Regulation of the p27(Kip1) tumor suppressor by miR-221 and miR-222 promotes cancer cell proliferation

Abstract

MicroRNAs (miRNAs) are potent post-transcriptional regulators of protein coding genes. Patterns of misexpression of miRNAs in cancer suggest key functions of miRNAs in tumorigenesis. However, current bioinformatics tools do not entirely support the identification and characterization of the mode of action of such miRNAs. Here, we used a novel functional genetic approach and identified miR-221 and miR-222 (miR-221&222) as potent regulators of p27(Kip1), a cell cycle inhibitor and tumor suppressor. Using miRNA inhibitors, we demonstrate that certain cancer cell lines require high activity of miR-221&222 to maintain low p27(Kip1) levels and continuous proliferation. Interestingly, high levels of miR-221&222 appear in glioblastomas and correlate with low levels of p27(Kip1) protein. Thus, deregulated expression of miR-221&222 promotes cancerous growth by inhibiting the expression of p27(Kip1).

Figure 1

A genetic screen to identify miRNA suppressors of p27^Kip1. (A) A schematic representation of p27Sen, a retroviral vector for stable expression of GFP under the control of p27-3′UTR. (B) A clonal population of GFP-p27-3′UTR expressing cells was subjected to transduction with the entire miRNA library (miR-Lib) in a single-well format, selected and then mixed for further analysis. Two independent experiments were performed. Subsequently, low expressing GFP cells were sorted and the abundance of miRNA inserts was compared to the total population using miR-Array. (C) A GFP expression profile of control and miR-Lib-transduced cells. The region marked ‘low GFP' was sorted out of the entire population of cells. (D) A representative MA plot showing the signal and change of signal of each miRNA insert. (E) The top six hits found enriched in the low expressing GFP cells.

Figure 2

miR-221 inhibits translation of p27^Kip1. (A) Luciferase reporter experiments were performed with Firefly-luciferase-p27-3′UTR, control Renilla-luciferase and the indicated miRNA constructs. The luciferase ratio between the Firefly and Renilla of the control sample was adjusted to 1. A summary of three independent experiments is shown. (B) A stable p27Sen-HeLa cell line was transduced with miR-221 and control expressing vectors, and drug-selected for a week. Polyclonal cell populations were analyzed by flow cytometry a week later. (C) HeLa cells were transduced with control or miR-221 expressing vectors and drug-selected for a week. Subsequently, total RNA was extracted from the stable cells and RPA was performed with miR-221 and control cyclophilin probes. The lane P contains probes without RNase treatment, Y is a lane where yeast RNA was used as control. (D) An immunoblot analysis with p27 and CDK4 antibodies on the same cell populations as in panel B. Quantification was performed using Tina 2.0 software. (E) QRT-PCR was performed on the same RNA extracts used in panel C. (F) HeLa cells stably expressing control or miR-221 vectors were treated with 100 μg/ml cyclohexamide. At the indicated time points, whole-cell extracts were prepared and analyzed by immunoblot with p27 and control tubulin antibodies. Band intensities were quantified using Tina 2.0 software, and the resulting p27-tubulin ratios plotted in a regression plot.

Figure 3

Specificity of p27^Kip1 inhibition by miRNA-221 and miR-222. (A) A schematic representation showing the mature miR-221 and miR-222 sequences (in red, identical sequences). In addition, the genomic locus of miR-221&222 cluster with the inserts used in the miRNA vectors is also shown. Red boxes represent miRNA precursor positions. Also, the 3′UTR of p27^Kip1 is shown. The green boxes represent the two predicted miR-221&222 targeting sequences (as predicted by Pictar and TargetScanS software). (B) Luciferase reporter experiments were performed as in Figure 2A. 221SM and 222SM are constructs in which the seed sequence of the miRNAs was altered. (C) Luciferase reporter experiments were performed as in Figure 2A. p27-3′UTR-DM (double mutant) is a construct where both predicted miR-221&222 targeting sequences were modified (see Materials and methods). (B, C) The histograms show a summary of the results of three independent experiments.

Figure 4

Inactivation of miR-221&222 with corresponding antagomiRs. (A) A diagram showing the design of antagomiR-221 and antagomiR-222. (B, C) MCF-7 cells were transfected with the indicated luciferase reporter constructs, as in Figure 2A. Twenty-four hours following transfection, antagomiR-222 (25 μg) was added to the culture medium. Luciferase activity was measured 48 h later. Results show a summary of three independent experiments. (D) Stable control and miR-221 expressing HeLa cell populations were treated with antagomiR-221 for a period of 24 h. Whole cell extracts were prepared and analyzed by immunoblot with p27 and control tubulin antibodies. Bands were quantified using Tina 2.0 software.

Figure 5

Oncogenic addiction of certain cancer cell lines to miR-221&222 activity. (A) RPA was performed with miR-221 probe using total RNA from the indicated cell lines. Sizes of probes and mature miR-221 are indicated. (B) The indicated cancer cell lines were grown in the presence or absence of a mixture of antagomiR-221 and anagomiR-222 oligos. Images were obtained 4 days after treatment. (C) The growth of the indicated cancer cell lines, either exposed or not to antagomiR treatment, was measured using a 3T3 protocol. (D) The indicated cell lines were either treated or not (NT, non-treated) with a mixture of antagomiR-221 and antagomiR-222 oligos. Forty-eight later, cells were split and nocodazole was added (+Noc). Flow cytometry analysis was performed 24 h later. (E) Cells were treated as in panel B, whole cell lysate was extracted 3 days after treatment and subjected to immunoblot analysis with antibodies to detect the indicated proteins. (F) U87 cells were transfected with either control or a p27-shRNA expressing construct. Efficiency of transfection was at least 80%, as determined by GFP treatment (data not shown). After 24 h, cell populations were treated with a mixture of antagomiR-221 and antagomiR-222 and proliferation was determined as in panel C. An immunoblot analysis for p27^Kip1 and control tubulin proteins was performed on wild-type and p27^kd-transfected U87 cells. Bands were quantified using Tina 2.0 software.

Figure 6

Deregulated miR-221 and miR-222 expression in glioblastomas correlate with p27^Kip1 levels. (A) QRT-PCR for the expression of miR-221 and miR-222 was performed on RNA extracted from five tumor frozen tissues and five frozen peripherial tissue (0.5 cm from core tumor). As controls, we used the miR-221&222-negative HeLa cells, HeLa with miR-Vec-221 (HeLa-221) and U87, which are positive for both miRNAs. In parentheses are the standard deviations of three experiments. (B) Core and periphery material of tumor #4 was stained brown with Ki-67 and p27 antibodies. Nuclei were stained blue. (C) Whole-cell extracts were obtained from core (C) and periphery (P) material of tumors #1, 2, 3 and 5. A 20 μg weight of each sample was loaded on gel and subjected to either immunostaining with p27 antibody or silver staining, to demonstrate equal loading.