ATRX represses alternative lengthening of telomeres

Abstract

The unlimited proliferation of cancer cells requires a mechanism to prevent telomere shortening. Alternative Lengthening of Telomeres (ALT) is an homologous recombination-mediated mechanism of telomere elongation used in tumors, including osteosarcomas, soft tissue sarcoma subtypes, and glial brain tumors. Mutations in the ATRX/DAXX chromatin remodeling complex have been reported in tumors and cell lines that use the ALT mechanism, suggesting that ATRX may be an ALT repressor. We show here that knockout or knockdown of ATRX in mortal cells or immortal telomerase-positive cells is insufficient to activate ALT. Notably, however, in SV40-transformed mortal fibroblasts ATRX loss results in either a significant increase in the proportion of cell lines activating ALT (instead of telomerase) or in a significant decrease in the time prior to ALT activation. These data indicate that loss of ATRX function cooperates with one or more as-yet unidentified genetic or epigenetic alterations to activate ALT. Moreover, transient ATRX expression in ALT-positive/ATRX-negative cells represses ALT activity. These data provide the first direct, functional evidence that ATRX represses ALT.

Keywords: ALT; ATRX; immortalization; telomere.

Figure 1. An ATRX knockout is compatible with telomerase activity

A. PCR across the region of the ATRX gene targeted by CRISPR/Cas9 treatment in HCT116 cells shows that all clones contain modifications that result in disruption of the SmlI restriction enzyme recognition sequence compared to wild-type (WT). Note that ATRX is located on the X chromosome and HCT116 is derived from a male, making the gene hemizygous. B. PCR confirming correct gene targeting by rAAV using ATRX IntF and ATRX IntR primers (Supplementary Table 1). Wild-type cells should show a band of 1217 bp, while correctly targeted clones yield of band of 2186 bp. C. Western blot analysis was conducted to confirm the lack of ATRX expression in all CRISPR/Cas9 and rAAV correctly-targeted HCT116 clones. Ku70 was used as a loading control. Telomere maintenance status was analyzed (D. C-circle, E. TRAP and F. TRF assays) in both wild-type and ATRX-knockout HCT116 cells. DNA from U-2 OS cells was used as a positive control for the C-circle and TRF assays and a negative control for the TRAP assay. All HCT116 clones were telomerase-positive/ALT-negative by these criteria.

Figure 2. Depletion of ATRX does not induce ALT in epithelial cells

A. the indicated Bre80 epithelial cultures were analyzed for ATRX and DAXX protein expression at a time point after immortalization; γ-tubulin was used as a loading control. B. growth curves of each immortalized Bre80 T5 and Bre80 T8 cell line. The number of days in culture since transduction is indicated on the x-axis, while the y-axis indicates the number of PDs the culture has undergone post-transduction. C. telomerase activity was assessed in each immortal culture using the TRAP assay. U-2 OS and HeLa cell lysates were used as negative and positive controls, respectively. D. the presence of C-circles in each immortal Bre80 T5 and Bre80 T8 cell line was determined. DNA from GM847 cells was used as a positive control.

Figure 3. ATRX loss promotes ALT activation in breast fibroblasts

A. ATRX and DAXX protein expression was analyzed in mortal (pre-crisis) Fre80-3T and Fre80-4Tii fibroblasts transduced with the indicated vectors, as well as untransduced parental Fre80-4Tii cells. 293 cells were included as a positive control and γ-tubulin was used as a loading control. B. all immortal cultures were assessed for the levels of ATRX and DAXX protein expression by Western blot. γ-tubulin was used as a loading control and HT1080 cells were used as a positive control. C. all immortal fibroblasts lines were examined for the presence of telomerase activity using the TRAP assay. GM847 and U-2 OS cells were used as negative controls and HT1080 cells used as a positive control. D. the presence of C-circles was determined in each immortal cell line. GM847 cells were included as a positive control. Samples with C-circle levels above background (−Φ29 negative control) were regarded as ALT-positive.

Figure 4. Spontaneous loss of ATRX during immortalization

A. JFCF-6/T.1/P and JFCF-6/T.5K cells that were unmodified (parental) or transduced with an empty vector (vector), a scrambled shRNA control (sc), shATRX or shDAXX were analyzed by Western blot for the expression of ATRX and DAXX proteins. All cultures were analyzed at a mortal and immortal time point, as indicated by the “+” above each lane. γ-tubulin was used as a loading control. B. growth curves of each immortalized JFCF-6/T.1/P and JFCF-6/T.5K cell line. The days in culture and cumulative PDs were calculated subsequent to SV40 transformation.

Figure 5. ATRX loss corresponds to a period of growth crisis

ATRX and DAXX protein levels of three immortal JFCF-6/T.1/P cell lines (parental, vector and sc1) were analyzed at multiple PDs indicated on the corresponding growth curves; crisis was defined as the cell culture undergoing less than 1 PD/7 days. γ-tubulin was used as a loading control.

Figure 6. Loss of ATRX promotes fibroblast immortalization

TLM status of JFCF-6/T.1/P and JFCF-6/T.5K cell lines treated as described for Figure 4 was analyzed. A. telomerase activity was assessed using the TRAP assay. Each sample was assessed at both mortal and immortal time points, indicated by the “+” above each lane. GM847 or U-2 OS cell lysates were used as negative controls and HT1080 served as a positive control, B. the presence of C-circles was assessed at both mortal and immortal time points in each cell line. C. mean telomere length analysis using the TRF assay was performed on each cell line for at least two time points (mortal and immortal). The triangle above each set of samples indicates increasing PDs.

Figure 7. ATRX expression represses the ALT mechanism

A. ATRX and DAXX protein analysis of untreated control, FuGENE-treated, empty vector (EV)- or ATRX-transfected cultures at days 2, 4, 6 and 8 post-transfection in GM847, JFCF-6/T.5K-sc1 or U-2 OS cells. γ-tubulin was used as a loading control. The blots shown are representative of at least three separate transfections. B. the level of C-circles was assessed in each untreated control, FuGENE-treated, EV- or ATRX-transfected GM847, JFCF-6/T.5K-sc1 or U-2 OS cell culture at 2, 4, 6 and 8 days post-transfection. C-circle levels were normalized to the quantity of DNA used for each reaction, followed by normalization to the EV-transfected control at the relevant day post-transfection. Bars indicate the mean ± SEM; n = 3. *P < 0.05, **P < 0.005 and ***P < 0.0001. C. APB-positive nuclei were quantified in untreated, FuGENE-treated and EV- or ATRX-transfected GM847, JFCF-6/T.5K-sc1 or U-2 OS cells at the indicated day post-transfection. A nucleus was scored APB-positive when TRF2 and PML co-localized; ATRX/APB-positive nuclei also showed nuclear ATRX staining. The bars represent the mean ± SEM; n = 3 to 5; at least 100 nuclei were counted in the untreated, FuGENE-treated or EV-transfected cultures and at least 100 ATRX-positive nuclei were counted in the ATRX-transfected cultures. *P < 0.05, §P < 0.01, **P < 0.005, ^§§P < 0.001, ***P < 0.0005, ^§§§P < 0.0001. nd = not determined due to lack of ATRX-positive nuclei. D. mean telomere length analysis using the TRF assay in GM847, JFCF-6/T.5K-sc1 or U-2 OS cell cultures treated as indicated above each lane.