In both mice and humans, cellular senescence is thought to contribute to tumor suppression and tissue aging . However, humans and mice show very different lifespans and tumor suppression capabilities. Specifically, humans can live 80 years or more and often suppress the onset of cancer for that long, whereas mice live 2-3 years and can often be afflicted by cancer in that time. That the basis for these profoundly important differences can be at least partly explained at a cellular level, rather than a systemic one, is underscored by the finding that mouse cells are much easier to transform in vitro than are human cells [28, 29]. Of note, these seminal studies were performed in fibroblasts. In light of this, it is important to describe and understand the differences between the cellular senescence programs of mouse and human fibroblasts. To that end, several differences have already been identified, notably in the relative roles of the pRB pathway and in shortened telomeres.
In this study, we have compared the regulation of the Wnt2-HIRA-SAHF pathway in senescent mouse and human fibroblasts. In doing so, we have uncovered additional differences between these two model systems of senescence. First, mouse fibroblasts differ from human fibroblasts in their ability to form robust punctate SAHF. Specifically, unlike human fibroblasts, the majority of mouse fibroblasts do not exhibit pronounced changes in chromatin as they become senescent, as judged by formation of DAPI- and macroH2A-stained foci. Interestingly, however, a considerable proportion of the population does show increased macroH2A staining, without marked change in organization revealed by DAPI-staining. A small population, which did not reach statistical significance in these studies, did appear to exhibit nuclear reorganization and chromatin compaction, based on DAPI and macroH2A staining. Second, senescent mouse fibroblasts do not recruit the histone chaperone HIRA to PML bodies. Third, MEFs that have been induced to senesce in response to activated Ras do not downregulate expression of Wnt2. Fourth, while ectopic expression of uninhibitable GSK3β in human fibroblasts is sufficient to trigger recruitment of HIRA to PML bodies and formation of SAHF in human cells, it fails to do so when ectopically expressed in MEFs. Together, these results indicate that, compared to human fibroblasts, mouse fibroblasts are markedly impaired in their ability to activate and signal through the Wnt2-HIRA-SAHF pathway.
More specifically, these results point to at least two differences between human fibroblasts and mouse fibroblasts in this Wnt2-HIRA-SAHF pathway. First, senescence of mouse fibroblasts is not associated with the trigger that is responsible for activation of this pathway in human cells, namely repression of Wnt2. Second, mouse fibroblasts and human fibroblasts also differ downstream of Wnt2, as indicated by the ability of GSK3βS9A to induce SAHF in human, but not mouse, fibroblasts. Why senescent mouse fibroblasts fail to repress expression of Wnt2, and the difference between mouse and human fibroblasts downstream of Wnt2, are issues that remain to be resolved. Regardless of the basis for these differences, it will ultimately be important to determine whether the relative ease of transformation of mouse cells, compared to human cells [28, 29], stems, at least in part, from their failure to regulate this Wnt2-HIRA-SAHF axis, a candidate tumor suppressor pathway.
The mechanism by which HIRA's recruitment to PML bodies contributes to formation of SAHF is not well understood. Previous studies have indicated that HIRA's relocation to this subnuclear organelle is required for formation of SAHF, because blocking its relocalization, either with a dominant negative HIRA mutant or with the PML-RARα fusion protein which disrupts PML bodies, abolishes formation of SAHF [20, 21]. This study further emphasizes the link between HIRA's localization to PML bodies and formation of SAHF, because the failure of mouse fibroblasts to obviously form punctate SAHF correlates with a failure of HIRA to enter PML bodies in these cells. Conversely, these results imply that recruitment of HIRA to PML bodies is not directly linked to regulation of macroH2A. This is consistent with HIRA being primarily involved in deposition of histone (H3/H4)2 heterotetramers, rather than H2A/H2B heterodimers .
In this study, we note a level of similarity between the appearance of DAPI-stained nuclei of proliferating non-senescent MEFs and senescent human cells. This similarity at the microscopic level is not founded at the molecular level. Specifically, the bright-stained DAPI puncta in mouse fibroblasts reflect the DNA sequence and structure of mouse pericentromeric heterochromatin [22, 23]. However, the SAHF of senescent human cells largely exclude pericentromeres and telomeres [20, 21]. The similarity at the microscopic level is potentially confusing, and we caution that a punctate DAPI stain should not be used to score senescence of MEFs, as it sometimes is in human cells.