1206

Martina Porstner et al.

DOI: 10.1002/eji.201444637

Eur. J. Immunol. 2015. 45: 1206–1215

miR-148a promotes plasma cell differentiation and targets the germinal center transcription factors Mitf and Bach2 Martina Porstner1 , Rebecca Winkelmann1 , Patrick Daum1 , Julia Schmid1 , Katharina Pracht1 , Joana Cˆ orte-Real1 , Sandra Schreiber1 , 2 Claudia Haftmann , Andreas Brandl1 , Mir-Farzin Mashreghi2 , Kolja Gelse3 , Manuela Hauke1 , Ina Wirries1 , Markus Zwick1 , Edith Roth1 , Andreas Radbruch2 , J¨ urgen Wittmann ∗1 and Hans-Martin J¨ ack ∗1 1

Division of Molecular Immunology, Department of Internal Medicine III, Nikolaus-Fiebiger-Center for Molecular Medicine, University of Erlangen-N¨ urnberg, Erlangen, Germany 2 Deutsches Rheuma-Forschungszentrum Berlin (DRFZ), Berlin, Germany 3 Department of Surgery, University of Erlangen-N¨ urnberg, Erlangen, Germany B cells undergo affinity maturation and class switch recombination of their immunoglobulin receptors during a germinal center (GC) reaction, before they differentiate into longlived antibody-secreting plasma cells (PCs). Transcription factors such as Bach2 and Mitf are essential during this process, as they delay premature differentiation of GC B cells by repressing Blimp-1 and IRF4, two transcription factors required for terminal PC differentiation. Therefore, Bach2 and Mitf expression must be attenuated in activated B cells to allow terminal PC differentiation, but the precise mechanism remains enigmatic. Here, we provide evidence that miR-148a, a small noncoding microRNA, fosters PC differentiation and survival. Next-generation sequencing revealed that miR-148a is the most abundant microRNA in primary human and murine PCs, and its expression is upregulated in activated murine B cells and coincides with Blimp-1 synthesis. miR-148a targets Bach2, Mitf and proapoptotic factors such as PTEN and Bim. When prematurely expressed, miR-148a promotes the differentiation and survival of plasmablasts and reduces frequencies of IgG1+ cells in primary B-cell cultures. In summary, we propose that miR-148a is a new player in the regulatory network controlling terminal PC differentiation and could, therefore, be a therapeutic target for interfering with PC differentiation and survival.

Keywords: B cells r Blimp-1 r MicroRNAs r miR-148a r Plasma cell differentiation r Bach2, Mitf See accompanying article by Haftmann et al.



Additional supporting information may be found in the online version of this article at the publisher’s web-site

Introduction The generation of antibody-secreting plasma cells (PCs) and memory B cells is essential for eliminating pathogens and establishing ¨ Correspondence: Prof. Hans-Martin Jack e-mail: [email protected]  C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

effective protection after vaccination. After antigen encounter, B cells are clonally expanded in the T-cell zone of a peripheral lymphatic organ and differentiate quickly into IgM-secreting short-lived PCs (extrafollicular pathway). If an antigen-activated ∗

These authors contributed equally to this work.

www.eji-journal.eu

Eur. J. Immunol. 2015. 45: 1206–1215

B cell receives T-cell help, some of these cells move from the Tcell zone to a B-cell follicle located in the B-cell zone. There, they are further expanded, resulting in the formation of a germinal center (GC). During a GC reaction, B cells undergo somatic hypermutation (SHM) of their immunoglobulin variable region exons and class switch recombination (CSR). After B cells with highaffinity and class-switched antigen receptors have been selected, they leave the GC and differentiate either into memory B cells or long-lived PCs [1–3]. The differentiation of mature B cells into long-lived PCs via the GC pathway is tightly controlled by a regulatory network of mutually exclusive transcription factors; one set maintains the GC reaction, and the other is required for PC differentiation (for reviews, see [3–5]). During a GC reaction, Pax5 and IRF4 cooperate to produce activation-induced cytidine deaminase (AID), a DNA mutator required for somatic hypermutation and CSR [6]. In addition, Bcl6 serves to maintain a centroblast in the cell cycle despite the presence of DNA damage introduced by AID [7], and Bach2 [8, 9] and Mitf [10] delay the premature differentiation of a GC B cell into a plasmablast by repressing the PC regulator Blimp-1 and preventing further upregulation of IRF4, respectively [10, 11], thereby increasing the time window for affinity maturation and CSR. In support of these roles, Bach2- and Mitf-deficient B cells differentiate more rapidly into PCs, resulting in fewer classswitched B cells and increased amounts of antigen-specific serum IgM [8, 10]. In contrast, IRF4 and Blimp-1 are both essential for initiating and terminating PC differentiation as well as maintaining the PC phenotype by controlling the expression of critical PC genes such as Xbp1 and downregulating GC B-cell identity genes such as Pax5 and Bcl6 (for reviews, see [2, 4, 5]). However, before a GC B cell can enter the PC pathway, Blimp-1- and IRF4-repressing transcription factors, such as Bach2 and Mitf, need to be downregulated, and the mechanism responsible for this process remains unknown. Small, noncoding microRNAs (miRNAs) are good candidates for fine-tuning the abundance of a transcription factor during the differentiation of a B cell into a PC. miRNAs regulate gene expression mainly at the posttranscriptional level by binding to target sequences mostly in the 3 untranslated region (UTR) of a mRNA, which results either in the attenuation of translation or the accelerated degradation of the respective mRNA (for review, see [12]), and it has been demonstrated that they play crucial roles during B-cell maturation and humoral immune responses (reviewed in [13–16]). Here, we present in vitro experiments with primary murine and human B cells that identified miR-148a as a factor that could promote terminal PC differentiation and survival. miR-148a was increased upon B-cell activation and was the most abundant miRNA in mature human and murine PCs. Intriguingly, this PC signature miRNA was found to downregulate several factors that are critical for PC maturation, among them are Bach2, Mitf, and the proapoptotic factors Bim and PTEN. Most importantly, premature expression of miR-148a by retroviral transduction favored plasmablast differentiation and the survival of in vitro activated primary murine B cells.  C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Molecular immunology

Figure 1. miR-148a is a plasma cell signature miRNA. RNA was isolated from magnetic and fluorescence-activated cell sorting of (A) murine pro-B- (pro-B), pre-B (pre-B), immature B (immature), transitional type 1 B (B1 B), marginal zone B (MZ B), follicular B (FO B), and plasma cells (PCs) from the spleen (spl) and BM of NZB/W- and C57BL/6 mice, as well as (B) from human CD22+ CD19+ blood B cells (blood B), blood CD38++ CD27++ CD19int. plasmablasts (PB), and BM CD138++ /CD38++ plasma cells (BM PCs). The miRanalyzer version 0.2 (miRBase 16) was used to determine the frequency of mature miR-148a reads (% of reads) from the next-generation sequencing data of RNA isolated from murine and human B-cell subsets (top). n = 1 for murine subsets; n = 2 for human blood B and BM PCs. Human blood PBs were pooled from four donors. The expression of the relative miR-148a abundance in the indicated B-cell subsets was confirmed by TaqMan quantitative (q) RT-PCR (bottom). The relative mean abundance of miR-148a ± SEM from triplicate measurements is displayed. n = 1 for murine pro-B to B1 B; n = 3 for murine FO B, spl PCs, and BM PCs; n = 2 for human PBs; n = 3 for human blood B and BM PCs. *p < 0.05, **p < 0.01; paired, two-tailed t-test.

Results miR-148a is upregulated upon naive B-cell activation and is the most abundant miRNA in BM PCs To identify miRNAs that are involved in B-cell activation and PC differentiation, we isolated miRNAs from various primary human and murine B-cell subsets, including mature BM PCs (Supporting Information Table 1), subjected them to next-generation sequencing (NGS), and quantified the sequence reads with the miRAnalyzer online tool [17]. Comparison of miRNA expression profiles from naive mature B-cell subsets with BM PCs identified a single miRNA, miR-148a, which made up as much as 30% of all expressed miRNAs in different PC isolates from mice (Fig. 1A) and humans (Fig. 1B). In contrast, miR-148a was barely expressed in other B-cell stages (Fig. 1A). NGS results for miR-148a were confirmed with TaqMan quantitative RT-PCR (Fig. 1, lower www.eji-journal.eu

1207

1208

Martina Porstner et al.

Figure 2. miR-148a expression is upregulated in activated splenic B cells and coincides with Blimp-1:GFP reporter gene expression. Primary CD43-negative B cells were isolated by magnetic cell sorting from the spleens of (A) C57BL/6 mice and (B, C) Blimp-1:GFP reporter mice and stimulated as indicated. (A) The kinetics of miR-148a expression in B cells after activation with LPS was evaluated. miRNAs were detected by northern blotting with specific 32 P-labeled DNA probes; miR-16 signals served as a control for the integrity and quantity of loaded RNA (left). The expression of miR-148a was confirmed by TaqMan qRT-PCR analysis (right). The miR-148a abundance relative to RNU6B signals is displayed as mean + SEM from n = 5 independent experiments, each measured in triplicate; **p < 0.01, Mann–Whitney U-test. (B) Analysis of miR-148a expression and Blimp-1 abundance in in vitro activated B cells. RNA was freshly isolated from B cells (0 h) and B cells stimulated for 72 h, as indicated. The relative miR-148a expression + SEM (left) and frequency of Blimp-1:GFP+ cells (right) were determined by TaqMan qRT-PCR and flow cytometric analysis, respectively. Data are shown as mean + SEM from n = 3 independent experiments, each performed in triplicate; *p < 0.05; ANOVA. (C) Freshly isolated B cells from Blimp-1:GFP reporter mice were activated for 72 h with LPS (left) or anti-CD40 + IL-21 (right). RNA was isolated from sorted GFP+ or GFP− cells, and the relative miR-148a expression was determined by Taqman qRT-PCR analysis. Data shown as mean from n = 3 experiments, each performed in triplicate; *p < 0.05; paired, two-tailed t-test.

diagrams). Compared to other PC-expressed miRNAs miR-148a was by far the most abundantly expressed miRNA (Supporting Information Table 2). To assess the kinetics of miR-148a expression during B-cell activation, CD43-negative naive splenic B cells were isolated from C57BL/6 mice and activated with LPS for several days. Northern blot and TaqMan qRT-PCR analyses (Fig. 2A, left and right panels) clearly revealed a significant upregulation of miR-148a after 48 h. Other stimulation approaches via various cytokine receptors,  C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Eur. J. Immunol. 2015. 45: 1206–1215

CD40 and TLR9, resulted only in a modest increase in miR-148a expression (Fig. 2B, left panel). Interestingly, LPS-induced activation of miR-148a expression could be blocked by cross-linking antibodies against IgM. As anti-IgM stimulation strongly interferes in vitro with LPS-induced plasmablast differentiation [18] (Fig. 2B, right panel) and miR-148a expression, one could argue that miR-148a controls plasmablast differentiation. In this case, B cells that have upregulated miR-148a expression should be determined for the PC differentiation pathway and, in line with this, should have induced the expression of Blimp-1, a PC signature transcription factor. To address this point, we isolated naive B cells from the spleens of mice carrying a GFP reporter knock-in under control of the Blimp-1 promoter [19] and monitored the changes in the frequency of Blimp-1:GFP-positive cells and miR-148a abundance in cultures before and 72 h after the addition of various B-cell stimuli (LPS, LPS+anti-IgM, CpG, anti-CD40+anti-IgM+IL-4 or anti-CD40+IL21). TaqMan qRT-PCR and GFP fluorescence analyses clearly revealed that the extent of miR-148a upregulation (Fig. 2B, left panel) correlated well with the induction of Blimp1:GFP reporter gene expression (Fig. 2B, right panel). In addition, we observed in sorted GFP-positive and -negative cells from cultures stimulated for 72 h either with LPS- (strong miR-148a induction) or with anti-CD40/IL21 (very weak miR-148a induction) very similar amounts of miR-148a in Blimp-1:GFP-positive cells in both settings (Fig. 2C). In contrast, the abundance of miR-148a in Blimp-1:GFP-negative cells was as low as in unstimulated B cells. Based on the findings that miR-148a is the most abundantly expressed miRNA in both human and murine longlived PCs and that its expression coincides with that of Blimp1, we hypothesized that miR-148a is part of a molecular circuit that controls the differentiation of antigen-activated B cells into antibody-secreting PCs.

miR-148a downregulates transcription factors that block premature PC maturation and favor cell death If miR-148a expression favors PC differentiation, this PC signature miRNA should control the expression of genes that delay PC differentiation. To test this prediction and to address the molecular mechanisms of miR-148a action, we used the miRNA target prediction algorithms TargetScan [20] and PicTar [21] to identify mRNAs whose expression could be controlled by miR-148a. The output target lists were filtered for mRNAs encoding proteins that are produced in B cells and are involved in B-cell activation and terminal PC differentiation. These filter criteria identified several interesting candidates (Supporting Information Fig. 1), including the PC inhibitory transcription factors Bach2 and Mitf, the proapoptotic proteins PTEN and Bim (BCL2L11) and DNA methyltransferase 1 (DNMT1), an enzyme that silences gene transcription by cytosine methylation associated with CpG islands. To biochemically confirm the in silico predicted targets, we cloned portions of the 3 UTRs of predicted candidate mRNAs into a dual luciferase vector (Fig. 3A, left), transiently www.eji-journal.eu

Eur. J. Immunol. 2015. 45: 1206–1215

Molecular immunology

Figure 3. miR-148a targets plasma cell delaying GC transcription factors Bach2 and Mitf. (A) Verification of in silico identified miR-148a targets with dual luciferase assays. Schematic presentation of the dual luciferase assay (left). HEK293 cells were transfected with a miR-148a expression construct together with a vector carrying the firefly luciferase gene as well as the Renilla luciferase gene fused to the indicated 3 UTR containing either the predicted miR-148a binding site (WT) or the mutated seed sequence (mut). The assays were performed 48 h later, and the ratio of firefly to Renilla luciferase activities in cells transfected with an empty expression vector (Control) was set to 100% (right). The mean and SEM values were calculated from five independent experiments performed in triplicate. The significance was tested using the two-tailed t-test. p-Values below 0.05 were considered significant. **p < 0.01; ***p < 0.001; n.s., not significant. (B-D) Verification of the luciferase-confirmed miR-148a targets DNMT1 (calculated molecular mass: 183 kDa), Bim (22 kDa), Mitf (52 kDa), and Bach2 (92 kDa) at the protein level by Western blotting in (B–C) primary murine B cells and (D) a human B-cell line DG75. β-Actin (42 kDa) or HSP90 (90 kDa) signals served as controls for the integrity and quantity of loaded protein. (B) LPS-preactivated splenic B cells were transduced with GFP-containing control or miR-148a-virus (encodes a mature sequence that is identical to that of murine miR-148a), sorted for GFP expression 24-h postinfection (p.i.) and analyzed for protein abundance 2 and 3 days postinfection. The myeloma cell lines JK6L and RPMI8226 were used as negative controls for Mitf expression. (C) For transient knockdown of mature miR-148a, murine splenic B cells were pulsed with scrambled control or miR-148a-specific anta miR-148a and cultured in the presence of LPS for 48 h. (D) The human Burkitt lymphoma cell line DG75 was transiently transfected with GFP-containing control or miR-148a expression plasmids. Forty-eight hours later, GFP-positive cells were isolated and analyzed by Western blotting. Data shown are from a single experiment representative of three performed.

transfected them together with a miR-148a expression vector or the respective control vector into HEK293 cells and performed dual luciferase assays. Among the seven 3 UTRs that displayed a significant reduction in luciferase activity in the presence of miR-148a expression were Mitf, PTEN, and Bach2, factors known  C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

to be involved in PC differentiation and survival (Fig. 3A, right; Supporting Information Fig. 2). To confirm that the observed miR-148a-mediated repression of Mitf, PTEN, and Bach2 was due to direct binding of the miRNA to its target mRNA, we repeated the assay with mutated miRNA www.eji-journal.eu

1209

1210

Martina Porstner et al.

seed sequences in the respective 3 UTRs. The repression of the luciferase reporter by miR-148a was significantly abrogated in all three cases, indicating a direct binding of miR-148a to the 3 UTR sequences of the respective mRNA targets (Fig. 3A, right). The regulation of DNMT1 by miR-148a has been reported in cholangiocytes and chronically activated CD4-positive T cells [22, 23], and miR-148a-mediated reduction of Bim abundance was recently confirmed in chronically activated T cells [24] and glioblastoma tumor cells [34]; thus, those miR-148a targets were not included in the luciferase assays. To determine whether miR-148a mRNA targets that were confirmed in HEK293 cells by a dual luciferase assay are also reduced in B cells upon miR-148a expression, we either overexpressed miR-148a or prevented its upregulation in activated B cells and determined the effect on the abundance of the corresponding proteins by Western blotting. The analysis of GFP+ cells from miR-148a/GFP- or control/GFP-vector transduced 1-day-old splenic B-cell cultures confirmed that DNMT1, Bim, and Mitf were reduced at protein levels in LPS blasts overexpressing miR-148a when compared to the corresponding β-actin loading control signals (Fig. 3B). In a reverse experiment, signals for Mitf, Bim, DNMT1, and PTEN were increased in LPS-activated splenic B cells in which the binding of mature miR-148a to its mRNA target sites was prevented by transfecting cells with cholesterol-tagged anta miR-148a (Fig. 3C). As no commercial anti-murine-Bach2 antibody was available for analyses, we transiently transfected the human Burkitt lymphoma cell line DG75, which barely expresses miR-148a, with miR-148a/GFP- and control/GFP-vectors (Supporting Information Fig. 3). As expected, Western blotting revealed decreased signals for Bach2, DNMT1, and Bim in sorted GFP+ DG75 cells when compared to signals in control vector-transfected GFP+ cells (Fig. 3D).

Ectopic miR-148a enhances IRF4 and Blimp-1 expression and promotes PC differentiation and survival miR-148a reduces the abundance of the proapoptotic polypeptides PTEN and Bim and the PC inhibitory factors Mitf and Bach2, which transcriptionally repress the expression of IRF4 and Blimp-1, respectively (Fig. 3). Therefore, we would expect that ectopic premature expression of miR-148a should favor PC maturation, increase the survival of LPS-activated B cells, and enhance IRF4 and Blimp-1 expression. To address these predictions, we retrovirally transduced LPS-pre-activated primary murine B cells with a vector encoding miR-148a and a GFP reporter, sorted GFP+ cells 1 day after infection and monitored CD138 expression, IgM secretion, and frequencies of antibody-secreting cells, all of which are hallmarks of PC differentiation [2]. Premature ectopic expression of miR-148a in transduced GFP+ -sorted splenic B cells was confirmed by Northern blot and TaqMan qRT-PCR analysis (Supporting Information Fig. 4A and B). Compared with controlGFP+ transduced cells, ectopic expression of miR-148a in splenic B cells indeed resulted in the accelerated appearance of CD138++  C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Eur. J. Immunol. 2015. 45: 1206–1215

plasmablasts (Fig. 4B). Furthermore, ELISA (Fig. 4C) and ELISpot (Fig. 4D) analyses of sorted GFP+ cells indicated that the frequency of IgM-secreting cells was also increased in cultures transduced with miR-148a. To confirm the positive effect of premature miR-148a expression on PC differentiation, we treated B cells from Blimp-1:GFP reporter mice with synthetic cholesterol-tagged miR-148a stem-loop precursor mimics (ConMiR-148a), measured the efficiency of premature miR-148a overexpression by TaqMan qRT-PCR analysis (Supporting Information Fig. 5A), and analyzed the effect on the frequency of CD138- and Blimp-1:GFP-positive cells over 4 days. As anticipated, we observed the accelerated appearance of CD138++ /Blimp-1:GFP+ cells in cultures treated with ConMiR-148a compared to cultures pulsed with a scrambled RNA control (Supporting Information Fig. 5B and C). To further corroborate the importance of miR-148a during PC differentiation, we used a complementary approach by reducing the abundance of miR-148a with cholesterol-tagged antagomirs [25] in LPS-activated CD43-negative splenic B cells from C57BL/6 mice. Compared to the results obtained in the experiments with prematurely expressed miR-148a, the knockdown approach revealed, as expected, the opposite effect: a small but significant reduction in the frequency of CD138+ B cells and IgM secretion following miR-148a knockdown (Supporting Information Figs. 6–8). Based on the finding that Bach2-deficient mice have fewer class-switched B cells [9] and that miR-148a targets Bach2 (Fig. 3D), premature expression of miR-148a should be accompanied with a reduction in the frequency of class-switched B cells. As predicted, we observed a significant decrease in the frequency of IgG1-switched cells in LPS/IL4-stimulated murine B-cell cultures after ectopic miR-148a expression (Fig. 4E). Because miR-148a targets the proapoptotic factor Bim (Fig. 3B and C and [24, 34]), we next analyzed whether premature expression of miR-148a affects the survival of LPS-activated B cells. Indeed, ectopic miR-148a expression in LPS-activated B-cell cultures resulted in a significant decrease in AnnexinV-positive cells (Fig. 4F). Premature expression of miR-148a, however, did not interfere with LPS-induced B-cell proliferation, because miR-148a and control vector-transduced splenic B cells, loaded with eFluor670 membrane dye before transduction, similarly lost eFluor670 membrane staining over time (Supporting Information Fig. 9). Therefore, we conclude that miR-148a affects survival, but does not interfere with the proliferative capacity of B cells. Finally, we found that the expression of the PC-promoting factors Blimp-1 and IRF4 was upregulated at the transcriptional level in miR-148a-transduced splenic B cells when compared to control vector-transduced B cells (Fig. 4G and Supporting Information Fig. 10). This supports the idea that miR-148a enhances plasmablast differentiation by targeting the IRF4- and Blimp-1repressors Mitf and Bach2 (Fig. 3B and D, respectively). In summary, we have identified miR-148a as a new PC signature miRNA and demonstrate that the expression of this miRNA is induced after B-cell activation. Most importantly, ectopic expression of miR-148a supports the differentiation of in vitro activated B cells into IgM-secreting plasmablasts, likely via the downregulation of the GC-maintaining transcription factors Bach2 and Mitf, www.eji-journal.eu

Eur. J. Immunol. 2015. 45: 1206–1215

Molecular immunology

Figure 4. Ectopic miR-148a expression favors plasmablast differentiation, induces Blimp-1 and IRF4 expression, and reduces the frequencies of IgG1-positive B cells. (A) Experimental set-up for the retroviral transduction of murine splenic B cells. Magnetic cell-sorted and purified CD43negative B cells were preactivated with LPS for 24 h and transduced with GFP-encoding retroviruses lacking or containing the gene for human miR-148a (encodes a mature sequence that is identical to that of murine miR-148a). (B–F) Twenty-four hours later, GFP+ cells were sorted (d1 p.i.) and analyzed on day 2 and 3 postinfection (d2 p.i. and d3 p.i.) for plasmablast differentiation by assessing (B) CD138 surface expression, (C, D) IgM secretion, (E) IgG class switching, and (F) apoptosis. (B) Representative flow cytometric analysis of CD138 fluorescence intensities (FI) among live-gated control (gray) and miR-148a (black) virus-transduced B cells on d3 p.i. is shown (left). Quantification of the mean frequency + SEM of CD138+ cells on d2 and d3 p.i., gated as displayed in the histogram (left) in 11 independent experiments (right, ***p < 0.001, Mann–Whitney U-test). (C) Supernatants from GFP+ -sorted control or miR-148a-transduced B cells were analyzed on d2 and d3 p.i. for the relative amounts of IgM by ELISA. The mean relative OD + SEM from five independent experiments performed in triplicate is displayed, **p < 0.01; paired, two-tailed t-test. (D) Triplicate LPS cultures of GFP+ -sorted control or miR-148a virus transduced B cells were analyzed on d2 and d3 p.i. for the frequency of IgM-secreting cells by ELISpot. The results of one representative ELISpot experiment and the mean number of spots + SEM (at a dilution of 250 cells/well) from five independent experiments performed in triplicate are shown (**p < 0.01, Mann–Whitney U-test). (E) Control and miR-148a virus transduced B cells were cultured in LPS/IL4, and the frequencies of gamma1-positive cells were determined on d2 and d3 p.i. using flow cytometry. Representative histograms displaying the frequencies of gamma1-positive cells among the GFP-positive fraction of control- (gray histogram) and miR-148a (black histogram)-transduced B-cell cultures and the quantification of the mean frequencies + SEM from five independent experiments are shown (**p < 0.01, Mann–Whitney U-test). (F) miR-148a-transduced B cells were stimulated with LPS and analyzed for AnnexinV-positive cells by flow cytometry. Representative histograms of the AnnexinV frequencies among control- (gray histogram) and miR-148a (black histogram) transduced GFP-positive cells are shown. The bar charts display the mean frequencies ± SEM of five independent experiments performed in triplicate (**p < 0.01). (G) LPS preactivated (1 day) CD43-negative B cells from spleens of C57BL/6 mice were infected with an eGFP-encoding controlor miR-148a retrovirus. RNA was isolated from sorted eGFP-positive cells 16 h after infection. mRNA abundance of Blimp-1 and IRF4 was measured by SYBR green qRT-PCR analysis and normalized to HPRT expression. The fold increase in miR-148a-tranduced B cells compared to control-vector transduced B cells from three independent transduction experiments (performed in triplicate) are presented as mean + SEM.  C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

www.eji-journal.eu

1211

1212

Martina Porstner et al.

thereby releasing the repression of the PC signature transcription factors Blimp-1 and IRF4, respectively.

Discussion Based on our findings, we suggest miR-148a is part of a hub of transcription factors and other miRNAs that in concert control and fine-tune the maturation of activated B cells and GC B cells into PCs and propose the following model: A GC B cell that receives a strong signal either via its high-affinity B-cell receptor during the interaction with its cognate antigen presented on a follicular dendritic cell or during the interaction with its cognate follicular T helper cell upregulates miR-148a. This results in the downregulation of Mitf and Bach2, and subsequently, the release of Mitf- and Bach2-mediated repression of IRF4 [10] and Blimp1 [8, 11, 26], which finally drives terminal PC differentiation [18, 27, 28]. Our data and model also allow for full integration of other miRNAs, which are either down- or upregulated during a GC reaction, in the fine-tuning of GC reactions, and establishment of memory B and long-lived PCs. For example, miR-125b, which prevents the premature expression of Blimp-1, is downregulated in GC B cells [29], and miR-24-3p, which promotes interleukin-6-mediated survival of PCs, is upregulated in activated human B cells [30]. Further, both miR-181b and miR-155 reduce the abundance of AID in GC B cells [31, 32], resulting in the inhibition of CSR. Additionally, miR-155 was proposed to control the maturation of B cells bearing high-affine B-cell receptors [33]. As miR-148a is the most abundant miRNA in mature murine and human BM PCs (Fig. 1 and Supporting Information Table 2), it could also play a pivotal role in maintaining the identity and life span of terminally differentiated PCs by preventing re-expression of Bach2 or Mitf and reducing the abundance of the proapoptotic proteins Bim and PTEN. Any potential miR-148a-mediated effects from targeting Bim and PTEN during a GC reaction may also contribute to an earlier stage of terminal PC differentiation by facilitating the survival of cells that have successfully upregulated miR-148a. In support, Bim has been identified as a miR-148a target in chronically activated T cells [24] and in glioblastoma [34]. Because DNMT1 negatively regulates miR-148a expression, for example in pancreatic and gastric cancer cells by methylation of CpG islands located upstream of the mature miR-148a sequence [35, 36], yet another function of miR-148a in B cells and very likely in mature PCs might be to downregulate its own repressor, that is DNMT1, during PC differentiation. This would reduce the abundance of DNMT1, which imparts methylation patterns during cell divisions (reviewed by [37]) to levels that do not interfere with the expression of miR-148a and other genes required for PC survival. Deregulated miR-148a expression can be detected in many tumors, for example from lung, stomach, prostate, brain, and pancreas, potentially affecting survival, proliferation, and cell migration by influencing the abundance of apoptotic proteins, transcrip-

 C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Eur. J. Immunol. 2015. 45: 1206–1215

tion factors (such as Mitf and Runx3), and the feedback regulatory miR-148a/DNMT1 axis [38–40]. miR-148a is also expressed in normal tissues, mainly in salivary glands and insulin-producing pancreatic β cells [41, 42], which are designated to secret large amounts of protein with a specialized function. This may imply a role for miR-148a not only in differentiation and survival, but also in the secretory mechanisms of cells destined to produce and secrete large amounts of a particular protein.

Materials and methods Construction of a retroviral miR-148a expression vector To construct a retroviral miR-148a expression vector, a fragment spanning 250 bp up- and downstream of the mature human miR-148a sequence (encodes a mature miRNA sequence that is identical to that of the murine sequence) was PCR-amplified using the primer set listed in Supporting Information Table 3, sequenced and cloned into the BamHI/EcoRI restriction sites of the GFPencoding retroviral pCLEP vector [43].

Construction of dual luciferase 3 UTR reporter gene constructs Fragments of the 3 UTRs of mRNAs predicted to contain potential miR-148a target sites by TargetScan [20] and PicTar [21] were PCR-amplified with primers containing XbaI or SpeI and NotI restriction sites (Supporting Information Table 3) and cloned into a modified psiCHECK-2 vector (Promega, Mannheim, Germany). The mutagenesis of the predicted miR-148a seed sequences in the Mitf (2x), PTEN (2x), and Bach2 3 UTR constructs was achieved by overlap PCR using the primers described in Supporting Information Table 5. Fragments with wild-type seed sequences for Mitf, PTEN, and Bach2 in the modified psiCHECK-2 vector were exchanged for fragments with mutated seed sequences by restriction digest cloning. The sequence identity was confirmed by sequencing (Seqlab, G¨ ottingen, Germany).

Dual luciferase reporter assay R For dual luciferase assays, a Dual-Luciferase Reporter Assay System (Promega, Mannheim, Germany) was used according to the manufacturer’s protocol. Briefly, 1 day prior to transfection, 0.9 × 105 HEK293 cells were seeded in antibiotic-free medium and then transfected in triplicate with the modified psiCHECK-2 vector containing the appropriate 3 UTR fragments and the miR-148aexpressing pcDNA3 vector or the corresponding empty vector using Lipofectamine 2000 (Invitrogen, Karlsruhe, Germany) according to the manufacturer’s protocol. Forty-eight hours after transfection, firefly and Renilla luciferase activities in the cell

www.eji-journal.eu

Eur. J. Immunol. 2015. 45: 1206–1215

lysates were measured after appropriate substrate addition using a Sirius Luminometer (Berthold Detection Systems, Pforzheim, Germany). The experiments were repeated independently five times in triplicate.

Retroviral transduction of splenic B cells Retroviral infection of splenic B cells was performed as described previously [43, 44]. Briefly, splenic B cells (0.5 × 106 /mL) were pre-stimulated with 10 μg/mL LPS (Sigma-Aldrich, Steinheim, Germany) for 24 h. Then, 1 × 106 B cells were resuspended in 2 mL of miR-148a-GFP or control-GFP virus supernatants containing 4 μg/mL polybrene (Sigma-Aldrich) (see supplements) and were centrifuged for 3.5 h at 1700 × g and 33°C. Next, the cells were cultured in fresh medium containing 10 μg/mL LPS. GFP+ cells were sorted 24 h after infection and kept in fresh medium containing 10 μg/mL LPS until analysis.

Molecular immunology

were PCR-amplified to approximately 20–30 ng/μL using a highfidelity DNA polymerase. Next, cDNAs were mixed in approximately equal amounts, and products of the appropriate length were fractionated on a preparative 6% polyacrylamide gel by electroelution. The cDNAs were 454- and Solexa-sequenced by GATC Biotech (Konstanz, Germany). Bioinformatic analysis was performed with the miRAnalyzer software [17].

Supplemental online procedures Statistical methods and standard procedures such as animal handling, cell culture, flow cytometry, RNA and protein analyses, ELISA and ELISpot and the eFluor670, AnnexinV and IgG1 staining procedures are described in the online Supporting Information.

Transient overexpression or knockdown of miR-148a For transient miR-148a overexpression, 1 × 106 human DG75 B cells were transfected with 2 μg of miR-148a-GFP or controlGFP expression plasmids using the Lonza cell nucleofection system according to manufacturer’s instructions (Lonza, Cologne, Germany). After 48 h, the cells were sorted for GFP expression and subjected to RNA or protein extraction. The transient knockdown of miR-148a in primary splenic B cells was performed with cholesterol-tagged miR-148a-specific anta miRs or scrambled control RNA (Riboxx, Radebeul, Germany) by pulsing 10 × 106 splenic B cells directly after isolation in 1 mL of RPMI1640 starvation medium without any supplements or stimuli with 3 nmol of oligos for 1.5 h at 37°C. After pulsing, the cells were cultured in complete medium supplemented with 10% FCS and stimulated as described. Alternatively, splenic B cells (5 × 106 cells/mL) were incubated in serum-free Accell medium (Dharmacon/Thermo Scientific, Lafayette, CO, USA) supplemented with 1 μM anta miR-148a or a scrambled control oligonucleotide (Dharmacon/Thermo Scientific) for 1.5 h at 37°C and 5% CO2 , diluted with RPMI1640 medium [including 10% FCS, supplements and LPS (10 μg/mL)] to 0.25 × 106 cells/mL and cultured in a 24-well plate prior to analysis.

Acknowledgments: This work was supported in part by the Interdisciplinary Center for Clinical Research (IZKF) Erlangen (projects D7, J21) to J.W., R.W., and H.-M.J. and research grants GRK1660, FOR832 (JA968/4), and TRR130 from the Deutsche Forschungsgemeinschaft (DFG) to H.-M.J. J.S., and K.P. received fellowships from the DFG doctoral training grant GRK1660, and J.C.-R. is a fellow from the Alexander-von-Humboldt Foundation. M.-F.M. was supported by the “e:Bio Innovationswettbewerb Systembiologie” program of the Federal Ministry of Education and A.R. from the ERC Advanced Grant (ERC-2010-AdG 20100317 Grant 268987). We thank Maureen Grady for bioinformatic support, Alexandra Klej for technical assistance, Uwe Appelt and Markus Mrotz for MoFlo cell sorting, and Stephen Nutt (WEHI, University of Melbourne) for kindly providing us with Blimp-1:GFP reporter mice.

Conflict of interest: The authors declare no financial or commercial conflict of interest.

References 1 Shapiro-Shelef, M. and Calame, K., Regulation of plasma-cell develop-

NGS

ment. Nat. Rev. Immunol. 2005. 5: 230–242. 2 Oracki, S. A., Walker, J. A., Hibbs, M. L., Corcoran, L. M. and Tarlinton, D. M., Plasma cell development and survival. Immunol. Rev. 2010. 237:

For NGS, small RNA libraries of sorted B-cell subsets were generated by Vertis Biotechnologie AG (Freising, Germany). Briefly, small RNAs were isolated from sorted cells (see Supporting Information Table 1 for details) using the mirVana miRNA isolation kit (Ambion Life Technologies, Karlsruhe, Germany) and separated on a denaturing 15% polyacrylamide gel. RNAs with lengths of 19–29 bases were extracted and eluted, and poly(A) tails were added. Then, a 5 -RNA adapter containing a specific “bar code” sequence for each sample was ligated. After first-strand cDNA synthesis using an oligo(dT)-adapter primer, the resulting cDNAs  C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

140–159. 3 Shlomchik, M. J. and Weisel, F., Germinal center selection and the development of memory B and plasma cells. Immunol. Rev. 2012. 247: 52–63. 4 Nutt, S. L., Taubenheim, N., Hasbold, J., Corcoran, L. M. and Hodgkin, P. D., The genetic network controlling plasma cell differentiation. Semin. Immunol. 2011. 23: 341–349. 5 Alinikula, J. and Lassila, O., Gene interaction network regulates plasma cell differentiation. Scand. J. Immunol. 2011. 73: 512–519. 6 Muramatsu, M., Kinoshita, K., Fagarasan, S., Yamada, S., Shinkai, Y. and Honjo, T., Class switch recombination and hypermutation require

www.eji-journal.eu

1213

1214

Martina Porstner et al.

activation-induced cytidine deaminase (AID), a potential RNA editing enzyme. Cell 2000. 102: 553–563.

Eur. J. Immunol. 2015. 45: 1206–1215

T-bet and promotes the survival of T helper 1 cells by regulating the proapoptotic gene Bim. Eur. J. Immunol. 2015. 45: 1192–1205.

7 Fukuda, T., Yoshida, T., Okada, S., Hatano, M., Miki, T., Ishibashi, K.,

25 Stittrich, A. B., Haftmann, C., Sgouroudis, E., Kuhl, ¨ A. A., Hegazy, A.

Okabe, S. et al., Disruption of the Bcl6 gene results in an impaired ger-

N., Panse, I., Riedel, R. et al., The microRNA miR-182 is induced by IL-2

minal center formation. J. Exp. Med. 1997. 186: 439–448.

and promotes clonal expansion of activated helper T lymphocytes. Nat.

8 Muto, A., Tashiro, S., Nakajima, O., Hoshino, H., Takahashi, S., Sakoda, E., Ikebe, D. et al., The transcriptional programme of antibody class switching involves the repressor Bach2. Nature 2004. 429: 566–571. 9 Muto, A., Ochiai, L., Kimura, Y., Itoh-Nakadai, A., Calame, K. L., Ikebe,

Immunol. 2010. 11: 1057–1062. 26 Ochiai, K., Katoh, Y., Ikura, T., Hoshikawa, Y., Noda, T., Karasuyama, H., Tashiro, S. et al., Plasmacytic transcription factor Blimp-1 is repressed by Bach2 in B cells. J. Biol. Chem. 2006. 281: 38226–38234.

D., Tashiro, S. et al., Bach2 represses plasma cell gene regulatory net-

27 Klein, U., Casola, S., Cattoretti, G., Shen, Q., Lia, M., Mo, T., Ludwig, T.

work in B cells to promote antibody class switch. EMBO J. 2010. 29:

et al., Transcription factor IRF4 controls plasma cell differentiation and

4048–4061.

class-switch recombination. Nat. Immunol. 2006. 7: 773–782.

10 Lin, L., Gerth, A. J. and Peng, S. L., Active inhibition of plasma cell devel-

28 Sciammas, R., Shaffer, A. L., Schatz, J. H., Zhao, H., Staudt, L. M. and

opment in resting B cells by microphthalmia-associated transcription

Singh, H., Graded expression of interferon regulatory factor-4 coordi-

factor. J. Exp. Med. 2004. 200: 115–122.

nates isotype switching with plasma cell differentiation. Immunity 2006.

11 Ochiai, K., Muto, A., Tanaka, H., Takahashi, S. and Igarashi, K., Regulation of the plasma cell transcription factor Blimp-1 gene by Bach2 and Bcl6. Int. Immunol. 2008. 20: 453–460. 12 Winter, J., Jung, S., Keller, S., Gregory, R. I. and Diederichs, S., Many roads to maturity: microRNA biogenesis pathways and their regulation. Nat. Cell. Biol. 2009. 11: 228–234. 13 Belver, L., Papavasiliou, F. N. and Ramiro, A. R., MicroRNA control of lymphocyte differentiation and function. Curr. Opin. Immunol. 2011. 23: 368–373. 14 Fernando, T. R., Rodriguez-Malave, N. I. and Rao, D. S., MicroRNAs in B cell development and malignancy. J. Hematol. Oncol. 2012. 5: 7. 15 Li, J., Wan, Y., Ji, Q., Fang, Y. and Wu, Y., The role of microRNAs in B-cell development and function. Cell. Mol. Immunol. 2013. 10: 107–112. 16 Baumjohann, D. and Ansel, K. M., MicroRNA regulation of the germinal center response. Curr. Opin. Immunol. 2014. 28C: 6–11. 17 Hackenberg, M., Sturm, M., Langenberger, D., Falcon-P ´ erez, ´ J. M. and Aransay, A. M., miRanalyzer: a microRNA detection and analysis tool for next-generation sequencing experiments. Nucleic Acids Res. 2009. 37(Web Server issue): W68–W76. 18 Berberich, I. and Schimpl, A., Regulation of immunoglobulin gene expression in normal lymphocytes. II. Mechanisms of down-regulation of immunoglobulin secretion after engagement of the B cell antigen receptor. Eur. J. Immunol. 1992. 22: 525–529. 19 Kallies, A., Hasbold, J., Tarlinton, D. M., Dietrich, W., Corcoran, L. M., Hodgkin, P. D. and Nutt, S. L., Plasma cell ontogeny defined by quantitative changes in blimp-1 expression. J. Exp. Med. 2004. 200: 967–977. 20 Lewis, B. P., Burge, C. B. and Bartel, D. P., Conserved seed pairing, often flanked by adenosines, indicates that thousands of human genes are microRNA targets. Cell 2005. 120: 15–20. 21 Krek, A., Grun, ¨ D., Poy, M. N., Wolf, R., Rosenberg, L., Epstein, E. J., MacMenamin, P. et al., Combinatorial microRNA target predictions. Nat. Genet. 2005. 37: 495–500. 22 Braconi, C., Huang, N. and Patel, T., MicroRNA-dependent regulation of DNA methyltransferase-1 and tumor suppressor gene expression by interleukin-6 in human malignant cholangiocytes. Hepatology 2010. 51: 881–890.

25: 225–236. 29 Gururajan, M., Haga, C. L., Das, S., Leu, C. M., Hodson, D., Josson, S., Turner, M. et al., MicroRNA 125b inhibition of B cell differentiation in germinal centers. Int. Immunol. 2010. 22: 583–592. 30 Gabler, J., Wittmann, J., Porstner, M., Renz, H., Jack, ¨ H. M., Abram, M. and Zemlin, M., Contribution of microRNA 24-3p and Erk1/2 to interleukin-6mediated plasma cell survival. Eur. J. Immunol. 2013. 43: 3028–3037. 31 de Yebenes, ´ V. G., Belver, L., Pisano, D. G., Gonzalez, ´ S., Villasante, A., Croce, C., He, L. et al., miR-181b negatively regulates activation-induced cytidine deaminase in B cells. J. Exp. Med. 2008. 205: 2199–2206. 32 Thai, T. H., Calado, D. P., Casola, S., Ansel, K. M., Xiao, C., Xue, Y., Murphy, A. et al., Regulation of the germinal center response by microRNA155. Science 2007. 316: 604–608. 33 Basso, K., Schneider, C., Shen, Q., Holmes, A. B., Setty, M., Leslie, C. and Dalla-Favera, R., BCL6 positively regulates AID and germinal center gene expression via repression of miR-155. J. Exp. Med. 2012. 209: 2455–2465. 34 Kim, J., Zhang, Y., Skalski, M., Hayes, J., Kefas, B., Schiff, D., Purow, B. et al., microRNA-148a is a prognostic oncomiR that targets MIG6 and BIM to regulate EGFR and apoptosis in glioblastoma. Cancer Res. 2014. 74: 1541–1553. 35 Hanoun, N., Delpu, Y., Suriawinata, A. A., Bournet, B., Bureau, C., Selves, J., Tsongalis, G. J. et al., The silencing of microRNA 148a production by DNA hypermethylation is an early event in pancreatic carcinogenesis. Clin. Chem. 2010. 56: 1107–1118. 36 Zhu, A., Xia, J., Zuo, J., Jin, S., Zhou, H., Yao, L., Huang, H. et al., MicroRNA-148a is silenced by hypermethylation and interacts with DNA methyltransferase 1 in gastric cancer. Med. Oncol. 2012. 29: 2701–2709. 37 Jurkowska, R. Z., Jurkowski, T. P. and Jeltsch, A., Structure and function of mammalian DNA methyltransferases. Chembiochem 2011. 12: 206–222. 38 Chen, Y., Song, Y. X. and Wang, Z. N., The microRNA-148/152 family: multi-faceted players. Mol. Cancer. 2013. 12: 43. 39 Xia, J., Guo, X., Yan, J. and Deng, K., The role of miR-148a in gastric cancer. J. Cancer Res. Clin. Oncol. 2014. 140: 1451–1456. 40 Chen, Z., Ma, T., Huang, C., Zhang, L., Xu, T., Hu, T. and Li, J., MicroRNA148a: a potential therapeutic target for cancer. Gene 2014. 533: 456–457. 41 Melkman-Zehavi, T., Oren, R., Kredo-Russo, S., Shapira, T., Mandel-

23 Pan, W., Zhu, S., Yuan, M., Cui, H., Wang, L., Luo, X., Li, J. et al.,

baum, A. D., Rivkin, N., Nir T. et al., E, miRNAs control insulin content

MicroRNA-21 and microRNA-148a contribute to DNA hypomethylation

in pancreatic β-cells via downregulation of transcriptional repressors.

in lupus CD4+ T-cells by directly and indirectly targeting DNA methyl-

EMBO J. 2011. 30: 835–845.

transferase 1. J. Immunol. 2010. 184: 6773–6781. 24 Haftmann, C., Stittrich, A. B., Zimmermann, J., Fang, Z., Hradilkova, K., Bardua, M., Westendorf, K. et al., miR-148a is upregulated by Twist1 and

 C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

42 Kuchen, S., Resch, W., Yamane, A., Kuo, N., Li, Z., Chakraborty, T., Wei, L. et al., Regulation of microRNA expression and abundance during lymphopoiesis. Immunity 2010. 32: 828–839.

www.eji-journal.eu

Eur. J. Immunol. 2015. 45: 1206–1215

43 Brandl, A., Wittmann, J. and Jack, ¨ H. M., A facile method to increase titers of miRNA-encoding retroviruses by inhibition of the RNaseIII enzyme Drosha. Eur. J. Immunol. 2011. 41: 549–551. 44 Brachs, S., Turqueti-Neves, A., Stein, M., Reimer, D., Brachvogel, B., Bosl, ¨ M., Winkler, T. et al., Swiprosin-1/EFhd2 limits germinal center responses and humoral type 2 immunity. Eur. J. Immunol. 2014. 44: 3206– 3219.

Abbreviations: AID: activation-induced cytidine deaminase · CSR: class switch recombination · DNMT1: DNA methyltransferase 1 · miRNA: microRNA · NGS: next-generation sequencing · PC: plasma cell · UTR: untranslated region

 C 2015 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Molecular immunology

¨ Full correspondence: Prof. Hans-Martin Jack, Division of Molecular Immunology, Department of Internal Medicine III, Nikolaus-Fiebiger-Center for Molecular Medicine, University of ¨ Erlangen-Nurnberg, D-91054 Erlangen, Germany Fax: +49(0)9131-8539343 e-mail: [email protected] See accompanying article: http://dx.doi.org/10.1002/eji.201444633 Received: 8/3/2014 Revised: 9/1/2015 Accepted: 9/2/2015 Accepted article online: 12/2/2015

www.eji-journal.eu

1215

miR-148a promotes plasma cell differentiation and targets the germinal center transcription factors Mitf and Bach2.

B cells undergo affinity maturation and class switch recombination of their immunoglobulin receptors during a germinal center (GC) reaction, before th...
1MB Sizes 1 Downloads 5 Views