Epidermolysis Bullosa Dystrophica, Autosomal Recessive

A number sign (#) is used with this entry because autosomal recessive dystrophic epidermolysis bullosa (RDEB) and the RDEB localisata variant are caused by homozygous or compound heterozygous mutation in the gene encoding type VII collagen (COL7A1; 120120) on chromosome 3p21.

Description

Autosomal recessive dystrophic epidermolysis bullosa is a severe skin disorder beginning at birth and characterized by recurrent blistering at the level of the sublamina densa beneath the cutaneous basement membrane. This results in mutilating scarring and contractures of the hands, feet, and joints. Patients also developed strictures of the gastrointestinal tract from mucosal involvement, which can lead to poor nutrition. Affected individuals have an increased risk of developing aggressive squamous cell carcinoma (Christiano et al., 1996; Varki et al., 2007).

Allelic disorders include autosomal dominant DEB (DDEB; 131750), in which the phenotype is less severe, and nonsyndromic congenital nail disorder-8 (NDNC8; 607523), which has been found to segregate as an autosomal dominant trait in heterozygous carriers in some families with recessive DEB.

Clinical Features

Christiano et al. (1995) reported 3 Japanese brothers, aged 20, 16, and 13 years, with autosomal recessive DEB. All had extreme fragility of the skin since birth. The skin involvement led to extensive mutilating scarring, loss of nails, fusion of the digits, and joint contractures. The patients also had blistering of the mucous membranes in the oral cavity and esophageal strictures that caused severe malnutrition and anemia, which led to death in the oldest brother at age 21 years. Skin biopsies showed subbasal lamina dermal-epidermal separation with no anchoring fibrils.

Christiano et al. (1996) reported 4 unrelated families in which 5 individuals had autosomal recessive DEB. Two of the families were consanguineous. All presented at birth or soon after with skin blistering on the fingers, lips, oral mucosa, and ears, which later became widespread. Older patients had multiple erosions, scarring, mitten deformities of the hands from fusion, and joint contractures. Other features included loss of nails and esophageal strictures. Electron microscopy showed hypoplastic anchoring fibrils and cleavage at the level of the sublamina densa, consistent with dystrophic EB. One patient had skin missing from the left thumb and both feet at birth, showing phenotypic overlap with Bart syndrome (132000). Obligate heterozygous parents were clinically unaffected.

Recessive Dystrophic Epidermolysis Bullosa Inversa

The inversa subtype of autosomal recessive dystrophic epidermolysis bullosa is a rare variant characterized by lesions involving primarily the flexural areas of the body with sparing of the fingers and toes (Wright et al., 1993). Gedde-Dahl (1971) first described EBD inversa in 13 patients from 6 Norwegian families and noted the difference in distribution of skin involvement and in the course of the disease, including corneal, perianal and perivulvar involvement, compared to the Hallopeau-Siemens type of DEB. Hashimoto et al. (1976) described the disorder in 2 sisters.

Pearson and Paller (1988) described 4 American patients with DEB inversa and emphasized the occurrence of severe oral and esophageal mucosal involvement. Fingernails were normal or minimally involved, whereas toenails were mildly to moderately dystrophic or atrophic. The microscopic changes were said to be similar to those of the Hallopeau-Siemens form of epidermolysis bullosa.

Wright et al. (1993) reported 10 patients with RDEB inversa in whom the diagnosis was confirmed by tissue separation below the lamina densa and the clinical presentation of blister formation that typically localized to flexural areas. Although there was clinical variability in the severity and distribution of skin involvement, none of the patients showed pronounced digital webbing, severe generalized blistering, or growth retardation characteristic of Hallopeau-Siemens DEB. All patients had oral involvement, including ankyloglossia, loss of tongue papillae, and obliteration of the oral vestibule between the lips and gingiva. The oral opening was significantly reduced in older patients compared to controls. The teeth were not clinically abnormal or malformed and showed no evidence of generalized enamel hypoplasia. Wright et al. (1993) concluded that the inversa form of RDEB presents with oral findings that are similar to but milder than those seen in the Hallopeau-Siemens variant.

Hovnanian et al. (1994) reported 2 unrelated patients with recessive DEB inversa. An 11-year-old girl had neck, axilla, groin, and oral blistering with sparing of the hands and feet as well as sparing of the rest of the body. She had had severe and recurrent esophageal stenosis. The other patient had a similar clinical course. Skin biopsies of both patients showed cleavage beneath the lamina densa, absence of normal anchoring fibrils, and small numbers of rudimentary fibrils on electron microscopy.

Lin et al. (1995) reported 2 cases of dystrophic epidermolysis bullosa inversa. One patient had finger web scarring that required surgical correction and also had mild syndactyly of toes. The parents of one of the patients were cousins.

Other Features

Destro et al. (1987) reported a 40-year-old woman with recessive DEB and ocular manifestations. She presented with lid ulcerations, chronic conjunctivitis, diffuse subepithelial corneal scarring, corneal ulceration, and cataracts. Management with intensive lubricant therapy, soft-bandage contact lenses, and cataract extraction successfully restored her sight. Histologic examination via light and electron microscopy revealed blister formation and scarring beneath the epithelial basement membrane of both the skin and cornea, confirming the diagnosis of RDEB. Other features included fusion of all fingers and toes into mittenlike deformities and severe contractures of all 4 limbs. She had survived a spontaneous esophageal perforation and had had 15 squamous cell carcinomas removed from the limbs. A similarly affected sister died at the age of 26 years from metastatic squamous cell carcinoma.

On the basis of an analysis of 246 patients with epidermolysis bullosa of various types, Travis et al. (1992) reported that dysphagia developed in 76% of those with recessive dystrophic EB, in 20% of those with dominant dystrophic EB, in 15% of those with junctional EB (see, e.g., 226700), and in 2% of those with simplex forms (see, e.g., 131950). Lingual adhesions or microstomia occurred in dystrophic epidermolysis bullosa only, and were 8 times more common in the recessive form than in the dominant form. These lesions were provoked by the trauma of eating and reduced food intake, which exacerbated constipation caused by anal blisters and resulted in malnutrition. Stricture of the esophagus was frequent, with single or multiple esophageal webs.

Bass et al. (1993) described a prematurely born female with this disorder whose mother had strikingly elevated mid-trimester serum and amniotic fluid concentrations of alpha-fetoprotein (AFP; 104150), a positive amniotic fluid acetylcholinesterase band, and negative serial ultrasound studies.

Bourke et al. (1995) observed fatal systemic amyloidosis in 2 sisters with recessive DEB. One was 22 years old when the diagnosis of amyloidosis was made. Despite rapidly deteriorating renal function, dialysis was deemed impossible because of her extensive cutaneous infection. The older sister had negative findings of a search for amyloidosis at the age of 26 years. Although her skin disease was equally as severe as her sister's, she did not develop amyloid nephropathy until the age of 35 years.

On the basis of an analysis of 246 patients with epidermolysis bullosa, Melville et al. (1996) reported 2 unrelated children with autosomal recessive dystrophic EB who developed fatal dilated cardiomyopathy. Both were malnourished and showed severely retarded growth. The authors suggested that the likely cause for the cardiomyopathy was a micronutrient deficiency, most probably selenium deficiency, because the serum selenium level was reduced in the case in which they measured it, and also in 14 of 25 other children with dystrophic epidermolysis bullosa. Echocardiographic screening of 18 other patients with recessive dystrophic epidermolysis bullosa showed no evidence of cardiomyopathy.

Diagnosis

Prenatal Diagnosis

Anton-Lamprecht et al. (1981) achieved prenatal diagnosis of the Hallopeau-Siemens type of epidermolysis bullosa dystrophica by inspection of the skin through the fetoscope, confirmed by electron microscopic examination of a skin biopsy.

Hovnanian et al. (1995) used COL7A1 gene analysis for successful first-trimester prenatal diagnosis in 6 families at risk for recurrence of generalized recessive DEB. The disorder was of the severe Hallopeau-Siemens form in 5 families and the generalized nonmutilating form in 1. In all cases analysis of fetal DNA from amniotic fluid cells showed that the fetus had inherited at least one normal COL7A1 allele.

Mapping

Hovnanian et al. (1992) demonstrated linkage between a PvuII polymorphic site in the COL7A1 gene on chromosome 3p21 and recessive dystrophic epidermolysis bullosa in 19 informative families (maximum lod score of 3.95).

Ryynanen et al. (1991) and Uitto et al. (1992) demonstrated linkage between a PvuII RFLP of the COL7A1 gene and dominant DEB, suggesting that the autosomal dominant and autosomal recessive disorders are due to mutations in the same gene.

Molecular Genetics

In an African American family in which 4 individuals related as first cousins once removed had autosomal recessive epidermolysis bullosa dystrophica, Christiano et al. (1993) identified a homozygous mutation in the COL7A1 gene (M2798K; 120120.0001). The unaffected mother and half-brother were heterozygous for the mutation.

In 3 Japanese brothers with autosomal recessive DEB, Christiano et al. (1995) found compound heterozygosity for 2 truncating mutations in the COL7A1 gene (120120.0005; 120120.0006). The unaffected parents were each heterozygous for 1 of the mutations.

Christiano et al. (1996) identified glycine substitution mutations in the COL7A1 gene in affected members of 4 unrelated families with RDEB. Two families were compound heterozygous for a glycine substitution and a premature termination mutation (see, e.g., 120120.0036; 120120.0037), whereas the other 2 families were homozygous for a glycine substitution (see, e.g., 120120.0038). In all 4 recessive families, the glycine substitution mutation was silent in heterozygous carriers who had no disease manifestations. Christiano et al. (1996) stated that the COL7A1 gene is thus unique among the collagen genes in that different glycine substitutions can be either silent in heterozygotes or can result in a dominantly inherited DEB. Inspection of the location of the glycine substitutions did not show a positional effect in terms of phenotype or pattern of inheritance.

In a patient with RDEB previously reported by Hatta et al. (1995), Shimizu et al. (1999) identified compound heterozygosity for 2 mutations in the COL7A1 gene G2316R (120120.0042) and G2287R (120120.0023). Heterozygous carriers of the G2287R allele had normal skin but isolated toenail dystrophy, also called nonsyndromic congenital nail dystrophy-8 (NDNC8; 607523).

Sato-Matsumura et al. (2002) studied 2 unrelated Japanese families with RDEB in which isolated toenail dystrophy also segregated as an autosomal dominant trait. In family members with dystrophic changes limited to the toenails but without skin fragility, they identified heterozygosity for the glycine substitutions G1595R (120120.0024) and G1815R (120120.0025), respectively. The patients with RDEB in each family were compound heterozygous for 1 of these mutations, respectively, in combination with a nonsense (120120.0043) or frameshift mutation (120120.0006) in COL7A1. These results supported the idea that certain glycine substitutions in the collagenous domain of COL7A1 cause a limited nail deformity, and that these alleles can also contribute to variable degrees of skin fragility when present in combination with nonsense or frameshift mutations in COL7A1.

Varki et al. (2007) analyzed the COL7A1 gene in 310 patients with dystrophic epidermolysis bullosa. Mutations were found in 1 or both alleles in 243 (78.4%) patients, comprising 355 mutant alleles of the anticipated 438 (81.1%) mutant alleles. The authors reviewed the spectrum of COL7A1 mutation and genotype-phenotype correlations, noting that patients with severe recessive DEB tended to have truncating mutations, whereas those with milder dominant DEB tended to have glycine substitutions. Seven patients had features of both dominant and recessive forms of disease and were found to carry both dominant and recessive mutations.

In 2 unrelated patients, one with recessive DEB inversa and another with classic RDEB, Hovnanian et al. (1994) identified a heterozygous mutation in the COL7A1 gene (R109X; 120120.0040). Although a second pathogenic mutation was not identified, the authors presenting convincing evidence that the disorder was recessive in both cases.

In 2 brothers with recessive DEB inversa, Kahofer et al. (2003) identified compound heterozygosity for 2 mutations in the COL7A1 gene (120120.0041; 120120.0045).

Modifier Genes

A defect in collagenase (MMP1; 120353) was implicated early on in the pathogenesis of dystrophic epidermolysis bullosa. Type VII collagen is susceptible to degradation by collagenase (Seltzer et al., 1989). Bauer (1977) found that procollagenase purified from fibroblasts of 2 patients with DEB was more thermolabile, showed decreased calcium affinity, and had decreased activity in vitro compared to control values. Bauer et al. (1977) postulated a structural gene mutation, defective posttranslational modification of the enzyme, or a mutation in a gene regulating normal degradation of collagenase.

Bauer and Eisen (1978) observed enhanced collagenase production by cultured skin fibroblasts in 8 of 10 patients with autosomal recessive dystrophic epidermolysis bullosa. Increased levels of immunoreactive collagenase were found in unaffected and affected areas of the skin. However, Winberg et al. (1989) found collagenase overexpression in only 4 of 18 RDEB patients. Bauer et al. (1986) found that enhanced expression of collagenase by fetal recessive dystrophic epidermolysis bullosa skin fibroblasts could serve as a biochemical adjunct and possibly an alternative to morphologic examination of tissue for antenatal diagnosis. Phenytoin, which was found to inhibit synthesis or secretion of collagenase, had been thought to be effective in the systemic treatment of RDEB (Bauer et al., 1980); however, in a controlled study, Caldwell-Brown et al. (1992) showed that it was without effect.

Titeux et al. (2008) found a significant association between a functional SNP (rs1799750) in the MMP1 gene (120353.0001) and disease severity in 3 affected members of an RDEB family who were discordant for the SNP. The observations were confirmed in a cohort of 31 unrelated French RDEB patients: the functional SNP resulting in increased collagenase activity was associated with more severe phenotype (p = 6.27 x 10(-5)). Titeux et al. (2008) concluded that increased MMP1 leads to increased collagen degradation and worsening disease severity, suggesting that MMP1 is a modifier gene in RDEB.

Genotype/Phenotype Correlations

Van den Akker et al. (2011) reviewed the 29 known full genotypes associated with RDEB inversa from their study and the literature and found that the functional genotype in the disorder is a homozygous, compound heterozygous, or hemizygous missense mutation within the triple helical domain of COL7A1. Of the 19 known missense mutations, all involved substitutions of arginine or glycine. Three of the 5 arginine substitutions (e.g., R2063G) and 9 of the 14 glycine substitutions (e.g., G1907E) were specific to the inversa form of RDEB.

Clinical Management

Gene Therapy

Chen et al. (2002) delivered and expressed full-length COL7A1 to human skin cells in cell culture using a self-inactivating minimal lentivirus-based vector. Transduction of lentiviral vectors containing the COL7A1 transgene into keratinocytes and fibroblasts from patients with RDEB and absence of type VII collagen resulted in persistent synthesis and secretion of type VII collagen. Unlike parent cells from these patients, the gene-corrected cells had normal morphology, proliferative potential, matrix attachment, and motility. Chen et al. (2002) used these gene-corrected cells to regenerate human skin on immune-deficient mice. Human skin regenerated by gene-corrected cells had restored expression of type VII collagen and formation of anchoring fibrils at the dermal-epidermal junction in vivo.

Ortiz-Urda et al. (2002) found that bacteriophage integrase-based gene transfer stably integrated the COL7A1 cDNA into genomes of primary cultured epidermal progenitor cells from 4 unrelated RDEB patients. Skin regenerated using these cells displayed stable correction of hallmark RDEB disease features, including type VII collagen protein expression, anchoring fibril formation, and dermal-epidermal cohesion. Ortiz-Urda et al. (2003) stably transfected recessive dystrophic epidermolysis bullosa fibroblasts with a plasmid encoding human COL7A1 cDNA, which led to the expression of type VII collagen protein at a level quantitatively higher than that of normal keratinocytes. Intradermal injection of these RDEB(+) cells into intact mouse skin resulted in correct localization of human type VII collagen to the epidermal-dermal basement membrane zone. This expression was stable for the 16-week duration of the experiment. Injection of RDEB(+) fibroblasts into RDEB human skin tissue regenerated on immune-deficient mice restored type VII collagen at the cutaneous basement membrane zone and corrected subepidermal blistering.

Wagner et al. (2010) reported the results of allogeneic hematopoietic stem-cell transplantation after immunomyeloablative chemotherapy in 6 children with autosomal recessive dystrophic epidermolysis bullosa. They showed variable reductions in blister formation between 30 and 130 days after transplantation. Two patients had rapid and substantial clinical improvement, 1 had slow improvement with only a modest overall benefit, 1 had rapid improvement on short-term follow-up, and 1 had a recurrence of blistering after an early period of almost no blistering. One recipient died at 183 days after transplant from graft rejection and infection. Skin biopsies showed a substantial proportion of donor cells in the skin and mucosa. The cells appeared to ba hematopoietic in origin, but their identity could not be fully determined. The authors suggested that the donor cells secreted type VII collagen that was subsequently incorporated into the lamina densa. Type VII collagen deposition could be detected in skin biopsies after treatment, but anchoring fibrils never appeared normal. Wagner et al. (2010) emphasized that bone marrow transplantation is a high-risk procedure, but noted that it may offer some benefits to patients with autosomal recessive dystrophic epidermolysis bullosa.

History

In the family with 3 affected sibs in which linkage studies excluded the involvement of the collagenase locus on 11q22, Hovnanian et al. (1991) found high levels of collagenase mRNA in only 2 of the 3 affected sibs, suggesting that it was not the primary defect in that family. By linkage studies, Colombi et al. (1992) also excluded the interstitial collagenase gene as the one responsible for severe generalized recessive epidermolysis bullosa dystrophica. They also excluded stromelysin I (MMP3; 185250) and stromelysin II (MMP10; 185260). The same family had other members affected with a form of cerebellar ataxia of postpubertal onset. The 3 genes, as well as fibronectin (FN1; 135600) on chromosome 2, were excluded as being involved in both phenotypes.

Animal Model

In an inbred breed of golden retriever dogs with RDEB and aberrant expression of collagen type VII reported by Palazzi et al. (2000), Baldeschi et al. (2003) isolated and analyzed the 9-kb dog COL7A1 cDNA and identified a 5716G-A transition in exon 68, resulting in a gly1906-to-ser (G1906S) substitution at a conserved residue. Highly efficient transfer of the wildtype COL7A1 cDNA to both dog RDEB and human primary RDEB COL7A1-null keratinocytes, using recombinant retrovirus vectors, achieved sustained and permanent expression of the transgene product. The expression and posttranslational modification profile of the recombinant collagen type VII was comparable to that of the wildtype counterpart. The recombinant canine collagen type VII in human RDEB keratinocytes and dog cells corrected the observable defects caused by RDEB keratinocytes in cell cultures and in vitro reconstructed skin. Baldeschi et al. (2003) concluded that not only infection efficiency but also high expression levels may be required to ensure therapeutic efficacy in the presence of mutated gene products.

Fritsch et al. (2008) developed a transgenic mouse model with conditional inactivation of Col7a1 expression resulting in a Col7a1 hypomorphic animal expressing about 10% of normal Col7a1 levels. Homozygous mice appeared normal at birth, but developed blisters on the paws by 24 to 48 hours after birth. Hypomorphic mice showed poor general condition resulting from poor nutrition and blisters of the tongue. A liquid diet resulted in increased survival. Mitten deformities of the paws were found to result from soft tissue accumulation and contraction due to aberrant fibrosis that accompanied wound healing. The phenotype resembled the human recessive disorder, including skin fragility, nail dystrophy, pseudosyndactyly, and growth retardation. Intradermal injection with wildtype fibroblasts restored Col7a1 deposition and function and resulted in phenotypic improvement.