Characterization of Cellular Senescence in Primary Human Astrocytes
Characterization of Cellular Senescence in Primary Human Astrocytes
Woodham, T. A.; Kelsey, M. M. G.; Sedivy, J. M.
AbstractSenescent astrocytes have been identified in the brains of patients with neurodegenerative disorders, including Alzheimer's disease, yet the molecular characteristics of replicative senescence in human astrocytes remain largely unexplored. Prior work has been hampered by the low proliferative capacity and limited telomere shortening of primary human astrocytes in culture. Here, we describe a culture system in which primary human astrocytes propagated under physiological (3%) oxygen reach canonical telomeric replicative senescence after extensive expansion (up to ~76 population doublings). Senescence was confirmed through multiple biomarkers, including reduced EdU incorporation, elevated senescence-associated beta-galactosidase (SA-{beta}-gal) activity, persistent DNA damage foci ({gamma}H2AX and 53BP1) predominantly localized to telomeres, and nuclear accumulation of p53. RNA sequencing across a 12-week time course revealed early upregulation of young LINE-1 (L1HS) retrotransposon transcripts, type-I interferon (IFN-I) and senescence-associated secretory phenotype (SASP) pathway genes, alongside downregulation of cell-cycle and DNA repair programs. To resolve L1HS expression at individual locus resolution, we performed Nanopore DNA sequencing to generate a custom reference genome incorporating non-reference LINE-1 insertions. Applying our TE-Seq pipeline, we identified two full-length intergenic L1HS elements consistently upregulated across the replicative senescence time course, one of which, L1HS_9q22.32_2, retained intact ORF1 and ORF2 open reading frames, indicating potential retrotransposition competence. To contextualize the astrocyte replicative senescence program, we compared it to three additional conditions. First, parallel astrocyte cultures maintained under normoxic (20%) oxygen entered senescence earlier and showed stronger SASP upregulation. Second, DNA damage-induced senescence (DDIS) triggered by etoposide treatment produced a stronger pro-inflammatory transcriptional signature than replicative senescence, including elevated IL6, IL1A, and IL1B expression. DDIS also upregulated L1HS_9q22.32_2 as well as a second intact element, L1HS_14q23.2_3, which we have previously identified among the small number of intact L1HS loci activated during replicative senescence in fibroblasts. The convergent activation of these intact elements across cell types and senescence modalities reinforces L1HS-driven IFN-I signaling as a conserved feature of the senescent program. Third, comparison with replicatively senescent fibroblasts revealed cell-type-specific SASP regulation: the pro-inflammatory cytokines IL6 and CCL2 were downregulated in senescent astrocytes relative to proliferating cells, opposite to their behavior in fibroblasts. Together, these data establish the first comprehensive transcriptomic profile of replicative senescence in human astrocytes, offering a resource for understanding brain aging and senescence-associated neurodegeneration.