| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1*0102 Molecule1
* Institut National de la Santé et de la Recherche Médicale, Unité 519, Faculté de Médecine et de Pharmacie, Rouen, France;
Département d’Ingénierie et d’Etudes des Protéines, Commissariat à l’Energie Atomique-Saclay, Gif sur Yvette, France;
Institut National de la Santé et de la Recherche Médicale, Unité 413, Mont Saint Aignan, France; and
Laboratoire d’Immunologie et d’Histocompatibilite, Institut National de la Santé et de la Recherche Médicale, Unité 396, Paris, France
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
1*0102 molecule as shown in a HLA-DR peptide-binding assay, and induces PF T cell proliferation. These data provide an illustration of an autoantigen encoded by alternative spliced transcript that may participate in the pathogenesis of the disease by bearing PF-associated HLA class II restricted-epitope. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
Self-Ags play a major role in the occurrence of autoimmune diseases because they participate in both the initiation and maintenance of the autoimmune response and the development of lesional immune effectors (2). Several mechanisms have been proposed to explain why self proteins may become immunogenic (3). Among these, alternative splicing might operate through the production at the periphery of novel B or T epitopes to which lymphocytes have not been tolerized during their maturation in central lymphoid organs. The role of alternative spliced variants of protein self Ags in the breakage of immune tolerance has been demonstrated in several autoimmune diseases (4, 5).
Desmogleins (Dsg)3 consist of four different desmosomal transmembrane glycoproteins belonging to the desmosomal cadherin family (Dsg1 to Dsg4), which are encoded by separate genes located within the 250-kb cadherin cluster of the 18q12.1 chromosome (6). Desmosomal cadherins comprise five extracellular (EC) domains, a transmembrane domain, and a cytoplasmic region containing five domains, and allow a calcium-dependent cell-cell adhesion critical for desmosomal junction. Dsg2 is detected in all desmosome-possessing tissues, whereas Dsg1 and Dsg3 are restricted to stratified squamous epithelia-like epidermis (7). Dsg1 is the autoantigen of pemphigus foliaceus (PF), an autoimmune blistering skin disease characterized by an autoantibody response directed against conformational calcium-dependent epitopes of the Dsg1 EC domains, particularly the N-terminal adhesive region (EC1 and EC2 domains) (8, 9, 10). Anti-Dsg1 autoantibodies are pathogenic because, when transferred to normal mice, their in vivo binding to Dsg1 leads to a loss of adhesion between keratinocytes called acantholysis and the formation of intraepidermal blisters (11, 12). The tolerance breakdown against Dsg1 is not restricted to PF, because anti-Dsg1 Abs can also be found in other forms of pemphigus, including pemphigus vulgaris (PV) and paraneoplastic pemphigus (13, 14). The production of anti-Dsg1 Abs is dependent on Dsg1-specific Th lymphocytes, which exhibit a memory T cell phenotype and a Th2-like cytokine profile and are detected in PF patients (15). Like in most autoimmune diseases, HLA class II genes contribute to genetic susceptibility of PF, particularly DR1, DR4, and DR14. Several alleles of these haplotypes are associated with PF, including DRB1*0102, which has been found recurring in different studies (16, 17, 18, 19), notably those involving French PF patients (17, 19). The major mechanism proposed to explain the association between MHC class II genes and autoimmune diseases is that disease-associated alleles can efficiently accommodate autoantigen-derived peptides initiating and sustaining the autoreactive T cell response. Thus, in PF, a major objective is to identify T cell epitope peptides of Dsg1 that could initiate the autoimmune response (15).
In this study, we describe alternative transcript of DSG1 gene (DSG1-AT) containing an intronic insertion of 101 bp in normal human skin. This insertion introduces a stop codon that leads to the synthesis of a truncated isoform of Dsg1 carrying amino acid sequences encoded by intronic sequences and absent from the native Dsg1. This Dsg1-truncated isoform was shown to be expressed in normal human epidermis and to bear a specific peptide sequence binding to the PF-associated DRB1*0102 molecule and inducing a T cell response in PF patients.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
Peripheral blood was obtained from 21 adult patients with PF (n = 8) or PV (n = 13) as well as from 6 patients with bullous pemphigoid (BP) (Table I), who were seen in Dermatology Departments of the French Bullous Study Group. All patients gave written consent to participate in this study. Clinical diagnosis of PF, PV, and BP was confirmed by histopathology, direct immunofluorescence microscopy on perilesional skin, and serum analysis including indirect immunofluorescence microscopy on monkey esophagus (The Binding Site), or salt-split human foreskin sections, immunoblotting on human epidermis extract, and commercial ELISAs (Medical and Biological Laboratories), to detect circulating anti-Dsg1/3 and anti-BP180 autoantibodies in pemphigus and BP patients, respectively. Active-onset of the disease was defined as de novo development of blisters/erosions on skin and/or mucosa of untreated patients. Chronic active pemphigus was defined as incomplete remission or relapse in patients with immunosuppressive treatment. Clinical remission was defined as treated patients who had not experienced new cutaneous and/or mucosal lesions for 
6 mo before the study.
|
Genomic DNA was extracted from whole blood (2 ml) of pemphigus patients using QIAamp DNA Blood Midi Kit (Qiagen) and used for HLA class II DR1 typing as described previously (19).
RT-PCR analysis
Mammary surgery specimens of normal skin were used as source of epidermis. Total RNA was extracted from nine human epidermis using TRIzol reagent (Invitrogen Life Technologies). The cDNA was obtained by transcription of random-primed RNA by Moloney murine leukemia virus reverse transcriptase (Invitrogen Life Technologies), and then subjected to PCR amplification (RT-PCR). Primers used to amplify specific gene products were as follows: GAPDH sense, 5'-TGC CAT CAA CGA CCC CTT CA-3'; GAPDH antisense, 5'-TGA CCT TGC CCA CAG CCT TG-3'; DSG1 sense, 5'-AAT GGC TAC ATT TGC AGG ACA-3'; DSG1 antisense, 5'-ATC TCG GTC AGA GCC TCT TAC A-3'; DSG1-AT sense, 5'-TAG ACA GAG AGA TTT ATG TAA ACG TTG-3'; DSG1-AT antisense, 5'-TAT TAC AGG CAA TAT CAC ACT TGA C-3'. PCR conditions for GAPDH comprised one cycle of 94°C for 5 min, 35 cycles of 94°C for 1 min, 58°C for 30 s, and 72°C for 30 s, and a final elongation step of 72°C for 4 min. For PCR of DSG1 cDNA, conditions were identical except for the time of initial denaturation step and the annealing temperature, which were 4 min and 58°C, respectively. To amplify DSG1-AT cDNA, PCR conditions comprised 94°C for 4 min, 40 cycles of 94°C for 1 min, 58°C for 30 s, and 72°C for 30 s, and a final elongation step of 72°C for 4 min. All PCR were performed in a volume of 50 µl using an Eppendorf thermocycler, with 2.5 U of Taq polymerase (Promega). Amplification products (15 µl) were separated and visualized by ethidium bromide-stained 1.5% agarose gel electrophoresis.
Quantitative real-time PCR analysis
The cDNA synthesis of the nine human epidermis samples is described above. Quantitative real-time PCR were performed using the LightCycler thermocycler system (Roche Molecular Biochemicals). Amplification reactions were conducted using the FastStart DNA Master SYBR Green I kit (Roche) according to the manufacturer’s instructions in a total volume of 20 µl in glass capillaries, containing 2 µl of cDNA sample, 2 µl of FastStart DNA Master SYBR Green I, 2 mM MgCl2, and 10 pmol of each primer. PCR amplifications of DSG1 and DSG1-AT cDNAs were conducted simultaneously and comprised one cycle of 94°C for 10 min, 50 cycles of 94°C for 20 s, 61°C for 10 s, and 72°C for 12 s. PCR conditions for GAPDH amplification comprised one cycle of 94°C for 10 min, 40 cycles of 94°C for 20 s, 62°C for 10 s, and 72°C for 22 s. At the end of amplification cycles, PCR products were subjected to melting temperature analysis in 3 steps: 1) 10 s for 94°C (20°C/s), 2) 30 s for 66°C (20°C/s), and 3) a slight temperature increase of 0.1°C/s until 94°C to generate fusion curves. Data acquisition and analysis were conducted using LightCycler 3.5 software.
To determine the absolute copy number of target transcripts, purified GAPDH, DSG1, and DSG1-AT cDNAs were used to generate calibration curves. cDNAs were amplified by classic PCR as described above, purified using the Qiaquick Gel Extraction Kit (Qiagen), and quantified with both measure of the absorbance at 260 nm and comigration on a 1.5% agarose gel with the Mass Ruler DNA Ladder (MBI Fermentas). Conversion of micrograms to copy number was done using the following formula: n (copy) = [(m x 1515)/Nbp] x 10–12 x N (m: µg of dsDNA, Nbp: length of the amplicon in bp, N: the constant of Avogadro). Serial dilutions were used to create calibration curves by plotting the threshold cycle vs the known copy number of each purified amplimer in the dilutions (20, 21). PCR efficiency was calculated using calibration curves with the following formula: e = 10–1/s (s: slope) (with an efficiency equal of 2 (s = –3.32) that give an optimal value of 100%). For each human epidermis sample, calibrators spanning five dilution points, GAPDH or DSG1 and AT-DSG1 cDNAs, and nontemplate controls were amplified in triplicate, simultaneously in one LightCycler run. Following each run, amplification reactions and amplimer sizes were checked by ethidium bromide-stained 1.5% agarose gel electrophoresis.
To correct for differences in both RNA quality and quantity between samples, data were normalized by dividing the copy number of DSG1 or DSG1-AT cDNA by the copy number of GAPDH (22). Quantitative results are presented as copies of the target cDNA per 1000 copies of GAPDH (23, 24). Relative quantity of DSG1-alternative transcript was determined, dividing the absolute copy number of these mRNAs by those of the total Dsg1 mRNAs (normal and alternative).
Peptides
Six peptides of the Dsg1-truncated isoform were synthesized, purified by HPLC (with purity >99%), and controlled by mass spectrometry (Institut de la Santé et de la Recherche Médicale, Unité 413). Two peptides were used in the immunization protocol (see below): the INT6 peptide (IYVNVEPTFQRTLHKTK) encoded by the intron 6 insertion is specific to the Dsg1-truncated isoform; and the EC2 (184–193) peptide (ADEPNNLNSK) is present in both Dsg1 and Dsg1-truncated proteins, and was chosen on the basis of antigenicity, accessibility, and hydrophilicity criteria. Four overlapping 20-mer peptides between the EC2 domain and INT6 peptide of the Dsg1-truncated isoform, EC2/INT6 211–230 (INRNTGEIRTMNNFLDREIY), EC2/INT6 216–235 (GEIRTMNNFLDREIYVNVEP), EC2/INT6 221–240 (MNNFLDREIYNVEPTFQRT), and EC2/INT6 226–245 (DREIYVNVEPTFQRTLHKTK), were used in the HLA-DR peptide-binding assays.
Animals and immunization
New Zealand (NZ) White rabbits and CD1 mice were purchased from Charles River Laboratories. For animal immunization, EC2 (184–193) and INT6 synthetic peptides were coupled to bovine thyroglobulin (BT) using glutaraldehyde. Mice and rabbits were primed respectively with 40 µg (i.p.) and 200 µg (s.c.) of BT-coupled peptides emulsified in CFA and boosted eight times every 3 wk for 7 mo with BT-coupled peptides in IFA. IgG anti-EC2 (184–193) or anti-INT6 peptides were detected by ELISA using the corresponding peptide as Ags and immunoblotting of purified recombinant Dsg1-truncated isoform synthetized by in vitro transcription/translation (EC1/2-INT6).
Peptide ELISA
Ninety-six-well microtiter plates (Maxisorb Nunc) were coated with 100 µl of EC2 (184–193) or INT6 synthetic peptides (10 µg/ml), in PBS and incubated overnight at 4°C. After washing three times with PBS-0.1% Tween 20 (PBST), wells were blocked with PBS-1% Tween 20–5% sucrose-3% milk powder for 1 h at room temperature. After removal of excess blocking solution, mouse or rabbit sera (1/100 in PBST-1% BSA) were added and incubated 1 h at room temperature. After washing with PBST, the plates were incubated with 1/10,000-fold diluted biotin-conjugated goat anti-mouse or anti-rabbit IgG (Caltag Laboratories) for 1 h. After three washes, 1/10,000-fold diluted alcalin phosphatase-conjugated streptavidin (Caltag Laboratories) was added for 30 min at room temperature. The plates were washed, and labeling was revealed by adding 100 µl of p-nitrophenyl phosphate (Sigma-Aldrich). OD were determined at 405 nm with an ELISA reader (Labosystems). All assays were run in duplicate, and background values given by each animal serum tested in uncoated wells were subtracted. To control the specificity of the assay, preimmune sera were included in each experiment.
Indirect immunofluorescence assay
Mouse and rabbit sera (1/50) were analyzed by indirect immunofluorescence on human foreskin sections using FITC-conjugated rabbit anti-mouse IgG Abs or FITC-conjugated goat anti-rabbit IgG Abs (Sigma-Aldrich) as the tracer. Sections were examined using a fluorescence microscope (Zeiss).
In vitro transcription translation
The Dsg1-truncated isoform was obtained as a recombinant polypeptide, called EC1/2-INT6, using an in vitro transcription-translation system. cDNA coding for EC1/2-INT6 was prepared from human epidermis RNAs,and was amplified with a specific pair of primers: 5'-GAA TGG ATC AAG TTC GCA GCA G-3' (sense); 5'-TTA CTT TGT CTT ATG TAA AGT TCT TTG-3' (antisense). The purified PCR product was cloned in-frame in the pSPUTK in vitro translation vector (Stratagene) at the ApaI and BamHI restriction sites and sequenced with an automated sequencer (PerkinElmer Applied Biosystems), using the SP6 and T7 primers. Sequences were aligned and analyzed using the GenBank/EMBL databases (National Center for Biotechnology Information/Basic Local Alignment Search Tool network service). Then, 2 µg of plasmid preparation were used for in vitro expression in the TNT6 SP6 Quick Coupled Transcription-Translation System (Promega) according to the manufacturer’s instructions. To obtain purified product, 2 µl of Transcend Biotin-Lysyl tRNA (Promega) were added to the reaction mixture, and the biotinylated polypeptides were captured by SoftLink Avidin Resin (Promega) and eluted with Laemmli solution. The production and purity of EC1/2-INT6 were controlled by SDS-PAGE analysis on a 14% separating gel followed by immunoblotting with a peroxidase-conjugated streptavidin (Promega).
Immunoblot analysis
Mammary surgery specimens of normal human skin were used as the source of epidermis. Protein extraction was performed as described previously (25). Total protein extracts (75 µg/well) of normal kidney and small intestine were also used as the other source of epidermal cells. Proteins were separated by SDS-PAGE with a 14% separating gel according to the Laemmli’s method and electrotransferred onto nitrocellulose membranes. The filters were then saturated for 1 h in TBS-0.05% Tween 20 (TBST) –5% dry milk, and incubated with rabbit or mouse sera (1/50) in TBST-5% dry milk for 2 h. After washing, filters were incubated for 1 h with 1/20,000-diluted peroxidase-conjugated goat anti-rabbit IgG or 1/80,000-diluted peroxidase-conjugated rabbit anti-mouse IgG (Sigma-Aldrich) in TBST-5% dry milk. The filters were washed and revealed by chemiluminescence reaction (ECL Supersignal; Amersham Biosciences). For infrared fluorescent (IRF) immunoblot experiments performed with the Odyssey Infrared Imaging system, procedures were followed as described previously (26). Briefly, after a saturating step, the membranes were incubated simultaneously with rabbit anti-EC2 and mouse anti-INT6 sera (1/50) for 2 h, then the membranes were incubated simultaneously with biotin-conjugated goat anti-rabbit IgG (Caltag Laboratories) and Alexa Fluor 680-conjugated goat anti-mouse IgG (Molecular Probes) (1/6,000) for 1 h. Finally, membranes were incubated for 1 h with IRDye 800-conjugated streptavidin (Rockland) (1/6,000), and examined with the Odyssey Infrared Imaging system (LI-COR Biosciences).
Two-dimensional-PAGE immunoblot analysis
Human epidermis was ground with a liquid nitrogen-cooled mortar and pestle, and proteins were extracted from the resulting powder as previously described by Görg et al. (27). The protein extract (60 µl) was then subjected to two-dimensional-PAGE in which the first dimension was conducted on ReadyStrip IPG strips (11 cm, nonlinear pH 3–10 gradients; Bio-Rad) using the PROTEAN IEF cell system (Bio-Rad) according to the manufacturer’s instructions. After IPG strip equilibration, the second dimension was run on 14% polyacrylamide gel (11 cm x 8 cm x 1 mm) using the Criterion Dodeca Cell system (Bio-Rad). Finally, gels were transferred onto nitrocellulose membranes. The IRF immunoblot analysis on protein-map replica was performed as described above.
HLA-DR peptide-binding assays
HLA-DRB1*0101, *0401, *0102, and *0402 molecules were immunopurified using L243 monomorphic Abs from HOM-2, BOLETH (a gift from Dr. C. de Toma, Centre d’Etude du Polymorphisme Humain, Paris, France), MZ07, and YAR cell lines, respectively. The binding capacities of the peptides were investigated in a competitive ELISA as described previously (28, 29). Briefly, the biotinylated peptide was incubated at pH 6 with a dose range of peptides to be tested and a fixed quantity of HLA-DR molecules. After 24–72 h, HLA-DR peptide complexes were trapped by L243 Abs adsorbed on ELISA plates. After washing, presence of the biotinylated peptide was revealed using streptavidin-alkaline phosphatase conjugate (Amersham Biosciences) and 4-methylumbelliferyl phosphate substrate. Emitted fluorescence was measured at 450 nm upon excitation at 365 nm. The biotinylated peptides used were as follows: HA 306–318 (PKYVKQNTLKLAT) for HLA-DRB1*0101 (1 nM) and HLA-DRB1*0401 (30 nM); and ALK (AALAAAKAAAAAA) for HLA-DRB1*0102 (10 nM) and HLA-DRB1*0402 (10 nM). Data were expressed as the concentration of peptide that prevented binding of 50% of the labeled peptide (IC50). Each value is the mean from two independent experiments.
Proliferation assays
PBMC from pemphigus patients (2 PF and 1 PV) were isolated by Ficoll-Hypaque Plus (Amersham Biosciences) density gradient separation. Two x 105 PBMC were cultured in triplicate in 200 µl of RPMI 1640 (Cambrex BioScience) supplemented with 10% autologous plasma and containing either EC2/INT6 216–235 peptide, EC2/INT6 226–245 peptide (20 µg/ml), or PHA (Sigma-Aldrich) at 5 µg/ml for 6 days at 37°C in 5% CO2. PBMC cultures were preformed without peptide and served as negative controls in each experiment. Cells in each individual well were pulsed with 1 µCi of [3H]thymidine (Amersham Biosciences) during the last 18 h of incubation, and were then harvested using an automated cell harvester (PerkinElmer Applied Biosystems). Proliferation of PBMC was determined by measuring the [3H]thymidine uptake (cpm) on a beta counter (Wallac). Data were presented both as average cpm ± SD and as stimulation index (SI): cpm of cells treated with EC2/INT6 peptides/cpm of cells without peptide. A SI 2 was considered as a positive response.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
To produce recombinant Dsg1, we previously sequenced a cDNA fragment encoding the entire EC domain of the protein (EC1–EC5) that was obtained by transcription of oligo(dT)-primed keratinocyte RNAs followed by nested PCR using primers located respectively at the 5' and 3' end of the nucleotide sequences encoding EC1 and EC5 (26). Surprisingly, we also found cDNAs corresponding to alternative DSG1 mRNAs that contained a 101-bp insertion corresponding to the 3' end of intron 6 localized between exon 6 and exon 7, which both code with exon 5 for the EC2 EC domain of Dsg1 (Fig. 1A). The intron 6 insertion contains a stop codon that could lead to a premature stop of the translation and thus to the production of a truncated isoform of Dsg1. If synthesized, this isoform would be composed of the EC1 domain, a part of the EC2 domain (158–227 aas), and an additional 17-mer peptide encoded by the intron 6 insertion (INT6) and absent from the transmembrane Dsg1 (Fig. 1B). To confirm the existence of DSG1-alternative transcript (DSG1-AT) in human epidermis, we amplified a specific cDNA sequence (amplimer of 101 bp) obtained from epidermis mRNAs, by PCR using primer sense (DSG1-AT sense) and primer antisense (DSG1-AT antisense) hybridizing the exon 6-intron 6 junction and the intron 6, respectively (Fig. 1A). Alternative transcripts of DSG1 were detected in all human epidermis samples tested (n = 9) (Fig. 2A). Moreover, both DSG1-transcripts were also detected in cultured keratinocytes from human epidermis (data not shown). Therefore, alternative mRNAs of DSG1 are physiologically expressed by keratinocytes of human epidermis.
|
|
To quantify the expression of DSG1-alternative transcripts in human epidermis, we performed quantitative real-time PCR using LightCycler thermocycler. The efficiency of PCR amplifications was checked using standard curves generated by serial dilutions of purified DSG1-AT, DSG1, and GAPDH amplimers. PCR were very efficient, reaching an efficiency of 99.5% for GAPDH, DSG1, and DSG1-AT amplifications (Fig. 2B), and allowing the quantification of the two DSG1 transcripts. Quantitative real-time PCR were then conducted for the nine human epidermis samples, and the specificity of each amplified product was confirmed by the fusion curve analysis (Fig. 2, C and D). Data obtained were normalized with GAPDH to evaluate the absolute copy number of both DSG1 and DSG1-AT transcripts in each of the nine samples (Fig. 2E). The relative quantity of alternative transcript of Dsg1 expressed in human epidermis ranged from 6 to 30% with an average of 
15% (Fig. 2F). Therefore, one can conclude to an interindividual variability of the DSG1-alternative transcript expression in normal human skin.
Production and characterization of polyclonal Abs directed against EC2 and INT6 peptides
All CD1 mice immunized with the INT6 peptide, specific of the Dsg1-truncated isoform, and the NZ rabbit immunized with the EC2 (184–193) peptide, present in both Dsg1 and Dsg1-truncated proteins, produced IgG reacting with INT6 and EC2 peptides in a solid phase ELISA, respectively (Fig. 3A). Anti-peptide sera were then tested for their capacity to react with the recombinant truncated isoform of Dsg1 (EC1/2-INT6) by immunoblot analysis. Both anti-INT6 and anti-EC2 sera reacted with EC1/2-INT6 recombinant polypeptide that exhibited a molecular mass of 25-kDa (Fig. 3B). Next, to determine whether anti-EC2 peptide Abs also recognized the 184–193 aa sequence present in native Dsg1, indirect immunofluorescence analysis of human foreskin and immunoblotting experiment of human epidermis were performed with the rabbit antiserum. IgG anti-EC2 bound to the intercellular space of human epidermis (Fig. 4B) with a labeling pattern similar to that obtained with PF anti-Dsg1 Abs on human foreskin sections (Fig. 4A). Interestingly, anti-INT6 mouse sera gave a labeling pattern of human epidermis located at the suprabasal keratinocyte level (Fig. 4C). IgG anti-EC2 also reacted with a 160-kDa protein present in human epidermis extract that comigrated with Dsg1 (Fig. 5A).
|
|
|
Once characterized, anti-EC2 and anti-INT6 sera were used as specific probes in immunoblot experiments on epidermal extract to determine the existence of the truncated isoform of Dsg1 in human epidermis. All four anti-INT6 sera reacted with a 25-kDa polypeptide doublet; whereas anti-EC2 serum reacted with a single 25-kDa protein. None of the preimmune sera bound to epidermal proteins with this molecular mass (Fig. 5A). Immunoblotting with anti-INT6 sera showed that the 25-kDa polypeptide doublet was also present in a cultured-keratinocyte extract (data not shown). When the replica was incubated with anti-INT6 and anti-EC2 sera simultaneously, the IRF immunoblot showed that the lower band of the 25-kDa polypeptide doublet was recognized by both sera (Fig. 5B, lane 3, yellow band) as well as the 25-kDa recombinant Dsg1-truncated isoform (Fig. 5C, lane 3, yellow band). In contrast, both Dsg1 and Dsg1-truncated isoform were not detected in other tested human epithelial tissues like kidney and small intestine (data not shown). Thus, the immunological identity of the 25-kDa protein present in human epidermal extract with the recombinant Dsg1-truncated isoform constitutes a first strong argument speaking for the Dsg1-truncated isoform expression in human epidermal cells. This argument was further supported by an IRF immunoblot analysis of a two-dimensional-PAGE-separated human epidermis protein map with anti-INT6 and anti-EC2 sera as specific probes of the truncated isoform of Dsg1. Both sera bound to a single epidermal protein (Fig. 6), which had molecular mass and isoelectric point (pI) coordinates (25-kDa, pI 4.7) close to those theoretically calculated for the truncated isoform of Dsg1 (22.4-kDa, pI 5.5) (Fig. 6C, yellow spot) (data obtained in triplicate), and that correspond to the lower band of the 25-kDa polypeptide doublet recognized by both anti-INT6 and anti-EC2 sera in IRF immunoblotting experiments.
|
1*0102 molecule
The truncated isoform of Dsg1 bears amino acid sequences that are absent from the transmembrane Dsg1, particularly the INT6 peptide and peptide-overlapping sequences with INT6 and the EC2 domain. This observation is reminiscent of the two isoforms of myelin basic protein (MBP) (21.5-kDa and 20.2-kDa), which are expressed during the remyelinization process, and contain an exon 2-encoded protein, X2MBP, which is absent from the major isoform of MBP (18.5-kDa). In multiple sclerosis, specific T cell responses directed against X2MBP were associated with the disease (30). To test the possibility that peptides of the Dsg1-truncated isoform bind to HLA class II molecules that predispose to PF, we synthetized four overlapping 20-mer peptides between EC2 domain and INT6 peptide (EC2/INT6 211–230, EC2/INT6 216–235, EC2/INT6 221–240, and EC2/INT6 226–245) and evaluated by a competitive ELISA, their relative binding capacity to HLA class II, DR1*0102, and DR1*0402, which were previously demonstrated to be associated with the disease in French PF patients (17, 19). Interestingly, the EC2/INT6 216–235 peptide bound to the DR1*0101, *0102, *0401, and *0402 molecules, because it exhibited toward all these molecules IC50 values inferior to the binding activity threshold of 1000 nM (Table II). Moreover, as compared with the reference peptides that are very good binders, its IC50 was of 3.4 toward DRB1*0102 and less high a factor of 20 toward DRB1*0101, *0401, and *0402 molecules. It can therefore be considered as a good peptide binder for these molecules.
|
To determine whether T cells proliferate in response to the EC2/INT6 216–235 peptide, we performed proliferation assays with PBMC from PF (n = 8) and PV (n = 13) patients (Table I). PBMC from four PF (PF1, PF2, PF5, and PF7) (50%) and one PV (PV4) (7.7%) patients showed a proliferative response to the EC2/INT6 216–235 peptide, whereas cells from only one PF patient (PF5) proliferated in presence of the EC2/INT6 226–245 peptide (Fig. 7). Interestingly, two PF patients (PF1 and PF7) whose PBMC showed a positive response to the EC2/INT6 216–235 peptide (SI = 3.2 and 3, respectively), carry the DR1*0102 allele (Fig. 7C). PBMC from PF2 patient, who carries the DR1*0101 allele previously reported to be overrepresented in PF (31) and shown to bind efficiently to EC2/INT6 216–235 peptide in our study, proliferated in response to this peptide (SI = 2.2). Only one PV patient carrying the DR1*0402 allele, which was detected in 8 of 13 PV patients (Fig. 7B), exhibited a significant T cell proliferation induced by the EC2/INT6 216–235 peptide (SI = 5) (Fig. 7B). No correlation was observed between the capacity of T cells to proliferate in response to the EC2/INT6 216–235 peptide and the presence of anti-Dsg1 Abs at the time of the study. Finally, T cells from BP patients (n = 6) did not proliferate in presence of the EC2/INT6 226–245 peptide (Fig. 7C).
|
1*0102 molecule also can induce the proliferation of T cells from PF patients. |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
This truncated isoform of Dsg1 may have different roles. First, from a physiological viewpoint, this truncated isoform that lacks the major part of the EC domain, the transmembrane domain, and the entire intracellular domain of the native Dsg1, could be soluble, unable to exert cell-cell adhesion function, and thus could induce a fragility of desmosomal junction. Indeed, it was previously shown that in vitro-synthesized truncated form of Dsg1 compromises desmosomes when added to keratinocyte tumor cell line culture (33). Second, and from an immunological viewpoint, it is well established that the translation of autoantigen isoforms encoded by alternative transcripts can play a role in the initiation and maintenance of systemic or organ-specific autoimmune responses through different mechanisms (32). A first mechanism proposes that an autoantigen isoform expressed in the thymus and lacking a T cell epitope borne by the full autoantigen expressed at the periphery, may prevent clonal deletion of autoreactive T cells. This is illustrated by murine experimental allergic encephalomyelitis in which the thymic expression of DM20, a shorter form of the proteolipid protein lacking 35-aa stretch, has been correlated with the presence of autoreactive T cells specific to the complete form of proteolipid protein expressed in the CNS but excluded from the thymus (4). Similarly, an alternatively spliced transcript of the pancreatic islet cell Ag 512 (ICA512/IA-2), which lacks sequence that encodes several T cell epitopes and is exclusively expressed in thymus, has been described and could be responsible for a tolerance breakdown against ICA512/IA-2 in type 1 diabetes (5). According to another mechanism, autoantigen isoforms could contain new epitopes susceptible to be recognized specifically by autoreactive T cells or autoantibodies. For example, in multiple sclerosis, specific T cell responses against X2MBP, an exon 2-encoded protein expressed by two isoforms of MBP during remyelinization process (21.5- and 20.2-kDa) and, absent from the major isoform of MBP (18.5-kDa), were correlated with disease progression (30). Closer to our observation is the demonstration that a truncated and soluble isoform of the IL-6 signal-transducing molecule gp130, which is translated from alternatively spliced mRNA, contains a specific epitope frequently targeted by the autoimmune response in patients with rheumatoid arthritis (34). One should add that the in silico study cited above (32) showed, by using appropriate algorithms, that most alternative spliced isoform regions of autoantigens encode potential MHC class I- and/or class II-restricted T cell epitopes.
These considerations prompted us to examine the potential role of the truncated isoform of Dsg1 in the T cell autoimmune response occurring in the course of PF. We tested the possibility that specific amino acid sequences of the truncated protein could bind to PF-associated HLA class II molecules. We thus set up an HLA-DR peptide-binding assay to evaluate the relative binding affinities of four peptides overlapping EC2 domain and INT6 sequences to four HLA class II molecules. The overlapping peptide EC2/INT6 216–235 of the Dsg1-truncated isoform bound to all HLA-DR molecules tested and to the PF-related DR1*0102 allele with the highest affinity. Moreover, the EC2/INT6 216–235 peptide was able to induce T cell proliferation in four PF patients (50%). Three of them carried the HLA class II alleles (DR1*0102 and *0101) shown to accommodate with a high affinity the selected peptide in our binding experiments. Although the subtype of the DR14 molecule expressed by the fourth PF patient, whose PBMC proliferated was not determined, one should mention that several PF-associated DR14 molecules, namely *1401 and *1406, share amino acid residues (LLEQRRAA) at critical positions of the peptide-binding groove with DR1*0102 and *0101 (31). One should also note that T cells from a DR1*0101 PF patient (PF6) did not proliferate when stimulated by the selected EC2/INT6 peptide, although the patient had a high anti-Dsg1 Ab titer (202 U/ml). As a matter of fact, T cell proliferation not only requires peptide presentation by APCs but also a sufficient number of peptide-specific T cells with adequate functional properties. Thus, on the one hand and according to their HLA class II genotype and T cell repertoire, not all PF patients may have the capacity to mount an autoimmune response to the T cell epitope borne by the Dsg1-truncated isoform, and, on the other hand, the clinical status (remission vs acute onset) and the therapeutic regimen may modulate the proliferative capacity of T cells. Among 13 PV patients, only one had a T cell proliferative response to the EC2/INT6 216–235 peptide. Interestingly, this patient had a mucocutaneous phenotype of the disease and carried the DR1*0402 allele that also was shown to bind the Dsg1-truncated isoform-derived peptide. PBMC obtained from other mucocutaneous PV bearing the DR1*0402 allele did not proliferate, indicating that this PV-associated HLA class II allele is per se not sufficient to trigger T cell proliferation and that other molecular and cellular factors are required.
Therefore, on the basis of results observed in PF patients, one can hypothesize that the soluble Dsg1-truncated isoform is processed by APCs bearing DR1*0102 molecules, which then present specific epitope EC2/INT6 216–235 to autoreactive T cells. Once primed, these T cells could activate Dsg1-specific B-cells and, finally, trigger the production of anti-Dsg1 autoantibodies.
In conclusion, we demonstrated the existence of a truncated isoform of Dsg1 in human epidermis, encoded by alternative transcript, and possessing a specific peptide bound by PF-associated HLA class II DR1*0102 molecules that could participate in the tolerance breakage to Dsg1.
|
Acknowledgments |
|---|
|
Disclosures |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
|
Footnotes |
|---|
1 This work was supported by funds from the Institut National de la Santé et de la Recherche Médicale. H.M. is the recipient of a fellowship from the Conseil Régional de Haute-Normandie. 
2 Address correspondence and reprint requests to Dr. François Tron, Institut National de la Santé et de la Recherche Médicale, Unité 519, Faculté de Médecine et de Pharmacie, 22 boulevard Gambetta, 76183 Rouen Cedex 1, France. E-mail address: francois.tron{at}chu-rouen.fr
3 Abbreviations used in this paper: Dsg, desmoglein; EC, extracellular; PF, pemphigus foliaceus; PV, pemphigus vulgaris; BP, bullous pemphigoid; BT, bovine thyroglobulin; IRF, infrared fluorescent; SI, stimulation index; pI, isoelectric point; MBP, myelin basic protein; NZ, New Zealand.
Received for publication December 23, 2005. Accepted for publication August 7, 2006.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionDisclosuresReferences |
|---|
/MYH11 fusion transcript by real-time RT-PCR in patients with INV(16) acute myeloid leukemia. Leukemia 15: 1072-1080. [Medline]This article has been cited by other articles:
|
|
 |
|
  J. Yamagami, S. Kacir, K. Ishii, A. S. Payne, D. L. Siegel, and J. R. Stanley Antibodies to the Desmoglein 1 Precursor Proprotein but Not to the Mature Cell Surface Protein Cloned from Individuals without Pemphigus J. Immunol., November 1, 2009; 183(9): 5615 - 5621. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) or preimmune sera (
)), biotin-conjugated goat anti-mouse or anti-rabbit IgG, and phosphatase-conjugated streptavidin. The reaction was revealed by adding p-nitrophenyl phosphate. OD were determined at 405 nm. NZ White rabbit serum; CD1, CD1 mouse sera. B, The purified biotinylated-EC1/2-INT6 recombinant protein was electrotransferred onto nitrocellulose filters that were then saturated, incubated with mouse or rabbit serum (1/50) and with peroxidase-conjugated goat anti-mouse or anti-rabbit IgG, and revealed by chemiluminescence reaction. Lane 1, peroxidase-conjugated streptavidin; lane 2, anti-EC2-NZ rabbit serum; lanes 3–6, anti-INT6-CD1 mouse sera; lane 7, preimmune NZ rabbit serum; lanes 8–11, preimmune CD1 mouse sera.
), with the EC2/INT6 226–245 peptide (
) or 216–335 (
)) and cultures without peptides are shown in C. A SI >2 was considered as a positive response. Cells from four PF (50%) and one PV (7.7%) patients showed a proliferative response to the EC2/INT6 216–235 peptide, whereas cells from the BP patients did not.