Abstract
purpose. To demonstrate the specific binding of autoantibodies present in the
sera of patients with ocular cicatricial pemphigoid (OCP) to human β4
integrin present in the normal human conjunctiva (NHC) and to study the
role of OCP autoantibodies and antibody to human β4 integrin in the
pathogenesis of subepithelial lesion formation in OCP.
methods. Indirect immunofluorescence assay and in vitro organ culture method
using NHC were used. Sera and IgG fractions from 10 patients with OCP;
immunoaffinity-purified OCP autoantibody; antibodies to human β4,β
1, α6, and α5 integrins; and sera from patients with pemphigus
vulgaris, bullous pemphigoid (BP), and chronic atopic and chronic
ocular rosacea cicatrizing conjunctivitis; and normal human serum (NHS)
were used.
results. Nine of 10 OCP sera or IgG fractions, immunoaffinity-purified OCP
autoantibody, antibodies to human β4 and α6 integrins, and sera
from patients with BP showed homogenous, smooth linear binding along
the basement membrane zone (BMZ) of the NHC. NHS, antibodies to other
integrins, and sera from patients with chronic cicatrizing
conjunctivitis from other causes showed no such binding. When NHC was
first absorbed with OCP sera and then reacted with anti-β4 antibodies
or vice versa, the intensity of the BMZ binding was dramatically
reduced or completely eliminated, indicating that there were
autoantibodies in OCP sera specific for the β4 integrin. BMZ
separation developed 48 to 72 hours after addition of total OCP sera,
IgG fractions from OCP sera, immunoaffinity-purified autoantibodies
from sera of patients with OCP, or anti-β4 antibodies to the NHC
cultures, but not after addition of normal control sera, sera from
patients with chronic cicatrizing conjunctivitis from causes other than
OCP, or sera from patients with OCP in clinical remission.
conclusion. Circulating anti-β4 integrin antibody may have an important role in
the pathogenesis of OCP.
Ocular cicatricial pemphigoid (OCP) is an uncommon, chronic,
subepithelial, scarring systemic autoimmune disease that mainly affects
the conjunctiva and other mucous membranes derived from stratified
squamous epithelium and (occasionally) the skin.
1 2 3 4 5 The
pathologic process of chronic conjunctivitis and the accompanying
progressive subepithelial fibrosis, results in trichiasis,
distichiasis, conjunctival keratinization, xerosis, and, eventually,
blindness secondary to corneal damage as a consequence of these changes
in the ocular environment.
OCP has some pathophysiological features in common with other bullous
diseases, such as linear IgA bullous disease, epidermolysis bullosa
acquisita, bullous pemphigoid (BP), and cicatricial pemphigoid (CP).
The process of identifying anti-BMZ antibodies in the sera of many
patients with CP is difficult and variable. It is likely that a variety
of anti-BMZ autoantibodies with different specificities recognize
different target molecules present in the BMZ. This may account in part
for the wide spectrum of clinical manifestations and clinical courses
of these bullous diseases. For example, BP sera bind to 180-kDa
hemidesmosome (BPAg2) and to 230-kDa desmoplakin (BPAg1) proteins. On
direct immunofluorescence examination of perilesional conjunctiva of
patients with OCP, deposition of immunoglobulin and/or complement along
the BMZ, although pathognomic, can often be difficult to demonstrate or
inconclusive.
Circulating anti-BMZ antibodies have been observed in the sera of
patients with OCP.
5 6 7 8 The autoantigen against which these
anti-BMZ antibodies are produced has not been well defined or fully
characterized. In our earlier studies we demonstrated that sera of
patients with OCP bind to a 205-kDa protein in human skin, conjunctiva,
and tumor cell lysates in an immunoblot.
6 7 In our further
studies we observed that the antibody to a 205-kDa protein in OCP sera
recognizes the cytoplasmic domain of human β4 integrin.
6 These observations collectively suggest that sera of patients with OCP
specifically contain antibodies to human β4 integrin. A subset of
patients with OCP with clinical features similar to CP has been
characterized by the presence of autoantibodies against epiligrin,
which is now identified as the α3-subunit of laminin
5,
9 10 a ligand for α6β4 integrin.
11 Binding of oral pemphigoid autoantibodies to the α6 integrin subunit
has been reported.
12
The purpose of this study was to investigate whether the anti-BMZ
autoantibodies in sera of patients with OCP specifically bind to the
human β4 integrin present in human conjunctiva. We describe an in
vitro model that may facilitate understanding of some of the events
that produce vesicular lesions in conjunctival epithelium in OCP.
Normal human bulbar conjunctiva was obtained during cataract
surgery from six patients after informed consent and ethical permission
was obtained for their use in a human organ culture model. Conjunctiva
was used immediately after surgical excision or snap frozen in liquid
nitrogen, embedded in compound (Optimum Cutting Temperature;
Tissue-Tek, Miles Scientific, Elkhart, IN), and stored at –70°C
until sectioning.
Blocking studies were performed as follows: Multiple sections of
NHC were preabsorbed for 60 minutes repeatedly with the following six
reagents: sera of patients with OCP that contained high titer (1:1000
by immunoblot) of anti-BMZ antibodies, the IgG fraction from OCP sera,
sera from patients with BP containing antibodies to BPAg1 and BPAg2,
monoclonal antibodies to human α6 integrin, monoclonal antibody to
human β1 integrin, and normal human serum. These sections were then
washed four times with PBS and incubated with monoclonal or polyclonal
antibodies to the human β4 integrin. The sections were then stained
with the appropriate FITC-conjugated secondary antibodies and viewed
under a fluorescence microscope to assess binding at the BMZ. In the
reverse experiment, multiple sections of NHC were repeatedly
preabsorbed with monoclonal and polyclonal antibodies to human β4
integrin, washed four times with PBS, and reacted with the six reagents
listed earlier in the paragraph. The sections were washed and then
stained with the appropriate FITC-conjugated secondary antibodies and
viewed under the fluorescence microscope for BMZ staining.
OCP sera, immunoaffinity-purified OCP antibodies, and
antibody to human β4 integrin absorbed with lysates of β4
integrin–expressing cell lines (UM-SC-22) did not produce blister
formation in the in vitro model. These antibodies did not bind to the
205-kDa protein in an immunoblot assay using normal human conjunctiva
and skin as substrate.
Sera from patients with PV demonstrated binding to 130-kDa protein in
the same substrates. When sera from patients with PV were similarly
absorbed with lysate of β4 integrin–expressing cell lines and used
in the in vitro model, it resulted in production of acantholysis of
conjunctival epithelial cells, identical with that produced by
unabsorbed sera. No BMZ separation was seen in NHC cultures treated
with absorbed or unabsorbed PV sera.
In this study, we focused on two specific issues in the
pathogenesis of OCP. First is the observation that the autoantibody in
OCP was targeted against human β4 integrin in the conjunctival BMZ.
Second, an in vitro organ culture model provided evidence that OCP
autoantibodies and antibodies to human β4 integrin were capable of
causing conjunctival BMZ separation that histologically resembled OCP.
We realize that unequivocal demonstration of blister formation may not
be possible in many patients with OCP. Nonetheless, this model provided
an opportunity to study the mechanism of separation of epithelial cells
from underlying subepithelial structures and in so doing to provide
indirectly information on factors that contribute to the integrity of
conjunctival BMZ.
Indirect immunofluorescence and blocking experiments using IIF assays
clearly indicated that OCP sera, IgG fractions from OCP sera, and
immunoaffinity-purified OCP autoantibodies from OCP sera bound to the
target antigen identically. The IIF assay also showed that conjunctival
BMZ contained α6, BPAg1, and BPAg2, along with other integrins and
adhesion molecules, and that OCP sera did not cross-react with them.
This experimental strategy was similar to that previously published in
studying the process of acantholysis in PV. In those studies, sera from
patients with PV bound to the intercellular cement substance
(desmoglein III) of the skin in a pattern identical with that seen in
vivo in patients with PV. The validity of our model is further enforced
by the observation that when NHC was incubated with serum from patients
with PV, the pemphigus autoantibody bound in a manner identical with
that seen on the normal human epidermis or in vivo in patients with PV.
Furthermore, BP sera bound to conjunctival BMZ in a smooth linear
manner identical with that of the skin. Normal human sera and sera of
patients with other cicatrizing ocular diseases did not bind to the
conjunctiva in a similar pattern. Observations in this report may
suggest that the subset of patients with OCP presented herein may be a
distinct subset, because they had a high frequency of extra ocular
involvement and universal presence of immunoreactants on conjunctival
BMZ. Therefore, it appears that the possibility of such patients having
a higher concentration of circulating autoantibodies in their sera to
different specificities was more likely to be detected in our assay
system than from other patients with OCP.
Interestingly, several investigators have demonstrated that sera of
patients with OCP contain antibodies to BPAg1 and BPAg2. However, our
studies clearly indicated that the antibodies to BPAg1 and BPAg2 did
not block the binding of the antibodies to conjunctival BMZ by OCP
sera. The binding sites for BP antibodies and CP antibodies were on the
different epitopes of BPAg2.
20
Studies indicate that β4 and α6 form a heterodimeric molecule
associated with hemidesmosomes.
21 Yet, our blocking
experiments demonstrated that, within the limitations of the IIF assay,
OCP sera and anti-β4 antibodies did not block the binding of
anti-α6 integrin antibody to conjunctival BMZ. This observation
further supports the hypothesis that production of anti-β4 antibodies
in patients with OCP is a specific, early, disease-related event and is
not a consequence of damage or immune injury. Additional support for
this notion of a pathogenic role of anti-β4 integrin antibodies in
OCP comes from the observation that such antibodies were not present in
the sera of patients with disease in prolonged clinical remission who
were receiving no systemic therapy.
In our in vitro conjunctival organ culture model, antibodies to β4
integrin were capable of causing BMZ separation that was histologically
similar to that produced by incubation of NHC with OCP sera, IgG from
patients with OCP, and immunoaffinity-purified OCP antibodies. These
microvesicles created in vitro had histology fairly identical with or
similar to that observed in the conjunctiva of patients with OCP early
in the course of the disease. Interestingly, antibody to α6 integrin
did not produce any microvesicles, indicating that antibody to β4
integrin was specific in its action. Further evidence for the
specificity of antibody to human β4 integrin–producing blister came
from the blocking experiments. OCP sera, immunoaffinity-purified OCP
antibodies, and antibody to human β4 integrin absorbed with lysate ofβ
4 integrin–expressing cell lines did not produce lesions seen in
incubation with unabsorbed antibodies. These experiments, therefore,
within the limitation of the organ culture system, provide additional
potent evidence that anti-β4 integrin antibody may play an important
role in the production of vesicles or bullae in the conjunctiva of
patients with OCP.
The concept that this in vitro model is a reasonable mechanism to study
events that may mimic the in vivo phenomenon in tissue dysadhesion is
valuable. Evidence for such a hypothesis comes from the experiments in
which typical acantholysis, which is a histologic hallmark of PV, was
observed in conjunctiva incubated with PV sera. The results observed in
the conjunctiva were identical with previously published observations
when normal human skin was incubated with PV sera.
19 Absorption of PV sera with β4-expressing cell lines did not affect
the ability of PV sera to produce acantholysis. The earlier in vitro
model culture for the study of PV eventually provided valuable insights
in understanding the molecular pathogenesis of PV.
In our conjunctival organ culture, we did not observe any vesicle
formation when NHC was incubated with BP sera. Using normal human skin,
Gammon et al.
22 have advanced our understanding of the
pathogenesis of BP, showing that BP sera alone is incapable of causing
skin BMZ separation but requires the addition of complement and
polymorphonuclear leukocytes for lesion production. The data on the
relationship between the duration of incubation and concentration of
antibodies indicate that these events require time for the biologic
processes to evolve. The process appeared to be dose dependent,
indicating that an appropriate concentration of antibody is required to
bind to available antigen sites to accomplish the separation of the
epithelium from the submucosa. Similarly, the process of blister
formation was time dependent, because no separation was seen in 12 to
24 hours. Our organ culture model permits the evaluation of the
specific roles of polymorphonuclear leukocytes, complement, cytokines,
and other biologic agents in the process of BMZ separation mediated by
an autoantibody.
The pathogenesis of OCP is a complicated process that is probably
heterogeneous and may involve multiple immunologic events and biologic
agents. Our in vitro organ culture model is simply one that allowed us
to study the binding of circulating antibodies to the conjunctiva and
the phenomenon associated with the separation of epithelial cells from
the underlying submucosa and tissue matrix. There are no animal models
for OCP. At least two interesting possible explanations emerge from our
observations on BMZ separation: first, that autoantibody in patients’
sera or exogenously added anti-β4 antibody dissolves β4 integrin
and that vesicles are a result of physical dissolution of the binding
of β4 integrin molecule to its ligand; and second, that binding of
antibody to β4 integrin on the conjunctival epithelial cell surface
may trigger an intracellular signal that ultimately results in the
movement of the epithelial cells away from the basement membrane, a
consequence of multiple intracellular events. Precedence for such
observations has been reported recently. Antibodies to α6β4
integrin heterodimer can influence the movement of tumor
cells.
23 Earlier observations in the study of PV
demonstrated that the breaking of cell adhesion in the epidermis
(acantholysis) is not caused by dissolution of desmoglein III, but
rather by activation of plasminogen to plasmin.
24 This
activation occurs as a direct consequence of unregulated serine
protease production in the suprabasal epidermal cells. The signal for
the upregulation of serine protease production is generated by the
binding of PV antibodies to desmoglein III on epidermal cell surfaces.
The in vitro model presented in this study allows investigators to
examine closely the consequences of the binding of anti-β4 integrin
antibodies to β4 integrin on conjunctival epithelial cell surfaces
and the processes that occur secondary to such binding.
RYC and KB contributed equally to this study.
Supported by Grant EY08378 from the National Institutes of Health.
Submitted for publication January 14, 1999; revised April 5, 1999;
accepted May 3, 1999.
Proprietary interest category: N.
Corresponding author: A. Razzaque Ahmed, Department of Oral Medicine
and Diagnostic Sciences, Harvard School of Dental Medicine, 188
Longwood Avenue, Boston, MA 02115.
Table 1. Relationship of Duration of Incubation and Concentration of
Autoantibody in Subepithelial Blister Formation in an In Vitro Culture
Model
Table 1. Relationship of Duration of Incubation and Concentration of
Autoantibody in Subepithelial Blister Formation in an In Vitro Culture
Model
Incubation Time (h) | OCP SERA* | | | Immunoaffinity-Purified OCP Antibody* | | | Antibody to Human β4 Integrin* | | |
| 10%, † | 20% | 30% | 100 μg, ‡ | 200 μg | 300 μg | 10 μg, ‡ | 20 μg | 50 μg |
12 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | 0 |
24 | 0 | 5 | 10 | 0 | 0 | 0 | 0 | 0 | 0 |
48 | 10 | 40 | 65 | 5 | 25 | 45 | 0 | 15 | 40 |
72 | 5 | 45 | 60 | 5 | 35 | 55 | 0 | 20 | 45 |
The authors thank Tong-zhen Zhao in the Hilles Laboratory of the
Massachusetts Eye and Ear Infirmary for valuable help.
Foster CS. Cicatricial pemphigoid: thesis for the American Ophthalmological Society. Trans Am Ophthalmol Soc
. 1986;84:527–663.
[PubMed]Ahmed AR, Hombal SM. Cicatricial pemphigoid. Int J Dermatol
. 1986;25:90–96.
[CrossRef] [PubMed]Ahmed AR, Kurgis BS, Rogers RS. Cicatricial pemphigoid. J Am Acad Dermatol. 1991;4:987–1001.
Ahmed AR, Foster CS, Zaltas M, et al. Association of DQw7 (DQB1*0301) with ocular cicatricial pemphigoid. Proc Natl Acad Sci USA
. 1991;88:11579–11582.
[CrossRef] [PubMed]Mondino BJ, Brown SI, Lempert S, Jenkins MS. The acute manifestations of ocular cicatricial pemphigoid: diagnosis and treatment. Ophthalmology
. 1979;86:543–555.
[CrossRef] [PubMed]Tyagi S, Bhol K, Natarajan K, et al. Ocular cicatricial pemphigoid antigen: partial sequence and biochemical characterization. Proc Natl Acad Sci USA
. 1996;93:14714–14719.
[CrossRef] [PubMed]Mohimen A, Neumann R, Foster CS, Ahmed AR. Detection and partial characterization of ocular cicatricial pemphigoid antigens on COLO and SCaBER tumor cell lines. Curr Eye Res. 1992;12:741–752.
Franklin RM, Fitzmorris CT. Antibodies against conjunctival basement membrane zone: occurrence in cicatricial pemphigoid. Arch Ophthalmol
. 1983;101:1611–1613.
[CrossRef] [PubMed]Domloge-Hultsch N, Gammon WR, Briggaman RA, et al. Epiligrin, the major human keratinocyte integrin ligand, is a target in both an acquired autoimmune and an inherited subepidermal blistering skin disease. J Clin Invest
. 1992;90:1628–1633.
[CrossRef] [PubMed]Gudula Kirtschig M, Marinkovich P, Burgeson RE, Yancey KB. Anti-basement membrane autoantibodies in patients with anti-epiligrin
cicatricial pemphigoid bind the α subunit of laminin 5. Invest Dermatol
. 1995;105:543–548.
[CrossRef] Niessen CM, Hogervorst F, Jaspars LH, et al. The α6β4 integrin is a receptor for both laminin and kalinin. Exp Cell Res
. 1994;211:360–367.
[CrossRef] [PubMed]Jones J, Downer CS, Speight PM. Changes in the expression of integrins and basement membrane proteins in benign mucous membrane pemphigoid. Oral Dis
. 1995;1:159–165.
[PubMed]Bhol K, Mohimen A, Neumann R, et al. Differences in the anti-basement membrane zone antibodies in ocular and pseudo-ocular cicatricial pemphigoid. Curr Eye Res
. 1996;15:521–532.
[CrossRef] [PubMed]Gammon WR, Briggaman RA, Inman A, Queen LL, Wheeler LE. Differentiating
anti-lamina lucida and anti-sub lamina densa anti-BMZ antibodies by
indirect immunofluorescence on 1.0 M sodium chloride-separated skin. J Invest Dermatol
. 1984;82:139–144.
[CrossRef] [PubMed]Ahmed AR, Workman S. Pemphigus-like antibodies. Arch Dermatol
. 1983;119:17–21.
[CrossRef] [PubMed]Olmsted JB. Affinity purification of antibodies from diazotized paper blots of heterogeneous protein samples. J Biol Chem
. 1981;256:11955–11957.
[PubMed]Ahmed AR, Khan KNA, Wells P, et al. Preliminary serological studies comparing immunofluorescence assay with radioimmunoassay. Curr Eye Res
. 1989;8:1011–1014.
[CrossRef] [PubMed]Bhol K, Natarajan K, Nagarwalla N, Mohimen A, Aoki V, Ahmed AR. Correlation of peptide specificity and IgG subclass with pathogenic and nonpathogenic autoantibodies in pemphigus vulgaris: a model for autoimmunity. Proc Natl Acad Sci USA
. 1995;92:5239–5243.
[CrossRef] [PubMed]Jeffes EB, Kaplan RP, Ahmed AR. Acantholysis produced in vitro with
pemphigus serum hydrocortisone inhibits acantholysis, while dapsone and
6-mercaptopurine do not inhibit acantholysis. J Clin Immunol
. 1984;4:359–363.
[CrossRef] [PubMed]Balding SD, Prost C, Diaz LA, et al. Cicatricial pemphigoid autoantibodies react with multiple sites on the BP 180 extracellular domain. J Invest Dermatol
. 1996;106:141–146.
[CrossRef] [PubMed]Stepp MA, Spurr-Michaud S, Tisdale A, Elwell J, Gipson I. Alpha 6 beta 4 integrin heterodimer is a component of hemidesmosome. Proc Natl Acad Sci USA
. 1990;87:8970–8974.
[CrossRef] [PubMed]Gammon WR, Merritt CC, Lewis DM, et al. An in vitro model of immune complex-mediated basement membrane zone separation caused by pemphigoid antibodies, leukocytes, and complement. J Invest Dermatol
. 1982;78:285–290.
[CrossRef] [PubMed]Rabinovitz I, Mercurio AM. The integrin α6β4 functions in carcinoma cell migration on laminin-1 by mediating the formation and stabilization of actin-containing motility structures. J Cell Biol
. 1997;139:1873–1884.
[CrossRef] [PubMed]Singer KH, Lazarus GS. Molecular mechanism of pathogenesis. Razzaque A eds. Pemphigus, Clinics in Dermatology. 1983;106–121. J. B. Lippincott Philadelphia.