NF-kB and c-Jun induce the expression of the oncogenic miR-221 and miR-222 in prostate carcinoma and glioblastoma cells

Abstract

MicroRNAs (miRNAs) are potent negative regulators of gene expression involved in all aspects of cell biology. They finely modulate virtually all physiological pathways in metazoans, and are deeply implicated in all main pathologies, among which cancer. Mir-221 and miR-222, two closely related miRNAs encoded in cluster from a genomic region on chromosome X, are strongly upregulated in several forms of human tumours. In this work, we report that the ectopic modulation of NF-kB modifies miR-221/222 expression in prostate carcinoma and glioblastoma cell lines, where we had previously shown their oncogenic activity. We identify two separate distal regions upstream of miR-221/222 promoter which are bound by the NF-kB subunit p65 and drive efficient transcription in luciferase reporter assays; consistently, the site-directed mutagenesis disrupting p65 binding sites or the ectopical inhibition of NF-kB activity significantly reduce luciferase activity. In the most distal enhancer region, we also define a binding site for c-Jun, and we show that the binding of this factor cooperates with that of p65, fully accounting for the observed upregulation of miR-221/222. Thus our work uncovers an additional mechanism through which NF-kB and c-Jun, two transcription factors deeply involved in cancer onset and progression, contribute to oncogenesis, by inducing miR-221/222 transcription.

Figure 1.

NF-kB modulates miR-221/222 expression through two genomic sites upstream of miR-221/222 transcriptional unit. (A) Schematic diagram of Regions A (RegA) and B (RegB) genomic loci on human chromosome X derived from Human Mar. 2006 (NCBI36/hg18) assembly UCSC Genome Browser. The direction of miR-221 and miR-222 transcription is indicated by an arrow. (B) NF-kB activity assay. LNCaP cells were starved for 24 h and then stimulated with TNFα for 3 h. PC3 and U87 cells were transfected for 48 h with the pCMV-IκBαM (+IκBαM) or pCMV (−IκBαM) plus pEFGP, to assess transfection efficiency. Nuclear protein extracts were harvested and NF-κB activity was measured by transAM NFκB p65 protein kit. (C) Total RNA from LNCaP, PC3 and U87 cells treated as in (B) was extracted for quantitative real time polymerase chain reaction (qRT-PCR) analysis, miRNAs were normalized to snoRNA U6. For each protocol, three experiments were performed independently and the mean values (±SE) among experiments are presented. (D) p3′-UTR-p27 construct harboring the 3′-UTR of p27^kip1, was transfected into PC3 and U87 cells after 24 h of control (p3′-UTR-p27) or pCMV-IkBαM (p3′-UTR-p27 + IκBαM) plasmids transfection. (E) Left panel: empty reporter plasmid (c) or pRegA and pRegA-mp65 luciferase constructs containing, respectively, a wild-type and a mutated Region A were transfected into PC3 and U87 cells after 24 h of control (−) or pCMV-IkBαM (+IκBαM) plasmids transfection. Another complete set of samples was treated with TNFα for 3 h (+TNFα) in each cell line. Luciferase activity was determined 48 h after reporter plasmid transfection in all cases. The ratio of normalized sensor to control luciferase activity is shown. Right panel: the same set of experiments was performed for constructs pRegB and pRegB-mp65 harboring a wild-type or a mutated Region B, respectively. Data are presented as mean ± SE from three separate experiments with at least n = 3 for each experiment.

Figure 2.

NF-kB binding to Regions A and B. (A) Electrophoretic mobility shift assay. Nuclear extracts were prepared as described in the text, and assayed for protein binding to sequence bsA, corresponding to NF-kB binding site in the Region A. An unlabeled bsA probe (lane 4) or mut-bsA probe mutated in NF-kB site (lane 3) were used as competitors. Where indicated, an antibody recognizing the NFk-B p65 subunit was included (lane 5). Products of binding reactions were resolved by electrophoresis on 4% polyacrylamide gels and detected by autoradiography. (B) The same set of experiments was performed for Region B; competition with unlabeled wt probe (bsB) and mutated probe (mut-bsB) are indicated in lanes 5 and 3, respectively. Supershift with p65 antibody is evidenced in lane 4. (C) ChIP assay of chromatin isolated from PC3 and U87 cell lines transfected with control (bsA and bsB) or pCMV-IkBαm (bsA+ IkBαm and bsB+ IkBαm) plasmids, and immunoprecipitated by anti-p65 or control IgG, followed by qPCR analysis with primers targeted to sequences bsA and bsB and to a region on chromosome 1 used as the negative control (16). The data show occupancy relative to control IgG and represent mean ± SE of three independent experiments.

Figure 3.

C-Jun binding to Region B. (A) Schematic representation of NF-kB (white rectangles) and c-Jun (white triangle) binding sites, inside Regions A and B. (B) PC3 and U87 cells were transfected with control siRNAs or anti-c-JUN siRNAs and after 48 h total RNA and proteins were extracted. Levels of miR-221 and miR-222 were analyzed by qRT-PCR (left and central panels). The efficiency of siRNA-mediated depletion of c-Jun was assessed by western blot (right panel). Western blot analysis of β-actin is shown as a loading control. (C) p3′-UTR-p27 construct harboring the 3′-UTR of p27^kip1, was transfected into PC3 and U87 cells after 24 h of control siRNA (p3′-UTR-p27) or anti-c-JUN siRNAs (p3′UTR-p27 + siRNA c-jun) transfection. (D) ChIP assay of chromatin isolated from PC3 and U87 cell lines transfected with control (bsB) or anti-c-Jun (bsB +siRNA c-jun) siRNAs and immunoprecipitated by anti-c-Jun or control IgG, followed by qPCR analysis with primers targeted to sequence bsB or to a sequence 130 nt upstream of the 5′-end of miR221/222 (−130 nt, black triangle in A) that represents a previously identified c-Jun site (3). The same region shown in Figure 2C was used as the negative control. The data show occupancy relative to control IgG and represent mean ± SE of three independent experiments.

Figure 4.

C-Jun is the main activator of Region B. (A) Empty reporter plasmid (c) or pRegB, pRegB-mc-jun and pRegB mp65/c-jun luciferase constructs containing, respectively, a wild-type, a mutated Region B in c-Jun site or a mutated Region B in both p65 and c-Jun sites were transfected into PC3 and U87 cells. pRegB was also transfected into cells previously treated with anti-c-Jun siRNAs (pRegB +siRNA c-jun). Luciferase activity was determined 48 h after transfection. The ratio of normalized sensor to control luciferase activity is shown and data are represented as mean ± SE from three separate experiments with at least n = 3 for each experiment. (B) ChIP assay was performed as described in Figure 3C but the chromatin was immunoprecipitated by anti-p65 antibody and the data regarding bsA (bsA) or bsA in the presence of anti-c-Jun siRNAs (bsA + siRNA c-jun) were included. (C) Sequential ChIP assay performed in U87 cells on bsA (bsA) or bsB (bsB), by a preliminary immunoprecipitation with anti-c-Jun antibodies, and a subsequent one with anti-p65 antibodies. The data show occupancy relative to control IgG and represent mean ± SE of three independent experiments.

Figure 5.

Epigenetic modifications and RNA Polymerase II recruitment at miR-221/222 regulatory regions. (A) ChIP assay of chromatin isolated from U87 cells transfected with control (TSS), pCMV-IkBαM (TSS+IkBαm) or anti-c-Jun siRNAs (TSS+siRNA c-jun) and immunoprecipitated by anti-POLII or control IgGs, followed by qPCR analysis with primers targeted to miR-221/222 TSS or a negative control region (16). The same chromatin as in (A) was immunoprecipitated by anti-H3K4me1 (B) or anti-H3K9ac (C) antibodies and qPCR analysis was performed with primers targeted to the sequences bsA, bsB, TSS and negative control. The data show occupancy relative to control IgGs and represent mean ± SE of three independent experiments.