Multiple myeloma is an incurable refractory hematologic malignancy arising from plasma cells in the bone marrow. Here we investigated miR-26a function in multiple myeloma and tested single-wall carbon nanotube delivery of miR-26a in vitro and in vivo. miR-26a was downregulated in patients with multiple myeloma cells compared with plasma cells from healthy donors. miR-26a overexpression inhibited proliferation and migration and induced apoptosis in multiple myeloma cell lines. To identify the targets of miR-26a, RPMI8226-V-miR-26-GFP and RPMI8226-V-GFP cells were cultured using stable isotope labeling by amino acids in cell culture (SILAC) medium, followed by mass spectrometry analysis. In multiple myeloma cells overexpressing miR-26a, CD38 protein was downregulated and subsequently confirmed to be a direct target of miR-26a. Depletion of CD38 in multiple myeloma cells duplicated the multiple myeloma inhibition observed with exogenous expression of miR-26a, whereas restoration of CD38 overcame the inhibition of miR-26a in multiple myeloma cells. In a human multiple myeloma xenograft mouse model, overexpression of miR-26a inhibited CD38 expression, provoked cell apoptosis, and inhibited cell proliferation. Daratumumab is the first CD38 antibody drug for monotherapy and combination therapy for patients with multiple myeloma, but eventually resistance develops. In multiple myeloma cells, CD38 remained at low level during daratumumab treatment, but a high-quality response is sustained. In daratumumab-resistant multiple myeloma cells, CD38 expression was completely restored but failed to correlate with daratumumab-induced cell death. Therefore, a therapeutic strategy to confer selection pressure to maintain low CD38 expression in multiple myeloma cells may have clinical benefit.
These results highlight the tumor suppressor function of miR-26a via its targeting of CD38 and suggest the therapeutic potential of miR-26a in patients with multiple myeloma.
Multiple myeloma, a cancer of terminally differentiated postgerminal center B cells or plasma cells, is the second most frequently diagnosed hematologic cancer in the United States (1). Despite recent advances in treatment, including proteasome inhibitors, immunomodulatory drugs, and CD38 mAb therapies, patient survival has increased by only about 1.5 years, from 3.5 to 5.5 years; further, it remains incurable due to frequent relapse and emergence of drug resistance. Fully understanding the mechanisms of multiple myeloma development, progression, and drug resistance are essential to developing new treatments that will improve patient outcome.
The miRNAs are short noncoding RNAs that suppress the expression of protein-coding genes by partial complementary binding, particularly to the 3′-untranslated regions of mRNAs. Disturbance of miRNA expression is involved in the tumorigenesis and metastasis of human cancers (2, 3), including multiple myeloma (4, 5). Studies of the patient with multiple myeloma miRNA signature have consistently shown that several miRNAs are commonly upregulated in multiple myeloma, such as miR-21, miR-221/222, and miR-181a/b, but that others are downregulated, such as miR-30s and miR-15/16 (6–8). Functionally, miR-21 has been reported to promote proliferation by inhibiting PTEN (9), and miR-30s represses multiple myeloma growth via inhibiting Wnt/BCL9 (10). In addition, miRNAs can be therapeutic targets for multiple myeloma—for instance, inhibitors of miR-34a and miR-21 reduced multiple myeloma proliferation in vitro and in vivo (10–12).
The human miR-26 family is composed of three members, miR-26a-1, miR-26a-2, and miR-26b, which are located on chromosomes 3, 12, and 2, respectively. miR-26a-1 and miR-26a-2 have identical sequences, which differ by two nucleotides from miR-26b. miR-26a is usually dysregulated in cancer: it induces cell-cycle arrest through downregulating CCND2 and CCNE2 in hepatocellular cancer (13); functions as a tumor suppressor in breast carcinogenesis by repressing MTDH and EZH2 (14); and in leukemia, inhibits proliferation, migration, invasion, angiogenesis, and metabolism through targeting EZH2, CDK6, and Mcl1 (15, 16). miR-26a delivered via adeno-associated virus suppresses proliferation and promotes apoptosis in xenograft mouse models, suggesting its potential clinical use (13).
In this study, we first investigated the function of miR-26a in cell proliferation and apoptosis in multiple myeloma and identified CD38 as its direct target in vitro and in vivo. Furthermore by knocking down and then restoring CD38 expression, we verified that miR-26a performed its function directly by targeting CD38 and observed a synergistic effect between miR-26a and bortezomib or melphalan treatments, providing proof-of-concept support for systemic delivery of miR-26a as a specific anti-multiple myeloma treatment.
Materials and Methods
miR-26 expression profiling data analysis
We downloaded the GSE16558 dataset (17), and normalized expression levels of miR-26a and miR-26b to the mean Ct value of RNU44 and RNU48, which are consistently expressed across the dataset. Relative quantification of miR-26 expression was calculated with the 2−ΔΔCt method, where ΔCt = Ct(miR-26) − Ct((RNU44+RNU48)/2) and ΔΔCt = ΔCt(MM) − ΔCt(average normal PC) (18).
Patient tissue preparation and cell line culture
Bone marrow, lymph node, and peripheral blood specimens were obtained from healthy donors and from patients with multiple myeloma and chronic lymphocytic leukemia (CLL) in accordance with Cleveland Clinic Foundation Institutional Review Board approval, and written informed consent was obtained in compliance with the Declaration of Helsinki. Normal plasma cells were purified using CD138 magnetic beads (Miltenyi Biotec) as described previously (19). The multiple myeloma cell lines H929, MM.1S, RPMI8226, U266, and the 293T cell line were purchased from ATCC. OPM1, OPM2, and KMS26 were kindly provided by Dr. Teru Hideshima at the Dana-Farber Cancer Institute (Boston, MA). All cell lines were routinely genotyped using the Human Cell Line Genotyping System (Promega) and Mycoplasma tested. Myeloma cells were cultured in RPMI1640 medium, and the 293T cells were cultured in DMEM medium, supplemented with 10% FBS in 5% CO2 in a humidified incubator at 37°C.
Primary human B-cell purification and in vitro differentiation assay
Peripheral blood mononuclear cells were isolated from healthy volunteers, and the naïve B cells were purified by negative selection using magnetic cell separation by using the Naive B Cell Isolation Kit II (Miltenyi Biotec) with anti-CD10 antibodies (Miltenyi Biotec). The purity of the isolated CD19+ CD27− naïve B-cell population was routinely >95%. The cells were labeled with 1 μmol/L CFSE (Invitrogen) in serum-free medium at 37°C for 10 minutes and washed in complete medium so as to monitor cellular division. Purified naive B cells were cultured at 7.5 × 105 cells/mL in 24-well plates and stimulated during 4 days with 2.6 μg/mL F(ab')2 fragment goat anti-human IgA+IgG+IgM (H+L; Jackson ImmunoResearch Laboratories), 100 ng/mL recombinant human soluble CD40L (Millipore), 1.0 mg/mL CpG oligodeoxynucleotide 2006 (Invivogen), and 50 U/mL recombinant IL2 (R&D Systems). Day 4-activated B cells were washed and cultured at 4 × 105 cells/mL for up to 3 days with 50 U/mL IL2, 50 ng/mL IL6, 50 ng/mL IL10, and 2 ng/mL IL12 (R&D Systems). At day 7 of culture, cells were washed and cultured with 50 ng/mL IL6, 10 ng/mL IL15 and 500 U/mL IFNα for 3 days. The cells were subjected to flow cytometry to analyze cell surface CD38, CD138. qRT-PCR was used to examine the expression of BLIMP1 mRNA and miR-26a.
Lentivirus packaging, infection, and transient transfection
Lentivirus packaging and infection of multiple myeloma cells were performed using pPACKH1 HIV Lentivector Packaging Kit according to the manufacturer's protocol (System Biosciences). miR-26a overexpression or control plasmid (System Biosciences) was transfected with packaging plasmids into 293T cells using Lipofectamine 2000 (Thermo Fisher Scientific), which were labeled as V-miR-26a-GFP and V-GFP, respectively. Lentiviral particles were then collected from the culture supernatant at 48 hours intervals and filtered with 0.22 μm filters. To obtain stable cell lines overexpressing miR-26a, multiple myeloma cells were infected with recombinant lentivirus in the presence of polybrene overnight; GFP-positive cells were sorted by flow cytometry 4 days after infection. The expression levels of miR-26a in stable clones and transiently transduced cells were verified with real-time reverse transcriptase quantitative PCR (qRT-PCR). CD38 was knocked down using specific siRNA (Santa Cruz Biotechnology), and protein levels were detected by Western blot analysis.
To obtain CD38 overexpression vector, full length of CD38 ORF was cloned into pCDH-MSCV-MCS-EF1-copGFP-T2A-Puro plasmid (System Biosciences) between XbaI and NotI sites. Primer sequences were:
qRT-PCR analysis was used to determine the relative expression levels of miR-26a and its downstream mRNAs. Total RNA was extracted from cultured cells or healthy donor samples using TRIzol (Thermo Fisher Scientific) according to the manufacturer's instructions. For miRNA analysis, cDNA was synthesized using the TaqMan MicroRNA Reverse Transcription Kit (Thermo Fisher Scientific) and then amplified with TaqMan Universal PCR Master Mix (Thermo Fisher Scientific) together with miR-26a-5p primer (Thermo Fisher Scientific). RNU44 was used as internal control for miR-26a expression. For CD38 and Blimp1 mRNA quantification, cDNA was synthesized using the High Capacity cDNA Reverse Transcription Kit (Thermo Fisher Scientific) and then amplified using SYBR Green PCR Master Mix (Thermo Fisher Scientific) with gene-specific primers. GAPDH was used as the internal control. Gene relative expression was calculated with the 2−ΔΔCt method. Primer sequences were:
In situ hybridization
Slides were deparaffinized and rehydrated through immersion in xylene and an ethanol gradient and then digested with 20 μg/mL proteinase K in prewarmed 50 mmol/L Tris for 20 minutes at 37°C. After fixation in 4% paraformaldehyde for 5 minutes at room temperature, slides were dehydrated by immersion in an ethanol gradient and air drying; slides were prehybridized using DIG Easy Hyb (Roche) at 50°C for 1 hour. The 10 pmol digoxin-labeled miR-26a locked nucleic acid (LNA) probe (5′-AGCCTATCCTGGATTACTTGAA-3′) was denatured in hybridization buffer at 95°C for 2 minutes and then chilled on ice. The probe was diluted in 250 μL prewarmed hybridization buffer. Each sample was covered with 50 to 100 μL diluted probe and incubated in a humidified hybridization chamber at 50°C overnight. Slides were washed twice in 50% formamide in 4 × SSC at 37°C for 30 minutes, and then washed three times in 2 × SSC at 37°C for 15 minutes. After washing twice with maleic acid buffer containing Tween-20 (MABT), slides were blocked using blocking buffer (Roche) at room temperature for 30 minutes, the blocking buffer was drained off, and the samples were incubated with 1:250 diluted anti-digoxigenin-AP fab fragments (Roche) at 37°C for 1 hour. After washing twice in MABT and once in detection solution (0.1M Tris-HCl, 0.1M NaCl, pH 9.5), the slides were stained with freshly diluted NBT/BCIP detection solution (Roche) and incubated at 37°C for 30 minutes. Slides were washed in PBS twice, air dried for 30 minutes, and then mounted with Eukitt quick-hardening mounting medium (Sigma Aldrich). Images were obtained under phase-contrast microscopy (Leica DM2000 LED) and a digital camera (Leica DMC 2900).
Formalin-fixed paraffin-embedded sections were deparaffinized and then incubated with rabbit anti-CD38 polyclonal antibody (Cell Signaling Technology, #14637S) or rabbit anti-Ki-67 polyclonal antibody (Cell Signaling Technology, #9027S) or rabbit anti-cleaved caspase-3 mAb (Cell Signaling Technology, #9664S) at 4°C overnight. After incubation with horseradish peroxidase (HRP)-conjugated goat anti-rabbit secondary antibody, signal was detected using a DAB Substrate Kit (Abcam, ab64238) following the manufacturer's instructions. Images were obtained using phase contrast microscopy (Leica DM2000 LED) and a digital camera (Leica DMC 2900).
Cell viability assay
To determine the effect of miR-26a on multiple myeloma proliferation, 5 × 103 V-miR-26a-GFP/V-GFP-infected RPMI8226, MM.1S, and H929 cells were seeded in 96-well plates and incubated with bortezomib in DMSO (0, 4, or 8 nmol/L for RPMI8226; and 0, 2, or 4 nmol/L for MM.1S and H929), or melphalan in DMSO (0, 10 or 20 μmol/L). Cell viability was evaluated at 24 hours intervals using the CellTiter-GloLuminescent Cell Viability Assay Kit (Promega) according to the manufacturer's protocol. All treatments and measurements were performed in three independent experiments with four replicates of each cell line.
Cell apoptosis analysis
RPMI8226, MM.1S, and H929 cells infected with V-miR-26a-GFP/V-GFP cells were cultured in six-well plates (5 × 105 cells/well); in experiments to test for synergy, cells were also exposed to bortezomib or melphalan for 48 hours. To quantify apoptosis, cells were washed with PBS and stained with Annexin V–APC and propidium iodide (Biolegend). Stained cells were analyzed by flow cytometry (BD FACSCalibur), 10,000 cells were recorded, and the data were statistically evaluated using FlowJo software OSX 10.6 (TreeStar). Some cells were harvested for Western blot analysis to detect expression of cleaved PARP1 and caspase-3 (c-PARP1 and c-caspase-3).
For CD38 knockdown experiments, RPMI8226, MM.1S, and H929 cells were transfected with 50 nmol/L CD38 siRNA or control siRNA (Santa Cruz Biotechnology) for 5 days. Cells were stained with Annexin V–FITC and propidium iodide (Biolegend) and then subjected to flow cytometry (BD FACSCalibur) to analyze cell apoptosis.
Transwell cell migration assay
For migration studies, a transwell assay was conducted in 24-well plates with 8-μm pore inserts (Corning). The bottom chambers were filled with culture medium containing 10% serum. Then 3 × 104 RMPI8226-V-miR-26a-GFP/V-GFP, MM.1S-V-miR-26a-GFP/V-GFP, or H929-V-miR-26a-GFP/V-GFP cells were plated onto the upper chambers with serum-free medium. Monitoring of migration was initiated immediately using an IncuCyte ZOOM system (Essen Bioscience); the bottom transwell chambers were imaged every 2 hours for a total of 9 hours. The scanned images were analyzed using Essenbio software (version 2016B).
Western blot analysis
Total protein samples were isolated with RIPA buffer, separated on 4% to 12% Bis-Tris polyacrylamide gels (Thermo Fisher Scientific), and electroblotted onto nitrocellulose membranes. Membranes were blocked with 5% nonfat milk and incubated overnight with rabbit anti-PARP1 antibody (9532S; Cell Signaling Technology), rabbit anti-caspase-3 antibody (9662; Cell Signaling Technology), rabbit anti-CD38 antibody (14637; Cell Signaling Technology), or actin-HRP (sc-1615; Santa Cruz Biotechnology), as appropriate. Then, membranes were washed and incubated with HRP-linked anti-rabbit IgG secondary antibody (7074; Cell Signaling Technology) or HRP-linked anti-mouse IgG secondary antibody (7076S; Cell Signaling Technology).
Stable isotope labeling with amino acids in cell culture
Stable isotope labeling with amino acids in cell culture (SILAC) experiments were performed using the Pierce SILAC Protein Quantitation Kit (Thermo Fisher Scientific). Briefly, RPMI8226-V-miR-26a-GFP cells were cultured in heavy SILAC medium (100 mg/L 13C6 l-lysine-HCl and 100 mg/L l-arginine-HCl), whereas the RPMI8226-V-GFP cells were cultured in light medium (100 mg/L l-lysine-HCl and 100 mg/L l-arginine-HCl). After 6 doublings, cells were lysed with RIPA buffer. After incorporation efficiency was confirmed by mass spectrometry, 20 μg protein from each cell lysate was mixed in a 1:1 (heavy:light) ratio, electrophoresed on 4% to 12% Bis-Tris polyacrylamide gels, and stained with Coomassie Brilliant Blue for band visualization. Protein bands were cut out and digested with trypsin, followed by liquid chromatography–mass spectrometry (LC-MS) analysis performed in the Proteomics and Metabolomics Core at the Cleveland Clinic Lerner Research Institute. Proteins were identified if their detection probability was greater than 99.9% and at least 2 specific peptides for 1 protein could be identified.
3′-UTR luciferase reporter assays
To construct the CD38 reporter vector, a 25-bp segment of the CD38 mRNA 3′UTR that contained the predicted miR-26a binding site was cloned into pmirGLO vector after the firefly luciferase coding sequence. For the dual luciferase assay, the CD38 reporter was transiently transfected into HEK293T cells with miR-26a overexpression (pCDH-CMV-MCS-EF1-copGFP-miR26a) or knockdown vector (pmiRZIP-26a). The cells were lysed, and luciferase activity was detected 48 hours after transfection. Results are presented as mean ± SD of three independent experiments.
Multiple myeloma mouse xenograft model
A total of 5 × 106 RMPI8226-V-26a-GFP/V-GFP or MM.1S-V-26a-GFP/V-GFP cells in 100 μL PBS together with an equal volume of Matrigel basement membrane matrix (BD Biosciences) was subcutaneously injected into the shoulder of Fox Chase SCID beige mice (strain code: 250; Charles River) to establish a human multiple myeloma xenograft model. The two orthogonal diameters (x and y) of tumors were monitored once a week, and xenograft volume was calculated as V = xy2/2. Mice were sacrificed at 28 days postinjection. RNA was isolated from xenograft determine miR-26a level, protein was extracted for CD38 expression, and remaining tissue was paraffin-embedded for IHC analysis (Ki-67 and c-caspase-3).
Functionalization of purified single-wall carbon nanotubes
The 6,5-enriched single-wall carbon nanotubes [(6,5) SWCNT] were purified by a polymer aqueous two-phase separation method using the DNA sequence: T(TA)2T4(AT)2T (20). The (6,5) SWCNTs were mixed with 10 mg PL-PEG2000-NH2 (Avanti Polar Lipids, 880128P) in 5 mL double-distilled water in a glass scintillation vial. The vial was sonicated in a bath sonicator (97043-992, VWR) for 1 hour at room temperature with water changes every 20 minutes to avoid overheating. The SWCNT suspension was centrifuged at 24,000 × g for 6 hours at room temperature, and the supernatant was collected. The SWCNT supernatant, 1 mL, was washed five times, by adding 1 mL SWCNT supernatant to a 4 mL centrifugal filter (Amicon; Millipore Sigma, UFC910008) and 3 mL double-distilled water, and centrifuging for 10 minutes, 4,000 × g, room temperature. After the final wash, the SWCNT concentration was measured using a UV/VIS spectrometer (Thermo Fisher Scientific, accuSkan GO UV/Vis Microplate Spectrophotometer) with an extinction coefficient of 0.0465 L/mg/cm at 808 nm. The SWCNT concentration was adjusted to approximately 50 mg/L by adding the required amount of double distilled water.
Conjugation of miR-26a to SWCNTs through cleavable disulfide bond
Functionalized SWCNTs, 500 μL, were mixed with 0.5 mg of Sulfo-LC-SPDP (c1118; ProteoChem). Fifty microliters of 10× PBS was added and incubated for 2 hours at room temperature. After incubation, the SWCNT solution was washed five to six times using a centrifugal filter (Amicon) by adding 3 to 4 mL DNase/RNase-free water and centrifuging for 6 to 8 minutes at 10,000 × g each time. Fifteen microliters of miR-26a (100 μmol/L) was mixed with 1.5 μL DTT solution (Sigma, #43815), incubated for 1.5 hours at room temperature, and then DTT-treated miR-26a was purified using a NAP-5 column (GE Healthcare, 17-0853-01) following the manufacturer's protocol. Five hundred microliters of miR-26a was eluted and collected from the column with DNase/RNase free 1× PBS. The activated SWCNTs were suspended with the 500 μL purified miR-26a solution, and the conjugation was allowed to proceed for 24 hours at 4°C.
Delivery of SWCNT-miR-26a to disseminated multiple myeloma mouse model
A murine disseminated model of human multiple myeloma cells were established in 8-week-old female NOD.CB17-Prkdcscid/J mice (Charles River). All mice were irradiated and then intravenously injected with 5 × 106 MM.1S-Luc-GFP cells and were randomized to separate to control and treatment groups. Mice were subsequently injected with 100 μL (40 mg/mL) SWCNT-miR-26a or SWCNT-ctrl, or bortezomib (0.5 mg/kg) plus SWCNT-ctrl, or bortezomib (0.5 mg/kg) plus SWCNT-miR-26a once a week through the tail veins in a masked fashion, then observed daily and sacrificed once mice developed hind limb paralysis (endpoint). Images were acquired using an in vivo imaging system (IVIS; PerkinElmer). Hind limb paralysis was used as the end point in this disseminated disease model. All experiments involving animals were preapproved by the Cleveland Clinic Institutional Animal Care and Use Committee.
Statistical analysis was performed using SPSS (version 17.0). Comparisons between two independent groups were performed using a two-tailed Student t test. In our mouse model, time to our endpoint of hind limb paralysis was measured using the Kaplan–Meier method, with Cox proportional hazard regression analysis for group comparisons. P ≤ 0.05 was considered as statistically significant. Correlation analysis was performed using the Pearson correlation test; R2 > 0.3 was considered as positive. Isobologram analysis was performed using the CompuSyn software program (ComboSyn, Inc.). A combination index (CI) less than 1.0 indicates synergism, and a CI of 1 indicates additive activity (21, 22).
miR-26a inhibited cell proliferation and migration and induced apoptosis in multiple myeloma
Analysis of the GSE16558 dataset (60 patients with multiple myeloma and five healthy donors; ref. 17) revealed that miR-26a, but not miR-26b, expression was significantly downregulated in patients with multiple myeloma compared with healthy donors (Fig. 1A). We confirmed this result in CD138+ plasma cells from the healthy donors and multiple myeloma cell lines (Fig. 1B). As posttranscriptional regulators, miRNAs may inhibit protein expression without influencing mRNA level (23, 24). Thus, to identify the downstream targets of miR-26a in multiple myeloma, we performed SILAC combined with LC/MS instead of mRNA microarray analysis to uncover all proteins regulated by miR-26a. RPMI8226-V-miR-26a-GFP and RPMI8226-V-GFP cells were cultured in heavy or light medium separately using SILAC, followed by protein separation and MS-LC. A total of 2,724 unique proteins were recognized, of which, 180 were upregulated (68 proteins) or downregulated (112 proteins; Supplementary Table S1). Because miRNAs are negative regulators of gene expression, we further screened the 112 downregulated proteins using web-based query tools (TargetScan Release 7.1 and miRBase), and identified CD38 (H/L = 0.49; P = 0.02) as potential target of miR-26a (Fig. 1C).
To investigate the function of miR-26a in multiple myeloma, we overexpressed miR-26a in RPMI8226 (Fig. 2A), MM.1S (Fig. 2B), and H929 (Fig. 2C) cells by infecting them with V-miR-26a-GFP/V-GFP. The overexpression of miR-26a was confirmed by qRT-PCR. To determine the influence of miR-26a on cell growth, proliferation was assessed in these stable cell lines. Results showed that overexpressed miR-26a slowed proliferation significantly. To assess the effect of miR-26a on migration, a transwell migration assay was performed using the IncuCyte live cell analysis system, which showed that multiple myeloma migration was inhibited by overexpressed miR-26a in both RPMI8226 (Fig. 2A), MM.1S (Fig. 2B), and H929 (Fig. 2C) lines. To uncover the apoptotic status of cells with miR-26a overexpression, we performed Western blots for PARP1/c-PARP1 and caspase-3/c-caspase-3, two proteins that reflect apoptotic status, and found miR-26a induced greater expression of c-PARP1 and c-caspase-3 in RPMI8226, MM.1S, and H929 cells (Fig. 2A–C). Furthermore, a reporter assay demonstrated that miR-26a directly regulated expression of CD38 mRNA through binding to its 3′UTR (Fig. 2D). In addition, overexpressed miR-26a inhibited CD38 protein expression in RPMI8226, MM.1S, and H929 cells (Fig. 2D).
miR-26a directly inhibited CD38 translation in multiple myeloma cells
We did not find a direct association between levels of miR-26a and CD38 mRNA in our analysis of the GSE16558 dataset (Supplementary Fig. S1A; R2 = 0.126; P = 0.0037). However, our qRT-PCR results revealed that the CD38 mRNA level increased after miR-26a overexpression in RPMI8226, but decreased in MM.1S cells, which indicated that the inhibition of CD38 protein expression caused by miR-26a is independent of its mRNA level (Supplementary Fig. S1B).
To further verify whether miR-26a repressed multiple myeloma cell growth through CD38 inhibition, we knocked down CD38 expression by siRNA in RPMI8226, MM.1S, and H929 cells. Five days after CD38 siRNA transfection, apoptosis was assayed (Fig. 3A) and found to be increased significantly by about 3- to 5-fold in CD38-knockdown cells compared with cells transfected with control siRNA, indicating that CD38 knockdown directly increased apoptosis in multiple myeloma. Protein expression also decreased after CD38 knockdown (Fig. 3A). In addition, we restored CD38 expression by infecting miR-26a mimic-treated MM.1S cells with V-CD38; V-GFP was the control (Fig. 3B). Restoring CD38 expression diminished the apoptosis induced by miR-26a overexpression, indicating the suppression of cell growth and accumulation of apoptosis caused by miR-26a were mediated by CD38 inhibition.
miR-26a synergistically enhanced the effects of bortezomib or melphalan in multiple myeloma
Our results indicate that miR-26a inhibits multiple myeloma growth through directly inhibiting CD38. To investigate whether miR-26a improved the therapeutic effect of bortezomib or melphalan in multiple myeloma cells, RPMI8226, MM.1S, and H929 cells stably transduced with V-miR-26a-GFP and V-GFP were treated with various doses of bortezomib in DMSO or melphalan in DMSO. Cells were collected for apoptosis determination by flow cytometry and Western blot analysis, and viability was evaluated using a CellTiter-Glo Luminescent Kit. Exogenous expression of miR-26a markedly increased bortezomib- or melphalan-induced apoptosis as measured by the percentage of Annexin V-APC–positive cells (Fig. 4A; Supplementary Fig. S2A) and protein levels of c-PARP1 and c-caspase-3 (Fig. 4B; Supplementary Fig. S2B). As shown in Fig. 4C and Supplementary Fig. S2C, enforced expression of miR-26a strongly increased growth inhibition induced by bortezomib or melphalan at 24 hours after treatment, and the inhibition became more pronounced over time. These findings indicate that overexpression of miR-26a in multiple myeloma cells act synergistically improve the curative effect of bortezomib or melphalan (Fig. 4D; Supplementary Fig. S2D).
MiR-26a decelerated multiple myeloma growth in xenograft murine models
To validate our in vitro findings that a miR-26a mimic can inhibit multiple myeloma, we examined its ability to suppress tumor growth in an in vivo xenograft murine model. RPMI8226 or MM.1S cells infected with V-miR-26a-GFP/V-GFP were injected subcutaneously into SCID mice. Xenograft growth of both cell lines was strikingly inhibited in the V-miR-26a mice compared with V-GFP control mice (Fig. 5A and B). V-miR-26a-transduced xenografts weighed significantly less than controls when they were sacrificed on postinjection day 28. RNA was extracted for miR-26a expression, tissues were fixed, and embedded for IHC for Ki-67 and c-caspase-3, and protein was extracted to determine expression of CD38. miR-26a overexpression not only inhibited CD38 expression and reduced proliferation as indicated by Ki-67 staining, but also induced c-caspase-3 in vivo (Fig. 5A and B), as we had observed in vitro, indicating that miR-26a played a tumor suppressor role in vivo through inhibiting the function of CD38, suppressing proliferation, and provoking cell apoptosis in multiple myeloma.
SWCNT-miR-26a delivery prolongs the lifespan in a disseminated human multiple myeloma cell mouse model
Our data indicated that miR-26a was a robust tool to provoke cell apoptosis in multiple myeloma by targeting CD38. In our next step, we delivered miR-26a in vivo to mimic drug delivery in the clinical setting. We had used SWCNT to deliver antisense nucleic-acid drugs effectively and efficiently with good tolerability and minimal toxicity in vitro and in vivo (25). To test SWCNT delivery of miRNA, we conjugated SWCNTs with miR-26a (SWCNT-miR-26a; Fig. 6A; Supplementary Fig. S3; refs. 26, 27), then added the conjugated SWCNT to culture medium of H929, RPMI8226, and MM.1S cells to validate delivery efficiency. As shown in Fig. 6B, SWCNT-miR-26a was delivered into the multiple myeloma cells efficiently and led to substantial overexpression of the endogenous miR-26a in RPMI8226, MM.1S, and H929 cells.
To further estimate the treatment potential of SWCNT-miR-26a in vivo, we intravenously injected MM.1S-Luc-GFP cells through the tail vein of SCID mice to establish a human multiple myeloma cells dissemination murine model (Fig. 6A). At day 7 after tumor cell injection, SWCNT-miR-26a or SWCNT-ctrl oligos or bortezomib (0.5 mg/kg) plus SWCNT-ctrl, or SWCNT-miR-26a combined with bortezomib (0.5 mg/kg) were intravenously injected once a week. We observed tumor burden with IVIS after luciferin injection, and found the miR-26a level was dramatically high in tumor cells isolated from SWCNT-miR-26a-treated mice compared with the cells isolated from SWCNT-ctrl–treated mice (Fig. 6B). The luciferin signal was significantly lower in SWCNT-miR-26a-treated mice compared with the SWCNT-ctrl-treated group and even further lower when combined with bortezomib treatment (Fig. 6C). We found SWCNT-miR-26a treatment extended time to hind limb paralysis significantly compared with SWCNT-ctrl-treated group. We also found that SWCNT-miR-26a and bortezomib combination therapy, compared with SWCNT-miR-26a or bortezomib single treatment, dramatically prolongs the lifespan of the treated multiple myeloma mice (Fig. 6D).
miR-26a regulates CD38 expression during human B-cell development to plasma cells
To understand whether the control exerted by miR-26a on CD38 expression is involved in normal B-cell differentiation toward plasma cells, naïve B cells were isolated from peripheral blood obtained from healthy donors. The purified naïve B cells were differentiated to plasma cells in vitro, and then subjected to qRT-PCR to examine miR-26a expression. As shown in Fig. 7A and B, naïve B cells were successfully isolated and differentiated to CD38-positive and CD138-positive plasma cells. The miR-26a level was markedly reduced in the differentiated plasma cells (Fig. 7C). We also determined the CD38 and miR-26a levels in serial histological sections of normal human tonsil tissues, finding the miR-26a level was low in germinal center B cells and plasma cells whereas the CD38 signal was extremely high (Fig. 7D).
PR domain zinc finger protein 1, also known as BLIMP1, is a key transcription factor that controls the differentiation of activated B cells into plasma cells by silencing several important B cells genes, including Pax5 (28), Myc (29), and Bcl6 (30). We examined BLIMP1 expression in day-10 differentiated plasma cells and found the mRNA level of BLIMP1 was increased compared with naïve B cells (Fig. 7E). To investigate whether BLIMP1controls the expression of miR-26a, we knocked down the BLIMP1 level with siRNA in day 3 activated B cells. After another 7 days’ differentiation, the cells were subjected to flow cytometry, qRT-PCR, and immunoblotting. As shown in Fig. 7F, BLIMP1 mRNA and protein levels were both reduced in the siRNA-treated cells compared with control siRNA-treated cells. The miR-26a level was significantly increased in BLIMP1 knockdown cells (Fig. 7G), indicating that BLIMP1 negatively regulated miR-26a expression. At the same time, the number of CD38-positive cells was significantly reduced in BLIMP1 knockdown groups (Fig. 7H), confirming that BLIMP1 potentially regulates miR-26a and CD38 during B-cell differentiation. To understand whether miR-26a was functionally associated with plasma cell differentiation in human B cells, we ectopically overexpressed miR-26a in B cells that had differentiated into plasma cells. Surprisingly, we found that even though CD38 expression was inhibited in miR-26a-transfected differentiated plasma cells, the total CD138+ plasma cell number was not reduced (Fig. 7I).
miR-26a has been reported recently to be downregulated in patients with CLL (31) and can predict prognosis after adjustment for covariate confounders (e.g., FISH analyses, IGHV mutational status, and ZAP-70 or CD38 expression; ref. 32). Whether miR-26a level and CD38 expression correlate is unknown. The topographic distribution of CD38 expression in CLL, which is often characterized by the formation of CD38-high hot spots within a CD38 negative milieu offer an optimal ground to investigate topographic distribution of the miR-26a expression. We first analyzed the GSE51529 microarray dataset (210 patients with CLL) reported by Maura and colleagues (33), and found a significant inverse correlation between miR-26a level and CD38 mRNA expression level (P = 0.0004; Supplementary Fig. S4A). In situ hybridization and IHC studies for miR-26a and CD38 in consecutive histologic sections from 25 patients with CLL lymph node revealed five cases with CD38 patchy positive staining and two cases with strong CD38+ cell clusters. miR-26a level and CD38 expression inversely correlated in some CD38+ cells spots on serial histologic sections (Supplementary Fig. S4B). The data also supported the rationale of previous miR-26a delivery studies using CD38 conjugated nanoparticles in CLL (34).
Daratumumab-resistant patient multiple myeloma cells express high level of CD38 and low level of miR-26a
In patients, surface CD38 level was decreased 90% in nondepleted myeloma cells after the first daratumumab infusion by selective killing, CD38 microvesicle release, and internalization of CD38-daratumumab (35). CD38 returned to basal level after about 6 months treatment, indicating the development of daratumumab resistance (35). To further study CD38 and miR-26a expression after disease becomes daratumumab resistant, we examined all the patients with multiple myeloma treated and followed up at the Cleveland Clinic Tausig Cancer Center from 2015 to2019 and who donated tissue to our multiple myeloma tissue bank. Of the 20 patients identified, two have experienced relapse and bone marrow from one of these patients was biopsied successfully. IHC revealed that the multiple myeloma cell surface CD38 level was even higher in the relapsed bone marrow biopsy sample compared with pre-daratumumab treatment. At the same time, miR-26a levels were undetectable in both pre- and post-daratumumab treatment multiple myeloma cells (Fig. 7J). Consistent with previous studies (35), greater CD38 expression that daratumumab no longer killed these multiple myeloma cells, even when combined with other drug treatment including dexamethasone, pomalidomide, ixazomib, cafizomib, and panobinostat.
Multiple myeloma is a refractory plasma cell malignancy characterized by accumulation of antibody-secreting malignant plasma cells and osteolytic lesions in the bone marrow. The genetic heterogeneity of multiple myeloma presents obstacles to treatment. New treatments, such as proteasome inhibitors and immunomodulatory agents, have increased patient survival to a limited extent, from 3.5 to 5.5 years (36–38), and patients with relapse have few treatment choices. Multiple myeloma is a disease with complex and variable genomic changes. Consequently, effective treatments thus far developed, including bortezomib, lenalidomide, and dexamethasone, are multiple target drugs. Therefore, we decided to focus on miRNAs, which are natural small RNAs with multiple targets. By analyzing established datasets, we found that miR-26a was markedly downregulated in multiple myeloma cells. miR-26a has been verified in patient tissues as a tumor suppressor in multiple cancers, including osteosarcoma (39), lymphoma (40), colorectal cancer, (41), and others. Given our database findings, we believed miR-26a also plays a role in multiple myeloma.
A type II transmembrane glycoprotein, CD38 is uniformly overexpressed on multiple myeloma cells and regulates signal transduction of specific cell surface proteins (42). The CD38-specific mAb, daratumumab (43), is FDA approved for refractory multiple myeloma and has produced impressive therapeutic effects in clinic. Daratumumab induces multiple myeloma cell death mainly via antibody-dependent cell-mediated cytotoxicity, complement-dependent cytotoxicity, and antibody-dependent cellular phagocytosis, but it also can induce apoptosis through acting as an ectoenzyme (44–46). We showed that knocking down CD38 induces apoptosis in multiple myeloma cells; furthermore, restoring CD38 expression in multiple myeloma cells overexpressing miR-26a suppressed apoptosis significantly. These results indicate that insufficient CD38 expression will cause multiple myeloma cell death without the need for Fc-mediated immunoreactions. Some studies have reported that in vitro, antibody-dependent cell-mediated and complement-dependent cytotoxicity of IgG1-based mAbs may translate into only limited clinical anti-multiple myeloma activity when used alone because patients with multiple myeloma are usually immunocompromised, especially when their disease is resistant to current treatments (47, 48). However, another CD38 mAb, isatuximab, has showed shown efficacy and tolerability as a mono- and combination therapy in phase I and phase II studies in patients with relapse or refractory multiple myeloma; it is in phase III clinical trials now. Isatuximab has strong proapoptotic activity via inhibiting CD38 ectoenzyme function, which is independent of CD38 inter-crosslinking (49–51).
miRNA regulates protein expression at the posttranscriptional level and so can induce cleavage of mRNA or inhibit translation without reducing mRNA level (23, 24). For this reason, we used SILAC to screen for downstream target screening. We are the first to report that miR-26a inhibits multiple myeloma cell proliferation and migration and promotes cell apoptosis by targeting CD38 in multiple myeloma. As we were preparing this manuscript, a report was published in Cancer Research that uncovered a novel CD38 function in which CD38 drives mitochondrial trafficking to promote bioenergetic plasticity in multiple myeloma (52). Consistent with what these investigators found, we found that siRNA mediated knockdown of CD38 improved animal survival. Thus, we believe miR-26a mimic oligonucleotide treatment will have more consistent effects in patients with multiple myeloma.
We also showed that miR-26a enhanced bortezomib and melphalan treatment effects in vitro and in vivo. The FDA has approved several antisense oligonucleotide drugs, including nusinersen for spinal muscular atrophy (53), mipomersen for homozygous familial hypercholesterolemia (54), fomivirsen for cytomegalovirus retinitis (55), and eteplirsen for Duchenne muscular dystrophy (56). As a natural oligonucleotide, a miR-26a mimic will have advantages because it would not induce an immune response or deposit in organs due to inappropriate metabolism. In our mouse xenograft experiments, miR-26a inhibited growth of RPMI8226 and MM.1S xenografts as compared with V-GFP infected xenografts. In addition, CD38 protein level was inhibited significantly in miR-26a-transduced xenografts, and Ki-67 was suppressed, whereas c-caspase-3 was intensely induced, all of which are consistent with our in vitro results. These results indicated that the miR-26a treatment alone effectively inhibits growth of multiple myeloma cells with (RPMI8226) or without (MM.1S and H929) p53 mutation (57). Thus, miR-26a is potentially an effective in multiple myeloma as a single treatment or in combination with standard chemotherapeutics.
All-trans retinoic acid (58), the histone deacetylase inhibitor panobinostat (59), and recently DNA methyltransferase inhibitors (60) are being used as a combination therapy to overexpression CD38 to increase the myeloma cell sensitivity to daratumumab. However, this strategy may not work after once daratumumab resistance develops because CD38 expression level is high in myeloma cells from patients with daratumumab resistance. Loss of cell surface CD38 expression on multiple myeloma cells during long-term daratumumab treatment does not underlie the resistance mechanism that develops. Although CD38 expression remained low during daratumumab treatment, the response was high as others have reported (35). A strategy that uses miR-26a to multiple myeloma cells under pressure to maintain CD38 expression at a certain level may resensitize resistant cells to daratumumab treatment and have clinical benefit.
Disclosure of Potential Conflicts of Interest
E.D. Hsi is a consultant at Seattle Genetics, Miltenyi and reports receiving other commercial research support from Abbvie and Cellerant. K.C. Anderson is a consultant at Janssen and Sanofi Aventis. N.C. Munshi is a consultant at Janssen, Amgen, Abbvie, Takeda, Karyopharm, Adaptive, and Oncopep, reports receiving other commercial research support from BMS, and has ownership interest (including patents) in Oncopep. No potential conflicts of interest were disclosed by the other authors.
Conception and design: Y. Hu, H. Liu, J.P. Maciejewski, J.N. Valent, J. Lin, J. Zhao
Development of methodology: Y. Hu, C. Fang, J. Lin, J. Zhao
Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): Y. Hu, H. Liu, C. Fang, H. Dysert, J. Bodo, G. Habermehl, B.E. Russell, W. Li, M. Chappell, S.L. Ondrejka, E.D. His, G. Ao, J. Lin
Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): Y. Hu, H. Liu, C. Li, W. Li, K.C. Anderson, N.C. Munshi, G. Ao, J. Lin, J. Zhao
Writing, review, and/or revision of the manuscript: Y. Hu, H. Liu, E.D. Hsi, K.C. Anderson, N.C. Munshi, G. Ao, J.P. Maciejewski, J.N. Valent, J. Lin, J. Zhao
Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): C. Fang, H. Dysert, J. Bodo, S.L. Ondrejka, Q. Yi, J. Zhao
Study supervision: G. Ao, J. Zhao
The authors thank the Lerner Research Institute proteomic, genomic, and imaging cores for their assistance and support and Dr. Cassandra Talerico provided editorial assistance and helpful comments. This work was financially supported by grant from: NCI R00 CA172292 (to J. Zhao) and start-up funds (to J. Zhao) and two Core Utilization Pilot Grants (to J. Zhao) from the Clinical and Translational Science Collaborative of Cleveland, V Foundation Scholar Award (to J. Zhao), NIH training grant T32 CA094186, Training in Computational Genomic Epidemiology of Cancer (CoGEC) career development program (to J. Lin), 5UL1TR002548 from the National Center for Advancing Translational Sciences (NCATS) component of the NIH and NIH roadmap for Medical Research. The Orbitrap Elite instrument was purchased via an NIH shared instrument grant, 1S10RR031537-01.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.