Different MT-hTers have distinct effects on cancer cell proliferation and telomere fusion.
(A and B) UM-UC-3 bladder cancer (A) and LOX melanoma (B) cells were infected with lentivirus expressing the indicated shRNAs and hTERs. Cell number was determined 8 days after infection with hTER-expressing lentivirus. Cell counts were normalized to samples treated with the control shRNA and wild-type hTER. Values indicate the mean ± standard deviation of three independent hTER infections run in parallel (A) or three independent experiments (B). Rescue of proliferation of MT-hTer-47A-treated cells by prior ATM depletion was significant in both cell lines (p < 0.001 (A) and p = 0.01 (B) with the Student's t test).
(C) UM-UC-3 cells 7 days after infection with hTER-expressing lentivirus. Phase-contrast images were obtained at identical magnification; the scale bar is 100 μm.
(D) Different types of MT-hTer-induced telomere fusions in LOX cells. DAPI-stained chromosomes were labeled with probes for wild-type telomeric sequence (red) and Alu repeats (green).
(E) LOX cell metaphase spreads were prepared 5 days after infection with hTER-expressing lentivirus. Between 45 and 67 metaphase spreads for each sample were scored for the presence of the different fusion types shown in (D). For clarity, only data for chromosome-type fusions and sister chromatid fusions are shown here. Data for other fusion types are included in Fig. S1E.
(F) Average number of different telomere fusion types per metaphase.

Proliferation of ATM-depleted cancer cells is efficiently blocked by self-complementary MT-hTers
(A) Potential intramolecular and intermolecular annealing of self-complementary MT-hTer-AU5 mutant telomeric repeats.
(B) Self-complementarity of tested MT-hTers. “Maximum length” indicates the longest stretch of uninterrupted self-complementarity within each 6-base-pair repeat, while “Total” indicates the maximum total number of self-complementary base pairs within each 6-base-pair repeat. Maximum self-complementarity was determined by considering all possible alignments of tandem telomeric repeats. The last column shows the predicted free energy of self-hybridization for 8 single-stranded tandem repeats of the indicated telomeric sequences, determined using DINAMelt (Markham and Zuker, 2005).
(C and D) UM-UC-3 (C) and LOX (D) cells were infected with lentivirus expressing the indicated shRNAs and hTERs. Cell counts were obtained 9 days after infection with the hTER-expressing virus. Values indicate the mean ± standard deviation of three independent hTER infections run in parallel. Student's t test was used to identify significant differences between matched control and ATM-depleted samples (* = p <0.05, ** = p < 0.005, *** = p < 0.0005).
(E) Correlation between predicted free energy of self-hybridization and proliferative rescue by ATM depletion. All MT-hTers from panels (C) and (D) are included except for MT-hTer-TAT4 and MT-hTer-TTCTTG, which were excluded from the graph due to their minimal effect on proliferation of UM-UC-3 control cells. For UM-UC-3 infections with MT-hTer-47A and MT-hTer-AU5, results from the two graphs in (C) were averaged and included as single points in (E). Values indicate the mean ± standard deviation of three independent hTER infections run in parallel.

Self-complementary MT-hTers induce a distinct pattern of telomere fusion.
(A) LOX cell metaphase spreads were prepared 5 days after infection with hTER-expressing lentivirus. Between 64 and 71 metaphase spreads for each sample were scored for the presence of the different fusion types shown in Fig. 1D. For clarity, only data for chromosome-type fusions and sister chromatid fusions are shown here. Data for other fusion types are included in Fig. S3A.
(B) Average number of different telomere fusion types per metaphase in LOX cells.
(C) MRC-5-TERT cell metaphase spreads were prepared 5 days after infection with hTER-expressing lentivirus. Between 25 and 35 metaphase spreads were scored for the presence of the different fusion types shown in Fig. 1D. For clarity, only data for chromosome-type fusions and sister chromatid fusions are shown. The average number of different telomere fusion types per metaphase spread is shown, while the full distribution of these fusions is presented in Fig. S3D.