Inhibition of miR-9 de-represses HuR and DICER1 and impairs Hodgkin lymphoma tumour outgrowth in vivo


MicroRNAs are important regulators of gene expression in normal development and disease. miR-9 is overexpressed in several cancer forms, including brain tumours, hepatocellular carcinomas, breast cancer and Hodgkin lymphoma (HL). Here we demonstrated a relevance for miR-9 in HL pathogenesis and identified two new targets Dicer1 and HuR. HL is characterized by a massive infiltration of immune cells and fibroblasts in the tumour, whereas malignant cells represent only 1% of the tumour mass. These infiltrates provide important survival and growth signals to the tumour cells, and several lines of evidence indicate that they are essential for the persistence of HL. We show that inhibition of miR-9 leads to derepression of DICER and HuR, which in turn results in a decrease in cytokine production by HL cells followed by an impaired ability to attract normal inflammatory cells. Finally, inhibition of miR-9 by a systemically delivered antimiR-9 in a xenograft model of HL increases the protein levels of HuR and DICER1 and results in decreased tumour outgrowth, confirming that miR-9 actively participates in HL pathogenesis and points to miR-9 as a potential therapeutic target.


Chronic inflammation is a well-recognized cause of cancer and up to 25% of all cancers are thought to be induced by either chronic infections or autoimmune diseases.1 Inflammation not only works as a tumour-promoting agent but also influences other steps of tumorigenesis by inducing DNA damage, angiogenesis, invasion and metastasis.2

Hodgkin lymphoma (HL) is one of the most frequent lymphomas in the western world3 and it can be divided in several subtypes based on the morphology, composition of the infiltrates and the phenotype of the immune cells. The nodular sclerosis subtype is the most common subtype of classical HL (cHL) and accounts for about 60–70% of the cHL cases, followed by mixed cellularity cHL accounting for 20–25%.

cHL is an unusual B-cell malignancy in which the tumour cells represent only 1% of the tumour bulk, whereas the vast majority of the cells are a mixture of infiltrating immune cells and fibroblasts actively attracted via chemokine secretion by the malignant cells, the so-called Hodgkin and Reed–Sternberg cells.3 Several lines of evidence indicate that the infiltrating cells are necessary for the survival of the HL cells by providing growth and survival signals, and protection from NK cells.3, 4

MicroRNAs (miRNAs) are small non-coding RNAs that regulate gene expression at the posttranscriptional level by base-pairing to target mRNAs and promoting transcript instability and/or translational repression.5 Mature miRNAs derive from long hairpin precursors that are sequentially processed by the RNase-III-type enzymes DROSHA and DICER1, respectively.5 miRNAs have been implicated in several physiological and pathological events and can act as both tumour suppressors and oncogenes.6

miR-9 is encoded by three different loci in the mouse and human genomes and its expression is altered in several cancer forms. Recently, Ma et al.7 demonstrated that miR-9 overexpression increases the metastatic potential of breast cancer cells through regulation of E-cadherin. Interestingly, miR-9 is induced in monocytes and neutrophils during the immune responses induced by LPS stimulation, and one of the validated miR-9 targets is NF-kb1, further supporting a role of this miRNA in the regulation of inflammation.8 Moreover, several groups have reported elevated levels of miR-9 in HL cell lines and primary HL cases.9, 10, 11, 12 Among these, Nie et al.12 identified the B-cell differentiation regulator PRDM1 as a miR-9 target.

Here we show that miR-9 directly targets DICER-1 and HuR in vitro and in vivo. Interestingly, although the function of DICER1 as a key enzyme in miRNA biogenesis is well characterized, recent reports have demonstrated that DICER acts as a haploinsufficent tumour suppressor in lung cancer and retinoblastoma.13, 14 On the other hand, HuR is a well-documented regulator of cytokine expression acting via direct binding to ARE sequence motifs. AREs are short AU-rich sequences in the 3′ untranslated region (UTR) of several mRNAs, especially cytokines and growth-related genes that affect mRNA stability and translation.15 The fate of these mRNAs depends on the protein complexes assembling at the ARE sequence motifs. Assembly of the ARE protein complex on cytokine mRNAs appears to be organized by HuR that promotes or inhibits translation depending on the interaction with other ARE-binding proteins.16

Here we report that miR-9 overexpression is important for pathological mechanisms underlying HL. We identify two novel direct targets, DICER1 and HuR, and we show that miR-9 affects the ability of HL cells to secrete cytokines and to attract normal blood cells in a HuR-dependent fashion. Finally, we provide evidence that pharmacological inhibition of miR-9 by tiny antimiR-917 in vivo results in derepression of DICER1 and HuR, and reduces the outgrowth of human HL tumours established in immunocompromised mice.


miR-9 inhibition modulates genes implicated in B-cell activation and inflammation

As miR-9 expression is upregulated in HL cells compared with germinal center (GC) B-cells,9 we sought to unveil an etiological role for miR-9 in HL. To identify genes and the pathways regulated by miR-9, we inhibited its expression in the L428 HL cell line using a seed-targeting tiny antimiR-917 and performed expression array analyses. Compared with cells treated with LNA control inhibitor, we identified 1145 genes that were deregulated (FC below −1.1 and above 1.1) in HL cells treated with the antimiR-9 (Supplementary Table 1). Gene ontology and network analyses revealed that the regulated genes could be divided into four major groups, namely ‘apoptosis’, ‘B- and T-cell activation’, ‘inflammation and cytokine signalling’ and ‘signal transduction’ (Supplementary Figure 1a). Array results were validated by qPCR for selected transcripts (Supplementary Figure 1b).

Surprisingly, we did not observe any enrichment for miR-9-binding sites in the 3′ UTRs of the upregulated transcripts in this experiment (Supplementary Figure 2). These results suggest that miR-9 could indirectly regulate several genes of potential importance for HL development.

Inhibition of miR-9 downregulates cytokine secretion

Many of the differentially regulated genes, such as NFAT5 (logFC=−0.24), STAT6 (logFC=−0.19), IRF5 (logFC=−0.277) and NFKBIZ (logFC=−0.22), are known to be important for cytokine signalling and production.18, 19 Given the importance of cytokines in the assembly of the HL environment, we tested the secretion of a panel of cytokines with relevance for HL development in L428 cells following miR-9 inhibition. Interestingly, enzyme-linked immunosorbent assay tests demonstrated significant reductions in IL-6 (P=0.0024), CCL-5 (P=0.0018) and TNF-alpha (P=0.0239) secretion, together with the complete absence of IL-5 (P=0.0009) in samples treated with antimiR-9 relative to the control (Figure 1). In contrast, no significant changes in GM-CSF secretion were observed (Figure 1). Similar results (Supplementary Figure 3a) that showed L428 to be derived from nodular sclerosis cHL were obtained in L540 cells. Thus, inhibition of miR-9 decreases the secretion of cytokines important for cHL development in the nodular sclerosis subtype, thereby influencing the assembly of the inflammatory environment.

Figure 1

miR-9 inhibition reduces cytokine secretion from L428 cells. Supernatants derived from 12.5 × 106 L428 HL cells transfected an LNA control or with an LNA-modified antimiR-9 were used to measure the amount of CCL-5, GM-CSF, IL-5, TNF-alpha and IL-6, respectively, by enzyme-linked immunosorbent assays (ELISAs). Error bars represent s.e. between three independent experiments. P-values were calculated using a t-test, *P<0.05, **P<0.01 and ***P<0.001. The asterisks refer to comparison between the treated samples with the control. Cytokine secretion values in pg/ml are listed in the table.

HuR and DICER1 are direct targets of miR-9

To identify possible regulators of cytokine secretion responsible for the effect of miR-9, we analysed the list of deregulated transcripts following miR-9 inhibition and noted that HuR (ELAVL1) was upregulated (0.24 logFC). Interestingly, the 3′ UTR of HuR harbours four putative miR-9-binding sites, two 8-mer and two 7A-mer sites, respectively (Supplementary Figure 4). In addition, we noticed that DICER1 was also regulated and it could also be a miR-9 target having a 7-mer seed match site (Supplementary Figure 4).

To demonstrate that miR-9 can directly target the HuR and DICER1 mRNAs, we employed an affinity purification approach in which low amounts of a biotin-tagged and thiouridine-labelled miR-9 were transfected into L428 cHL cells. Following in vivo cross-linking, miR-9 and its targets were purified using streptavidin beads. This method has previously been used to identify and validate targets for miR-10a, miR-34a and miR-191.20, 21, 22 Using qRT-PCR, we showed that biotin-tagged and thiouridine-labelled miR-9 can efficiently sequester endogenous HuR and DICER1 transcripts, in contrast to a similarly modified miR-34a control, thus implying that HuR and DICER1 are direct targets of miR-9 (Figures 2a and b). To functionally validate these interactions, we transfected luciferase 3′ UTR reporters for HuR and DICER1 into U87MG cells and assayed the ability of miR-9 to directly regulate these 3′ UTRs. The U87MG glioblastoma cell line is easier to transfect compared with HL cell lines, but have comparable amounts of miR-9 (Supplementary Figure 5S). In accordance with the results obtained using affinity purification, luciferase expression from these reporters was repressed following miR-9 overexpression but not by expression of a miR-191 control (Figures 2c and d).

Figure 2

(ad) HuR and DICER1 are direct targets of miR-9. The experiment was repeated at least three times. Error bars represent s.e. between three independent experiments. (a, b) Biotinylated miR-9 or miR-34a were transfected into L428 cells and their ability to pull-down HuR (a) or DICER1 (b) mRNAs were assessed using qPCR. Relative enrichment of HuR and DICER1 mRNA, respectively, was measured by qPCR, normalizing the results to the input. P-values were calculated using ANOVA test. (c, d) A luciferase reporter containing the HuR (c) or DICER1 (d) 3′ UTRs was transfected into U87MG cells together with synthetic miR-9 or miR-191 as a negative control. (e) Three different HL cell lines were transfected with antimiR-9 or with LNA scramble control. The expression of HuR and DICER1 was assessed by western blot analysis. PRDM5 was used as a loading control and the band intensities were measured using the ImageJ software ( Ratios between the targets and the loading control are indicated above each figure.

Conversely, transfection with the antimiR-9 LNA resulted in a significant increase in the luciferase activity of both reporters (Supplementary Figure 4b and c; pHuR-UTR P=0.011 and pDICER1-UTR P=0.017).

Importantly, we confirmed the regulation of the endogenous HuR and DICER1 proteins by miR-9 in three different HL cell lines using western blot, as demonstrated in Figure 2e (also see Supplementary Figure 5S) by using the miR-9 inhibitor. This resulted in marked derepression of both the targets at the protein level. Taken together, our data indicate that HuR and DICER1 are bona fide targets that are repressed in cHL cells by endogenous miR-9.

miR-9 inhibition leads to deregulation of genes with HuR-binding sites

Most of the genes deregulated, following miR-9 inhibition, in L428 cells do not contain putative miR-9-binding motifs (that is, matches to the miR-9 seed sequence). These changes in transcript levels are therefore likely secondary effects and we speculated that deregulation of HuR could, at least in part, be responsible for the modulation. With a known role in post-transcriptional regulation of inflammation-related genes16 HuR is a good candidate, so we scanned the list of genes deregulated following miR-9 inhibition for the presence of the consensus HuR-binding motif.23 Intriguingly, using the Asap software package,24 we found a highly significant overrepresentation of the HuR-binding motif among the deregulated transcripts compared to a group of transcript that were unaffected by miR-9 inhibition (P=7.56 × 10−4 for upregulated genes and P=3.66 × 10−3 for downregulated genes) (Table 1). Hence, the direct regulation of HuR by miR-9 is reflected in the transcriptome following miR-9 inhibition, suggesting that a large proportion of the effects seen in the arrays upon miR-9 inhibition were HuR-mediated.

Table 1 Enrichment of HuR-binding motifs in down and upregulated transcripts compared with the no-change dataset

The regulation of cytokines by miR-9 is HuR-dependent

HuR is a known regulator of inflammation by mediating the stability and the translation of mRNAs holding ARE motifs,16 moreover, all the cytokines in this study deregulated by miR-9 contain several repeats of the HuR-binding motif identified recently by PAR-CLIP25, 26 (Supplementary 3B). Because we found HuR to be a direct target of miR-9, we next analysed the role of HuR in the regulation of cytokines by miR-9.

For this purpose, we used a reporter plasmid in which the expression of luciferase is under control of the 3′ UTR of the IL-5 gene containing a class II ARE motif. As expected, transfection of antimiR-9 resulted in a repression of the reporter (P=0.039) (Figure 3a). Importantly, this effect was completely abolished by co-transfection with a HuR-specific siRNA, whereas inhibition of HuR alone mediated an increased expression of the reporter (P=0.032)(Figure 3a). A similar effect was demonstrated using the pGFP-TNFARE reporter, in which the enhanced green fluorescent protein expression is under control of the TNFalpha-ARE (Figure 3b). In addition, enzyme-linked immunosorbent assays on L428 supernatants transfected with a siRNA against HuR show upregulation of three out of the four cytokines (Figure 3c). Furthermore, the effect of HuR on cytokines can be reverted by co-transfection of antimiR-9 LNA (Figure 3c). We therefore conclude that miR-9 can regulate the expression of cytokine transcripts in a HuR-dependent manner.

Figure 3

(a, b) miR-9 modulates IL-5 and TNF-alpha expression via HuR. (a) A luciferase reporter containing the IL-5 3′-UTR was used to measure the effect of siHuR and antimiR-9 on cytokine production in HEK293 cells. antimiR-9 significantly reduced activity of the luciferase reporter, whereas siRNAs targeting HuR increased this activity. Transfection of both abolishes the effect of miR-9 inhibition. Error bars represent s.e.m. between three independent experiments. P-values were calculated by t-test. *P<0.05, ***P<0.001 and ns=not significant. If not differently indicated, the asterisks refer to comparison between the treated samples with the control. (b) An enhanced green fluorescent protein reporter under control of the TNF-ARE was used to measure the effect of HuR and/or miR-9 inhibition on cytokine production in HEK293 cells. Effects of miR-9 and/or HuR inhibition on TNF-ARE were detected using western blot for enhanced green fluorescent protein. PRDM5 was used as a loading control and intensity was analysed using the ImageJ software. Ratios between the targets and the loading control are indicated above each figure. (c) Effects of HuR-silencing (with or without antimiR-9 treatment) on IL5, IL6, TNF and CCL-5 were analysed in supernatants by ELISA. P-values were calculated by ANOVA: IL-6=ns; IL-5=0.0302; CCL-5=0.0368; and TNF=0.0186.

Chemo-attractant activity of L428 cells is inhibited following miR-9 inhibition

To functionally validate that cytokine reduction upon miR-9 silencing can influence the inflammatory environment in cHL, we sought to mimic in vitro the attraction exerted by cHL cells on normal cells. To achieve this aim, we tested the ability of L428 supernatants to attract normal blood cells when miR-9 and/or HuR were inhibited. As expected, supernatant from cells treated with antimiR-9 showed a significantly lower ability to attract normal blood cells in Boyden chamber migration assays (P=0.0071), whereas HuR-silencing resulted in a moderate increase in the number of migrating cells. Importantly, the miR-9 inhibitory phenotype was almost abolished following concomitant siHuR transfection (P=0.0013) (Figure 4). Hence, we found that miR-9 inhibition can reduce the pool of migrating blood cells with about 22% and this reduction is partially dependent on HuR.

Figure 4

(a, b) antimiR-9-mediated upregulation of HuR impairs the ability of HL cells to attract normal blood cells. The ability of L428 cells to attract normal cells with or without miR-9 inhibition was assayed in a Boyden chamber assay using the supernatant from L428 cells and normal blood cells (a). A reduced migration activity in presence of a miR-9 inhibitor was detected. Reduced migration is only partially dependent on HuR (and thus on cytokine secretion) (b). The graph is representative of three biological experiments. Error bars represent s.e. between three independent experiments. P-values were calculated using a t-test. *P<0.05, **P<0.01. If not differently indicated, the asterisks refer to comparison between the treated samples and the control.

Pharmacological inhibition of miR-9 in vivo decreases tumour growth

To ask whether miR-9 overexpression is relevant to the outgrowth of cHL tumours, we injected L428 cells in the severely immunocompromised NOG mouse strain previously shown to support cHL engraftment.27 NOG mice lack B- and T-cell immunity and have defects in NK, DC and complement functions.27 These mice develop large tumour masses in the liver and sporadically in the spleen at 30 days after intravenous injection of cHL cells. To assess the importance of miR-9, we allowed the tumours to establish for 2 weeks after which the mice were treated with saline, antimiR-9 or LNA scramble control compounds. Each treatment group consisted of 10 mice and the LNA compounds were delivered subcutaneously twice weekly for 4 weeks after which the animals were killed. All the mice developed liver tumours and the pathological analysis showed that these tumours were necrotic and composed of large to medium size cells with sporadic Hodgkin and Reed–Sternberg cells. Moreover, immunohistochemistry revealed that the tumours lack typical mature B cell markers (CD20) and expressed CD30, the activation marker characteristic of cHL (Supplementary Figure 6).

Analyses of the dissected livers demonstrated a significant increase in liver weight caused by the presence of multiple large tumour nodules (Figure 5a). Notably, although treatment with the control LNA at 25 mg/kg did not significantly affect tumour burden, as estimated from the overall liver weights, treatment of mice with the antimiR-9 at the same dose level significantly reduced the tumour burden (P=0.002) (Figure 5a). To further substantiate this finding, we quantified the tumour burden from liver sections and found a significant decrease (P=0.026) in the presence of tumour cells in livers from mice receiving high doses of the antimiR-9 compound relative to the LNA scramble control-treated group (Figure 5b). Moreover, we found that treatment with the antimiR-9 at 25 mg/kg induced a significant and consistent increase in the number and intensity of DICER1- and HuR- positive cells (P=0.001 and 0.0001, respectively) (Figures 5c–f) compared with mice treated with LNA scramble control. Taken together, our data show that HuR and DICER1 are targeted by miR-9 in vivo during cHL tumorigenesis and that pharmacological inhibition of miR-9 by a seed-targeting tiny antimiR leads to significantly reduced tumour outgrowth in a mouse model of cHL.

Figure 5

(af) Silencing of miR-9 by antimiR-9 reduces HL tumour growth in NOG mice. The effects of miR-9 inhibition in vivo were tested on NOG mice injected i.v. with 4 × 106 L428 cells. Treatment of the mice for 4 weeks with antimiR-9 decreased the liver weight in a dose-dependent manner compared with saline vehicle control and LNA scramble-treated mice (a). P-values were calculated using ANOVA (P<0.0001) with Bonferroni correction. *P<0.05 and ***P<0.001. If not differently indicated, the asterisks refer to comparison between the treated samples with the control. The decrease in liver weight is the result of the impaired tumour growth in antimiR-9-treated mice (P<0.05) (b). The nodule surface was calculated using the grid method. Statistical significance was calculated using t-test. The expression of the validated miR-9 targets DICER1 (c) and HuR (e) was analysed by immunohistochemistry in samples from NOG mice injected with L428 HL cells (magnification × 20). The percentage of DICER1- and HuR- positive tumour cells (d and f, respectively), and the intensity of the staining was evaluated. Intensity was set to 1 for not or low expression level, or to 2 for high expression level. Statistical significance was calculated by two-way ANOVA.


In the present study, we identified two novel targets of miR-9 and validated them both in vitro and in vivo. In addition, we showed that miR-9 influenced cytokine secretion and chemo-attractant ability of cHL cells in vitro and impaired tumour outgrowth in vivo.

It is becoming increasingly clear that miRNAs have important roles in cancer-related processes.28 Apart from cancer-associated deregulation of individual miRNAs, genome-wide studies have demonstrated a general downregulation of mature miRNAs in cancers.29 miRNA downregulation is likely a cause rather than a by-product of cancer, because lowered miRNA levels have been found to increase the transforming capacity and the invasive behaviour of a lung adenocarcinoma cell line.30 Furthermore, hemizygous deletion of DICER1 occurs in 27% of the human cancers,14 and recent studies have demonstrated that DICER1 is a haploinsufficent tumour suppressor in both lung cancer and retinoblastoma.13, 14 This is supported by a recent report, in which DICER was shown to be specifically targeted by the miR-103/107 family in breast cancer, where overexpression of miR-103/107 members induced epithelial-to-mesenchymal transition via reduced levels of miR-200b.31 Therefore, direct regulation of DICER1 levels by miR-9 overexpression could contribute to transformation and invasiveness in HL.

cHL is characterized by the presence of an inflammatory environment that is essential for the pathogenesis of this disease.3 Inflammatory cells are actively attracted in the tumour by cytokine secretion, and in turn provide protection and survival signals to the malignant cells.3 To ensure a rapid switch in cytokine production, most cytokine-coding genes share an AU-rich sequence (ARE) in their 3′ UTR, which guides their rapid degradation or stabilization.15, 32 HuR is a ubiquitous member of the Hu family of proteins33 that can regulate mRNA stability and translation by binding to ARE-motifs and interacting with other ARE-binding proteins, such as TTP, AUF1 and KH-type splicing regulatory protein.34, 35 Despite the fact that HuR is often reported as a stabilizing factor for ARE-containing mRNAs, its ability to synergize with TIA-1 and to inhibit translation of cytokines is also well known.36

In the present study, we show that miR-9 directly targets HuR and Dicer1 for posttranscriptional repression. Consequently, inhibition of miR-9 by a seed-targeting antimiR-9 in cultured cells and in HL xenograft tumours resulted in upregulation of both the targets. In particular, derepression of HuR results in change in a strong decrease in cytokine secretion as demonstrated by enzyme-linked immunosorbent assays in cell supernatants. Notably, we observed reduction in the cytokines TNFalpha, CCL-5, IL-6 and IL-5 that have well-established roles in HL pathogenesis, where they are known to attract fibroblasts, eosinophils and mast cells to the tumour.37, 38, 39 Regulation of the cytokines is likely a posttranscriptional event, because a siRNA construct targeting HuR could rescue the effect of miR-9 inhibition in enzyme-linked immunosorbent assay on L428 supernatants and on reporters harbouring IL-5 and TNF-alpha ARE sequences. Moreover, miR-9 inhibition in HL cells resulted in the failure to efficiently attract normal inflammatory cells and this is partially due to the effect of miR-9 on HuR. Hence, miR-9 overexpression supports the presence of infiltrating inflammatory cells in the tumours.

Finally, to ask whether pharmacological inhibition of miR-9 could affect the outgrowth of cHL cells in vivo, we injected NOG mice with established xenograft HL tumours27 with a saline-formulated antimiR-9 compound. Several previous studies have suggested that individual miRNAs could constitute targets for therapeutic intervention of cancers.40 However, reports that demonstrate an effect upon systemic delivery of miRNA inhibitors on outgrown tumours are still limited.41 Although xenograft models for cHL rely on the use of severely immunocompromised mouse strains, such as the NOG mouse model used here, and therefore cannot fully recapitulate cHL pathogenesis, our data show that therapeutic silencing of miR-9 by subcutaneously delivered antimiR-9 inhibitors significantly limits the outgrowth of HL tumour nodules in the livers of NOG mice. We speculate that the effects seen on tumour growth probably result from the combined miR-9 repression of several different targets, including NF-kb and PRDM1,8, 12 both have a well-established role in B-cell development and pathogenesis.

In summary, our findings support the idea that miR-9 has an important role in the aetiology of cHL and highlight miR-9 as a potential therapeutic target for treatment of cHL.

Materials and methods

Cell lines and transfection

The cHL cell lines L428, L540 and KM-H2 were grown at standard conditions in 5% CO2 at 37 °C in RPMI 1640-glutamax (Gibco, Invitrogen, Paisley, UK) supplemented with 10% FBS (Hyclone, Thermo Scientific, Waltham, MA, USA).

For the array experiments, and to produce the supernatants for the migration assays, the L428 were transfected by nucleofection using solution L and program X-001 (Amaxa, Basel, Switzerland) and Nucleofector II (Amaxa). Cells were harvested at 9 h after transfection and RNA was extracted using Trizol (Invitrogen). For all the other experiments, the cells were transfected via passive perfusion42 by adding the antimiR-9 or LNA scramble control oligonucleotides directly to the culture medium at a final concentration of 10 μM and leaving the cells for 24–72 h in the medium. Subsequently, the cells were washed extensively in PBS and transfection efficiencies were checked by flow cytometry (Supplementary Figure 7) and microscopy of a parallel cell population treated with FITC-labelled LNA oligonucleotide.

U87MG cells used in the Luciferase assay were grown in DMEM-glutamax (Gibco, Invitrogen), supplemented with 10%FBS (Hyclone, Thermo Scientific).

The HEK cells used in the reporter assays were grown in DMEM-glutamax (Gibco, Invitrogen), supplemented with 10% FBS (Hyclone, Thermo Scientific). For the experiments combining miR-9 and HuR inhibition, the cells were transfected with either 50 μM scrambled siRNA or siHuR, together with antimiR-9 or LNA scramble control oligonucleotide (30 μM) using lipofectamine 2000 (Invitrogen). The pIL5-UTR Luciferase reporter was transfected 24 h later. Luciferase assay was performed 48 h after the first transfection.

Data processing of microarray profiles

For the transcriptome analyses, three biological replicates of total RNA extracted from cells treated with either antimiR-9 or LNA scramble control were analysed on Affymetrix HG-U133 Plus 2.0 arrays (Affymetrix, Santa Clara, CA, USA). The data were processed using the BioConductor ‘affy’ package.43 Probe set intensities were summarized using the Robust Multichip Average method, transformed to the generalized log values with the variance stable VSN method44 and mapped to Ensembl transcripts using BioMart. Probe sets that mapped to more than two different Ensembl genes were discarded. For genes with more than one probeset mapped to it, the probeset with the largest inter quartile range of expression intensity was selected. Nonspecific filtering was used to remove genes with low variance between arrays using a cut-off of 0.25. This left 1763 genes that were used for the following analysis. Differentially expressed genes were identified using limma.45 Genes with a fold-change above 1.1 or below −1.1 were regarded as upregulated and downregulated, respectively. A third set of genes that did not change from control to experiment was defined by selecting genes with a log2 fold-change centred at 0. The microarray data are MIAME compliant and were deposited in the GEO database with the accession no. GSE27529. The microarray results were validated using qPCR for three of the genes upregulated in the arrays (Supplementary Figure 2).

HuR motif search

The 3′ UTR sequences of the upregulated, downregulated and no-change transcript sets were used to search for the previously published HuR-binding motif.23 The count matrix was log-transformed into a position-specific weight matrix using a background nucleotide frequency of 25% and pseudo count of 1. The search for matches to this position-specific weight matrix was performed with the Asap software package24 using a threshold of 0.8. As control, the HuR motif was shuffled 100 times. The percentage of sequences with hits for the HuR motif and the shuffled motif sites was compared between the down and the no-change sets or between the up and no-change sets. The P-values were calculated testing the null –hypothesis, which showed that the proportion of sequences with hits was the same in the two compared gene sets.

cHL xenograft model in NOG mice

Treatment of mice was in compliance with: (a) institutional guidelines; (b) the Guide for the Care and Use of Laboratory Animals (National Academy of Sciences, 1996); and (c) the Association for Assessment and Accreditation of Laboratory Animal Care International. In addition, specific consent was obtained from the Central Ethical Committee of the Danish Government for all animal experiments. NOG mice (Taconic, Hudson, NY, USA) were injected intravenously with 4 × 106 L428 cells. After 2 weeks, the mice were treated twice weekly with saline, or with saline-formulated LNA scramble control (25 mg/kg) or tiny seed-targeting LNA-antimiR-9 (5 or 25 mg/kg dose levels). Each treatment group contained 10 mice.

The mice were killed 4 weeks after the first LNA injection. Livers were dissected and weighted.

Accession codes


Gene Expression Omnibus


  1. 1

    Schetter AJ, Nguyen GH, Bowman ED, Mathe EA, Yuen ST, Hawkes JE et al. Association of inflammation-related and microRNA gene expression with cancer-specific mortality of colon adenocarcinoma. Clin Cancer Res 2009; 15: 5878–5887.

  2. 2

    Grivennikov SI, Greten FR, Karin M . Immunity, inflammation, and cancer. Cell Mol Life Sci [Review] 2010; 140: 883–899.

  3. 3

    Kuppers R . The biology of Hodgkin's lymphoma. Nat Rev Cancer 2009; 9: 15–27.

  4. 4

    Brauninger A, Schmitz R, Bechtel D, Renne C, Hansmann ML, Kuppers R . Molecular biology of Hodgkin's and Reed/Sternberg cells in Hodgkin's lymphoma. Int J Cancer 2006; 118: 1853–1861.

  5. 5

    Krol J, Loedige I, Filipowicz W . The widespread regulation of microRNA biogenesis, function and decay. Nat Rev Genet 2010; 11: 597–610.

  6. 6

    Esquela-Kerscher A, Slack FJ . Oncomirs—microRNAs with a role in cancer. Nat Rev Cancer 2006; 6: 259–269.

  7. 7

    Ma L, Young J, Prabhala H, Pan E, Mestdagh P, Muth D et al. miR-9, a MYC/MYCN-activated microRNA, regulates E-cadherin and cancer metastasis. Nat Cell Biol 2010; 12: 247–256.

  8. 8

    Bazzoni F, Rossato M, Fabbri M, Gaudiosi D, Mirolo M, Mori L et al. Induction and regulatory function of miR-9 in human monocytes and neutrophils exposed to proinflammatory signals. Proc Natl Acad Sci USA 2009; 106: 5282–5287.

  9. 9

    Lawrie CH, Saunders NJ, Soneji S, Palazzo S, Dunlop HM, Cooper CD et al. MicroRNA expression in lymphocyte development and malignancy. Leukemia 2008; 22: 1440–1446.

  10. 10

    Van Vlierberghe P, De Weer A, Mestdagh P, Feys T, De Preter K, De Paepe P et al. Comparison of miRNA profiles of microdissected Hodgkin/Reed-Sternberg cells and Hodgkin cell lines versus CD77+ B-cells reveals a distinct subset of differentially expressed miRNAs. Br J Haematol 2009; 147: 686–690.

  11. 11

    Navarro A, Gaya A, Martinez A, Urbano-Ispizua A, Pons A, Balague O et al. MicroRNA expression profiling in classic Hodgkin lymphoma. Blood 2008; 111: 2825–2832.

  12. 12

    Nie K, Gomez M, Landgraf P, Garcia JF, Liu Y, Tan LH et al. MicroRNA-mediated down-regulation of PRDM1/Blimp-1 in Hodgkin/Reed-Sternberg cells: a potential pathogenetic lesion in Hodgkin lymphomas. Am J Pathol 2008; 173: 242–252.

  13. 13

    Lambertz I, Nittner D, Mestdagh P, Denecker G, Vandesompele J, Dyer MA et al. Monoallelic but not biallelic loss of Dicer1 promotes tumorigenesis in vivo Dicer1 is a haploinsufficient tumor suppressor. Cell Death Differ 2010; 17: 8.

  14. 14

    Kumar MS, Pester RE, Chen CY, Lane K, Chin C, Lu J et al. Dicer1 functions as a haploinsufficient tumor suppressor. Genes Dev 2009; 23: 4.

  15. 15

    Khabar KS . Post-transcriptional control during chronic inflammation and cancer: a focus on AU-rich elements. Cell Mol Life Sci 2010; 67: 2937–2955.

  16. 16

    Anderson P . Post-transcriptional control of cytokine production. Nat Immunol 2008; 9: 353–359.

  17. 17

    Obad S, dos Santos CO, Petri A, Heidenblad M, Broom O, Ruse C et al. Silencing of microRNA families by seed-targeting tiny LNAs. Nat Genet 2011; 43: 371–378.

  18. 18

    Lopez-Rodriguez C, Aramburu J, Jin L, Rakeman AS, Michino M, Rao A . Bridging the NFAT and NF-kappaB families: NFAT5 dimerization regulates cytokine gene transcription in response to osmotic stress. Immunity 2001; 15: 47–58.

  19. 19

    Medzhitov R, Horng T . Transcriptional control of the inflammatory response. Nat Rev Immunol 2009; 9: 692–703.

  20. 20

    Ørom UA NF, Lund AH . MicroRNA-10a binds the 5′UTR of ribosomal protein mRNAs and enhances their translation. Mol Cell 2008; 30: 11.

  21. 21

    Christoffersen NR, Shalgi R, Frankel LB, Leucci E, Lees M, Klausen M et al. p53-independent upregulation of miR-34a during oncogene-induced senescence represses MYC. Cell Death Differ 2010; 17: 236–245.

  22. 22

    Wynendaele J, Bohnke A, Leucci E, Nielsen SJ, Lambertz I, Hammer S et al. An illegitimate microRNA target site within the 3′ UTR of MDM4 affects ovarian cancer progression and chemosensitivity. Cancer Res 2010; 70: 9641–9649.

  23. 23

    Lopez de Silanes I, Zhan M, Lal A, Yang X, Gorospe M . Identification of a target RNA motif for RNA-binding protein HuR. Proc Natl Acad Sci USA 2004; 101: 2987–2992.

  24. 24

    Marstrand TT, Frellsen J, Moltke I, Thiim M, Valen E, Retelska D et al. Asap: a framework for over-representation statistics for transcription factor binding sites. PLoS One 2008; 3: e1623.

  25. 25

    Mukherjee N, Corcoran DL, Nusbaum JD, Reid DW, Georgiev S, Hafner M et al. Integrative regulatory mapping indicates that the RNA-binding protein HuR couples pre-mRNA processing and mRNA stability. Mol Cell 2011; 43: 327–339.

  26. 26

    Lebedeva S, Jens M, Theil K, Schwanhausser B, Selbach M, Landthaler M et al. Transcriptome-wide analysis of regulatory interactions of the RNA-binding protein HuR. Mol Cell 2011; 43: 340–352.

  27. 27

    Dewan MZ WM, Ahmed S, Terashima K, Horiuchi S, Sata T, Honda M et al. Hodgkin's lymphoma cells are efficiently engrafted and tumor marker CD30 is expressed with constitutive nuclear factor-kappaB activity in unconditioned NOD/SCID/gammac(null) mice. Cancer Sci 2005; 96: 7.

  28. 28

    Ventura A, Jacks T . MicroRNAs and cancer: short RNAs go a long way. Cell 2009; 136: 586–591.

  29. 29

    Lu J, Getz G, Miska EA, Alvarez-Saavedra E, Lamb J, Peck D et al. MicroRNA expression profiles classify human cancers. Nature 2005; 435: 834–838.

  30. 30

    Kumar MS, Lu J, Mercer KL, Golub TR, Jacks T . Impaired microRNA processing enhances cellular transformation and tumorigenesis. Nat Genet 2007; 39: 673–677.

  31. 31

    Martello G, Rosato A, Ferrari F, Manfrin A, Cordenonsi M, Dupont S et al. A MicroRNA targeting dicer for metastasis control. Cell 2010; 141: 1195–1207.

  32. 32

    Dean JL, Sully G, Clark AR, Saklatvala J . The involvement of AU-rich element-binding proteins in p38 mitogen-activated protein kinase pathway-mediated mRNA stabilisation. Cell Signal 2004; 16: 1113–1121.

  33. 33

    Lu JY, Schneider RJ . Tissue distribution of AU-rich mRNA-binding proteins involved in regulation of mRNA decay. J Biol Chem 2004; 279: 12974–12979.

  34. 34

    Kawai T, Lal A, Yang X, Galban S, Mazan-Mamczarz K, Gorospe M . Translational control of cytochrome c by RNA-binding proteins TIA-1 and HuR. Mol Cell Biol 2006; 26: 3295–3307.

  35. 35

    Linker K, Pautz A, Fechir M, Hubrich T, Greeve J, Kleinert H . Involvement of KSRP in the post-transcriptional regulation of human iNOS expression-complex interplay of KSRP with TTP and HuR. Nucleic Acids Res 2005; 33: 4813–4827.

  36. 36

    Katsanou V PO, Milatos S, Blackshear PJ, Anderson P, Kollias G, Kontoyiannis DL . HuR as a negative posttranscriptional modulator in inflammation. Mol Cell 2005; 19: 12.

  37. 37

    Aldinucci D, Lorenzon D, Cattaruzza L, Pinto A, Gloghini A, Carbone A et al. Expression of CCR5 receptors on Reed-Sternberg cells and Hodgkin lymphoma cell lines: involvement of CCL5/Rantes in tumor cell growth and microenvironmental interactions. Int J Cancer 2008; 122: 769–776.

  38. 38

    Skinnider BF, Mak TW . The role of cytokines in classical Hodgkin lymphoma. Blood 2002; 99: 4283–4297.

  39. 39

    Khan G . Epstein-Barr virus, cytokines, and inflammation: a cocktail for the pathogenesis of Hodgkin's lymphoma? Exp Hematol 2006; 34: 399–406.

  40. 40

    Kota SK, Balasubramanian S . Cancer therapy via modulation of micro RNA levels: a promising future. Drug Discov Today 2010; 15: 733–740.

  41. 41

    Ma L, Reinhardt F, Pan E, Soutschek J, Bhat B, Marcusson EG et al. Therapeutic silencing of miR-10b inhibits metastasis in a mouse mammary tumor model. Nat Biotechnol 2010; 28: 341–347.

  42. 42

    Stein CA, Hansen JB, Lai J, Wu S, Voskresenskiy A, Hog A et al. Efficient gene silencing by delivery of locked nucleic acid antisense oligonucleotides, unassisted by transfection reagents. Nucleic Acids Res 2010; 38: e3.

  43. 43

    Gentleman R, Carey V, Huber W, Irizarry R, Dudoit S . Bioinformatics and Computational Biology Solutions Using R and Bioconductor [Book] 2005; 473.

  44. 44

    Huber W, von Heydebreck A, Sultmann H, Poustka A, Vingron M . Variance stabilization applied to microarray data calibration and to the quantification of differential expression. Bioinformatics 2002; 18 (Suppl 1): S96–S104.

  45. 45

    G S . Limma: linear models for microarray data. In: Gentleman RCV, Dudoit S, Irizarry R, Huber W (eds). Bioinformatics and Computational Biology, Solutions using R and Bioconductor [Chapter] 2005; 23.

Download references


This work was supported by The Danish National Advanced Technology Foundation, The EC FP7 ONCOMIRS consortium (Grant agreement number 201102: this publication reflects only authors' views;the commission is not liable for any use that may be made of the information herein), The Novo Nordisk Foundation, The Lundbeck Foundation, The Danish Cancer Society and the Danish National Research Foundation. Dr Bellan's and Professor Leoncini's work is supported by the Monte dei Paschi di Siena Foundation. Dr Leucci is supported by a grant from the Danish Medical Research Council.

Author contributions: EL and AHL designed the overall study. SO and SK designed and provided the tiny anti-miR-LNAs. LHG performed all the bioinformatic analyses. LL and CB performed the analysis on NOG MICE. KTJ coordinated the mouse experiments; AZ and EL performed all the experiments. EL and AHL wrote the paper.

Author information

Correspondence to A H Lund.

Ethics declarations

Competing interests

Dr Obad and Professor Kauppinen are employees of Santaris Pharma. The other authors declare no conflict of interest.

Additional information

Supplementary Information accompanies the paper on the Oncogene website

Supplementary information

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Leucci, E., Zriwil, A., Gregersen, L. et al. Inhibition of miR-9 de-represses HuR and DICER1 and impairs Hodgkin lymphoma tumour outgrowth in vivo. Oncogene 31, 5081–5089 (2012).

Download citation


  • miR-9
  • cytokines
  • Hodgkin lymphoma
  • HuR
  • DICER1

Further reading