Many skin diseases are characterized by single gene mutations, either dominant or recessive, and can therefore possibly be corrected by destruction of the mutant gene product or introduction of a wild-type gene product. In addition to being used to correct inherited skin diseases, the skin can be used as a delivery system to secrete various polypeptides, such as enzymes, growth factors, and cytokines, into the systemic circulation. Furthermore, taking advantage of the potent antigen-presenting dendritic cells in the epidermis (Langerhans cells) and dermis, the skin can be used as a route of immunization against tumor-associated or infectious-associated antigens.
Table 3 Physical Methods for Cutaneous Gene Transfer Technique
Direct intradermal injection Topical application
Particle-mediated gene transfer ("gene gun") Puncture-mediated gene transfer Electroporation Ultrasounda
Hydrodynamic injectiona aCurrent value for cutaneous gene transfer is minimal.
A. Cutaneous Gene Therapy for Inherited Skin Diseases
The molecular basis of a number of inherited skin diseases has been elucidated in recent years. These include genetic lesions leading to skin blistering, abnormal cutaneous cornifi-cation, and predisposition for cancer. In the epidermis, mutations in the genes expressed in the basal keratinocytes usually lead to skin fragility and blistering. Examples include, K5 and K14 mutations in epidermolysis bullosa (EB) simplex (88,89), laminin 5 mutations in a subset of junctional EB (90,91), and defects in type VII collagen in dystrophic EB (92,93).
Mutations in genes expressed in the suprabasal keratino-cytes lead to abnormal epidermal terminal differentiation and are often manifested as keratinization disorders known as ich-thyoses, characterized by thickened and scaly skin (94). Examples include transglutaminase 1 (TGasel) mutations in lamellar ichthyosis (LI) (95), and K1 and K10 mutations in epidermolytic hyperkeratosis (EHK) (96-98). Genetic defects in genes expressed in the skin have also been linked to cancer predisposition, as in the cases of patched mutations in basal cell nevus syndrome (99,100) and mutations in XP genes in xeroderma pigmentosa (101).
Understanding of the molecular basis of disease pathogen-esis provides the cornerstone in designing rational gene therapy strategies. Correction of recessive phenotypes, in principle, requires the correct expression of the wild-type gene product where it was previously absent or defective. Much progress has been made in such attempts and success has been recently achieved in model systems (20-22,102). Correction of dominant-negative phenotypes, however, presents more of a challenge. The dominant-negative gene product has to be efficiently disrupted before the normal functions of the wildtype gene product can be restored. Several strategies are under development to circumvent the negative effects of mutant gene products.
On understanding the mechanisms of a particular skin disease, an appropriate model has to be established to test possible gene therapy strategies before they are applied to patients. Such a model can be either a human tissue/animal chimeric model or an entirely animal model.
For recessive diseases, the pathologic phenotypes develop only when both alleles are mutated or deleted from the genome. Although mutant cells retaining disease characteristics from patients can be used to test therapeutic approaches in vitro or grafted onto an animal (103), such material is often limited. The development of mouse embryonic stem (ES) cell techniques and the use of homologous recombination made it possible to obtain virtually unlimited material if a mouse model could be generated for a particular disease. The ES cell techniques and homologous recombination allow manipulation of the mouse genome in a finely controlled manner, where mutations can be made in a particular gene to observe the consequence of its alteration during development and differentiation. Mouse models for several recessive skin diseases have been made by making a null mutation in epidermal genes, such as TGase 1 gene in LI (104) and type VII collagen gene in dystrophic EB (105). These mouse models not only provided much information on the mechanisms of disease pathologic study, but are also invaluable model systems to test gene therapy approaches for these diseases.
For dominant diseases caused by haploinsufficiency, as in the case of striate palmoplantar keratoderma (106,107), introduction of a second wild-type allele for desmoplakin may eliminate the disease phenotype, as in the case for recessive diseases. For diseases caused by a dominant-negative mutation, traditional transgenic mouse models were very informative in helping us to understand the disease mechanisms. In this approach, a mutated transgene is introduced into the mouse genome, and when it is expressed at a high enough level to compete with gene products from both wild-type alleles, a phenotype is observed. Mouse models were made in this manner for epidermolysis bullosa simplex (EBS) (108) and EHK (109). However, such mouse models cannot be used to test gene therapy approaches. First, because both wild-type alleles are present, the mutant gene product has to compete with wild-type product from both alleles. Second, the transgene is integrated into the genome randomly; therefore, its expression level and consequently the severity of the phenotype are affected by the surrounding sequences. Thus, the ratio of wildtype to mutant gene product, a crucial factor in judging the success of the therapy, remains variable in these models.
We took advantage of the mouse ES cell techniques and homologous recombination, and generated several mouse models for dominant skin diseases. ''Hot spot'' mutations in the 1A region of the rod domain in K14 and K10 have been linked to EBS and EHK, respectively (110,111). Using a knock-in/replacement strategy, we replaced a wild-type K14 or K10 allele with a mutant allele containing a ''hot spot'' mutation (112,113). Heterozygous mutant mice developed phenotypes similar to EBS and EHK, respectively, as expected for the dominant mutations. These mouse models mimic the diseases at both the genetic and phenotypic level, and will be the ideal systems to test gene therapy approaches for EBS and EHK.
3. Progress in Gene Therapy of Specific Skin Disorders a. Lamellar Ichthyosis. The autosomal recessive ich-thyoses are a clinically heterogeneous family of diseases characterized by abnormal cornification, and comprise LI and congenital ichthyosiform erythroderma (114). In LI, patients are born encased in a ''collodian'' membrane that is later shed and followed by development of large, thick scales of varying degrees of erythema. Palmar and plantar hyperkeratoses are often present.
Defects in the gene encoding keratinocyte TGase 1 were identified in a number of patients with LI (95,115). TGase 1 is normally expressed in differentiated keratinocytes and catalyzes cross-linking of cornified envelope precursor molecules, such as involucrin, loricrin, and small proline-rich proteins (116,117). With loss of TGase 1 function in the formation of insoluble cornified envelope, the barrier function of the outer epidermis is disrupted (94,118). Therefore, restoration of the TGase 1 enzymatic activity may represent a possible means of correcting the LI disorder.
Correction of recessive phenotypes requires the introduction of a wild-type gene product. A high-efficiency retroviral vector containing a wild-type TGase 1 gene was used to transduce mutant keratinocytes form LI (21). More than 98% of the primary cells expressed wild-type TGase 1, as measured by the proportion of keratinocytes positive for the transferred gene compared with the total number of cells determined by propidium iodide counterstaining (21). TGase 1 enzymatic activity was restored to normal levels and was targeted to the membrane fraction. In addition, transduced keratinocytes also demonstrated restored involucrin cross-linking and normal cornification (21). When these transduced keratinocytes were grafted to immunodeficient mice, the regenerated skin displayed restored TGase 1 protein expression in vivo and was normalized at the levels of histology, clinical surface appearance, and barrier function (20). However, TGase 1 expression in the human skin graft was lost in a month because of silencing of the vector (22).
Besides transplantation of genetically modified cells, another way to deliver genes to the skin is through direct administration to the intact tissue. Choate and Khavari (119) regenerated skin from patients with LI on nude mice to examine the corrective impact of direct injection of naked plasmid DNA. Regenerated LI skin received repeated in vivo injections with a TGase 1 expression plasmid, and restoration of TGase 1 expression in the correct tissue location in the suprabasal epidermis was observed. However, unlike LI skin regenerated from keratinocytes first transduced in vitro with a retrovirus carrying TGase 1 prior to grafting, directly injected LI skin displayed a nonuniform TGase 1 expression pattern (119). In addition, direct injection failed to correct the central histologic and functional abnormalities of LI. These results show that partial restoration of gene expression can be achieved through direct injection of naked DNA into the human skin diseased area, but underscore the need for new advances to achieve efficient and sustained plasmid-based gene delivery to the skin. b. Junctional EB. There are 2 major forms of junctional EB, the Herlitz variant or EB letalis, and generalized atrophic benign EB (GABEB). In addition, there are several other variants (120). Both forms are transmitted in an autosomal recessive manner, and both have onset at birth and are associated with marked skin fragility and generalized blister formation. Although scarring and milia formation are usually absent in both forms of junctional EB, skin atrophy is a characteristic finding in GABEB patients who also tend to have marked dystrophic or absent nails, palmoplantar hyperkeratosis, and significant scarring alopecia of the scalp (120). A characteristic feature of Herlitz disease is the development of large, non-healing areas of granulation tissue; common sites include the perioral and perinasal areas, trunk, and nape of the neck. In addition, extracutaneous involvement may occur in both forms of junctional EB. In Herlitz disease, findings may include oral blisters and erosions, dysplastic teeth, marked growth retardation, and severe anemia. In contrast, patients with GABEB have milder mucosal involvement (oral cavity, conjunctiva, and esophagus) and early loss of permanent teeth, but neither growth retardation nor anemia (120). Unlike dystrophic EB, musculoskeletal abnormalities are absent in both forms of junctional EB.
Junctional EB involves dissociation of the dermal-epidermal junction, which occurs beneath the basal cell layer, but above the lamina densa. On examination by electron microscopy, junctional EB is caused by dissolution of the lamina lucida. Mutations in a number of genes encoding vital structural proteins, including laminin 5 components a3, p3, 7 2 (90,91), BP180 (type XVII collagen or BPAG2) (121), inte-grin p4 (122), and plectin (123) have been identified in junctional EB.
Several somatic gene therapy approaches have been tested to correct the junctional EB phenotype. In one study, when keratinocytes from a patient with Herlitz junctional EB with a mutation in laminin 5 p3 were transduced with a p3 transgene, the transduced keratinocytes synthesized p3 peptide that assembled with the endogenous a3 and 72. They assembled into biologically active laminin 5, which was secreted, processed, and deposited into the extracellular matrix. Reexpression of laminin 5 induced cell spreading, nucleation of semidesmosomal-like structures, and enhanced adhesion to culture substrate (124). Organotypic cultures with the transduced keratinocytes reconstituted epidermis closely adhering to the mesenchyme and presenting mature hemidesmosomes, bridging the cytoplasmic intermediate filaments of the basal cells to the anchoring filaments of the BM (124).
A mouse line with targeted disruption of laminin a3 (LAMA3) (laminin 5 a3) was recently created (125). Although the mutation in homozygous pups caused neonatal lethality, cells isolated from these pups could be used to test therapeutic approaches in vitro. In a study carried out to restore BP180 function in cultured junctional EB patient keratinocytes and skin graft through gene transduction (126), the transduced cells had normalization of their adhesion parameter. In addition, a revertant mosaicism was reported recently in a patient with GABEB, representing a ''natural gene therapy'' (127). Importantly, reversion of the affected genotype to carrier (heterozygote in a recessive disease) genotype in about 50% of the keratinocytes was sufficient for the normal functioning of the skin.
A risk for reintroducing a therapeutic wild-type protein into patients with a recessive disease is that immunologic responses may be elicited. Extra caution has to be taken to decrease such possibility.
c. Epidermolysis Bullosa Simplex. Epidermolysis bul-losa (EB) is a group of hereditary mechanobullous disorders with at least 11 distinct forms, 7 of which are dominantly inherited. The EBS subtype is characterized by intraepidermal blistering, and most cases are due to dominant keratin mutations. The estimated incidence for EBS is 10 per 1 million births in the United States (128), with a considerable perinatal mortality rate because of electrolyte imbalance, marked protein loss and sepsis. The most severe form of EBS, epidermolysis bullosa herpetiformis (EB herpetiformis) or Dowling-Meara variant (EBS-DM), presents at birth with generalized blistering (129). Blisters occur characteristically in groups (herpetiform) on the trunk and extremities, including palms and soles, and usually heal without scarring. Development of hyperkeratoses starts later in childhood.
Approximately 70% of the reported mutations in EBS-DM occur at the same mutational ''hot spot,'' codon 125, an argi-nine located at the beginning of the rod domain of K14. On ultrastructural examination, EBS-DM cells have perinuclear aggregates of keratin filaments in the basal cells instead of keratin bundles throughout the cytoplasm in normal basal cells (128). The weakened basal cells rupture on mild mechanical trauma, and the clinical phenotype is blistering within the basal cell layer. Because K14 is only expressed in the basal keratinocytes, the suprabasal cells appear normal and undergo normal terminal differentiation.
Two mouse models were previously developed, including a transgenic model (108) and a K14 null model (130). Although both models helped us to understand the disease mechanisms, and the K14 null model also provided important insight into K15 functions, neither mimics EBS-DM at the genetic level and therefore cannot be used to test gene therapy approaches for the diseases. We recently developed a mouse model for EBS-DM, where a wild-type K14 allele was replaced with a mutant allele in the mouse germline (112). The presence of a neomycin-resistance cassette in an intron affected expression from the mutant K14 allele, and the heterozygotes had subclinical phenotypes. But homozygous pups developed extensive blisters and died shortly after birth. When the neo selection cassette was removed by Cre-mediated excision, the resulting heterozygous pups developed large blisters, as was expected for this dominant mutation. To our knowledge, this is the first mouse model that mimics EBS-DM at both the genetic and phenotypic level. Unfortunately, the pups died because of severe blistering.
To overcome such problems, we recently developed a transgenic mouse model that allows the focal deletion of geno-
mic sequence via Cre-mediated excision. A transgenic mouse line carries a Cre recombinase fused to a truncated progesterone receptor, driven by a K14 promoter (131). These mice were crossed with mice heterozygous for the mutant K14 allele with the neo cassette flanked by loxP sites, and heterozygous mutant mice carrying the transgene were obtained. On topical treatment with an antiprogestin, the neo cassette can be deleted in focal areas by activated Cre. We expect that in such areas the mutant K14 expression would be comparable with that of the wild-type K14, therefore causing blistering in the skin. We were indeed able to induce blisters using this system, and the mice remained viable (112). This transgenic mouse model is ideally suited to test gene therapy approaches for EBS-DM. In this mouse model, the mutant K14 allele can be focally activated in epidermal stem cells, and following topical administration of an inducer, blisters develop in treated areas. However, after a few weeks, blisters heal and never reappear. Some skin disorders are characterized by a mosaic pattern, with alternating stripes of affected and unaffected skin that follow the lines of Blaschko. These nonrandom patterns are believed to be caused by postzygotic mutations that occur during embryogenesis. Interestingly, a mosaic form of EBS has never been reported. It has been suggested that basal stem cells carrying a postzygotic mutation in K5 or K14 would have a selective disadvantage and be rapidly displaced by wild-type basal stem cells, which can move laterally. Using laser capture microdissection, we have shown that the induced blisters healed by migration of surrounding nonphenotypic stem cells into the wound bed. Thus, our model predicts that if EBS stem cells could be corrected, they will have a selective growth advantage when introduced into areas prone to blistering. This observation provides an explanation for the lack of mosaic forms of EBS-DM. In addition, it has important implications for gene therapy because it predicts that defective EBS stem cells will be replaced by nondefective stem cells. Another unexpected observation from this mouse model was the discovery that mice that express the mutant K14 allele at levels approximately 50% of wild-type K14 do not exhibit a skin phenotype. Previously, it had been assumed that gene therapy approaches for dominant disorders like EBS must aim to either correct the mutant allele, or completely inhibit its expression. Our model predicts that the EBS phenotype may be eliminated by overexpression of the normal K14 allele or partial suppression of the mutant K14 allele, thus increasing the ratio of wild-type to mutant protein. Technically, it is much easier to design gene therapy strategies that would achieve a partial suppression of a mutant dominant allele rather than a complete correction or suppression. Therefore, we will try to correct the EBS-DM phenotype by altering the ratio of wildtype to mutant K14 by increasing expression of wild type K14. This will be done using a viral vector expressing wildtype K14. In addition as an alternative strategy to correct the EBS-DM phenotype, we will transduce EBS-DM keratino-cytes with a viral vector expressing a ribozyme designed to specifically degrade mutant K14 transcripts. d. Epidermolytic Hyperkeratosis. This disease, also called bullous congenital ichthyosiform erythroderma, is in herited in an autosomal dominant mode, with an incidence of 1 in 200,000 to 300,000 newborns (128). Up to 50% of the reported cases arise sporadically. Affected children present at birth with erythroderma, blistering, and peeling. Erythroderma and blistering diminish during the first year of life, and hyperkeratoses develop, predominantly over the flexural areas of the extremities. On histopathologic examination, findings consist of hyperkeratosis and parakeratosis, lysis of the su-prabasal keratinocytes, and perinuclear vacuolar degeneration. The basal keratinocytes appear normal but exhibit hyperproli-feration (129). The transit time for keratinocytes to move from the basal layer to the stratum corneum is remarkably shortened and takes only 4 days in EHK patients instead of the normal 4 weeks (132). Mutations to K1 and K10 have been linked to EHK, and a mutational ''hot spot'' at Arg 156 in K10 has been identified in most severe cases of EHK (98,133). Three mouse models were previously generated, including 2 transgenic models (109,134) and a K10 null model (135). However, because they do not mimic EHK at the genetic level, none can be used as a model system to test gene therapy for this disease.
We recently generated a mouse model that mimics EHK at both the genetic and phenotypic levels by replacing a wildtype K10 allele with one that has a point mutation at Arg 156 (113). We will attempt correction of the phenotype by altering the ratio of wild-type to mutant K10 by increasing expression of wild-type K10. To do this, we will collect epidermal stem cells from heterozygous mutant mice, and transduce them with viral vectors (both retroviral and lentiviral vectors) to overexpress the wild-type K10 allele. In addition, we will attempt to correct the EHK phenotype by suppressing the expression of the mutant K10 allele by transducing EHK cells with a viral vector expressing a ribozyme specifically designed to cleave the mutant K10 transcripts. As opposed to EBS, in which a correction of the mutation in the basal keratins translates into a growth advantage, corrected and mutant stem cells in EHK will proliferate at similar rates because the mutant K10 is not expressed in stem cells, but only in the progeny of stem cells after they have differentiated and moved into the suprabasal layers. There is no selection against epidermal stem cells with K10 mutations, and these stem cells continue to give rise to defective differentiated progeny. It has recently been demonstrated that topical selection with colchicine can be used to select, in vivo, for human keratinocyte stem cells transduced with the multidrug-resistance gene (MDR) (136). Therefore, to efficiently amplify the corrected cell population, we will include the MDR as a selection marker in our targeting strategy to ablate defective epidermal stem cells and allow their replacement with corrected stem cells, which are expected to repopulate the epidermis.
e. X-linked Ichthyosis. X-linked ichthyosis (XLI) is caused by a deficiency in steroid sulfatase (STS), which leads to the accumulation of cholesterol sulphate and results in abnormal scaling skin (137). It is inherited in a pseudoautosomal mode, escaping X inactivation in both humans and mice. Transduction using a retroviral expression vector in vitro with the STS gene leads to restoration of STS protein expression as well as enzymatic activity (102). In the same study, it was shown that transduced XLI keratinocytes from XLI patients regenerated epidermis histologically indistinguishable from that formed by keratinocytes from patients with normal skin, after grafting onto immunodeficient mice. In addition, the transduced XLI epidermis presented a return of barrier function parameters to normal.
f. Xeroderma Pigmentosa. These are a group of autosomal recessive disorders associated with defects in a number of DNA repair genes, and are characterized by inadequacies in DNA repair after UV injury (101,138). XP genes that are involved in a specific mechanism of DNA repair called nu-cleotide excision repair, fall into 7 complementation groups (XP-A to XP-G), and all have been cloned (139). Xeroderma pigmentosa (XP) patients have increased susceptibility to epidermal neoplasms, such as basal and squamous cell carcinomas, and malignant melanoma (140). XP affects both the basal and suprabasal compartments, including melanocytes. Because every cell that is unable to repair the UV-induced DNA damage should be considered a potentially tumoral cell, any treatment to be curative, must target every cell in the epidermis, including every stem cell and melanocytes. Although promising attempts to correct the genetic defects in this condition have been performed (141-143), XP represents a challenging condition for cutaneous targeted gene therapy because high-efficiency targeting of melanocytes remains to be developed.
g. Other Conditions. In addition to the dominant-negative mutations in keratins that cause various blistering diseases and hyperkeratoses, several other diseases are inherited in a dominant mode through haploinsufficiency, such as pal-moplantar keratoderma that results from desmoplakin mutations (106,107). Gene therapy strategies could involve introduction of a wild-type allele, and host immunologic reactions will not be a concern in this case.
B. Cutaneous Gene Therapy for Systemic Diseases: The Skin as a Bioreactor
Epidermal keratinocytes secrete a variety of proteins such as collagen VII, laminin, proteinases, proteinase inhibitors, growth factors, and cytokines (144 and references therein). Studies using cultured human keratinocytes, grafted onto athymic mice and rats, demonstrated that keratinocytes secrete proteins that reach the systemic circulation. One of the first studies of the secretory function of keratinocytes was carried out by Lorne Taichman and colleagues. They first monitored secretion of human apolipoprotein E (apoE) in cultured human keratinocytes (145). The protein was identified as apoE on the basis of molecular weight, isoform pattern, and immuno-reactivity. When they grafted human keratinocytes onto athymic mice and rats, human apoE was detected in the systemic circulation of graft-bearing animals as long as the graft remained on the animals (146). Within 24 h of graft removal, human apoE was not detected in the plasma, indicating that human apoE in the plasma resulted from continuous produc tion of the protein by grafted human keratinocytes. These results showed that proteins as large as apoE (299 amino acids) can transverse the epidermal-dermal barrier and achieve systemic circulation.
Once it was established that epidermis-secreted proteins could reach the central circulation, genetically modified kera-tinocytes were used to test whether they could deliver transgene products into the bloodstream. Subsequent experiments using both in vivo and ex vivo approaches have been successful in delivering different polypeptides, such as growth hormone, erythropoietin (Epo), factor VIII and IX, leptin, and interleukin 10 (IL-10), to the circulation. In our laboratory, we have further enhanced the usefulness of the skin as a biore-actor by developing a bigenic gene switch system that allows focal induction of transgene expression via topical administration of an inducer (147). Therefore, because of its ability to deliver various polypeptides into the systemic circulation, its accessibility and abundant vascularization, added to the gene switch system development, the skin is a very attractive tissue to test gene therapy strategies for systemic conditions that respond to delivery of polypeptides into the circulation.
1. Progress in Cutaneous Gene Therapy of Specific Systemic Disorders a. Hemophilia A. This is an X-linked inherited disease caused by deficiency of factor VIII and has an incidence of 1 in 5000 male live births (148). This condition occurs in mild, moderate, and severe forms, reflecting the mutational heterogeneity seen in the factor VIII gene, and symptoms vary from excessive bleeding only after trauma or surgery to frequent episodes of spontaneous or excessive bleeding after minor trauma, particularly into joints and muscles (148-149). In a recent study, factor VIII-deficient transgenic mice expressing human factor VIII under the control of the involucrin promoter were generated (150). Plasma factor VIII activity and correction of the phenotype were seen in this mouse model. In the same study, skin explants from these transgenic mice were grafted into factor VIII double knockouts, which showed plasma factor VIII activity of 4% to 20% normal and had improved whole blood clotting (150).
b. Leptin Deficiency. Leptin, a 16-kDa protein hormone, is involved in the regulation of body weight in mammals (151,152). It is secreted primarily by adipocytes, and it has been shown to regulate food intake and neuroendrocrine function through its action in the hypothalamus (152). In accordance with the phenotype seen in the ob/ob mice (153), it has been determined that congenital leptin deficiency is associated with early-onset obesity in humans (154). Leptin replacement therapy has provided encouraging results in clinical studies (155); however, the need for repetitive dosing has prompted an alternative approach using gene therapy to correct this condition. In a recent study, a cutaneous gene therapy approach for leptin deficiency was successful in correcting the mouse ob/ob phenotype (156). Here, immunodeficient ob/ob mice grafted with skin implants from mice overexpressing leptin, reached body weight equivalent to that of wild-type animals. In addition, immunosupressed ob/ob mice that were trans planted with skin grafts made of human keratinocytes transduced with a leptin cDNA-carrying retroviral vector, showed weight reduction concomitant with a decrease in blood glucose and food intake (156).
c. Anemia due to Erythropoietin Deficiency. Epo is a kidney-produced glycoprotein that regulates red cell production. It binds to its receptor found in erythroid progenitor cells, activating a signaling pathway that leads to the increase of survival of these cells by inhibiting apoptosis (157). In 1987, recombinant human Epo (rHuEpo) was approved in the United States for the treatment of anemia of end-stage renal disease (157). In addition, its therapeutic use has been extended to other conditions, such as anemia associated with Zidovudine treatment of patients with AIDS, anemia secondary to chemotherapy in the treatment of cancer, anemia of pregnancy, anemia of prematurity, myelodysplastic syndrome, and bone marrow transplantation (157 and references therein). In a recent study, a lentiviral vector encoding HuEpo was delivered by single intracutaneous injection into human skin grafts on immune-deficient mice (158). The investigators demonstrated that HuEpo was present in serum, and its levels increased in a dose-dependent fashion. In addition, the hematocrit improved within 1 month after lentiviral injection and remained stable for almost 1 year.
d. Contact Hypersensitivity. Contact hypersensitivity (CHS) responses are regulated by T cells that release cytokines and attract other inflammatory cells after reacting with antigen. IL-10 has been known to be a key regulatory cytokine in both inflammatory and immune responses. Studies have demonstrated that IL-10 is involved in the regulation of the hypersensitivity response because recombinant IL-10 (rIL-10) prevented the elicitation of CHS in previously sensitized mice (159). Meng et al. injected a DNA plasmid containing human IL-10 into the dorsal skin of hairless rats (160). They detected local expression of mRNA and protein in a dose-dependent manner. Further, they showed that the transduced keratinocytes produced and released IL-10 into the circulation by detecting it in the bloodstream, and by quantifying a reduced response to challenge in distant areas from the injection site of previously sensitized animals.
e. Growth Hormone Deficiency. Original experiments of the release of an exogenous growth hormone by transduced keratinocytes into the circulation were performed by Morgan et al. (161). Using the recently developed bigenic gene switch mouse model, we showed that after a single induction, high levels of the therapeutic protein, human growth hormone (hGH), were released from keratinocytes into the circulation (147). The serum levels of hGH were dependent on the amount of inducer applied, and repeated induction resulted in increased weight gain by transgenic vs. control mice. Furthermore, physiological levels of hGH were detected in the serum of nude mice after topical induction of small transgenic skin grafts. These results clearly demonstrate the feasibility of using the gene switch system to regulate the delivery of growth hormone into the circulation for the treatment of growth hormone deficiency.
f. Other Conditions. Additional examples of the skin as a bioreactor approach and its potential use have been performed. For example, two forms of apoE, both the endogenous human apoE and a recombinant form from a transfected vector, were detected in the serum of athymic mice bearing grafts of modified human keratinocytes (162). When human kera-tinocytes in culture were transduced with a retroviral vector carrying factor IX gene, they secreted active factor IX into the medium (163). When they were grafted onto nude mice, small quantities of factor IX were detected in the bloodstream (163).
C. Other Applications of Cutaneous Gene Therapy
In addition, cutaneous gene therapy has also been applied to other physiopathological processes, such as wound healing, immunoregulation, and cancer.
The process of wound healing involves 3 stages: inflammatory reaction, formation of granulation tissue, and tissue remodeling (164). All these events are known to be regulated by different cytokines and growth factors. There are obvious advantages of gene transfer techniques for treatment of wound healing abnormalities or to enhance the wound healing rate because the epidermal barrier is defective, only limited target gene expression is needed, and the treated area is usually localized (165). Tyrone et al. successfully treated ischemic dermal ulcers in rabbits by topical application of platelet-derived growth factor (PDGF)-A or-B-DNA plasmids embedded within a collagen latice. They showed that PDGF-A and PDGF-B DNA substantially increased the formation of new granulation tissue, epithelialization, and wound closure (166). Subcutaneously injected liposomal insulin-like growth factor 1 (IGF-1) cDNA construct was shown to effectively promote reepithelialization of burn wounds, by decreasing prolonged local inflammation through modulation of the expression of pro- and anti-inflammatory cytokines (167). Intradermal injection of an AAV vector expressing human vascular endothelial growth factor A (VEGF-A) to full thickness excisional wounds in rats was found to induce new vessel formation and enhance wound healing rate (168). Similar results were obtained by Romano Di Peppe et al. using topical application of an adenovirus vector to deliver VEGF on excisional wounds of streptozotocin-induced diabetic mice (169). In other studies, liposomal keratinocyte growth factor cDNA gene delivery to acute wounds in rats, enhanced wound healing by increasing cell proliferation, reepithelialization, and neovascularization, by reducing cell apoptosis, and by activating mesenchymal cells through the induction of IGF-1 expression (170). Other groups have tested different modalities of gene transfer to skin wounds, such as the gene gun approach and the use of genetically modified cultured skin substitutes, with promising results (171,172).
2. Cutaneous Immunomodulation and Cancer Gene Therapy
Gene transfer has become a practical method to induce an immune response. The skin is rich in antigen-presenting cells that are able to initiate and control a specific immune response. It has been demonstrated that injection of naked DNA that encodes antigenic epitopes can induce specific humoral immune responses (173). The advantages of immunomodulation by genetic vaccination are evident; for example, there is no need to isolate and purify protein for vaccination, and it circumvents the use of life or attenuated viral vaccines. In addition, cutaneous transfer of plasmid DNA or mRNA allows for the concurrent delivery of genetic material encoding antigenic epitopes and immunomodulators (174,175). All these make cutaneous genetic transfer a desirable method for engineering specific immune responses. This strategy has been used to induce immune responses with different objectives, such as antitumoral immunotherapy, treatment/prevention of infections, and treatment of autoimmune diseases (175). Cutaneous gene transfer studies that induce an immune response and hold promise as plausible options in the treatment of certain malignancies are briefly presented. The gene transfer by gene gun, and subsequent expression in the skin of different the antitumoral cytokines IL-2, IL-6, and tumor necrosis factor-a and-p (TNF-a and TNF-P), led to tumor regression and/ or increased survival in mice with subcutaneously implanted tumors (176). Regression of established primary and meta-static murine tumors was documented after cutaneous gene gun delivery of a plasmid carrying the mouse IL-12 (177). In the same study, the researchers were able to demonstrate that a tumor-specific immunological response had been induced in the treated mice. In a clinical trial conducted by Thurner et al., dendritic cells were isolated, propagated ex vivo, and pulsed with a Mage-3A1 peptide. Subsequently, the cells were injected subcutaneously to advanced melanoma patients; regression of some metastases was seen in 6 of 11 patients, and expansion of specific cytotoxic T cells was seen in 8 of 11 patients (178).
3. Cutaneous Gene Therapy Approaches for Melanoma and Squamous Cell Carcinoma
The incidence of malignant melanoma (MM) has been on the rise in the last years. Squamous cell carcinoma (SCC) is a common malignancy of the epidermal keratinocytes. Due to the advantages that make the skin an amenable tissue for gene transfer, MM and SCC are frequently used as models for tumor-specific gene therapy. Different strategies have been applied for both conditions. For instance, adenoviral-mediated transfer of the herpes simplex thymidine kinase ''suicide'' gene (tk) by direct intratumoral injection, followed by gan-ciclovir administration, was used to treat human MM established in nude mice (179). In a different study using the B16 melanoma model, a synergistic effect was observed when combination therapy by adenovirus-mediated transfer of tk and IL-2 or tk and granulocyte-macrophage colony-stimulating factor was used (180). Similar approaches have been em ployed in studies with SCC models (181,182). Besides gene transfer mediated by adenoviral vectors, an in vivo liposomal-mediated approach with tk and mIL-2, and a transgenic model constitutively expressing a costimulatory molecule, have been performed for treatment of SCC (19 and references therein). In addition, other genetic transfer approaches have been employed in MM models, such as genetically altered fibroblasts and the use of helper virus-dependent, HSV-1 amplicon vectors on human MM xenografts (19 and references therein, 183).
Was this article helpful?