High level of telomerase RNA gene expression is associated with chromatin modification, the ALT phenotype and poor prognosis in liposarcoma

Abstract

Telomere length is maintained by two known mechanisms, activation of telomerase or alternative lengthening of telomeres (ALT). The ALT pathway is more commonly activated in tumours of mesenchymal origin, although the mechanisms involved in the decision of a cell to activate either telomerase or ALT are unknown at present and no molecular markers exist to define the ALT phenotype. We have previously shown an association between chromatin remodelling, telomerase gene expression and ALT in cell line models. Here, we evaluate these findings and investigate their prognostic significance in a panel of liposarcoma tissue samples to understand the biology underlying the ALT phenotype. Liposarcoma samples were split into three groups: telomerase positive (Tel+); ALT positive; ALT-/Tel-. Differences in telomerase gene expression were evident between the groups with increased expression of hTR in ALT and Tel+ compared to ALT-/Tel- samples and increased hTERT in Tel+ samples only. Investigation of a small panel of chromatin modifications revealed significantly increased binding of acetyl H3 in association with hTR expression. We confirm that the presence of the ALT phenotype is associated with poor prognosis and in addition, for the first time, we show a direct association between hTR expression and poor prognosis in liposarcoma patients.

Figure 1

Telomerase gene expression in liposarcoma samples. (A) hTR expression taken as a percentage of the riboprotein S15; (B) total hTERT expression in ng per μl. The grey box defines 25 and 75% quartiles, while error bars represent the first and 99th percentiles of the distribution. Dots represent outliers and the black line defines the median of the distribution. P-values were calculated using the Kruskal–Wallis test, which tests the likelihood of all medians being the same. (C) Four major hTERT splice variant PCR products were quantified using the Agilent Bioanalyser.

Figure 2

Chromatin modifications associated with low and high hTR expressing liposarcoma samples represented as a percentage of input chromatin. (A) Association of acetyl histone H3; (B) association of acetyl histone H4; (C) association of tri-methyl K9 of H3; (D) association of tri-methyl K20 of H4. The grey box defines 25 and 75% quartiles, while error bars represent the first and 99th percentiles of the distribution. Dots represent outliers and the black line defines the median of the distribution. P-values were calculated using the Mann–Whitney test, which tests the likelihood of all medians being the same.

Figure 3

Chromatin modifications associated with liposarcoma samples expressing wild-type hTERT or not represented as a percentage of input chromatin. (A) Association of acetyl histone H3; (B) association of acetyl histone H4; (C) association of tri-methyl K9 of H3; (D) association of tri-methyl K20 of H4. The grey box defines 25 and 75% quartiles, while error bars represent the first and 99th percentiles of the distribution. Dots represent outliers and the black line defines the median of the distribution. P-values were calculated using the Mann–Whitney test, which tests the likelihood of all medians being the same.

Figure 4

Chromatin modifications at the hTR promoter in samples grouped by phenotype represented as a percentage of input chromatin. (A) Association of acetyl histone H3; (B) association of acetyl histone H4; (C) association of tri-methyl K9 of H3; (D) association of tri-methyl K20 of H4. The grey box defines 25 and 75% quartiles, while error bars represent the first and 99th percentiles of the distribution. Dots represent outliers and the black line defines the median of the distribution. P-values were calculated using the Kruskal–Wallis test, which tests the likelihood of all medians being the same.

Figure 5

Overall survival as a function of TMM. Comparison between patients with ALT−/Tel− tumours and those with telomerase-positive or ALT tumours.

Figure 6

Overall survival as a function hTR expression. Comparison among patients with tumours expressing low, intermediate and high hTR levels. High level of telomerase RNA classes were defined on the basis of its frequency distribution, using the first and the third quartiles as cutoff values.