The telomere length landscape of prostate cancer

Abstract

Replicative immortality is a hallmark of cancer, and can be achieved through telomere lengthening and maintenance. Although the role of telomere length in cancer has been well studied, its association to genomic features is less well known. Here, we report the telomere lengths of 392 localized prostate cancer tumours and characterize their relationship to genomic, transcriptomic and proteomic features. Shorter tumour telomere lengths are associated with elevated genomic instability, including single-nucleotide variants, indels and structural variants. Genes involved in cell proliferation and signaling are correlated with tumour telomere length at all levels of the central dogma. Telomere length is also associated with multiple clinical features of a tumour. Longer telomere lengths in non-tumour samples are associated with a lower rate of biochemical relapse. In summary, we describe the multi-level integration of telomere length, genomics, transcriptomics and proteomics in localized prostate cancer.

Fig. 1. Tumour telomere length (TL) is associated with genomic features.

a Spearman correlation between tumour TL and non-tumour TL. b Spearman correlation between tumour TL and TL ratio (tumour TL / non-tumour TL). c Spearman correlation between non-tumour TL and TL ratio. d Tumour TL is ranked in descending order of length (kbp). The association of tumour TL and measures of mutational burden (PGA; percent genome altered), TMPRSS2:ERG (T2E) fusion status, as well as known prostate cancer genes with recurrent copy number aberrations (CNAs), coding single-nucleotide variants (SNVs), indels (insertion and/or deletion) and genomic rearrangements (GRs) are shown. Bar plots indicate the statistical significance of each association (Methods).

Fig. 2. Mutational landscape differs with telomere length.

a, b Correlation between the number of single-nucleotide variants (SNVs) and a tumour telomere length (TL) b TL ratio. c, d Correlation between the number of indels and c tumour TL d TL ratio. e, f Correlation between the number of genomic rearrangements (GRs) and e tumour TL f TL ratio. g, h Correlation of percentage of the genome altered (PGA) and g tumour TL h TL ratio. i, j Correlation between the number of gene fusions and I tumour TL j TL ratio. Spearman’s ρ and P-values are displayed, two-sided.

Fig. 3. The genomic correlates of TERT abundance.

a Correlation of TERT RNA abundance with tumour telomere length (TL) and TL ratio. Spearman’s ρ and P-values are displayed, two-sides. b–e Correlation of TERT abundance and b the number of GRs, c number of SNVs, d number of indels and e PGA. Spearman’s ρ and P-values are displayed, two-sided. f Spearman’s correlation of significantly associated methylation probes with RNA abundance and tumour TL. Probes within the promoter are labelled in red while the rest are located in the gene body. Dot size indicated the magnitude of correlation. Background colour indicates unadjusted P-values. Methylation probes are ordered by their correlation between TERT RNA abundance from negative to positive.

Fig. 4. Association of methylation, RNA abundance, protein abundance and TL.

a Positive correlation of methylation and tumour telomere length (TL), but negative correlation of RNA and protein abundance. b Negative correlation of methylation and tumour TL, but positive correlation of RNA and protein abundance. Darker purple dots represent undetected, imputed protein abundance measures. Spearman’s ρ and P-values are displayed, two-sided.

Fig. 5. Telomere length differs by copy number status.

a, b Difference in a tumour TL and b TL ratio between samples (n = 381) with a copy number aberration and those without in prostate cancer-related genes and associated genes. Q values are from a two-sided Mann–Whitney U test and are bolded when significant (FDR < 0.05). Box plots depict the upper and lower quartiles, with the median shown as a solid line; whiskers indicate 1.5 times the interquartile range (IQR).

Fig. 6. Telomere length is associated with clinical features and biochemical relapse.

a, b Correlation of age at diagnosis with a tumour telomere length (TL) and b TL ratio. Spearman’s ρ and P-values are displayed, two-sided. c, d Correlation of pre-treatment prostate specific antigen (PSA) with c tumour TL and d TL ratio. Spearman’s ρ and P-values are displayed, two-sided. e, f Association of ISUP (International Society of Urological Pathology) grade with e tumour TL and f TL ratio. P-value is from a one-way ANOVA, n = 381. Box plots depict the upper and lower quartiles, with the median shown as a solid line; whiskers indicate 1.5 times the interquartile range (IQR). g, h Association of T category with g tumour TL and h TL ratio. P-value is from a one-way ANOVA, n = 381. Box plots depict the upper and lower quartiles, with the median shown as a solid line; whiskers indicate 1.5 times the interquartile range (IQR). On all plots, green indicates TL ratio, while orange indicates tumour TL. i, k Cox proportional hazard models were created for i non-tumour TL, j tumour TL and k TL ratio with biochemical relapse (BCR) as the endpoint. Samples (n = 290) were split into two groups based on the optimal cut point analysis (Methods).