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

Abstract

The impact of exosome-mediated crosstalk between multiple myeloma (MM) cells and osteoclasts (OCs) on bone lesions remains to be investigated. Here, we identified NSUN2 and YBX1-mediated m5C modifications upregulated LncRNA MALAT1 expression in MM cells, which could be transported to OCs via exosomes and promote bone lesions. Methodologically, RNA-seq was carried out to detect the cargoes of exosomes. TRAP staining and WB were used to evaluate osteoclastogenesis in vitro. Micro-CT and bone histomorphometric analyses were performed to identify bone destruction in vivo. RNA pull-down, RIP, MeRIP, and luciferase reporter assays were used to test the interactions between molecules. The clinical features of MALAT1, NSUN2 and YBX1 were verified through public datasets and clinicopathological data analyses. Mechanistically, MALAT1 was the highest expressed lncRNA in U266 exosomes and could be transported to RAW264.7 cells. MALAT1 could enhance the differentiation of RAW264.7 cells into OCs by stimulating RANKL expression and its downstream AKT and MAPKs signaling pathways via a ceRNA mechanism. Additionally, MALAT1 could be modified by NSUN2, an m5C methyltransferase, which in turn stabilized MALAT1 through the "reader" YBX1. Clinical studies indicated a notable positive correlation between MALAT1, NSUN2, YBX1 levels and bone destruction features, as well as with RANKL expression.

Fig. 1. MM exosomes promote bone lesion formation.

A The shapes and sizes of U266 exosomes were observed by TEM. B The average diameter and concentration of exosomes were calculated by NTA. C U266 exosomes were labeled with PKH26 (red) and co-cultured with RAW264.7 cells. The nuclei of RAW264.7 cells were stained with DAPI (blue). Scar bar = 10 μm. D TRAP staining was performed on RAW264.7 cells under the following treatment conditions: (1) untreated cells (negative control), (2) cells treated with sRANKL alone, (3) cells treated with sRANKL and MM-conditioned culture medium (MM-CCM), (4) cells treated with sRANKL and MM exosomes, (5) cells treated with sRANKL and MM-CCM collected from MM cells pretreated with GW4869, and (6) cells treated with sRANKL, MM-CCM collected from MM cells pretreated with GW4869, and MM exosomes. TRAP-positive multinucleated cells containing more than three nuclei were counted and analyzed (n = 3 independent experiments; plotted as mean ± s.d. after normalization to the negative control group). Scar bar = 50 μm. P-value by one-way ANOVA followed by Tukey’s multiple comparison test. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Fig. 2. MM exosomes promoted RANKL expression and induced AKT and MAPK activation.

A Western blot result of RANKL, p-p-38, p-38, p-ERK, ERK, p-JNK, JNK, p-AKT, and AKT of RAW264.7 cells treated with nothing (negative control), sRANKL, sRANKL and U266-CCM, sRANKL and U266 exosomes, sRANKL and U266-CCM collected from U266 cells pretreated with GW4869 and sRANKL, U266-CCM collected from U266 cells pretreated with GW4869, and U266 exosomes. B, C Densitometric analysis of protein expression (n = 3 independent experiments; plotted as mean ± s.d. after normalization to β-actin and negative control group). D The mRNA expression of RANKL (n = 3 independent experiments; plotted as mean ± s.d. after normalization to GAPDH and negative control group). E Western blot results of RANKL of RAW264.7 cells treated with nothing (negative control), sRANKL, sRANKL and MM1S/RPMI8226-CCM, sRANKL and MM1S/RPMI8226 exosomes, sRANKL and MM1S/RPMI8226-CCM collected from MM cells pretreated with GW4869 and sRANKL, MM1S/RPMI8226-CCM collected from MM cells pretreated with GW4869 and MM1S/RPMI8226 exosomes. F Densitometric analysis of protein expression (n = 3 independent experiments; plotted as mean ± s.d. after normalization to GAPDH and negative control group). P-value by one-way ANOVA followed by Tukey’s multiple comparison test. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Fig. 3. LncMALAT1 was identified as the most highly expressed lncRNA in U266 exosomes and upregulated RANKL.

A MALAT1 levels were significantly increased in MM samples. Patients were designated healthy donors with normal BM plasma cells (NP, n  =  22) or multiple myeloma (MM, n  =  559). B FISH analysis of MALAT1 expression in BM tissues from healthy donors (n = 4) and BM tissues from MM patients (n = 36) (plotted as mean ± s.d. after normalization to the control group). Scar bar = 20 μm. C Intercellular trafficking of MALAT1 packaged into exosomes and taken up by RAW264.7 cells by isolated U266 exosomes labeled with PKH26 dye and MALAT1 labeled with GFP and cultured with RAW246.7 cells. The nuclei of RAW264.7 cells were stained with DAPI (blue) (n = 3 independent experiments). Scar bar = 25 μm. D, E Relative MALAT1, and RANKL RNA expression in RAW264.7 cells were detected by qRT-PCR (n = 3 independent experiments; plotted as mean ± s.d. after normalization to GAPDH and control group). The groups were as follows: Control: Cells treated with sRANKL alone. EXO: Cells treated with sRANKL and exosomes derived from untreated U266 cells. EXO-siNC: Cells treated with sRANKL and exosomes derived from U266 cells transfected with a negative control siRNA. EXO-siMALAT1: Cells treated with sRANKL and exosomes derived from U266 cells transfected with siRNA targeting MALAT1. EXO-Vector: Cells treated with sRANKL and exosomes derived from U266 cells transfected with an empty vector. EXO-MALAT1: Cells treated with sRANKL and exosomes derived from U266 cells transfected with a vector overexpressing MALAT1. F Western blot results of RANKL in RAW264.7 cells (n = 3 independent experiments). G TRAP staining of RAW264.7 cells. TRAP-positive multinucleated cells containing more than three nuclei were counted and analyzed (n = 3 independent experiments; plotted as mean ± s.d. after normalization to the control group). Scar bar = 50 μm. P-value by Welch’s t-test (A, B), or one-way ANOVA followed by Tukey’s multiple comparison test (D, E, G). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Fig. 4. NSUN2-mediated m5C modification enhances the stability of MALAT1 in MM.

A An RNA pulldown experiment was conducted to examine the interaction between NSUN2 and MALAT1. Antisense MALAT1 was used as a negative control (n = 3 independent experiments). B The m5C MeRIP assay followed by qRT-PCR was used to assess the m5C level of MALAT1 in U266 cells transfected with control or NSUN2 siRNAs (n = 3 independent experiments; plotted as mean ± s.d. after normalization to input RNA levels and IgG group). C siRNA targeting NSUN2 was transfected into U266 cells, and MALAT1 was tested using qRT-PCR (n = 3 independent experiments; plotted as mean ± s.d. after normalization to GAPDH and siNC group). D RIP assays using NSUN2 antibody. IgG was used as a negative control (n = 3 independent experiments; plotted as mean ± s.d. after normalization to input RNA levels and IgG group). E U266 cells transfected with control or NSUN2 siRNAs were exposed to actinomycin D (5 μg/mL), and cellular RNA was isolated at the indicated times. qRT-PCR was performed to assess the half-life of MALAT1 (n = 3 independent experiments; plotted as mean ± s.d. after normalization to the 0-hour time point). F siRNA targeting YBX1 was transfected into NSUN2-overexpressing U266 cells, and the expression level (n = 3 independent experiments; plotted as mean ± s.d. after normalization to GAPDH and control group). G The stability of MALAT1 in NSUN2-overexpressing U266 cells transfected with YBX1 siRNA was assessed (n = 3 independent experiments; plotted as mean ± s.d. after normalization to the 0-hour time point). H NSUN2 and YBX1 RNA levels were significantly increased in MM samples. Patients were designated healthy donors with normal BM plasma cells (NP, n  =  22) or multiple myeloma (MM, n  =  559). I Representative IHC images of healthy donors (n = 3) and MM patients (n = 36) BM locating NSUN2 and YBX1 (plotted as mean ± s.d. after normalization to the control group). Scar bar = 50 μm. Colorectal cancer tissue was used as a positive control. Negative controls were bone marrow tissues stained with the primary antibody omitted and replaced with PBS (Additional file 3). P-value by two-tailed Student’s t-test (B–E, H), Welch’s t-test (I), or one-way ANOVA followed by Tukey’s multiple comparison test (F, G). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001. ▲P <  0.05 versus the NSUN2-OE group.

Fig. 5. MALAT1, NSUN2, and YBX1 were positively correlated with bone destruction features in MM patients.

To study the correlation between MALAT1, RANKL, YBX1, NSUN2, and bone lesion formation, we detected their expression in BM tissues of NDMM patients (n = 36) by FISH or IHC. MALAT1 expression positively correlated with bone lesion formation (A), corrected serum calcium (B), and RANKL level (C). D MALAT1 expression positively correlated with YBX1 and NSUN2 levels. NSUN2 (E) and YBX1 (F) expression positively correlated with bone lesion formation, RANKL level, and corrected serum calcium (G). H The accuracy of MALAT1, NSUN2, and YBX1 in predicting the death outcome of MM patients with bone lesions. I Representative IHC (scar bar = 50 μm) and FISH (scar bar = 10 μm) images of high-risk and low-risk MM patients (n = 36) BM locating NSUN2, YBX1, and MALAT1 (plotted as mean ± s.d. after normalization to low-risk group). For FISH, MALAT1 expression was assessed specifically in myeloma cells by co-staining with CD138. For IHC, colorectal cancer tissue was used as a positive control. Negative controls were bone marrow tissues stained with the primary antibody omitted and replaced with PBS (Additional File 3). J Cumulative survival function of overall survival analyses comparing MALAT1, NSUN2, and YBX1 high-risk and low-risk groups (P < 0.01). P-value by Welch’s t-test (I). Pearson correlation analyses (A–G). Log-rank test (J). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Fig. 6. MALAT1 sponges miR-17 to promote RANKL expression in MM.

A Relative expression levels of miR-17, miR-34a, miR-146a, miR-125a, and miR-101-1 were measured in RAW264.7 cells under various treatment conditions by qRT-PCR (n = 3 independent experiments; plotted as mean ± s.d. after normalization to U6 and control group). The groups were as follows: Control: Cells treated with sRANKL alone. EXO: Cells treated with sRANKL and exosomes derived from untreated U266 cells. EXO-siNC: Cells treated with sRANKL and exosomes derived from U266 cells transfected with a negative control siRNA. EXO-siMALAT1: Cells treated with sRANKL and exosomes derived from U266 cells transfected with siRNA targeting MALAT1. EXO-Vector: Cells treated with sRANKL and exosomes derived from U266 cells transfected with an empty vector. EXO-MALAT1: Cells treated with sRANKL and exosomes derived from U266 cells transfected with a vector overexpressing MALAT1. B The direct interaction of miR-17 with MALAT1 and RANKL was verified by dual-luciferase reporter assays (n = 4 independent experiments; plotted as mean ± s.d. after normalization to the control group). C Western blot results of RANKL in RAW264.7 cells treated with sRANKL, sRANKL + U266 exosomes, sRANKL + U266-siNC exosomes + miR-NC inhibitors, sRANKL + U266-siMALAT1 exosomes + miR-17 inhibitors, sRANKL + U266-vector exosomes + miR-NC mimics or sRANKL + U266-MALAT1 exosomes + miR-17 mimics, and densitometric analysis of protein expression (n = 3 independent experiments; plotted as mean ± s.d. after normalization to β-actin and negative control group). D The mRNA expression of RANKL (n = 3 independent experiments; plotted as mean ± s.d. after normalization to GAPDH and control group). E Cells were fixed, subjected to TRAP staining, and observed under a light microscope. TRAP-positive multinucleated cells containing more than three nuclei were counted and analyzed (n = 3 independent experiments; plotted as mean ± s.d. after normalization to the control group). Scar bar = 50 μm. P-value by two-tailed Student’s t-test (B), or one-way ANOVA followed by Tukey’s multiple comparison test (A, C–E). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Fig. 7. MM exosomes impair osteogenesis and exacerbate bone lesions in an MM mouse model.

Representative photomicrographs of micro-CT scans of the femur (A), spine (B), and skull (C) of mice from the control group, model group, U266 exosomes group, and U266-shMALAT1 exosomes group. Trabecular bone volume (BV/TV), trabecular thickness (Tb.Th), trabecular number (Tb.N), trabecular separation (Tb.Sp), bone-specific surface (BS/BV) and cranial suture of femur (D), spine (E), and skull (F) were assessed by micro-CT (n = 6 independent experiments; plotted as mean ± s.d. after normalization to control group). Representative H&E (scar bar = 50 μm) (G) and TRAP (scar bar = 20 μm) (H) staining of femur sections collected from the indicated groups of mice (n = 3 independent experiments; plotted as mean ± s.d. after normalization to the control group). P-value by one-way ANOVA followed by Tukey’s multiple comparison test. *P < 0.05, **P < 0.01, ***P < 0.001.

Fig. 8. Exosomal MALAT1 promoted RANKL expression and induced AKT and MAPK activation in vivo.

A FISH analysis of MALAT1 expression in femur sections of mice from different groups (n = 6 independent experiments; plotted as mean ± s.d. after normalization to the control group). Scar bar = 20 μm. B IHC analysis was used to detect the presence of p-p38, p-ERK, p-JNK, p-AKT, and RANKL in femur sections collected from the indicated groups of mice (n = 3 independent experiments; plotted as mean ± s.d. after normalization to the control group). Scar bar = 20 μm. Representative images are shown for each group. P-value by one-way ANOVA followed by Tukey’s multiple comparison test. *P < 0.05, **P < 0.01, ***P < 0.001.