RAB22A as a predictor of exosome secretion in the progression and relapse of multiple myeloma

Source of data

We obtained 78 samples from GSE5900 dataset in GEO database, including 22 normal people, 44 MGUS patients and 12 SMM patients. We also extracted the data of 414 newly diagnosed MM patients from GSE4581 dataset for the analysis of RAB22A expression level and biological function. In addition, we extracted the MSCs expression data of 12 newly diagnosed and 9 relapsed MM patients from GSE146649 dataset, and analyzed the expression levels of RAB22A and exosome-related genes. We use “sva” package to correct batch effect in these data sets.

Analysis of differentially expressed genes (DEGs) of RAB22A

We extracted the RAB22A expression level of 414 newly diagnosed MM patients with GSE4581 by using the “limma” package, and divided the samples into high RAB22A expression group and low RAB22A expression group according to the median value. The heatmap and volcano plot showed the DEGs in the high and low RAB22A expression groups. Gene Ontology (GO) enrichment analysis and Kyoto encyclopedia of genes and genomes (KEGG) enrichment analysis were carried out to further explore the potential biological functions.

Correlation analysis of RAB22A with EMT, exosome, immune checkpoints and m6A related genes

In order to further clarify the biological function of RAB22A in MM, we selected EMT, exosome, immune checkpoint and m6A related gene sets. We obtained EMT-related genes from the website of dbEMT2.0, and intersected EMT genes with the DEGs related to RAB22A. Exosome gene set is divided into three modules, including genes related to extracellular vesicle, exosome marker genes and genes regulating exosome secretion [20–23] (see in Supplementary Table 1). We selected 25 immune checkpoint genes (Supplementary Table 2). In addition, we selected 14 m6A-related genes for analysis (GSE4581 dataset lacks some m6A-related genes), as shown in Supplementary Table 3. Wilcox. test was used to calculate the correlation between gene sets and RAB22A, and the results were presented by boxplot and heatmap.

Analysis of immune cell infiltration

We used the “CIBERSORT” package for immune cell infiltration analysis. The degree of immune cell infiltration in high and low RAB22A expression groups was analyzed, and the “ggplot2” and “ggpubr” packages were used to draw borplot and barplot. In addition, we further drew scatterplots between the expression level of RAB22A and the infiltration level of immune cells.

Calculation of EMT and exosome scores

We downloaded the “IOBR” package with “calculate_sig_score” function and used “PCA” and “ssGSEA” algorithm to calculate EMT scores and exosomes scores of different RAB22A groups.

Clinical samples and cell lines

We collected the clinical information and bone marrow samples of 71 normal controls and 30 newly diagnosed MM patients in the hematology department of the Affiliated Hospital of Guizhou Medical University. All participants obtained informed consent, and the research was approved by the ethics committee of the Affiliated Hospital of Guizhou Medical University. Human MM cell lines RPMI 8226 and U266 were provided by Hematopoietic Stem Cell Transplantation Center Laboratory in Guizhou province. All human cell lines were tested for mycoplasma contamination and verified by short tandem repeats. Ficoll gradient centrifugation was used to separate MSCs, and the 2-4 generation cells were selected for further experiments. Cell lines were cultured in RPMI-1640 (Gibco, USA) medium with 10% fetal bovine serum (FBS), penicillin (100 units/mL) and streptomycin (100 mg/mL) at 37° C and 5% CO₂ with saturating humidity. MSCs were cultured in complete medium of bone marrow mesenchymal stromal cells (BM-MSCs) (OriCell, China), and 48 hours before exosomes were extracted, they were changed to medium containing exosome-depleted FBS (OriCell, China).

Quantitative real-time PCR (qRT-PCR)

Total RNA was extracted from cells by using Trizol reagent (Invitrogen, USA) according to the manufacturer’s instruction. SYBR Green PCR Master Mix (MCE, China) and PRISM 7500 real-time PCR detection system (ABI, USA) were used for real-time PCR detection. The primers used in this study are as follows: for RAB22A, forward: 5′-TACCAAAGAGGCAAAGCATGT GCG-3′, reverse: 5′-AGACACCATGCAATCACCAACAGC-3′; for β-actin, forward: 5′-GAGACCTTCAACACCCCAGC-3′, reverse:5′-ATGTCACGCACGATTTCCC-3′. The relative expression of mRNA was normalized to β-actin and calculated by 2^-ΔCT.

Western blotting analysis

Western blotting was used to detect the expression of RAB22A in normal people and MM patients, and normalized it to β-actin. BCA protein assay kit (Pierce, Hercules, USA) was used to determine the protein concentration. The protein samples were loaded on 10% SDS-PAGE gel and transferred to PVDF membranes. After being sealed with 5% skim milk, the membrane was incubated overnight at 4° C with the following antibodies: anti-RAB22A (1:500, Affinity, China) and anti β-actin (1:3000, Proteintech Group, China). Protein bands were observed using ECL kit (7 Sea Biotechnology, China).

Lentiviral transduction

Human RAB22A-RNAi was purchased from Genechem Co., Ltd. (Shanghai, China). Transfection of RAB22A was performed according to the manufacturer’s protocol. Cells (RPMI 8226, U266 cell line and MSCs) transfected with empty vector (EV) were used as controls. After amplification and maintenance in culture medium, stable RPMI 8226, U266 cell lines and MSCs expressing si-RAB22A were selected by puromycin (1ug/ml).

Extraction and identification of exosomes

According to the manufacturer’s instructions, the supernatant of MSCs cultured in medium containing exosome-depleted FBS was collected. Exosome concentration solution (MCE, China) was added to MSC supernatant, and crude exosomes were obtained after centrifugation. Purified exosomes were obtained by centrifugation with exosome purification filter (MCE, China). Western blotting was used to identify exosome surface markers CD63 (Immunoway, USA) and TSG101 (Immunoway, USA).

Immunofluorescence

The exosomes resuspended in 200ul phosphate buffer saline were incubated with PKH26 (Solarbio, China) in the dark for 5 minutes. MM cell lines were inoculated into 24-well plates, and the labeled MSCs-derived exosomes were added. After 24 hours, the cells were collected and fixed with formaldehyde, then 4’,6-diamidino-2-phenylindole (DAPI) was added for nuclear staining. After that, the cells were observed under a fluorescence microscope.

Cell activity detection

Cells were inoculated into a 96-well plate, and exosomes filtered with a 0.22um filter were added. Cell Counting Kit-8 (CCK-8) was added at different time points and incubated at 37° C for 2 hours. The absorbance at 450nm was measured by spectrophotometer.

Statistical analysis

R software package (version 4.1.3) was used for statistical analysis and visualization of results. Shapiro-Wilk normality test was used to test the normality of variables. Unpaired student’s t-test was used to evaluate the statistical significance of the data that conforms to the normal distribution. Using 1: 1 properness score matching (PSM) to eliminate the bias between the control group and the newly diagnosed MM group. P<0.05 indicated statistical significance.

Data availability statement

All data supporting the results of this study can be obtained from the corresponding authors according to reasonable requirements. The datasets presented in this study can be found in online repositories. The names of the repository can be found below: https://portal.gdc.cancer.gov.

The expression of RAB22A increased in MM

Firstly, we explored the expression level of RAB22A in normal people, MGUS, SMM and MM patients, and found that the expression level of RAB22A in SMM and MM patients was significantly higher than that in normal people and MGUS patients (Figure 1A). In order to further clarify the prognostic role of RAB22A in MM, we analyzed the expression level of RAB22A in newly diagnosed and relapsed MM patients. The results showed that the expression level of RAB22A in relapsed MM patients was significantly higher than that in newly diagnosed MM patients (Figure 1B). These results showed the prognostic value of RAB22A for MM patients, which urged us to further explore the biological function of RAB22A.

Figure 1. Differential analysis of expression level of RAB22A. (A) Differential expression analysis of RAB22A in normal people, MGUS, SMM and MM patients. (B) Differential expression of RAB22A in MSC of newly diagnosed and relapsed MM patients.

Analysis of DEGs related to RAB22A

According to the median value of RAB22A in GSE4581 dataset, we divided the samples into high RAB22A expression group and low RAB22A expression group. The DEGs of the two groups were presented in the heatmap and volcano plot, and we marked EMT genes (Figure 2A, 2B) in the volcano plot. Then, we carried out GO and KEGG enrichment analysis on the DEGs. The results of GO enrichment analysis showed that biological process (BP) was mainly related to membrane lipid metabolic process, sphingolipid metabolic process and positive regulation of leukocyte mediated immunity. Cellular component (CC) was mainly related to cell leading edge, bicellular tight junction and apical junction complex. And Molecular function (MF) was related to cell adhesion mediator activity, integrin binding and microtubule motor activity (Figure 2C, 2D). KEGG results are shown in the Supplementary Figure 1. These results suggest that DEGs may be involved in intercellular communication, migration and immunity of MM.

Figure 2. Differentially expressed genes and enrichment analysis in different RAB22A groups. (A, B) The differentially expressed genes of high and low RAB22A groups were presented by heatmap and volcano plot. (C, D) GO enrichment analysis was performed on differentially expressed genes.

Correlation between RAB22A and EMT

We obtained the DEGs in the high and low RAB22A expression groups, and intersected them with the EMT gene set to obtain 15 EMT genes related to RAB22A. We showed the EMT genes expression levels (Figure 3A, 3B) among different RAB22A expression groups by using heatmap and boxplot. The results showed that the expression levels of MYBL2, ITGB1, CDK5, CBR1, EPCAM, S100A4, FOXM1, TIMP1, BHLHE40 and AHR in high RAB22A expression group were higher than those in low RAB22A expression group, while the expression levels of DKK1, TNFSF10, CCR5, PCDH9 and KIT in high RAB22A expression group were lower. In addition, we drew scatterplots of the correlation between RAB22A and EMT genes (Figure 3D–3R). Finally, we calculated the EMT scores of different RAB22A groups through the “IOBR” package, and the results showed that the EMT scores of high RAB22A expression group were significantly higher than those of low RAB22A expression group (Figure 3C).

Figure 3. Correlation analysis between RAB22A and EMT. (A, B) EMT-related genes with significantly different expression levels between high and low RAB22A groups. (C) EMT scores of high and low RAB22A groups were calculated. (D–R) Scatterplots and correlation of RAB22A and EMT-related genes.

Correlation between RAB22A expression and exosome scores

Exosomes contain microRNAs (miRNAs), mRNAs, proteins, lipids and so on, which are used for information transmission between cells. MSCs are an important part of bone marrow microenvironment, and BM-MSCs can mediate drug resistance of many hematological neoplasms [24]. BM-MSCs can influence tumor cells through various information interaction ways, and induce tumor cells to escape the killing of chemotherapy drugs. Previous studies have found that exosomes secreted by BM-MSCs in tumor microenvironment affect the progression of leukemia by transporting microRNA-7-5p into acute myeloid leukemia (AML) cells [25]. Therefore, the exosomes secretion of MSCs may be an important factor leading to the progress of hematological tumors, which urges us to pay attention to the genes regulating exosomes secretion.

We extracted the expression data of MSCs from 12 newly diagnosed MM patients and 9 relapsed MM patients in GSE146649 dataset, and calculated the exosome scores with “ssGSEA” algorithm. The results showed that the exosome scores of MSC in relapsed patients was significantly higher than that in newly diagnosed patients (Figure 4A). Then, we further analyzed the expression levels of exosome-related genes in newly diagnosed and relapsed patients. The results showed that the expression levels of CFL1, ACTB, CCT4, RDX, TSG101, FLOT1, CTTN, and STAT3 in relapsed patients were significantly higher than those in newly diagnosed patients (Figure 4B, 4C). In order to further clarify the effect of RAB22A on exosomes secretion, we extracted the expression data of RAB22A from newly diagnosed and relapsed patients, and divided the samples into high and low RAB22A expression groups according to the median value. The results showed that the exosome scores of high RAB22A expression group was significantly higher than that of low RAB22A expression group (Figure 5A). Heatmap and boxplot showed that the expression levels of exosome-related genes in high RAB22A expression group and low RAB22A expression group. The results showed that the expression levels of CFL1, ACTB, CCT4, RDX, TSG101, CD81, FLOT1, PIKFYVE, CTTN and STAT3 in high RAB22A expression group were significantly higher than those in low RAB22A expression group (Figure 5B, 5C). Finally, we showed the correlation between the expression level of RAB22A and exosome genes by scatterplots (Figure 5D–5N).

Figure 4. Relationship between MM disease state and exosomes secretion. (A) Exosome scores of MSCs in newly diagnosed and recurrent patients. (B, C) Heatmap and boxplot showed the expression level of exosome-related genes in MSCs of newly diagnosed and relapsed MM patients.

Figure 5. Relationship between RAB22A and exosomes secretion. (A) Exosome scores in high and low RAB22A groups. (B, C) Expression level of exosome-related genes in high and low RAB22A groups. (D–N) Correlation and scatterplots between RAB22A and exosome-related genes.

The correlation between RAB22A and immune cell infiltration and immune checkpoint genes

We explored the infiltration levels of immune cells in different RAB22A expression groups in the samples of GSE4581 data set. The infiltration levels of CD56 dim natural killer cell, gamma delta T cell, myeloid-derived suppressor cell (MDSC), plasmacytoid dendritic cell, type 1 T helper cell and type 2 T helper cell in the high RAB22A expression group were significantly higher than those in the low RAB22A expression group (Figure 6A, 6B). In order to further clarify the correlation between RAB22A and immune cell infiltration, we drew scatterplots. The results showed that monocyte, NK cells resting, eosinophils, T cells regulatory (Treg) and T cells CD4 memory activated were positively correlated with RAB22A. While plasma cells and T cells gamma delta were negatively correlated with RAB22A (Figure 6C–6J).

Figure 6. Relationship between RAB22A and immune cell infiltration. (A, B) Infiltration level of immune cells in high and low RAB22A groups. (C) Correlation between RAB22A and infiltration level of immune cells. (D–J) Scatterplots of RAB22A and immune cell infiltration.

In addition, we selected 25 immune checkpoint genes and analyzed the expression levels of these genes in high and low RAB22A expression groups. The results showed that the expression levels of BTNL3, BTNL9, C10orf54, CD209, CD274, CD47, CD80, CTLA4, LAG3, PVR, TDO2, TNFRSF14 and TNFSF14 in the high RAB22A group were lower than those in the low RAB22A group, while CD276, HLA-B, HLA-DMA, HLA-DRA, SIRPA and TNFRSF4 were higher in the high RAB22A group (Figure 7A, 7B).

Figure 7. Relationship between RAB22A and immune checkpoint genes. (A) Heatmap showed the gene expression level of immune checkpoint genes in high and low RAB22A groups. (B) Boxplot showed the gene expression level of immune checkpoint genes in high and low RAB22A groups.

RAB22A was associated with m6A RNA methylation regulators

M6A RNA methylation is the most common internal modification in mammalian mRNA, which plays a key biological role by regulating important cellular processes, especially in cancer, where the modification disorder of m6A methylation is often observed. We analyzed the expression levels of m6A-related genes in different RAB22A expression groups. The results showed that the expression levels of METTL14, VIRMA, RBM15 and FMR1 in high RAB22A expression group were significantly higher than those in low RAB22A expression group (Figure 8A, 8B), and the scatterplots showed that these genes were positively correlated with RAB22A (Figure 8C–8G). In addition, we constructed a nomogram (Figure 8H) based on the m6A methylation genes significantly related to RAB22A, which was used to predict the risk value of overexpression of RAB22A. The calibration curve showed that the nomogram had good predictive ability (Figure 8I).

Figure 8. Relationship between RAB22A and m6A methylation. (A, B) Expression level of m6A methylation-related genes in different RAB22A groups. (C–G) Correlation and scatterplots between RAB22A and m6A methylation-related genes. (H, I) Nomogram and calibration curve of the risk of expressing RAB22A.

Drug sensitivity testing

We screened sensitive drugs according to the difference of half maximum inhibition concentration (IC50) between high and low RAB22A expression groups. After consulting the literature, we focused on the drugs that have been used or have prospects in MM. It was worth noting that the IC50 values of doxorubicin, bcl-2 inhibitor (Navitoclax) and cisplatin in low RAB22A expression group were significantly lower than those in high RAB22A expression group (Figure 9A–9C).

Figure 9. (A–C) Drugs with significantly different IC50 values between high and low RAB22A groups.

The effect of RAB22A on exosome secretion and cell proliferation

We collected the clinical information of 71 normal controls and 30 newly diagnosed MM patients. After 1: 1 propensity score matching (PSM) was performed to eliminate the bias caused by sex and age, 25 normal controls and 25 newly diagnosed MM patients were included for follow-up analysis (Supplementary Table 4). We used western blot and qRT-PCR to verify the difference in the expression level of RAB22A. The results showed that the expression level of RAB22A in MM was significantly higher than that in normal people (Figure 10A, 10B, 10E), and the expression level in MM-BMSC was significantly higher than that in Normal-BMSC (N-BMSC) (Figure 10C, 10D, 10F). Then, we down-regulated RAB22A expression in U266, RPMI 8226 cell lines and MM-MSCs by lentivirus transfection. After down-regulating RAB22A in MM cell line, the cell proliferation was significantly decreased (Figure 10G). We extracted the exosomes of MSCs from the control group and the si-RAB22A group, identified them by exosome markers (Figure 11A), and co-cultured the exosomes from MSCs with MM cell lines. Immunofluorescence staining results showed that the secretion of exosomes decreased after inhibiting the expression of RAB22A in MSCs (Figure 11C). The results of CCK-8 assay showed that compared with the exosomes of MSCs that down-regulated RAB22A, the exosomes of MSCs in control group significantly promoted the proliferation of MM cells (Figure 11B).

Figure 10. To verify the expression level of RAB22A in clinical samples and cell lines. (A–D) Western blotting results showed that the expression level of RAB22A protein in MM patients was significantly higher than that in normal people. (E, F) qRT-PCR results showed that the expression level of RAB22 gene in MM patients was significantly higher than that in normal people. (G) CCK-8 assay showed that the OD value decreased after down-regulating the expression of RAB22A in MM cell lines.

Figure 11. Effects of regulating RAB22A on exosomes secretion and cell proliferation. (A) Verification of exosomes by western blotting. (B) The exosomes of MSCs in control group and down-regulated RAB22A group were co-cultured with MM cell line, and the OD values of MM cells at different time points were detected by CCK-8 method. (C) The exosomes of control group and down-regulated RAB22A group were collected and stained with PKH26. The stained exosomes were co-cultured with MM cells, and the endocytosis of exosomes was observed under fluorescence microscope.