ΔNp63α-mediated epigenetic regulation in keratinocyte senescence

ABSTRACT Keratinocyte senescence contributes to skin ageing and epidermal dysfunction. According to the existing knowledge, the transcription factor ΔNp63α plays pivotal roles in differentiation and proliferation of keratinocytes. It is traditionally accepted that ΔNp63α exerts its functions via binding to promoter regions to activate or repress gene transcription. However, accumulating evidence demonstrates that ΔNp63α can bind to elements away from promoter regions of its target genes, mediating epigenetic regulation. On the other hand, several epigenetic alterations, including DNA methylation, histone modification and variation, chromatin remodelling, as well as enhancer-promoter looping, are found to be related to cell senescence. To systematically elucidate how ΔNp63α affects keratinocyte senescence via epigenetic regulation, we comprehensively compiled the literatures on the roles of ΔNp63α in keratinocyte senescence, epigenetics in cellular senescence, and the relation between ΔNp63α-mediated epigenetic regulation and keratinocyte senescence. Based on the published data, we conclude that ΔNp63α mediates epigenetic regulation via multiple mechanisms: recruiting epigenetic enzymes to modify DNA or histones, coordinating chromatin remodelling complexes (CRCs) or regulating their expression, and mediating enhancer-promoter looping. Consequently, the expression of genes related to cell cycle is modulated, and proliferation of keratinocytes and renewal of stem cells are maintained, by ΔNp63α. During skin inflammaging, the decline of ΔNp63α may lead to epigenetic dysregulation, resultantly deteriorating keratinocyte senescence.


Introduction
The largest organ of the human body, skin, acts as a protective barrier on our surface. It is also responsible for maintaining homoeostasis of various substances as well as sensory perception and temperature regulation [1,2]. Keratinocyte is the predominant cell type in the epidermis, which is the outermost layer of the skin. Senescence of keratinocytes is an important part of skin ageing, which may impair the functions of skin. Numerous studies show that epigenetic alterations are involved in skin ageing and cell senescence [3][4][5]. p53 family proteins are key transcription factors controlling cell cycle [6]. p63 belongs to p53 gene family and encodes multiple protein isoforms, among which ΔNp63α is the predominant species expressed in the skin tissue [7]. As a key transcription factor in keratinocyte differentiation and renewal, ΔNp63α plays key roles to maintain proliferation of keratinocytes and to prevent them from undergoing senescence. It is traditionally accepted that ΔNp63α exerts proliferative functions via binding to promoter regions of various cell cycle arresting genes (e.g., Pten, p16 INK4a and p21 WAF1/CIP1 ) and consequently downregulating their expression [7][8][9]. Recently, ΔNp63α has been found to act as a pioneer transcription factor to regulate gene expression via binding to regions other than promoters [10,11]. In this review, we focus on the epigenetic regulation of ΔNp63α in preventing keratinocyte senescence.

ΔNp63α and senescence of keratinocytes
The skin consists of three layers, the epidermis, dermis, and hypodermis, each of which is separate but functionally interdependent. The outermost layer, epidermis, is predominantly composed of keratinocytes, which provide hardness and moisture-lock properties to the skin [1]. Keratinocytes are continuously proliferating to renew the outer skin barrier. They migrate upwards and differentiate into cells that comprise the epidermis [12][13][14].
Skin undergoes intrinsic or chronological ageing, as a consequence of internal and external changes due to passage of time [15]. Owing to extensive contact with the outside environment, epidermal tissues (especially keratinocytes) are continuously exposed to diverse external stimuli, which can cause inflammaging [16][17][18]. Alongside inflammaging of the skin, the chronic low-level pro-inflammatory cytokines, in concert with the external stimuli, may induce and deteriorate senescence of keratinocytes and other skin-resident stromal cells. As a continuously renewing epithelium, the epidermis relies heavily on the proliferative potential of keratinocytes to work as a functional barrier of our body [19]. Senescence of keratinocytes may lead to decline of their physiological functionality, which accelerates the ageing process of skin tissues [13,20]. In particular, these senescent keratinocytes produce a complex secretome, which is termed senescence-associated secretory phenotype (SASP) and leads to some detrimental effects, such as paracrine senescence, immune evasion, inflammation, and tumorigenesis (depicted as Figure 1a) [21].
Epidermal stem cells are important for replacing the cells lost during keratinocyte differentiation and senescence. It is documented that ΔNp63α keeps the regenerative capacity of stratified epithelial structure via maintaining epithelial stemness [37]. ΔNp63α is essential for the asymmetric division of epithelial stem cells to maintain the stem cell pool [38], which can be exhausted via symmetric division into transit-amplifying cells in the absence of p63 gene, resulting in epithelial defects [39][40][41]. ΔNp63α may contribute to the stemness phenotype via controlling multiple signalling pathways (depicted as Figure 1b): (1) transactivating Shh, Gli2 and Ptch1 to activate the Hedgehog (Hh) pathway [42,43]; (2) stimulating the Notch pathway via positively regulating the expression of a number of components of this pathway, including Notch1, Notch3, Jag1, Jag2, and Hes1 [39,[44][45][46][47]; (3) activating Wnt signalling pathway through the direct upregulation of transcription of the Fzd7 receptor as well as the Wnt5a ligand [48]. Via modulating these signalling pathways in different scenarios, ΔNp63α governs the balance between self-renewing and transit-amplifying cells [37,49].
The essential roles of ΔNp63α in antagonizing keratinocyte senescence are supported by data from mouse models and clinical studies. Keyes WM et al. found that accelerated skin ageing is induced in mice conditionally deleted with p63 in the epidermis, where ΔNp63α is the predominant isoform [50]. Melino G group demonstrated that the expression of p63 in skin keratinocytes of elderly people is significantly lower than that of youths, and downregulation of ΔNp63α accelerates skin ageing [19]. In one of our recent studies, we found that ΔNp63α is downregulated in either mouse skin or human keratinocytes during photoaging induced by ultraviolet B (UVB) [51].
Downregulation of ΔNp63α during skin ageing may be due to the activation of Akt/Foxo3a, p38 MAPK or other signalling pathways mediated by ROS and DNA damage, according to the experimental results of our laboratory and other research groups (depicted as Figure 1c) [31,[51][52][53]. Ultraviolet radiation (UVR) can activate p38 MAPK pathway, which in turn phosphates and stabilizes p53 in keratinocytes; consequently, p53 binds to ΔNp63α and direct a caspase-1-mediated degradation of the latter protein [52]. In our previous studies, we demonstrated that the treatment with DNA damage drugs down-regulates ΔNp63α at the mRNA level, which is independent of p53 [31]. Recently, it is reported that UVB radiation can activate PI3K/Akt pathway [54], which may increase FOXO3a phosphorylation and nuclear export, resulting in downregulation of ΔNp63α transcription [53].

Epigenetics in cell senescence
Epigenetics is a class of mechanisms generating heritable phenotypic changes independently of DNA sequence changes. Mechanisms of epigenetics include DNA and histone modifications, chromatin remodelling, as well as other processes such as enhancer-promoter looping. Via these mechanisms, environment can regulate gene expression to produce visible phenotypes [55,56]. As a state of cell cycle arrest is regulated by a network of molecules, cell senescence is driven by diverse epigenetic changes [5,57].

DNA methylation
The templates of transcription, DNA, randomly contains cytosines, which can be potentially methylated by DNA methyltransferases (DNMTs), i.e., DNMT1, DNMT3A and DNMT3B. 5-methyl cytosine (5mC) is enriched in CpG dinucleotides and directly affects the ability of transcription factors and other DNA binding proteins to access DNA. Therefore, hypermethylation of CpGs in gene promoters generally results in silencing or low expression of downstream genes, while hydroxymethylation upregulates gene transcription [58,59].

Histone modification
In eukaryotic chromatin, DNA strands are packaged into nucleosomes by histones including H2A, H2B, H3, and H4 [68]. Diverse modifications, including different types of acetylation and methylation at specific residues of histones, can change chromatin figuration and regulate gene expression via affecting DNA accessibility for the transcription machinery [69].

Enhancer-promoter looping
Chromatin structure in mammals can also regulate transcription by modulating three-dimensional interactions between enhancers and promoters, which can be facilitated by chromatin loops [94]. CCCTC-binding factor (CTCF) is a transcription factor which works together with the cohesion complex to drive the formation of chromatin loops [95]. Recently, Olan I et al reported that enhancerpromoter interactions are extensively altered during cell senescence. These rewired enhancer-promoter interactions may modulate the expression of genes related to cell senescence, such as IL1B [96].

ΔNp63α-mediated epigenetic regulation and keratinocyte senescence
As a traditional transcription factor, ΔNp63α is well established to bind to canonical p53 DNA binding sites in promoter regions of its target genes and thus compete with TAp63/p73/p53, resulting in transinhibition (Figure 2a). ΔNp63αpromoter association can also activate gene expression via recruiting other transcription factors (TFs) and RNA polymerase II (RNApol II) (Figure 2a) [7]. Recently, accumulating data demonstrate that ΔNp63α directly binds to DNA regions other than promoters and consequently regulates expression of genes related to keratinocyte differentiation and proliferation [10,[97][98][99]. This indicates that, like its homologue p53, ΔNp63α may also function as a pioneer transcription factor to modulate chromatin structure and to regulate diverse biological processes, including keratinocyte senescence [8,11,100].

ΔNp63α regulates DNA modification to promote keratinocyte proliferation
Using epigenome profiling of differentiating human primary epidermal keratinocytes, Kouwenhoven EN et al. found that ΔNp63α binds to a batch of enhancers [10]. Rinaldi L et al found that via binding to the enhancers, ΔNp63α recruits DNA methyltransferase 3A (DNMT3A) to the centre of the enhancers. Then, TET2 mediates the sequential hydroxymethylation of cytosine there. This can activate the expression of diverse genes, such as KRT5, KRT14 and ΔNp63 itself. These genes are related to differentiation and proliferation of epidermal cells [58,101]. These data indicate that ΔNp63α may promote keratinocyte proliferation and antagonize senescence via facilitating cytosine hydroxymethylation at the enhancers of proliferative genes and upregulating their expression (depicted as Figure 2b). It remains to be further investigated whether ΔNp63α regulates other types of DNA modification via recruiting other enzymes.

ΔNp63α regulates histone modifications to prevent keratinocyte senescence
Recent investigations suggest that ΔNp63α affects histone acetylation and nucleosome repositioning and changes DNA accessibility to regulate gene expression (depicted as Figure 2b). Among the ΔNp63α-bound enhancers identified by Kouwenhoven EN et al, a half is active as defined by histone modification H3K27ac, which is an active enhancer mark. These data indicate that ΔNp63α upregulates nearby genes (e.g., KRT5, KRT10 and TGM1) related to keratinocyte proliferation and differentiation via promoting histone acetylation [10]. In another study, Hamdan FH et al identified numerous ΔNp63α-occupied distal elements away from the transcription start sites of genes in pancreatic cancer cells. Many of these elements are active enhancers intersected with H3K27ac mark and locate in open chromatin regions identified by assay for transposase-accessible chromatin sequencing (ATAC-seq). This distal pattern of occupancy implies the function of ΔNp63α in enhancer activation to promote cell proliferation [102]. In line with both reports, Yu X et al found that ΔNp63α can bind to inaccessible chromatin regions, which is characterized by unmodified histones; after ΔNp63α binding, histone modifications occur in these regions [11]. These data indicate that ΔNp63α induces proliferative genes via increasing chromatin accessibility and this effect may be exerted in various situations.
ΔNp63α-induced histone modifications may be not limited to increasing chromatin accessibility. Data from Yu X et al show that histone modifications induced by ΔNp63α include not only transcriptional active histone modifications (H3K9ac, H3K27ac,  H3K4me1, H3K4me2, H3K4me3), but also repressive ones (H3K27me3 and H3K9me3) [11]. Consistent with the repression effect of p63 binding, Pattison JM et al found that p63 induces trimethylation of histone H3 lysine 27 (H3K27me3) and closes chromatin accessibility in keratinocytes [98].
These modifications may be due to ΔNp63αmediated recruitment of epigenetic enzymes. Lin-Shiao E et al found that the histone methyltransferase KMT2D interacts with ΔNp63α on chromatin and at a broad array of enhancers, resulting in histone modifications, including H3 lysine 4 monomethylation (H3K4me1) and H3 lysine 27 acetylation (H3K27ac), in keratinocytes. This may maintain transcription of hundreds of genes (e.g., KRT5, KLF4 and ZNF750), which are essential for p63 to keep proliferative capacity of keratinocytes through positively regulating genes critical for epithelial development, differentiation, and adherence [97]. Ramsey MR et al reported that ΔNp63α recruits histone deacetylases HDAC1 and HDAC2 to erase the acetylation of histones. This histone deacetylation increases histone-DNA affinity and mediates transcriptional repression of genes (e.g., Puma) responsible for cell proliferation and tumorigenesis [103,104].

ΔNp63α coordinates with chromatin remodeling complexes or regulates their expression to prevent keratinocyte senescence
ΔNp63α can also coordinate chromatin remodelling complexes (CRCs) to orchestrate the remodelling (depicted as Figure 2d). BRG1/BRM-associated factor (BAF) belongs to the SWI/SNF CRC family [105]. Using assay for transposase-accessible chromatin with high throughput sequencing (ATAC-seq), Bao X et al revealed that BAF and p63 mutually recruit each other to maintain 14,853 open chromatin regions and cooperatively position nucleosomes, resultantly recruiting transcriptional machinery and controlling transcription during epidermal differentiation. Chromatin remodelling coordinately mediated by p63 and BAF upregulates multiple genes related to epidermal differentiation and cell proliferation, including LCE3D, KRT1 and KRT10 [106].
The interaction and cooperation between ΔNp63α and CRCs in chromatin remodelling is also found in tumorigenesis. ΔNp63α may interact with actin-like protein 6A (ACTL6A), which is a subunit of SWI/SNF CRC and co-amplified with ΔNp63α in head and neck squamous cell carcinoma (HNSCC); this collaboration leads to a downregulation of tumour suppressors such as WWC1 and GPRC5A [107]. In addition, ΔNp63α can participate in the formation of another CRC called SRCAP (SNF2-related CBP Activator Protein), which belongs to the INO80 family and facilitates the substitution of histone H2A to its variant H2A.Z in nucleosomes; as a result, anti-proliferative genes (e.g., SAMD9L, ZHX2, and IGFBP3) are repressed at transcription level [108][109][110]. It remains obscure whether ΔNp63α modulates keratinocyte senescence via these mechanisms.
Apart from the direct interactions with CRCs, ΔNp63α can also modulate chromatin remodelling via regulating gene expression of multiple remodellers or related proteins (depicted as Figure 2a&d). Keyes WM et al previously reported that ΔNp63α drives proliferation of skin stem cells via transactivating lymphoid-specific helicase (LSH) [111], which belongs to the SNF2 family of chromatin-remodelling ATPase and promotes cell proliferation in multiple scenarios [112,113]. Intriguingly, Mardaryev AN et al found that ΔNp63α can bind to intron 1 of BRG1 gene and upregulate its transcription in epidermal progenitor cell during skin development [114]. As mentioned above, BRG1 is the core subunit of SWI/SNF-like BAF remodeller, which directly binds to p63 protein and cooperatively modulates DNA accessibility during epidermal differentiation [106]. These reports indicate that ΔNp63α may promote BAF-mediated chromatin remodelling from two dimensions: transactivating BRG1 gene and coordinating with the BAF complex. In addition, Fessing MY et al found that ΔNp63α controls expression of genes encoding several regulators of the higher-order chromatin structure and ATP-dependent chromatin remodelling [115]. Among which, special AT-rich binding protein 1 (SATB1) acts as a docking site for several chromatin remodelling enzymes to regulate gene expression [116]. According to Fessing MY's data, ΔNp63α directly binds to the promoter of SATB1 gene and upregulates its expression in keratinocytes, hence controlling chromatin remodelling during development of the epidermis [115].

ΔNp63α prevents keratinocyte senescence via mediating enhancer-promoter looping
Several recent studies suggest that ΔNp63α promotes keratinocyte proliferation via mediating enhancer-promoter looping in cooperation with other transcription factors (depicted as Figure 2c). Qu J et al found that ΔNp63α and CTCF are cooperatively involved in chromatin looping to regulate epidermal genes [117]. CTCF is found to consistently bind to promoters of epidermal genes (e.g., PIGV and KRT5) [118]. This consistent association of CTCF acts as a barrier for gene expression under most scenarios [94]. However, the p63-CTCF cooperation assists the looping of p63-bound enhancers to gene promoters and turns them on during keratinocyte differentiation and proliferation [117]. It is reported that CTCF mediates chromatin remodelling via physically interacting with BRD9, which is a component of non-canonical BAF complex [119]. It remains unclear whether this SWI/SNF chromatin remodelling complex is involved in p63-CTCF cooperation in enhancer-promoter looping. According to a recent study conducted by Kurinna S et al,p63 can also mediate enhancerpromoter looping in cooperation with another transcription factor, nuclear factor erythroid 2-related factor 2 (NRF2). This can activate genes including CDK12 and miR29a and promotes keratinocyte proliferation in the epidermis [99].

Discussion
Keratinocytes compose the epidermis and keep exuberant proliferating capacity to renew this skin layer [19]. Senescence of keratinocytes heavily impairs skin functionality and seriously deteriorates the skin ageing process [13,20]. As a crucial transcription factor in epidermal differentiation and cell proliferation, ΔNp63α has been reported to play key roles in delaying keratinocyte senescence [19]. Generally, ΔNp63α is assumed to specifically bind to elements in the promoter regions of its target genes to activate (e.g., Fos, c-Jun, HK2, and KRT14) or repress (e.g., Pten, p16 INK4a , p21 WAF1/CIP1 and Puma) their expression (depicted as Figure 2a) [7,9]. As a consequence, ΔNp63α promotes epidermal differentiation and keratinocyte proliferation. Mechanisms of ΔNp63α-mediated epigenetic regulation in preventing keratinocyte senescence. (a) As a traditional transcription factor, ΔNp63α binds to promoter regions, antagonizing its homologues (TAp63, p73 or p53), as well as recruiting other transcription factors (TFs) and RNA polymerase II (RNApol II), regulating genes related to cell cycle and chromatin remodeling complex (CRC). (b) ΔNp63α recruits diverse epigenetic enzymes (e.g., DNMT3A, KMT2D, HDAC1/2), leading to either active or repressive modification of DNA and histones. (c) ΔNp63α cooperates with CTCF or NRF2 to mediate enhancer-promoter looping. (d) ΔNp63α can also recruit CRCs (e.g., BAF, ACTL6A, SRCAP) to change chromatin configurations, which may either increase or decrease DNA accessibility for transcription machinery. Resultantly, proliferative genes are upregulated, while anti-proliferative genes are downregulated. Therefore, proliferation of keratinocytes is maintained and senescence is prevented.

Article highlights
• Knowledges of ΔNp63α and keratinocyte senescence were introduced • General epigenetic alterations in cellular senescence were summarized • Mechanisms of ΔNp63α-mediated epigenetic regulation in preventing keratinocyte senescence, including DNA and histone modification, enhancer-promoter looping, directly and indirectly promoting chromatin remodelling, were reviewed.

Disclosure statement
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Author contributions
LK drafted the manuscript. CL revised the manuscript. Both authors read and approved the submitted version.

Data availability statement
Not applicable.

Funding
This work was supported by National Natural Science Foundation of China (#32070747 to Li) and Department of Science and Technology of Sichuan Province (#2022NSFSC0771 to Li and #2023NSFSC1699 to Kuang).