Longevity Papers

Cellular senescence refers to a permanent state of cell cycle exit and plays a key role in blocking the proliferation of compromised cells. Despite diverse drivers, senescent cells often share common features, one of which is their increased size. Though previously viewed as a passive outcome of cell growth during senescence, recent work shows that excess cell size can actively promote senescence via the p53-p21 pathway, making it both a consequence and a cause of cell cycle arrest. How cells sense their own size and communicate this information to the cell cycle control apparatus remains unresolved. Using genome-wide proximity mapping, we show that excess cell size globally redistributes chromatin between nuclear speckles and the nuclear lamina. These changes correspond with altered gene expression profiles, including the de-repression of typically silenced lamina-associated genes, some of which are involved in the senescence-associated secretory phenotype (SASP). Consistent with altered nuclear speckle function, we find that excess cell size perturbs normal splice isoform usage. In particular, MDM4, a negative regulator of p53 signaling, is differentially spliced when cells become enlarged to yield a non-functional variant, MDM4-S. Re-introducing correctly spliced MDM4 transcripts restores G1 cell cycle progression, thereby reverting senescent phenotypes in enlarged cells. Together, these data indicate that excess cell size is sensed via changes in mRNA homeostasis associated with a reorganized genome, driving both the secretory phenotype and permanent cell cycle exit that are characteristic of cellular senescence.

Autophagy is a conserved cellular recycling pathway linked to longevity and regulated by autophagy-related (ATG) protein. Many of these proteins serve a key function in ATG8 lipidation onto autophagosomal membranes, a step that precedes subsequent lysosomal degradation. Recently, mammalian ATG16L1, a member of ATG8 lipidation machinery, have been shown to also direct ATG8 lipidation onto other membranes via its C-terminal WD40 domain in acutely stressed cells. However, whether such non-canonical functions of the ATG16L1 WD40 domain influence longevity remain largely unknown. Here, we endogenously deleted the WD40 domain of atg-16.2, the C. elegans ortholog of ATG16L1, and examined autophagy, healthspan and longevity. The ATG-16.2 WD40 domain was dispensable for healthspan measures and for ATG8 lipidation in whole animals and neurons but was required for autophagic flux across all major tissues examined. Notably, the ATG-16.2 WD40 domain was also critical in limiting lysosomal membrane damage after acute heat stress. Because long-lived C. elegans mutants are typically heat resistant, we asked whether this lysosomal function supported their thermotolerance and found the WD40 domain to be selectively required for thermotolerance induced by germline removal, but not by reduced insulin/IGF-1 signaling or dietary restriction. In germline-less animals, loss of the ATG-16.2 WD40 domain reduced nuclear TFEB/HLH-30, reduced neuronal autophagic degradation, altered neuronal lysosomal pH, and sensitized animals to heat-induced lysosomal damage. Collectively, these findings establish the ATG16 WD40 domain as a regulator of autophagic degradation and lysosomal integrity in vivo, and show that this role is associated with longevity in a pathway-selective manner.

The cGAS-STING pathway is a major driver of chronic inflammation. While acute STING inhibition can reduce neuroinflammation and tissue damage, its long-term impact on longevity remains unknown. Here, we demonstrate that genetic deletion of STING extends median lifespan in mice. This survival benefit is accompanied by a marked resistance to age-related hepatic decline. Our findings demonstrate that targeting STING is a promising therapeutic target to extend healthy lifespan.

Identifying lifespan-extending compounds is an important objective in ageing research, but experimental screening remains costly and time consuming. This study evaluates supervised machine learning for predicting lifespan-extending compounds in Caenorhabditis elegans using compound-protein interaction features derived from the STITCH database. Three feature representations were compared: continuous interaction probabilities, binary features generated using a fixed threshold of 0.5, and binary features generated using an optimised threshold selected within cross-validation. Random Forest and XGBoost classifiers were evaluated using stratified 10-fold cross-validation, with performance assessed by accuracy, area under the receiver operating characteristic curve (AUROC), and area under the precision recall curve (AUPRC). Model interpretation was examined using Mean Decrease in Impurity (MDI), SHAP, TreeSHAP, and LIME, with Jaccard similarity used to compare the top-ranked protein features. Random Forest achieved its highest accuracy (78.84%), AUROC (0.7391), and AUPRC (0.5202) using the fixed threshold binary representation. XGBoost achieved its highest accuracy (80.36%) using the optimised threshold representation, whereas continuous features yielded its highest AUROC (0.7605) and AUPRC (0.5332). SHAP and TreeSHAP showed the strongest agreement in protein rankings, producing identical top-10 feature sets across all three XGBoost representations. These findings demonstrate that feature representation influences both predictive performance and model interpretation, highlighting the importance of evaluating classification metrics alongside explainability when prioritising candidate compounds for further ageing research.