Gene therapy serves as a promising therapy in the pipeline for treatment of epidermolysis bullosa (EB). However, with great promise, the risk of autoimmunity must be considered. While EB is a group of inherited blistering disorders caused by mutations in various skin proteins, autoimmune blistering diseases (AIBD) have a similar clinical phenotype and are caused by autoantibodies targeting skin antigens. Often, AIBD and EB have the same protein targeted through antibody or mutation, respectively. Moreover, EB patients are also reported to carry anti-skin antibodies of questionable pathogenicity. It has been speculated that activation of autoimmunity is both a consequence and cause of further skin deterioration in EB due to a state of chronic inflammation.
1. Introduction
Autoimmune blistering diseases (AIBD) are rare diseases with significant morbidity and mortality
[1][2]. AIBD are caused by autoantibodies targeting various skin antigens. In contrast, epidermolysis bullosa (EB) is a group of inherited blistering disorders caused by mutations in various skin proteins
[3]. AIBD and EB often have the same protein targeted through antibody or mutation, respectively ().
Table 1. Select autoantigens shared between pemphigoid diseases and epidermolysis bullosa.
Antigen |
EB Subtype |
AIBD Subtype |
BP230 (dystonin) |
EBS |
Bullous pemphigoid |
Collagen XVII (BP180) |
JEB |
Bullous pemphigoid, Pemphigoid gestationis |
Laminin 332 |
JEB |
Mucous membrane pemphigoid |
α6β4 integrin |
JEB |
Mucous membrane pemphigoid |
Collagen VII |
DEB |
Epidermolysis bullosa acquisita |
EB encompasses a group of inherited skin fragility diseases marked by blisters which may erode and lead to ulcers in the skin and mucous membranes
[4]. Recently, EB has been grouped into four expansive categories based on the location of tissue separation within the basement membrane zone (BMZ)
[4]. These categories include the simplex forms (EBS), the junctional forms (JEB), the dystrophic forms (DEB), and the newest subtype, known as the Kindler syndrome
[4]. In this review, we will focus on the JEB and DEB forms which involve tissue separation, namely within the lamina lucida, within and below the lamina densa confined to the upper papillary dermis, respectively
[4].
While EB and AIBD are distinct entities, there are reports of spontaneous development of AIBD in patients with pre-existing EB. Likewise, several groups have noted the prevalence of anti-skin antibodies in EB patients. Although the pathogenicity of these antibodies remains elusive, some speculate that chronic inflammation in EB combined with antigen unmasking can lead to a break in immune tolerance, resulting in development of AIBD.
Gene therapy serves as a promising therapy in the pipeline for treatment of EB. Through various vectors, different technologies lead to correction or expression of functional portions of collagen
[3][5]. However, with great promise, the risk of autoimmunity must be considered. For example, corrections of highly immunogenic portions of protein may confer a greater risk towards developing AIBD. For illustration, details the location of mutations in collagen VII and collagen XVII with respect to the immunogenic epitopes in AIBD. As such, we reviewed the literature concerning the development of AIBD and anti-skin in EB patients.
Figure 1. (
A) Collagen mutation profile and immunogenic domains of epitope reactivity. This schematic representation of collagen VII consists of noncollagenous-1 (NC1, shown as a blue elliptical bar), triple-helix collagenous (shown as a blue rectangular bar), and noncollagenous-2 domains (NC2, shown as a blue oval bar). A crucial region within NC1 is the fibronectin-III-like domains 1–9 (shown as blue vertical bars). The asterisks indicate the approximate location of the mutations alongside intervals of 100 amino acids within the collagen VII polypeptide chain. The size of the asterisk corresponds to the number of mutations detected within the specified interval. The second half of this image depicts the combined results of studies measuring the reactivity of sera from patients with epidermolysis bullosa acquisita to epitopes alongside collagen VII. The intensity of the color relates to the percent reactivity identified within the individual domains. The areas of highest immunogenicity in a majority of patients include fibronectin-III-like domains 4–6 of collagen VII (approximately AA 500–800), and various regions within the NC1 and collagenous domains. As such, gene therapy must target the greatest number of mutations, while avoiding highly immunogenic areas of epitope binding
[6][7][8]. (
B) In this schematic representation of collagen XVII, the extracellular domain consists of stretches of noncollagenous domains and a series of 15 collagenous domains (shown as blue vertical bars). Also represented are the intracellular (shown as a blue elliptical bar) and transmembrane domains (shown as a black curved bar). The asterisks indicate the approximate location of the mutations alongside intervals of 50 amino acids within the collagen XVII polypeptide chain. The size of the asterisk corresponds to the number of mutations detected within the specified interval. The second half of this image depicts the combined results of studies measuring the reactivity of sera from patients with bullous pemphigoid to epitopes alongside collagen XVII. The intensity of the color relates to the percent reactivity identified within the individual domains. The NC16a region of collagen XVII (AA 490–566) was identified as having the highest immunogenicity in the majority of patients but reactivity to various subdomains within the intracellular region must be considered. As such, gene therapy must target the greatest number of mutations, while avoiding highly immunogenic areas of epitope binding
[9][10][11][12][13].
2. Presence of Anti-Skin Antibodies in EB
Although EB is an inherited disorder caused by genetic defects, autoantibodies to skin antigens have been demonstrated in the sera of patients with EB
[14][15][16][17]. Overall, patients with EB have significantly higher antibody titers against components of molecules responsible for cell adhesion, such as desmoglein 1, desmoglein 3, collagen XVII, BP230, and collagen VII compared to controls ()
[7]. Similarly, 40.6–90.6% of patients with EB have significant antibody titers against collagen type III, IV, and V, as well as laminin
[14]. Different subtypes of EB have varying prevalence of these circulating autoantibodies
[14][15][16]. For instance, patients with the generalized forms of EB had significantly higher antibody titers (2–5-fold) against desmoglein 1, desmoglein 3, collagen XVII, BP230, and collagen VII compared to patients with other EB forms
[15][16]. This difference in antibodies concentrations was accentuated when the sera of patients with recessive DEB (the most severe EB subtype clinically) was compared to that of patients with EBS, as recessive DEB patients had 7–11-fold higher antibody titers against collagen XVII, BP230, and collagen VII
[15]. Furthermore, a direct correlation was found between disease severity (Birmingham EB Severity scores) and the concentrations of the antibody titers against collagen XVII, BP230, and collagen VII
[15][16].
Table 2. Presence of autoantibodies in patients with EB.
|
EB Subtype |
n |
Autoantigen |
|
|
|
Collagen |
FN |
LAM |
Dsg1 |
Dsg3 |
Collagen XVII/ BP180 |
BP230 |
I |
II |
III |
IV |
V |
VI |
VII |
[14] |
EBA |
2 |
0.0% |
0.0% |
100.0% |
0.0% |
100.0% |
0.0% |
|
0.0% |
50.0% |
|
EBS |
20 |
0.0% |
0.0% |
85.0% |
60.0% |
85.0% |
0.0% |
0.0% |
40.0% |
JEB |
4 |
0.0% |
0.0% |
100.0% |
50.0% |
100.0% |
25.0% |
0.0% |
0.0% |
DEB |
6 |
0.0% |
16.7% |
83.3% |
33.3% |
100.0% |
16.7% |
16.7% |
66.7% |
Total |
32 |
0.0% |
3.1% |
87.5% |
50.0% |
90.6% |
6.3% |
3.1% |
40.6% |
[15] |
RDEB |
19 |
|
4.96 U/mL |
|
5.62 U/mL |
6.14 U/mL |
14.2 U/mL |
12.7 U/mL |
Other EB |
23 |
1.08 U/mL |
2.67 U/mL |
2.8 U/mL |
5.7 U/mL |
3.7 U/mL |
Healthy Controls |
38 |
0.26 U/mL |
2.12 U/mL |
1.58 U/mL |
1.82 U/mL |
1.68 U/mL |
[16] |
RDEB |
17 |
88% |
|
combined percentage of 88% |
EBS |
10 |
10% |
combined percentage of 50% |
The implications of elevation in serum concentrations of autoantibodies in EB are not yet known, but autoantibodies against collagen VII have been shown to induce blistering in humans and experimental models
[18]. IgG4 dominated the autoimmune response in patients with collagen-VII-specific antibodies
[19]. High levels of IgG4 autoantibodies and other circulating anti-collagen VII autoantibodies have been detected in a majority of recessive DEB patients, independent of the
COL7A1 mutation type or quantitative collagen VII levels
[16][20][21]. In epidermolysis bullosa acquisita (EBA), autoantibody-induced tissue damage against collagen VII contributes to blistering
[22]. Complement activation through both the classical and alternative pathways have been implicated in disease pathogenesis but the alternative pathway appears to be predominant
[23][24]. Autoreactive IgG and immune complex-FcγR binding initiate an inflammatory complement cascade resulting in extravasation of neutrophils, release of proteolytic enzymes, and reactive oxygen species
[25]. Moreover, T cells may perpetuate tissue damage in EBA through association with immune complexes and neutrophils
[26].
Anti-collagen XVII antibodies trigger subepidermal blistering in this bullous pemphigoid (BP) model via complement activation and non-complement patterns
[27][28]. The IgG4 subtype was found to induce inflammation by activating leukocytes in a non-complement fixing pattern or by binding collagen XVII in a Fc-independent manner causing dermo–epidermal junction (DEJ) separation in BP
[29]. Elevated anti-collagen XVII autoantibodies are even linked to more active and severe disease, as well as poorer prognosis
[30][31][32]. However, negative direct and indirect immunofluorescence test results in most cases suggest that circulating autoantibodies are not pathogenic
[33].
Moreover, immune-mediated complications and disease pathology have been described in EB patients, including celiac disease, amyloidosis, post-infectious glomerulonephritis, and IgA nephropathy
[34][35][36][37]. Among cutaneous disease, a few reported cases of autoantibodies causing concurrent AIBD in patients with inherited EB include: EBA in a patient with dominant DEB
[20], EBA in a patient with recessive DEB
[38], and BP in a patient with JEB
[39] (). In each of these cases, the acquired AIBD was resistant to common therapy, patients had minimal clinical improvement, and one patient with dominant DEB even died from severe hypoalbuminemia and anemia
[20]. The authors speculate that genetic modifiers or environmental factors may help explain why some patients with positive serology exhibit clinical disease while others do not. Although not fully understood, the authors propose that chronic blistering and inflammation due to altered protein synthesis and structure in EB
[16][38] contribute to the immunologic recognition of “self.” Alternatively, it is plausible that activation of autoimmunity is both a consequence and cause of further skin deterioration in EB due to a state of chronic inflammation. Herein, we review the factors that facilitate the initiation of autoimmune and inflammatory responses to help understand the pathogenesis and therapeutic implications of the overlap between EB and AIBD.
Table 3. Reported cases of confirmed cases of AIBD arising in patients with EB.
Year |
Author |
EB Type |
AIBD Type |
Workup |
2016 |
Hayashi |
DDEB |
EBA |
DIF: Linear deposits of IgG and C3 at the DEJ IIF: Linear deposition of IgG at the dermal side of the DEJ Immunoblot analysis: Reactive to collagen type VII and its NC1 domain. Non-reactive to laminin 322 Mutations: c.7868G > A in the COL7A1 gene |
2018 |
Guerra |
RDEB |
EBA |
DIF: Linear deposition of IgG with a u-serrated pattern along the cutaneous BMZ IIF: IgG binding to the dermal side of the salt-split skin ELISA: Positive for anti-collagen type VII, anti-BP180, and anti-BP230 Immunoblot Analysis: Reactive to laminin 332 Mutations: c.410G > A and c.3674C > T in the COL7A1 gene |
2019 |
Fania |
JEB |
BP |
DIF: Linear IgG and C3 deposits in an n-serrated pattern at the DEJ IIF: Epidermal staining of the salt-split skin ELISA: Positive for anti-BP180. Negative for anti-BP230 Immunoblot analysis: Reactive to BP180 and its LAD-1 domain. Not reactive to laminin 332 Mutations: c.1132 + 5G > A in the LAMB3 gene |