STAT5-mediated expression of oncogenic miR-155 in cutaneous T-cell lymphoma

Abstract

The pathogenesis of cutaneous T-cell lymphoma (CTCL) remains elusive. Recent discoveries indicate that the oncogenic microRNA miR-155 is overexpressed in affected skin from CTCL patients. Here, we address what drives the expression of miR-155 and investigate its role in the pathogenesis of CTCL. We show that malignant T cells constitutively express high levels of miR-155 and its host gene BIC (B cell integration cluster). Using ChIP-seq, we identify BIC as a target of transcription factor STAT5, which is aberrantly activated in malignant T cells and induced by IL-2/IL-15 in non-malignant T cells. Incubation with JAK inhibitor or siRNA-mediated knockdown of STAT5 decreases BIC/miR-155 expression, whereas IL-2 and IL-15 increase their expression in cell lines and primary cells. In contrast, knockdown of STAT3 has no effect, and BIC is not a transcriptional target of STAT3, indicating that regulation of BIC/miR-155 expression by STAT5 is highly specific. Malignant proliferation is significantly inhibited by an antisense-miR-155 as well as by knockdown of STAT5 and BIC.

In conclusion, we provide the first evidence that STAT5 drives expression of oncogenic BIC/miR-155 in cancer. Moreover, our data indicate that the STAT5/BIC/miR-155 pathway promotes proliferation of malignant T cells, and therefore is a putative target for therapy in CTCL.

Introduction

miRNAs are short (20–25 nt) post-transcriptional regulators of gene expression. Their activity plays a role in a multitude of cellular events, including the orchestration of the immune response.^1,2 Dysregulation of miRNA expression can result in various pathologies, including cancer.³ miR-155 is one of the most studied oncogenic miRNAs. It has been found overexpressed in hematological^4,5 as well as a number of solid cancers.^6-9 miR-155 is localized within an exon of its precursor gene BIC (B cell integration cluster or pri-miR-155).⁴ Transiently elevated expression of miR-155 plays a role in the activation of several types of immune cells,^2,10-12 whereas constitutively high levels of this miRNA can result in the development of malignancies by increase of genomic instability^13,14 and sustained proliferation and survival^5,15 of malignant cells. Thus, miR-155 is often referred to as a “bridge between inflammation and cancer.”¹³ Chronic inflammation is a hallmark of cutaneous T-cell lymphoma (CTCL), a clinically heterogeneous group of lymphoproliferative malignancies derived from skin-homing T cells.¹⁶ Recently, miR-155 has been found overexpressed in CTCL and provides diagnostic value distinguishing between benign and malignant inflammation.^17,18

The most common subtypes of CTCL are mycosis fungoides (MF) and Sézary syndrome (SS), an aggressive CTCL characterized by peripheral blood involvement. Early MF lesions often present as flat erythematous patches or plaques. In progressive disease, skin lesions develop into overt tumors, and the malignant T cells may spread to lymph nodes and internal organs.¹⁶

The etiology of CTCL is unknown, but evidence suggests that the IL-2 receptor (IL-2R) complex and the associated Janus kinases (JAKs) and signal transducers and activators of transcription (STATs) play a key role in the disease pathogenesis. Thus, malignant T cells display a constitutive expression of the high-affinity IL-2R complex [consisting of IL-2Rα (CD25), IL-2Rβ (CD122) and IL-2Rγ (CD132)]^19-21 and respond to IL-2 family cytokines (IL-2,²² IL-7,^23,24 IL-15,²⁵ and IL-21²²) by activation of STATs, resulting in malignant cell proliferation and survival.^23-25 In advanced disease, malignant T cells display a cytokine-independent, constitutive activation of JAK3, which drives a constitutive activation of STAT3 that, in turn, increases survival and resistance to apoptosis in malignant T cells,^26,27 and induces production of cytokines involved in eosinophilia and erythroderma (IL-5), as well as T helper 2 (Th2),²⁸ and Th17²⁹ cytokines. Furthermore, the pathway is believed to play a role in creating a pro-oncogenic inflammatory environment via production of VEGF³⁰ and IL-10³¹ and induction of suppressor of cytokine signaling-3 (SOCS-3),³² which confers resistance to IFNα in malignant T cells. Interestingly, miR-21, another oncogenic miRNA that is significantly upregulated in CTCL patients,^17,18 was shown to be modulated by STAT3.^33,34

Much less is known about STAT5, another downstream effector of the IL-2R/JAK3 complex. A previous study indicated that STAT5 is aberrantly activated in CTCL, but only a small number of patients was analyzed, and the pathological relevance was unclear.³⁵ Others obtained evidence that STAT5 was truncated and functioned as a transcriptional repressor in malignant T cells in SS,³⁶ whereas another study²³ reported that STAT5 is involved in survival of malignant T cells via induction of Bcl-2.

While early stage CTCL can be treated successfully, there is no cure or standard therapy that has proven to definitively prolong survival of late stage CTCL patients, where treatments are mostly palliative. Recurrent challenges are serious side effects and the development of resistances to systemic treatments.³⁷ Accordingly, it is of great importance to thoroughly characterize aberrantly activated molecular pathways and their role in the disease.

The present investigation was undertaken to address the abnormal expression and function of the oncogenic miR-155 in CTCL. We provide the first evidence that the miR-155 host gene BIC is a transcriptional target of STAT5, and that miR-155 is involved in proliferation of malignant T cells, suggesting that the STAT5/BIC/miR-155 pathway is a putative target for therapy.

Results

The expression of miR-155 is enhanced in affected skin from CTCL patients when compared with healthy individuals and patients with benign inflammatory disorders.^17,18 Here, we provide the first evidence that miR-155-5p (from here on referred to as miR-155) and its precursor BIC are expressed at significantly higher levels in malignant CTCL cells than in non-malignant T cells (Fig. 1A and B). Thus, miR-155 expression in the malignant T cell lines MyLa2059, PB2B and MAC2A was 5–9 times higher than in a non-malignant T cell line (MySi). Also, the SS cell lines SeZ4.1 and SeAx showed 3–5 times higher expression than the non-malignant cells (Fig. 1A). Likewise, BIC is strongly expressed in all malignant T cells but not in non-malignant T cells (Fig. 1B), confirming a constitutive BIC/miR-155 expression in malignant T cells.

Figure 1. BIC/miR-155 expression in CTCL. (A) miR-155 expression in non-malignant (MySi) and malignant (SeAx, SeZ4.1, MyLa2059, PB2B, MAC2A) CTCL T cell lines as measured by qPCR. (B) Expression of the miR-155 host gene BIC in the same cell lines obtained by RT-PCR. Reference GAPDH. miR-155 expression levels are significantly higher in the malignant cell lines and highest in the MF cell line MyLa2059 and the ALCL cell line MAC2A.

To address which signaling pathways drive BIC/miR-155 expression, we analyzed the miR-155 host gene (MIR155HG/BIC) promoter for potential transcription factor binding sites.^38,39 The region contains a putative STAT binding site (Fig. 2A, blue) as well as binding sites for members of the NFκB family (not shown). In other human cell types, miR-155 was induced by synergistic activity of IFNγ and TNFα. This induction was partially inhibited by NFκB (p65) knockdown, whereas the role of STAT proteins was not addressed.⁴⁰ Both STAT3 and NFkB are known to be constitutively activated in CTCL.^26,27,41 Likewise, we observed expression of the active form of STAT5 (pY-STAT5) in malignant T cells in situ and in cell lysates from malignant T cell lines (Fig. S1A–D and E, respectively), confirming and extending other studies,³⁵ and suggesting that STAT5 is constitutively activated in CTCL.

Figure 2. miR-155 precursor BIC is a transcriptional target of STAT5. (A) Analysis of the BIC promoter region yielded one STAT binding site (highlighted in blue). Primers flanking the binding site that were used for PCR analysis of ChIP samples are indicated in red. The beginning of the BIC transcript is highlighted in bold script. (B) ChIP-seq reads from the BIC (MIR155HG) gene promoter region in malignant MyLa2059 cells. Reads (76 bases) obtained from immunoprecipitation of STAT5, STAT3, RelA and a negative control (rabbit IgG, bottom). The chromosomal positions of MIR155HG refer to hg19. Forward reads are indicated in green and reverse reads in red. (C) PCR analysis of ChIP samples using the primer set indicated in (B). The 190 bp amplicon was only detected in the STAT5-precipitated samples and the positive control (histone H3). (D) Similar results were obtained by qPCR: STAT5-precipitated samples displayed a 10-fold enrichment of the amplified sequence in relation to a negative control, whereas detection in STAT3 precipitated samples ranged at background level. U6 rRNA was used for normalization. (E) Luciferase assay for BIC promoter activity. CHO-K1 cells were transiently transfected with pCMV-LacZ, empty pGL3 or pGL3-BIC-WT, together with STAT5A, STAT5A-CA, STAT5B or STAT5B-CA (CA, constitutively active). Transfection with the constitutively active STAT5A-CA and STAT5B-CA resulted in a 28.6- (p < 0.001) and 29-fold (p = 0.002) induction of BIC promoter activity, respectively. Data were obtained from three independent experiments.

To elucidate whether the above pathways regulated miR-155 induction in CTCL, we performed chromatin immunoprecipitation followed by DNA sequencing (ChIP-seq) to identify transcriptional targets of STAT and Rel transcription factors in CTCL. ChIP-seq analysis of STAT5-precipitated chromatin from malignant MyLa2059 cells yielded an enrichment of reads comprising a region of the miR-155 host gene promoter (BIC, Fig. 2B). In contrast, no reads for the BIC promoter were detected in chromatin precipitated with STAT3 antibody (Fig. 2B). Also, precipitation with RelA (p65) did not yield any reads for the BIC promoter region and, in line with that, NFκB inhibitors had no effect on miR-155 expression (Fig. 2B and data not shown). Based on the information obtained from ChIP-seq, we designed primers (Fig. 2A, red) flanking the STAT-binding site in the promoter. These were used to detect enrichment of this sequence in chromatin from malignant MyLa2059 precipitated with STAT3 and STAT5 antibodies relative to a negative control antibody (rabbit IgG) by PCR. In line with the results from ChIP-seq, conventional PCR (Fig. 2C) as well as qPCR (Fig. 2D) showed an enrichment of the sequence representing the BIC promoter in samples precipitated with STAT5 antibody, whereas STAT3-precipitated samples were negative or equal to background levels. In support of these data, expression of constitutive active forms of STAT5A and STAT5B in CHO cells resulted in an increased transcription from the BIC promoter as judged from luciferase activity (Fig. 2E). Taken together, these findings indicate that BIC, the transcript encoding the mature miR-155, is a transcriptional target of activated STAT5.

To verify the results above, we studied BIC/miR-155 expression levels after functional inhibition of the JAK/STAT pathway and following siRNA-mediated knockdown of STAT3 or STAT5 (Fig. 3). Knockdown of STAT5A, STAT5B and both STAT5A and STAT5B (Fig. 3A) resulted in a 50% inhibition of BIC and 40% inhibition of miR-155 expression (Fig. 3B). Essentially identical results were obtained in three independent experiments and two different malignant T cell lines (MyLa2059 and MAC2A, Fig. 3B left and right, respectively). It was a repeated observation that STAT5A and STAT5B were only partially depleted by siRNA treatment (Fig. 3A), which might explain why the inhibition of miR-155 expression was only partial. STAT3 was recently shown to regulate miR-21 expression in malignant T cells.^33,34 As siRNA directed against STAT3 resulted in an almost complete depletion of STAT3 without modulating the expression of BIC and miR-155 (Fig. 3A and B), it appears that STAT3 and STAT5 have distinct and specialized roles regulating miR-21 and miR-155, respectively.

Figure 3. miR-155 expression is regulated by the JAK/STAT pathway. (A) Representative western blot showing siRNA-mediated knockdown of STAT3, STAT5A, STAT5B and STAT5A+B in malignant MyLa2059/MAC2A cells. (B) Relative quantification of miR-155 (top) and BIC (bottom) in malignant MyLa2059 (left) and MAC2A (right) cells as measured by qPCR and determined by the ddCt method. miR-155 expression was decreased by 40% after knockdown of STAT5A and STAT5B, but not STAT3 in both cell lines; nt, non-targeting control. p values were obtained from Student’s t-test. *Indicates a significance of p < 0.05. (C) Incubation with JAK3 inhibitor (JAK3i) CP-690550 (50 µM) decreased pY-STAT5 (top) and resulted in decreased miR-155 expression in malignant MyLa2059 cells. miR-155 expression was decreased by 60–80% when cells were treated with JAK3i for 24, 48 or 72 h. Control DMSO. (D) JAK3 inhibition with CP-690550 (0.1 and 50 µM, 24 h) in PBMCs from a patient with SS resulted in a decreased STAT5 phosphorylation (top) and a 50% inhibition of miR-155 expression.

As Janus kinases such as JAK3 and JAK1 are constitutively active in malignant T cells,^30,42 we examined the effect of JAK inhibition on STAT5 activation and miR-155 expression in malignant T cell lines and primary PBMCs from a SS patient. As shown in Figure 3C and D, a selective inhibitor of JAK kinases (CP-690550) strongly inhibited the activation of STAT5 and triggered a decreased expression of miR-155, supporting the idea that STAT5 drives BIC and miR-155 expression.

IL-2 and IL-15 have been implicated in the pathogenesis of CTCL and are believed to shape the malignant phenotype through signaling via the IL-2Rβ (CD122)/γc (CD132)/JAK3 complex.²² As these cytokines activate STAT5 in normal T cells, we investigated whether IL-2 influenced BIC/miR-155 expression in malignant and non-malignant CTCL T cells. When we incubated non-malignant cells (MyLa1850) with IL-2 for 24 h, pY-STAT5, BIC and miR-155 expression levels increased significantly (Fig. 4A, upper; p < 0.05). A similar increase in BIC/miR-155 was also observed when malignant SS cells (SeAx) were incubated with IL-2 (Fig. 4A, lower, p < 0.02). Of note, the IL-2-dependent induction of miR-155 transcription was abolished by simultaneous administration of JAK3 inhibitor (Fig. S2). Importantly, an IL-2-dependent induction of miR-155 was also observed in primary T cells obtained from the blood of SS patients (Fig. 4B, P3 and P4) and sorted into CD4⁺/CD26⁻ (malignant) and CD4⁺/CD26⁺ (non-malignant) populations. Following culture with and without IL-2 for 24 h after sorting, T cells cultured with IL-2 displayed increased levels of miR-155 (Fig. 4B) when compared with the cytokine-starved T cells. IL-2-dependent induction of miR-155 was more pronounced in both fractions obtained from P4 than in those from P3, suggesting heterogeneity among patients. To address whether IL-2-driven miR-155 expression was a unique feature of CTCL, PBMCs from healthy donors were incubated with IL-2 for 24 h and assayed for miR-155 expression. Indeed, miR-155 expression in these cells was increased 2-fold by IL-2 (Fig. 4C). As shown in Figure 4D, IL-15 also induced an enhanced expression of BIC and miR-155 in non-malignant T cells (Fig. 4D, left and right, respectively). Of note, IL-21 did not affect miR-155 expression levels (data not shown), indicating that only cytokines engaging both the IL-2Rβ- and γc- chains (IL-2 and IL-15) induce BIC/miR-155 expression in T cells.

Figure 4. miR-155 expression in response to STAT5-activating cytokines. (A) Incubation with IL-2 (10³U/ml) for 24 h increased expression of BIC (left) and miR-155 (right) in non-malignant MyLa1850 (top, 1.7-fold increase, p < 0.05) as well as malignant SeAx T cells (bottom, 1.5-fold increase, p < 0.02) as shown by qPCR, and was accompanied by induction of STAT5 phosphorylation. (B) In primary T cells obtained from two SS patients (P3 and P4), incubation with IL-2 (2 × 10³ U/ml) for 24 h resulted in an increase of miR-155 in both malignant (CD26⁻) and non-malignant (CD26⁺) populations. The increase was stronger in T cells isolated from P4; however, both samples showed elevated levels of miR-155 in response to IL-2. (C) In PBMCs from a healthy donor, a 24 h incubation with IL-2 (2 × 10³U/ml) led to a 2.3-fold increase of miR-155. (D) Incubation of non-malignant MySi T cells with IL-15 (25 ng/ml) for 24h resulted in increased levels of BIC (2-fold) and miR-155 (1.5-fold). p values were obtained from Student’s t-test. *Indicates a significance of p < 0.05.

miR-155 has been assigned the status of an oncogenic miRNA, because it induces genomic instability,^13,14 and elevated expression levels are observed in hematological^4,5,17 as well as different types of solid tissue cancer.^6-9 Sustained proliferation is a hallmark of transformed cells. Knockdown of BIC in malignant MyLa2059 cells entailed a 40% decrease of miR-155 expression (Fig. 5A, p = 0.02) and a 20–25% decrease in spontaneous proliferation (Fig. 5B, p = 0.01) measured by ³H-thymidine uptake. Similar results were achieved using miRNA inhibitors of miR-155 (antagomiR-155, Fig. 5C, p < 0.001). Likewise, STAT5 knockdown resulted in a 20–25% decrease in proliferation (Fig. 5D, STAT5A p = 0.004). These data indicate that BIC/miR-155 is involved in malignant cell growth and suggest that the inhibitory effects of STAT5 knockdown on proliferation might be mediated through the inhibition of miR-155. Taken together, this study provides the first evidence for a STAT5-dependent induction of oncogenic miR-155 in cancer.

Figure 5. miR-155 is involved in malignant proliferation. (A) siRNA-mediated knockdown of BIC in malignant MyLa2059 cells resulted in a 40% decrease of mature miR-155 levels after 48 h (p = 0.02). (B) Proliferation of malignant cells in the same samples was reduced by 20% (p = 0.01) determined by methyl-³H-thymidine incorporation. (C) Transfection of MyLa2059 with small RNA inhibitors targeting the mature miR-155 (antagomiR-155) yielded similar results, entailing a 20% reduction in cell proliferation. p < 0.001. (D) Likewise, knockdown of STAT5A (p = 0.004) or STAT5B (ns) entailed an approximately 25% decrease of proliferation as determined by ³H-thymidine uptake. nt = non-targeting control. p values were obtained by Student’s t-test. *Indicates a significance of p < 0.05.

Discussion

There is plenty of data indicating that the STAT5 transcription factor has transforming and oncogenic potential. STAT5 plays a crucial role in lymphocyte development and has been shown to sustain survival and proliferation of transformed cells. It has been found aberrantly activated in leukemia and lymphoma as well as solid cancers as breast and prostate cancer (reviewed in refs. 43–45). In this light, it is surprising how little is known about the function of STAT5 and its downstream targets in CTCL. We provide the first evidence that the JAK/STAT5 pathway drives expression of BIC and miR-155. Thus, (1) STAT5 binds to the BIC promoter; (2) siRNA-mediated STAT5 knockdown inhibits BIC and miR-155 expression; (3) JAK inhibition blocks STAT5 activation and inhibits BIC/miR-155 expression; and (4) IL-2Rβγ cytokines enhance STAT5 activation and BIC/miR-155 expression. In contrast to STAT5, STAT3 does not bind the BIC promoter, and knockdown of STAT3 has no effect on the expression levels of BIC and miR-155, indicating a non-redundant role of STAT3 and STAT5 in the regulation of BIC/miR-155 expression. This conclusion is in agreement with previous findings that STAT3 modulates miR-21 expression^33,34 and our observation that STAT5 knockdown has no effect on miR-21 expression (data not shown).

Although an aberrant activation of STAT5 is confined to malignant T cells, JAK/STAT5-mediated BIC/miR-155 expression was not confined. In contrast, non-malignant T cells from CTCL patients as well as PBMCs from healthy individuals also display a significant upregulation of BIC/miR-155 upon STAT5 activation by IL-2Rβγ cytokines (IL-2 and IL-15), indicating that transcriptional regulation of BIC by STAT5 is not a unique feature of malignant T cells, but a common regulatory pathway.

In addition, NFκB has been identified as a transcriptional regulator of miR-155 in human cells. However, NFκB blockage only partly inhibited cytokine-mediated miR-155 expression, indicating that other signaling pathways were also involved.⁴⁰ Interestingly, IFNγ and TNFα synergistically induced miR-155 expression,⁴⁰ but the role of STAT proteins was not investigated. Kutty et al. obtained circumstantial evidence that STAT1 regulates miR-155 expression in epithelial cells.⁴⁶ Besides the canonical activation of STAT1, IFNγ induces activation of STAT5.⁴⁷ Given our findings of STAT5 as a regulator of miR-155 expression, it is tempting to speculate that STAT5 is also involved in IFNγ- mediated miR-155 expression. We looked for, but did not find, involvement of NFkB in miR-155 expression in malignant CTCL T cells. Thus, NFκB pathway inhibitors did not inhibit miR-155 expression, and precipitation of chromatin with RELA (p65) antibody did not result in an enrichment of reads for the miR-155 promoter region despite enrichment of known NFkB target genes. This conclusion is in line with the observation that the expression and regulation of specific miRNAs is heavily dependent on the cellular context and can vary greatly among different cell types and tissues.

STAT5 is known to be involved in the transcription of anti-apoptotic (bcl-2, bcl-xl) genes as well as genes such as cyclin D and c-myc that drive cell cycle progression and proliferation and, hence, sustain survival and cell growth.²³ We also observed that knockdown of STAT5 inhibited proliferation of malignant T cells, albeit only by 20–25%. The effect of STAT5 inhibition was paralleled by a 20–25% inhibition of proliferation following BIC knockdown and inhibition of mature miR-155 with antagomiRs. However, our inability to obtain a complete knockdown of STAT5 (as seen in Fig. 3A) and BIC limits the conclusions, which can be drawn regarding the role of STAT5/BIC/miR-155 in malignant proliferation. Still, a decrease in proliferation of malignant cells was observed in two distinct assays, both resulting in miR-155 inhibition: knockdown of BIC as well as inhibition of mature miR-155 by treatment with an antisense miRNA. Although the present findings might underestimate the importance of STAT5 and BIC/miR-155, our study clearly shows that inhibition of STAT5A and miR-155 triggers a significant decrease in the proliferation of malignant T cells. Given the enhanced expression of miR-155 in CTCL,^17,18 our findings therefore provide the first evidence that miR-155 functions as an oncomiR in CTCL.

In conclusion, we show that STAT5 drives the aberrant BIC/miR-155 expression and provide the first evidence that the STAT5/BIC/miR-155 pathway is involved in malignant proliferation. These data further support a crucial role of the JAK/STAT pathway in the pathogenesis of and as potential therapeutic target in CTCL.

Materials and Methods

Patients and ethics

The study includes formalin-fixed and paraffin-embedded biopsies from patients diagnosed with CTCL during the period from 1979–2004. Samples were drawn from the archives of the Departments of Pathology at Rigshospitalet (Copenhagen University Hospital) and Bispebjerg Hospital and have been described in detail elsewhere.⁴⁸ For immunohistochemistry, 47 cases were selected for analysis. Fourteen patients had MF with patches, and 21 had plaques. Four patients displayed tumors, and five had MF with transformation to a large T-cell lymphoma. Three patients with SS were included. The study was approved by the Ethics Committee of Copenhagen and Frederiksberg (journal no. 01 284225) and the Danish Data Protection Agency (Datatilsynet, journal no. 2005-41-5930).⁴⁸

Cell lines and culture

Malignant T cells lines (MyLa2059, PB2B, SeAx and SeZ4.1) and non-malignant T cell lines (MySi, MyLa1850) were obtained from CTCL patients and have been described perviously.^26,29,30,49 MAC2A is an anaplastic large T-cell lymphoma (ALCL) cell line established from a skin tumor in the progressive phase of the disease.⁵⁰ The Jurkat cell line (J-Tag) has been described elsewhere.⁴¹ All cell lines were cultured as described previously.^{26,29,30,49-51} Peripheral blood mononuclear cells (PBMCs) from a healthy donor were isolated by density gradient centrifugation using Lymphoprep (Axis-Shield, PoC AS) as described before.⁵² PBMCs from three patients diagnosed with SS were isolated as described before²⁹ and sorted to yield CD4+/CD26⁻ as well as CD4+/CD26+ populations. After sorting, both populations were cultured with and without IL-2 (2 × 10³U/ml, Proleukin, Chiron) for 24 h.

Reagents and antibodies

JAK3 inhibitor CP-690550 (Pfizer), DMSO (Sigma-Aldrich, D2438). rhIL-2 (Proleukin, Chiron) and rhIL-15 (Peprotech, #200-15). Anti-CD4-FITC and anti-CD26-APC antibodies were from BD Biosciences (#345768) and BioLegend (BA5b, #302710), respectively. STAT5A and STAT5B antibodies were from Santa Cruz Biotechnology (L-20, sc-1081; C-17, sc-835-G), py-STAT5, STAT5 and STAT3 antibodies from Cell Signaling Technology (#9351, #9363, #4904). GAPDH antibody from Abcam (Ab9485), and anti-actin antibody from Sigma (A4700, AC-40).

Fluorescence-activated cell sorting (FACS)

PBMCs from SS patients were labeled with antibodies for CD4 and CD26, as described before.²⁶ Subsequently, cells were sorted in a FACSAria cell sorter (BD) to obtain CD4+/CD26− as well as CD4+/CD26+ populations.

qRT-PCR (TaqMan)

Total RNA was purified with miRNeasy Kit (Qiagen), cDNA was transcribed from 10ng RNA using TaqMan^® miRNA Reverse Transcription Kit (Applied Biosystems), and real-time PCR was performed using TaqMan^® miRNA assays (Applied Biosystems) for miR-155-5p according to the manufacturer’s instructions. U6 rRNA was included as a reference besides the miRNA of interest. For analysis of BIC (pri-miR-155) expression, cDNA was transcribed using the High Capacity cDNA Reverse Transcription Kit followed by PCR analysis using TaqMan^® pri-miRNA assays (Applied Biosystems) according to the manufacturer’s instructions. GAPDH, POP4 and PSMC4 (TaqMan^® Gene Expression Assays, all Applied Biosystems) were used as references. Amplification was performed in an Mx3000P real-time thermal cycler (Stratagene) on standard settings. Data presented here was obtained from three independent experiments. Each experiment included three technical replicates. Results are presented as relative quantity to the control sample determined by the ddCt method.

qRT-PCR (SYBR)

mRNA was purified with mRNeasy Kit (Qiagen) and transcribed to cDNA as described previously.⁴⁹ Quantitative analysis of BIC expression was performed using DyNAmo^™ Flash SYBR^® Green qPCR Kit (Finnzymes) with BIC (pri-miR-155)-specific primers (forward: 5′-CCCAATGGAGATGGCTCTAA-3′ ; reverse: 5′-AGGAGTCAGTTGGAGGCAAA-3′) according to the manufacturer’s instructions. Amplification was performed in an Mx3000P real-time thermal cycler (Stratagene). Each experiment included three technical replicates. Results are presented as relative quantity to the control sample determined by the ddCt method.

RT-PCR

cDNA was transcribed from mRNA purified with mRNeasy Kit (Qiagen). Amplification with BIC (pri-miR-155) (forward: 5′-CTCTAATGGTGGCACAAA-3′, reverse: 5′-TGATAAAAACAAACATGGGCTTGAC-3′) and GAPDH (forward: 5′-CCATGGAGAAGGCTGGGG-3′, reverse: 5′-CAAAGTTGTCATGGATGACC-3′)-specific primers was performed as described previously.⁴⁹

Chromatin immunoprecipitation (ChIP)

For ChIP analysis, SimpleChIP^® Enzymatic Chromatin IP Kit (Agarose Beads) from Cell Signaling Technologies was used according to the manufacturer’s instructions. In brief, cells were cross-linked in 1% formaldehyde. After cell lysis and sonication, the chromatin was fragmented to an average size of 500 bp and captured with antibody against STAT3, STAT5, RELA, histone H3 (positive control) or rabbit IgG (all Cell Signaling Technologies) at 4°C overnight. Cross-links for enriched and input DNA were then reversed, and DNA was treated with RNase A and proteinase K before purification using spin columns. Immunoprecipitated and input DNA were subjected to PCR and qPCR using PCR primers for BIC (MIR155HG) promoter upstream regions flanking STAT5 binding sites: BIC promoter forward, 5′-GAAAGGGAAAGGGGAAAACA-3′, and reverse, 5′-CGAACGTGCGACCCTTTTAT-3′. PCR annealing temperature was 56°C, and the amplicon size was 190 bp. qRT-PCR (SYBR) was performed as described above. Quantitation of transcription factor binding was expressed as enrichment ratios of antibody over the IgG control.

ChIP-Seq

Library construction for ChIP sequencing was performed by usage of the Illumina’s “ChIP-Seq DNA Sample Prep Kit, IP-102-1001” with 10 ng of immunoprecipitated DNA as starting material and with gel size selection of the 150–250 bp after adaptor ligation. Sequencing was performed as single read sequencing (76 cycles) on an Illumina GAIIx. ChIP-seq reads were aligned to hg19 (on average 21 million reads/library) using CLC Genomics Workbench.

Luciferase assay for BIC promoter activity

CHO-K1 cells were transiently transfected with 0.1 μg pCMV-LacZ, 0.5 μg of either empty pGL3 or pGL3-BIC-WT, together with 1.0 μg STAT5A, STAT5A-CA, STAT5B or STAT5B-CA, respectively, using 4 μl of Lipofectamine 2000 transfection reagent (Invitrogen); CA, constitutively active. Cells were rested for 24 h before luciferase, and β-galactosidase activities were assayed according to the manufacturer’s protocol (Promega). The β-galactosidase activity was used as internal control for normalization of luciferase activity. Fold induction was calculated as pGL3/pGL3-BIC-WT for each of the four STAT5 expression plasmids. Data was obtained from three independent experiments. The pGL3 reporter construct containing BIC wild type promoter (pGL3-BIC-WT) was a generous gift from Dr Erik Flemington and has previously been described.⁵³ The STAT5A or STAT5B expression plasmids [wild type, or constitutively active (CA)] were a kind gift from Dr Peter Lobie and have previously been described.⁵⁴

Transient transfection with siRNA/miRNA inhibitors (antagomiRs)

Cells were transfected with small interfering RNA (siRNA) against STAT3, STAT5A, STAT5B and non-targeting control #1 (ONtarget PLUS, smart pool, Dharmacon) or siRNA directed against BIC and universal negative control #1 (Sigma) using 0.5 nmol of the respective siRNA on 2 × 10⁶ cells with an Amaxa Nucleofector as described previously.⁴⁹ For proliferation assays, the same protocol was applied to transfect cells with miR-155 antagomiRs and negative control (mirVana^™ miRNA Inhibitors, Ambion^®) using a concentration of 300 nM.

[Methyl-³H]-thymidine proliferation assay

Proliferation of malignant MyLa2059 cells was determined by [methyl-³H]-thymidine incorporation as described previously.⁴⁹

Protein extraction, SDS-PAGE and western blotting

1 × 10⁶ cells were pelleted, lysed and protein samples were subjected to SDS-PAGE and western blotting as described previously.⁴⁹

Statistics

Data presented in this report were obtained from three independent experiments each including technical replicates. For statistical analysis, a two-tailed Student’s t-test with a significance level of 0.05 was used. The asterisk symbol (*) denotes a significant difference (p < 0.05) between the indicated sample and the control sample.

Glossary

Abbreviations:

CTCL

cutaneous T-cell lymphoma

MF

mycosis fungoides

SS

Sézary syndrome

JAK

Janus kinase

STAT

signal transducers and activators of transcription

BIC

B cell integration cluster

miRNA

microRNA

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Supplemental Materials

Supplemental materials may be found here: 
www.landesbioscience.com/journals/cc/article/24987