Live-cell imaging reveals the dynamics and function of single-telomere TERRA molecules in cancer cells

Abstract

Telomeres cap the ends of eukaryotic chromosomes, protecting them from degradation and erroneous recombination events which may lead to genome instability. Telomeres are transcribed giving rise to telomeric repeat-containing RNAs, called TERRA. The TERRA long noncoding RNAs have been proposed to play important roles in telomere biology, including heterochromatin formation and telomere length homeostasis. While TERRA RNAs are predominantly nuclear and localize at telomeres, little is known about the dynamics and function of TERRA molecules expressed from individual telomeres. Herein, we developed an assay to image endogenous TERRA molecules expressed from a single telomere in living human cancer cells. We show that single-telomere TERRA can be detected as TERRA RNA single particles which freely diffuse within the nucleus. Furthermore, TERRA molecules aggregate forming TERRA clusters. Three-dimensional size distribution and single particle tracking analyses revealed distinct sizes and dynamics for TERRA RNA single particles and clusters. Simultaneous time lapse confocal imaging of TERRA particles and telomeres showed that TERRA clusters transiently co-localize with telomeres. Finally, we used chemically modified antisense oligonucleotides to deplete TERRA molecules expressed from a single telomere. Single-telomere TERRA depletion resulted in increased DNA damage at telomeres and elsewhere in the genome. These results suggest that single-telomere TERRA transcripts participate in the maintenance of genomic integrity in human cancer cells.

Keywords: DNA damage; TERRA; Telomeres; cancer; live-cell imaging.

Figure 1.

Generation and characterization of cell lines expressing MS2-tagged TERRA molecules from subtelomere 15q. (A) Sequencing analyses confirm the presence of 10×MS2 repeats at subtelomere 15q in two clones. Sequence alignment proceeds 5′ to 3′. The MS2 sequences and the loxp site present downstream of the 10xMS2 repeats are shown. (B) Southern blot screening of neomycin-resistant clones. Genomic DNA was digested with BamHI or NcoI restriction enzymes. After blotting, hybridization was performed using a radioactive MS2 sequence-specific probe. A single band of the expected size was detected in two clones. A schematic of subtelomere 15q containing the MS2 cassette is shown. Unprocessed images are shown in Figure S7. (C) RT-qPCR analyses of Tel15q-TERRA and Tel15q-TERRA-MS2 expression in WT cells and the two TERRA-MS2 clones. Data shown represent mean ± SD from two biological replicates. (D) RT-PCR analyses of Tel15q-TERRA-MS2 expression in WT cells and the two TERRA-MS2 clones. Genomic DNA extracted from clone 1 was used used as positive control.

Figure 2.

Imaging Tel15q TERRA-MS2 molecules in living AGS cells. (A) Images of AGS WT (control), clone 1 and clone 2 cells expressing MS2-GFP. Tel15q TERRA-MS2 foci are indicated by arrows (yellow for cluster and white for particles). Scale bar: 5 µm. (B) Two populations of Tel15q TERRA-MS2 foci are detected in AGS clones. (C) Expression of Tel15q TERRA-MS2 particles in AGS WT, clone 1 and clone 2 cells (N = 88 to 157 cells). (D) Expression of Tel15q TERRA-MS2 clusters in AGS WT, clone 1 and clone 2 cells (N = 95 to 154 cells). (E) Diffusion coefficients (in log D) of Tel15q TERRA-MS2 particles and clusters from clones 1 and 2. Diffusion coefficients of telomeres (TRF1-mCherry) and Cajal bodies (Coilin-mCherry) are included for comparison. (F) Overlay between the frequency distributions of Tel15q TERRA-MS2 clusters, particles and telomeres diffusion coefficients (log D).

Figure 3.

Colocalization between Tel15q TERRA clusters and telomeres. Quantification of colocalization events between TRF1-mCherry labelled telomeres and Tel15q TERRA-MS2 clusters observed during time-lapse imaging. TERRA-MS2 clusters are indicated by yellow arrows. Asterisk marks the co-localization event. Representative images of MS2-GFP and TRF1-mCherry signals co-localizing (bottom) or not (top) are shown. Frequency of occurrence is indicated as percentage (N = 55 cells). Scale bar: 5 µm.

Figure 4.

Validation of Tel15q MS2-TERRA depletion upon transfection of antisense oligonucleotides (ASO) in AGS clones and analyses of DNA damage by immunofluorescence experiments. (A) RT-qPCR analysis of Tel15q TERRA-MS2 expression in the two MS2 clones. RNA was extracted from cells upon 48 hours from transfection of ASO-MS2 or ASO scrambled. Values shown were normalized with actin. N = 2. (B) RT-PCR analysis of Tel15q TERRA-MS2 expression in the two MS2 clones upon transfection of the ASO, as in A. Upon RT reaction, a specific set of primers was used for PCR reaction to detect a 500 bp long fragment which includes the 10xMS2 repeats. (C) RT-qPCR analyses of Te1-2-10-13q TERRA in the two MS2 clones upon transfection of ASO, as in A. (D) TERRA-MS2 clones and AGS WT cells were transfected with MS2 antisense oligonucleotides (ASO) or scrambled ASO as negative control. Immunofluorescence experiments were performed using γH2AX antibody to analyse DNA damage. Nuclei were stained by DAPI. Representative images for each transfection are shown. Scale bar: 1 µm (E) Quantification of the number of γH2AX-positive cells upon depletion of TERRA-MS2 transcripts. At least 300 cells were analysed in two different experiments. **: p-value< 0.0002; *: p-value < 0.003. P-values were calculated using an unpaired t-test.

Figure 5.

Immunofluorescence experiments show increase in γH2AX foci co-localizing with telomeres upon depletion of TERRA-MS2 transcripts. (A) AGS WT and TERRA MS2 clones were transfected with ASO scrambled or MS2-specific ASO. Immunofluorescence experiments were performed using γH2AX and Rap1 specific antibodies. The co-localization between γH2AX and Rap1 were analysed by two colours Z-stack imaging. γH2AX foci are in green, Rap1 foci are in red. White arrows indicate representative co-localization events detected. Nuclei were stained by DAPI. (B) Quantification of colocalization occurrence between γH2AX foci and Rap1 signals. Results are shown as percentage of γH2AX-positive cells in which colocalization between γH2AX foci and Rap1 signals was detected. N = 2. (C) Quantification of the number of γH2AX foci colocalizing with Rap1 foci per cell. The Y axis shows percentage of cells where the indicated number of γH2AX foci colocalizing with Rap1 foci were detected.