Aberrant NSUN2-mediated m5C modification of exosomal LncRNA MALAT1 induced RANKL-mediated bone destruction in multiple myeloma.

Background

Multiple myeloma (MM) is an incurable plasma cell malignancy in which osteoclast-mediated bone destruction is a hallmark of active disease 1 . In a retrospective study, approximately 80% of patients with myeloma suffered pathological fractures, and 90% developed osteolytic bone lesions during the course of the disease 2 , which was associated with poor survival and mortality 3 . The pathogenesis of myeloma-related bone disease (MBD) is an imbalance and uncoupling of bone remodeling processes resulting from dysregulation of the bone marrow microenvironment 4 , 5 , increased osteoclast formation and activity, and osteoblast inhibition 6 . The direct and indirect interactions between MM cells and bone mesenchymal stem cells, osteoclasts (OCs), osteoblasts, osteocytes, and immune cells are also involved in the complex pathogenesis event 7 . Tumor-derived exosomes have received extensive research attention over the past decade, which are small extracellular vesicles loaded with proteins, DNA, mRNAs, miRNAs, and LncRNAs 8 – 10 . They originate from intracellular multivesicular bodies and are released into the extracellular microenvironment by exocytosis, acting as signal transducers or messengers in cell–cell communication for the changing phenotypes in recipient cells 11 , 12 . MALAT1 is a widely studied lncRNA in cancer and is important for cancer-related pathway regulation. In MM, MALAT1 is implicated in tumorigenesis and progression 13 , impacting tumor proliferation, invasion, metastasis, and migration 14 , 15 , but its role in bone destruction remains unclear.

RNA modification has been reported to play a vital role in the regulation of post-transcriptional gene expression in a large variety of cancers. N6-methyladenosine (m6A) and 5-methylcytosine (m5C) methylation are two major types of RNA methylation. However, m6A has been extensively studied in MM, but m5C is rare. M5C is a modification form of the fifth N of cytosine and is widely distributed in various RNAs including mRNA, tRNA, rRNA, and LncRNA, which affects RNA metabolism, such as RNA stability, recognition, export, and translation 16 , 17 . For example, aberrant NSUN2-mediated m5C modification of H19 LncRNA was associated with poor differentiation of hepatocellular carcinoma 18 . Yan et al. reported that lnc FOXC2-AS1 stabilized FOXC2 mRNA via association with NSUN2 in gastric cancer cells 19 . Here, we set out to investigate the contributions of MM-derived exosomal LncRNAs to OCs differentiation and the possibility that LncRNA expression might be modulated by m5C methylation.

Results

MM exosomes promote bone lesion formation

To test whether exosomes derived from MM cells promote bone lesions, we isolated exosomes from the supernatant of U266, MM1S, and RPMI8226 cells. TEM was used to detect exosomes with a morphological size of 30–150 nm and a cup‐plate‐like structure (Fig. 1A ). In addition, NTA proved that the exosomes were approximately 100 nm in diameter (Fig. 1B ). Laser scanning confocal microscopy was used to demonstrate the uptake of labeled exosomes by RAW264.7 cells (Fig. 1C ). To observe the effect of exosomes on osteoclastogenesis, we performed TRAP staining. OCs induced from RAW264.7 cells were quantified. Exosomes significantly increased the number of OCs (Fig. 1D ). GW4869 is a well-recognized exosome inhibitor that reduces exosome release 20 . The number of TRAP-positive OCs was decreased in co-cultured GW4869-treated culture-conditioned medium (CCM) of U266 and MM1S cells compared with WT CCM (Fig. 1D ). Consistently, rescue experiments also confirmed that OC differentiation was augmented in the presence of exosomes (Fig. 1D ). The above results suggested that exosomes derived from MM cells could be taken up by RAW264.7 cells, promoting their differentiation into OCs.

MM exosomes promoted RANKL expression and induced AKT and MAPK activation

To investigate the signaling pathways involved in the positive effects of exosomes on OC differentiation, we examined RANKL expression and the phosphorylation of AKT and MAPKs, JNK, ERK, and p38 in RAW264.7 cells. U266-CCM and exosome extract treatment increased the level of RANKL expression and phosphorylated AKT and MAPKs. Pretreatment with GW4869, which impairs U266 exosome synthesis, ameliorated the expression of RANKL and the activation of downstream pathways, while exosomes reversed these effects (Fig. 2A–D ). Furthermore, exosomes derived from MM1S and RPMI8226 cells similarly enhanced RANKL expression in RAW264.7 cells (Fig. 2E, F ). These results suggested that RANKL and its downstream AKT and MAPK signaling pathways were involved in the promotion of OC differentiation by MM-derived exosomes.

MALAT1 was identified as the most highly expressed lncRNA in U266 exosomes and upregulated RANKL

The U266 cell line was used for the subsequent mechanism exploration. To investigate the information transferred by MM-derived exosomes, high‐throughput sequencing was conducted on U266 exosomes. MALAT1 displayed the highest expressed level among exosomal lncRNAs (Supplementary Table 1 ). Sequence similarity searches with BLAST were used to prove the homology of human and murine MALAT1 . Subsequently, we assessed MALAT1 expression in MM cohorts. Intriguingly, the expression of MALAT1 was significantly elevated in MM cells compared with normal plasma (NP) cells in the GSE24080 and GSE5900 patient cohorts (Fig. 3A ). FISH examination showed an obvious abundance of MALAT1 within the bone marrow (BM) of newly diagnosed MM (NDMM) patients (Fig. 3B ). To investigate whether MALAT1 in exosomes can be taken up by OCs, PKH26 dye and GFP were used to label exosomes and MALAT1, respectively, that were then co-cultured with RAW264.7 cells. As presented in Fig. 3C , the majority of the recipient cells exhibited a yellow signal in the cytoplasm under the confocal microscope, suggesting that MALAT1 could be packaged into exosomes and taken up by RAW264.7 cells, which was also confirmed by qRT-PCR (Fig. 3D ).

To investigate the exact function of MALAT1 in RAW264.7 cells, we transfected si-MALAT1 or its corresponding negative control (siNC) into U266 cells. Exosomes were then isolated from the culture media of these transfection groups and used to treat RAW264.7 cells. Exo-siMALAT1 abolished the upregulation of MALAT1 , and RANKL expression caused by exosomes in RAW264.7 cells (Fig. 3D–F ). TRAP-positive OCs were also restrained, while the overexpression of MALAT1 resulted in the opposite effect (Fig. 3D–G ). We also directly influenced MALAT1 expression in RAW264.7 cells, and qRT-PCR and TRAP assays showed similar results (Supplementary Fig. 1 ). Taken together, these data proved that MALAT1 was upregulated in MM patients and cell lines and could be transmitted to OCs via exosomes, which in turn promoted bone lesions and upregulated RANKL expression.

NSUN2 mediates MALAT1 expression in an m5C‑dependent manner

To investigate the reason for MALAT1 upregulation in MM cells, RNA pull-down was performed. The results showed that NSUN2, the main methyltransferase of m5C methylation, was markedly pulled down by biotin-labeled MALAT1 but not by its antisense transcript (Fig. 4A ). To test whether and how m5C modification regulates MALAT1 expression, we transfected U266 cells with siRNA targeting NSUN2. Knockdown of NSUN2 significantly reduced the m5C levels of MALAT1 (Fig. 4B ), leading to a notable reduction in MALAT1 levels (Fig. 4C ). RIP assay demonstrated that MALAT1 could be significantly enriched by NSUN2 antibody compared to negative control IgG (Fig. 4D ), which was consistent with the results of the pull-down assay. The RNA stability assay showed that depletion of NSUN2 resulted in the destabilization of MALAT1 (Fig. 4E ), suggesting that m5C regulates the stability of MALAT1 . YBX1 is a well-characterized m5C reader that recognizes and stabilizes m5C-modified RNA 21 . Hence, we explored whether YBX1 plays a critical role in the NSUN2-mediated stabilization of MALAT1 . A rescue experiment showed that MALAT1 expression and stability increased by NSUN2 were reversed when transfected with YBX1 siRNA (Fig. 4F, G ). Furthermore, the mRNA levels of NSUN2 and YBX1 in the GEO cohort were significantly increased in CD138+ cells ( n = 559) from MM patients compared to healthy controls ( n = 22) (Fig. 4H ). The IHC images also showed the apparent abundance of both in the BM of NDMM patients (Fig. 4I ). Together, our findings indicate that NSUN2 and YBX1 were upregulated in MM cells and that NSUN2-mediated m5C modification of MALAT1 increases its stability in a YBX1-dependent manner.

MALAT1 , NSUN2, and YBX1 were positively correlated with bone destruction features in MM patients

To further verify the relationship among MALAT1 , NSUN2, YBX1, and bone lesions, and evaluate their prognostic value, BM samples were collected from 36 NDMM patients. The patients’ features are shown in Supplementary Table 2 . FISH and IHC were used to evaluate the levels of MALAT1 specifically in MM cells, NSUN2, and YBX1 in the BM tissues, respectively. Pearson correlation analyses were performed. These analyses revealed that MALAT1 levels were strongly correlated with the number of bone lesions ( R 2 = 0.81, P < 0.001) (Fig. 5A ) and corrected serum calcium levels ( R 2 = 0.705, P < 0.001) (Fig. 5B ), as well as to RANKL ( R 2 = 0.799, P < 0.001), NSUN2 ( R 2 = 0.758, P < 0.001), and YBX1 ( R 2 = 0.748, P < 0.001) levels (Fig. 5C, D ). NSUN2 and YBX1 showed a high correlation with the number of bone lesions ( P < 0.001) and RANKL levels ( P < 0.001) (Fig. 5E, F ), and a moderate correlation with corrected serum calcium levels ( P < 0.001) (Fig. 5G ).

Additionally, the area under the receiver operating characteristic (ROC) curve (AUC) of MALAT1 , NSUN2, and YBX1 for overall survival was 0.876, 0.873, and 0.848, respectively (Fig. 5H ). MM patients were divided into high- and low-risk subgroups on the basis of the best Youden Index of MALAT1 , NSUN2, or YBX1 as the cut-off value. Figure 5I showed the expression levels of MALAT1 , NSUN2, and YBX1 in patients from different subgroups. K-M analyses demonstrated that compared to MM patients in the low-risk subgroup, those in the high-risk subgroup possessed worse OS (Fig. 5J ). Univariate Cox regression analyses showed that the expression levels of MALAT1 , NSUN2, and YBX1 were significantly negatively correlated with the survival rate of patients (Table 1 ). Together, our findings indicate that MALAT1 , NSUN2, and YBX1 may serve as independent prognostic factors and their expression was positively correlated with bone destruction features in MM patients.

MALAT1 sponges miR-17 to promote RANKL expression in MM

We then explored the potential mechanisms between MALAT1 and RANKL. It has been reported that lncRNAs function as miRNA sponges to regulate their targeted mRNA expression 22 . Thus, we performed bioinformatic analysis to predict the potential binding miRNAs of MALAT1 in the miRcode, miRanda, and RNAhybrid and the potential binding miRNAs of RANKL in TargetScan databases. Total 5 miRNAs (miR-17, miR-34a, miR-146a, miR-125a and miR-101-1) were selected. Exosomes were isolated from U266 cells with MALAT1 knockdown or overexpression, and the effect of exosomal MALAT1 on miRNA expression in RAW264.7 cells was detected. As expected, MALAT1 knockdown and overexpression increased or reduced the levels of all five miRNAs that were downregulated by exosomes (Fig. 6A ). Among them, miR-17 showed the greatest significance and stability. Therefore, miR-17 was chosen for the following experiments. Next, we investigated whether MALAT1 directly binds miR-17. Fifteen binding sites were predicted between MALAT1 and miR-17 (Supplementary Table 3 ), and seven luciferase reporter plasmids encoding truncated MALAT1 were constructed. We transfected miR-17 mimics into HEK293T cells and observed that luciferase activity from the different MALAT1-WT reporters was significantly lower than that from the control except for MALAT1-4 WT (Fig. 6B ). In addition, the protein level and mRNA expression level of RANKL were unchanged by co-transfection with miR-17 and MALAT1 compared to levels after treatment with exosomes alone (Fig. 6C, D ). A similar result was obtained for OC differentiation based on TRAP assays (Fig. 6E ). These data indicated that MALAT1 functions as a sponge for miR-17 to promote RANKL expression and then increase OC differentiation.

MM exosomal MALAT1 impairs osteogenesis and exacerbates bone lesions in an MM mouse model

The ability of MM exosomes to destroy bone was then verified in vivo. NOG mice injected with the U266 cell line were randomly divided into three separate treatment groups: 200 μg exosomes from U266-shNC cells, 200 μg exosomes from U266-shMALAT1 cells, and equal dose PBS. After 4 weeks, mice were euthanized, and bone samples were collected. The extent of bone disease in each model was assessed by micro-CT. Parametric analysis using micro-CT revealed a significant decrease in the bone volume (bone volume/tissue volume), trabecular thickness, and trabecular number by exosomes. In contrast, the trabecular separation, bone-specific surface (bone surface/bone volume), and cranial suture were increased by exosomes (Fig. 7A–F ). However, when injecting exosomes from U266-shMALAT1 cells, we observed an increased bone volume level compared with that of the exosome group. Similarly, both trabecular thickness and trabecular number were also increased in the Exo-shMALAT1 group, although no significant difference in some groups was observed. Simultaneously, the exosome-increased trabecular separation, bone-specific surface, and cranial suture were decreased to the level of the model group (Fig. 7A–F ). In addition, H&E staining (Fig. 7G ) and TRAP staining (Fig. 7H ) of femur tissue indicated more severe bone destruction and more OCs in the exosome group than in the model and Exo-shMALAT1 groups. These results suggest that injection of exosomes increases MM-induced OC differentiation and bone lesions, in part through MALAT1 .

Exosomal MALAT1 promoted RANKL expression and induced AKT and MAPK activation in vivo

The promotion effect of MALAT1 on the phosphorylation of AKT and MAPKs was further confirmed in the following animal tests. As expected, the expression level of MALAT1 was increased in the femur tissue of the group tail vein-injected with exosomes from U266-shNC cells but not MALAT1 knockdown cells (Fig. 8A ). The IHC images showed that compared with those of the model group, obviously more RANKL and phosphorylation of AKT, JNK, ERK, and p38 were observed in the exosome-injected mice (Exo-shNC group). The expression of RANKL and phosphorylation of AKT, JNK, ERK, and p38 were significantly decreased by knockdown of MALAT1 in the Exo-shMALAT1 group compared with the Exo-shNC group (Fig. 8B and Supplementary Fig. 2 ). Consistent with the WB results, IHC data suggested that MALAT1 promoted OC differentiation in mice by positively regulating RANKL expression and its downstream AKT and MAPK signaling pathways.

Discussion

Bone destruction is the hallmark of MM, significantly affecting patients’ quality of life. Bisphosphonates are the treatment of choice for MBD, but osteolytic lesions are still a critical issue in the current management of MM patients 23 . Emerging evidence indicates that exosomes function as intercellular communicators within the tumor microenvironment, regulating OC differentiation 24 , 25 . In this study, we innovatively combined aberrant m5C RNA modification, which is involved in the regulation of cell proliferation and metastasis formation and is upregulated in multiple cancers 26 , 27 , but no research is executed in MM, with exosomal LncRNA, to investigate how myeloma cell-derived exosomal LncRNA effect OC differentiation and the role m5C plays in it.

Here, in vitro experiments showed that MM cell-secreted exosomes promote OC differentiation. In addition, they increased the levels of RANKL expression and phosphorylated AKT and MAPKs, whereas pretreatment with exosome inhibitors ameliorated these effects. RANKL is one of the cytokines needed to regulate OCs. Its binding to RANK stimulates the signals necessary for OC differentiation and for the survival of multinucleated OCs 28 . Downstream of RANKL signaling, the activation of the AKT and MAPK pathways are vital in regulating cell proliferation, differentiation, and apoptosis, and their phosphorylation induces the expression of downstream molecules to drive OC formation 29 – 31 . These results suggested that MM-derived exosomes promote OC differentiation by stimulating RANKL expression and its downstream AKT and MAPK signaling pathways.

The use of exosomes as biological delivery vehicles is of considerable interest for cancer due to a stable environment for a variety of RNAs. MALAT1 is a highly conserved LncRNA that has been widely studied in cancers and is associated with poor prognosis in MM patients 32 , 33 . However, its function through exosomes on bone lesions of MM has not been previously described. Here, we identified MALAT1 as a risk factor in patients with MBD. Its expression level showed a positive correlation with clinical features of bone destruction and the level of RANKL expression. Furthermore, MALAT1 was the highest expressed lncRNA in exosomes derived from MM cells and could be transported to recipient RAW264.7 cells through exosomes. Mechanistic studies in vitro and in vivo MM models directly demonstrated that overexpression of exosomal MALAT1 induced OC differentiation, and bone lesions formation, whereas knockdown of exosomal MALAT1 had the opposite effects.

Notably, LncRNAs mainly target miRNAs in regulating MM progression. In this study, MALAT1 -overexpressing exosomes markedly suppressed the expression of miR-17, miR-34a, miR-125a, miR-146a, and miR-101-1, and knockdown of MALAT1 resulted in the opposite effects. In addition, our data on truncated MALAT1 and RANKL dual-luciferase assays demonstrated that miR-17 can directly bind with MALAT1 through almost 14 binding sites and directly bind with RANKL . Moreover, qRT-PCR, WB, and TRAP assays indicated that MALAT1 upregulated RANKL expression via functions as a sponge for miR-17, thereby regulating OC differentiation. MiR-17 belongs to the miR-17-92 cluster, is oncogenic, and is a valuable therapeutic target in c-MYC (MYC) -driven malignancies such as MM 34 . In bone diseases, previous studies have shown that miR-17 promotes osteogenic differentiation through Smurf1 -mediated regulation of Runx2 35 , which is consistent with our study that miR-17 inhibits osteoclast differentiation. However, our study focused more on MALAT1 , and the role of miR-17 in MM bone disease should be further explored separately.

In recent years, accumulating evidence has demonstrated that dysregulation of RNA modifications is implicated in tumorigenesis, progression, and drug response 36 – 38 . Research on RNA methylation mainly focuses on the m6A modification of mRNA, but very little is known about the m5C modification. We found MALAT1 could be directly bound with NSUN2, an m5C methyltransferase of mammalian RNA, which was reported to be highly expressed in multiple tumors 26 , 39 . Depletion of NSUN2 led to reduced methylation of MALAT1 , which was also accompanied by decreased expression of MALAT1 . Many previous reports indicated that NSUN2-mediated RNA m5C modification may mainly exert its function by affecting the translation of RNA rather than altering RNA levels 40 , 41 . A recent study showed that in human bladder urothelial carcinoma (UBC), NSUN2-mediated m5C modification stabilizes oncogene RNA through the “reader” YBX1, and hypermethylated m5C levels correlate with oncogene mRNA overexpression in UBC cells 21 . NSUN2 may also play a similar role in LncRNAs 18 . Here, we showed that NSUN2 deficiency significantly decreased the half-life of MALAT1 . A rescue experiment demonstrated that MALAT1 expression and stability increased by NSUN2 were reversed when transfected with YBX1 siRNA. Clinical data analysis also indicated that NSUN2 and YBX1 may serve as independent prognostic factors in MM.

Undeniably, this study has several limitations. The first is that the experimental results need to be verified on more myeloma cell lines, preferably primary MM plasma cells. In addition, due to limitations of experimental conditions, the experiments were conducted on a murine macrophage cell line, which can differentiate towards OCs. Although sequence similarity searches prove the homology of human and murine MALAT1 , future studies should include human OCs and human osteoclast progenitor cells to better validate the findings and ensure their relevance to human physiology. Furthermore, regarding the limitations on healthy control sample size, future studies should include more controls to ensure a comparable sample size for experimental groups.

Conclusion

In conclusion, our study demonstrates that the myeloma cell-derived exosomal MALAT1 is involved in bone lesion formation. Functionally and mechanistically, NSUN2 and YBX1-mediated m5C modifications upregulate MALAT1 expression in MM cells, which can be transported to OCs via exosomes and promote OCs activation through the miR-17/ RANKL axis. Additionally, this study explored the role of m5C modification in MM and proved the existence of m5C modification on LncRNA, which provides a fresh perspective for studying the role of LncRNA, but the clinical sample size of this study is insufficient, and further exploration is still needed.

Methods

Patients and clinical specimens

BM tissues and clinical parameters were obtained from NDMM patients and healthy donors at the Department of Hematology, Affiliated Hospital of Shandong University of Traditional Chinese Medicine and Department of Oncology and Hematology, the Second Affiliated Hospital of Shandong University of Traditional Chinese Medicine from 2020 to 2023. Written informed consent was obtained from all participants. This study was approved by the ethics committee of the Affiliated Hospital of Shandong University of Traditional Chinese Medicine (2020) ethical review No. (010)-KY) and the Second Affiliated Hospital of Shandong University of Traditional Chinese Medicine (2023 ethical review-KY-001). All ethical regulations relevant to human research participants were followed. Immediately after removal, each tissue sample was fixed in 4% formalin and embedded in paraffin.

Cells and exposure

Human MM cell lines U266, MM1S, and RPMI8226, along with RAW264.7 macrophages and 293T cells, were purchased from the American Type Culture Collection (ATCC). The human MM cell lines were cultured in RPMI-1640 medium supplemented with 10% fetal bovine serum (FBS) (Clark, USA), 100 U/mL penicillin, and 100 μg/mL streptomycin (HyClone, USA). RAW264.7 macrophages and 293 T cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) (Gibco, USA), also supplemented with 10% FBS, 100 U/mL penicillin, and 100 μg/mL streptomycin. All cells were maintained at 37 °C in a humidified chamber of 5% CO 2 .

To investigate the molecular mechanisms, GW4869 (Selleck, USA) was used to assess the impact of exosomes derived from myeloma cells on OCs activation. GW4869 is a potent and specific non-competitive inhibitor of membrane-neutral sphingomyelinase, known to significantly reduce exosome release while exhibiting minimal toxicity to cultured cells 20 . The inhibitor was initially dissolved in DMSO. Myeloma cells were pretreated with GW4869 at a concentration of 5 μM for 24 h. Control cells received an equivalent volume of DMSO (vehicle control) under the same conditions. Following the treatment, the supernatant was collected for subsequent experiments.

Isolation, validation, and high-throughput sequencing of exosomes

After U266 cell lines were incubated with exosome-depleted FBS for 48 hours, the conditioned medium underwent differential centrifugation (800 × g for 5 min; 2000 × g for 20 min) (Beckman Optima L-100XP). A 0.22-μm filter (Merck Millipore) was used to remove detached desquamated vesicles and other large vesicles. Then, the cell-free medium was centrifuged at 20,000 × g for 70 min. The exosomes were resuspended in PBS and centrifuged at 100,000 × g again for 1.5 h. Transmission electron microscopy (TEM) and nanoparticle tracking analysis (NTA) analyses were conducted to confirm the successful separation of exosomes.

High-throughput sequencing was carried out by Shandong JieluoXuan Biological Technology Co., Ltd. (Jinan, China). In brief, ribosomal RNA was removed from the total RNA of each sample using the Ribo-Zero rRNA Removal Kit (Illumina, USA) prior to RNA-seq library construction. RNA libraries were then prepared following the manufacturer’s protocol for the TruSeq Stranded Total RNA Library Prep Kit (Illumina). The quantity and quality of the RNA libraries were assessed with a BioAnalyzer 2100 system (Agilent Technologies, USA). Subsequently, the libraries were reverse transcribed into cDNAs, captured on Illumina flow cells, amplified in situ, and sequenced on a HiSeq 4000 sequencing system (Illumina) to generate 150-bp paired-end reads.

Transfection

For transfection, si-MALAT1/si-NC, si-YBX1/si-NC, sh-MALAT1/sh-NC, MALAT1 overexpression vector/control vector, NSUN2 overexpression vector/control, and miR-17 mimics, inhibitors, and their respective controls were obtained from GENECHEM (China). Lipofectamine 2000 (Invitrogen, USA) was used following the manufacturer’s instructions.

U266 cells were seeded in 6-well plates at 5 × 10 5 cells per well and grown overnight to 70-80% confluency. Transfection complexes were prepared by diluting 5 μL Lipofectamine 2000 in 250 μL Opti-MEM (Gibco, USA) and incubating for 5 min. Separately, 50 nM siRNA or 2 μg plasmid DNA was diluted in 250 μL Opti-MEM. These were combined, mixed, and incubated for 20 min to form transfection complexes, then added dropwise to the cells. After 6 hours, the medium was replaced with fresh RPMI-1640 with 10% FBS and antibiotics.

After stimulation, RAW264.7 cells were seeded in 6-well plates at 2 × 10 5 cells per well, and cultured overnight. For transfection, 50 nM miR-17 mimics, inhibitors, or their controls were diluted in 250 μL Opti-MEM. These were combined with diluted Lipofectamine 2000, mixed, and incubated for 20 min to form transfection complexes, then added dropwise to the cells. After 6 h, the medium was replaced with fresh DMEM with 10% FBS and antibiotics.

PKH26 staining

100 μL of PKH26 (Sigma, USA) was mixed with 100 μL of Diluent C. Then, 50 μg of exosomes derived from MM cells was added to the mixture and incubated for 3 min. Then, 200 μL of 0.5% exosome-depleted FBS was added and incubated for 5 min to stop the reaction. Then, labeled exosomes were washed twice with PBS and co-cultured with recipient cells in one well of a 6-well plate for 12 h prior to imaging. Imaging was performed using a fluorescence microscope with appropriate filters for PKH26 (red) and DAPI (blue).

RNA fluorescent in situ hybridization (RNA-FISH)

The expression levels and cellular localization of MALAT1 in MM BM samples, mouse femoral tissues, and RAW264.7 cells were examined by FISH analysis. A detailed description of FISH analysis was conducted as previously reported 20 . GFP-labeled MALAT1 probes and CY3-labeled MALAT1 probes were obtained from Beijing Dingguo Changsheng Biotech Co., Ltd., were synthesized for FISH analysis. All fluorescence images were taken with a confocal laser scanning microscope (Leica, Germany). The probe sequence is listed in Supplementary Information: Supplementary Table 4 .

Tartrate-resistant acidic phosphatase (TRAP) staining

RAW264.7 cells were cultured in the presence of sRANKL (50 ng/mL, Sigma, USA) for 5 days. Medium and sRANKL were replaced on day 3. They were also co-cultured with U266 exosomes at a concentration of 50 μg/mL. At the end of the culture period, the cells were first placed in 4% paraformaldehyde (Sigma, USA) and then incubated with naphthol-AS-BI-phosphate (Sigma, USA), pararosaniline (Sigma, USA), and tartrate at 37 °C in acetate buffer (30 μM) for 45 min 42 . TRAP-positive cells with three or more nuclei were identified as OCs and were counted and photographed using a Nikon DS-U3 camera attached to a Nikon Eclipse E100 (Nikon, Japan). Quantification was performed by counting the number of TRAP-positive multinucleated cells in multiple fields. The results were normalized to the average number of TRAP-positive OCs in the negative control or control group.

Quantitative real-time PCR (qRT-PCR)

RNA was isolated from cells using TRIzol reagent (Invitrogen, USA), while TRIzol LS reagent (Invitrogen, USA) was used for RNA extraction from exosomes. For the reverse transcription and qRT-PCR of lncRNAs, we utilized the RiboSCRIPTTM mRNA/lncRNA qRT-PCR Kit (RiboBio, China). The expression of miRNAs was determined using the Mir-X miRNA First-Strand Synthesis and TB Green™ qRT-PCR Kit (TaKaRa, Japan). To analyze mRNA expression, we employed the iScript cDNA Synthesis Kit and iQ SYBR Green SuperMix reagents (Bio-Rad, USA). The primers used for qRT-PCR are listed in Supplementary Information: Supplementary Table 5 . qRT-PCR was conducted on the CFX Connect Real-Time System (Bio-Rad, USA). The relative expression levels of the indicated genes were calculated using the 2 −∆∆Ct method, normalized to GAPDH or U6, and compared each treatment group to the control group.

Western blot

Total proteins were extracted with RIPA lysis buffer (Beyotime) containing a proteinase inhibitor cocktail (Roche). After quantification with a BCA protein assay kit (Beyotime), the proteins were separated using SDS‒PAGE gels and then transferred onto PVDF membranes (Millipore). After the membrane was blocked with 5% skimmed milk for 1 h at room temperature, it was incubated with the indicated primary antibodies at 4 °C overnight followed by incubation with goat anti-rabbit/anti-mouse secondary antibody, HRP conjugate (ZSGB, China) at dilutions 1:5000. The protein bands were visualized using an enhanced chemiluminescence kit (Beyotime, China), and intensities were measured by ImageJ 6.0 software. The primary antibodies are listed in Supplementary Information: Supplementary Table 6 .

Luciferase reporter assay

To investigate the interaction between RANKL mRNA, MALAT1 , and miR-17, we conducted a dual-luciferase reporter assay. Wild-type sequences of RANKL mRNA and MALAT1 that bind to miR-17 were subcloned into a 3’ untranslated region (UTR) of the Renilla luciferase gene within the psiCHECK2 plasmid (Promega, USA). These constructs were co-transfected with a miR-17 agomir or a negative control miRNA into 293T cells using Lipofectamine 2000 (Invitrogen, USA) according to the manufacturer’s instructions. After 48 h of transfection, luciferase activities were measured using the Dual-Luciferase Reporter Assay System (Promega, USA) according to the manufacturer’s protocol. Firefly luciferase activity was normalized to Renilla luciferase activity to control for transfection efficiency. The relative luciferase activity (luciferase ratio) was calculated by comparing the normalized luciferase activity of the experimental group to that of the control group.

Bioinformatics

GSE24080 and GSE5900 from the GEO repository were utilized to acquire clinical parameters and normalized gene expression data 43 , 44 . Gene expression profiles and associated clinicopathological data of MM patients were retrieved from the Multiple Myeloma Research Foundation (MMRF) program ( https://research.themmrf.org and http://www.themmrf.org ). KEGG analyses were performed using the clusterProfiler R package 45 .

Mouse and treatment schedule

Animal studies were approved and performed in accordance with the guidelines of the Institutional Animal Care and Use Committee of Shandong University of Traditional Chinese Medicine. The maximal tumor burden permitted by the Committee is up to 10% of the animal’s body weight, and none of the experiments exceeded this limit. We have complied with all relevant ethical regulations for animal use. NOD/Shi-scid-IL2rgamma (null) (NOG) female mice aged 10 weeks were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. (Beijing, China). Each group consisted of 10 mice that were administered by intravenous injection via the tail vein with 100 μL of phosphate-buffered saline (PBS, negative control) and 2 × 10 6 U266 cells within four hours after irradiation with 1.5 Gy. 200 μg of exosomes collected from U266-shNC cells or U266-shMALAT1 cells were quantified and injected intravenously three times per week from day 0. Four weeks later, mice were euthanized to assess bone disease and tumor burden.

Micro-CT analysis

At sacrifice, the femur, spine, and skull were dissected free of soft tissue and fixed in 10% formalin before micro-CT analysis. Micro-CT analysis was performed as described previously 46 . The SkyScan 1176 micro-CT scanner (Bruker, Germany) was used to analyze bones (9-mm voxel size, 59 KVp, 127 mA, 0.48 rotation step). Image acquisition and data analysis were conducted using Cone-beam CT Reconstruction Software (version 2.6).

Bone histomorphometry

After femoral decalcification, wax embedding, and sectioning, the specimens were processed with H&E and TRAP staining according to the manufacturer’s instructions. Slide Converter 2.3 (3D HISTECH, Budapest, Hungary) was used to scan and analyze all stained slides. Active OCs were defined as multinucleated ( ≥ 3) TRAP-positive cells in contact with the bone surface 47 . OC surface/bone surface (Oc.N/BS) was calculated using a BioQuant system (BioQuant, Nashville, TN, USA). Quantification involved counting the number of TRAP-positive multinucleated cells and normalizing them to the bone surface area.

Immunohistochemical (IHC) assay

IHC assays were performed according to the manufacturer’s protocol. Samples were incubated overnight with rabbit primary antibody at 4 °C and then treated with a goat anti-rabbit HRP secondary antibody (ThermoFisher, 31460, 1:20) for 30 min at 37 °C, and nuclei were counterstained with hematoxylin. P-AKT, p-JNK, p-ERK, p-p38, RANKL, YBX1, and NSUN2 expression in mouse femur tissues or human BM tissues was semiquantitatively evaluated. Quantification was performed by counting the number of positive cells and total cells in several randomly selected high-power fields (HPFs) of the stained sections. Images were captured at HPFs magnified 400×. The results were normalized to the control group by calculating the ratio of positive cells in the experimental group to the positive cells in the control group. Immunoreactivity scores were classified as negative/weak (0-2 points), moderate (3–4 points), and strong (5–6 points). The primary antibodies are listed in Supplementary Information: Supplementary Table 6 .

RNA pulldown assay

RNA pulldown assays were performed as previously described 48 . Briefly, after synthesizing biotinylated wild-type or mutant MALAT1 with a T7 Transcription Kit (ThermoFisher Scientific, USA), they were treated with RNase-free DNase I (Roche, Switzerland) and purified with an RNeasy Mini Kit (Qiagen, Germany). For the pulldown assay, 1 mg of whole cell lysate was incubated with 3 μg of purified biotinylated transcripts at 25 °C for 1 h. The RNA-protein complexes were then captured using streptavidin agarose beads (Invitrogen, USA). After washing the beads to remove non-specifically bound proteins, the RNA-associated proteins were eluted and analyzed by Western blotting to detect the protein pulled down by MALAT1 .

RNA immunoprecipitation

The RIP assays were performed using a Magna RIP TM RNA-Binding Protein Immunoprecipitation Kit (Millipore, USA). U266 cell lysates were prepared using RIP lysis buffer. The lysates were then immunoprecipitated with RIP buffer containing NSUN2 antibodies (Abcam, ab259941, 1:30). For the immunoprecipitation, a mixture of cell lysate and antibody was incubated with Protein A/G magnetic beads at 4 °C overnight. The beads were washed and the RNA-protein complexes were eluted. The precipitated RNAs were purified and analyzed by qRT-PCR. IgG was used as the negative control to ensure the specificity of the immunoprecipitation.

RNA stability detection

Cells were treated with the RNA synthesis inhibitor actinomycin D (Act D, 5 μg/mL). RNA was extracted at different time points (0, 6, 12, 18, and 24 h) post-treatment using TRIzol reagent. The extracted RNA was reverse transcribed into cDNA using the iScript™ cDNA Synthesis Kit (Bio-Rad, USA). The half-life of MALAT1 was then measured by qRT-PCR using SYBR Green Supermix (Bio-Rad, USA), with GAPDH as a normalization control.

Methylated RNA immunoprecipitation

For the analysis of m5C levels in MALAT1, the MeRIP assay was performed using the Magna MeRIP™ mRNA Methylation Kit (Millipore, USA) according to the manufacturer’s protocol. In brief, total RNA was isolated from cells using TRIzol (Invitrogen). Then, 5 μg RNA was fragmented into ~200 nucleotide-long fragments. The fragmented RNA was incubated with a mixture of Magna Protein G beads and anti-m5C antibodies (Abcam, ab10805) at 4 °C overnight. The RNA-antibody complexes were then immunoprecipitated, and the RNA was purified from the beads. The enriched RNA fragments containing m5C modifications were analyzed by qRT-PCR.

Statistics and reproducibility

Data are displayed as the mean ± SD. All experiments were repeated at least three times. The data were analyzed by SPSS Statistics 26.0. For comparisons between the two groups, a two-tailed Student’s t -test or Welch’s t-test was used. For comparisons involving more than two groups, one-way ANOVA followed by Tukey’s multiple comparison test was used to determine statistically significant differences. Kaplan–Meier analyses were performed via SPSS Statistics 26.0 and the survival and survminer R packages. In the univariate Cox regression model, we chose a P -value threshold of ≤0.1 to identify potential predictors that might not reach the conventional levels of statistical significance but could still be of biological or clinical relevance, while a P -value ≤ 0.05 was considered a statistically significant difference in other statistical analyses.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Supplementary information

Peer Review File

Supplementary Information

Description of Additional Supplementary File

Supplementary Data 1

Reporting summary

Acknowledgements

The study was supported by the National Natural Science Foundation of China (No.82074348, No. 82274491), the Taishan Scholar Program (No. tsqn201812145), Shandong Provincial Natural Science Foundation Innovation Development Joint Fund (No. ZR2023LZL009), the Natural Science Foundation of Shandong Province (No. ZR2020MH388), and Shandong Traditional Chinese Medicine Science and Technology Project (M-2022158).

Author contributions

X.C. and M.Y. conceptualized and designed this study. M.Y., Y.Z., and J.F. performed experiments and analyzed the data. Z.C. and J.Z. supervised this project and the manuscript. M.Y. drafted the manuscript. X.C. critically reviewed and edited the manuscript.

Peer review

Peer review information

Communications Biology thanks Flavia Pichiorri, Manuela Mania, and Sorina Dinescu. for their contribution to the peer review of this work. Primary handling editors: Martina Rauner and Christina Karlsson Rosenthal. A peer review file is available.

Data availability

The sequencing data of all samples in this study have been uploaded to NCBI’s SRA database, with access number PRJNA1155007 . All other data are available from the corresponding author upon reasonable request. The uncropped images are available in Supplementary Information and source data of all figures are available in Supplementary Data 1 .

Code availability

Codes are available upon reasonable request to the corresponding author.

Competing interests

The authors declare no competing interests.

Footnotes

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

Supplementary information

The online version contains supplementary material available at 10.1038/s42003-024-06918-8.

References

1. Gai, D. et al. CST6 suppresses osteolytic bone disease in multiple myeloma by blocking osteoclast differentiation. J. Clin. Invest. 132 , 10.1172/JCI159527 (2022).

2. Melton, L. J. 3rd, Kyle, R. A., Achenbach, S. J., Oberg, A. L. & Rajkumar, S. V. Fracture risk with multiple myeloma: a population-based study. J. Bone Min. Res. 20 , 487–493 (2005).

3. Saad, F. et al. Pathologic fractures correlate with reduced survival in patients with malignant bone disease. Cancer 110 , 1860–1867 (2007).

4. Terpos, E., Ntanasis-Stathopoulos, I. & Dimopoulos, M. A. Myeloma bone disease: from biology findings to treatment approaches. Blood 133 , 1534–1539 (2019).

5. Yen, C. H., Hsu, C. M., Hsiao, S. Y. & Hsiao, H. H. Pathogenic Mechanisms of Myeloma Bone Disease and Possible Roles for NRF2. Int. J. Mol. Sci . 21 , 10.3390/ijms21186723 (2020).

6. Giuliani, N. et al. Increased osteocyte death in multiple myeloma patients: role in myeloma-induced osteoclast formation. Leukemia 26 , 1391–1401 (2012).

7. Gau, Y. C., Yeh, T. J., Hsu, C. M., Hsiao, S. Y. & Hsiao, H. H. Pathogenesis and treatment of myeloma-related bone disease. Int. J. Mol. Sci . 23 , 10.3390/ijms23063112 (2022).

8. Krylova, S. V. & Feng, D. The machinery of exosomes: biogenesis, release, and uptake. Int. J. Mol. Sci . 24 , 10.3390/ijms24021337 (2023).

9. Thery, C., Zitvogel, L. & Amigorena, S. Exosomes: composition, biogenesis and function. Nat. Rev. Immunol. 2 , 569–579 (2002).

10. Shah, R., Patel, T. & Freedman, J. E. Circulating extracellular vesicles in human disease. N. Engl. J. Med. 379 , 2180–2181 (2018).

11. Madeo, M. et al. Cancer exosomes induce tumor innervation. Nat. Commun. 9 , 4284 (2018).

12. Thery, C. et al. Minimal information for studies of extracellular vesicles 2018 (MISEV2018): a position statement of the International Society for Extracellular Vesicles and update of the MISEV2014 guidelines. J. Extracell. Vesicles 7 , 1535750 (2018).

13. Liu, N. et al. Long non-coding RNA MALAT1 facilitates the tumorigenesis, invasion and glycolysis of multiple myeloma via miR-1271-5p/SOX13 axis. J. Cancer Res. Clin. Oncol. 146 , 367–379 (2020).

14. Sun, Y. et al. LncRNA MALAT1/miR-181a-5p affects the proliferation and adhesion of myeloma cells via regulation of Hippo-YAP signaling pathway. Cell Cycle 18 , 2509–2523 (2019).

15. Herrero, A. B. et al. FAM46C controls antibody production by the polyadenylation of immunoglobulin mRNAs and inhibits cell migration in multiple myeloma. J. Cell Mol. Med. 24 , 4171–4182 (2020).

16. Trixl, L. & Lusser, A. The dynamic RNA modification 5-methylcytosine and its emerging role as an epitranscriptomic mark. Wiley Interdiscip. Rev. RNA 10 , e1510 (2019).

17. Yang, X. et al. 5-methylcytosine promotes mRNA export—NSUN2 as the methyltransferase and ALYREF as an m(5)C reader. Cell Res. 27 , 606–625 (2017).

18. Sun, Z. et al. Aberrant NSUN2-mediated m(5)C modification of H19 lncRNA is associated with poor differentiation of hepatocellular carcinoma. Oncogene 39 , 6906–6919 (2020).

19. Yan, J., Liu, J., Huang, Z., Huang, W. & Lv, J. FOXC2-AS1 stabilizes FOXC2 mRNA via association with NSUN2 in gastric cancer cells. Hum. Cell 34 , 1755–1764 (2021).

20. Zhang, J. et al. Circular RNA_LARP4 inhibits cell proliferation and invasion of gastric cancer by sponging miR-424-5p and regulating LATS1 expression. Mol. Cancer 16 , 151 (2017).

21. Chen, X. et al. 5-methylcytosine promotes pathogenesis of bladder cancer through stabilizing mRNAs. Nat. Cell Biol. 21 , 978–990 (2019).

22. Tay, Y., Rinn, J. & Pandolfi, P. P. The multilayered complexity of ceRNA crosstalk and competition. Nature 505 , 344–352 (2014).

23. Bolzoni, M. et al. Possible targets to treat myeloma-related osteoclastogenesis. Expert Rev. Hematol. 11 , 325–336 (2018).

24. Yuan, X. et al. Breast cancer exosomes contribute to pre-metastatic niche formation and promote bone metastasis of tumor cells. Theranostics 11 , 1429–1445 (2021).

25. Liu, Z. et al. Exosomes from LSD1 knockdown breast cancer cells activate osteoclastogenesis and inhibit osteoblastogenesis. Int. J. Biol. Macromol. 235 , 123792 (2023).

26. Hu, Y. et al. NSUN2 modified by SUMO-2/3 promotes gastric cancer progression and regulates mRNA m5C methylation. Cell Death Dis. 12 , 842 (2021).

27. Fang, L. et al. m5C-methylated lncRNA NR_033928 promotes gastric cancer proliferation by stabilizing GLS mRNA to promote glutamine metabolism reprogramming. Cell Death Dis. 14 , 520 (2023).

28. Boyce, B. F. & Xing, L. Functions of RANKL/RANK/OPG in bone modeling and remodeling. Arch. Biochem. Biophys. 473 , 139–146 (2008).

29. Boyle, W. J., Simonet, W. S. & Lacey, D. L. Osteoclast differentiation and activation. Nature 423 , 337–342 (2003).

30. Ouyang, Z. et al. Hypericin suppresses osteoclast formation and wear particle-induced osteolysis via modulating ERK signalling pathway. Biochem. Pharm. 90 , 276–287 (2014).

31. Monje, P., Hernandez-Losa, J., Lyons, R. J., Castellone, M. D. & Gutkind, J. S. Regulation of the transcriptional activity of c-Fos by ERK. A novel role for the prolyl isomerase PIN1. J. Biol. Chem. 280 , 35081–35084 (2005).

32. Qin, J. et al. Aberrantly expressed long noncoding RNAs as potential prognostic biomarkers in newly diagnosed multiple myeloma: a systemic review and meta-analysis. Cancer Med. 12 , 2199–2218 (2023).

33. Hu, Y. et al. Targeting the MALAT1/PARP1/LIG3 complex induces DNA damage and apoptosis in multiple myeloma. Leukemia 32 , 2250–2262 (2018).

34. Morelli, E. et al. Therapeutic vulnerability of multiple myeloma to MIR17PTi, a first-in-class inhibitor of pri-miR-17-92. Blood 132 , 1050–1063 (2018).

35. Liu, Y. et al. MiR-17 modulates osteogenic differentiation through a coherent feed-forward loop in mesenchymal stem cells isolated from periodontal ligaments of patients with periodontitis. Stem Cells 29 , 1804–1816 (2011).

36. Lin, S., Choe, J., Du, P., Triboulet, R. & Gregory, R. I. The m(6)A methyltransferase METTL3 promotes translation in human cancer cells. Mol. Cell 62 , 335–345 (2016).

37. Hussain, S. et al. The nucleolar RNA methyltransferase Misu (NSun2) is required for mitotic spindle stability. J. Cell Biol. 186 , 27–40 (2009).

38. Ma, Y., Yang, J., Ji, T. & Wen, F. Identification of a novel m5C/m6A-related gene signature for predicting prognosis and immunotherapy efficacy in lung adenocarcinoma. Front. Genet. 13 , 990623 (2022).

39. Chen, T. et al. NSUN2 is a glucose sensor suppressing cGAS/STING to maintain tumorigenesis and immunotherapy resistance. Cell Metab. 35 , 1782–1798.e1788 (2023).

40. Xing, J. et al. NSun2 promotes cell growth via elevating cyclin-dependent kinase 1 translation. Mol. Cell Biol. 35 , 4043–4052 (2015).

41. Li, Q. et al. NSUN2-mediated m5C methylation and METTL3/METTL14-mediated m6A methylation cooperatively enhance p21 translation. J. Cell Biochem. 118 , 2587–2598 (2017).

42. Song, C. et al. Evaluation of efficacy on RANKL induced osteoclast from RAW264.7 cells. J. Cell Physiol. 234 , 11969–11975 (2019).

43. Zhan, F. et al. Gene-expression signature of benign monoclonal gammopathy evident in multiple myeloma is linked to good prognosis. Blood 109 , 1692–1700 (2007).

44. Mitchell, J. S. et al. Genome-wide association study identifies multiple susceptibility loci for multiple myeloma. Nat. Commun. 7 , 12050 (2016).

45. Yu, G., Wang, L. G., Han, Y. & He, Q. Y. clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS 16 , 284–287 (2012).

46. Li, B. et al. Exosome-mediated transfer of lncRUNX2-AS1 from multiple myeloma cells to MSCs contributes to osteogenesis. Oncogene 37 , 5508–5519 (2018).

47. Li, C. H. & Amar, S. Morphometric, histomorphometric, and microcomputed tomographic analysis of periodontal inflammatory lesions in a murine model. J. Periodontol. 78 , 1120–1128 (2007).

48. Yang, F. et al. Long noncoding RNA high expression in hepatocellular carcinoma facilitates tumor growth through enhancer of zeste homolog 2 in humans. Hepatology 54 , 1679–1689 (2011).

Associated Data

Supplementary Materials

Peer Review File

Supplementary Information

Description of Additional Supplementary File

Supplementary Data 1

Reporting summary

Data Availability Statement

The sequencing data of all samples in this study have been uploaded to NCBI’s SRA database, with access number PRJNA1155007 . All other data are available from the corresponding author upon reasonable request. The uncropped images are available in Supplementary Information and source data of all figures are available in Supplementary Data 1 .

Codes are available upon reasonable request to the corresponding author.