Burkitt lymphoma (BL) is a highly aggressive B-cell lymphoma that includes two forms of BL differing in Epstein–Barr virus (EBV) infection status, EBV-positive and EBV-negative. Although many efforts, such as high-intensity, short-duration combination chemotherapy, have been devoted to improving therapy for this rapidly proliferating neoplasm, there are still significant treatment-associated toxicities. Therefore, there remains a need for novel effective therapeutic strategies. MicroRNAs play a role in “fine tuning” the physiological and pathological differentiation process, by which cells can rapidly regulate dynamic events such as cell-lineage decisions and morphogenesis. This unique miRNA feature shifts the traditional one drug target paradigm to a novel one drug multiple targets paradigm. Here, we found that BL cell lines showed an extremely low expression of microRNA-150 (miR-150), and then restored miR-150 expression at physiologic levels in BL cell lines Daudi, Raji, BJAB, and Ramos. The results showed that re-expression of miR-150 reduced proliferation of Daudi and Raji cells. Furthermore, Daudi and Raji, both of which are of EBV-positive germinal center B-cell origin, transduced with miR-150 can be rescued to differentiate toward B-cell terminal stage. However, no significant changes were observed in BJAB or Ramos cells, which are of EBV-negative germinal center B-cell origin. Of note, re-expression of miR-150 also resulted in decreasing c-Myb protein levels. Additionally, c-Myb knockdown in Daudi and Raji cell lines recapitulated the partial characteristics similar to that caused by re-expression of miR-150. Taken together, our findings show that miR-150 can induce EBV-positive BL differentiation by targeting c-Myb.
Burkitt lymphoma is a highly aggressive B-cell malignancy, identified and described for the first time by Dennis Burkitt in 1958. According to the EBV infection status, BLs are divided into two subgroups, EBV-positive and EBV-negative. Burkitt lymphomas have a tendency to morphologically resemble germinal center (GC) cells, and to immunophenotypically express characteristic GC cell markers such as CD10 and BCL6, and coexpress B-cell markers CD19 and CD20, which suggests follicle center B-cell origin for this lymphoma. In recent years, efforts have focused on improving therapy for this, the fastest growing human tumor, while minimizing treatment-associated toxicity. Outcome with intensive chemotherapy has improved and is now excellent in children, but the prognosis is poor in elderly adults. In addition, there are significant associated toxicities, such as frequent myelosuppression, mucositis, neuropathy, and complication of tumor lysis syndrome.[5, 6] Therefore, it is imperative and important to pursuit more effective therapies.
MicroRNAs, a novel class of small non-coding RNAs of approximately 19–26 nt, exert multiple cellular functions and play critical roles in cellular proliferation, apoptosis, cellular differentiation, and tumorigenesis.[7-9] MicroRNA expression profiling studies have found that in hematopoiesis, certain miRNAs are expressed in a stage-specific fashion in the lymphoid hematopoietic system.[10-14] Therefore, deregulations of their expression may result in the development of hematopoietic malignancies.
MicroRNA-150 has attracted our great attention in hematopoiesis among a large number of miRNAs in recent years, which has been studied in B, T, and NK/T cells.[16-20] There is strong evidence that miR-150 is preferentially expressed in mature lymphocytes, but not their progenitors. Ectopic miR-150 expression of murine hematopoietic stem cells impairs the transition of pro-B to pre-B stage. In addition, miR-150 is found to target c-Myb, which is highly expressed in lymphocyte progenitors, downregulated during maturation, and again increased after activation of the mature cells. It has been shown that expression of miR-150 was consistently downregulated in diffuse large B-cell lymphoma cell lines compared with centroblasts and in BL compared with the normal tissues. Furthermore, aberrantly low expression of miR150 is identified in Sezary syndrome and NK/T-cell lymphoma. Re-expression of miR-150 in NK/T-cell lymphoma cells increases the incidence of apoptosis and reduces cell proliferation, suggesting that miR-150 functions as a tumor suppressor.
The discovery of miRNAs has added an entirely new dimension to antitumor therapeutic approaches. Depending on miRNA function and status in tumor, miRNAs are generally classified as tumor suppressors or oncomiRs. Therefore, from the therapeutic point of view, tumor suppressor-miRNAs can be induced by ectopic expression, and those oncogenic-miRNAs can be inhibited. Re-expressing lost miRNA in a cell can deliver a dramatic effect, because miRNAs regulate a vast number of genes and pathways. The ectopic expression of miR-223, miR15/miR-16, let7, mir-342, miR-29a, and miR-142-3p in acute myeloid leukemia could stimulate myeloid differentiation of leukemia cells.[26-30] Therefore, differentiation induction by miRNAs is proposed to be a potentially attractive strategy for leukemia. Recently, our group found that CD99 can trigger Hodgkin/Reed–Sternberg cells redifferentiation by upregulating miR-9-modulated PRDM1/BLIMP1. Here, we found that BL cell lines showed an extremely low expression of miR-150. Consequently, testing whether transfection of miR-150 into BL cell lines can shift the regulators' expression profiles toward that of the terminal B-cell is one purpose in our work. In the immature hematopoietic system of a mouse model, miR-150 functions by targeting c-Myb, however, it remains unknown whether there is a similar relationship between miR-150 and c-Myb in BL.
Materials and Methods
The BL cell lines Raji (EBV+), Daudi (EBV+), BJAB (EBV−), and Ramos (EBV−), the plasmacytoma cell line KM3, the acute T-cell leukemia cell line Jurkat, the anaplastic large cell lymphoma cell line KARPAS-299, and the classic Hodgkin's lymphoma cell line L428, were cultured in RPMI-1640 (Gibco, Los Angeles, CA, USA) supplemented with 10% heat-inactivated FBS (HyClone, Logan, UT, USA). All cell lines were cultured under humidified air containing 5% CO2 at 37°C.
Normal B-cell subpopulations
Tissue samples from patients undergoing routine tonsillectomy were collected from the Nanfang Hospital affiliated to South Medical University (Guangzhou, China). Informed consent was obtained from patients. Normal B lymphocyte subpopulations from three different tonsil samples were sorted by FACS, as described earlier.[32, 33] The detailed procedure is available in Figure S1. The antibodies used for FACS are available in Table S1.
Transfection and cell sorting
Stable expression lentiviral vectors for miR-150 were provided by Shanghai GeneChem (Shanghai, China), and the lentiviral supernatant was synthesized by GeneChem according to standard molecular biology techniques as previously described. The following lentiviral vectors were used: (i) PP-GFP bearing a cDNA fusion of GFP and multiple clone site under the CMV promoter; and (ii) a control vector bearing only the CMV-GFP cassette. The pre-miR-150 containing the hairpin region of miR-150 and ~200 bp flanking sequences on each side was amplified by PCR from human genomic DNA with primers comprising 5′-linker sequences harboring relevant digestion sites (core primer sequences, 5′-CAGCATAGGGTGGAGTGGGT-3′ and 5′-TACTTTGCGCATCACACAGA-3′). The miR-150 expression cassette was cloned into the XhoI and EcoRI sites in the lentiviral vector PP-GFP.
For cell sorting after transfection, cells were suspended at the concentration of 1 × 107/mL in PBS then sorted for GFP expression on a FACSAria (BD Biosciences, San Diego, CA, USA).
For RNAi, siRNA sequences targeting the human c-Myb and c-Myc were designed and synthesized by Sigma (St. Louis, MO, USA) (siRNA ID, SASI_Hs01_00127047 and SASI_Hs01_00222676, respectively). A scrambled siRNA was included as a negative control. The miR-Ribo miRNA inhibitor, a competitive inhibitor with hsa-miR-150, was designed and synthesized by Ribobio (Guanghzou, China), and miR02101 miR-Ribo inhibitor NControl#22 was used as a negative control. The Lipofectamine 2000 system (Invitrogen, Los Angeles, CA, USA) was used for transient transfection according to the manufacturer's instructions.
RNA isolation, reverse transcription, and quantitative real-time PCR analysis
Total RNA extraction, reverse transcription, and qRT-PCR were carried out as previously described. Primers specific for human and reaction conditions are available in Table S2. GAPDH was used as an internal control.
For quantitative detection of miR-150, 2 μg total RNA was reversely transcribed. The qRT-PCR was carried out as previously described. The sequence of miR-150 is available in Table S2. RNA-U48 was used as an internal control.
The relative levels were calculated using the 2−ΔΔCt method. The qRT-PCR experiments were run independently three times using synthesized cDNAs.
Cell viability assay
The MTT assay was used for cell viability experiments, as previously described.
All flow cytometric immunophenotypic analyses of cells were carried out on a Beckman Coulter FC500 MPL at Kingmed Medical Test Center (Guangzhou, China); cell sorting was carried out at the Chinese Academy of Sciences (Guangzhou, China). A panel of antibodies used for flow cytometric analysis is available in Table S1.
Western blot analysis
The procedure was carried out as previously described. The antibodies are listed in Table S3.
Confocal microscopy analysis
The procedure was carried out as previously described.
All images, such as Western blot and flow cytometry, are representative of at least three independent experiments. Quantitative RT-PCR assays were carried out in triplicate for each experiment. The data shown are presented as the mean ± SD for three independent experiments. The spss software version 13.0 (SPSS Inc., Chicago, IL, USA) was used for statistical analyses. Differences were considered statistically significant at P < 0.05, assessed using the two-tailed paired Student's t-test for assays with qRT-PCR, and flow cytometry or using the two-tailed independent Student's t-test for MTT assays.
Expression profiles of miR-150 and c-Myb in B-cell subpopulations from tonsils and in human peripheral lymphoma cell lines
It has been documented that miRNAs were expressed in a stage-specific fashion in the lymphoid hematopoietic system,[10-14] so we used qRT-PCR to analyze the expression patterns of miR-150 in B-cell subsets: naïve B-cells; GC centroblasts; GC centrocytes; and memory B-cells, all of which were obtained from tonsils (Fig. 1a). The results showed that the expression pattern of miR-150 of four B-cell subpopulations was dynamic, and there were low levels of miR-150 in centroblasts and centrocytes, but high levels in memory B-cells with respect to naïve cells. To obtain insights into the relationship between miR-150 and c-Myb, we determined the expression levels of c-Myb in the four B-cell subsets (Fig. 1b). A reciprocal correlation of expression patterns of miR-150 and c-Myb was found.
It is well known that miRNAs are involved in carcinogenesis and lymphomagenesis. Therefore, we detected the levels of expression of miR-150 in human lymphoma cell lines Daudi, Raji, BJAB, Ramos, KM3, L428, KARPAS-299, and Jurkat. The result showed that miR-150 expression was significantly reduced in all cell lines with respect to normal B-cells (Fig. 1c). To further explore the association between miR-150 with c-Myb, we analyzed the expression of c-Myb in lymphoma cell lines (Fig. 1d,e). The results showed that most of these lymphoma cells strongly expressed c-Myb compared with B-cells.
Thus, the complementary expression pattern of miR-150 and c-Myb might suggest their intrinsic correlation, and deregulated miR-150 and c-Myb activity may contribute to human lymphomagenesis.
Re-expression of miR-150 disadvantages EBV-positive BL growth
Considering the dynamic and differential expression patterns of miR-150 during the B-cell development and its low expression or absence in human lymphoma cell lines, we explored the potential role of miR-150 in BL in vitro. We produced lentiviral vector constitutively expressing pre-miR-150 labeled with GFP, then stably transfected it into Daudi, Raji, BJAB, and Ramos cells along with a control vector labeled with GFP. The efficiency of transfection was measured by qRT-PCR (Fig. 2a). To assess whether lentiviral vector produced physiologically relevant levels of mature miR-150, we compared the expression of miR-150 in lentivirally infected Daudi, Raji, BJAB, and Ramos cells to one in normal GC centroblasts. The results showed that transfected cells expressed miR-150 at approximately physiological levels (Fig. 2a).
It has been documented that ectopic miR-150 expression increased cell death of pro-B cells in vitro and susceptibility to apoptosis and reduced cellular proliferation in NK/T-cell lymphoma cell lines. Consequently, we tested whether there was a similar effect in our experiment. Given that re-expression of miR-150 can lead to a decrease in proliferation of transduced tumor cells, keeping the cells in culture for a longer time would result in positive selection of those with low or no GFP, which finally outnumbered the GFP positive cells. To this end, combined with the method described by Watanabe et al., we monitored weekly the GFP expression in the mixed cultures with GFP-miR-150 positive (GFP-miR-150+) cells and GFP negative (GFP−) cells by flow cytometer. Our results showed that there was a significant reduction in the percentage of GFP-miR-150+ compared with GFP− cells in Daudi-miR-150 and Raji-miR-150 but not in BJAB-miR-150 or Ramos-miR-150 (Fig. 2b,c). The effect of reduced growth in transduced Daudi and Raji cell lines was confirmed by MTT assay (Fig. 2d,e). These results indicate that miR-150 possesses a tumor-suppressing property that is consistent with a previous report.
Next, we evaluated whether miR-150 re-expression could promote apoptosis. However, a strong increase in the early apoptotic fraction and late apoptotic cells compared with control was only observed in Daudi-miR-150 (Fig. S2).
Re-expression of miR-150 induces EBV-positive BL differentiation in vitro
We then explored whether restoration of miR-150 could trigger differentiation of BL. Both BCL6 and PRDM1 are critical markers of mature B-cell differentiation. Additionally, a high bcl-2 expression has been found in the more mature B lymphocytes, particularly in plasma cells that are long-living cells.[37, 38] Our results showed that re-expression of miR-150 resulted in a consistent downregulation of BCL6 in the mRNA and protein levels compared with controls in Daudi-miR-150 and Raji-miR-150 (Fig. 3a,b). Conversely, re-expression of miR-150 induced PRDM1 and bcl-2 expression in Daudi-miR-150 and Raji-miR-150 (Fig. 3a,b). The results were further confirmed by confocal analysis in Raji-miR-150 (Fig. 3c). No obvious alterations were observed in BJAB-miR-150 and Ramos-miR-150 about expression of BCL6 and PRDM1 and bcl-2 compared with controls (data not shown).
Next, we used a flow cytometer to determine the immunophenotypic profiles of transduced cell lines. The maturation status of peripheral B-cells can be classified based on their expression of CD19, CD10, CD38, and CD138, following the pathway of GC cells (CD19+CD10+CD38+CD138−), to preplasmablasts or plasmablasts (CD19+CD10−CD38++CD138+/−), then to plasma cells (CD19−CD10−CD38+CD138+). Our results indicated that there was a significant increase in the percentage of CD19−CD10−CD138+ cells in Daudi-miR-150, and CD10−CD38++ cells in Raji-miR-150 compared with empty vectors (GFP) (Fig. 3d–f). No changes were observed in the immunophenotypic profiles of BJAB-miR-150 or Ramos-miR-150 compared with GFP (data not shown). Then, by analyzing SS and FS, the optical physical parameters of flow cytometry, we found that intracellular organelles were significantly increased in both Daudi-miR-150 and Raji-miR-150 compared with GFP (Fig. 3g,h). Furthermore, Raji-miR-150 showed increased cellular size (Fig. 3g,h).
C-Myb is a potential target of miR-150 in EBV-positive BL
Having shown that there was an inverse correlation between expression of miR-150 and c-Myb in mature B-cell subpopulations and human peripheral lymphoma cell lines, we reasoned that c-Myb may have potential to function as target of miR-150. To test this assumption, we detected the expression of c-Myb after miR-150 re-expression in BL cell lines. The results showed that c-Myb protein was downregulated in Daudi-miR-150 and Raji-miR-150 compared with controls (Fig. 4a,b). In terms of mRNA levels, c-Myb was downregulated in Raji-miR-150 whereas no obvious alteration was observed in Daudi-miR-150 compared with controls (Fig. 4a,b). The confocal analysis also indicated that c-Myb protein was reduced in Raji-miR-150 compared with GFP (Fig. 4c). No miR-150-mediated effect on c-Myb in BJAB-miR-150 and Ramos-miR-150 could be found.
To corroborate that the differentiation effect mediated by re-expression of miR-150 might be the consequence of c-Myb protein repression, we carried out transient transfections with inhibition of miR-150 in Daudi-miR-150 and Raji-miR-150 as well as c-Myb in Daudi and Raji cell lines. The results showed that the expression of c-Myb protein increased again compared with controls when miR-150 expression was inhibited in Daudi-miR-150 and Raji-miR-150 (Fig. 4d). The efficacy of c-Myb siRNA was confirmed by qRT-PCR and Western blot analysis (Fig. 4e). Given that miR-150 rescued EBV-positive BL by regulating c-Myb, tumor cells with downregulation of c-Myb protein levels would be predicted to recapitulate the partial phenocopies, similar to that caused by re-expression of miR-150. As expected, the c-Myb siRNA Daudi and Raji cells showed downregulation of BCL6 and upregulation of bcl-2 in the levels of mRNA and protein compared with controls (Fig. 4e). Although PRDM1 mRNA was induced in c-Myb siRNA Raji and Daudi cells, Western blot analysis showed that PRDM1 protein levels were only observed in c-Myb siRNA Daudi cells.
In this study, we investigated the importance of miR-150 in inducing BL differentiation and attempted to elucidate the underlying mechanism by which miR-150 functions in BL. Re-expression of miR-150 by modulating c-Myb partially drove plasmacytic differentiation of EBV-positive BL cell lines (Daudi and Raji), but this was not observed in EBV-negative BL cell lines (BJAB and Ramos). Daudi cells transduced with miR-150 lost GC markers BCL6 and CD10, even pan-B cell marker CD19, and acquired plasmacytic markers PRDM1 and CD138, as well as mature marker bcl-2, presenting with abundant intracellular organelles, all of which were consistent with the features of plasma cells. Raji cells showed downregulated BCL6 and CD10, upregulated PRDM1 and bcl-2, stronger CD38 expression, larger cellular size, and abundant intracellular organelles compared with controls, suggesting preplasmablasts or plasmablast differentiation.
A characteristic feature of leukemia or lymphoma cells is that they are blocked at a distinct stage in cellular maturation and fail to differentiate into functional mature cells. During the 1970s and 1980s, several scientific achievements, such as the differentiating capability of DMSO on erythropoiesis, controlling the differentiation of myeloid leukemia, and the differentiating properties of retinoic acid,[42, 43] popularized the strategy of inducing malignant cells to overcome their block of differentiation and enter the apoptotic pathways. This intervention could theoretically limit exposure to unwanted side-effects of cytotoxic chemotherapy and, more importantly, improve complete remission and cure rates. In recent years, the advent of miRNAs holds much promise for new therapeutic strategies. The rationale for considering miRNAs as potential therapeutic targets is that miRNAs are expressed in a tissue-specific manner and dysregulated in tumor cells, as well as their ability to modulate hundreds of transcripts and to act as non-toxic differentiating agents. Differentiation induction by miRNAs has been successfully achieved in a few cases.[26-31] In this experiment, we found that BL cell lines expressed miR-150 in very low levels compared with normal B lymphocytes. In addition, restoration of miR-150 disadvantaged cell growth and induced tumor cells to differentiate into a more mature stage in EBV-positive BL cell lines.
However, the question is why re-expression of miR-150 only triggered differentiation in EBV-positive BL cell lines, but not in EBV-negative ones. A recent investigation suggested that EBV-positive and EBV-negative BL might differ in their cells of origin and probably in their pathogenetic mechanisms, and it seems that there are two B-cell differentiation stages to which BL could correspond: early centroblasts for EBV-negative BL; and a memory B-cell or a late GC B-cell for EBV-positive BL. CD77 has been used as a marker differentiating GC centrocytes from GC centroblasts, in that the latter express CD77 more strongly than the former. Using flow cytometric analysis, we observed a significantly greater frequency of CD77+ cells in BJAB and Ramos cell lines than that in Daudi or Raji cell lines (Fig. S3), which further supported the different origins of EBV-positive and EBV-negative BL. These may explain why Daudi and Raji cell lines, in which the differentiation toward a plasma cell has been initiated, can be triggered to differentiate into more mature stages in the set of re-expression of miR-150 whereas BJAB and Ramos cell lines cannot.
In terms of the targets of miR-150, different conclusions have been drawn in different cellular systems and different microenvironments.[17, 19, 20, 34, 46-49] This is not surprising, as each miRNA can have one or more target transcripts that exert diverse roles in a given cell type. Among the targets of miR-150, transcription factor c-Myb, a top predicted target of miR-150, has been well studied. However, the previous studies mostly focused in mice. Some differences have been found in expression patterns and their targets between human and mouse hematopoietic cells even for the same miRNA, and this should be taken into consideration when extrapolating results from research in mice to humans. Although in the present study we did not provide direct evidence that miR-150 exerted its roles in EBV-positive BL through suppression of c-Myb, this hypothesis was supported by the observed inverse expression pattern of miR-150 and c-Myb in B-cell subsets and human lymphoma cell lines as well as decreased expression of c-Myb protein in the cell lines transduced with miR-150. Importantly, c-Myb was induced again by downregulating miR-150 in Daudi-miR-150 and Raji-miR-150. Furthermore, partial ablation of c-Myb recapitulated some characteristics similar to that caused by re-expression of miR-150 in Daudi and Raji cell lines. The reason that c-Myb knockdown did not induce the expression of PRDM1 protein in Raji cells was probably due to an insufficient absolute amount of PRDM1 mRNA. The working style of miR-150 regulating c-Myb has been reported through control of the post-transcriptional levels.[17, 46, 49, 52] Our results also suggest that miR-150 modulates c-Myb expression post-transcriptionally by miRNA-mediated transcript decay (in Raji-miR-150) or inhibition of translation (in Daudi-miR-150). The explanation why miR-150 cannot modulate c-Myb expression in EBV-negative BL could be that the relationship between miR-150 and c-Myb depends on cellular systems and microenvironments. Another possible explanation is that viral miRNAs and other viral products in BL might interfere in the regulation of human mRNAs and miRNAs.
One of important mechanisms driving dysregulated changes in miRNAs could be c-Myc activation, which has been shown to repress a large number of miRNAs, including miR-150, in tumor development.[36, 54] In accordance with this, our results also suggest that downregulation of miR-150 is virtually due to repression by c-Myc (Fig. S4A,B). The relationship between c-Myc and the regulatory factors that control the differentiation of GC B-cells to plasma cells, such as BCL6 and PRDM1, has been reviewed in detail.[55, 56] As expected in our experiment, c-Myc protein was reduced by PRDM1-mediated repression in Daudi-miR-150 and Raji-miR-150 compared with controls (Fig. S4C).
In summary, by way of a feedback loop, re-expression of miR-150 induces EBV-positive BL differentiation towards a more mature stage by miR-150 and c-Myb regulating each other's expression (Fig. 5). This further broadens the significance of differentiation induction by miRNAs in lymphoma, especially in BL.
We thank the Hematology Department personnel of the Kingmed Medical Test Center (Guangzhou, China) for their assistance in flow cytometry. This work was supported by grants from the National Natural Science Foundation of China (Grant No. 81272634).