Cross-tissue comparison of telomere length and quality metrics of DNA among individuals aged 8 to 70 years

Abstract

Telomere length (TL) is an important biomarker of cellular aging, yet its links with health outcomes may be complicated by use of different tissues. We evaluated within- and between-individual variability in TL and quality metrics of DNA across five tissues using a cross-sectional dataset ranging from 8 to 70 years (N = 197). DNA was extracted from all tissue cells using the Gentra Puregene DNA Extraction Kit. Absolute TL (aTL) in kilobase pairs was measured in buccal epithelial cells, saliva, dried blood spots (DBS), buffy coat, and peripheral blood mononuclear cells (PBMCs) using qPCR. aTL significantly shortened with age for all tissues except saliva and buffy coat, although buffy coat was available for a restricted age range (8 to 15 years). aTL did not significantly differ across blood-based tissues (DBS, buffy coat, PBMC), which had significantly longer aTL than buccal cells and saliva. Additionally, aTL was significantly correlated for the majority of tissue pairs, with partial Spearman's correlations controlling for age and sex ranging from ⍴ = 0.18 to 0.51. We also measured quality metrics of DNA including integrity, purity, and quantity of extracted DNA from all tissues and explored whether controlling for DNA metrics improved predictions of aTL. We found significant tissue variation: DNA from blood-based tissues had high DNA integrity, more acceptable A260/280 and A260/230 values, and greater extracted DNA concentrations compared to buccal cells and saliva. Longer aTL was associated with lower DNA integrity, higher extracted DNA concentrations, and higher A260/230, particularly for saliva. Model comparisons suggested that incorporation of quality DNA metrics improves models of TL, although relevant metrics vary by tissue. These findings highlight the merits of using blood-based tissues and suggest that incorporation of quality DNA metrics as control variables in population-based studies can improve TL predictions, especially for more variable tissues like buccal and saliva.

Fig 1

Biological variation in aTL with chronological age (A) and tissue type (B) for individuals ranging from 8 to 70 years old. Note that buffy coat and PBMC are exclusive to child and adult cohorts, respectively.

Fig 2. Partial Spearman’s correlations of aTL among tissue types, which account for age and sex.

Ellipse shape and color denotes the strength and direction of correlations. Asterisks indicate significant p-values after adjusting for multiple comparisons using the Benjamini-Hochberg method and controlling false discovery rate (FDR) at < 0.01.

Fig 3

Variation in metrics of DNA integrity (A-D), purity (E-F), and quantity (G-I) across tissue types.

Fig 4. Partial Spearman’s correlations among DNA metrics for each tissue type, after accounting for age and sex of participants.

Spearman’s ⍴ values range from -1 to 1 on the y-axis. Asterisks indicate significant p-values after adjusting for multiple comparisons using the Benjamini-Hochberg method and controlling false discovery rate (FDR) at < 0.01.

Fig 5. Partial Spearman’s correlations between aTL and metrics of DNA integrity, purity, and quantity, adjusted for age and sex and split by tissue type.

Spearman’s ⍴ values range from -1 to 1 on the y-axis. Asterisks indicate significant p-values after adjusting for multiple comparisons using the Benjamini-Hochberg method and controlling false discovery rate (FDR) at < 0.01.