Ex vivo gene modification therapy for genetic skin diseases—recent advances in gene modification technologies and delivery

Abstract Genetic skin diseases, also known as genodermatoses, are inherited disorders affecting skin and constitute a large and heterogeneous group of diseases. While genodermatoses are rare with the prevalence rate of less than 1 in 50,000 – 200,000, they frequently occur at birth or early in life and are generally chronic, severe, and could be life‐threatening. The quality of life of patients and their families are severely compromised by the negative psychosocial impact of disease, physical manifestations, and the lack or loss of autonomy. Currently, there are no curative treatments for these conditions. Ex vivo gene modification therapy that involves modification or correction of mutant genes in patients’ cells in vitro and then transplanted back to patients to restore functional gene expression has being developed for genodermatoses. In this review, the ex vivo gene modification therapy strategies for genodermatoses are reviewed, focusing on current advances in gene modification and correction in patients’ cells and delivery of genetically modified cells to patients with discussions on gene therapy trials which have been performed in this area.

to prevent graft-versus-host disease (GVHD), a lethal complication of allogeneic bone marrow transplantation where genetically disparate host cells get attacked by the immunocompetent donor T cells. 9 In addition, studies on the disease of dystrophic epidermolysis bullosa showed that only dysfunctional Type VII Collagen (C7) protein secreted by patients' epithelial cells could be detected following allogenic BMC or MSC transplantation, suggesting that a short-term clinical improvement following BMCs or MSCs transplantation was more likely due to an anti-inflammatory property of these cells. 4,6,7 In comparison with allogenic gene therapy, ex vivo autologous gene modification therapy, where patients' cells are genetically modified in vitro and then grafted or injected back to patients, overcomes the disadvantages seen in allogenic cell therapy and has become a favourite gene therapy strategy for genodermatoses, particularly in those diseases with severe epidermal loss, such as junctional epidermolysis bullosa (JEB) and recessive dystrophic epidermolysis bullosa (RDEB).
Topical delivery or intravenous/intradermal injection of therapeutic vectors directly into the skin or patients' body, known as in vivo gene therapy, has become another attractive therapeutic strategy for genodermatoses. This strategy is more suitable for those skin diseases with skin lesions covering whole body, in which local skin grafts and injections have their limitations. Several topical gene therapies using modified herpes simplex virus (HSV) carrying wildtype genes such as COL7A1 for RDEB or transglutaminase 1 (TGM1) for autosomal recessive congenital ichthyosis have been developed. 10,11 The outcomes from in vivo phase 1 topical delivery gene therapy were promising and encouraging, although the expressions of therapeutic genes in the host skin/cells were transient due to nonintegrating feature of herpes simplex virus, and multiple administrations of therapeutic vector in this in vivo gene therapy are required.
In addition, topically delivered vectors are difficult to target keratinocyte stem cells (KSCs) as these cells are scattered in the lowest layer of the epidermis. Without targeting long-lived stem cells, the durability of therapeutic effect is limited. There were few reported in vivo gene therapy studies using intravenous delivery approach for genodermatoses as extensive studies are required to demonstrate the risk and safety issues related to vectors through circulation system and the efficiency of therapeutic vectors homing to the skin.
Skin is an attractive target for gene therapy. The epidermis can be readily harvested and manipulated ex vivo, and the effects of the therapy can be easily monitored. Advances in culture techniques have allowed keratinocytes including KSCs or holoclones to be cultured in vitro. [12][13][14] Sheets of epithelium can be generated and have primarily been used as autografts in patients with severe burns for more than three decades. 15 All this has led to the development of ex vivo gene-modified epidermal sheet therapy for genodermatoses. The first ex vivo genetically modified epidermal sheet therapy for JEB was reported in 2006 following retroviral transduction of KSCs. 16 This landmark study provided important proof-of-principle evidence that correction of skin stem cells can afford effective therapy for genodermatoses. Since then, a number of ex vivo genemodified epidermal sheet therapies have been performed. The most remarkable one was the life-saving epidermal sheet therapy for a JEB patient with complete epidermal loss across 80% of body surface area. 14 Recently, phase I gene-modified epidermal sheet therapies for RDEB and Netherton syndrome (NS) were also reported and showed restoration of transgene expressions and functions to some extent 17,18 (Table 1).
The success of ex vivo genetically modified epidermal sheet therapy relies on the effective gene modification in patients' KSCs, allowing durable transgene expression in grafted epidermal sheet generated using heterogeneous keratinocyte populations including KSCs. There are, therefore, two challenges in the development of an ex vivo genetically modified epidermal sheet therapy, that is correction of mutant genes in primary keratinocytes obtained from patients and retaining KSC stemness in epidermal sheet following gene modification and sheet culture. Several gene modification strategies have been developed. This includes (i) adding a wild-type gene to patients' cells harbouring recessive mutations 14,[16][17][18] ; (ii) introducing antisense oligonucleotides or small interfering RNAs into patients cells/tissues with dominant mutation to silence mutant alleles 19-21 ; and (iii) editing genomic DNA in patients' cells using recently developed genome-editing toolkits such as transcription activatorlike effector nucleases (TALENs), Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR) and CRISPR associated protein 9 (CRISPR/Cas9), programmed base editing, and prime editing. [22][23][24][25][26][27][28][29] Following gene modification, administration or delivery of genetically modified cells to patients is the next important step for a successful ex vivo gene therapy. The common approach to this is to graft epidermal sheets generated using genetically modified patients' keratinocytes. This requires primary keratinocyte culture in vitro to expand cells including KSCs and then grow these cells as epidermal sheets. Several cultivation systems have been used for culturing primary keratinocytes, but the traditional culture system developed by Rheinwald and Green is the only one that has confirmed to retain KSCs population in order to achieve a long-term survival of the engraftment with durable transgene expression. 14,30 This review discussed issues in the development of ex vivo autologous gene modification therapy for genodermatoses, focusing on gene modification and correction of patients' cells, administration of gene-modified cells to patients with discussions on gene therapy trials which have been performed in this area.

| Gene-addition therapy and genome editing using CRISPR/Cas9
Adding an exogenous wild-type gene into patients' cells to replace mutant/defective endogenous gene and restore functional gene expression is a common gene modification strategy for recessive genetic skin diseases. The vectors that deliver wild-type genes into patients' keratinocytes and fibroblasts are commonly retro-and lentiviral vectors. 31 Both vectors have a transgene packaging capacity of ~8 kb and are able to stably integrate transgenes into host genome. 16 Retro-and lentiviral vectors can also transduce hard-to-transfect cells such as keratinocytes with high transduction efficiency and have been used in phase I/II gene therapy trials. 31 In spite of its success, gene-addition therapy has its limitations, not least the unregulated expression of genes and ectopic sites of gene insertion. In addition, in some cases, efficiency of transgene transfer has been challenging if the diseasecausing gene is large. For example, mutations in the gene COL7A1 that causes RDEB are more than 9 kb in size which is more than the packaging capacity of both retro-and lentiviral vectors, resulting in low viral titre and transduction efficiency. 32 Genome editing, aiming to directly repair mutations at genomic level, is therefore more powerful. In early years, genome-editing toolkits such as Zing-finger nucleases and TALENs were used to correct mutations in COL7A1 in keratinocytes. 25,26,33,34 However, these technologies are not easily adoptable due to either complicated design implicating protein engineering for each target gene or low editing efficiency in certain cells or tissue types. 35 The recent development of CRISPR/Cas9 technology has revolutionized gene modification approach, bringing a new era in genome editing as it comprises an easy, cheap and universal editing machinery to achieve gene correction.
CRISPR relies on the activity of a DNA nuclease Cas9 which can be recruited to the target DNA sequence by a short RNA fragment known as guide RNA (gRNA). A Protospacer Adjacent Motif (PAM) which has to be adjacent to the target DNA sequence is essential for Cas9 binding to exert its DNA cleavage function. 36 Following DNA cleavage, cellular DNA repair mechanisms are triggered to fix Cas9 induced DNA double-strand break (DSB). The most prevalent repair mechanism is non-homologous end joining (NHEJ) in which the broken DNA ends are ligated back together with no regard to homology. NHEJ causes insertions or deletions (indels) of nucleotides that can cause frame shifts/ premature termination codons on target gene sites, resulting in gene disruption and knockout, and this genome-editing strategy has widely been used for generation of disease models with gene knockout.
On the contrary, when a DNA template that shares high degree of homology with the target sequence is provided, homology-directed repair (HDR) pathway is involved, resulting in precise genetic modifications. This genome-editing approach that uses CRISPR/Cas9 mediated HDR has been experimentally used for correction of mutant genes in patients' keratinocytes and fibroblasts with genodermatoses.
For example, using the dual sgRNA CRISPR/Cas9 deletion strategy of COL7A1 exon 80, RDEB keratinocytes harbouring a homozygous carrier of the c.6527insC mutation restored the C7 protein expression. 23 Studies have also shown correction of a splice-site mutation (c.425A > G) in exon 3 or a null mutation in exon 2 (c.189delG) of COL7A1 using CRISPR/Cas9 mediated HDR approach with restoration of the C7 protein expression and anchoring fibril formation. 24,25 However, it has been noticed that the efficiency of CRISPR/Cas9 mediated HDR was low around 11 -15%, and this limited CRISPR/Cas9 mediated HDR genome editing for clinical application. 24 Apart from low editing efficiency, HDR events are often outnumbered by NHEJ, causing a remarkably challenging environment for donor template knock-in and precise genome editing. Furthermore, the promiscuous activity of Cas9 can cause random DSBs at nonspecific sites (off-targets), unexpected NHEJ at off-target sites and induce p53 and apoptosis pathways. 37,38 All these require further refinements of the conventional CRISPR/Cas9 technology.

| Programmed base editing
The approach of programmed single base editing (BE) has been developed recently. 39 It allows precise, irreversible deamination of one nucleotide to another in a programmable manner, obviating the need for DNA double-strand break and HDR used in conventional CRISPR/Cas9 genome editing. In the BE toolbox with versions 3 TA B L E 1 Ex vivo epidermal sheet graft therapies discussed in this review.

| Delivery of genome-editing toolkit into keratinocytes and fibroblasts
The common approaches to deliver genome-editing toolkits to pa- was also supported by another study in which the expression of integrin β1, one of KSC markers, was not changed in primary keratinocytes after electroporation. 54 All these indicate that electroporation could be a reliable and feasible approach to deliver genome-editing toolkits into patients' keratinocytes with higher delivery efficiency and retention of KSC's stemness. Nevertheless, low cell viability was reported with nearly 20% cell death after electroporation. 54 To prevent cell death, rho-associated protein kinase inhibitor (ROCKi) has been widely used in cell culture medium following electroporation.
This application is based on the discovery that ROCKi could help in the growth of stem cells and prevent cell death. 55,56 However, the precise mechanism of ROCKi on cell survival and its cytotoxicity on cells is not clear and needs to be further demonstrated.

| Generation of genetically modified epidermal sheet
Ex vivo gene therapy using genetically modified epidermal sheet grafts has been applied for the treatment of JEB and RDEB. This is because generalized blistering, recurrent wounds and fragile skin are so severe in these patients, and ex vivo autologous gene-modified epidermal sheet therapy not only provides a long-term functional gene/protein expression, but also covers wounds to support wound healing. 57 In ex vivo epidermal sheet graft therapy, patients' keratinocytes are isolated from a small skin biopsy, expanded, genetically modified, and cultured as epidermal sheet in vitro and then grafted back to the patients (Figure 1). Mavilio et al. carried out the first ex vivo gene-modified autologous epidermal sheet therapy for JEB. 16 In this study, keratinocytes including KSCs isolated from an adult JEB patient with laminin β3 (LAMβ3) deficiency were transduced with the retroviral vector carrying a wild-type LAMβ3 cDNA. in SFM has been suggested to be beneficial for keratinocyte proliferation and prevention of cell differentiation. 79,80 However, there are conflicting opinions on feeder-free culture for retention of KSCs.
Studies showed that KSCs growing in feeder-free culture might lose their stem cell properties such as limited proliferation ability and increased tendency towards differentiation and senescence. 81,82 In the trial of epidermal sheet graft therapy for RDEB, SFM and feeder-free system was used and the outcomes showed a relatively short-term transgene expression. 18 It might be because the SFM and feeder-free culture system was selective or favourite for transient amplifying cells but not KSCs, but this speculation remains unconfirmed as the trial for RDEB did not assess the KSC population after transduction and sheet culture. However, there are learning points from these studies, such as it is important to check the proportion of KSCs before grafting cells or epidermal sheet to the patients and this should be one of release criteria for investigational medicinal product (IMP).

| Genetically modified fibroblasts
Intradermal injection of genetically modified dermal fibroblasts is another approach of ex vivo gene therapy for genodermatoses such as RDEB. Dermal fibroblasts are easy to isolate from the skin, require less complicated culture process compared to keratinocytes and can be extensively expanded in vitro. 83,84 As fibroblasts secrete C7 protein, the major component of dermal anchoring fibrils, intradermal injection of fibroblasts is, therefore, a relatively straightforward gene therapy approach for RDEB patients with the absence of C7 expression. 5 A phase I/II clinical trial for RDEB utilized intradermal injection of gene-modified autologous fibroblasts showed functional C7 expression and new anchoring fibrils formation. 85 Another intradermal injection of autologous gene-modified fibroblast for RDEB also revealed a significant increase in the expression of C7 at the site of injection for at least 12 months, although functional anchoring fibrils formation was not observed. 32 However, it has to be aware that this therapy is unable to provide a long-term therapeutic solution, and multiple intradermal injections of genetically modified fibroblasts would be required in order to maintain a durable therapeutic effect.

| Future perspectives of gene therapy for genodermatoses
Since CRISPR/CAS9 was elucidated to be potential for genome editing in 2012, 86   In summary, there is no doubt that mutation correction using recently developed genome-editing technology is the future direction of gene therapy for genodermatoses. Currently, ex vivo genemodified therapy for genodermatoses is still at its earlier stage, but it has the potential to treat those devastating genetic skin conditions in the near future.

CO N FLI C T O F I NTE R E S T
None declared.

AUTH O R CO NTR I B UTI O N S
VJ contributes to writing and revision; EK contributes to writing and revision; WQ contributes review and comments; WLD contributes construction, writing and revision. All authors have read and approved the final manuscript.