m6A reader hnRNPA2B1 drives multiple myeloma osteolytic bone disease

Rationale: Bone destruction is a hallmark of multiple myeloma (MM) and affects more than 80% of patients. Although previous works revealed the roles of N6-methyladenosine (m6A) reader hnRNPA2B1 in the development of tumors, whether hnRNPA2B1 regulates bone destruction in MM is still unknown. Methods: Alizarin red S staining, TRAP staining, ELISA and quantitative real-time PCR assays were used to evaluate osteogenesis and osteoclastogenesis in vitro. X ray and bone histomorphometric analysis were preformed to identify bone resorption and bone formation in vivo. Exosome isolation and characterization were demonstrated by transmission electron microscopy, dynamic light scattering, immunofluorescence and flow cytometry assays. The interactions between hnRNPA2B1 and primary microRNAs were examined using RNA pull-down and RIP assays. Coimmunoprecipitation assay was used to test the interaction between hnRNPA2B1 and DGCR8 proteins. Luciferase assay was established to assess miRNAs target genes. Results: Here we show that myeloma cells hnRNPA2B1 mediates microRNAs processing and upregulates miR-92a-2-5p and miR-373-3p expression. These two microRNAs are transported to recipient monocytes or mesenchymal stem cells (MSCs) through exosomes, leading to activation of osteoclastogenesis and suppression of osteoblastogenesis by inhibiting IRF8 or RUNX2. Furthermore, clinical studies revealed a highly positive correlation between the level of myeloma cells hnRNPA2B1 and the number of osteolytic bone lesions in myeloma patients. Conclusions: This study elucidates an important mechanism by which myeloma-induced bone lesions, suggesting that hnRNPA2B1 may be targeted to prevent myeloma-associated bone disease.


Introduction
Multiple myeloma (MM) arises from malignant plasma cells within bone marrow and remains an incurable disease. One hallmark of MM is osteolytic bone disease. More than 80% of patients suffer from bone destruction, which includes pathological fracture, severe bone pain, spinal cord compression, and hypercalcemia, greatly reduces their quality of life and has a severe negative impact on survival [1].

Ivyspring International Publisher
Bone is a dynamic tissue that is constantly being remodeled by bone-resorbing osteoclasts and bone-forming osteoblasts [2]. Osteoclasts arise from hematopoietic monocytic precursors and resorbs bone. The formation of osteoclasts requires soluble cytokines such as receptor activator of nuclear factor-κB ligand (RANKL) and macrophage colony-stimulating factor (M-CSF). RANKL enhances the expression of nuclear factor of activated T-cells, cytoplasmic 1 protein (NFATc1), which upregulates the expression of osteoclast differentiation-associated genes, such as tartrate-resistant acid phosphatase (TRAP), calcitonin receptor (CALCR), and cathepsin K (CTSK), whereas the transcriptional factor interferon regulatory factor 8 (IRF8) can suppress RANKLinduced NFATc1 expression [2]. The next player in the remodeling cycle is the osteoblasts, which are differentiated from mesenchymal stem cells (MSCs). Osteoblast's differentiation requires the activation of core-binding factor α-1/runt-related transcription factor 2 (RUNX2) and osterix, which stimulate the expression of osteoblast differentiation-associated genes, such as bone gamma-carboxyglutamic acidcontaining protein (BGLAP), alkaline phosphatase (ALP), and collagen type I α1 (COL1A1) [2,3]. This delicate balance is disrupted in certain types of malignancies, including MM and solid tumors, such as breast and lung cancer [4,5]. Myeloma cells can stimulate production of several cytokines such as RANKL, macrophage inflammatory protein-1α (MIP-1α), and monocyte chemoattractant protein-1 (MCP-1) and thus enhance osteoclast differentiation and bone resorption activity [4,6]. On the other hand, myeloma cells can also secrete dickkopf-related protein 1 (DKK1), which inhibits the Wnt/β-catenin signaling pathway and suppresses maturation of MSCs into osteoblasts [7].
N6-methyladenosine (m6A) modification is the most abundant modification on eukaryotic RNA. The modification is composed of three classes of protein factors: methylate adenosine at N6 position (writers), demethylate m6A for reversible regulation (erasers) and effectors that recognize and bind to m6A motif, regulating RNA stability, splicing, trafficking, and mRNA translation and others (Readers) [8]. The heterogeneous nuclear ribonucleoprotein A2B1 (hnRNPA2B1) is one of the nuclear readers of m6A, which is highly expressed in many types of cancers and accelerate mRNA processing via RNA binding, indicating an important role in the development of tumors [9,10]. For instance, hnRNPA2B1 results in pyruvate kinase isozymes M2 (PKM2) cumulation and promotes glioma cells proliferation [11]. In pancreatic ductal adenocarcinoma cells, hnRNPA2B1 interacts with Kirsten rat sarcoma viral oncogene (KRAS) and modulated cell proliferation, mobility, and apoptosis [12]. During epithelial mesenchymal transition (EMT) in tumor cells, hnRNPA2B1 up-regulates vimentin and N-cadherin and downregulates E-cadherin, promotes cell invasion and metastasis in various cancers [13]. In MM, recent work reveals m6A-dependent effect of hnRNPA2B1 on activating AKT signaling pathway and promoting MM progression [14]. However, hnRNPA2B1 has never been implicated in the regulation of bone resorption or formation in tumors. In this work, we hypothesized that the hnRNPA2B1 plays a role in the pathogenesis of cancer-induced bone destruction in myeloma.
Through a combination of in vitro, in vivo, and patient samples study, we reported that hnRNPA2B1 has a unique role in myeloma-induced bone disease. Our results showed that myeloma cell hnRNPA2B1-DGCR8 (DiGeorge syndrome critical region 8) complex upregulates miR-92a-2-5p and miR-373-3p in myeloma cell exosomes. Exosomes miR-92a-2-5p inhibits the expression of interferon regulatory factor 8 (IRF8) and thereby activates RANKL-induced NFATc1 expression, leading to an increase in osteoclastogenesis and bone resorption. Exosomes miR-373-3p suppresses osteoblastogenesis and bone formation by downregulating the expression of RUNX2 in human MSCs. Our findings not only elucidate a mechanism of cancer-induced bone destruction, but also implicate a potential therapeutic approach for cancer patients with osteolytic bone lesions by targeting hnRNPA2B1.

hnRNPA2B1 enhances lytic bone lesions and tumor progression
We analyzed the alteration of gene expression in plasma cells of myeloma patients, compared to plasma cells of healthy donors, using GEO dataset (GSE6691) [15]. Computational overlapping of genes with the Molecular Signatures Database (Broad Institute) hallmark gene sets suggested significant enrichment of genes in mRNA stability and metabolic process ( Figure 1A). As m6A modification proteins (METTL3, WTAP, FTO, hnRNPC, hnRNPA2B1, YTHDF3) are well known regulators on eukaryotic RNA processing. Paired differential analysis identified three of them (hnRNPA2B1, YTHDF3 and hnRNPC) are significantly upregulated in plasma cells of myeloma patients compared to normal plasma cells (Fold Change > 2) ( Figure 1B). Unsupervised hierarchical clustering of microarray data also suggested that these three genes expression were increased in the plasma cells of myeloma patients ( Figure 1C-D and Figure S1A). Additionally, to understand whether m6A modification proteins may also affect myeloma cell-induced bone lesions, we compared their expression levels in myeloma cells of patients with or without bone lesions from a published dataset (GEO: GSE755) [7]. Among the three candidate genes, the levels of hnRNPA2B1 expression were higher in myeloma cells of patients with bone lesions compared with those without (Figure 1E and Figure S1B). Western blot results also showed that hnRNPA2B1 was expressed in most of the bone marrow aspirates of primary myeloma cells and in most of the established human myeloma cell lines, but not in aspirates of plasma cells from normal subjects ( Figure 1F). We next analyzed the association of hnRNPA2B1 and other m6A modification proteins with myeloma disease using dataset from Multiple Myeloma Research Foundation (MMRF) coMMpass study IA15, and found that patients with high levels of hnRNPA2B1 in myeloma cells displayed shorter overall survival or progression free survival than those with low expression (Figure 1G and Figure  S1C). Based on these results, hnRNPA2B1 was selected as candidate gene, which is highly likely a proliferation-related and bone lesion-related gene in myeloma.
To examine the functional role of hnRNPA2B1 in tumor growth and bone lesions, we knocked down its expression in RPMI8226 myeloma cells using small hairpin RNAs (shRNAs) against human hnRNPA2B1 and over expressed hnRNPA2B1 complementary DNA (cDNA; A2B1) in MM.1S myeloma cells ( Figure  S2A). We observed decreased colony formation and growth in shA2B1 myeloma cells compared with shCtrl cells (Figure 1H and Figure S2B-G), and increased colony formation and growth in A2B1 myeloma cells compared with Vec cells (Figure 1H and Figure S2B-G). Furthermore, we found a strong positive correlation between the level of hnRNPA2B1expression in myeloma cells and bone lesion numbers in patients ( Figure 1I). hnRNPA2B1 and CD138 (myeloma cells surface marker) expression were higher in myeloma cells from patient with high bone lesion numbers (P1) than in those from patient with low lesion numbers (P2) (Figure 1J). Representative images of magnetic resonance imaging and X ray scanning showed more lytic lesions in the spine ( Figure 1K) and skull ( Figure 1L) of P1 than P2. These results indicate the association of hnRNPA2B1 to myeloma tumorigenesis and bone disease.
To determine the functional role of myeloma expressed hnRNPA2B1 in lytic bone lesions, we injected shA2B1 RPMI8226 cells into mouse femurs and caused fewer lytic lesions than did shCtrl RPMI8226 cells. Conversely, A2B1 MM.1S cells caused more femur lesions than did Vec MM.1S cells (Figure  2A). To assess the role of myeloma-expressed hnRNPA2B1 in osteoclast-mediated bone resorption in vivo, we examined the levels of mouse serum procollagen type I N-terminal propeptide (PINP), a bone formation marker and C-telopeptide of type I collagen (CTX-1), a bone resorption marker. We found higher PINP levels and lower CTX-1 levels in shA2B1 or Vec group, compared with shCtrl or A2B1 group (Figure 2B-C). We also stained myeloma-bearing mouse femurs for TRAP and Toluidine blue, bone histomorphometric analysis demonstrated a lower bone volume/total volume (BV/TV) (Figure 2D

hnRNPA2B1 enhances RANKL-induced osteoclastogenesis and inhibits osteoblastogenesis through exosomes
Bone remodeling is maintained by a balance between osteoclast-mediated resorption and osteoblast-mediated bone formation. To examine whether myeloma cells hnRNPA2B1 can regulate this balance, we first assessed their effects on osteoclast differentiation. In the presence of RANKL, coculture of precursors of osteoclasts (preOCs) with myeloma cells expressing high levels of hnRNPA2B1 (shCtrl RPMI8226 or A2B1 MM.1S) induced more multinuclear tartrate-resistant acid phosphatasepositive (TRAP + ) cells formation (Figure 3A-B), TRAP 5b secretion (Figure 3C), and osteoclast gene expression ( Figure 3D-E) than in those cocultured with low hnRNPA2B1 myeloma cells (shA2B1 or Vec). These results indicate that myeloma cells hnRNPA2B1 enhances osteoclastogenesis.
To assess the effect of myeloma cells hnRNPA2B1 on osteoblast formation, we cocultured osteoblast progenitors, MSCs, in osteoblast medium with myeloma cells. MSCs cultured alone in this medium served as a positive control. Cocultured with low hnRNPA2B1-expressing myeloma cell lines (Vec MM.1S, or shA2B1 RPMI8226) had comparatively more mature osteoblasts (Figure 3F-G), higher ALP activities (Figure 3H), and higher expression of osteoblast differentiation-associated genes ( Figure  3I-J) than those with high levels of hnRNPA2B1.
Exosomes are small membrane vesicles (30-150 nm) derived from the luminal membranes of multivesicular bodies. Exosomes mediate local and systemic cell communication during tumor growth and progression through the horizontal transfer of information, such as mRNAs, microRNAs (miRNAs) and proteins [16,17]. In myeloma, other studies indicated that myeloma-exosomes modulate osteoclast and osteoblast function and differentiation, but the mechanisms are still unknown [18,19]. We isolated exosomes from cell culture medium using ultracentrifugation and confirmed their identity by electron microscopy and dynamic light scattering analysis ( Figure 4A). This was further confirmed by the expression of exosome markers (Figure 4B-D). Osteoclastogenesis and osteoblastogenesis assay indicated that myeloma cells exosomes promote osteoclast differentiation ( Figure 4E-F and Figure  S3A-B) and inhibit osteoblast differentiation ( Figure  4G-H and Figure S3C-D). To investigate whether exosomes can be taken up by precursors of osteoclasts or MSCs, a Dil dye was used to label the exosomes that were then co-cultured with target cells. Confocal microscopy showed that Dil signals were detected in cytoplasm of precursors of osteoclasts or MSCs (Figure 4I-J). Moreover, exosomes isolated from myeloma cells expressing high levels of hnRNPA2B1 (shCtrl RPMI8226 or A2B1 MM.1S) induced more TRAP + cells formation ( Figure 4K) and less mature osteoblasts than those with low levels of hnRNPA2B1 ( Figure 4L). We pre-treated MM.1S cells (Vec and A2B1) with or without GW4869 (Inhibitor of exosome biogenesis/release), and collected the conditioned medium (CM). Precursors of osteoclasts or MSCs were cultured with or without CM. The results showed that GW4869 significantly reversed CM induced osteoclastogenesis or inhibited osteoblastogenesis, and there is no difference between Vec and A2B1 groups pre-treated with GW4869 ( Figure  4M-N). These experiments demonstrate that hnRNPA2B1 enhances osteoclastogenesis and inhibits osteoblastogenesis through exosomes.

hnRNPA2B1 upregulates exosomes miR-92a-2-5p and miR-373-3p expression and correlates with patient's bone lesions
hnRNPA2B1 was identified as a mediator of m6A-dependent primary miRNAs processing events [10]. Thus, we hypothesized that hnRNPA2B1 regulates exosome miRNAs expression to induce bone lesions. We then divided myeloma patients into two groups on whether they suffered from bone lesions and cultured them to isolate exosomes. Using real time PCR analysis, we identified top two miRNAs: miR-92a-2-5p and miR-373-3p, which were upregulated in bone lesions group ( Figure 4O). We confirmed the results by another 10 patient samples with bone lesions and 10 patient samples without bone lesions. The levels of miR-92a-2-5p and miR-373-3p expression were higher in myeloma cells of patients with bone lesions compared with those without (Figure 4P-Q). Furthermore, we found a strong positive correlation between the levels of miR-92a-2-5p or miR-373-3p in myeloma cells and bone lesion numbers in patients (Figure 4R-S). Real time PCR analysis indicated that miR-92a-2-5p or miR-373-3p upregulated in myeloma cell lines exosomes than healthy plasma cell exosomes ( Figure  4T-U). To investigate the potential effect of miR-92a-2-5p on osteoclast differentiation, monocytes were transfected with miR-92a-2-5p mimics or control miRNA ( Figure S4A). TRAP staining and TRAP 5b levels indicated that miR-92a-2-5p enhanced osteoclast differentiation (Figure 4V-W). In addition, MSCs were transfected with miR-373-3p mimics ( Figure S4B). Alizarin red S staining and ALP activities experiments showed suppressed osteoblast differentiation (Figure 4X-Y).
Next, we investigated the molecular mechanism of how hnRNPA2B1 regulates miR-92a-2-5p or miR-373-3p expression. Their expressions are higher in both the RPMI8226 and MM.1S cells and their corresponding exosomes as compared with that in monocytes or MSCs (Figure S5A-B). Previous study indicated that hnRNPA2B1 recruited microRNA microprocessor complex protein DGCR8 to a subset of precursor miRNAs and facilitates their processing into mature miRNAs [10]. We used coimmunoprecipitation assays to interrogate the interaction between hnRNPA2B1 and DGCR8 proteins. We immunoprecipitated myeloma cells lysates with an anti-DGCR8 antibody and observed the presence of hnRNPA2B1 proteins in the lysates ( Figure 5A). We also detected DGCR8 proteins in immunoprecipitates using an anti-hnRNPA2B1 antibody (Figure 5B), indicating that hnRNPA2B1 and DGCR8 form a complex. Using immunofluorescent staining, we observed a co-localization of hnRNPA2B1 and DGCR8 in myeloma cells ( Figure 5C). Western blot analysis of enriched proteins after RNA pull-down indicated that primary miR-92a-2-5p or miR-373-3p bound specifically to hnRNPA2B1 (Figure 5D-E). Data are averages ± SD. Each experiment was repeated three times. *P < 0.05; **P < 0.01. All P values were determined using one-way ANOVA.
In addition, we found a strong positive correlation between the levels of miR-92a-2-5p or miR-373-3p with hnRNPA2B1 mRNA in patient myeloma cells ( Figure  5J-K).  Data are averages ± SD. Each experiment was repeated three times. **P < 0.01; ****P < 0.0001. P values were determined using one way ANOVA.
We also analyzed TCGA database and found upregulated expression of hnRNPA2B1 in some malignancies including breast, colon, lung and liver cancer (Figure S9A-D). To further confirm our findings, we knocked down or over expressed hnRNPA2B1 in breast cancer cell line MCF7 ( Figure  S10A), we observed decreased growth in shA2B1 MCF7 cells compared with shCtrl group (Figure  S10B), and increased growth in A2B1 MCF7 cells compared with Vec group (Figure S10B). Similar to myeloma cells, exosomes isolated from MCF7 cells expressing high levels of hnRNPA2B1 (shCtrl, A2B1) induced more TRAP + cells formation ( Figure S10C) and less mature osteoblasts than those with low levels of hnRNPA2B1 (Vec or shA2B1) (Figure S10D). These findings may have broader implications for the genesis of bone lesions caused by these and other tumors.

Discussion
Osteolytic bone lesion is a hallmark in the vast majority of myeloma patients. Myeloma cells disrupt the delicate balance between bone formation and resorption, leading to debilitating osteolytic bone lesions. Although previous studies established that myeloma cell exosomes enhance osteoclast differentiation and inhibit osteoblast differentiation, the mechanism underlying remains elusive. In this study we clearly demonstrated that myeloma cells hnRNPA2B1 may be responsible, at least in part, for promoting bone destruction in vivo through myeloma cell exosomes. hnRNPA2B1-DGCR8 complex mediates m6A-dependent primary microRNA processing events and upregulates miR-92a-2-5p and miR-373-3p expression. These two miRNAs are packed into exosomes and transported to recipient monocytes or MSCs, leading to activating osteoclastogenesis and suppressing osteogenesis by inhibiting IRF8 or RUNX2 (Figure 7H). Our clinical studies examining the correlation between the level of myeloma cell hnRNPA2B1 and the number of osteolytic bone lesions in newly diagnosed patients support this conclusion. Thus, this study reveals a novel mechanism that explains how myeloma cells induce bone destruction. Our study also suggests that hnRNPA2B1 may be a therapeutic target for bone disease in patients with myeloma. miRNAs widely found in eukaryotes, are a class of noncoding RNAs (19-25 nucleotides in length). They usually bind to the 3' UTR regions of their target mRNAs, and lead to translation inhibition or degradation [20]. miRNAs play important roles in physiology and pathophysiology, including development, apoptosis, tumor development, and so on [16,21,22]. Previous studies also showed that miRNAs are involved in osteoclastogenesis and osteoblastogenesis [23][24][25]. More than 20 miRNAs function has been identified in osteoclast generation, and showed positive or negative feedback to regulate osteoclastogenesis. For instance, miR-340 inhibits osteoclast differentiation through repression of microphthalmiaassociated transcription factor (MITF) [26], miR-29 family play a critical role in early phase of osteoclastogenesis by targeting NFIA (nuclear factor I A) [27]. But the function of miR-92a-2-5p in osteoclast differentiation remains unknown. We reported that miR-92a-2-5p enhances osteoclast differentiation by inhibiting IRF8. IRF8 is a well-known transcription factor which can suppress RANKL-induced NFATc1 expression [28]. In osteoblastogenesis, some studies have discovered multiple miRNAs to be important regulators of bone-forming genes, including essential transcription factors and developmental signaling molecules and their receptors that are required for the complex process of osteoblastogenesis. miR-26a negatively regulates osteoblast differentiation by targeting the SMAD family member 1 (SMAD1) [29]. miRNA-133a-5p and miRNA-132-3p inhibits osteoblast differentiation by targeting the 3' UTR of RUNX2 or E1A binding protein P300 (EP300) directly [30]. miR-15b promotes osteoblast differentiation by protecting RUNX2 protein from SMAD specific E3 ubiquitin protein ligase 1 (Smurf1) mediated degradation [31]. In our study, we found miR-373-3p binds directly to the 3' UTR of RUNX2 and inhibits osteoblastogenesis. It is a helpful addition to the available research.
Exosomes are small membrane vesicles of endocytic origin derived from various cell types and released by fusion with the recipient cell membrane. Exosomes mediate cell-to-cell communication by transferring mRNAs, miRNAs, long non-coding RNAs (lncRNAs) and proteins. Exosomes play multiple roles in immune response, antigen presentation, tumor development, tumor metastasis, cell migration, cell differentiation, and angiogenesis, and so on [16,17]. But the effect of myeloma cell released exosomes in bone lesion remains poorly understood. A previous study suggested that myeloma cell exosomes modulate osteoclast's function and differentiation [18], but the mechanism is unknown. Another work reported that myeloma cell exosomes induced MSCs miR-103-3p expression and inhibited osteoblast differentiation [19]. Our study elucidated that myeloma cells hnRNPA2B1 upregulates exosomes miR-92a-2-5p and miR-373-3p expression, which enhances osteoclastogenesis and inhibits osteoblastogenesis and thus lead to bone destruction. Furthermore, miRNA-92a-2-5p and miR-373-3p expression levels are much lower in MSCs or monocytes compared with myeloma cells. Myeloma cell packed these miRNAs and transferred to recipient cells through exosomes, which provided a new possibility of how tumor cells or other stromal cell in tumor microenvironment to mediate bone lesions.
Collectively, our results elucidate a new mechanism by which myeloma induced bone lesions. At present, lack effective targeted drug is a major barrier to myeloma bone lesion therapy. Attempts to target RANKL and DKK1 have achieved only modest success. For an example, the anti-resorptive agent denosumab (a monoclonal antibody against RANKL) only showed a moderate effect in a Phase III trial [2]. BHQ880 (a monoclonal antibody against DKK-1) fails to restore new bone formation in a Phase I/II study. Bisphosphonates can suppress osteoclast function, but which are less than fully effective and cause osteonecrosis of the jaw in part of treated patients [32]. Because there is no hnRNPA2B1 inhibitor in commercial, the therapeutic effect of these drugs in myeloma-induced bone lesions remains unknown. But our data make a compelling case for a role of hnRNPA2B1 in myeloma-induced bone disease and thus encourage evaluating of these inhibitors. More importantly, hnRNPA2B1 is often upregulated by other malignancies including breast, colon, liver, and lung cancer, our findings also have broader implications for the mechanisms of bone metastasis caused by these and other tumors.

Antibodies, plasmids, and reagents
The plasmids hnRNPA2B1 and control vector were purchased from GeneCopoeia. MDH1-PGK-GFP 2.0 retroviral vector was purchased from Addgene (#11375). Except where specified, all chemicals were purchased from Sigma-Aldrich, and all antibodies for western blot analysis were purchased from Cell Signaling Technology. shRNAs against hnRNPA2B1 and non-target control were purchased from Sigma-Aldrich. ELISA kits were purchased from Immunodiagnostic Systems.

In vitro osteoblast and osteoclast formation and function assays
MSCs were obtained from the bone marrow, and mature osteoblasts were generated from MSCs with osteoblast differentiation medium as described previously [33]. The bone formation activity of osteoblasts was determined using Alizarin red S staining (Sigma-Aldrich). Human monocytes were isolated from peripheral blood mononuclear cells and cultured to obtain the precursors of osteoclasts. The precursors derived from human monocytes were cultured in M-CSF (25 ng/ml), a low dose of RANKL (10 ng/ml) and cocultured with or without myeloma cells for 7 days to induce mature osteoclast formation. For the detection of mature osteoclasts, TRAP staining was performed using a leukocyte acid phosphatase kit (Sigma-Aldrich).

Quantitative real-time PCR of mRNAs
Total RNA was isolated using a RNeasy kit (QIAGEN). An aliquot of 1 g of total RNA was subjected to reverse transcription (RT) with a SuperScript II RT-PCR kit (Invitrogen). Quantitative PCR was performed using SYBR Green Master Mix (Life Technologies) with the QuantStudio 3 Real-Time PCR System (Life Technologies). The reaction was performed with the following settings: 95 °C for 10 min, followed by 40 cycles of 95 °C for 15 s and 60 °C for 60 s. The primers used are listed in Table S1.

Quantitative real-time PCR of miRNAs
Total RNA was isolated with the mirVana miRNA Isolation Kit (Ambion). Quantification of the mature form of miRNAs was performed with a Bulge-LoopTM miRNA Quantitative real time PCR primer kit (RiBoBio, Guangzhou, China). The U6 small nuclear RNA was used as an internal control.

Cell viability, Cell cycle, soft agar colony formation assays and ELISA
For viability assays, cells were plated at 1 × 10 4 cells/well in triplicate. Assays were performed using CellTiter-Glo Luminescent Cell Viability Assay Kit (Promega) or Cell Counting Kit-8 (CCK-8) (Dojindo). Cell proliferation was also measured using the 5-ethynyl-2'-deoxyuridine (EdU) assay kit (RiboBio) according to the manufacturer's instructions. For the cell cycle analysis, the cells were fixed and then stained with PI staining buffer (Multisciences Biotech) for 30 min at room temperature in the dark, and measured using flow cytometry. Soft agar colony formation assays were performed as previously described. Briefly, the 5 × 10 4 cells in 0.4% Noble agar were plated on top of the 0.8% Noble agar bottom layer in a 6-well plate. After 3 weeks, they were stained with 1 mg/ml p-iodonitrotetrazolium chloride for visualization and counting. In addition, mouse serum was collected and measured using an ELISA kit (Immunodiagnostic Systems) according to the manufacturer's instructions.

Immunohistochemistry
Formalin-fixed, paraffin-embedded sections of bone marrow biopsy samples obtained from patients with myeloma were deparaffinized and stained. Slides were stained with anti-CD138 (LS-B9360, LifeSpan BioSciences) and hnRNPA2B1 (LS-B10604, LifeSpan BioSciences) antibody using an EnVision System (#K5361, DAKO) following the manufacturer's instructions and counterstained with hematoxylin.

Transfection of miRNA mimics or inhibitors
Cells were transfected with the indicated miRNA mimics or inhibitors (50 nM) or control oligonucleotides (50 nM) (Thermo Fisher Scientific) using the Oligofectamine reagent (Invitrogen). At 48 hours after transfection, cells were harvested for RNA and protein analyses.

Fluorescent Staining
MM cells were fixed with 4% formaldehyde and permeabilized with 0.3% Triton X-100 in 1 × PBS. After blocking with 2% goat serum, the cells were stained with antibodies against hnRNPA2B1 (Santa Cruz) or DGCR8 (Abcam) at 4°C overnight, followed by incubation with Alexa 594-or Alexa 488-conjugated secondary antibodies (Abclonal) for 30 min at room temperature and the cell nuclei were stained with DAPI and mounted with antifade reagent (Molecular Probes). Immunofluorescent images were acquired with an IX71 confocal microscope system (Olympus).

Luciferase assay in vitro
The wild type and mutated 3' UTR of human IRF8 and RUNX2 were subcloned into the pGL2 vector (Addgene). Mutant forms of 3' UTR were mutated from wild type using the QuikChange site-directed mutagenesis kit (StrataGene). The constructs (2 ng) were co-transfected into HEK293T cells in 96-well plates together with 200 ng of control plasmid or plasmids expressing miR-92a-2-5p or miR-373-3p and Renilla plasmid (0.2 ng). Luciferase activity was measured 48 hours after transfection using a Dual-Luciferase Reporter Assay System (Promega) according to the manufacturer's instructions. The primers used in the subcloning are listed in Table S2.

Exosome isolation and characterization
Primary myeloma cells, myeloma cell lines and breast cancer cell line were used for exosome production. In brief, cells were cultured in respective media with microvesicle-free fetal bovine serum for 48 hours. Conditioned media was collected, centrifuged twice at 3000 rpm for 10 min to remove debris. The supernatant was centrifuged at 100000 × g for 60 min to collect exosomes.
The pellet containing exosomes was resuspended in 1 × PBS buffer. They were examined by transmission electron microscopy (High Resolution Electron Microscopy Facility at Xiamen University). The hydrodynamic size distribution of exosomes were determined by dynamic light scattering (DLS) system. Purified exosomes were incubated with 4 μmdiameter aldehyde/sulphate latex beads (Interfacial Dynamics) in PBS buffer overnight at 4°C. Exosome could be stained with exosome marker antibodies, anti-CD9 (BioLegend) or anti-CD63 (BioLegend) for 30 min at 4°C, and analyzed using flow cytometer (BD Biosciences).

Exosome uptake assay
Dil cell-labeling solution (Thermo Fisher Scientific) was used to label exosomes. Briefly, The Dil-labeled exosomes were added to the culture of precursors of osteoclasts or MSCs. After 24 h, cells were collected, fixed with 4% paraformaldehyde, stained with DAPI, and then observed under confocal microscopy.

RNA pull-down and RIP assays
The interactions between hnRNPA2B1 and primary miR-92a-2-5p or miR-373-3p were examined using RNA pull-down assays according to the instructions of the Pierce Magnetic RNA-Protein Pull-Down Kit (Thermo Fisher Scientific). Biotinylated primary miR-92a-2-5p or miR-373-3p and antisense sequences were synthesized using a TranscriptAid T7 High Yield Transcription Kit (Thermo Fisher Scientific). The nuclear proteins obtained using a NE-PER Nuclear Protein Extraction Kit (Thermo Fisher Scientific) was incubated overnight with biotinylated primary miR-92a-2-5p or miR-373-3p, followed by precipitation with streptavidin magnetic beads. The retrieved protein was eluted from the RNA-protein complex and analyzed by immunoblotting.
The RIP assays were performed using an EZ-Magna RIP kit (Millipore). Lysates of RPMI8226 or MM.1S cells obtained using RIP lysis buffer were immunoprecipitated with RIP buffer containing anti-hnRNPA2B1 or m6A antibody-conjugated magnetic beads (Abcam). The precipitated RNAs were analyzed by Quantitative real time PCR. IgG was used as the negative control.

In vivo mouse experiments, measurement of tumor burden, radiography and bone histomorphometry
CB.17 SCID mice purchased from Charles River Labs, Beijing, China, were maintained in Xiamen University Animal Care-accredited facilities. The mouse studies were approved by the Institutional Animal Care and Use Committee of Xiamen University. Myeloma cells (RPMI8226 or MM.1S) (5 × 10 5 cell/mouse) were injected into the femurs of 8-week-old SCID mice. To monitor the tumor burden, serum samples were collected from the mice weekly and tested for myeloma-secreted M proteins using ELISA analysis. To examine the lytic bone lesions, radiographs were scanned with a Bruker In-Vivo Xtreme imaging system. Bone tissues were fixed in 10% neutral-buffered formalin and decalcified, and sections of them were stained with toluidine blue or TRAP following standard protocols. Both analyses were done using the BIOQUANT OSTEO (v18.2.6) software program (BIOQUANT Image Analysis Corporation).

Statistical analysis
Statistical significance was analyzed using the Graphpad (Version 9.0) program with two tailed unpaired Student t-tests for comparison of two groups, and one-way ANOVA with Tukey's multiple comparisons test for comparison of more than two groups. P values less than 0.05 were considered statistically significant. All results were reproduced in at least three independent experiments.