Epigenetic Mechanisms, Assessment and Therapeutics of Epidermal Stem Cells in Skin Aging

Aging precipitates a significant thinning of the epidermis, documenting a 20%–30% reduction in thickness by age 70 [11]. This attenuation stems from a diminished EpSCs‐driven renewal of keratinocytes, observable in hematoxylin and eosin (HE)‐stained sections as fewer stratum corneum layers and a sparser basal cell population [12, 13]. The functional fallout is profound, as the skin's barrier weakens, leading to increased transepidermal water loss, persistent dryness (xerosis), and greater susceptibility to infections [14]. Concurrently, the dermal‐epidermal junction (DEJ) loses its characteristic undulations, flattening due to impaired basal cell turnover. This structural change, represented as a straighter DEJ contour, undermines the mechanical stability and nutrient exchange between the epidermis and dermis [15]. These deficits manifest most strikingly in wound healing. Aged mice with sluggish EpSCs mobilization exhibited closure rates that were reduced by up to 50% compared with their younger counterparts. Such delays elevate the risk of chronic wounds in the elderly [16, 17]. And external factors like ultraviolet (UV) exposure make this worse. UV accelerates EpSCs decline and ushers in photoaging presented as pronounced thinning and loss of elasticity [18, 19].

The senescence of EpSCs underpins the macroscopic changes observed in aging skin. Thus, it is vital to understand the changes in EpSCs at the cellular level. Within the basal layer, senescent EpSCs exhibit nuclear alterations, including increased chromatin condensation and nucleolar fragmentation, which appear as darker, more clumped nuclear staining [20]. These changes signal DNA damage accumulation and a shift toward a senescent state, impairing the regenerative potential of the cells. Additionally, the reduction in organelle content, particularly mitochondria and endoplasmic reticulum (ER), is inferred from paler cytoplasmic staining, suggesting diminished metabolic activity [21]. In contrast to the uniform, densely packed basal layer of youthful skin, aged skin reveals larger, and darker EpSCs nuclei along with reduced cell density [22].

EpSCs anchor to the basement membrane via integrins, but aging reduces both the expression and function of these receptors. This decline lays the groundwork for the DEJ collapse [23, 24]. Diminished α3β1 integrin levels compromise stable keratinocyte adhesion to the laminin‐rich matrix, preventing the formation of an ordered basal layer [25]. Concurrent defects in α6β4 integrin weaken hemidesmosomal attachments to Laminin‐5 and suppress Rac1 activation [26]. In the absence of these critical mechanical cues, coordinated collective migration is disrupted, resulting in erratic cell dispersal and impaired wound closure [26, 27]. Simultaneously, chronic inflammation in aged EpSCs activates the Jak‐STAT signaling pathway. This activation, in turn, drives the upregulation of EMT factors such as Snail, Zeb1, and Twist‐1, which should heighten cell motility. This potential is, however, undermined by flawed integrins that prevent the cells from securing the firm adhesion needed for effective migration [28 –30].

This incoordination is exacerbated by imbalances in intracellular signaling. In the aged, pro‐inflammatory microenvironment, Protein Kinase C (PKC) exhibits atypical activation [31], and contrary to its typical role, inhibits growth factor (e.g. EGF)‐induced keratinocyte migration [32]. Furthermore, the damage‐associated keratins KRT6 and KRT16, while critical for mechanical strength, adopt a dysregulated mode of action under chronic inflammation that disrupts directional sensing [33, 34]. As a result, EpSCs are stalled in a hypermetabolic, inflammatory state that lacks the coordination required for effective tissue repair [35, 36].

There are changes in some factors and pathways associated with EpSCs during skin aging. Elevated levels of cytokines, such as IL‐6, regulate EpSCs behavior through signaling pathways including Jak‐STAT, Hippo and Notch potentially diminishing their regenerative capacity [13, 37 –39]. The inhibition of Jak‐STAT signaling exacerbates regenerative decline, mirroring the delayed wound healing observed in aged murine skin [16, 40]. Moreover, the secretory activities of other cell types can modulate the extracellular matrix (ECM) to influence EpSCs in aged skin. For instance, fibroblasts produce migrasomes, a process mediated by Tetraspanin 4 (TSPAN4) that becomes impaired during skin aging [41]. Consequently, young fibroblast‐derived migrasomes have been demonstrated to enhance the migration of senescent HaCaT cells, reduce reactive oxygen species (ROS) levels and downregulate pro‐inflammatory factors including IL‐1β, IL‐6, and MMP‐14, thereby facilitating epidermal repair [42].

In human studies, the Hippo and Notch signaling pathways, which are crucial for maintaining skin integrity, become dysregulated with age [39, 43]. Aging‐associated environmental stressors such as UV radiation and oxidative stress, may lead to mutations in the LAMB3 gene, which encodes the β 3 chain. β 3 chain is a crucial component of Laminin‐332 [44, 45]. Such mutations result in a defective Laminin‐332 protein in the basement membrane, severely compromising the integrity of the ECM and the mechanosensing function of EpSCs. This loss of ECM integrity serves as a direct signal for the dysregulation of the downstream Hippo pathway [46 –48]. The activity of key Hippo effectors, yes‐associated protein (YAP) and its paralog TAZ decreases, disrupting the balance between stem cell proliferation and differentiation [38, 49]. This is exemplified in junctional epidermolysis bullosa (JEB), where YAP inactivation leads to EpSCs depletion [50]. In addition, declining Notch activity disrupts the conversion of EpSCs into mature keratinocytes [39]. These collective deficits in self‐renewal and differentiation contribute to the characteristic thinning and fragility of aged skin [51, 52]. These molecular dynamics underscore the complex interplay between cytokine signaling, ECM components, and EpSCs senescence in driving the aging phenotype (Figure 1).

There is ample evidence that epigenetic mechanisms contribute significantly to skin aging because of their sensitivity to lifestyle and environmental factor [53, 54]. Epigenetic mechanisms include histone modifications DNAm and changes in chromatin structure. Endogenous and exogenous factors collectively drive diverse epigenetic manifestations of senescent phenotypes at the cellular level during skin aging.

During senescence, the level of histone H4 on arginine 3 (H4R3me2as) demethylation diminished. Because protein arginine methyltransferase 1 (PRMT1)‐mediated asymmetric demethylation of H4R3me2as preserves the stability of H4, the reduction of H4R3me2as level in turn leads to the strengthened interaction between proteasome activator PA200 and histone H4, which ultimately catalyzes the polyubiquitin‐independent degradation of histone H4 [55]. Histones are involved in transcriptional regulation, DNA repair, DNA replication, and chromosome stability, and are a core component of nucleosomes. Thus, posttranslational modifications of histones and alterations in DNAm result in the widespread loss of heterochromatin and focal gains of heterochromatin during skin senescence [56, 57].

Meanwhile, DNAm is catalyzed by a conserved family of enzymes known as DNA methyltransferases (DNMTs), which cooperatively establish and maintain DNAm patterns during embryonic development and tissue homeostasis [58]. Among these, DNMT1 plays a pivotal role in preserving methylation patterns during cellular replication by copying methylation marks from the parental (methylated) DNA strand to the newly synthesized daughter strand [59, 60]. DNMT1 maintains a methylation state by adding methyl radicals to cytosine C5 and the 5‐methylcytosine (5‐mC) DNAm is one of the key mechanisms of epigenetics [61]. Notably, in skin epithelial cells, DNMT1 expression exhibits an age‐dependent decline, suggesting its potential contribution to epigenetic dysregulation in aging skin [62]. Beyond DNMTs, the methylcytosine dioxygenase TET3 also plays a critical role in aging through epigenetic regulation [63]. TET3 mediates active DNA demethylation by oxidizing 5‐mC, establishing and maintaining hypomethylated states at critical genomic regulatory regions [64].

UV radiation is a major driver of extrinsic aging, which causes epigenetic changes such as hypermethylation of tumor suppressor genes and hypomethylation of oncogenes [65]. Notably, UV exposure triggers global DNA hypomethylation in epidermal keratinocytes, indicative of genome‐wide epigenetic instability [64]. Furthermore, UV radiation modulates epigenetic regulation by altering the expression of key genomic regulators, including DNMT1. In UV‐irradiated human skin, upregulation of DNMT1 may drive TIMP2 promoter hypermethylation, leading to transcriptional silencing of TIMP2 [66]. In addition, poor nutrition, smoking, stress, and lack of sleep exacerbate skin aging by impairing skin repair (Figure 2).

Through epigenetic modifications such as methylation, phosphorylation, and changes in histones, the gene and transcription level also change accordingly during skin aging.

Concerning the regulation of phosphorylation states, the cellular senescence is marked by the upregulation of cyclin‐dependent kinases (CDK) inhibitors p21^CIPI and P16^INK4A (also known as CDKN1A and CDKN2A) [67], leading to the persistent hypo‐phosphorylation of RB family proteins, suppression of E2F transcription factor activity, reduced expression of the proliferation marker MKI67 [68], and subsequently cell cycle arrest in the G1 and S phase [69]. Additionally, epigenetic regulation through Sprouty1 (SPRY1) methylation contributes to the natural senescence of skin epidermal cells which are composed of ~95% of keratinocytes [64]. In the immortalized human keratinocyte cell line HaCaT and SPRY1 expression exhibited an age‐dependent increase [62].

In terms of EpSCs, a recent bioinformatics analysis study based on the analysis of the GEO database found the expressions of pro‐inflammatory genes IRF4 and Cxcl12 are upregulated in aging EpSCs, while Peg3 and Cybrd1 are down‐regulated [70]. The changes in these genes may be related to epigenetic regulation. The expression of the cell cycle inhibitor P16^INK4A is directly related to chronological aging of the skin in vivo. It has been shown that the number of cell cycle inhibitor P16^INK4A protein‐positive cells increases with aging in the epidermis, which may be related to epigenetic changes [71 –73]. Notably, senescence‐associated β‐galactosidase (SA‐β‐gal) activity is significantly upregulated during this process [74]. High mobility group box 1 (HMGB1), a member of the high mobility group protein family known for its role in modulating chromatin architecture and gene expression, is released from the nucleus into the extracellular environment during cellular senescence [75].

In addition, research findings have demonstrated that the level of SOX5 declines in a range of senescent cells and tissues [76]. Moreover, in recent years, it has been found that overexpression of transcription factor SOX5 is sufficient to trigger epigenetic and transcriptional remodeling, leading to the activation of genes such as high mobility group box 2 (HMGB2) and thus alleviating cellular senescence [76].

Epigenetic changes affect cell morphology through genes expression and transcription. Senescent EpSCs display enhanced degradation of the nuclear fiber layer and nuclear structural abnormalities. These include an enlarged nuclear size, loosened nuclear membranes, chromatin relaxation, decreased levels of Lamin B1 and Lamin‐associated peptide 2 (LAP2), and defects in most of the organelles such as lysosomes, ER, and mitochondria [67]. As a very important heterogeneous organelle for energy metabolism in the human body [77], mitochondria have attracted much attention for their relevance to stem cell senescence. Although inconsistent changes in different models, it has been shown that mitochondrial complex abnormalities lead to elevated intracellular mitochondrial NADH levels and ROS levels [78 –81].

Under the influence of epigenetics, EpSCs will ultimately affect cytokine release and pathway regulation through the above mechanisms. The aging EpSCs secrete a plethora of chemokines, cytokines, growth factors and proteases including IL‐1α, IL‐1β, IL‐6, CXCL8, CXCL10, MMP3, MMP9, and MMP1, referred to as the senescence‐associated secretory phenotype (SASP) [82]. IL‐1α can elevate the expression levels of TNF‐α, IL‐8, IL‐12, MMP‐2, MMP‐9 and IL‐1R1, as well as the ROS level [83]. In addition, aging is also marked by diminished expression of cell adhesion and ECM genes in EpSCs, notably changes in collagen type XVII α 1 (COL17A1) [84]. It has been confirmed that COL17A1 is specifically expressed in interfollicular EpSC niches and significantly reduced in naturally aged human skin in vivo [85]. As a binding partner of the aPKC‐PAR complex, COL17, has been implicated in altering cell polarity and aging of the epidermis (Table 1) [88].

Epigenetic mechanisms are recognized as highly promising approaches for assessing skin aging. Among them, DNAm is a well‐studied epigenetic modification with age‐related patterns of change that can be used as a surrogate indicator of biological aging in various tissues. The commonly used DNAm‐based age estimator is called the DNAm clock. In human tissues, epigenetic biomarkers, or biological "clocks" have been developed to reliably track actual age [89]. Epigenetic factors heavily influence aging process. Heritable changes in gene expression do not alter the underlying DNA sequence [6]. This revelation has sparked the introduction of epigenetic clocks as biomarkers of biological aging [10, 54]. The estimate of DNAm age (DNAmAge), also known as the epigenetic clock, is considered one of the most promising indices accurately measuring one's biological age [90, 91].

Horvath's clock is one of the most popular epigenetic clocks. It determines biological age in a variety of tissues, including skin, by measuring methylation at the CpG (cytosine–phosphate–guanine) sites in the human genome [7]. It has been supported by data from the field of aging [10, 92]. When it came to predicting skin age, the first generation of Horvath's clocks applied to pan‐tissues had certain limitations. A novel DNAm‐based biomarker (based on 391 CpGs) was presented using the second‐generation Horvath's clock. It was created to precisely determine the age of human fibroblasts, keratinocytes, buccal cells, endothelium cells, skin, and blood samples, which were cultivated ex vivo. The dynamic aging of these cells was properly followed by this extremely sensitive age estimator, which also showed that a constant increase in epigenetic age coincides with cell proliferation. Its use on fibroblasts from Hutchinson Gilford Progeria Syndrome patients is an example [93]. The above epigenetic clock uses a wide range of CpG sites, which is insufficiently targeted at skin cells, which may be the reason for its limitations in predicting skin age.

Therefore, to assess skin aging more accurately, incorporating additional skin cell‐specific methylation sites may be desirable. For instance, the EpSCs mentioned in this study could serve as an excellent sample source. A research team has successfully created VisAgeX, a skin‐specific epigenetic aging clock capable of detecting variations in facial visual aging patterns. Their findings show that DNAm profiles from epidermal samples can effectively predict wrinkle severity, perceived facial age, and visual aging progression [6]. These studies comprehensively illustrate how age‐related changes in DNAm might connect environmental exposures to biological skin aging. We compared the advantages and disadvantages of the updated epigenetic clocks and highlighted the significance of the specialized skin epigenetic clock (Table 2).

Based on the miRNA expression patterns of healthy individuals, a study developed an epigenetic molecular clock using machine‐learning algorithms. The models used 1856 distinct miRNAs to predict age with a mean absolute error of 10.89 years and 80% accuracy in identifying age groups. According to the findings, miRNA‐based epigenetic clocks are noticeable in the outer layers of the skin and correspond similarly to mRNA or DNAm, but they are more stable than mRNA [94].

We systematically reviewed therapeutic strategies that directly modulate epigenetic mechanisms in EpSCs. Emerging evidence indicates that certain pharmacological agents and bioactive compounds can benefit cutaneous antiaging and facial rejuvenation by targeting molecular mechanisms in EpSCs. Notably, retinoids modulate DNAm patterns to reverse age‐associated epigenetic alterations, demonstrating their capacity to reset epigenetic clocks and promote cutaneous rejuvenation [54]. In addition, some active skincare ingredients such as dihydromyricetin (DHM), curcumin, and genistein, have gained attention for their ability to inhibit DNMT1 [60, 95, 96]. It has been proven experimentally DHM showed robust inhibition of DNMT1 in biochemical assays [60]. And ectoine downregulates DNMTs (DNMT1, DNMT3a and DNMT3L), thereby reducing global DNAm and reactivating tumor‐suppressor genes. This mechanism suggests its potential as an antiaging agent [97].

In addition to sites that directly regulate epigenetics, we also have systematically summarized treatment strategies aimed at altering downstream pathways of epigenetic regulation.

Apremilast, a small molecular inhibitor of phosphodiesterase 4 (PDE4), protected EpSCs against IL‐1α‐induced impairment in capacities of EpSCs, which can offer the opportunity for cutaneous antiaging [83]. Activated Myd88/TRAF6/NF‐κB signaling pathway induced by stimulation with IL‐1α was significantly inhibited by the introduction of Apremilast [83]. And biofermented Aframomum angustifolium (BAA) extract contained specific organic acids such as lactic, gluconic, succinic acid, and polyphenols. Treating keratinocyte stem cells (KSCs)‐depleted skin equivalents with it exhibited higher mitotic activity in the epidermis basal layer including EpSCs [98].

Injectable skin fillers offer a wider range of options for delaying skin aging. The positive effects of PLLA microspheres which are increasingly favored as degradable and long‐lasting fillers on EpSCs have been shown on rat epidermis and EpSCs [99].

In terms of regulating gene expression, Amniotic membrane (AM) effectively reduces TGF‐induced phosphorylation of Smad2 and Smad3 in keratinocytes, thereby modulating the expression of cell cycle regulators CDK1A (p21) and CDK2B (p15). Given that these cell cycle factors (p21/p15) are crucial mediators of cellular senescence, their altered expression due to the interference of AM with the TGF‐β pathway suggests that AM may hold potential for preventing skin aging by delaying cellular senescence and promoting tissue repair [100]. In lipid‐deficient skin under oxidative stress, wound‐edge keratinocytes exhibit elevated p21^CIP expression, which impedes wound healing when overexpressed. Conversely, suppression of p21^CIP enhances keratinocyte proliferation and barrier repair, suggesting that small‐molecule p21^CIP inhibitors—similar to those used in chemotherapy‐resistant kidney carcinoma—could promote tissue regeneration [101]. Additionally, thyroxine (T4) treatment in organ‐cultured human skin downregulates aging‐related biomarkers, including P16^INK4A transcription [102], further supporting its potential role in mitigating age‐impaired repair mechanisms.

Triiodothyronine (T3) was discovered to dramatically increase sirt1 transcription in the human epidermis, a factor linked to genomic stability, in organ‐cultured human skin. Although proliferator‐activated receptor‐γ 1‐α (PGC1α) protein expression only demonstrated a nonsignificant rising trend, T3 also raised PGC1α mRNA levels. These modifications imply that T3 might affect the epigenetic modulators of mitochondrial activity [102].