DNA damage & repair

Nuclear DNA repair mutants exhibit progeroid symptoms. There are many types of DNA damage, and, accordingly, we have evolved mechanisms to deal with each type of damage. Nucleotide excision repair removes bulky adducts, and base excision repair removes damaged bases. Mismatch repair fixes nucleotides that aren’t matched in their correct A:T/G:C configuration. Lastly, non-homologous end joining and recombination can fix double stranded breaks. Deficiencies in several of these repair mechanisms have been implicated in aging, and they may play a role in age-related disease.

Repair mechanisms also exist for mitochondrial DNA. Mitochondria have robust base excision repair, and there is new evidence for mismatch repair. But do deficiencies in mtDNA repair play a similar role in aging? We’ve already seen that mitochondrial DNA damage accumulates with age. And calorie restriction, the gold standard of lifespan extension, prevents this increase in damage. A new report in Nucleic Acids Research looks at the controversy over when mitochondria DNA deletions occur: is it during replication or repair? Using a proofreading-deficient mitochondrial DNA polymerase (POLG), which causes premature aging phenotypes and early death, Bailey et al provide supporting evidence that the majority of damage comes from replication pausing and breakage at fragile sites.

This particular mutation in POLG results in high levels of point mutations and linear DNA (mtDNA is normally circular). Most of the linear fragments came from one particular region, and POLG is known to initiate and pause replication at specific mtDNA regions. This indicates a common location of breakage. The authors showed that the POLG mutant mouse had increased replication intermediates compared to wild type. Because of the high number of point mutations, POLG might be stalling at sites it is attempting to repair. In this manner, DNA damage is acting as a checkpoint for replication. POLG mutant mice mtDNA is also more sensitive to the single stranded nuclease S1, indicating chromosomal breakage. These single stranded ends can give rise to deletions through recombination.

The authors argue that the increased level of chromosomal breakage and the replicative pausing in the mutant mouse are responsible for the progeroid symptoms of the POLG mouse. In their view, mitochondrial DNA replication is actually upregulated in order to compensate for the reduction in replication capacity. Because of the high levels in point mutations, ox/phos activity would be decreased, which might lead to an even greater need for mitochondrial DNA replication.

And because DNA processing resources – nucleotide precursors as well as enzymes such as RNase H1, Flap endonuclease, and Brca1 – are shared by mitochondria and the nucleus, it is possible that there is a connection between the POLG mutator mouse and mutations in nuclear DNA repair proteins. The phenotype of DNA repair mutants could be caused not by mutations themselves, but by the effort it takes to prevent DNA mutation from occurring past some threshold which would cause cellular catastrophe. The authors note the similarities between POLG and WRN, a helicase in the nucleus. Like POLG, WRN is involved with both DNA replication and DNA repair. Mutations in WRN cause similar DNA breakage and lead to the human progeroid Werner syndrome.

What do you think? Is it possible that the problem in progeroid models is not due to the DNA damage itself, but to the energy required to prevent a catastrophic collapse of DNA integrity?

ResearchBlogging.orgBailey, L., Cluett, T., Reyes, A., Prolla, T., Poulton, J., Leeuwenburgh, C., & Holt, I. (2009). Mice expressing an error-prone DNA polymerase in mitochondria display elevated replication pausing and chromosomal breakage at fragile sites of mitochondrial DNA Nucleic Acids Research DOI: 10.1093/nar/gkp091


I’m sitting in an auditorium listening to a seminar by Laura Niedernhoefer from U. Pittsburgh. She’s telling us that Ercc1-/- mice, which are deficient in nucleotide excision repair, show transcriptional changes that mirror those found in old wildtype or unusually long-lived mutant animals. Her data is strongly reminiscent of recent findings that progeroid DNA repair mutants exhibit transcriptional similarities to aging calorie-restricted and dwarf animals. As in the earlier studies, the Ercc1-/- animals showed a marked downregulation of the somatotroph (GH/IGF-I) axis — suggesting that persistent DNA damage causes the body to direct energy away toward repair and away from growth.

Curiously, the Ercc1-/- mice are resistant to both cancer initiation and tumor progression — they rarely form tumors of their own, and even whopping doses of aggressive fibrosarcoma cells (from ERCC1 wildtype animals) don’t result in significant tumor growth. Largely on the basis of the latter finding, Niedernhoefer attributes the decrease in cancer to the downregulation of somatotrophic signaling; the idea here is that tumors can’t get off the ground without the cocktail of growth factors present in an animal with a normal GH/IGF-I axis.

The final chapter of the talk dealt with the mechanism of the ERCC1/XPF nuclease (an obligate heterodimer of the two proteins). Knockout animals showed a specific defect in accurately resolved experimentally induced double-stranded breaks with 3′ overhangs — other sorts of DSBs were processed equally efficiently in wildtype and knockout cells. Genetic crosses with other DSB repair mutants demonstrate that ERCC1 nuclease acts at a different step than either DNA-PKCS or Ku, two other key factors in the cellular response to DSBs.

Niedernhoefer concluded her talk with the provocative idea that DNA cross-links in real cells (as opposed to cells in an experimental setting) are largely the consequence of molecular species formed by lipid peroxidation: reactive oxygen species (ROS) oxidize polyunsaturated membrane lipids, which then fragment into reasonably stable multifunctional unsaturated compounds that are much more stable than the parent ROS and therefore more likely to make it to the nucleus, where they can modify DNA to form cross-links. Consistent with this, diets high in polyunsaturated fats dramatically decrease the lifespan of ERCC1 mutants (which are sensitive to all kinds of DNA cross-links).

In past years, I was fond of comparing biogerontology to the tale of the blind men and the elephant: everyone was approaching the problem from different directions, unable to see the big picture — and reaching conclusions that had more to do with the direction of approach (i.e., initial biases) than the fundamental importance of any given observation.

But this analogy is becoming increasingly less apt, and we may be on the verge of the era of unifying theories in the biology of aging.

What causes aging? The various subfields of biogerontology answer this question in very different ways. To vastly oversimplify: In one corner we have metabolism, including the related stories of sirtuins and calorie restriction; in another corner, we have DNA damage and stochasticity of gene expression. (Already we’re seeing the unifying tendency of recent findings: a few years ago I might have put those four items in four separate corners, but on the basis of recent reports I feel comfortable starting to bin them together. There are those, however, who would argue I’m being premature if not outright inaccurate in so doing.) In both corners, one could make a legitimate claim that the phenomenon in question has serious explanatory power regarding a fundamental mechanism of aging or longevity assurance — but is there a connection between the two?

Quite possibly. A major paper from Oberdoerffer et al. has investigated the role of sirtuins (specifically, yeast Sir2 and mammalian SIRT1) in chromatin. Over the course of their elegant study, the authors trace the parallels between sirtuins’ roles in yeast and human genome stability — and in the process, build a bridge between sirtuin activity, DNA damage, and transcriptional dysregulation:

SIRT1 Redistribution on Chromatin Promotes Genomic Stability but Alters Gene Expression during Aging

Genomic instability and alterations in gene expression are hallmarks of eukaryotic aging. The yeast histone deacetylase Sir2 silences transcription and stabilizes repetitive DNA, but during aging or in response to a DNA break, the Sir complex relocalizes to sites of genomic instability, resulting in the desilencing of genes that cause sterility, a characteristic of yeast aging. Using embryonic stem cells, we show that mammalian Sir2, SIRT1, represses repetitive DNA and a functionally diverse set of genes across the mouse genome. In response to DNA damage, SIRT1 dissociates from these loci and relocalizes to DNA breaks to promote repair, resulting in transcriptional changes that parallel those in the aging mouse brain. Increased SIRT1 expression promotes survival in a mouse model of genomic instability and suppresses age-dependent transcriptional changes. Thus, DNA damage-induced redistribution of SIRT1 and other chromatin-modifying proteins may be a conserved mechanism of aging in eukaryotes.

In both yeast and mammals, Sir2/SIRT1 relocalizes from its original location to sites of DNA damage. This may be a good thing — the sirtuin appears to be involved in recruiting DNA repair factors to break sites — but it has a negative consequence: The departure of Sir2/SIRT1 from its original perch can result in transcriptional derepression (recall that SIR genes were originally identified in S. cerevisiae as silent information regulators). The extent of such derepression will depend on SIRT1 levels/activity and also the extent of the damage — but it’s fairly obvious that the resulting transcriptional changes will have a stochastic component.

This probably isn’t such a bad thing when cells are young and damage is rare — but as either chronological or replicative age increases, damage becomes more prevalent (and even persistent), and the transcriptional changes due to SIRT1 relocalization could become quite severe.

That’s only part of the story: At damage sites, SIRT1 is promoting DNA repair and genomic stability (though evidence for the latter presented here is rather indirect, inferred from tumor spectra and cancer survival in a mouse model).

At both silenced promoters and DNA breaks, more SIRT1 (or SIRT1 activity; there’s a subtle but important distinction having to do with whether the protein is playing a stoichiometric or catalytic role) appears to be a good thing: More protein means that the “dilution” resulting from SIRT1 relocalization will result in less net loss from sites of transcriptional repression; similarly, higher levels of SIRT1 result in lower levels of genomic instability in response to DNA damage, presumably because there’s more SIRT1 around to do its thing at break sites. To complete their unifying bridge, the authors close the paper with a speculation that calorie restriction’s effect on lifespan might be mediated by increasing SIRT1 levels or activity in both arenas.

According to the authors’ interpretation, SIRT1 is performing two important functions — controlling transcription and ensuring efficient DNA repair — both of which (the authors claim) are essential for longevity assurance. This seems very reasonable, though the formal logic is slightly strained: just because SIRT1 improves genomic stability, prevents transcriptional derepression and extends lifespan doesn’t mean that the (now clearly related) former two functions are in any way related to the latter one. Sorting out causation will be a challenge, requiring some high-impact genetics — perhaps involving an approach that uncouples SIRT1’s silencing function at promoters with its affinity for or activity in damage sites. Probably crazy hard, but not impossible, and logically rather important to the model.

(See also the treatment of this article at ScienceNOW.)

Sometimes I feel like our field produces review articles faster than it produces good ideas. Certainly, biogerontology generates more reviews in a given week than truly significant papers, but the same might be said of any discipline.

I’ve been ambivalent about how to deal with reviews — I’ve considered ignoring them altogether, only covering the “important ones,” link-dumping a bunch of them whenever I was too lazy to write a real post, and various other hybrid strategies. Ignoring them seemed most attractive, since our main mission at Ouroboros is to review the primary literature, so reviewing reviews seemed pointless and derivative.

But a recent reader inquiry (from one of our junior colleagues who basically wanted me to do some of their homework for them; my response was basically “read a review and make up your own mind”) reminded me of the importance of review articles: They’re a great way for scientists who aren’t already expert in a field to figure out where the important questions are. The best ones also juxtapose the most current efforts in creative and interesting ways, adding value by pointing out non-obvious connections between subfields. If read closely and attentively, reviews can be the source of great inspiration.

So rather than treating the elements of the secondary literature like second-class citizens, I’m going to start a quasi-regular feature wherein I (or one of the other writers) compile a list of the most important and interesting reviews of the last couple of weeks, and link to them without much further comment (thereby avoiding the vaguely ridiculous feeling of reviewing reviews, which would make one — what? — the “tertiary literature”?). You, the reader, can do what you wish with them. This new feature of Ouroboros begins…NOW!



DNA damage & gene expression:





TOR signaling:


Like I said, I’ll do something like this every couple of weeks, or whenever the review folder gets full. That way we’ll never fall too far behind.

In this first entry on Session II, I’ll focus on the non-telomere-related talks from this session:

Jan Vijg gave a review of his published data about stochasticity and transcriptional noise in response to aging and damage. He also described some newer analysis suggesting that transcriptional noise is present in young animals as well, but that the variation for a given gene depends on the function of that gene — in other words, different functional categories of genes generally exhibited different variation profiles.

Christian Beauséjour argued that mice do not clear DNA-damaged cells — in particular, that neither the adaptive nor the innate immune system is involved in eliminating cells exhibiting signs of persistent DNA damage. This is a foray into controversial territory: Recently published data from other labs suggests that DNA damage causes cells to express ligands for immune cells, resulting in attack and clearance by e.g. NK cells. Beauséjour could reproduce these findings in marrow stem cells (the cell type used in the earlier studies) in vitro, but observed no evidence of clearance of damaged MSCs in vivo.

Philipp Oberdoerffer (from David Sinclair’s lab) asked: What drives chromatin reorganization and how does it relate to mammalian aging? Based on findings in yeast, he looked at mammalian SIRT1 relocalization in response to genotoxic stress, and found that SIRT1 is transiently associated with DNA double-strand break sites.

Session index:

Last year we heard about the counterintuitive observation that DNA repair mutants (which exhibit premature aging and shortened lifespans) have significant phenotypic and transcriptional overlap with genetically dwarfed or calorie restricted (CR) animals (which exhibit delayed aging and extended longevity).

The interpretation of those surprising results: Both DNA repair deficiency and CR cause organisms to divert resources away from reproduction and growth, and toward maintenance and repair. (It just happens to be fruitless for the DNA repair mutants, since they’re dumping energy into a compromised pathway.) Chalk one up for the disposable soma theory of aging.

Now, a follow-up paper from (basically) the same research group compares age-related transcriptional changes between mice aging normally, prematurely or slowly. The study, which includes data from multiple organ systems across the entire lifespan, confirms and expands on the observation that progeria and extended lifespan share common phenotypic features. From Schumacher et al.:

Delayed and Accelerated Aging Share Common Longevity Assurance Mechanisms

Mutant dwarf and calorie-restricted mice benefit from healthy aging and unusually long lifespan. In contrast, mouse models for DNA repair-deficient progeroid syndromes age and die prematurely. To identify mechanisms that regulate mammalian longevity, we quantified the parallels between the genome-wide liver expression profiles of mice with those two extremes of lifespan. Contrary to expectation, we find significant, genome-wide expression associations between the progeroid and long-lived mice. Subsequent analysis of significantly over-represented biological processes revealed suppression of the endocrine and energy pathways with increased stress responses in both delayed and premature aging. To test the relevance of these processes in natural aging, we compared the transcriptomes of liver, lung, kidney, and spleen over the entire murine adult lifespan and subsequently confirmed these findings on an independent aging cohort. The majority of genes showed similar expression changes in all four organs, indicating a systemic transcriptional response with aging. This systemic response included the same biological processes that are triggered in progeroid and long-lived mice. However, on a genome-wide scale, transcriptomes of naturally aged mice showed a strong association to progeroid but not to long-lived mice. Thus, endocrine and metabolic changes are indicative of “survival” responses to genotoxic stress or starvation, whereas genome-wide associations in gene expression with natural aging are indicative of biological age, which may thus delineate pro- and anti-aging effects of treatments aimed at health-span extension.

Those last two sentences are very important, in that they address a critical issue in studies of transcription (indeed of any phenotype) as it changes with age. Given the observation that expression of gene X (or hormone Z) changes with age, one must next ask: How do we know whether this change reflects a causative feature of aging, a defensive response to another age-related change, a passive response of no great import, an epiphenomenon, or an artifact of the experimental system? (I’ve discussed this concern before, in the context of age-specific regulation of micro-RNAs.)

The authors would argue that the changes that are common to both progeroid and long-lived animals represent true protective/defensive responses to age-related stresses (according to the same logic that underlies the interpretation of the earlier work, discussed above). In contrast, those features shared between natural aging and progeria — of which there are far more — are signs of deterioration and decrepitude, and thus reflect age-related decline.

This logic is powerful: Having distinguished between these two classes of age-related transcriptional change, we’re far better equipped to start meaningfully measuring biological age.

Telomeres — the structures at the end of chromosomes — have a long history in biogerontology. Telomeres shorten with every cell division, essentially providing a “clock” that ticks down until reaching some critical length, at which point the cell will undergo the permanent growth arrest known as senescence. Even though this clock is an important tumor suppression checkpoint (because it prevents cells that have divided many times from continuing to proliferate), senescent cells themselves contribute both directly and indirectly to aging (by diminishing regenerative capacity and secreting deleterious signaling molecules, respectively). Telomere length is also a useful biomarker: it is positively correlated with life expectancy, and appears to respond to environmental influences including chronic infection and psychological stress.

Telomeres and telomerase are therefore subjects of a great deal of active study, and if one isn’t careful one can quickly fall far behind the literature — as I have. Keeping up with the pace I set yesterday, then, here are a few of the many worthy telomere papers published over the past few months. Quoted passages are excerpts from the articles’ abstracts.

Senescence and its transcriptional profile: Telomere dysfunction in human keratinocytes elicits senescence and a novel transcription profile, Minty et al.:

Transcriptional profiling of TRF2-depleted keratinocytes showed a reproducible up-regulation of several genes. … This study has thus revealed highly sensitive and specific candidate indicators of telomere dysfunction that may find use in identifying telomere-mediated keratinocyte senescence in ageing, cancer and other diseases.

Telomere clocks and biological clocks: Telomerase reconstitution contributes to resetting of circadian rhythm in fibroblasts, Qu et al.:

We found that the response of rhythmic gene expression to serum stimulation was markedly attenuated in senescent fibroblasts, telomerase-reconstituted fibroblasts reset the circadian oscillation of rhythmic gene expression … These findings suggested that telomerase reconstitution might be a good way to reset synchronization of peripheral circadian rhythms disrupted in senescent tissues.

Telomere dysfunction and progeria: Mutations in the telomerase component NHP2 cause the premature ageing syndrome dyskeratosis congenita, Vulliamy et al.:

Dyskeratosis congenita is a premature aging syndrome characterized by muco-cutaneous features and a range of other abnormalities, including early greying, dental loss, osteoporosis, and malignancy. … Most of the mutations so far identified in patients with classical dyskeratosis congenita impact either directly or indirectly on the stability of RNAs. In keeping with this effect, patients with dyskerin, NOP10, and now NHP2 mutations have all been shown to have low levels of telomerase RNA in their peripheral blood, providing direct evidence of their role in telomere maintenance in humans.

Cancer prevention: Telomere dysfunction and tumour suppression: the senescence connection, Deng et al.:

Impaired telomere function activates the canonical DNA damage response pathway that engages p53 to initiate apoptosis or replicative senescence. Here, we discuss how p53-dependent senescence induced by dysfunctional telomeres may be as potent as apoptosis in suppressing tumorigenesis in vivo.

Cirrhosis and senescence: Telomere shortening in the damaged small bile ducts in primary biliary cirrhosis reflects ongoing cellular senescence, Sasaki et al.:

Telomere shortening and an accumulation of DNA damage coincide with increased expression of p16INK4a and p21WAF1/Cip1 in the damaged bile ducts, characterize biliary cellular senescence, and may play a role in the following progressive bile duct loss in PBC.

Telomere length in schizophrenia: Short telomeres in patients with chronic schizophrenia
who show a poor response to treatment
, Yu et al.:

Compared with the control group, a significant amount of telomere shortening was found in peripheral blood leukocytes from patients with schizophrenia who experienced poor response to antipsychotics (p< 0.001). Conclusion: Shortened telomere length in chronic schizophrenia may be a trait marker caused by oxidative stress, and the ensuing cellular dysfunction may be a factor contributing to the progressive deterioration in treatment-resistant schizophrenia.

And finally, two articles about the growing body of evidence that telomerase has functions other than making new telomeres:

Moonlighting, reviewed: Actions of human telomerase beyond telomeres, Cong and Shay:

…recent studies have led some investigators to suggest novel biochemical properties of telomerase in several essential cell signaling pathways without apparent involvement of its well established function in telomere maintenance. … This review will provide an update on the extracurricular activities of telomerase in apoptosis, DNA repair, stem cell function, and in the regulation of gene expression.

What’s going on in the nucleolus?: Nucleolar localization of TERT is unrelated to telomerase function in human cells, Lin et al.:

Here, we identify that residues 965-981 of the human TERT polypeptide constitute an active nucleolar-targeting signal (NTS) essential for mediating human TERT nucleolar localization. Mutational inactivation of this NTS completely disrupted TERT nucleolar translocation in both normal and malignant human cells. Most interestingly, such a TERT mutant still retained the capacity to activate telomerase activity, maintain telomere length and extend the life-span of cellular proliferation, as does wild-type TERT, in BJ cells (normal fibroblasts).

Tomorrow: protein degradation and, if I get to it, oxidation.

Next Page »