Aging counts on chromosomes

The acquisition of an abnormal number of chromosomes is a hallmark of many human cancers. A new study indicates that unequal segregation of genetic material to daughter cells during cell division can also lead to premature senescence and accelerated onset of a variety of aging phenotypes.

The French comedian Maurice Chevalier once said that “growing old isn't so bad when you consider the alternative.” It is becoming clear that the accumulation of DNA damage has a key role in age-related cellular degeneration. Normal aging itself is associated with increased mutation rates and genomic instability, which may also contribute to the increased cancer risk that comes with age^1,2. Furthermore, the inactivation of certain DNA repair pathways increases the rate at which aging phenotypes develop¹. At the cellular level, one hallmark of these 'accelerated' aging models is the increase in various types of chromosomal abnormalities, including mutations, translocations, fusions and fragmented chromosomes. Damaged chromosomes are detected by a protein network that triggers either cell death (apoptosis) or a permanent arrest of cell division (senescence), thereby eliminating damaged cells that might otherwise pose a risk for malignant transformation. In addition to DNA damage, there are many other stimuli that trigger the apoptotic and senescence end-point, including oxidative stress, replicative exhaustion, telomere dysfunction and oncogene activation. On page 744 of this issue, Darren Baker and colleagues³ illustrate that the accumulation of an abnormal number of chromosomes—even if these are structurally intact—also leads to cellular senescence and progeria in a mouse with a compromised mitotic checkpoint.

The mitotic checkpoint

Mitosis is the last gate at which the cellular checkpoint machinery can ensure that a proper genomic content is passed on to both daughter cells. The spindle or mitotic checkpoint^4,5 accomplishes this task by sensing the proper attachment of microtubules to protein complexes called kinetochores, as well as by monitoring the tension between the kinetochores and sister chromatids. The checkpoint signal emanating from the kinetochores delays entry into anaphase until each chromatid is properly attached to the microtubules. The proteins belonging to this checkpoint machinery include Bub1, Bub3, CENP-E, Mad1, Mad2, BubR1 (also called Mad3) and Mps1. By perhaps two independent mechanisms, the signal generated by these proteins leads to the inactivation of a complex with E3 ubiquitin ligase activity (anaphase-promoting complex) that promotes entry into anaphase⁶.

Defects in the spindle assembly checkpoint provoke chromosome mis-segregation, aneuploidy and cell death. For example, mice lacking Mad2 have enhanced rates of chromosome gain and loss and do not survive beyond embryonic day 6.5 as a result of extensive apoptosis⁷. Similarly, recent studies have shown that targeted disruption of Bub1b (encoding BubR1) leads to early embryonic lethality in mice⁸, and that severe depletion of BubR1 or Mad2 in human cancer cells inhibits tumor growth^9,10. In contrast, heterozygosity with respect to Bub1b or Mad2l1 (encoding Mad2) results in mis-segregation of chromosomes and an increase in aneuploidy without loss of viability. As a result, Bub1b^+/− and Mad2l1^+/− mice are more susceptible to tumorigenesis^11,12. This suggests that although the complete absence of the mitotic checkpoint is incompatible with life, partial disruption can contribute to the development of cancer. Consistent with this idea, levels of mitotic checkpoints proteins are diminished, but not completely absent, in several human cancers¹³.

Stand up and be counted

The study by Baker and colleagues³ takes this issue of dosage a step further by developing a hypomorphic mouse model (with genotype Bub1b^H/H) in which BubR1 levels are minimal. Bub1b^−/H mice, which express BubR1 at 4% of normal levels, die at birth, whereas Bub1b^H/H mice (which express BubR1 at 11% of normal levels) initially appear normal but show a short lifespan and premature onset of several progeroid features, including cataracts, weight loss, muscle atrophy, thinning of the skin, reduced ability to repair wounds and infertility. Consistent with other models of accelerated aging, senescent cells accumulated in tissues as well as in mouse embryonic fibroblast (MEF) cultures from Bub1b^H/H mice. There were no detectable defects in chromosomal repair or associated structural aberrations, despite the fact that the canonical senescence response proteins p53, p21, p16 and p19 were activated. What, then, is the lesion that triggers senescence, and ultimately aging? The only discernable defect in Bub1b^H/H MEFs was a high rate of chromosomal mis-segregation and an accumulation of aneuploid cells, and this defect was even more profound in Bub1b^−/H MEFs. Furthermore, BubR1 levels decreased during normal aging in several tissues from wild-type mice. Taken together, these results suggest that instability of chromosome number, in addition to chromosomal structure, might trigger senescence and aging (Fig. 1). It is noteworthy that in contrast to most human progeroid disorders, which have defects in DNA repair¹, Down syndrome is caused by chromosome dosage imbalance (trisomy for chromosome 21) and is also associated with accelerated aging phenotypes.

Figure 1: The consequences of chromosomal instability.

The model proposes that abnormalities in chromosomal numbers might have biological consequences similar to those triggered by abnormalities in chromosomal structure: cancer and aging.

Aneuploidy in cancer and aging

It is formally possible that senescence is not triggered by aneuploidy per se, but rather by an unidentified senescence-suppressor function mediated by BubR1. After all, haploinsufficiency of BubR1 also leads to a weakened spindle checkpoint and to frequent chromosomal mis-segregation with no increase in senescence or aging¹². On the other hand, the senescence response may be triggered only when abnormalities of chromosome number increase beyond a certain threshold. This checkpoint arrest would have to be overridden in cancer cells by an antiapoptotic mutation, since these often have high rates of chromosomal gains and losses. Notably, a p53-dependent checkpoint that eliminates polyploid cells has been recently identified, and the existence of an aneuploidy checkpoint has also been suggested¹⁴. Given that the tumor suppressors p53 and p19 are upregulated in Bub1b^H/H senescent cells, it would be interesting to determine whether loss of these in Bub1b^H/H mice would relax cell cycle control, promote tumorigenesis and perhaps alleviate some of the phenotypes associated with aging.

Whether aneuploidy is a specific driving force in the development of cancer is controversial. With the demonstration that severe disruption of Bub1b leads to aging-related phenotypes, the study by Baker et al.³ now adds aneuploidy and aging to the debate about the relationship between biological cause and consequence. Regardless of these hierarchical considerations, it is apparent that although silencing of the mitotic checkpoint (by pharmacological agents, for example) may be beneficial for inhibiting the growth of cancer cells, the price we may pay is premature aging.