Effects of oxidative stress on telomere structure and telomeric epigenetic modifications : the role of telomere in chromosome instability

Abstract

Telomeres are nucleoprotein structures organised into heterochromatin domain and located at the end of linear chromosomes, which primary role is to maintain chromosome and genome stability. They consist of non-coding repetitive DNA sequences, such as TTAGGG in human, telomeres binding proteins and histone modifications marks. Telomeres getting shorter at each cell division, furthermore there are two different mechanisms able to maintain telomere length: I) telomerase, a ribonucleoprotein complex that regulates telomere-length maintenance by adding telomeric repeats to the chromosome 3’-end using an RNA template, II) Alternative Lengthening of Telomeres (ALT) that is a mechanism of telomere length maintenance based on recombination. Telomerase is inactive in somatic cells; however, it is active in 85% of cancer cells, instead ALT mechanism is active in the remaining 15% of cases. Oxidative DNA damage, in particular 8-oxoguanine (8-oxoG), represents the most frequent DNA damage in human cells. The high incidence of guanine residues in telomeric DNA sequences and the telomeric low efficiency in DNA damage repair, make telomere the preferential target for the accumulation of the oxidised bases. Furthermore the presence of 8-oxoG can interfere with the replication fork at telomeres; aborted replication may lead to strand breaks and the loss of telomere repeats. Previous studies stated that telomere shortening could be accelerated significantly by chemical and physical environmental agents. In order to confirm the effect of oxidative stress on telomere shortening in our experimental conditions, we treated human primary fibroblasts (MRC-5 cells) with two doses of hydrogen peroxide, 100 and 200 μM, for 1 hr. We performed the Q-FISH analysis to measure telomere length and our results confirmed a significant telomere shortening 48 hrs after treatment. At 72 and 96 hrs after treatment we observed a restoring of telomere length, indicating a transient effect of acute oxidative stress on telomere shortening. To evaluate the possibility of the activation of one of the two mechanisms of telomere maintaining, we performed the telomerase activity assay that resulted negative at each time post treatment for both doses of hydrogen peroxide, excluding that oxidative stress activated telomerase in this cells. Furthermore we performed the CO-FISH analysis to analyse the sister chromatid exchanges (T-SCE), a marker of ALT mechanism, but also in this case the results were negative indicating that hydrogen peroxide did not induced ALT activation. Therefore we speculated that there was a mechanism of cell selection that led to telomere length restoring. We performed the cell growth and the cell viability assay observing a significant growth delay starting at 48 hrs and persisting up to 96 hrs, indicating that H2O2 significantly reduced the MRC-5 proliferation rate. For all times analysed, cell viability was not different between treated and untreated cells, allowing us to exclude an effect on this end point in treated samples. The idea is that oxidative damage determines telomere shortening that in turn induces decreased growth rate and then cells with longer telomere are positively selected causing the observed telomere restoring. To evaluate and compare the genomic and telomeric DNA damage after oxidative stress we used FPG-enzyme to detect 8-oxoG base modifications by cutting the DNA in proximity of these lesions and creating SSBs. We performed two different analysis, on the whole genome and specifically at telomeric level. Results obtained by comet assay immediately after treatment, evidenced a significant increase of the genomic damage after oxidative stress, but 24 hrs later the genomic damage was completely repaired. On the other hand the analysis of the oxidative telomeric damage by quantitative PCR with telomere-specific primers demonstrated a significant increase of telomeric damage immediately after treatment whit both doses of hydrogen peroxide, and 24 hrs after treatment, conversely to the data obtained for genomic damage, we observed a significant persistent telomeric damage, leading us to hypothesise that it could be responsible of telomere shortening observed. Then we evaluated the activation of the genomic DNA damage response (DDR) by the analysis of 53BP1 foci, related with the presence of DSBs, and the phosphorylation of H2AX (ƴH2AX foci), that occurs after a DSBs but also with the block of the replication fork. We observed a significant increase of 53BP1 and ƴH2AX foci 24 hrs after treatment with both doses of hydrogen peroxide, moreover a significant increase of both foci was observed 48 hrs after 200 μM treatment. The activation of the DDR was also analysed at telomere by the immunoFISH staining that detects the telomere-dysfunction induced foci (TIFs); they correspond to the colocalisation between DNA damage markers (53BP1 and ƴH2AX) and telomeric sequences. Data demonstrated that the frequency of 53BP1 foci at telomeric level did not change at the different times post treatment; on the contrary, the frequency of ƴH2AX TIFs increased, in statistically significant manner, 48 hrs after treatment with both doses of H2O2 and after 72 hrs for the higher dose. The difference observed between the two DNA damage repair markers leads us to suppose that the principal effect of oxidative damage that occurs at telomeric level is the replication fork arrest rather than a DSBs. It is known from literature that the deletion in MEFs of TRF1 and TRF2, the most important proteins of telomeric loop, activated the DDR at telomere inducing the raise of TIFs. One other study showed that 8-oxoG in modified telomeric oligonucleotides may interfere with the recognition by TRF1 and TRF2 to telomere. Trying to elucidate the effects of oxidative stress at telomere level, we analysed telomere structure, in particular if TRF1 and TRF2 still bind telomeric sequences after oxidative stress. We performed the ChIP assay and our results demonstrated a significant decrease of both proteins 48 hrs after treatment for the two doses of hydrogen peroxide. These data allow us to hypothesise the mechanism by which oxidative stress influence telomere structure supposing that the persistent 8-oxoG at telomere could determine both the direct detachment of TRF1 and TRF2 and the stall of replication fork that in turn induce telomere shortening. The loss of telomeric repeats and the loss of protection by telomere-associated proteins could lead to dysfunctional telomeres that may give rise to chromosomes instability. Recently telomere-dysfunction dependent chromosomal instability was associated mainly by the formation of nuclear abnormalities such as micronuclei (MN), nucleoplasmic bridges (NPBs) and nuclear buds (NBUDs) and Pampalona et al., 2010 proposed NPBs as an aberration specifically linked to critically short telomeres. NPBs can break during mitosis giving rise to NBUDs and MN. For this reason we studied chromosome instability by the analysis of these abnormal nuclear morphologies at different times post treatment. The results showed an increase of NPBs 48 hrs after treatment for both doses of hydrogen peroxide that decrease in a time-dependent manner. For NBUDs and MN we observed, a fluctuating trend at 100 μM, instead at 200 μM there was a time dependent increase. These data indicate that there is an inverse relation between telomere length and NPBs indicating that this biomarker represents a good readout for telomere defects and confirming that oxidative stress is able to induce chromosome instability, in our idea through preferential telomeric damage. In the last part of this work we analysed the telomeric epigenetic modifications, in particular the density of a marker of heterochromatin, H3K9me3. We observed an increase of this marks for both doses of hydrogen peroxide, a new and unexplored effect of oxidative stress on this mark at telomere. This result could indicated more condensed status of telomeric region; in fact exist a phenomenon known as Telomere Position Effect (TPE), that is the ability of telomere to silence nearby genes. We could speculate that the observed higher chromatin condensation could influence the subtelomeric gene expression but this aspect should be deeper investigated. In conclusion, telomeric 8-oxoG not repaired that persist 24 hrs after treatment can induce telomere shortening, replication fork arrest and TRF1 and TRF2 detachment. All these in turn increase the telomere-dysfunction induced foci that lead to chromosome instability. Furthermore, oxidative stress could also be able to induce an alteration of near telomere gene expression, by inducing telomeric chromatin condensation.