8226, and OPM-2 cell lines. We also thank all members of our research team for their valuable contributions and technical support throughout this study.

Author contributions

JX, SC, ZH performed the experimental studies and analyzed the data, CZ performed the bioinformatics analysis, XL supervised the quality of clinical samples, YZ, XYand HW analyzed the data, prepared the manuscript and supervised the experimental study. All authors read and approved the manuscript.

Funding

This work was supported by The Shanxi Natural Science Foundation of China (NO.202203021222391),JXC, Associate chief physician, The Second Clinical Medical College of Shanxi Medical University; This research was funded by Shanxi BethuneTalent Foundation Project ( No.2021RC038),XLZ, The Shanxi Bethune Hospital, Shanxi Academy of Advanced Research and Innovation.

Data availability

The RNA sequencing datasets generated during this study are publicly available in the Gene Expression Omnibus (GEO) repository under the accession code GSE315264 (accessible via: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc= GSE315264 ). All other data supporting the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Competing interests

The authors declare no competing interests.

References

1. Global Burden of Disease Global, Regional, and National Cancer Incidence, Mortality, Years of Life Lost, Years Lived With Disability, and Disability-Adjusted Life-Years for 29 Cancer Groups, 1990 to 2017: A Systematic Analysis for the Global Burden of Disease Study JAMA Oncol. 2019 5 12 1749 1768 10.1001/jamaoncol.2019.2996 31560378 PMC6777271 Global Burden of Disease et al. Global, regional, and national cancer incidence, mortality, years of life lost, years lived with disability, and disability-adjusted life-years for 29 cancer groups, 1990 to 2017: A systematic analysis for the global Burden of disease study. JAMA Oncol. 5 (12), 1749–1768 (2019). 31560378 10.1001/jamaoncol.2019.2996 PMC6777271

2. Hameed A Brady JJ Dowling P Clynes M O’Gorman P Bone disease in multiple myeloma: pathophysiology and management Cancer Growth Metastasis 2014 7 33 42 10.4137/CGM.S16817 25187738 PMC4133035 Hameed, A., Brady, J. J., Dowling, P., Clynes, M. & O’Gorman, P. Bone disease in multiple myeloma: pathophysiology and management. Cancer Growth Metastasis . 7 , 33–42 (2014). 25187738 10.4137/CGM.S16817 PMC4133035

3. Eskelinen V First-Line Combination with Proteasome Inhibitor-Based Treatment and Zoledronic Acid Is Effective in Reducing Later Fractures in Multiple Myeloma Irrespective of Multiple Myeloma Bone Disease at Diagnosis Hematol. Rep. 2024 16 3 529 540 10.3390/hematolrep16030051 39189247 PMC11348108 Eskelinen, V. et al. First-line combination with proteasome inhibitor-based treatment and zoledronic acid is effective in reducing later fractures in multiple myeloma irrespective of multiple myeloma bone disease at diagnosis. Hematol. Rep. 16 (3), 529–540 (2024). 39189247 10.3390/hematolrep16030051 PMC11348108

4. Giuliani N Rizzoli V Roodman GD Multiple myeloma bone disease: Pathophysiology of osteoblast inhibition Blood 2006 108 13 3992 3996 10.1182/blood-2006-05-026112 16917004 Giuliani, N., Rizzoli, V. & Roodman, G. D. Multiple myeloma bone disease: Pathophysiology of osteoblast inhibition. Blood 108 (13), 3992–3996 (2006). 16917004 10.1182/blood-2006-05-026112

5. Alegre A Gironella M Bailén A Giraldo P Zoledronic acid in the management of bone disease as a consequence of multiple myeloma: a review Eur. J. Haematol. 2014 92 3 181 188 10.1111/ejh.12239 24330023 Alegre, A., Gironella, M., Bailén, A. & Giraldo, P. Zoledronic acid in the management of bone disease as a consequence of multiple myeloma: a review. Eur. J. Haematol. 92 (3), 181–188 (2014). 24330023 10.1111/ejh.12239

6. Huang H CLO25-072: A Multicenter, Randomized, Controlled, Non-Inferiority Study of Narlumosbart Compared With Denosumab in the Treatment of Bone Disease in Patients With Multiple Myeloma J. Natl. Compr. Canc Netw. 2025 23 3.5 CLO25 072 10.6004/jnccn.2024.7215 Huang, H. et al. CLO25-072: A multicenter, randomized, controlled, non-inferiority study of narlumosbart compared with Denosumab in the treatment of bone disease in patients with multiple myeloma. J. Natl. Compr. Canc Netw. 23 (3.5), CLO25–072 (2025).

7. Huang SY Denosumab Versus Zoledronic Acid in Bone Disease Treatment of Newly Diagnosed Multiple Myeloma: An International, Double-Blind, Randomized Controlled Phase 3 Study-Asian Subgroup Analysis Adv. Ther. 2020 37 7 3404 3416 10.1007/s12325-020-01395-x 32524500 PMC7467415 Huang, S. Y. et al. Denosumab versus Zoledronic acid in bone disease treatment of newly diagnosed multiple myeloma: An international, double-blind, randomized controlled phase 3 study-Asian subgroup analysis. Adv. Ther. 37 (7), 3404–3416 (2020). 32524500 10.1007/s12325-020-01395-x PMC7467415

8. Chen T Moscvin M Bianchi G Exosomes in the Pathogenesis and Treatment of Multiple Myeloma in the Context of the Bone Marrow Microenvironment Front. Oncol. 2020 10 608815 10.3389/fonc.2020.608815 33251153 PMC7674795 Chen, T., Moscvin, M. & Bianchi, G. Exosomes in the pathogenesis and treatment of multiple myeloma in the context of the bone marrow microenvironment. Front. Oncol. 10 , 608815 (2020). 33251153 10.3389/fonc.2020.608815 PMC7674795

9. Faict S Exosomes play a role in multiple myeloma bone disease and tumor development by targeting osteoclasts and osteoblasts Blood Cancer J. 2018 8 11 105 10.1038/s41408-018-0139-7 30409995 PMC6224554 Faict, S. et al. Exosomes play a role in multiple myeloma bone disease and tumor development by targeting osteoclasts and osteoblasts. Blood Cancer J. 8 (11), 105 (2018). 30409995 10.1038/s41408-018-0139-7 PMC6224554

10. Bian S Exosomal MiR-4261 mediates calcium overload in RBCs by downregulating the expression of ATP2B4 in multiple myeloma Front. Oncol. 2022 12 978755 10.3389/fonc.2022.978755 36091107 PMC9458875 Bian, S. et al. Exosomal MiR-4261 mediates calcium overload in RBCs by downregulating the expression of ATP2B4 in multiple myeloma. Front. Oncol. 12 , 978755 (2022). 36091107 10.3389/fonc.2022.978755 PMC9458875

11. Xu Y PAMM: a redox regulatory protein that modulates osteoclast differentiation Antioxid. Redox Signal. 2010 13 1 27 37 10.1089/ars.2009.2886 19951071 PMC2877117 Xu, Y. et al. PAMM: a redox regulatory protein that modulates osteoclast differentiation. Antioxid. Redox Signal. 13 (1), 27–37 (2010). 19951071 10.1089/ars.2009.2886 PMC2877117

12. Li W Oxidative stress-related gene PRXL2A may participate in the occurrence and development of steroid-induced osteonecrosis of the femoral head by regulating osteoclast differentiation: A multi-omics analysis Gene 2025 965 149676 10.1016/j.gene.2025.149676 40681124 Li, W. et al. Oxidative stress-related gene PRXL2A may participate in the occurrence and development of steroid-induced osteonecrosis of the femoral head by regulating osteoclast differentiation: A multi-omics analysis. Gene 965 , 149676 (2025). 40681124 10.1016/j.gene.2025.149676

13. Xu Y Morse LR da Silva RAB Wang D Battaglino RA A short report: PAMM, a novel antioxidant protein, induced by oxidative stress Redox Biol. 2015 6 446 453 10.1016/j.redox.2015.09.008 26402163 PMC4588419 Xu, Y., Morse, L. R., da Silva, R. A. B., Wang, D. & Battaglino, R. A. A short report: PAMM, a novel antioxidant protein, induced by oxidative stress. Redox Biol. 6 , 446–453 (2015). 26402163 10.1016/j.redox.2015.09.008 PMC4588419

14. Wang J Zweegman S Groen RWJ Models of myeloma bone disease: In vivo and in vitro approaches Bone 2026 205 117789 10.1016/j.bone.2026.117789 41525837 Wang, J., Zweegman, S. & Groen, R. W. J. Models of myeloma bone disease: In vivo and in vitro approaches. Bone 205 , 117789 (2026). 41525837 10.1016/j.bone.2026.117789

15. Mukkamalla SKR Malipeddi D Myeloma Bone Disease: A Comprehensive Review Int. J. Mol. Sci. 2021 22 12 6208 10.3390/ijms22126208 34201396 PMC8227693 Mukkamalla, S. K. R. & Malipeddi, D. Myeloma bone disease: A comprehensive review. Int. J. Mol. Sci. 22 (12), 6208 (2021). 34201396 10.3390/ijms22126208 PMC8227693

16. Safari MH Long non-coding RNAs and therapeutic resistance in multiple myeloma: from molecular insights to clinical applications Clin. Exp. Med. 2026 26 1 114 10.1007/s10238-025-02015-0 41547689 PMC12847185 Safari, M. H. et al. Long non-coding RNAs and therapeutic resistance in multiple myeloma: from molecular insights to clinical applications. Clin. Exp. Med. 26 (1), 114 (2026). 41547689 10.1007/s10238-025-02015-0 PMC12847185

17. Maleknia S Sex-specific dysregulation of exosomal non-coding RNAs drives multiple myeloma progression Blood Cancer J. 2025 15 1 162 10.1038/s41408-025-01362-1 41102192 PMC12533198 Maleknia, S. et al. Sex-specific dysregulation of exosomal non-coding RNAs drives multiple myeloma progression. Blood Cancer J. 15 (1), 162 (2025). 41102192 10.1038/s41408-025-01362-1 PMC12533198

18. Wu K Exosomal miR-19a and IBSP cooperate to induce osteolytic bone metastasis of estrogen receptor-positive breast cancer Nat. Commun. 2021 12 1 5196 10.1038/s41467-021-25473-y 34465793 PMC8408156 Wu, K. et al. Exosomal miR-19a and IBSP cooperate to induce osteolytic bone metastasis of estrogen receptor-positive breast cancer. Nat. Commun. 12 (1), 5196 (2021). 34465793 10.1038/s41467-021-25473-y PMC8408156

19. Wang Q Shen X Chen Y Chen J Li Y Osteoblasts-derived exosomes regulate osteoclast differentiation through miR-503-3p/Hpse axis Acta Histochem. 2021 123 7 151790 10.1016/j.acthis.2021.151790 34592492 Wang, Q., Shen, X., Chen, Y., Chen, J. & Li, Y. Osteoblasts-derived exosomes regulate osteoclast differentiation through miR-503-3p/Hpse axis. Acta Histochem. 123 (7), 151790 (2021). 34592492 10.1016/j.acthis.2021.151790

20. Alipoor SD Chang H Exosomal miRNAs in the Tumor Microenvironment of Multiple Myeloma Cells 2023 12 7 1030 10.3390/cells12071030 37048103 PMC10092980 Alipoor, S. D. & Chang, H. Exosomal miRNAs in the tumor microenvironment of multiple myeloma. Cells 12 (7), 1030 (2023). 37048103 10.3390/cells12071030 PMC10092980

21. Ma J Li TF Han XW Yuan HF Downregulated MEG3 contributes to tumour progression and poor prognosis in oesophagal squamous cell carcinoma by interacting with miR-4261, downregulating DKK2 and activating the Wnt/β-catenin signalling Artif. Cells Nanomed. Biotechnol. 2019 47 1 1513 1523 10.1080/21691401.2019.1602538 30990378 Ma, J., Li, T. F., Han, X. W. & Yuan, H. F. Downregulated MEG3 contributes to tumour progression and poor prognosis in oesophagal squamous cell carcinoma by interacting with miR-4261, downregulating DKK2 and activating the Wnt/β-catenin signalling. Artif. Cells Nanomed. Biotechnol. 47 (1), 1513–1523 (2019). 30990378 10.1080/21691401.2019.1602538

22. Jiao G Therapeutic Suppression of miR-4261 Attenuates Colorectal Cancer by Targeting MCC Mol. Ther. Nucleic Acids 2017 8 36 45 10.1016/j.omtn.2017.05.010 28918036 PMC5480279 Jiao, G. et al. Therapeutic suppression of miR-4261 attenuates colorectal cancer by targeting MCC. Mol. Ther. Nucleic Acids . 8 , 36–45 (2017). 28918036 10.1016/j.omtn.2017.05.010 PMC5480279

23. Inoue K Ng C Xia Y Zhao B Regulation of Osteoclastogenesis and Bone Resorption by miRNAs Front. Cell. Dev. Biol. 2021 9 651161 10.3389/fcell.2021.651161 34222229 PMC8249944 Inoue, K., Ng, C., Xia, Y. & Zhao, B. Regulation of osteoclastogenesis and bone resorption by miRNAs. Front. Cell. Dev. Biol. 9 , 651161 (2021). 34222229 10.3389/fcell.2021.651161 PMC8249944

24. Pourhanifeh MH MicroRNAs and exosomes: Small molecules with big actions in multiple myeloma pathogenesis IUBMB Life 2020 72 3 314 333 10.1002/iub.2211 31828868 Pourhanifeh, M. H. et al. MicroRNAs and exosomes: Small molecules with big actions in multiple myeloma pathogenesis. IUBMB Life . 72 (3), 314–333 (2020). 31828868 10.1002/iub.2211

25. Gong N Keratinocytes-derived exosomal miRNA regulates osteoclast differentiation in middle ear cholesteatoma Biochem. Biophys. Res Commun 2020 525 2 341 347 10.1016/j.bbrc.2020.02.058 32093888 Gong, N. et al. Keratinocytes-derived exosomal miRNA regulates osteoclast differentiation in middle ear cholesteatoma. Biochem. Biophys. Res Commun . 525 (2), 341–347 (2020). 32093888 10.1016/j.bbrc.2020.02.058

26. Guo F Adipocyte-derived PAMM suppresses macrophage inflammation by inhibiting MAPK signalling Biochem. J. 2015 472 3 309 318 10.1042/BJ20150019 26438880 PMC4754088 Guo, F. et al. Adipocyte-derived PAMM suppresses macrophage inflammation by inhibiting MAPK signalling. Biochem. J. 472 (3), 309–318 (2015). 26438880 10.1042/BJ20150019 PMC4754088

27. Lee KI Chen JH Chen KH Osteoporosis After Menopause and After Drug Therapy: The Molecular Mechanism of Bone Loss and Its Treatment Int. J. Mol. Sci. 2026 27 2 641 10.3390/ijms27020641 41596293 PMC12841279 Lee, K. I., Chen, J. H. & Chen, K. H. Osteoporosis after menopause and after drug therapy: The molecular mechanism of bone loss and its treatment. Int. J. Mol. Sci. 27 (2), 641 (2026). 41596293 10.3390/ijms27020641 PMC12841279

28. Wen P Liu L Functions of Macrophages, T Cells, and Neutrophils in the Synovial Microenvironment of Osteoarthritis J. Inflamm. Res. 2025 18 15671 15686 10.2147/JIR.S563253 41243987 PMC12619582 Wen, P. & Liu, L. Functions of macrophages, T cells, and neutrophils in the synovial microenvironment of osteoarthritis. J. Inflamm. Res. 18 , 15671–15686 (2025). 41243987 10.2147/JIR.S563253 PMC12619582

29. Ren X Antioxidant Gene Signature Impacts the Immune Infiltration and Predicts the Prognosis of Kidney Renal Clear Cell Carcinoma Front. Genet. 2021 12 721252 10.3389/fgene.2021.721252 34490047 PMC8416991 Ren, X. et al. Antioxidant gene signature impacts the immune infiltration and predicts the prognosis of kidney renal clear cell carcinoma. Front. Genet. 12 , 721252 (2021). 34490047 10.3389/fgene.2021.721252 PMC8416991

30. Jaiswal V Park M Lee HJ Comparative Transcriptome Analysis of the Expression of Antioxidant and Immunity Genes in the Spleen of a Cyanidin 3-O-Glucoside-Treated Alzheimer’s Mouse Model Antioxid. (Basel) 2021 10 9 1435 10.3390/antiox10091435 PMC8472539 34573067 Jaiswal, V., Park, M. & Lee, H. J. Comparative transcriptome analysis of the expression of antioxidant and immunity genes in the spleen of a cyanidin 3-O-glucoside-treated Alzheimer’s mouse model. Antioxid. (Basel) . 10 (9), 1435 (2021). 10.3390/antiox10091435 PMC8472539 34573067

31. Oh CK FAM213A is linked to prognostic significance in acute myeloid leukemia through regulation of oxidative stress and myelopoiesis Hematol. Oncol. 2020 38 3 381 389 10.1002/hon.2728 32124993 Oh, C. K. et al. FAM213A is linked to prognostic significance in acute myeloid leukemia through regulation of oxidative stress and myelopoiesis. Hematol. Oncol. 38 (3), 381–389 (2020). 32124993 10.1002/hon.2728

32. Anloague A A novel CCL3-HMGB1 signaling axis regulating osteocyte RANKL expression in multiple myeloma Haematologica 2025 110 4 952 966 10.3324/haematol.2024.286484 39605211 PMC11959238 Anloague, A. et al. A novel CCL3-HMGB1 signaling axis regulating osteocyte RANKL expression in multiple myeloma. Haematologica 110 (4), 952–966 (2025). 39605211 10.3324/haematol.2024.286484 PMC11959238

33. Lungu O Toscani D Giuliani N Mechanistic insights into bone destruction in multiple myeloma: Cellular and molecular perspectives J. Bone Oncol. 2025 51 100668 10.1016/j.jbo.2025.100668 40124903 PMC11928850 Lungu, O., Toscani, D. & Giuliani, N. Mechanistic insights into bone destruction in multiple myeloma: Cellular and molecular perspectives. J. Bone Oncol. 51 , 100668 (2025). 40124903 10.1016/j.jbo.2025.100668 PMC11928850

34. Yu T Polycomb-like Protein 3 Induces Proliferation and Drug Resistance in Multiple Myeloma and Is Regulated by miRNA-15a Mol. Cancer Res. 2020 18 7 1063 1073 10.1158/1541-7786.MCR-19-0852 32312841 Yu, T. et al. Polycomb-like protein 3 induces proliferation and drug resistance in multiple myeloma and is regulated by miRNA-15a. Mol. Cancer Res. 18 (7), 1063–1073 (2020). 32312841 10.1158/1541-7786.MCR-19-0852

35. Qin Z The use of THP-1 cells as a model for mimicking the function and regulation of monocytes and macrophages in the vasculature Atherosclerosis 2012 221 1 2 11 10.1016/j.atherosclerosis.2011.09.003 21978918 Qin, Z. The use of THP-1 cells as a model for mimicking the function and regulation of monocytes and macrophages in the vasculature. Atherosclerosis 221 (1), 2–11 (2012). 21978918 10.1016/j.atherosclerosis.2011.09.003

36. DU Y TIPE2 regulates periodontal inflammation by inhibiting NF-κB p65 phosphorylation J. Appl. Oral Sci. 2023 31 e20230162 10.1590/1678-7757-2023-0162 37493703 PMC10382077 DU, Y. et al. TIPE2 regulates periodontal inflammation by inhibiting NF-κB p65 phosphorylation. J. Appl. Oral Sci. 31 , e20230162 (2023). 37493703 10.1590/1678-7757-2023-0162 PMC10382077

37. Salamanna F Sustainable innovation with a method based on peripheral mononuclear cells to screen, monitor and stratify the population at risk of osteoporosis and fractures - a multicenter cross-sectional trial protocol Front. Endocrinol. (Lausanne) 2025 16 1647800 10.3389/fendo.2025.1647800 41113706 PMC12527827 Salamanna, F. et al. Sustainable innovation with a method based on peripheral mononuclear cells to screen, monitor and stratify the population at risk of osteoporosis and fractures - a multicenter cross-sectional trial protocol. Front. Endocrinol. (Lausanne) . 16 , 1647800 (2025). 41113706 10.3389/fendo.2025.1647800 PMC12527827

38. Freeman FE Localized Nanoparticle-Mediated Delivery of miR-29b Normalizes the Dysregulation of Bone Homeostasis Caused by Osteosarcoma whilst Simultaneously Inhibiting Tumor Growth Adv. Mater. 2023 35 23 e2207877 10.1002/adma.202207877 36994935 Freeman, F. E. et al. Localized nanoparticle-mediated delivery of miR-29b normalizes the dysregulation of bone homeostasis caused by osteosarcoma whilst simultaneously inhibiting tumor growth. Adv. Mater. 35 (23), e2207877 (2023). 36994935 10.1002/adma.202207877

39. Li ZH High-dose PMA with RANKL and MCSF induces THP 1 cell differentiation into human functional osteoclasts in vitro Mol. Med. Rep. 2017 16 6 8380 8384 10.3892/mmr.2017.7625 28983613 Li, Z. H. et al. High-dose PMA with RANKL and MCSF induces THP 1 cell differentiation into human functional osteoclasts in vitro. Mol. Med. Rep. 16 (6), 8380–8384 (2017). 28983613 10.3892/mmr.2017.7625

40. Yao H Guizhi Shaoyao Zhimu granules attenuate bone destruction in mice with collagen-induced arthritis by promoting mitophagy of osteoclast precursors to inhibit osteoclastogenesis Phytomedicine 2023 118 154967 10.1016/j.phymed.2023.154967 37490802 Yao, H. et al. Guizhi Shaoyao Zhimu granules attenuate bone destruction in mice with collagen-induced arthritis by promoting mitophagy of osteoclast precursors to inhibit osteoclastogenesis. Phytomedicine 118 , 154967. (2023). 37490802 10.1016/j.phymed.2023.154967

41. Norton A Thieu K Baumann CW Lowe DA Mansky KC Estrogen regulation of myokines that enhance osteoclast differentiation and activity Sci. Rep. 2022 12 1 15900 10.1038/s41598-022-19438-4 36151243 PMC9508086 Norton, A., Thieu, K., Baumann, C. W., Lowe, D. A. & Mansky, K. C. Estrogen regulation of myokines that enhance osteoclast differentiation and activity. Sci. Rep. 12 (1), 15900 (2022). 36151243 10.1038/s41598-022-19438-4 PMC9508086