| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Departments of
*
Microbiology/Immunology and
Dermatology, College of Medicine, University of Illinois, Chicago, IL 60612; and
Department of Biochemistry/Molecular Biology, Mayo Clinic and Medical School, Rochester, MN 55905
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To assess the function of Dsg 3 in vivo, Allen et al. (13) produced transgenic mice expressing NH2-terminally truncated Dsg 3. Within days of birth these mice showed swollen paws, flakiness on their backs, and blackening of the tail tip. These gross changes were accompanied by histological changes including increased intracellular spaces and disruption of desmosomes. Certain regions showed loss of desmosomes accompanied by epidermal thickening. Affected areas showed a dramatic increase in cell proliferation. Together these data showed that mutation in Dsg 3 can disturb adhesion of cells in the epidermis resulting in changes in the skin. To more directly address the importance of Dsg 3, Koch et al. (14) developed a mouse with targeted disruption of Dsg 3. These mice, although normal at birth, showed runting and lower weight by about 2–3 wk. Histopathological examination showed acantholysis in the oral cavity and traumatized skin, typical symptoms of PV. The phenotype of these mice was identical to that of bal mice, which lack Dsg 3. Based on the results from these studies, the authors concluded that autoantibodies might interfere directly with the adhesive function of Dsg 3 in the stratified squamous epithelium and cause disruption of the suprabasilar layer resulting in blisters.
Earlier, we and others have shown that the recombinant extracellular domain of Dsg 3 (E-Dsg 3) and full-length Dsg 3 can react with and neutralize blister-causing Abs in PV patient sera (15, 16, 17). We have further demonstrated that rabbits immunized with the full-length recombinant Dsg 3 and not with the ectodomain can produce pathogenic Abs, as determined by passive transfer to neonatal mice or by incubation with human skin (17, 18). However, to date, the pathogenesis of PV is not fully understood because of a lack of an appropriate animal model that can be used to study actively induced immune response to Dsg 3. To understand immunoregulation and to address the role of the immunogenetic component in the production of pathogenic Abs that are characteristic of PV, we turned to the development of a mouse model. Four strains of female mice were screened for their ability to produce pathogenic anti-Dsg 3 Abs.
Sera from immunized BALB/c, DBA/1, SJL/J, and HRS/J mice were tested for their Ab titers against native Dsg 3, recombinant Dsg 3, recombinant E-Dsg 3, and synthetic peptides spanning the entire Dsg 3. We evaluated Abs from different strains of mice for their pathogenic ability both in vivo and in vitro. Our results showed that only sera from BALB/c mice immunized with Dsg 3 are capable of inducing lesions.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Six- to 8-wk-old female BALB/c, DBA/1, SJL/J, and HRS/J mice were purchased from The Jackson Laboratory (Bar Harbor, ME). Groups of 10 mice were immunized four times with 20 µg/mouse of purified Dsg 3 protein in CFA (on days 1, 10, 20, and 30), four times with 20 µg/mouse of E-Dsg 3 (on days 40, 50, 60, and 70), and twice with 20 µg/mouse of refolded Dsg 3 (on days 80 and 90); then they were boosted two more times with 20 µg/mouse of E-Dsg 3 in IFA by i.p. inoculation (on days 97 and 104). Control groups of mice were similarly immunized with BSA.
Gel purification and refolding of Dsg 3 and E-Dsg 3
Recombinant Dsg 3 and E-Dsg 3 produced in insect cells were purified on SDS-PAGE, as described previously (17, 18). In short, enriched insect cell lysates containing either Dsg 3 or E-Dsg 3 were loaded onto a 7.5% preparative polyacrylamide gel flanked with broad-range, prestained molecular mass markers (Bio-Rad, Hercules, CA). Proteins were separated at 100 V for 5 h, and a 5 mm-wide band, corresponding to either the 115- to 130-kDa or 70- to 85-kDa range, was excised. Gels containing 130- and 115-kDa Dsg 3, and the 70- and 85-kDa E-Dsg 3 were incubated for 2 h at 37°C in bicarbonate elution buffer (50 mM ammonium bicarbonate, 0.1% SDS). The eluted Dsg 3 and E-Dsg 3 were lyophilized and the SDS was removed by repeated washing with cold 80% acetone. These gel pure proteins were used for immunizations. For renaturing Dsg 3, 1 mg of gel pure protein was solubilized in 40 ml of 6 M guanidine-HCl, neutralized with 20 ml of 1 M Tris-HCl (pH 7.5), and then diluted to 160 ml with water containing equimolar concentrations of cystine and cysteine (final concentrations of 1 mM each). The solution was kept at 4°C overnight and then dialyzed against 4 liters of 0.1% ammonium bicarbonate, with three changes of buffer. The dialyzed solution was lyophilized and resuspended in 100 mM Tris-HCl (pH 7.5) and 2 mM CaCl2 (17).
Titration of Abs to Dsg 3, E-Dsg 3, and synthetic peptides
The titers of Abs against both human and mouse Dsg 3, E-Dsg 3, and peptides were determined using an ELISA (19). To summarize, polystyrene Immulon 2 microtiter plates (Dynatech Laboratories, Chantilly, VA) were coated with 50 ng/well of either pure Dsg 3, E-Dsg 3, or peptides using 100 µl of carbonate/bicarbonate buffer (pH 9.6) and incubated overnight at 4°C. The plates were washed three times using PBS containing 0.05% Tween 20 in a microplate washer (Titertek, Huntsville, AL). Then the following reagents were added in succession with washing between each step: 100 µl of sera for 1 h at room temperature; 100 µl of HRP-conjugated goat anti-mouse IgG diluted 1:3000 (Caltag, Burlingame, CA) for 1 h at room temperature; and 100 µl of substrate solution containing 2.2'-azino-bis-(3-ethylbenzthiazoline-6-sulfonic acid) (Sigma, St. Louis, MO) and 0.03% of 30% hydrogen peroxide in 0.1 M citrate buffer (pH 4.2). Then the plates were incubated at room temperature for 10 min. The reaction was stopped by adding 100 µl of 1% SDS to each well. Substrate conversion was measured at 405 nM using a model 550 microplate reader (Bio-Rad).
Determination of isotype and subclass of anti-Dsg 3 and E-Dsg 3 Abs
Relative concentrations of IgG, IgM, IgG1, IgG2a, IgG2b, and IgG3 were measured by ELISA. The assays were performed as described above, using isotype or subclass-specific HRP-conjugated second Abs (Caltag) at a dilution of 1:2000.
Preparation of synthetic peptides
Forty-six peptides spanning the entire primary structure of the
human Dsg 3 (see Table III
) were synthesized as previously described
(20). Peptides designated 1 through 28 are derived from
the predicted extracellular domain (aa 1–592) of the human Dsg 3.
Peptide 00 (aa 593–616) represents the transmembrane region and
peptides 29–45 are derived from the predicted intracellular region (aa
617–976). Numbering does not include the first 23 amino acids, which
represent the putative signal peptide. In short, each peptide was
purified to homogeneity by reversed-phase HPLC, and a single
homogeneous peak of each peptide was collected and lyophilized. After
purification, the identity of each peptide was confirmed by amino acid
composition analysis and mass spectroscopy.
|
Abs from immunized mice were tested by indirect immunofluorescence staining on commercially available sections of monkey esophagus (17) (Inova, San Diego, CA) and sections of mouse tongue prepared in our laboratory. The bound Abs were detected using FITC-conjugated goat anti-mouse IgG diluted 1:40 (Caltag, South San Francisco, CA) as secondary Abs. Slides were examined under a Zeiss (Oberkochen, Germany) fluorescence microscope equipped with epi-illumination. To demonstrate specificity of Ab binding, antisera were preincubated with either Dsg 3 or a control protein before staining.
Induction of human skin lesion in vitro
Organ culture of normal human foreskin was used in these experiments as previously described (21). Fresh human foreskin was cut into small, 2 x 2-mm pieces and floated dermal side down in tissue culture dishes containing RPMI 1640 medium. We then added mouse serum. The culture was incubated at 37°C in a humidified CO2 incubator for periods ranging from 2 to 6 days. The skin explant specimens were processed for routine histologic study. Hematoxylin-eosin-stained sections were examined for the presence of acantholytic cells and suprabasilar cleft formation.
Passive transfer of Abs to neonatal mice
Igs from 10 ml of pooled serum from mice immunized with Dsg 3 were precipitated with 40% ammonium sulfate, dialyzed against PBS twice, lyophilized, and reconstituted in water to 1 ml. One hundred microliters of the reconstituted Abs/mouse was injected subcutaneously into neonatal BALB/c mice. Mice were examined 18–24 h postinjection for blister formation. Cross sections containing the blister and comparable areas in control animals were biopsied and frozen sections were prepared for routine histologic examination.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Groups of BALB/c, DBA/1, SJL/J, and HRS/J mice were immunized with
a combination of full-length Dsg 3 and E-Dsg 3, or with BSA. Mice
belonging to each strain mounted Ab responses against the recombinant
full-length Dsg 3 at variable titers (Fig. 1). SJL/J mice exhibited the highest
response, with DBA/1 mice showing the lowest titers. BALB/c and HRS/J
mice showed moderate responses. Because functionally relevant Abs react
primarily with the E-Dsg 3, we tested the sera against E-Dsg 3. Again
SJL/J mice had the highest titer, with DBA/1 mice showing the lowest
titer. Both HRS/J and BALB/c mice had modest responses but less than
that seen in SJL/J mice. The control sera reacted very weakly at a
concentration of <1:40 against both Dsg 3 and E-Dsg 3 (data not
shown). Although Ab responses against both Dsg 3 and E-Dsg 3 consisted
of IgG1, IgG2a, and IgG2b (Table I), the
Ab response to the E-Dsg 3 was predominantly of the IgG1 subclass
(Table I). In spite of high levels of Abs against E-Dsg 3, none of the
adult mice exhibited mucocutaneous lesions.
|
|
).
|
Sera from all four strains of mice were tested by indirect
immunofluorescence staining of frozen sections of mouse tongue (a
source of native Dsg 3) for their ability to react with native mouse
Dsg 3. As shown in Fig. 2, sera from all
four strains of mice were capable of reacting with the native mouse
Dsg 3.
|
, sera from all four different strains
of mice showed considerable reactivity against the mouse Dsg 3.
Moreover, with the exception of HRS/J mice, which had a slightly lower
reactivity, the other three strains of mice exhibited very similar
titers.
|
To assess the differences in binding specificity of anti-Dsg 3
Abs, we tested pooled sera obtained from the test and control mice (10
mice from each group) for their reactivities against a panel of 46
synthetic peptides spanning the entire Dsg 3 (Table III
). Each strain of mice exhibited its
own pattern of epitope preference. All four strains mounted an Ab
response against peptide 7, derived from the extracellular domain.
SJL/J mice mounted the broadest Ab response and showed reactivity
against peptides 1, 7, 10, 12, 23, 30, and 38. In contrast, sera from
DBA/1 mice recognized only peptide 7. BALB/c sera reacted strongly and
uniformly against peptides 7, 12, and 30, exhibiting the highest titer
against peptide 30 derived from the cytoplasmic region (Fig. 4). HRS/J mice exhibited only moderate
reactivity against peptides 7 and 23.
|
Sera from immunized mice were diluted in medium and incubated with
human foreskin. Only sera from BALB/c mice immunized with Dsg 3
specifically caused acantholysis, whereas sera from the other three
strains had no apparent effect (Fig. 5).
Histopathological examination of the lesion showed suprabasilar
separation of keratinocytes, which is characteristic of PV. To
determine the specificity of pathogenic Abs, sera were preincubated
either with E-Dsg 3 or a control Ag before testing for their ability to
induce acantholysis in human foreskin (Fig. 6). As expected, the pathogenic Abs were
neutralized only by E-Dsg 3 and thus failed to cause acantholysis. In
contrast, preincubation with a control Ag had no effect. This in vitro
finding was further corroborated using the neonatal passive transfer
mouse model (Fig. 7). Concentrated Igs
obtained from pooled sera of each mouse strain were inoculated into
neonatal mice and the mice were observed for blister formation. Only
neonatal mice which received Ig from Dsg 3-immunized BALB/c mice
developed the lesion.
|
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
There are two major issues concerning the development of an animal model for PV. One concerns induction of Abs that can react with native Dsg 3, and the other concerns the ability of Abs to cause lesions in adult or neonatal mice. Our earlier data from rabbits (17, 18) and the data we now have generated using mice clearly show that experimentally induced Abs react with native Dsg 3 and can cause blisters in neonatal mice. These observations are analogous to those made earlier with patient Abs (8, 9). Identification of a "susceptible strain" will allow studies to be undertaken aimed at understanding the molecular interactions leading up to the production of pathogenic Abs. Inability to induce blisters in adult mice by passive transfer of either patient or experimental animal sera, or by active immunization, could suggest several possibilities: 1) architecture of the skin in adult mice is quite different from that in neonatal mice and thus may resist lesion development; 2) amount of Abs required to induce lesions in adult mice might be significantly more than that required in neonatal mice; or 3) an appropriate strain of mice has not yet been identified. Future systematic studies could provide insights into factors that affect lesion development in neonatal vs adult mice.
In this study, we demonstrated that only BALB/c mice can mount a
pathogenic Ab response capable of causing acantholysis of human
foreskin in culture and blisters in neonatal mice. This observation
suggested that either H-2d or the BALB background
contains the immunogenetic makeup necessary for the production of
pathogenic anti-Dsg 3 Abs. SJL/J mice had the highest titer of Abs
against Dsg 3, and HRS/J mice had Ab levels comparable to the levels in
BALB/c mice as measured by ELISA. In spite of this, antisera from HRS/J
and SJL/J mice were nonpathogenic. This finding suggests that Abs in
SJL/J and HRS/J mice might not react as well with native Dsg 3 as the
Abs in BALB/c mice did. However, our results from indirect
immunofluorescence studies clearly show that there were no significant
differences in the Ab titer against the native autoantigen. Moreover,
these Abs were capable of reacting with mouse Dsg3 as demonstrated by
specific indirect immunofluorescence staining of mouse tongue (Fig. 2).
In addition, the level of reactivity among different strains of mice
was very similar (Fig. 3), with the exception of HRS/J mice, which
showed somewhat lower reactivity.
Because different subclasses of Abs can have different functional
effects and because a particular subclass is often associated with the
pathogenic potential, we tested for subclass-specific Ab responses. It
is clear from the data shown in Table I that there was no apparent
correlation between either the subclass distribution or the levels of
different subclasses of Abs, and the pathogenic potential. In this
study we were unable to purify sufficient amounts of different
subclasses of Abs to test whether they were responsible for
differential pathogenicity. Therefore, additional studies are needed to
fully resolve this. Moreover, it is possible that the available Ig
repertoire in BALB/c mice is different from those available in other
strains of mice.
Next, to see whether differences in fine specificities of Abs against
Dsg 3 could explain the difference in pathogenic ability of Abs, we
tested antisera against peptides spanning the entire Dsg 3. SJL/J mice
had a weak response to peptide 30, but had a stronger response to
peptide 12. DBA/1 mice had a strong reactivity only to peptide 7. HRS/J
mice, on the other hand, had a modest response to peptide 23 and a weak
response to peptide 7. BALB/c mice produced Abs reactive to peptides 7,
12, and 30, with the highest response against peptide 30 (Fig. 4),
which is derived from the cytoplasmic domain of Dsg 3. It is likely
that Abs reactive with either peptide 7 or 12 are not involved in
pathogenicity because higher levels of Abs against these peptides are
present in DBA/1 and SJL/J mice, respectively, yet sera from these
later mice were nonpathogenic. Therefore, peptide 30 (aa 638–658)
seems to be the most specific for BALB/c, yet it is unlikely to be a
pathogenic epitope because E-Dsg 3 (aa 1–592) alone is capable of
adsorbing out all pathogenic Abs (Fig. 6). These results are consistent
with our earlier observations of removal of pathogenic Abs from patient
sera by either Dsg 3 or E-Dsg 3 (17, 18).
Inability to identify critical Ab epitopes using synthetic peptides was not entirely surprising because they might not have displayed the relevant epitopes. Due to the large size of the protein, we synthesized contiguous instead of overlapping peptides and thus it is possible that the boundaries of the peptides might have been inappropriately selected. Because Dsg 3 is a glycoprotein, it is likely that glycosylation plays a critical role in the formation of B cell epitopes but that the synthetic peptides are devoid of sugars. Alternatively, the pathogenic Abs primarily recognize conformational epitopes formed by either contiguous or noncontiguous amino acids. Supporting this notion is a study by Amagai et al. (11), which showed that an amino terminal fragment of Dsg 3 expressed in bacteria can only partially adsorb pathogenic Abs in the sera of patients with PV and that the nonadsorbed Abs were still pathogenic, suggesting that Abs reactive to conformational epitopes were important. Consistent with this observation, an earlier study of ours showed the ability to induce pathogenic Abs only when rabbits were immunized with the renatured full-length protein, but not when immunized with the denatured protein (17, 18). Currently, efforts are underway to generate larger overlapping recombinant fragments of E-Dsg 3 and E-Dsg 3/E-Dsg 1 chimeric proteins, and mAbs capable of causing blistering, so that conformational epitopes that might be involved in pathogenic Ab binding can be identified.
|
Footnotes |
|---|
2 Current address: Department of Medicine, University of Chicago, Chicago, IL 60370.
3 Address correspondence and reprint requests to Dr. B. S. Prabhakar, Department of Microbiology/Immunology (MC790), College of Medicine, University of Illinois, E-709 Medical Science Building, 835 South Wolcott Avenue, Chicago, IL 60612. E-mail address:
4 Abbreviations used in this paper: PV, pemphigus vulgaris; Dsg 3, desmoglein 3; E-Dsg 3, extracellular domain of Dsg 3.
Received for publication April 21, 1999. Accepted for publication September 14, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  K. Tsunoda, T. Ota, M. Aoki, T. Yamada, T. Nagai, T. Nakagawa, S. Koyasu, T. Nishikawa, and M. Amagai Induction of Pemphigus Phenotype by a Mouse Monoclonal Antibody Against the Amino-Terminal Adhesive Interface of Desmoglein 3 J. Immunol., February 15, 2003; 170(4): 2170 - 2178. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Sekiguchi, Y. Futei, Y. Fujii, T. Iwasaki, T. Nishikawa, and M. Amagai Dominant Autoimmune Epitopes Recognized by Pemphigus Antibodies Map to the N-Terminal Adhesive Region of Desmogleins J. Immunol., November 1, 2001; 167(9): 5439 - 5448. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. T. Nguyen, A. Ndoye, and S. A. Grando Pemphigus Vulgaris Antibody Identifies Pemphaxin. A NOVEL KERATINOCYTE ANNEXIN-LIKE MOLECULE BINDING ACETYLCHOLINE J. Biol. Chem., September 15, 2000; 275(38): 29466 - 29476. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
