Cortical neurons develop a senescence-like phenotype promoted by dysfunctional autophagy

Abstract

Senescent cells accumulate in various tissues and organs with aging altering surrounding tissue due to an active secretome, and at least in mice their elimination extends healthy lifespan and ameliorates several chronic diseases. Whether all cell types senesce, including post-mitotic cells, has been poorly described mainly because cellular senescence was defined as a permanent cell cycle arrest. Nevertheless, neurons with features of senescence have been described in old rodent and human brains. In this study we characterized an in vitro model useful to study the molecular basis of senescence of primary rat cortical cells that recapitulates senescent features described in brain aging. We found that in long-term cultures, rat primary cortical neurons displayed features of cellular senescence before glial cells did, and developed a functional senescence-associated secretory phenotype able to induce paracrine premature senescence of mouse embryonic fibroblasts but proliferation of rat glial cells. Functional autophagy seems to prevent neuronal senescence, as we observed an autophagic flux reduction in senescent neurons both in vitro and in vivo, and autophagy impairment induced cortical cell senescence while autophagy stimulation inhibited it. Our findings suggest that aging-associated dysfunctional autophagy contributes to senescence transition also in neuronal cells.

Proteostasis failure and cellular senescence in long‐term cultured postmitotic rat neurons

“…Therefore, we verified that PCNs underwent senescence and proteostasis failure during LTC. We confirmed that, as with PHNs, PCNs exhibited the conventional senescent phenotypes—SA‐β‐gal activity (Moreno‐Blas et al, ; Piechota et al, ), p16 upregulation, lamin B1 reduction, and SASP induction ( Cxcl1 , Igfbp2 , and Igfbp4 )—as well as elevated levels of nuclear REST (Piechota et al, ) by LTC (~ 28 DIV; Figure a–d). These cells also exhibited proteostasis failure, as demonstrated by the accumulation of insoluble Ub‐conjugates (Figure e).…”

“…We observed robust BafA‐dependent punctate p62 signals at 7 and 14 DIV, indicating an active autophagic flux (Figure g, h). In contrast, the BafA‐dependent signal was less pronounced at 21 DIV and was eventually lost at 28 DIV (Figure g, h), suggesting that the autophagic activity becomes impaired over time in vitro as recently reported in senescence‐like PCNs at 26 DIV (Moreno‐Blas et al, ). Together, these results suggest that proteostasis failure occurred in LTC‐PHNs as evidenced by accumulation of poly‐Ub proteins and amyloid aggregates.…”

“…Together, these results suggest that proteostasis failure occurred in LTC‐PHNs as evidenced by accumulation of poly‐Ub proteins and amyloid aggregates. It was likely caused via impairment of autophagic clearance during in vitro aging, as is observed in physiological brain aging (Moreno‐Blas et al, ; Yang et al, ).…”

Amitosenescence and Pseudomitosenescence: Putative New Players in the Aging Process

“…Although later expanded to oncogene-induced senescence (OIS) and stress-induced premature senescence (SIPS), the dogma has persisted that neurons are spared from senescence as they do not apply to the basal concepts of the paradigm. Such a view has now been challenged by the discovery that terminally differentiated neurons acquire a senescence-like phenotype as assessed by different markers, such as SA-β-Gal, 4-HNE, γH2AX, mH2A, p-p38, IL-6, and MCP-1 [2,3]. The appearance and features of cellular senescence in postmitotic or rarely dividing tissues, a phenotype recently termed 'postmitotic cellular senescence' (PoMiCS) [4] and that we here specify as 'amitosenescence' particularly for postmitotic neurons, has been demonstrated in several cell populations, including neuronal cells in tissues of the central nervous system (CNS) [2], cardiomyocytes, osteocytes, and adipocytes [5][6][7].…”

“…Whether a corresponding SASP in neurons might elicit a canonical or rather alternative factor profile, is still ill-defined. In favor of a highly cell type-specific SASP response, both senescent-like cortical neurons and cardiomyocytes apparently differ in their profile of SASP-associated components secreted, as Edn3, Tgfb2, and Gdf15 have been highlighted for cardiomyocytes [5], whereas GATA4 and MCP-1 were identified to support the SASP in cortical neurons [3]. Interestingly, in both senescent-like postmitotic cell entities classical SASP components, such as IL-1α and IL-1β, were not significantly involved [3,5].…”