Alteration of MicroRNAs Regulated by c-Myc in Burkitt Lymphoma

Background Burkitt lymphoma (BL) is an aggressive B-cell lymphoma, with a characteristic clinical presentation, morphology and immunophenotype. Over the past years, the typical translocation t(8;14) and its variants have been considered the molecular hallmark of this tumor. However, BL cases with no detectable MYC rearrangement have been identified. Intriguingly, these cases express MYC at levels comparable with cases carrying the translocation. In normal cells c-Myc expression is tightly regulated through a complex feedback loop mechanism. In cancer, MYC is often dysregulated, commonly due to genomic abnormalities. It has recently emerged that this phenomenon may rely on an alteration of post-transcriptional regulation mediated by microRNAs (miRNAs), whose functional alterations are associated with neoplastic transformation. It is also emerging that c-Myc modulates miRNA expression, revealing an intriguing crosstalk between c-Myc and miRNAs. Principal Findings Here, we investigated the expression of miRNAs possibly regulated by c-Myc in BL cases positive or negative for the translocation. A common trend of miRNA expression, with the exception of hsa-miR-9*, was observed in all of the cases. Intriguingly, down-regulation of this miRNA seems to specifically identify a particular subset of BL cases, lacking MYC translocation. Here, we provided evidence that hsa-miR-9-1 gene is heavily methylated in those cases. Finally, we showed that hsa-miR-9* is able to modulate E2F1 and c-Myc expression. Conclusions Particularly, this study identifies hsa-miR-9* as potentially relevant for malignant transformation in BL cases with no detectable MYC translocation. Deregulation of hsa-miR-9* may therefore be useful as a diagnostic tool, suggesting it as a promising novel candidate for tumor cell marker.


Introduction
The c-Myc transcription factor is pathologically activated in many human malignancies [1]. A paradigm for c-Myc deregulation is offered by Burkitt Lymphoma (BL), where chromosomal translocations that join MYC with immunoglobulin (Ig) heavy-(Igh) or light-chain (Igk, Igl) loci are the crucial initiating oncogenic events [2]. High levels of c-MYC have been clearly shown to have a tumour-promoting effect [3]. Just a 2-fold difference in c-Myc expression can affect cell size in flies or cell number in mice [4][5][6][7].
However, there is increasing evidence that less than 10% of classical BL cases lack an identifiable MYC rearrangement [8][9][10]. Interestingly, no significant difference of MYC expression between MYC translocation-positive and negative cases has been found, independently of genomic alterations [10]. This may suggest that additional mechanisms, alternative to chromosomal translocations, which may result in MYC deregulation also exist. c-Myc expression is strictly regulated by a feedback loop autoregulatory mechanism involving the transcription factor E2F1, whose loss impairs MYCmediated proliferation and lymphomagenesis [11]. c-Myc expression can be also regulated at the level of post-transcriptional control from a new class of non-coding RNA, miRNAs, able to inhibit mRNAs from being translated or causing them to be degraded [12].
MiRNAs play major roles in crucial processes as proliferation, differentiation and cell death [13]. It has become clear that alterations in the expression of miRNAs contribute to the pathogenesis of most, if not all, human malignancies [14]. MiRNA deregulation can be caused by various mechanisms, either genetic or epigenetic, including deletions, amplifications or mutations involving miRNA loci, epigenetic silencing or deregulation of transcription factors that target specific miRNAs [14].
In a previous study, we have analyzed a set of BL cases lacking MYC translocation, in which no other genomic aberrations, as increase of MYC copy number or aneuploidy were present, which showed high levels of MYC expression. We searched for alternative molecular alterations responsible for c-MYC deregulation in these cases and observed an altered expression of a specific miRNA, hsamir-34b, predicted to regulate MYC [10]. Being MYC a specific target of this miRNA, its deregulation may explain MYC altered expression in these cases [10].
In this study, we aimed at analyzing the expression of these specific miRNAs regulated by c-Myc in the previously described set of BL cases, based on the existence of a regulatory loop linking c-Myc and specific miRNAs. Our results show that a single miRNA, hsa-miR-9*, was found differentially expressed between BL cases carrying or not MYC translocation, being significantly down-regulated only in MYC translocation-negative cases. Intriguingly, we provide evidence that hsa-miR-9* is able to modulate E2F1 and c-Myc expression, suggesting down-regulation of hsa-miR-9* as a possible mechanism of c-Myc over-expression in BL cases negative for the translocation.
In summary, a better knowledge of miRNA alteration in such cases can potentially provide new markers to improve diagnosis and prognosis, as well as novel therapeutic approaches for BL treatment.

Ethics Statement
Ethics approval for this study was obtained from the Institutional Review Board at the University of Siena, University of Nairoby and at the CNIO. Informed written consent was obtained in all cases.

Cases selection and Immunophenotype
For this study BL specimens collected at the Department of Pathology, Nairobi Hospital, Kenya and the Department of Human Pathology and Oncology, University of Siena, Italy, with available remaining materials from our prior study, have been used [10]. Since MYC translocation-negative BL cases may represent a challenging diagnosis to discriminate them from DLBCL and Intermediate DLBCL/BL, additional formalin-fixed and paraffin-embedded (FFPE) specimens of MYC translocationnegative Diffuse Large B cell Lymphoma (DLBCL), Intermediate DLBCL/BL and BLs have been retrieved at the Department of Human Pathology and Oncology, University of Siena, Italy, at the Molecular Pathology Programme, CNIO, Spain and at the Aga Khan University, Nairobi, Kenya. Attention was given to select cases with immunophenotype suggestive of BL (CD10+, BCL6+, BCL2-or weakly positive). Cases were reviewed by expert pathologists (BC, LL) and diagnoses were confirmed by morphology on histological slides stained with HE, Giemsa and by immunophenotyping, according to the Word Health Organization (WHO) classification which reports that MYC translocationnegative BL cases must be otherwise completely typical to make a diagnosis of BL [19].
Immunohistochemical studies were performed on representative paraffin sections from each case using microwave pre-treatment of slides for antigen retrieval. A large panel of antibodies recognizing formalin-resistant epitopes of the various antigens was applied, in conjunction with the alkaline phosphatase antialkaline phosphatase (APAAP) method, to visualize antibody binding [20]. Reactive lymph nodes were used as a control.
The presence of the Epstein-Barr virus (EBV) was assessed by in situ hybridization for EBERs using Epstein-Barr Virus (EBER) PNA/Fluorescein (DAKO, Denmark), a mixture of PNA probes complementary to the two nuclear EBER RNAs encoded by EBV, in conjunction with DAKO PNA ISH Detection Kit (DAKO, Denmark). Five micrometer-thick paraffin sections were deparaffinized, rehydrated and processed according to manufacturer's instructions. A control slide, prepared from a paraffin-embedded tissue block containing metastatic nasopharyngeal carcinoma in a lymph node, accompanied each hybridization run.

Fluorescence in situ hybridization (FISH)
FISH analysis was performed following standard protocols used during EUROFISH and available at www.euro-fish.org. All reagents, instruments and split-signal probes were kindly provided by DakoCytomation (Glostrup, Denmark). Briefly, MYC rearrangements were sought using the MYC FISH DNA Probe-Split Signal using standard procedures. FISH was carried out as previously reported [10]. Dual color fluorescent in situ hybridization (FISH) for the MYC gene and the chromosome 8 centromere (CEN8), interpretation of results, classification of increased or not increased MYC gene copy number was carried out as previously described [10].

MiRNA extraction
FFPE sections of primary tumors and reactive lymph nodes as a control were treated with Xylene to eliminate paraffin and then miRNAs were extracted using TRIZOL (Invitrogen, Milan, Italy). Germinal centre cells were isolated by laser capture microdissection and miRNAs were then extracted using TRIZOL. DNase I (Promega, Milan, Italy) treatment was performed in all samples.

Analysis of miRNA expression
MiRNA expression was analyzed by qRT-PCR. Briefly, RNA samples were reverse transcribed using the Taqman MicroRNA reverse transcription kit (Applied Biosystems, Applera, Italy) and primers specific for each miRNA (hsa-miR-17-5p, hsa-miR-20a, hsa-miR-9, hsa-miR-9*), according to manufacturer's instructions. For each sample, 10 ng of total RNA were reverse transcribed. Real-time PCR was performed using Taqman probes specific for each miRNA and for RNU43, used as an endogenous control (Applied Biosystems, Applera, Italy). Amount and quality of RNA were evaluated measuring the OD at 260 nm, the 260/230 and the 260/280 ratios by Nanodrop (Celbio, Italy). Quality of RNA was also checked by BioAnalyzer (Agilent, CA).

Gene expression analysis
Relative quantification of gene expression for E2F1 and MYC was also carried out by Real-time PCR using FluoCycle SYBR green (Euroclone, Celbio, Italy) according to manufacturer's instructions.

DNA extraction and BSP
FFPE section (10 mM) of 8 cases of endemic BL, negative for MYC translocation, and two reactive lymph nodes were deparaffinized with Xylene and DNA extraction was performed with NucleoSpin kit (Macherey-Nagel), according to manufacturer's instructions.
PCR products were separated on a 2% agarose gel to confirm size, quantity and purity of each product, and then purified using NucleoSpin Kit (Macherey-Nagel), according to manufacturer's instructions. 2 ml aliquot of purified PCR products were cycle sequenced using Big Dye Terminator Mix (Applied Biosystem, Foster City, CA) in a total volume of 20 ml. Reaction products were purified with ethanol/MgCl 2 and finally 5 ml of product were resuspended in 15 ml of ultra pure water before being loaded on ABI capillary sequencer. Sequence reads were base-called using Chromas program. To explore in silico bisulfite modified DNA (either or not methylated at its CpG dinucleotides) methBLAST was used (http://medgen.ugent.be/methBLAST/), a sequence similarity search program.

Clinical, morphological features and immunophenotype
BL, DLBCL and Intermediate BL/DLBCL cases were diagnosed based on morphology, immunophenotype and genetic alterations, in accordance to the WHO classification [19]. The clinical and pathological data of these cases are summarized in the Table 1.
c-Myc-induced miRNAs: hsa-miR-17-5p, hsa-miR-20a, hsa-miR-9 and hsa-miR-9* The expression of hsa-miR-17-5p, hsa-miR-20a, hsa-miR-9 and hsa-miR-9*, reported to be regulated by c-Myc, was checked in all of BL cases. As expected, a strong up-regulation of hsa-miR-17-5p and hsa-miR-20a was found in all of the cases, independently of MYC translocation, which is in line with high expression of MYC in these cases (data not shown). Hsa-miR-9 showed an heterogeneous expression among all of BL cases, with no significant difference between MYC translocation-positive and negative BL cases (data not shown).
Interestingly, the expression of hsa-miR-9* was significantly different between MYC translocation-positive and negative cases, compared to normal controls (reactive lymph nodes and GC cells), being over-expressed in cases carrying MYC translocation (10/10) and strongly down-regulated in the others (8/9) ( Figure 1A).
Of note, one case negative for the translocation expressed hsa-miR-9* at high levels. It is worth noting that among the cases lacking MYC translocation, this was the only case positive for EBV, and occurred in an elderly patient ( Figure 1A).
The absence of MYC translocation in BL cases may represent a challenge in the diagnosis of BL, to discriminate them from DLBCL and cases with intermediate features between DLBCL and BL cases (DLBCL/BL). To further confirm a role of hsa-miR-9* as a possible molecular marker for MYC translocation-negative BL, we tested its expression in DLBCL GC-like and Intermediate DLBCL/BL MYC translocation-negative cases. Hsa-miR-9* resulted over-expressed in all of DLBCL cases (10/10). As expected, Intermediate DLBCL/BL cases showed an heterogeneous expression of hsa-miR-9*. In particular, hsa-miR-9* was over-expressed in 4/6 cases, whereas 2/6 showed low expression, similarly to BL cases lacking MYC translocation (Figure 2A). The median fold changes has been used as an indicator of the expression level in each category of lymphomas. To group the median fold changes, cut-offs have been defined as low (median fold changes ,0.5), medium (median fold changes between 0.5 and 1.5), and high (median fold changes .1.5) expression level. doi:10.1371/journal.pone.0012960.t002

E2F1 expression in primary tumors
We then investigated the molecular mechanism underlying such a difference in hsa-miR-9* expression. Using valid on-line database, we searched for mRNAs predicted to be hsa-miR-9* targets. Among those predicted, we found E2F1 as a possible intriguing candidate, because it is known to controls c-Myc expression [27,28]. First, we tested its expression in MYC translocation-negative BL cases in comparison to BL carrying . Functional in vitro study on hsa-miR-9*. (A) Relative expression by qRT-PCR of hsa-mir-9* in LCL transfected with increasing amounts of a synthetic hsa-mir-9*, compared to non-treated LCL and LCL transfected with a negative control (NC). The level of hsamir-9* increased in a dose-dependent manner. (B) Relative expression by qRT-PCR of hsa-mir-9* in LCL transfected with an hsa-miR-9* inhibitor, at the concentration of 50 nM, compared to non-treated LCL and LCL trasfected with a negative control (I NC). The level of hsa-mir-9* decreased significantly upon transfection of the inhibitor. (C) Relative quantification of E2F1 and MYC transcripts after transfection of increasing amounts of a synthetic hsa-mir-9*. A dose-dependent decrease of both genes is detected in the presence of synthetic hsa-mir-9*. (D) Relative quantification of E2F1 and MYC transcripts after transfection of a hsa-mir-9* inhibitor. Transfection of the hsa-mir-9* inhibitor is able to increase E2F1 and MYC expression, in comparison with the expression level of both proteins in non-treated LCL or cells transfected with a negative control (I NC). (E) E2F1 and c-Myc protein levels in LCL transfected with higher concentration of synthetic hsa-mir-9*, hsa-mir-9* inhibitor and their negative controls, evaluated by western blotting. MYC translocation and normal controls (reactive lymph nodes and GC cells). E2F1 was up-regulated only in BL cases lacking MYC translocation, in respect with both BL cases carrying MYC translocation and normal controls ( Figure 1B). To further confirm these results, we tested its expression also in DLBCL and Intermediate DLBCL/BL cases ( Figure 2B), where an inverse correlation between hsa-miR-9* and E2F1 was also found. Table 2 summarizes the level of the expression of hsa-miR-9* and E2F1 with the presence or not of MYC translocation in the different categories of analysed lymphomas.
E2F1 over-expression observed in MYC translocation-negative BL cases could be responsible for c-Myc over-expression, and to sustain it at high levels in the absence of translocation. This event may be relevant for malignant transformation, because of the reciprocal activation E2F1/c-Myc.
Functional in vitro study on hsa-miR-9* To confirm the bioinformatic predictions which indicate E2F1 as a possible target of hsa-miR-9* (Figure 3), we transfected a synthetic hsa-mir-9* ( Figure 4A) or its inhibitor ( Figure 4B) in a human B lymphoblastoid cell line, and then measured E2F1 expression both at the mRNA and protein levels by qRT-PCR and western blotting, respectively. We found a significant dosedependent decrease in E2F1 mRNA in cells transfected with hsamir-9* ( Figure 4C). Conversely, when a hsa-mir-9* inhibitor was transfected, an increase in E2F1 mRNA was observed ( Figure 4D).
As c-Myc expression can be induced by E2F1, we also checked MYC mRNA expression. Accordingly, MYC mRNA levels resulted downregulated after transfection of a synthetic hsa-miR-9* ( Figure 4C). On the other hand, we observed an increase of MYC mRNA expression after silencing of hsa-miR-9* ( Figure 4D).These findings were confirmed at the protein level, by western blotting ( Figure 4E).

Methylation analysis of hsa-miR-9-1 gene in primary tumors
Hypermethylation is described as possible mechanism of miRNA silencing. Due to the presence of a CpG island in hsa-miR-9-1 gene (Ensemble Genome browser, ENSG00000207933), we performed methylation analysis by bisulfite sequencing to assess whether down-regulation of hsa-miR-9-1 in those cases could rely on an epigenetic inactivation. Figure 5 shows the results, indicating hypermethylation in BL cases negative for MYC translocation, in respect with normal controls.

Discussion
The molecular feature of Burkitt lymphoma (BL) is the translocation that places MYC under the control of immunoglobulin gene regulatory elements. However, there is recent evidence that infrequent cases may lack an identifiable MYC translocation, the explanation for which is still uncertain, though suggesting the existence of pathogenetic mechanisms alternative to genetic alterations [8,9].
Recently, it has been suggested that the over-expression of MYC in the absence of translocation might be due to miRNA deregulation [10]. It has been recently demonstrated that hsa-miR-34b is expressed at different instances between MYC translocation-negative and positive BL cases [10]. In particular, it is down-regulated only in MYC translocation-negative BL cases, giving evidence that its deregulation could influence c-Myc expression in the absence of genomic translocation. On the other hand, c-Myc itself is able to activate the expression of specific miRNAs [12][13][14][15], suggesting the existence of a feedback loop between c-Myc and specific miRNAs, able to reciprocally control their expression. It may then be speculated that c-Myc overexpression induces a specific miRNA pattern that, in turn, might be responsible for a differential gene expression, and for functional alterations of tumor cells.
Here, we found a strong up-regulation of hsa-miR-17-5p and hsa-miR-20a, which correlates with high levels of MYC expression in these tumors, and is in line with the c-Myc-mediated induction of the miR-17-92 cluster. An heterogeneous expression of hsa-miR-9 was observed among BL cases, whereas hsa-miR-9* was the only miRNA strongly down-regulated only in BL MYC translocation-negative cases, with the exception of one case. Interestingly, this case was the only EBV-positive among the MYC translocationnegative BL cases. It might be argued that the virus may be able to modulate hsa-miR-9* expression in the absence of translocation, being now clear the existence of a delicate and complex interplay between viruses and cellular miRNAs [29][30][31][32].
The MYC translocation-negative BL cases may represent a challenging diagnosis to discriminate them from DLBCL and from cases with intermediate features between DLBCL and BL cases (DLBCL/BL) [8,19,33,34]. The latter is an heterogeneous category that is not considered a distinct disease entity, but it is useful in allowing the classification of cases not meeting criteria for classical BL or DLBCL. The diagnosis should not be made in cases of morphologically typical DLBCL that have a MYC rearrangement, or in otherwise typical BL in which a MYC rearrangement cannot be demonstrated. Improvements in appropriate therapeutic approaches should be established for this category which as extremely poor prognosis.
To further confirm that the hsa-miR-9* may be considered specific a molecular marker for BL cases lacking the typical translocation, we tested its expression also in DLBCL and Intermediate DLBCL/BL cases. We found that DLBCL cases showed a strong over-expression of hsa-miR-9*, whereas Intermediate DLBCL/BL cases an heterogeneous expression. In fact, among the latter hsa-miR-9* was over-expressed in 4/6 and down-regulated in 2/6, similar to BL cases lacking MYC translocation. These findings further suggest that Intermediate DLBCL/BL cases are an heterogeneous category and additional studies are necessary to demonstrate whether hsa-miR-9* may be and empty circles (#) indicate that CpG site is methylated or unmethylated, respectively. The BL cases negative for MYC translocation showed aberrant (cases 1-5) and partial (cases 6 and 7) methylation, in respect with the normal controls (C1 and C2 used as a specific marker to differentiate BL from DLBCL, and to identify cases that may benefit from a more aggressive therapy.
To unravel the pathological role of hsa-miR-9* down-regulation on MYC translocation-negative BL cases, we looked at mRNAs predicted to be target of hsa-miR-9*. Among those predicted, we found particularly interesting E2F1, which is essential for the G1-S1 phase passage. Expression of E2F1 is known to be induced by c-Myc, and in turn controls c-Myc expression [27,28].
Interestingly, an inverse correlation between hsa-miR-9* and E2F1 was found. In particular, E2F1 over-expression observed in BL cases lacking the typical translocation may represent itself an appealing possible mechanism able to determine c-Myc overexpression and sustain it at high levels in the absence of translocation, due to the existence of a feedback loop responsible for controlling the reciprocal expression of these transcription factors. On the other hand, hsa-miR-9* up-regulation observed both in MYC translocation-positive BL, DLBCL and in 4/6 Intermediate BL/DLBCL cases may be consistent with a regulatory loop between hsa-miR-9* and E2F1. To verify our hypothesis, based on bioinformatic predictions, functional in vitro studies were performed. E2F1 regulation by hsa-miR-9* was observed, either direct or indirect, as miR9* ectopic up-regulation reduces E2F1 levels. Conversely, miR9* silencing induces E2F1 expression. As a consequence, also the expression of MYC was affected, both at the mRNA and protein levels. These findings highlight that hsa-miR-9* inactivation may determine E2F1 upregulation and consequent MYC over-expression in BL lacking MYC translocation.
MiRNA silencing in cancer can be due to several mechanism, as a consequence of genomic abnormalities or epigenetic regulation [35]. Hsa-miR-9* down-regulation in MYC translocation-negative cases is due to an epigenetic mechanism, as an aberrant methylation of hsa-miR-9-1 gene was observed in these cases. Accordingly, inactivation of hsa-miR-9-1 gene due to hypermethylation has been suggested as an additional mechanism for miRNA inactivation also in human breast cancer [26].
Overall, we can conclude that several mechanisms may explain MYC over-expression. In particular, miRNA deregulation may have an important role in the pathogenesis of BL cases negative for MYC translocation, and may provide a new intriguing molecular mechanism other than typical translocation in affecting MYC expression ( Figure 6).
Hsa-miR-9* down-regulation seems to be specific for BL cases lacking MYC translocation, and identifies hsa-miR-9* as a possible novel candidate for a more accurate clinical diagnosis, helping to design miRNA-based gene therapy, which could be the future tool for gene therapy.