Telomere length regulates ISG15 expression in human cells

Abstract

Endogenous genes regulated by telomere length have not previously been identified in human cells. Here we show that telomere length regulates the expression of interferon stimulated gene 15 (ISG15, 1p36.33). ISG15 expression (RNA and protein) increases in human cells with short telomeres, and decreases following the elongation of telomeres by human telomerase reverse transcriptase (hTERT). The short-telomere-dependent up-regulation of ISG15 is not mediated by replicative senescence/DNA damage signaling or type I interferons. In human skin specimens obtained from various aged individuals, ISG15 is up-regulated in a subset of cells in older individuals. Our results demonstrate that endogenous human genes can be regulated by the length of telomeres prior to the onset of DNA damage signals, and suggest the possibility that cell turnover/telomere shortening may provide a mechanism for adjusting cellular physiology. The upregulation of ISG15 with telomere shortening may contribute to chronic inflammatory states associated with human aging.

Keywords: ISG15; aging; cancer; cell turnover; inflammation; telomere position effect.

Figure 1.. Up-regulation of gene expression with telomere shortening in human fibroblasts.

mRNA levels of ISG15 and agrin were assayed by qantitative PCR in human fibroblasts with different telomere lengths. Results from at least three separate experiments are shown as means ± SEM. GAPDH was used as an internal normalization control. All values were then normalized to the level (=100%) of mRNA in young cells with long telomeres (PD18) (see Table 1). Results show increase of ISG15 and agrin expression with telomere shortening. Similar results for ISG15 were also obtained in IMR90 cells, NHK and HME epithelial cells (Supplementary Figure 1).

Figure 2.. Replicative senescence and DNA damage signaling independent up-regulation of ISG15 expression in cells with short telomeres.

(A) Short and long-term expression of hTERT rescued cells from replicative senescence. BJ cells with short telomeres (BJ-80 and BJ13-141) exhibited significant increases in the number of SA-β-Gal positive cells; whereas, the cells with long telomeres (BJ-18) did not show SA-β-Gal staining. Exogenous telomerase rapidly eliminated senescent cells (BJ13-141+6H, when only the shortest telomeres had been lengthened) as well as after bulk telomere elongation had occurred (BJ13-141+53H). Rare fields with an SA-β-Gal staining positive cell were selected for the last two images to validate the staining procedure. The number in each image is a key to the cell lines used in B-D. (B) γ-H2AX staining shows that exogenous hTERT rapidly eliminates DNA damage signalling due to short telomeres. Approximately 500 nuclei of each cell line were analyzed using Metasystems software (Metasystems, Germany). (C) Western blot shows that p21, a transcriptional target of DNA damage-induced p53 signaling, rapidly disappeared following the introduction of telomerase to elongate the shortest telomeres. (D) Q-PCR showing that ISG15 expression remained high in BJ cells rescued from replicative senescence/DNA damage signaling after only a few doublings in the presence of exogenous telomerase when telomeres were still short (ISG15, column 4), while elongation of the telomeres after 53 doublings led to decreased expression (ISG15, column 5). In contrast, elimi-nating replicative senescence/DNA damage following a short exposure to telomerase caused a decrease in the expression of agrin (agrin, column 4). Agrin thus did not meet our criteria for telomere length regulation, since its increase in old cells (agrin, columns 2&3) is secondary to senescence and/or DNA damage (column 4). (E) BJ cells overexpressing hTERT and having long telomeres express low levels of ISG15.

Figure 3.. p53 is not involved in the up-regulation of ISG15 expression in cells with short telomeres.

(A) Western blot of ISG15 in human fibroblasts with different telomere lengths. Both free and conjugated (data not shown) ISG15 increase with telomere shortening (lanes 1 and 2) and in cells with short telomere (BJ13-141 before and after expressing telomerase for 6 doublings, lanes 5 and 6). Expression of HPV16 E6, which degrades p53, had no effect on ISG15 protein expression, while elongation of telomeres by the expression of telomerase for 53 doublings returned ISG15 levels to baseline. β-Actin served as a loading control. A typical result from three independent experiments is s shown. (B) Western bolt analysis of ISG15 and total p53 protein in young and old human fibroblasts with long and short telomeres, respectively. Stable expression of shRNA led to significant (> 80%) reduction in the level of p53 protein compared to those in parental and mock infection cells in both young and old cells. The reduction of p53 protein levels had no effect on the expression of ISG15. β-actin served as a loading control.

Figure 4.. The up-regulation of ISG1 5 expression in cells with short telomeres does not depend on interferon beta1 (INFB1).

(A) Q-PCR analysis of mRNA levels of INFB1 in human fibroblasts with different telomere length (BJ-18 to BJ-83). Results from at least three separate experiments are shown as means ± SEM. GAPDH was used as an internal normalization control. All values were then normalized to the level (=100%) of mRNA in young cells (PD18) with long telomeres. Results show an increase of INFB1 expression with telomere shortening. (B) Stable knock down of INFB1 by shRNA in BJ cells with short telomeres did not reduce the expression of ISG15. mRNA levels of ISG15 and INFB1 were quantified by q-PCR. (C) Western blot showing that blocking antibodies to both IFN α and β reduced the levels of both free and conjugated ISG15 in young BJ fibroblasts, but failed to reduce expression in old cells with short telomeres. Young (BJ-18) and old (BJ-76) cells were treated with neutralizing antibodies against INFA and INFB1.

Figure 5.. ISG15 is increased in human skin with aging.

(A) Immunofluorescence staining of ISG15 in BJ cells at different population doublings. The negative sample was treated identically except no primary antibody was added. Nuclei were stained with DAPI. Staining intensity increases in cells with short telomeres. (B) Immunochemical staining illustrating an age-dependent up-regulation of ISG15 expression in the dermis of human skin tissues. 2-4 cases were examined in each group. Infant, 0-1 year old; young adult, 20-24 year old; older adult, 53-68 year old. No primary antibody was added to the negative control. (C) Quantitation of the results of all the samples described above. 8-10 random fields were counted for each sample.