Proteins induced by telomere dysfunction and DNA damage represent biomarkers of human aging and disease

Abstract

Telomere dysfunction limits the proliferative capacity of human cells by activation of DNA damage responses, inducing senescence or apoptosis. In humans, telomere shortening occurs in the vast majority of tissues during aging, and telomere shortening is accelerated in chronic diseases that increase the rate of cell turnover. Yet, the functional role of telomere dysfunction and DNA damage in human aging and diseases remains under debate. Here, we identified marker proteins (i.e., CRAMP, stathmin, EF-1alpha, and chitinase) that are secreted from telomere-dysfunctional bone-marrow cells of late generation telomerase knockout mice (G4mTerc(-/-)). The expression levels of these proteins increase in blood and in various tissues of aging G4mTerc(-/-) mice but not in aging mice with long telomere reserves. Orthologs of these proteins are up-regulated in late-passage presenescent human fibroblasts and in early passage human cells in response to gamma-irradiation. The study shows that the expression level of these marker proteins increases in the blood plasma of aging humans and shows a further increase in geriatric patients with aging-associated diseases. Moreover, there was a significant increase in the expression of the biomarkers in the blood plasma of patients with chronic diseases that are associated with increased rates of cell turnover and telomere shortening, such as cirrhosis and myelodysplastic syndromes (MDS). Analysis of blinded test samples validated the effectiveness of the biomarkers to discriminate between young and old, and between disease groups (MDS, cirrhosis) and healthy controls. These results support the concept that telomere dysfunction and DNA damage are interconnected pathways that are activated during human aging and disease.

Fig. 1.

Identification of biomarkers of aging telomere-dysfunctional mice. (A) The pictures show contour plots of the three-dimensional CE-MS of polypeptides secreted into the medium from bone-marrow cells of 2- or 12-month-old mTerc^+/+ and G4mTerc^−/− mice. Molecular mass (0.8–20 kDa) is plotted against CE migration time (20–70 min). MS signal intensity is plotted on the z axis. Note that the expression of the peptides increases in culture medium of bone-marrow cells derived from 12-month-old G4mTerc^−/− mice. (B) Western blots from total bone-marrow cells of 2- and 12-month-old mice of the indicated genotypes (n = 5 mice per group). Note that CRAMP protein shows two bands; the lower band represents the active, cleaved form of the protein, which was up-regulated in 12-month-old G4mTerc^−/− mice.

Fig. 2.

Up-regulation of marker proteins in organs and plasma of aging telomere-dysfunctional mice. (A–F) The histograms show the mRNA expression of the indicated biomarker relative to GAPDH in the indicated organs. The analysis was conducted on 2-, 12-, and 24-month-old mTerc^+/+ mice and on 2- and 12-month-old G4mTerc^−/− mice (n = 5 mice per group). The bars show mean values. Error bars show standard deviation. (G) The dot plots show the expression of the marker proteins and chitinase enzyme activity in plasma of 2-, 12-, and 24-month-old mTerc^+/+ mice, and of 2- and 12-month-old G4mTerc^−/− mice (n = 5 mice per group). The lines show mean values.

Fig. 3.

Secreted proteins of telomere-dysfunctional mice are overexpressed in presenescent human cells and in response DNA damage. BJ fibroblasts were used for the analyses. Under our laboratory conditions, the proliferation of BJ cells slowed down at PD60 (presenescence). The cells were fully senescent at PD70. (A) The histogram shows mRNA expression in BJ cells at the indicated PD (n = 3 repeat experiments). The bars show mean values; error bars show standard deviation. (B) Representative Western blots on the expression of marker proteins in BJ cells at the indicated PD (n = 3 repeat experiments). (C–E) The histograms show protein expression (C and E) and enzyme activity (D) of the indicated markers in culture medium incubated for 4 h on BJ-cells at the indicated PD (n = 3 repeat experiments). The dots show mean values; the error bars show standard deviation. Note that the expression of marker proteins increases in presenescent BJ-cells at PD60. (F) The histogram shows mRNA expression of the indicated biomarkers relative to GAPDH in irradiated (4 h after 4-Gy γ-irradiation) and nonirradiated BJ-cells at the indicated PD (n = 3 repeat experiments). The bars show mean values; error bars show standard deviation. (G) Representative Western blots on the expression of marker proteins in irradiated (4 h after 4-Gy γ-irradiation) and nonirradiated BJ-cells at the indicated PD (n = 3 repeat experiments). (H) The histograms show protein expression (CRAMP, stathmin, EF-1α) and enzyme activity (chitinase) in culture medium incubated for 4 h on irradiated (4-Gy γ-irradiation) and nonirradiated BJ-cells at the indicated PD.

Fig. 4.

Proteins induced by telomere dysfunction and DNA damage show increased expression during human aging. (A–F) The dot plots show the concentration of the indicated marker proteins in human blood plasma as measured by ELISA for CRAMP (A), EF-1α (B), and stathmin (C). (D) The dot plot shows chitinase enzyme activity in blood plasma. Note that all four marker proteins were up-regulated in unaffected aged individuals living in an elderly home (old; n = 20; mean age, 85 ± 8.1 years) and more pronounced in hospitalized geriatric patients (old patients; n = 72; mean age, 73 ± 8.5 years) compared with young individuals (young; n = 31; mean age, 30 ± 3.8 years). (E) The dot plot shows a combined analysis of the best two markers (CRAMP, chitinase) according to the regression analysis (Table S2) on the same set of samples. (F) The dot plot shows IL-6 level in plasma of the indicated cohorts. (G) The histograms show ROC analyses of the marker proteins comparing the indicated cohorts. (H) The histograms show ROC analyses of the marker proteins comparing young individuals to the combined cohort of old individuals consisting of unaffected old individuals and geriatric patients (n = 92). Note that the levels of CRAMP, EF-1α, and chitinase enzyme activity correlate better with human aging than IL-6.