A Novel Model System to Identify Cellular and Molecular Defects Underlying Rare Genetic Disorders

ABSTRACT

Ankyloblepharon‐ectodermal defects‐cleft lip/palate (AEC) is a disorder caused by autosomal‐dominant mutations in the TP63 gene. AEC is characterised by the presence of severe and painful skin erosions that can take years to heal. Current treatment options for these devastating lesions are limited, highlighting the need for new therapeutic strategies. We previously generated keratinocytes from patient‐derived induced pluripotent stem cells (iPSC‐K) and identified defects in several cell adhesion complexes, including desmosomes, hemidesmosomes and focal adhesions. In the present study, we developed a complementary in vitro model using NTERT keratinocytes transduced with lentiviral constructs expressing AEC‐related TP63 mutations (N‐AEC). This model allows for the large‐scale production of disease‐relevant material, overcoming the limitations of iPSC‐derived keratinocytes, which have the characteristics of primary keratinocytes, including limited cell doublings and lifespan. We demonstrate that N‐AEC keratinocytes exhibit key defects observed in AEC iPSC‐K and AEC patient skin, including downregulation of cell adhesion proteins. In addition, 3D epidermal equivalents generated from these cells replicate pathological features seen in AEC patient skin, such as intra‐epidermal cysts, reduced desmosomal protein expression and altered expression of differentiation markers. Our N‐AEC model provides a valuable tool for investigating the mechanisms underlying skin fragility in AEC and other genetic skin disorders and advances the potential for novel therapeutic development.

1. Background

Ankyloblepharon‐ectodermal defects‐cleft lip/palate (AEC, OMIM #106260) is a genetic disorder caused by autosomal dominant missense mutations in TP63, a gene that encodes a set of transcription factors with essential roles in skin development and homeostasis [1, 2, 3, 4]. AEC patients exhibit developmental abnormalities of many ectoderm‐derived structures, including skin, teeth, hair, limbs and nails. Among these disease characteristics, the severe skin erosions are considered one of the most medically challenging aspects of the disorder [1, 5]. Although often localised to the scalp, erosions can occur anywhere on the body and can cover up to 80% of the body surface in severely affected individuals. Current treatment options are mainly limited to standard wound care, and novel treatment strategies are desperately needed for this patient population.

We and others have previously analysed the histopathology of AEC patient skin, and documented acantholysis and impaired cell‐ECM adhesion. Additionally, these histological abnormalities were associated with abnormal suprabasal keratinocyte proliferation and delayed terminal differentiation [6, 7, 8, 9]. However, to develop effective treatment strategies for AEC, it is essential to dissect the cellular and molecular defects resulting from the expression of mutant TP63. Several mouse models have greatly contributed to our understanding of AEC disease processes [6, 10], but to further dissect the molecular mechanisms, in vitro model systems that faithfully mimic disease genetics and phenotypes are ideal. We have previously generated patient iPSC‐derived keratinocyte models (iPSC‐K) and corrected the disease‐causing TP63 mutation using Crispr/Cas gene editing [11] to create isogenic pairs of cells. By comparing patient‐derived iPSC‐K to gene‐corrected counterparts, we identified key defects in critical cell adhesion complexes [9, 11, 12, 13]. Specifically, we observed downregulation and abnormal localisation of key components of desmosomes, hemidesmosomes, and focal adhesions in both patient‐derived iPSC‐K and in AEC patient skin. Desmosomes function in cell–cell adhesion, while hemidesmosomes and focal adhesions mediate cell‐ECM adhesion [14, 15, 16, 17]. As defects in any of these junctions are known to cause skin fragility, these findings suggest that these cell adhesion defects contribute to the AEC skin phenotype.

The ability to compare patient iPSC‐K with their gene‐corrected counterparts makes this system ideal for performing molecular studies. However, the differentiation protocols used to generate iPSC‐K inherently produce a limited number of cells. Further, upon differentiation iPSC‐K act as normal primary keratinocytes and have a limited lifespan, posing challenges for performing experiments that require large cell numbers. To address these limitations and to enable large‐scale functional experiments, we have developed a complementary in vitro system that allows for the production of unlimited disease‐relevant material. To this end, we transduced NTERT keratinocytes with lentiviral constructs expressing different AEC‐related TP63 mutations (N‐AEC). To most closely mimic the allele ratio in AEC patient skin, we ensured an approximate 1:1 ratio of mutant to wild‐type TP63 expression. NTERT keratinocytes, an immortalised keratinocyte cell line that the Rheinwald laboratory developed [18], have been widely used to investigate keratinocyte biology in the context of normal and pathological conditions. Unlike most other human keratinocyte cell lines, NTERT keratinocytes maintain normal keratinocyte behaviour, can differentiate into 2D and 3D cultures, and are amenable to lentiviral transduction [19, 20, 21]. Upon generating N‐AEC keratinocytes, we found that they recapitulate the gene and protein expression abnormalities observed in AEC iPSC‐K and AEC patient skin. Further, 3D organotypic cultures generated from these cells replicate key features of AEC patient skin. Thus, we conclude that this model is highly relevant to the disease and can be used to investigate mechanisms underlying skin fragility in AEC patients.

2. Questions Addressed

The purpose of this study was to develop a disease‐relevant model system that allows for the unlimited production of keratinocytes, which can be used to investigate mechanisms underlying skin fragility in AEC patients.

3. Experimental Design

See Supporting Information for experimental procedures.

4. Results

To generate N‐AEC keratinocytes, we constructed lentiviral vectors that express the TP63 cDNA, either with or without an AEC‐causing mutation, under the control of a CAG promotor. For these experiments, we selected three TP63‐AEC mutations that are known to cause severe skin erosions in humans (denoted AEC 1–3 in this manuscript). Additionally, these mutations are identical to those identified in our previously established patient iPSC lines [9]. The TP63 cDNA was linked to a TdTomato reporter gene via a self‐cleaving T2A domain, thereby allowing for the expression of equimolar amounts of TP63‐AEC and TdTomato (Figure 1a). TdTomato expression allowed us to identify and isolate pure populations of transduced keratinocytes by fluorescence‐activated cell sorting (FACS) (Figure 1b). Additionally, to analyse the consequences of TP63‐AEC expression during keratinocyte differentiation, N‐AEC keratinocytes were exposed to cell culture media with high calcium concentrations for 24 h (T24) or 48 h (T48) (Figure 1b) [22, 23].

FIGURE 1.

N‐AEC have reduced cell–cell and cell‐ECM protein expression. (A) N‐AEC model system. TP63 mutations were engineered into lentiviral constructs carrying DNp63alpha cDNA driven by a CAG promoter and linked to a TdTomato reporter via a T2A self‐cleaving peptide domain. (B) Phase‐contrast images of N‐AEC exposed to 0 (T0), 24 (T24) or 48 (T48) hours of calcium to induce differentiation gene expression (left panels). TdTomato expression (red) shows transduced cells (right panels). (C) Western blot analysis showing that TP63 expression levels in transduced NTERT are approximately double that of non‐transduced NTERT. The approximate 1:1 ratio of mutant and wild‐type protein mimics the ratio present in AEC patient keratinocytes. TP63 expression was normalised to GAPDH. Numbers indicate the ratio of normalised signal intensity compared to non‐transduced NTERT keratinocytes. (D) Ratio of transgenic TP63 expression and total TP63 expression. Standard deviation of technical triplicates. (E) Western blot analysis of desmosomal, hemidesmosomal, and focal adhesion components in T0, T24 and T48 NTERT keratinocytes. In proliferating conditions (T0), ITGA6, ITGA2, ITGB1, DSC3 and p‐FAK were downregulated in N‐AEC compared to N‐WT. In high calcium media exposed keratinocytes (T24 and T48), DSG3, DSC3 and DSG1,2 were downregulated in N‐AEC. Protein expression was normalised to GAPDH. Numbers indicate the ratio of normalised signal intensity compared to N‐WT keratinocytes. (F) N‐AEC epidermal equivalents show abnormal tissue architecture compared to non‐transduced and N‐WT cultures. Staining for TP63 and KRT14 show normal expression while DSC3 show focal downregulation in N‐AEC and normal expression in N‐WT and non‐transduced NTERT. Tomato (Tom) marks transgene expression. Scale bars in B are 1000 μm and F are 100 μm. Arrows depict abnormal regions in epidermal equivalents. Staining of the stratum corneum is non‐specific background staining. DSC3, desmocollin 3; ITGA2, integrin alpha 2; ITGA6, integrin alpha 6; ITGB1, integrin beta 1; KRT14, keratin 14; p‐FAK, phospho‐focal adhesion kinase.

To ensure that our N‐AEC model replicates AEC patient keratinocytes, we optimised our lentiviral transduction scheme to achieve an approximate 1:1 ratio of wild‐type to mutant TP63 protein. This ratio reflects the predicted ratio of wild type to mutant TP63 expression in AEC patient keratinocytes, which carry a heterozygous mutation in TP63. Since antibodies that distinguish mutant from wild‐type TP63 are unavailable, we analysed total TP63 expression using Western blot analysis. We found that TP63 protein levels in N‐AEC keratinocytes were approximately twice those in non‐transduced NTERTs (Figure 1c). As the endogenous TP63 expression levels are expected to remain unchanged, the increased TP63 protein expression likely reflects lentiviral‐mediated expression TP63, resulting in an approximately 1:1 ratio of mutant to wild‐type protein. The gene expression ratio of transgenic to endogenous TP63 was also confirmed through qRT‐PCR analysis (Figure 1d). Additionally, we generated a NTERT line that expresses wild‐type TP63 (N‐WT) at similar levels to the N‐AEC lines to determine whether expression of ectopic TP63 affects cell properties.

To generate transduced NTERT cell lines, we performed mass selection using fluorescence‐activated cell sorting (FACS) to obtain a heterogeneous population of cells containing different lentiviral integration sites. To mitigate potential effects arising from variability in lentiviral integration sites, we performed technical and biological replicates. Additionally, because ectopic expression of DNp63alpha could lead to unintended cellular consequences, we compared NTERT keratinocytes with and without ectopic expression in our initial experiments and observed no differences in adhesion, migration or 3D tissue architecture, supporting the use of N‐WT cells as an appropriate control.

In our previous work, we demonstrated the downregulation of hemidesmosomal, desmosomal and focal adhesion components in AEC skin and in AEC iPSC‐K [9]. To determine if our NTERT‐based model system faithfully replicates AEC disease mechanisms, we first analysed the expression levels of these same proteins in N‐AEC keratinocytes (Figure 1e). We observed a downregulation of key cell adhesion proteins, including the hemidesmosomal protein ITGA6, the focal adhesion proteins ITGA2, ITGB1, and p‐FAK and the desmosomal protein DSC3, and p‐FAK in N‐AEC keratinocytes. In addition, calcium exposure failed to properly induce expression of the desmosomal proteins DSC3, DSG3 and DSG1 in N‐AEC keratinocytes. As these abnormalities are also observed in AEC patient skin and in AEC iPSC‐K, these data support the conclusion that the N‐AEC model mimics AEC disease phenotypes.

We have previously shown that the concurrent downregulation of multiple cell‐ECM adhesion genes leads to reduced adhesion of AEC iPSC‐K to several extracellular matrices [9, 12]. Therefore, we analysed the ability of N‐AEC keratinocytes to adhere to collagen 4 (COL4), laminin 332 (Lam332), laminin 511 (Lam511) and fibronectin (FN) (Figure 2a). Consistent with AEC iPSC‐K, N‐AEC keratinocytes showed a significantly reduced ability to adhere to COL4 and FN extracellular matrices. The similarities between AEC iPSC‐K and N‐AEC in protein expression and functional behaviour suggest that N‐AEC is a suitable model to further investigate the mechanisms underlying AEC skin fragility.

FIGURE 2.

N‐AEC keratinocytes mimic abnormalities seen in AEC patient skin. (A) The ability of N‐AEC and N‐WT ability to adhere to several extracellular matrices was analysed using adhesion assays. Extracellular matrices tested were: COL4, LAM332, LAM511, and FN. N‐AEC have a significantly reduced ability to adhere to COL4 and FN. (B) BrdU incorporation analysis was used to determine proliferation rates in N‐AEC and N‐WT. N‐AEC show reduced proliferation in 2D culture and 3D epidermal equivalents. (C) N‐AEC epidermal equivalents show reduced proliferation (left panels; BrdU incorporation (green)) and reduced ZNF750 expression (right panels; green) in N‐AEC epidermal equivalents. Staining of the stratum corneum is non‐specific background staining. (D) qRT‐PCR analysis showed a reduction of ZNF750 expression in T24 and T48 N‐AEC keratinocytes. qRT analysis was performed on biological triplicates. (E) Immunofluorescent staining in AEC patient skin showed focal downregulation of ZNF750 indicated by the arrow and regions of normal expression shown by the arrowhead. *p < 0.05. Scale bars in C and E are 100 μm. COL4, Collagen 4; FN, fibronectin; LAM332, laminin 332; LAM511, laminin 511; NHS, normal human skin.

Next, we tested the ability of N‐AEC keratinocytes to generate 3D organotypic skin that mimics AEC patient skin [20]. Histological analysis of the 3D organotypic cultures showed that N‐WT cells, that is, cells that express lentivirus‐transduced wild‐type TP63, generated 3D organotypic skin that were indistinguishable from 3D organotypic cultures generated by non‐transduced NTERTs (Figure 1f). These data suggest that overexpressing wild‐type TP63 at the level generated in the current study does not affect the ability of these cells to generate a well differentiated epidermal equivalent or express normal epidermal markers such as keratin 14 (KRT14). In contrast, histological analysis of 3D organotypic skin generated from N‐AEC cells revealed pathological features also observed in AEC patient skin, including intra‐epidermal cysts and focal downregulation of DSC3 for all three mutations (Figure 1f). The focal nature of DSC3 downregulation is consistent with previously published data showing focal downregulation of adhesion genes in AEC patient skin [9, 12]. The mechanism for this focal expression changes is currently unknown.

Abnormalities in keratinocyte proliferation have previously been observed in AEC patient skin [11]. Therefore, we wanted to determine whether N‐AEC keratinocytes also exhibit proliferation abnormalities. Analysis of BrdU incorporation in 2D and 3D cultures demonstrated reduced proliferation of N‐AEC keratinocytes (Figure 2b,c).

To further determine if AEC 3D epidermal equivalents resemble AEC patient skin, we analysed the expression of ZNF750, an epidermal differentiation gene that is regulated by TP63 and shown to be deregulated in AEC skin equivalents [24, 25, 26]. Immunofluorescent staining for ZNF750 on N‐WT epidermal equivalents showed the expected nuclear expression in the suprabasal epidermal layers (Figure 2c). In contrast, ZNF750 expression was dramatically reduced in N‐AEC 3D epidermal equivalents, and a more diffuse staining pattern was observed (Figure 2c). Consistent with these findings, ZNF750 gene expression was greatly reduced in N‐AEC keratinocytes cultured under high calcium conditions (Figure 2d). Finally, to determine whether these in vitro findings are relevant to AEC, we examined ZNF750 expression in human skin. In normal human skin (NHS), ZNF750 expression is primarily localised to the suprabasal layers of the epidermis and has a predominantly nuclear staining pattern. However, in AEC patient skin, ZNF750 staining was diffuse, with areas of focal downregulation (Figure 2e). We previously observed a focal downregulation of desmosomal and hemidesmosomal proteins in AEC patient skin [9, 12], suggesting that the mosaic nature of protein expression is inherent to the AEC skin phenotype.

5. Conclusion and Perspectives

In summary, we have generated a novel in vitro model system to investigate the consequences of mutant TP63 expression in AEC. Here, we demonstrate that this model mimics both AEC patient skin and AEC patient‐derived iPSC‐K. As this model allows for the rapid production of large numbers of cells, it is ideally suited for large‐scale functional experiments that cannot be performed using AEC iPSC‐K. This model will advance our understanding of the mechanisms underlying AEC skin fragility, and can easily be adapted for the investigation of other genetic skin disorders.

Our work contributes to the understanding of the molecular mechanisms leading to the severe skin erosions in AEC. This work complements that of other groups suggesting important roles for the misfolding and aggregation of mutant TP63 proteins, as well as for impaired FGF signalling [10, 27]. To further dissect the pathological mechanisms, our 3D epidermal equivalent model could be enhanced by incorporating AEC patient fibroblasts. This would dissect the potential signalling defects in the cross‐talk between keratinocytes and fibroblasts. Our NTERT disease model offers a clinically relevant system to elucidate these mechanisms at play and contribute to the development of novel therapeutic options for this patient population.

Author Contributions

M.I.K., P.J.K. and M.N.S. conceptualised and designed this work; M.N.S. generated and analysed transduced NTERT lines, generated and analysed 3D epidermal equivalents and performed immunostaining. M.N.S., S.W. and I.A.P. performed RNA and protein analyses. M.I.K. and M.N.S. analysed the results and wrote the draft of the manuscript. All authors have read and approved the final version of the manuscript.

Funding

This work was supported by National Institute of Arthritis and Musculoskeletal and Skin Diseases, R01AR072621, F31AR083693.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Data S1: exd70238‐sup‐0001‐Supinfo.docx.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.