The Cytoplasmic Ceiling: Golgi-Trafficking Machinery as the Division-Dependent Repository of Cellular Age Information

Background Induced pluripotent stem cell (iPSC)-derived neurons undergo complete rejuvenation regardless of donor age, while directly converted induced neurons (iNs) retain transcriptomic and functional aging signatures. Despite extensive investigation over the past decade, no mechanistic explanation has been established for why cell division-rather than transcription factor activity alone-is required to achieve cellular rejuvenation. This gap represents a fundamental barrier to developing targeted rejuvenation therapies. Methods We performed systematic reanalysis of published RNA-sequencing data from Mertens et al. (2015) (ArrayExpress E-MTAB-3037), which profiled age-associated differentially expressed (age-DE) genes across fibroblasts (n = 78 genes), iPSCs (n = 1 gene), and directly converted iNs (n = 202 genes) from donors spanning 0–89 years. We conducted hypergeometric enrichment analysis across eight literature-curated organelle gene sets, validated findings against an independent disease-focused comparison, and performed prospective hypothesis testing using GTEx v8 brain transcriptome data (n = 2,642 samples across 13 brain regions) to assess age-associated expression in human tissue. Results We identified a striking 202-fold asymmetry in age-associated gene expression between iPSCs and iNs. Systematic organelle analysis revealed that among eight cellular compartments tested, only Golgi-trafficking machinery shows statistically significant enrichment (4 observed vs. 1.95 expected genes; 2.05-fold enrichment; hypergeometric p = 0.021). The aging signature comprises seven functionally coherent genes: four Golgi/glycosylation enzymes (ST8SIA2, ST6GALNAC3, MANEA, TMED7), two vesicle trafficking scaffolds (FLNB, FLNC), and one nucleocytoplasmic transport receptor (RanBP17). None of these genes appeared among 778 genes differentially expressed between Alzheimer's disease and control iNs, demonstrating specificity for cellular age. Critically, prospective validation in GTEx brain tissue confirmed that ST8SIA2-the polysialyltransferase central to the hypothesis-shows robust age-associated decline across all brain regions (Frontal Cortex: r=-0.40, p = 1.6x10⁻⁹). The seven genes exhibit elevated co-expression (mean pairwise |r|=0.41), and a composite module score shows significant age association (r=-0.20, p = 0.004). Weaker individual correlations for other genes are predicted by the hypothesis: bulk tissue contains ~ 50% dividing glial cells that would dilute any division-dependent signal. Conclusions We propose the 'cytoplasmic ceiling' hypothesis: the Golgi apparatus and its associated trafficking machinery constitute a non-genetic, self-perpetuating repository of cellular age information that is diluted through cell division (explaining iPSC rejuvenation) but preserved in division-independent conversions (explaining iN aging retention). Independent validation of ST8SIA2 in GTEx provides the first prospective confirmation of this framework. This model provides a mechanistic explanation for the division requirement in cellular rejuvenation, unifies observations across five research domains, and identifies Golgi-targeted interventions-particularly ST8SIA2 modulation-as a novel therapeutic strategy.