Abstract The skin is composed of multiple fibroblast subpopulations with different functions in homeostasis and repair, but their role in skin diseases is largely unknown. Junctional epidermolysis bullosa (JEB) is a hereditary skin disorder characterised by severe skin fragility and aberrant granulation tissue formation, caused by loss-of-function variants in basement membrane proteins, including laminin-332. We developed JEB-like organotypic (OT) cultures with distinct fibroblast subpopulations and explored their role in an inducible JEB in vivo disease model, mimicking key features of the human disease. Mechanistically, papillary fibroblasts are highly increased in the granulation tissue of blistered JEB skin, promoting pathological αvβ6 integrin and TGFβ signalling in JEB keratinocytes. Treatment with the TGFβ receptor inhibitor RepSox not only normalised aberrant cell proliferation, differentiation, and cytokine signalling in JEB OTs but also reduced aberrant granulation tissue formation and skin blistering in laminin-332-depleted mice. Collectively, our study reveals that papillary fibroblasts promote JEB pathogenesis through increasing αvβ6 integrin and TGFβ signalling and disruption of these pathological signalling interactions significantly improved skin health and regeneration in JEB.
Abstract Introduction and aims Lipid nanoparticles (LNPs) are an attractive platform for delivering nucleic acid-based therapeutics, including gene editors (GEs), but their potential in treating skin diseases remains underinvestigated. Dystrophic epidermolysis bullosa (DEB) is a genetic skin disease caused by pathogenic variants in COL7A1, leading to collagen VII deficiency, chronic blistering, nonhealing wounds and eventually skin cancer. We hypothesize that LNP-mediated delivery of GEs, either topically or systemically, can treat DEB. Building on previous work demonstrating successful delivery of GEs to DEB fibroblasts using cationic LNPs, this work investigates the use of ionizable LNPs for mRNA-based GE delivery. Methods Ionizable LNPs encapsulating fluorescent reporter mRNA were formulated using ionizable lipids (MC3, DODMA, or ALC-0315), phospholipids (DOPE or DSPC), cholesterol, and PEG-2000-DMG with varying component ratios. The LNP’s ability to deliver fluorescent reporter mRNA was evaluated in vitro in primary human fibroblasts and NTERT keratinocytes, and three-dimensional (3D) skin constructs and in vivo onto wounds on a DEB mouse model. Results The LNPs exhibited favourable physicochemical properties, with small particle sizes, low polydispersity and high encapsulation efficiency. We identified several candidate LNPs, with ALC0315 or MC3 as ionizable lipids, capable of delivering fluorescent reporter mRNA to cells in vitro. Moreover, we have shown that efficacy and cell type selectivity is influenced by the choice of helper lipid and component ratios. For example, ALC0315/DOPE performs best in fibroblasts while ALC0315/DSPC favours keratinocyte delivery. We have begun applying the top LNP candidates in 3D human skin equivalents and into wounds on a DEB mouse model suitable for GE testing. Our preliminary results indicate successful delivery of LNPs into basal keratinocytes in the 3D constructs and into migrating cells within the mouse wound bed during healing. Conclusions These findings support the potential of LNPs as a versatile and translatable platform for GE therapies in DEB, providing a foundation for future therapeutic development.
Epidermolysis bullosa (EB) is a family of rare, incurable, inherited disorders characterized by extreme skin and mucosal fragility. The two most severe forms are dystrophic EB (DEB), caused by genetic variance in type-VII collagen, and junctional EB (JEB), caused by genetic variance in genes encoding basement membrane proteins laminin 332 (α3, β3, γ2), type XVII collagen or integrin α6β4. Despite clinical similarities, there are dramatic differences in the life expectancies of patients with these subtypes. Patients with RDEB survive to early or midadulthood, comparatively patients with JEB rarely survive beyond the first year of life. We aim to explore the disparity in life expectancy using a transcriptomic and an in vivo approach. We performed bulk RNA sequencing on 7 DEB and 13 JEB samples (3 LAMA3, 8 LAMB3 and 2 LAMC2) to identify molecular and cellular differences underpinning these diseases. Principal component analysis and differential gene expression (DEG) analysis was conducted to explore the relatedness between subtypes. Immunohistochemistry was used to validate these findings in mouse models of EB. Three distinct clusters of samples were grouped by condition; one DEB and two JEB clusters separated by LAMA3 and LAMC2 mutations. Pathway and gene ontology (GO) analysis of DEGs indicated an upregulation in cell death, proinflammatory pathways and interleukins (ILs) (IL-1, IL-11, CXCL12 and β-defensins) which stimulate the adaptive immune response. In addition, an abundance of transcriptional factors linked to wound healing (MMP2, COL1A1), collagen remodelling (LOX), fibrosis (TGFβ) and epidermal development (EPPK1, TGM1, FLG and KRT14) were upregulated in patients with JEB compared with those with DEB. These changes have been validated in mouse models of EB at Queen Mary University of London. We have identified and validated significant changes between EB subtypes. The use of pathway and GOterm analysis identified altered pathways and molecular drivers that could be used as a novel approach for therapeutic treatments for EB.
Epithelial integrin αvβ6 has many important roles in skin repair in particular the activation of latent transforming growth factor (TGF)β and its abnormal activity has been associated with chronic wounds. αvβ6 integrin is transiently upregulated in blistered skin; however, it becomes chronic in epidermolysis bullosa skin blistering diseases such as recessive dystrophic epidermolysis bullosa (RDEB) and junctional epidermolysis bullosa (JEB). Here we explore how pathological integrin αvβ6 upregulation in epidermolysis bullosa (EB) is driving defective wound healing, chronic inflammation and fibrosis and if its targeted inhibition improves skin EB symptoms using an inducible JEB disease mouse model. A tamoxifen-inducible JEB disease model with epidermis-specific LAMA3 knockout (KO) was utilized for identifying the αvβ6 integrin signalling changes by immunofluorescent staining and flow cytometry during skin homeostasis and wound healing. To inhibit αvβ6 signalling in JEB skin in vivo a novel αvβ6-blocking antibody 264RAD was injected intraperitoneally for 2 weeks prior skin analysis. αvβ6 integrin expression is significantly upregulated in epidermis of JEB KO mice, especially in the blistering areas in comparison with controls. The increased expression of αvβ6 correlates with enhanced inflammatory and fibrosis markers in blistered and wounded skin, suggesting a key role in the pathological mechanisms driving skin disease. The 264RAD treatment showed a strong decrease in integrin αvβ6 expression but also a modest increase in B-cell level accompanied by a rise in mononuclear immune cell populations. Importantly, a reduction in fibrosis around skin blisters and an improvement in wound healing has been observed in JEB KO mice treated with 264RAD. The obtained preclinical data reveals abnormal expression of αvβ6 integrin in JEB and indicates that inhibition of αvβ6 is a potential novel therapeutic target to improve skin health and regeneration in JEB and possibly other EB subtypes.
Abstract Introduction and aims Recessive dystrophic epidermolysis bullosa (RDEB) is a rare inherited blistering disorder caused by loss of type VII collagen (C7). RDEB is characterized by chronic wounds, fibrosis and aggressive cutaneous squamous cell carcinoma (SCC). Recent work from our group has identified nuclear envelope changes associated with C7 loss in cultured keratinocytes. We aimed to examine keratinocyte nuclear morphology following C7 loss in RDEB skin and organotypic models, and to assess spatial transcriptomic differences between RDEB skin, RDEB SCC and healthy skin. Methods Three-dimensional human skin organotypic models were generated using N/TERT-1 keratinocytes with stable short hairpin (sh)RNA-mediated C7 knockdown (shC7) or control shRNA (shC). Nuclear morphology was assessed by immunofluorescence staining for Lamin A/C, Lamin B1, Lamin B2, SUN1, SUN2 and Emerin, with quantitative analysis of keratinocyte nuclear area and mean fluorescence intensity (n = 4 organotypics per condition). Parallel analyses were performed on skin of patients with RDEB (n = 3) and healthy control skin (n = 4). Group comparisons were performed using an unpaired two-tailed Welch’s t-test. Spatial transcriptomic profiling (Visium HD) was performed on RDEB skin, RDEB SCC and healthy control skin, followed by clustering, manual annotation and spatial mapping. Results Loss of C7 was associated with increased keratinocyte nuclear area in organotypic cultures and epidermis of patients with RDEB. shC7 organotypic models showed reduced SUN2 immunofluorescence intensity and more basally distributed SUN2 and Lamin B1 staining compared with controls. Spatial transcriptomic analysis resolved distinct epidermal, dermal and tumour-associated cell populations. Stressed basal keratinocyte and immune cell clusters were prominent in RDEB skin and RDEB SCC compared with healthy controls. Nuclear lamina genes including LMNA, LMNB1 and LMNB2 were enriched in basal keratinocyte clusters, including stressed basal keratinocytes. Conclusions C7 loss is associated with altered keratinocyte nuclear morphology in human skin and organotypic models. Spatial transcriptomic analysis identifies disease-associated epidermal and immune cell populations RDEB skin and SCC.
Harlequin ichthyosis (HI) is the most severe form of autosomal recessive congenital ichthyosis (ARCI) with aberrant lipid transport in the epidermis leading to a severe barrier defect. It is caused by variants in ABCA12, a lipid transporter involved in the transport of glucosylceramides from the lamellar body to the lipid lamellae. The aim of this study was to explore further the effect of Janus kinase inhibitors (JAKis) on the epidermal barrier in in vitro and in vivo models of HI. Three-dimensional skin equivalents (3DSEs) were generated from human ABCA12 CRISPR-Cas9 knockout and wildtype keratinocytes and treated with the JAKi abrocitinib and upadacitinib (UPA) and bulk RNA sequencing (RNAseq) and lipidomics were performed. UPA was tested in an in vivo HI mouse model followed by haematoxylin and eosin, Nile red, ABCA12, glucosylceramide and immune cell staining. Bulk RNAseq on treated and control 3DSEs showed upregulation of genes involved in keratinocyte metabolism, differentiation and skin barrier function as well as downregulation in immune response genes post-JAKi. Lipidomic analysis showed upregulation of cholesterol sulfate and phosphatidylethanolamine post-JAKi. UPA was selected to test at a concentration of 5 mg kg−1 in 10% dimethyl sulfoxide/corn oil on adult male and female inducible Abca12flox/flox K14CreER and control mice (n = 4 per group, two repeats). One day after topical 4-OH-tamoxifen treatment, UPA or vehicle were administered via oral gavage for nine consecutive days before collecting back skin. The UPA-treated HI mice had significantly reduced epidermal thickness compared with controls. Nile red and glucosylceramide staining showed restoration of neutral and polar lipids and upper epidermal glucosylceramide in the UPA-treated mice compared with controls. A reduction in T cells and dendritic cells was also observed in the JAK1i-treated mice. These results suggest that UPA restores the skin barrier in severe ichthyosis and may be a promising treatment for this group of patients.
Abstract Introduction and aims Skin ageing is associated with visible phenotypical changes such as wrinkling. On a cellular level, the number of senescent cells accumulates with age, which can drive skin ageing phenotype. During ageing, there are also profound changes in cutaneous immunity which render older adults more susceptible to skin cancers and infections. We have previously shown that skin from older adults (≥ 65 years) has impaired antigen-recall response against pathogens. This age-associated impairment is due to recruitment of proinflammatory monocytes by activated senescent fibroblasts, these monocytes secret prostaglandin E2 and blunt antigen-specific immunity. However, the mechanisms of senescent fibroblast activation that induces the recruitment of inflammatory monocytes are currently unknown. Owing to ethical limitations of performing repeated biopsies on humans, a novel ageing skin model needs to be developed. The aim of this project is to build immune responsive ‘young’ and ‘old’ in vitro three-dimensional human skin equivalent (HSE) model to identify the cellular source of this age-associated skin inflammation. Methods We have built ‘young’ (100% proliferating fibroblasts) and ‘old’ (10% senescent; 90% proliferating) HSEs. The dermal fibroblasts were isolated from sun-protected healthy human skin, and N/TERTs were employed for the epidermis. Furthermore, we established immune-responsive HSEs with the addition of monocytes isolated from fresh peripheral blood to the dermal layer. Results The ‘old’ HSEs exhibited epidermal thinning and ECM disorganization. Additionally, the increase of senescent fibroblast percentage within the HSE dermis is associated with further impairment of epidermal development. The monocytes within the HSEs differentiated into dermal macrophages as defined by CD68 expression. Conclusions We have developed HSEs that recapitulate human skin ageing. Research is ongoing to determine the influence of the old skin environment on mononuclear phagocyte differentiation. This project will aid our understanding of age-related cutaneous immune decline, and ultimately, the identification of druggable pathways to help improve skin health in older adults.
Abstract Introduction and aims Gene editing holds strong therapeutic promise for inherited skin disorders, but translation to in vivo application requires clinically relevant animal models. Dystrophic epidermolysis bullosa (DEB), caused by pathogenic COL7A1 variants, is a severe condition for which existing mouse models are poorly suited to testing gene-editing therapies. This study aimed to establish and characterize a dominant DEB (DDEB) mouse model for in vivo gene-editing evaluation and to assess the feasibility of lipid nanoparticle (LNP)-mediated delivery to skin cells. Methods A DDEB mouse model heterozygous for the Col7a1 exon 73 c.6085G>C variant was utilized. Phenotypic assessment focused on blistering severity and wound-healing dynamics. Cells isolated from mouse skin biopsies were analysed by Western blot to assess intracellular accumulation and extracellular secretion of type VII collagen. Topical delivery of self-formulated LNPs encapsulating mCherry mRNA was tested on wounded mouse skin to evaluate cellular targeting and delivery efficiency. In parallel, two gene-editing strategies were established in vitro: a deaminase-free, glycosylase-based cytosine base editor (gCBE) to correct the C>G mutation, and CRISPR-Cas9-mediated knockout of the mutant allele to mitigate the dominant-negative effect. Results The DDEB mouse model exhibited a mild baseline blistering phenotype, enabling longer-term studies, but showed delayed wound healing following injury, providing a suitable context for therapeutic testing. Western blot analysis revealed intracellular accumulation of C7 with reduced extracellular secretion. Topical application of LNPs resulted in mCherry expression in basal keratinocytes within the migrating epithelial tongue of healing wounds, indicating effective delivery to a therapeutically relevant target cell population. Gene-editing approaches are being established in vitro, with ongoing optimization to identify the most efficient strategy. Conclusions This DDEB mouse model provides a clinically relevant platform for in vivo gene-editing studies. Our results demonstrate topical LNP-mediated delivery to basal keratinocytes, supporting further development of gene-editing therapies for DDEB and other inherited skin disorders.
Abstract Introduction and aims Junctional epidermolysis bullosa (JEB) is a severe skin blistering disease with impaired wound healing and high infection risk, caused by defects in laminin-332. The transcriptional coregulators Yes-associated protein (YAP)/TAZ are important for long-term self-renewal of epidermal stem cells. They are negatively regulated by the Hippo signalling pathway. Dysregulated YAP/TAZ activity in JEB skin was previously shown to decrease epidermal keratinocyte proliferation and tissue repair. This project aims to characterize the defective signalling mechanism that cause dysfunction of YAP/TAZ in JEB epidermis. Methods Immortalized N/TERT-1 keratinocytes with short hairpin RNA-mediated laminin-332 knockdown were evaluated as alternative in vitro models to replace primary JEB keratinocytes, using clonal growth assays, Western blot analysis, and reverse transcriptase quantitative polymerase chain reaction. Results LAMA3-depleted N/TERT-1 cells displayed reduced clonal growth, thus recapitulating a key phenotype of primary JEB keratinocytes. Cell density-dependent activation of the Hippo pathway was confirmed in parental N/TERT-1 keratinocytes. High cell density caused activation of the core Hippo kinase cascade, leading to inactivating phosphorylation of YAP, and decreased mRNA expression of YAP/TAZ target genes. LAMA3 knockdown in N/TERT-1 cells caused reduced expression of YAP at mRNA and protein level, along with reduced levels of YAP/TAZ target genes CCN1 and ANKRD1. Conclusions Our findings suggest that LAMA3-depleted N/TERT-1 keratinocytes recapitulate key features of primary JEB keratinocytes, supporting their use as an in vitro model for studying YAP/TAZ dysregulation in JEB. The observed decrease in YAP expression at both mRNA and protein levels indicates laminin-332-dependent regulation of YAP. Although preliminary, these results highlight a promising path for uncovering mechanisms underlying impaired Hippo/YAP/TAZ signalling in laminin-332-deficient keratinocytes and for guiding future therapeutic strategies. Next, we will test small-molecule YAP/TAZ activators in two-dimensional and three-dimensional organotypic cultures and in a JEB mouse model to evaluate whether they can restore normal YAP/TAZ activity in JEB skin.
Abstract Introduction and aims The skin microbiome is a dynamic community of microorganisms that plays a critical role in maintaining barrier integrity and overall skin health. Cutibacterium acnes, the most prevalent member of the skin microbiome abundant in sebaceous regions contributes to barrier function by metabolizing sebum into free fatty acids in addition to interacting with keratinocytes and immune cells to modulate antimicrobial peptide production and inflammatory responses. The role of individual microbial metabolites in supporting the epidermal differentiation and skin barrier has not been fully elucidated. It is also not yet fully understood how different bacterial species and strains are able to interact with the skin and elicit different effects. A better understanding of the skin microbiome and its metabolites is essential for strategies aimed at preserving barrier function while preventing disease. This project aimed to generate a model for evaluating the effect of relevant bacterial metabolites on epidermal differentiation and the skin barrier. Methods Using a combined bioinformatic and lab-based approach, Kyoto Encyclopedia of Genes and Genomes pathways was used to predict potential metabolites produced by C. acnes strains. Combined with a metabolomics approach, where C. acnes strains were grown in artificial sebum to collect metabolites produced in an environment that mimics the sebaceous gland. Metabolites were screened using two-dimensional proliferation and differentiation assays and further investigated in a fully immortalized three-dimensional skin organotypic model. Results Bacterial metabolites that altered keratinocyte proliferation, were identified. Investigation of these metabolites identified distinct effects on skin differentiation. Metabolites that decreased proliferation triggered early differentiation with altered patterns of ki67, involucrin, K10 and transglutaminase 1 expression, suggesting a direct role in shaping skin barrier architecture. Conclusions These findings highlight the importance of microbial metabolites as active regulators of skin homeostasis and provide a foundation for future research into microbiome-based strategies for enhancing barrier function and preventing skin disorders.
Dominant dystrophic epidermolysis bullosa (DDEB) is a congenital blistering skin disease caused by mutant collagen VII (C7) chains interfering with the folding of wildtype (WT) C7, which weakens dermoepidermal junctions. Base editors (BEs) offer unique advantage in precisely correcting the mutant COL7A1 allele while preserving the WT allele. While our group has previously demonstrated the use of adenine base editor 8e (ABE8e) to correct recessive dystrophic epidermolysis bullosa (RDEB), this is the first report of ABE8e being applied to DDEB. Primary fibroblasts from patients with DDEB carrying the COL7A1 exon 73 c.6127G>A mutation were electroporated with ABE8e mRNA and single guide (sg)RNA. Base editing efficiency was evaluated by semiquantitative analysis of Sanger sequencing chromatographs in EditR. Protein-level correction was assessed using a C7 Western blot. In preparation for future in vivo base editing, we validated a DDEB mouse model carrying the Col7a1 exon 73 c.6085C>G mutation using Sanger sequencing. Sanger sequencing analysis revealed that ABE8e achieved 19–41% editing efficiency in correcting the c.6127G>A mutation, with a 2 : 1 sgRNA-to-ABE8e ratio yielding higher efficiency than a 1 : 2 ratio. However, Western blot analysis did not show consistent protein-level correction. Sanger sequencing confirmed that our DDEB mouse model is heterozygous for the Col7a1 c.6085G>C mutation. This study demonstrated partial DNA-level correction of the COL7A1 c.6127 G>A allele in ABE8e-electroporated primary DDEB fibroblasts ex vivo. However, this editing did not result in increased C7 protein levels. Notably, Western blots assess differences in C7 quantity, but not stability – a key factor in DDEB pathology. Further replication of DNA-level correction and evaluation of C7 stability using a trypsin digestion assay will be essential. These efforts will lay the foundation for transitioning from ex vivo experiments to in vivo correction of the Col7a1 c.6085C>G mutation in our DDEB mouse model.
Junctional epidermolysis bullosa caused by loss-of-function variants in genes encoding the skin basement membrane proteins laminin 332, type XVII collagen, or integrin a6(34 affects patients from birth with severe blistering, eventually leading to scarring and early lethality. In this study, we have optimized a previously published junctional epidermolysis bullosa-knockout mouse model with weekly tamoxifen intraperitoneal injections, resulting in a more controllable and severe model. Owing to the titratable dosing, this model now recapitulates both early and advanced stages of the human disease, strengthening its use in therapeutic studies. The gradual loss of laminin-a3 in the skin of the mouse through weekly injections lead to generalized blistering and fibrotic dermal changes in multiple skin sites by week 12 after tamoxifen. Our findings demonstrate the usefulness of optimizing tamoxifen induction in Cre-loxP mouse models of extracellular matrix proteins, an approach that could be applicable to other emerging inducible transgenic disease models to improve their ability to mimic the human disease phenotype.
Abstract Introduction and aims Skin ageing is a universal process that can increase susceptibility to inflammation and disease. Basement membrane collagens (BMCs) are reduced during ageing and the consequences are poorly understood. BMC sensing of mechanical stress, vital for skin homeostasis, elicits signalling sensed via cytoskeletal actin filaments to the linker of nucleoskeleton and cytoskeleton (LINC) complex (composed of nesprin and sun proteins), nuclear lamina and lamina-associated chromatin. Our hypothesis is that BMC loss during ageing leads to a deformation of the nucleocytoskeleton and disrupts chromatin organization in keratinocytes, leading to impairment of epidermal differentiation and inability to withstand mechanical stress. Methods To explore the role of BMC in mechanical stress response, stable lentiviral short hairpin (sh)RNA knockdowns of Col7 (shCol7) and Col17 (shCol17) were generated in immortalized N/TERT keratinocytes. Western blot (WB) validated knockdowns were compared with shC nontargeting control. IncuCyte assays showed shCol17 had a hyperproliferative phenotype compared with shC. Sun2 was reduced in shCol17 (P < 0.001) compared with shC (n = 5 experimental replicates). The mechanical stretch effect on cells on a flexible membrane was studied using immunofluorescence and WB analysis. Images taken from the IN Cell Analyzer confocal microscope were analysed to detect changes in protein expression and localization. Nuclear and cytoplasmic fractions made before and after stretch were analysed by WB. Results We observed collagen-specific responses to stretch-induced stress with changes to actin, components of the nuclear membrane, and LINC complex proteins. Additionally, work on investigating chromatin changes with BMC knockdown using Cleavage Under Targets & Tagmentation and Cleavage Under Targeted Accessible Chromatin assays is ongoing. Conclusions Our data show an altered nucleocytoskeleton in the context of BMC loss and mechanotransduction in skin. These data could provide novel insights into epidermal regulation by BMC in ageing.