miR-128 inhibits telomerase activity by targeting TERT mRNA

Abstract

Telomerase is a unique cellular reverse transcriptase (RT) essential for maintaining telomere stability and required for the unlimited proliferation of cancer cells. The limiting determinant of telomerase activity is the catalytic component TERT, and TERT expression is closely correlated with telomerase activity and cancer initiation and disease progression. For this reason the regulation of TERT levels in the cell is of great importance. microRNAs (miRs) function as an additional regulatory level in cells, crucial for defining expression boundaries, proper cell fate decisions, cell cycle control, genome integrity, cell death and metastasis. We performed an anti-miR library screen to identity novel miRs, which participate in the control of telomerase. We identified the tumor suppressor miR (miR-128) as a novel endogenous telomerase inhibitor and determined that miR-128 significantly reduces the mRNA and protein levels of Tert in a panel of cancer cell lines. We further evaluated the mechanism by which miR-128 regulates TERT and demonstrated that miR-128 interacts directly with the coding sequence of TERT mRNA in both HeLa cells and teratoma cells. Interestingly, the functional miR-128 binding site in TERT mRNA, is conserved between TERT and the other cellular reverse transcriptase encoded by Long Interspersed Elements-1 (LINE-1 or L1), which can also contribute to the oncogenic phenotype of cancer. This finding supports the novel idea that miRs may function in parallel pathways to inhibit tumorigenesis, by regulating a group of enzymes (such as RT) by targeting conserved binding sites in the coding region of both enzymes.

Keywords: TERT; cancer; miR; miR-128; telomerase.

Figure 1. Identification of miR-128 as an inhibitor of telomerase activity

(A) Schematic of the anti-miR library screen to identify miRs that regulate telomerase activity. HeLa cells were transduced using a lentiviral-based, miR-neutralizing shRNA library, selected for puromycin resistance and clonally expanded. Each well represents the neutralization of a single endogenously expressed miR. Cells were then subjected to quantitative PCR-based TRAP assay (q-TRAP) analysis using the quantitative telomeric repeat amplification protocol. (B) Relative telomerase activity of HeLa cells after transduction with lentiviral miR-neutralizing anti-miR library, selection, and clonal expansion as measured by quantitative telomeric repeat amplification protocol (q-TRAP) as shown. Shown is a panel of miR controls (Controls), relative to samples transduced with the anti-miR library (Samples). (C) Secondary measurement of relative telomerase activity in select samples of anti-miR library-expressing HeLa cells, as described for Figure 1B. (D) Single high-titer miR-control, miR-128 and anti-miR-128 lentiviruses were generated and stable miR-modulated HeLa cells were assayed for relative telomerase activity measured by q-TRAP analysis. Results shown as percent change ± SEM (n = 3, independent biological replicates, ^*p < 0.05, ^****p < 0.0001).

Figure 2. miR-128 reduces the expression of TERT mRNA and protein

(A) miR-modulated HeLa cell lines were generated (over-expressing miR-128, anti-miR-128 or control miR). Relative expression levels of Tert RNA were determined and normalized to beta-2-microglobulin (B2M) expression levels. (B) Stably miR-128, anti-miR-128 or control miR control HeLa cells lines were analyzed by western blot analysis using antibodies against Tert and α-tubulin. One representative example of three is shown. Quantification of results (n = 3) normalized to α-tubulin is shown (right panel). (C) Stable miR-128, anti-miR-128 and control miR HeLa cell lines were analyzed by immunofluorescence for Tert expression and co-localization with DAPI. Quantification of results is shown (bottom panel) (n = 3). (D) Stable A549, SW620 and PANC1 miR-modulated (miR-128, anti-miR-128 and control miR) cell lines were generated and Western blot analysis were performed using Tert and α-tubulin antibodies. Quantification of results (n = 3) normalized to α-tubulin is shown (bottom panels). All results are shown as mean ± SEM, n = 3, independent biological replicates. ^*p < 0.05, ^**p < 0.01, ^***p < 0.001, by two-tailed Student’s t test.

Figure 3. miR-128 interacts with the seed site of the coding sequence of TERT mRNA

(A) Schematic representation of the predicted and functional miR-128 binding sites in long-interspaced element-1 (LINE-1, L1) CDS RNA compared to two predicted miR-128 binding sites in the CDS of TERT mRNA. (B) Schematic representation of the two putative non-perfect 7-mer miR-128 binding sites in the coding sequence of TERT mRNA. (C) Relative luciferase levels of HeLa cell transfected with constructs expressing either TERT binding sequences at site #1 or site #2 or a perfect miR-128 seed match (perfect seed) (Wildtype (WT) seed plasmids), along with miR-control (Control) or miR-128 (miR-128) mimics, measured 48 hours post transfection. Results shown as percent change ± SEM (n = 3 independent experiments, ^*p < 0.05, ^**p < 0.01, ^****p < 0.0001) (D) Relative luciferase levels of HeLa cell transfected with constructs expressing the wildtype (WT seed plasmid) or mutated (Mutated seed plasmid) 7-mer TERT binding sequences, along with miR-control (Control) or miR-128 (miR-128) mimics, measured 48 hours post transfection. Results shown as percent change ± SEM (n = 3 independent experiments, ^**p < 0.01).

Figure 4. miR-128 directly interacts with TERT mRNA in cells

(A) Schematic of Argonaute immunopurification strategy (Ago-RIP) strategy. HeLa or Tera cell lines were generated in which miR-128 was either neutralized or over-expressed (stably transduced with anti-miR-128 or miR-128-expressing constructs). If TERT is a direct target of the miR-128/Ago complex, then Ago immunopurification in cells with neutralized miR-128 will pull out less TERT mRNA compared to miR-128 expressing cells, which will bind TERT mRNA directly. (B) Relative levels of TERT mRNA were determined by q-PCR analysis and (normalized to B2M) in “input samples” of miR modulated HeLa cells. Relative TERT mRNA levels were next determined in IP fractions and normalized to input levels. TERT IP fractions are also shown as “corrected” levels, in which IP TERT levels were corrected for levels of miR-128 in HeLa samples. As a negative control the same samples were analyzed for relative expression of GAPDH levels. Finally, Ago-RIP in Tera cells was performed as described for HeLa cells. Results from 3 independent experiments are shown as mean of IP fraction ± SEM of three independent experiments (n = 3, ^*p < 0.05, ^**p < 0.01, ^***p < 0.001, ^****p < 0.0001).