miR‐17‐92 cluster‐BTG2 axis regulates B‐cell receptor signaling in mantle cell lymphoma

Abstract

B‐cell receptor (BCR) signaling is critically activated and stable for mantle cell lymphoma (MCL), but the underlying mechanism of the activated BCR signaling pathway is not clear. The pathogenic basis of miR‐17‐92 cluster remains unclear although the oncogenic microRNA (miRNA) miR‐17‐92 cluster is highly expressed in patients with MCL. We revealed that miR‐17‐92 cluster overexpression is partly dependent on SOX11 expression and chromatin acetylation of MIR17HG enhancer regions. Moreover, miR‐17‐92 cluster regulates not only cell proliferation but BCR signaling activation in MCL cell lines. To comprehensively identify miR‐17‐92 cluster target genes, we performed pulldown‐seq, where target RNA of miRNA was captured using the biotinylated miRNA mimics and magnetic bead‐coated streptavidin, and quantified using next‐generation sequencing. The pulldown‐seq identified novel miRNA target genes, including tumor suppressors such as BTG2 (miR‐19b), CDKN2A (miR‐17), SYNE1 (miR‐20a), TET2 (miR‐18, miR‐19b, and miR‐92a), TNFRSF10A (miR‐92a), and TRAF3 (miR‐17). Notably, the gene expression profile data of patients with MCL revealed that BTG2 expression was negatively associated with that of BCR signature genes, and low BTG2 expression was associated with poor overall survival. Moreover, BTG2 silencing in MCL cell lines significantly induced BCR signaling overactivation and cell proliferation. Our results suggest an oncogenic role of miR‐17‐92 cluster‐activating BCR signaling throughout BTG2 deregulation in MCL. Furthermore, this may contribute to the prediction of the therapeutic efficacy and improved outcomes of MCL.

SOX11 and enhancer‐regulated mechanism induce overexpression of miR‐17‐92 in mantle cell lymphoma (MCL). BTG2 was identified as a novel target of miR‐17‐92 cluster. miR‐17‐92 cluster activates B‐cell receptor signaling and cell proliferation by regulating BTG2 in MCL.

Abbreviations

BCL

B‐cell lymphoma

BCR

B‐cell receptor

BTK

Bruton's kinase

ChIP

chromatin immunoprecipitation

DLBCL

diffuse large B‐cell lymphoma

FL

follicular lymphoma

GC

germinal center

MCL

mantle cell lymphoma

miRNA

microRNA

PRMT1

protein arginine methyltransferase‐1

shRNA

short hairpin RNA

1. INTRODUCTION

Mantle cell lymphoma (MCL) is a mature B‐cell lymphoma (BCL) that accounts for 3%–6% of all malignant lymphomas. ¹ The advent of rituximab‐containing immunochemotherapy and molecular‐targeted therapeutics has improved the MCL outcome. Notably, targeting therapy for Bruton's kinase (BTK), which is a component of the B‐cell receptor (BCR) signaling pathway, such as ibrutinib, acalabrutinib, and zanubrutinib, is highly effective for MCL. However, MCL remained incurable despite the recent development of novel therapy. Moreover, the prognosis is generally very poor when MCL progresses to be refractory to target therapy, such as BTK inhibitors. ² , ³

The mechanisms of BCR activation in MCLs are not sufficiently understood although BCR signal activation is important for MCL pathogenesis. ³ , ⁴ , ⁵ , ⁶ Genomic abnormalities in various genes that compose the BCR signaling pathway, such as CD79A/CD79B, TNFAIP3, CARD11, and MYD88, are considered drivers of BCR signal activation in other types of BCL, such as diffuse large B‐cell lymphoma (DLBCL). ⁷ , ⁸ , ⁹ , ¹⁰ , ¹¹ In contrast, these genetic abnormalities are uncommon in MCL. Some MCLs show reactivity to autoantigens, suggesting chronic active BCR signaling activation. ¹² , ¹³ SOX11 overexpression and MCL tumor microenvironment have induced BCR activation. ¹² However, the detailed mechanisms of BCR activation remained unclear. Hence, knowledge of the pathogenic basis of BCR signal activation is urgently needed to overcome the resistance to BCR signal‐targeted therapies.

MicroRNAs (miRNAs) are small 20–24 nucleotide noncoding RNAs that induce mRNA degradation and suppress translation. ¹⁴ The miR‐17‐92 cluster consists of six miRNAs, miR‐17, 18a, 19a, 20a, 19b‐1, and 92a‐1, located on chromosome13q31.3, with two paralogs, miR‐106a‐363 and miR‐106b‐25. ¹⁵ , ¹⁶ These miRNAs play oncogenic roles in various cancers. ¹⁷ Additionally, the miR‐17‐92 cluster is constitutively overexpressed in MCLs, ¹⁸ and miR‐18b overexpression is associated with poor prognosis. ¹⁹ Importantly, the expression of miR‐17‐92 cluster is lower in naive B cells, which are considered as the origin of MCL, than in germinal center (GC) B cells and post‐GC activated B cells. ²⁰ We therefore hypothesized that the miR‐17‐92 cluster plays critical roles in tumor‐specific pathogenesis, which is different from the normal counterpart of MCL. However, the function and target genes of miR‐17‐92 cluster in MCL are not well known.

This study explored the fundamental roles of the miR‐17‐92 cluster through comprehensive analysis to identify the targets of these miRNAs in vitro and revealed that BTG2, which is a novel target of the miR‐17‐92 cluster, regulates proliferation and BCR signaling activation in MCL.

2. MATERIALS AND METHODS

2.1. Cell lines and reagents

This study used four human MCL cell lines, Jeko‐1, JVM2, and Z138 obtained from American Type Culture Collection (ATCC) and KPUM‐YY1 established from a patient with MCL in our laboratory, ²¹ and the human embryonic kidney cell‐derived cell line HEK293T from ATCC. MCL cells were maintained in RPMI‐1640 (Wako) containing 10% fetal calf serum (Life Technologies), 2 mM L‐glutamate, and penicillin/streptomycin at 37°C in humidified 95% air and 5% CO₂. In addition, 2% fetal calf serum medium was used in some BTG2 silencing experiments. HEK 293T cells were maintained in Dulbecco's modified Eagle's medium (Nacalai Tesque, Inc.) containing 10% fetal calf serum, 2 mM L‐glutamate, and penicillin/streptomycin. Normal B lymphocytes were collected from the blood of two healthy donors to analyze the expression level of the miR‐17‐92 cluster. CD19‐positive B lymphocytes were isolated using Lymphocyte Separation Solution (Nacalai Tesque, Inc.) and MACSprep™ Chimerism CD19 MicroBeads, human (Miltenyi Biotec).

2.2. Transfection of miRNA mimics and inhibitors

MCL cells were transfected with 1 μM miRCURY LNA™ Premium miRNA Mimic, miRCURY LNA™ miRNA Mimic Negative Control (Qiagen), 6 μM mirVana™ miRNA inhibitor or mirVana™ miRNA inhibitor negative control #1 (Life Technologies) using Hemagglutinating Virus of Japan (HVJ)‐envelope vector (Ishihara Sangyo Kaisha, Ltd.), following the manufacturers’ protocol.

2.3. Pulldown‐seq

The protocol was modified by previous reports. ²² , ²³ , ²⁴ The biotinylated miRNA mimic or negative control was introduced into Z138 cells, and target mRNA was captured. The complex of captured target mRNA and miRNA mimic was pulldown with magnetic bead‐coated streptavidin. The extracted RNA was sequenced and analyzed by next‐generation sequencing (NGS) at the NGS Core Facility of the Kyoto Prefectural University of Medicine. Details are described in Data S1.

2.4. Chromatin immunoprecipitation assay

Chromatin immunoprecipitation (ChIP) was performed using the SimpleChIP Enzymatic Chromatin IP Kit (Cell Signaling Technology) according to the manufacturer's protocol as previously described. ²⁵ For immunoprecipitation, antibodies specific for SOX11 (HPA000536; Sigma‐Aldrich), BRD4 (E2A7X; Cell Signaling Technology), and acetylated H3 (ab4279; Abcam) were diluted to 2 μg/mL. Normal rabbit IgG supplied by SimpleChIP kit was used as a background control. Extracted DNA was analyzed by quantitative PCR using Fast SYBR Green Master Mix (Thermo Fisher Scientific). The following primers were used: for SOX11 ChIP‐qPCR, MIR17HG_promoter‐forward (F): 5′‐GGG AGG TCG GAA GTA CTT TGT‐3′ and MIR17HG_promoter‐reverse (R): 5′‐GCT CCC GCC TCA ACG TAA AT‐3′; for BRD4 and H3K27AC ChIP‐qPCR, MIR17HG_E1‐F: 5′‐AAA TGC AGC TGG GCA TGA GA‐3′, MIR17HG_E1‐R: 5′‐ TTA ACC TGG CCG CGT GTA AA ‐3′, MIR17HG_E2‐forward (F): 5′‐TAA TGA GGG AGT GGG GCT TGT‐3′, MIR17HG_E2‐R: 5′‐CCT CGA AGG ACC ATG TGG GT‐3′, MIR17HG_E3‐F: 5′‐TGC CCG GTC TTC TGT TCC TA‐3′, MIR17HG_E3‐R: 5′‐TCC TGA TGG CAT GCC GTT AA‐3′.

2.5. Ca²⁺ flux assay

The strength of the BCR signaling was quantified by measuring intracellular Ca²⁺ concentration. Cells were incubated with 3 μM Fluo‐4AM (Dojindo Molecular Technologies) in Hanks' balanced salt solution (HBSS) with CaCl₂ at 37°C for 45 min. Cells were subjected to flow cytometric analysis using a FACS Celesta (BD Biosciences) after washing the cells with HBSS without CaCl₂. The measuring tube was removed 30 s after the start of the measurement, 10 μg/mL of anti‐IgM (Jackson ImmunoReserch Laboratories Inc.) was administered at 35 s, and measurements were restarted from 40 to 340 s. FlowJo™ software (BD Biosciences) was used to analyze data. The signal intensity was normalized with the mean of those between 0 and 30 s, and the area under the curve between 42.5 and 242.5 s was calculated.

2.6. Gene silencing using lentiviral infection

SOX11 and BTG2 silencing was accomplished by stable expression of inhibitory short hairpin RNAs (shRNAs). The target sequences for human SOX11 and BTG2 were 5′‐GCC TCT ACT ACA GCT TCA AGA‐3′ (shSOX11#1), 5′‐GCT GTT ATC TTA GTT TAA AGA‐3′ (shSOX11#2), 5′‐GCT GCA TTC GCA TCA ACC ACA‐3′ (shBTG2#1), and 5′‐GCA TTC GCA TCA ACC ACA AGA‐3′ (shBTG2#2). The sequences encoding shRNAs against SOX11 and BTG2 were cloned into lentivirus vector pLKO.1 puro (Addgene). The empty vector was used as negative control (mock). Lentivirus was produced by transient co‐transfection of HEK293T cells with a lentivirus vector, packaging plasmid pMDLg/pRRE, pRSV‐Rev, and pMD2.G using PEI MAX (Polysciences). Viral supernatant was collected at 48 and 72 h after transfection and was concentrated by ultracentrifugation. Infected MCL cells were selected with appropriate concentrations of puromycin from 48 h after infection and maintained under puromycin selection for 2 weeks in BTG2 silencing.

2.7. Microarray data analysis

Publicly available and deposited gene expression data in the National Center for Biotechnology Information Gene Expression Omnibus (NCBI GEO; accession number GSE93291) were analyzed to investigate the relationship between BTG2 gene expression level and MCL clinical characteristics. The expression data were normalized with the robust multichip array method. When there were more than one probes fot the same gene, a signle probe with the highest mean signal value was chosen.

The overall survival was compared within two subgroups defined by the BTG2 signal with the cut‐off of the median, using the log‐rank method. BCR signature gene expression was evaluated using the gene set from Herishanu et al. (CASP3, CCL4, CD83, CTLA4, DDIT3, DDX21, DUSP2, EGR1, EGR3, GFI1, GNPDA1, IRF4, KFKB1E, KLF10, LILRA4, LPL, LY9, NABP2, NAMPT, NHP2, NR4A3, OAS3, PAICS, PDGFA, RGS10, SLAMF7, TXNRD1). ²⁶

2.8. Luciferase reporter assay

BTG2 3′UTR sequences containing miR‐17/20a, miR‐18a, miR‐19a/19b, and miR‐92a biding regions were subcloned into pmirGLO Dual‐Luciferase miRNA target expression vector (Promega) digested by SacI and XbaI. The oligonucleotide sequences subcloned into the vector are shown in Figure S1B. HEK293T cells at 70% confluency of a 96‐well plate were transfected with 0.1 μg of plasmid DNA using Lipofectamine 3000 (Thermo Fisher Scientific) and cultured for 36 h. Firefly and Renilla luciferase activities were measured using the Dual‐Luciferase Reporter Assay kit (Promega) and GloMax Discover System plate reader (Promega). To normalize, ratios of firefly luciferase activity to Renilla luciferase activity were calculated. Measurements were performed with three technical replicates.

2.9. Statistical analysis

Values are reported as the mean and standard error of mean for normally distributed data. Student's t‐test was used to compare normally distributed continuous variables between groups. Statistical analyses were performed with R (version 4.0.5). A P value of <0.05 was considered significant.

Other methods are described in Data S1.

3. RESULTS

3.1. Mechanism of miR‐17‐92 cluster overexpression in MCL

First, each miR‐17‐92 cluster miRNA expression was confirmed to be significantly higher in the cell lines than in normal B lymphocytes using qRT‐PCR (Figure S2). Next, we hypothesized that SOX11 regulates miR‐17‐92 transcription because previous data of ChIP‐Seq of SOX11 in MCL cell lines showed that SOX11 binds upstream of the MIR17HG locus, ²⁷ and expression levels of miR‐17 and SOX11 were correlated in MCL patients. ²⁸ To validate this ChIP‐Seq data, we performed ChIP‐qPCR of SOX11 and revealed that the relative enrichment of SOX11 to normal IgG control at the MIR17HG promoter region was significantly higher in three SOX11‐positive cell lines (Jeko‐1, KPUM‐YY1, and Z138) than in SOX11‐negative JVM2 cells (Figure 1A). Silencing SOX11 repressed not only cell proliferation but the expression of each miR‐17‐92 in three SOX11‐positive MCL cell lines (Figure 1B–D). Moreover, we inferred other mechanisms of miR‐17‐92 overexpression because miR‐17‐92 is overexpressed in SOX11‐negative MCLs. To explore the SOX11‐independent and tumor‐specific background of miR‐17‐92 expression, we investigated the enhancer‐regulated mechanism using BRD4 ChIP‐seq data of JVM2 cells, which we had previously performed (Figure 1E). ²⁵ High BRD4 binding suggesting histone acetylation was found in the MIR17HG enhancer region in JVM2 cells. ChIP‐qPCR revealed that both BRD4 and H3K27Ac were highly enriched at the MIR17HG enhancer regions in MCL cell lines regardless of SOX11 expression (Figure 1F).

FIGURE 1.

Mechanism of miR‐17‐92 cluster overexpression in mantle cell lymphoma (MCL). (A) The results of SOX11 chromatin immunoprecipitation (ChIP)‐qPCR for SOX11‐negative (JVM2) and SOX11‐positive (Jeko‐1, KPUM‐YY1 and Z138) cells. Relative SOX11 enrichment at the MIR17HG promoter region compared with normal rabbit IgG is shown. P values are calculated by t‐test comparing SOX11‐negative and positive cells. (B) Expression of SOX11 in SOX11 short hairpin RNA‐transduced cells. Jeko‐1, KPUM‐YY1, and Z138 cells were transduced with mock (Ctrl), shSOX11#1 (sh#1), or shSOX11#2 (sh#2). Relative SOX11 transcriptional levels compared with those of mock are shown. (C) Cell proliferation assay in SOX11‐knocked down MCL cells. The x axis indicates the time after transfection and the y axis shows the viable cell count. Results are shown as mean ± standard error of the mean of three independent experiments. (D) Relative miR‐17‐92 levels in SOX11 knocked‐down MCL cells compared with those of mock. (E) Gene tracks of BRD4 signals of JVM2 at the MIR17HG locus. The vertical axis represents reads per million mapped reads. (F) Relative enrichment of BRD4 and H3K27Ac at three MIR17HG enhancer regions (E1, E2, and E3 represented on the gene track in Figure 1E) analyzed by ChIP‐qPCR. P values are calculated by t‐test. *P < 0.05, ***P < 0.001.

3.2. Functional role of miR‐17‐92 cluster in MCL pathogenesis

We performed a proliferation assay in MCL cell lines transfected with miRNA inhibitors, which are chemically modified, single‐stranded oligonucleotides designed to specifically bind to endogenous miRNAs, to confirm the oncogenic role of these miRNAs in MCL. The cell proliferation of inhibitor‐transfected cell lines was generally repressed compared to the negative control group in all MCL cell lines examined (Figure S3). The underlying mechanism affecting cell proliferation is not well revealed despite the miR‐17‐92 cluster being known as oncogenic miRNA. Then, we hypothesized that the miR‐17‐92 cluster might be associated with BCR signal pathway upregulation because both the miR‐17‐92 cluster and the BCR signal are aberrantly activated in MCL cells compared to the normal counterpart of naive B cells. ⁴ , ²⁰ We evaluated intracellular Ca²⁺ influx from ER stores in MCL cell lines crosslinked and stimulated with anti‐IgM BCR to assess the impact of miR‐17‐92 cluster on the BCR signal (Figure 2A). The calcium response on BCR stimulation was reduced in cells transfected with all miRNA inhibitors except for miR‐17 (Figure 2B). Western blot (WB) analyses showed the miR‐17‐92 inhibitors decreased the activity of Syk, BTK, and AKT (Figure 2C). These results suggest that the miR‐17‐92 cluster leads to cell growth promotion partly depending on BCR signal activation.

FIGURE 2.

Ca²⁺ flux assay in MCL after the introduction of miR‐17‐92 inhibitors. (A) The representative data of Ca²⁺ flux assay of Jeko‐1, JVM2, KPUM‐YY1, and Z138 cells. The assay was performed 24 h after transfection. (B) The relative area under the curve of each inhibitor compared to that of the negative control. Results are shown as mean ± standard error of the mean of five independent experiments, colored by four cell lines. P values are calculated by t‐test. NS, not statistically significant, *P < 0.05, **P < 0.01, ***P < 0.001. miR‐17‐I, miR‐17 inhibitor; miR‐18a‐I, miR‐18a inhibitor; miR‐19a‐I, miR‐19a inhibitor; miR‐19b‐I, miR‐19b inhibitor; miR‐20a‐I, miR‐20a inhibitor; miR‐92a‐I, miR‐92a inhibitor. (C) Western blot of BCR signaling components, with and without anti‐IgM antibody stimulation, in Z138 cells after introduction of negative control and miR‐17‐92 inhibitor. The assay was performed 48 h after transfection. Proteins were extracted 2 min after treatment with an anti‐IgM antibody. The expression levels of phosphorylated protein/total protein were compared using the negative control as a reference, under each band as measured by densitometry analysis. β‐actin (ACTB) was used as an internal control.

3.3. Investigation of targeted genes of the miR‐17‐92 cluster

The landscape of miR‐17‐92 cluster targets, especially in MCL, is not described although a part of the targets of the miR‐17‐92 cluster consists of several genes, such as FGFR2B and CD22, affecting the BCR signal in other lymphomas such as DLBCL. ²⁹ Moreover, miRNA targets are different among tissues and cancer types, therefore this study aimed to identify the target genes of each matured miRNA of the miR‐17‐92 cluster in MCL cells but not by in silico simulation. We performed pulldown‐seq, where we captured mRNA using biotinylated miRNAs and analyzed them with NGS to comprehensively identify miR‐17‐92 cluster‐targeted genes (Figure 3A). We used the cell line Z138, which had the lowest expression of miR‐17‐92 cluster among the cell lines used in this study, to minimize the impact of overexpression on endogenous miRNAs (Figure S1). The level of each miRNA was checked by RT‐qPCR against a negative control to determine whether the pulldown was successful (Figure S4). Quantitative analysis of the captured mRNA level comparing those of each miRNA mimic and negative control identified candidate targets, some of which are known as recurrently mutated genes in MCL, BCR signal‐related genes, and tumor suppressors, was as follows: BTG2 (miR‐19b), CDKN2A (miR‐17), ³⁰ SYNE1 (miR‐20a), ³⁰ TET2 (miR‐18, miR‐19b, and miR‐92a), ³¹ TNFRSF10A (miR‐92a), ³² and TRAF3 (miR‐17) ³³ (Figure 3B and Table S1).

FIGURE 3.

Summary of pulldown‐seq. (A) Overview of pulldown‐seq. The biotinylated miRNA mimic or negative control was introduced into the cells, and target mRNA was captured. The complex of captured target mRNA and miRNA mimic was pulldown with magnetic bead‐coated streptavidin. Extracted RNAs were analyzed by next‐generation sequencing. (B) MA plots of pulldown‐seq for miR‐17, miR‐18a, miR‐19a, miR‐19b, miR‐20a, and miR‐92a. The x axis indicates the mean amount of normalized miRNA levels and the y axis indicates relative miRNA levels compared with the negative control‐transfected samples. Significantly pulled‐down genes are colored pink (P value <0.05), and more stringently genes are colored red (adjusted P value <0.1).

3.4. Functional analysis of BTG2 in MCL

We adopted publicly available gene expression profile data derived from patients with MCL who received R‐CHOP therapy (GSE93291) ³⁴ to investigate the clinical significance of those candidate target genes in patients with MCL. Data for these candidate genes revealed that the expression of BTG2 was negatively correlated with that of BCR signature genes ²⁶ (P = 1.15e‐04, R = 0.342; Figure 4A). Moreover, low BTG2 expression was significantly associated with poor overall survival (P = 0.006; Figure 4B), therefore we inferred the important role of BTG2 in BCR signaling and MCL pathogenesis.

FIGURE 4.

Effects of BTG2 expression on B‐cell receptor (BCR) signaling and survival in mantle cell lymphoma (MCL) patients. We analyzed the deposited gene expression profiling data in Gene Expression Omnibus from patients with MCL (GSE93291). (A) Relationship between the mean z‐scores of BCR signature genes and those of BTG2 expression. (B) Overall survival analysis was performed according to BTG2 expression levels with a cut‐off of the median BTG2 expression value. P value was calculated by log‐rank analysis.

We evaluated the BTG2 expression at the protein level to validate the data of pulldown‐seq showing that BTG2 was a candidate for the target of the miR‐17‐92 cluster and confirmed BTG2 upregulation after the introduction of miR‐17‐92 cluster inhibitors in all four MCL cell lines examined (Figure 5). Increased protein level expression of BTG2 was observed in miR‐17 and miR‐18a, which could not be identified by pulldown‐seq, in common with the four cell lines. Moreover, we performed the luciferase assay for four predicted miR‐17‐92 binding sequences and paired mutated mismatch sequences (Figure S1A,B). Under endogenous miR‐17‐92 expression of HEK293T cells, the luciferase activities were significantly repressed in the miR‐17/20a, miR‐18a, and miR‐92a reporters and modestly reduced in the miR‐19a/19b reporter (Figure S1C). We knocked down BTG2 by RNA interference in four MCL line cells to explore the functional role of BTG2 (Figures 6A,B and S5). Under normal culture conditions, there was no difference between mock and knockdown cells in the three cell groups examined (Figure S6A,C). However, 2% medium cell proliferation was increased in the knockdown group because BTG2 is a known stress‐induced gene ³⁵ (Figure S6). KPUM‐YY1 was excluded because it could not tolerate reduced FBS medium. Moreover, the Ca²⁺ flux assay revealed that BTG2 knockdown enhanced BCR signaling (Figures 6C,D and S5C–E). The molecules that constitute the BCR signaling involved in BTG2 knockdown were examined in WB. Additionally, the involvement of the Syk, BTK, and NF‐κB pathways was considered. Conversely, Lyn showed little change between mock and knockdown groups, suggesting that it was unregulated by BTG2 (Figure 6E).

FIGURE 5.

BTG2 protein level after transfection of miR‐17‐92 inhibitors. (A) The representative data of western blot (WB) analysis for four cell lines: Jeko‐1, JVM2, KPUM‐YY1, and Z138. Protein was collected 24 h after miR‐17‐92 inhibitor transfection. Relative expression levels to negative control by densitometry are shown under each band. ACTB was used as an internal control. (B) The results of five independent WB experiences for four cell lines. BTG2 expression data relative to the negative control for each inhibitor were integrated and analyzed for four cells. Results are shown as mean ± standard error of the mean, colored by four cell lines. P values are calculated by t‐test. *P < 0.05, **P < 0.01, ***P < 0.001. miR‐17‐I, miR‐17 inhibitor; miR‐18a‐I, miR‐18a inhibitor; miR‐19a‐I, miR‐19a inhibitor; miR‐19b‐I, miR‐19b inhibitor; miR‐20a‐I, miR‐20a inhibitor; miR‐92a‐I, miR‐92a inhibitor.

FIGURE 6.

Effects of BTG2 silencing on B‐cell receptor (BCR) signaling. Z138 cells were transduced with mock (Ctrl), shBTG2#1 (sh#1), or shBTG2#2 (sh#2). (A, B) Expression of BTG2 in SOX11 short hairpin RNA (shRNA)‐transduced cells. (A) Relative BTG2 expression compared with those of mock by transcriptional level (A) and protein level (B) are shown. Relative expression levels compared with those of negative control by densitometry are shown under each band. (C, D) The representative data of Ca²⁺ flux assay (C) and relative area under the curves (AUCs) of sh#1 and sh#2 to mock (D). Relative AUCs are shown as mean ± standard error of the mean of three independent experiments. Each AUC is a mean of three technical replicates. P values are calculated by t‐test. *P < 0.05, **P < 0.01. (E) Western blot of BCR signaling components, with and without anti‐IgM antibody stimulation, in mock and two shRNA‐transfected Z138 cells. Proteins were extracted 5 min after treatment with an anti‐IgM antibody. The expression levels of phosphorylated protein/total protein were compared using mock as a reference under each band as measured by densitometry analysis. ACTB was used as an internal control.

4. DISCUSSION

This study revealed that the miR‐17‐92 cluster, which is known as overexpressed oncogenic miRNA in MCL, regulates BCR signaling by in vitro analyses. Moreover, we identified BTG2 as a novel target of the miR‐17‐92 cluster and, importantly, that BTG2 downregulation led to BCR signal activation (Figure 7).

FIGURE 7.

The correlation diagram of SOX11, MIR17HG, miR17‐92 cluster, BTG2, and B‐cell receptor (BCR) signaling. This figure shares parts of Figures 4 and 6 and Figure S6.

miRNAs are small noncoding RNAs that silence endogenous target genes throughout mRNA degradation and translation inhibition. Targeting of miRNAs generally depends on the interaction between seed sequences of miRNAs consisting of seven or eight nucleotides and complementary sequences mainly in the 3′UTR of mRNAs. However, the recent development of sequencing approaches, such as argonaute cross‐linking immunoprecipitation, has demonstrated that miRNA targets are determined by not only the canonical interaction of 3′UTR of mRNAs but 5′UTR and coding regions. ³⁶ Therefore, the classical approach to identifying the targets of miRNAs using mRNA expression analysis and in silico predicting tools using miRNA sequences might not fully capture the targets of each miRNA. Moreover, target mRNAs of each miRNA potentially vary among the cancer types because of the different endogenous expression levels of mRNAs and miRNAs, which constitute complex regulatory networks with each other. Therefore, our pulldown‐seq approach is suitable for comprehensive analysis of the miR‐17‐92 cluster in MCL. BTG2 was identified as a potential target for miR‐17, miR‐20a, and miR‐92a in TargetScan (https://www.targetscan.org/vert_80/) (Table S2) and miRDB (https://mirdb.org/), and consistently the luciferase reporter assay confirmed that the canonical matched seed sequences for miR‐17, miR‐20a, miR‐92a, and miR‐18a, which is not listed by these in silico predicting tools, exists in BTG2. Interestingly, although the pulldown method revealed miR‐19a/19b significantly binds to BTG2 and this finding is validated by the protein level in inhibitor‐transfected samples, the luciferase assay could identify only modest binding in the 3′UTR region of BTG2. These results suggest the interaction between miR‐19a/19b and BTG2 might be regulated by noncanonical seed‐matched BTG2 regions, which were undetectable by the reporter assay. On the other hand, we could not capture all the miR‐17‐92 cluster targets associated with the BCR pathway in the pulldown‐seq analysis, partly because of the sensitivity of this method and the difference in dependency of these molecules for the BCR signal in the Z138 cells which we used.

BCL subtypes are dependent on different pathogenic BCR signal modes. The tonic BCR signal in GCB‐DLBCL and follicular lymphoma (FL) is activated to engage the PI3K pathway. ¹³ , ³⁷ The chronic active BCR signal, which engages the BCR‐dependent NF‐κB and PI3K signal, is activated under genetic alterations, such as CD79A/CD79B, CARD11, MYD88 mutations, and loss of TNFAIP3, in ABC‐DLBCL. ⁷ , ⁸ , ⁹ , ¹⁰ , ¹¹ The chronic active BCR signal in MCL is generally considered active. IGH‐V restriction in MCL suggests self‐antigen‐driven BCR signal activation and some patients with MCL harbor LRPAP1 self‐antibody. ³⁸ SOX11 overexpression and CXCR4‐dependent microenvironment interaction are considered important for BCR signal activation. ³⁹ However, the background of BCR signal activation is not sufficiently resolved. miR‐17‐92 cluster overexpression has been previously reported to activate the PI3K/AKT/mTOR pathway by targeting the protein phosphatase PHLPP2, PTEN, and BIM in MCL. ⁴⁰ , ⁴¹ In DLBCL, the miR‐17‐92 cluster in the presence of Myc enhances BCR signaling activity via ITIM‐containing proteins such as CD22 and FCGR2B. ²⁹ We evaluated the effect of individual mature miRNAs and revealed that miR‐18a, miR‐19a, miR‐19b, miR‐20a, and miR‐92a but not miR‐17 regulate BCR signaling. Considering the results that the inhibition of all six miRNAs consisting of miR‐17‐92 cluster repressed cell proliferation, we consider that oncogenic roles of miR‐17‐92 that partly depend on dysregulation of BCR signaling and other mechanisms also exist.

BTG2 is a member of the antiproliferative (APRO) gene family (TOB1‐2, BTG1‐4), which is downregulated in many cancers as a tumor suppressor gene and is involved in cell differentiation, proliferation, DNA repair, and apoptosis. ³⁵ , ⁴² The underlying mechanism of BTG2 downregulation is not fully described, and our findings illustrate one of the causes of tumor‐specific decreased BTG2 expression. Regarding the function of BTG2 in blood cells, BTG2/protein arginine methyltransferase‐1 (PRMT1) complex is generally generated in B cells, and CDK4 methylation and CCND3 inhibition lead to cell proliferation and differentiation inhibition through cell cycle arrest, ⁴³ , ⁴⁴ and are involved in mRNA stability in T cells. ⁴⁵

No reports have examined the function of BTG2 in malignant lymphoma pathogenesis, although BTG2 mutations are frequently found in several lymphomas, such as DLBCL ⁴⁶ and FL, ⁴⁷ and this mutation was reported as a poor prognostic factor in primary testicular DLBCL. ⁴⁸ Recently, some studies have demonstrated that functional disruption of BTG1, which belongs to the APRO family and is frequently mutated in GCB‐DLBCL and FL, drives the formation of aggressive BCLs. ⁴⁹ , ⁵⁰ BTG1 acts as an immune gatekeeper in GC B cells and mutant BTG1 promotes aggressive BCLs in humans and mice. ⁴⁹ BTG1 inactivation also drives lymphomagenesis throughout the interaction with the scaffold protein BCAR1. ⁵⁰ Regarding the mechanism by which BTG2 regulates BCR signaling, PRMT1, which interacts with BTG2 and catalyzes asymmetric dimethylation of arginine‐residues localized within glycine arginine‐rich regions, has been previously reported to suppress BCR signaling through CD79A R198 methylation just below the ITAM region. ⁴⁴ , ⁵¹ , ⁵² BTG2 inhibition and PRMT1 interaction in the MCL result in CD79A methylation inhibition and BCR signaling activation.

The mechanism of miR‐17‐92 cluster overexpression in MCL is not well understood, although 13q31‐32 amplification or gain, where MIR17HG is located, is found in 10%–20% of MCLs. ³⁰ We found that SOX11 overexpression induced overexpression of miR‐17‐92 in MCL cell lines. Importantly, miR‐17 overexpression has been described in SOX11‐positive MCLs compared to SOX11‐low‐expressing or negative MCLs in primary cases. ²⁸ The detailed mechanism of SOX11‐dependent BCR activation is unknown, although previous studies have demonstrated that SOX11‐overexpressed B lymphocytes develop MCL‐like lymphoma under BCR activation. ⁵³ Our result revealing that SOX11 regulates miR‐17‐92 suggests part of the cause of SOX11‐dependent BCR activation. Further investigation, especially using patient‐derived samples, is needed. Moreover, we demonstrated the enhancer‐regulated mechanism of miR‐17‐92 overexpression in a SOX11‐independent manner. Previous studies demonstrated that the expression level of normal naive B cells is lower than that of GC B cells. ¹⁸ The data of ChIP‐seq for acetylated histone profiling of normal B cells (GSE62063) revealed low enrichment of acetylated histone at MIR17HG enhancer regions in normal naïve B cells (data not shown). BRD4 facilitates the transcription of genes important for neoplastic cells in a cancer‐type specific manner associated with acetylated histones. ⁵⁴ Hence, our data support miR‐17‐92 overexpression as the tumor‐specific pathogenic event in MCLs.

The role of BTK inhibitors, such as ibrutinib in MCL treatment, is currently significant, and elucidating the mechanism of BCR signaling activation in MCL is an important issue in overcoming the clinical unmet need for targeted therapy resistance in this pathway. This study, which revealed that the miR‐17‐92 cluster activates BCR signaling by regulating BTG2 in MCL, has the potential to contribute to future investigations of therapeutic response and improvement of therapeutic outcomes in MCL. Furthermore, comprehensive analysis of the target molecules of the miR‐17‐92 cluster is expected to elucidate the pathogenesis of not only MCL but also many other malignant tumors.

AUTHOR CONTRIBUTIONS

Yuka Kawaji‐Kanayama: Investigation; writing – original draft. Taku Tsukamoto: Conceptualization; formal analysis; funding acquisition; investigation; writing – original draft. Masakazu Nakano: Formal analysis; investigation; writing – review and editing. Yuichi Tokuda: Formal analysis; investigation; writing – review and editing. Hiroaki Nagata: Investigation. Kentaro Mizuhara: Investigation. Yoko Katsuragawa‐Taminishi: Investigation. Reiko Isa: Investigation. Takahiro Fujino: Investigation. Yayoi Matsumura‐Kimoto: Investigation. Shinsuke Mizutani: Investigation. Yuji Shimura: Investigation. Masafumi Taniwaki: Supervision. Kei Tashiro: Supervision. Junya Kuroda: Conceptualization; supervision; writing – review and editing.

FUNDING INFORMATION

The study was supported in part by grants from Nippon Shinyaku Research Grant and JSPS KAKENHI (19K17867) (TT).

CONFLICT OF INTEREST STATEMENT

Masafumi Taniwaki received research funding from Daiichi Sankyo Pharmaceutical. Junya Kuroda received research funding from Kyowa Kirin, Chugai Pharmaceutical, Daiichi Sankyo Pharmaceutical, Ono Pharmaceutical, Eisai, Taiho Pharmaceutical, Sumitomo Pharma, Shionogi Pharmaceutical, and Bristol Myers Squibb, has received honoraria from Ono Pharmaceutical, Sanofi, Bristol Myers Squibb, and Janssen Pharmaceutical, and is a consultant for Janssen Pharmaceutical, Bristol Myers Squibb, Asahikasei Pharma, and Pfizer. Yuka Kawaji‐Kanayama, Taku Tsukamoto, Masakazu Nakano, Yuichi Tokuda, Hiroaki Nagata, Kentaro Mizuhara, Yoko Katsuragawa‐Taminishi, Reiko Isa, Takahiro Fujino, Yayoi Matsumura‐Kimoto, Shinsuke Mizutani, Yuji Shimura, and Kei Tashiro have no conflict of interest.

ETHICS STATEMENT

Approval of the research protocol by an Institutional Reviewer Board: N/A.

Informed Consent: N/A.

Registry and the Registration No. of the study/trial: N/A.

Animal Studies: N/A.

Supporting information

DATA S1.

FIGURE S1.

FIGURE S2.

FIGURE S3.

FIGURE S4.

FIGURE S5.

FIGURE S6.

TABLE S1.

TABLE S2.

Kawaji‐Kanayama Y, Tsukamoto T, Nakano M, et al. miR‐17‐92 cluster‐BTG2 axis regulates B‐cell receptor signaling in mantle cell lymphoma. Cancer Sci. 2024;115:452‐464. doi: 10.1111/cas.16027

DATA AVAILABILITY STATEMENT

The sequencing data reported in this article have been deposited in the DNA data bank of Japan (DDBJ) (https://www.ddbj.nig.ac.jp/) (accession number DRA016021).