The authors frame skin aging as a two-front battle. Intrinsic aging unfolds over time through genetic and metabolic factors, producing atrophy, dryness and fine lines in areas the sun never touches. Extrinsic aging is driven by the environment, and ultraviolet radiation dominates it, accounting for roughly 80 percent of visible signs of skin aging. UVB rays, spanning 280 to 320 nanometres, strike the epidermis and can directly damage DNA, while longer UVA wavelengths penetrate to the dermis and generate reactive oxygen species that corrode cellular machinery. The hallmark of this photoaging is solar elastosis, an abundance of pathologically altered elastic fibres that leaves chronically sun-exposed skin leathery and lax. Air pollution and smoking add further oxidative insult, but it is the interplay between external damage and internal cellular failure that the review places at the centre of the story.

That interplay plays out through four principal cell types. Fibroblasts, the workhorses of the dermis, synthesize type I and type III collagen and elastin, the structural proteins that keep skin firm and resilient. When fibroblasts senesce, they stop producing these components and instead secrete matrix metalloproteinases, enzymes that chew through the extracellular matrix. The result is dermal atrophy, reduced elasticity and wrinkling. Intriguingly, fibroblast senescence has a dual character in wound healing: in young mice, a transient wave of senescent fibroblasts secretes platelet-derived growth factor AA that drives myofibroblast differentiation and efficient wound closure, but as p21-positive senescent fibroblasts accumulate with age, proliferation stalls and repair is delayed. Senescent myofibroblasts also reduce expression of alpha-smooth muscle actin, a key contractile protein, undermining wound contraction and limiting fibrosis in ways that complicate healing in older tissue.

Macrophages add an inflammatory dimension. These immune cells normally clear debris and pathogens, with pro-inflammatory M1 macrophages dominating the early phase of wound repair and anti-inflammatory M2 macrophages steering the transition to tissue rebuilding. In aged skin, this balance tips: macrophages show reduced proliferative and phagocytic capacity, polarization shifts toward M1, and pro-inflammatory gene expression rises. The consequences for structure are concrete—fewer M2 cells mean reduced synthesis of collagen types I, V and VI, while excess M1 activity ramps up metalloproteinase-mediated collagen degradation. Keratinocytes, the dominant epidermal cell, slow their renewal with senescence, thinning the epidermis, increasing moisture loss and opening the door to pathogens such as Staphylococcus aureus. Epidermal stem cells, meanwhile, lose the proliferative vigour needed for re-epithelialization; reduced PGC-1α impairs their differentiation in aged mice, and an age-related decline in SIRT7 disrupts Nfatc1-mediated quiescence of hair follicle stem cells, contributing to the hair thinning so characteristic of aging skin.

At the molecular level, the review identifies reactive oxygen species and DNA damage as primary instigators. Excess ROS activate the transcription factors AP-1 and NF-κB through MAPK signalling, driving expression of matrix metalloproteinases that degrade collagen and elastin while simultaneously suppressing synthesis of new matrix components. The damage feeds on itself: fragmented collagen reduces the mechanical resistance of the extracellular matrix, lowering tension inside fibroblasts, which impairs their function and spurs further ROS production—a positive feedback loop that accelerates aging. ROS also oxidize DNA bases and break strands, and although repair pathways such as nucleotide excision repair and base excision repair normally respond, excessive oxidative stress can damage the repair machinery itself. Persistent DNA damage signalling activates ATM and ATR kinases, triggering the p53–p21 axis that locks cells into permanent cycle arrest and senescence, complete with the senescence-associated secretory phenotype.

The SASP is where individual senescent cells become a tissue-wide problem. Senescent cells resist apoptosis, evade immune clearance and broadcast inflammatory cytokines, growth factors and extracellular vesicles to their neighbours. Among the messengers, TGF-β normally drives collagen synthesis through Smad2/3 signalling, but photoaging reduces TGF-β receptor expression and UV radiation induces the inhibitory Smad7, choking off matrix production. IL-1α, stored in keratinocytes and released by UV exposure, activates the IRAK–TRAF6–TAK1 cascade, igniting NF-κB and inducing IL-6 and IL-8—while NF-κB in turn upregulates IL-1α, creating a self-amplifying inflammatory loop. Perhaps most striking are the extracellular vesicles: senescent cells release more of them, and these vesicles can carry toxic nuclear DNA fragments that induce senescence in recipient cells via the cGAS–STING pathway, as well as microRNAs such as miR-30a, miR-326-3p and miR-451a that suppress proliferation, promote fibroblast senescence and elevate ROS. Senescence, in effect, is contagious.

The review also highlights an epigenetic layer: histone modifications that govern macrophage polarization in aging wounds. The demethylase JMJD3 promotes early inflammatory M1 responses by removing the repressive H3K27me3 mark, but in diabetic wounds its sustained overexpression prolongs NF-κB-driven inflammation and delays healing. SETDB2-mediated H3K9me3 normally represses inflammatory genes and favours M2 polarization, yet it is downregulated in diabetic tissue. Persistent MOF-driven H4K16 acetylation sustains inflammatory transcription, and the age-associated histone lactylation mark H3K18La has been shown to activate NF-κB signalling at the promoters of p65 and p50. These findings suggest that aging skews the epigenetic landscape of immune cells in ways that trap wounds in a chronic inflammatory state, offering new targets for intervention.

On the therapeutic front, senotherapeutics lead the charge, divided into senolytics that kill senescent cells and senomorphics that neutralize their secretions. Navitoclax (ABT-263) inhibits anti-apoptotic BCL-2 family proteins, selectively eliminating senescent dermal fibroblasts, reducing MMP secretion and improving wound healing, though senescent melanocytes protected by MCL-1 resist it and side effects including thrombocytopenia demand caution. The flavonoids quercetin and fisetin induce apoptosis through PI3K/AKT inhibition, with the dasatinib–quercetin combination showing clinical promise for removing senescent fibroblasts, and fisetin additionally suppressing TNF-α, IL-1β, IL-6 and MMPs via NF-κB and MAPK/AP-1 blockade. Rutin, a senomorphic flavonoid, reduced wrinkle length and improved elasticity in a randomized double-blind trial while scavenging ROS and activating Nrf2 defences. Rapamycin suppresses the mTOR-driven SASP, enhances autophagy and directly activated TGF-β receptors in a trial showing increased dermal collagen and improved fine lines, while metformin activates AMPK to inhibit mTOR, reduce collagen-damaging oxidative stress and accelerate wound healing.

Stem cell therapy offers a complementary regenerative route. Mesenchymal stem cells can differentiate into keratinocyte lineages, and their secretomes—rich in GDF-11, TGF-β1, FGF family members, EGF, VEGF and HGF—stimulate fibroblast proliferation, collagen synthesis, angiogenesis and hair growth. Clinical trials using microneedling to deliver adipose-derived stem cell secretomes, umbilical cord-conditioned media and amniotic membrane preparations have reported significant improvements in wrinkles, elasticity, hydration and pigmentation. Stem cell-derived extracellular vesicles transfer TIMP1 to inhibit matrix-degrading enzymes, carry antioxidant peroxiredoxins and can push macrophages toward anti-inflammatory states. Delivery remains a bottleneck—the skin barrier limits penetration of most compounds—but liposomes, solid lipid nanoparticles, hydrogel-encapsulated systems and EV-based dual-loading carriers are extending how deeply and stably anti-aging agents reach their targets.

Perhaps the most forward-looking section concerns artificial intelligence. With traditional drug development slow and costly, machine learning models trained on labelled compound datasets are compressing discovery timelines dramatically. One XGBoost model screened more than 4,000 compounds and identified ginkgetin, periplocin and oleandrin as validated senolytics; an ensemble predictor combining a support vector machine and multilayer perceptron surfaced panaxatriol and voclosporin from hundreds of database candidates. The authors argue that integrating multi-omics data with AI could resolve the cellular heterogeneity that currently limits senolytic precision, though they caution that high-resolution spatial transcriptomic datasets for human skin remain scarce. For a field in which few candidates have reached Phase III trials, the convergence of senolytic chemistry, stem cell secretomes, nanoscale delivery and AI-driven screening may prove the combination that finally turns the biology of skin aging into treatable medicine.

Subject of Research: Molecular mechanisms of skin aging and emerging senotherapeutic, stem cell and AI-based interventions

Article Title: Skin Aging: From Molecular Mechanisms to Therapeutic and Technological Innovations

Article References: Chen, S., Yang, Y., Pan, X., Pan, Y., & Cai, S. (2026). Skin Aging: From Molecular Mechanisms to Therapeutic and Technological Innovations. Journal of Cellular and Molecular Medicine, 30(18), Article e71357. https://doi.org/10.1111/jcmm.71357

Image Credits: AI Generated

DOI: 10.1111/jcmm.71357

Keywords: skin aging, cellular senescence, SASP, senolytics, photoaging, reactive oxygen species, extracellular vesicles, macrophage polarization, mesenchymal stem cells, rapamycin, nanocarriers, machine learning