| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Division of Rheumatology, Allergy and Clinical Immunology, University of California School of Medicine, Davis, CA 95616;
Department of Pathology, Emory University School of Medicine, Atlanta, GA, 30322;
Department of Pathology, Kanazawa University, School of Medicine, Kanazawa, Japan; and
§
Department of Microbiology, Monash University, Clayton, Victoria, Australia
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We have taken advantage of an expressed recombinant protein (pML-MIT3) that is a "trihybrid" protein containing the lipoyl domains of PDC-E2, OGDC-E2, and BCOADC-E2. Using this trihybrid, we immunized BALB/c mice to produce 35 new mAbs specific for one or more of the mitochondrial autoantigens. We report herein that 7 of 35 of these mAbs, including Abs reactive to the individual mitochondrial Ags as well as Abs reactive to all three, uniquely stained the apical region of BEC in PBC. Moreover, the majority of mAbs (28 of 35) to PDC-E2, OGDC-E2, and BCOADC-E2 do not show this altered level of staining. Thus, the material reactive in BEC of PBC tissue shows only some of the immunologic features of these three Ags. These data suggest that the AMA response occurs to an Ag that shares conformational determinants with the 2-oxoacid dehydrogenase enzymes. Enhanced responses to PDC-E2, BCOADC-E2, and OGDC-E2 may occur later by a process of determinant spreading (17). We believe this cross-reactive molecule is imported, not produced by BEC, and that the identification of the original immunogen of the AMA response will provide insight for the etiology of PBC.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Briefly, a 414-bp EcoRI fragment coding for the
PDC-E2 inner lipoyl domain and part of the outer lipoyl domain, amino
acid residues 91–228 of the previously mature PDC-E2, was amplified by
PCR and expressed as described (18, 19). Transformants were tested for
expression of recombinant PDC-E2 by a colony immunoassay, and the
presence of the cDNA inserts was determined with both DNA hybridization
using 32P-labeled human PDC-E2 cDNA fragment as a probe and
plasmid DNA analysis on 1% agarose gel electrophoresis. Similarly, the
OGDC-E2 epitope (amino acid residues 67–147) and the BCOADC-E2 epitope
(amino acid residues 1–118) were sequentially cloned into the
NotI and BamHI site of the same pGEX 4T-1.
Successful cloning and expression of OGDC-E2 and BCOADC-E2
were confirmed by DNA hybridization and immunoassay with rabbit
anti-BCOADC-E2 Abs and affinity-purified PBC patient serum against
OGDC-E2, respectively. Additionally, the OGDC-E2 and
BCOADC-E2 subclones were produced by the insertion of the
NotI cDNA fragment and the BamHI cDNA fragment,
which were used for the construction of the trihybrid, into pGEX 4T-1
for expression (Fig. 1
). As a control
immunogen, glutathione S-transferase (GST) protein was
prepared as described above.
|
The purified trihybrid molecule was suspended at 1 mg/ml in PBS and mixed with an equal volume of CFA (Difco, Detroit, MI). Emulsion (100 µl) was injected i.p. into BALB/c female mice. Thereafter, mice were boosted at the same dose but emulsified in IFA (Difco), three or four times at 2-week intervals. Immunized mice were killed, and spleen cell suspensions were made and washed with HBSS (JRH Biosciences, Lenexa, KS). Myeloma cells (P3X63-Ag4-653) were collected and washed with HBSS. Spleen cells were mixed with myeloma cells and centrifuged at 1500 rpm for 5 min. After removing the supernatant, 1 ml of 50% polyethylene glycol 1500 in PBS (Hybrimax, Sigma Chemical Co., St. Louis, MO) was added dropwise to the cell pellet and incubated for 5 min. After centrifugation, the cells were resuspended in Iscove’s modified Dulbecco’s medium (JRH Biosciences) containing hypoxanthine, aminopterin, and thiamine (Sigma) and seeded to the 96-well plates at 5 x 105 cells/well. Supernatant fluids of the growing hybridoma cells were screened for reactivity against the trihybrid fusion protein by ELISA, and reactive cultures were cloned by limiting dilution. The cloned cells were further screened for reactivity against each Ag by ELISA. Specific reactivities were also confirmed by immunoblotting as described below. For the ELISA screening process, the purified recombinant fusion protein and, for purposes of control, GST were coated onto microtiter plates (Falcon, Becton Dickinson, Mountain View, CA) at 10 µg/ml in PBS overnight at 4°C. After washing three times with 0.05% Tween-20 in PBS (PBS/Tween), the plates were blocked with 3% dry milk powder in PBS for 1 h at room temperature, and an ELISA was performed as described.
In efforts to define specificity, both immunoblotting of beef heart
mitochondrial fraction (BHM) and ELISA using individual recombinant
proteins were performed. BHM was prepared as described previously (9).
Briefly, BHM was resuspended in 250 µl of sample buffer (125 mM
Tris-HCl (pH 6.8) containing 4% SDS, 20% glycerol, and 5% 2-ME),
boiled for 3 min, and resolved by SDS-PAGE using 1.5 mm-thick slab gels
with a 4.75% stacking gel and a 12% separating gel. Separated
proteins were transferred electrophoretically to nitrocellulose filters
(Micron Separations, Westboro, MA). After transfer, nitrocellulose
filters were blocked in 3% milk powder in PBS for 1 h at room
temperature and probed by incubation for 1 h with mouse sera
diluted at 1:1000 in blocking solution or mAb supernatants at optimal
dilutions (varied for each Ab). After washing with PBS/Tween, the
strips were incubated for an additional hour with goat anti-mouse
polyvalent Ig Abs (1:2000) (Caltag Laboratories, South San Francisco,
CA) or (1:5000) (Zymed, South San Francisco, CA). The strips were
washed and visualized with 0.05% diaminobenzidine containing 0.05 M
hydrogen peroxide in PBS or enhanced chemiluminescent substrate for
detection of horseradish peroxidase (Pierce, Rockford, IL). Sera
from patients with PBC with known reactivities against BHM were used as
positive controls. Similarly, reactivities of mouse sera or mAbs were
tested by ELISA against recombinant BCOADC-E2, PDC-E2, OGDC-E2, and
BCOADC-E3 (20, 21). Purified GST and an irrelevant fusion protein WK1.1
(MetE1-GST fusion) were also used as controls to determine mAbs
reactive to GST alone and/or an irrelevant GST fusion protein. A
total of 44 mAb were finally obtained; 9 were found to react to
GST-MetE1, our irrelevant recombinant control, and 35 were found to be
specific for the mitochondrial autoantigens. These latter 35 mAbs are
the focus of this study. In all cases, the class and subclass
specificities of mAbs were determined by ELISA using a mouse Ig
isotyping kit (PharMingen, San Diego, CA) and rat anti-mouse µ-,
-, 
-, IgG1-, IgG2a-, IgG2b-, IgG3-, 
-, and 
-chain
Abs.
Immunohistochemistry
Two different immunohistochemical studies were performed in an effort to examine the pattern of staining of the reactive mAbs. First, a typical anti-mitochondrial assay was performed using HEp-2 cell slides (Antibodies Inc., Davis, CA). Sections were incubated at room temperature for 1 h with each individual mAb. After incubation, the slides were washed in PBS for 10 min. The reaction was followed by a 30-min incubation with FITC-conjugated goat anti-mouse polyvalent Abs (Caltag Laboratories) diluted at 1:30 in PBS. All sections were viewed by a fluorescent microscope.
Secondly, liver was obtained from patients with PBC and, as a control, from patients with primary sclerosing cholangitis (PSC). After collection, the liver tissues were fixed in 10% neutral-buffered formalin and embedded in paraffin; 6-µm sections were cut and mounted onto poly-L-lysine-coated slides to prevent tissue detachment. The sections were deparaffinized and rehydrated through xylenes and graded concentrations of ethanol. A total of six specimens were studied; four were from patients with PBC, and two were from patients with PSC. The data on individual patients were similar to those of other, comparable patients, and previously reported data; accordingly, the results were combined. Immunohistochemical staining was conducted using a standard avidin-biotin detection method (22). C355.1, an mAb to PDC-E2, was used as a positive control of the apical staining of BEC from PBC livers (23). Briefly, after washing with Tris-buffered saline (TBS), the sections were incubated for 30 min with diluted blocking serum (Vector Laboratories, Burlingame, CA) to block nonspecific background. Sections were then incubated overnight at 4°C, or 1 h at room temperature, with a predetermined optimal dilution of individual mAbs. After incubation, the slides were washed in TBS, followed by a 30-min incubation with biotinylated goat anti-mouse IgG (H+L chain) or IgM (µ chain specific) (Vector Laboratories) diluted at 1:500, or according to kit instructions, in TBS. Following appropriate incubation and washing, an alkaline phosphatase-conjugated streptavidin-biotin complex avidin-biotin complex-alkaline phosphatase reagent (Vector Laboratories) was applied to all sections for 30 min. After washing with TBS, the substrate Vector Red (Vector Laboratories) was applied to the sections and incubated for an additional 5–10 min. Levamisole (1 mM) was added to the substrate to block endogenous alkaline phosphatase activity. All sections were viewed by a light microscope and a Bio-Rad (Richmond, CA) MRC 600 laser confocal microscope.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
A total of 35 mAbs specific for the mitochondrial autoantigens
were obtained and their specificities defined by both ELISA and
immunoblotting using individual recombinant proteins. By ELISA, 12 of
35 reacted only with PDC-E2, 15 reacted only with OGDC-E2, 5 reacted
with only BCOADC-E2, 1 reacted with both PDC-E2 and OGDC-E2, and 2
reacted with all three Ags (Table I). By
immunoblotting, 9 of 35 reacted with only PDC-E2, 14 reacted
with only OGDC-E2, 5 reacted with only BCOADC-E2, 4 reacted to PDC and
OGDC, 1 reacted to all three, and 2 were nonreactive (Figs. 2 and 3).
|
|
Positive AMA staining using HEp-2 cells was obtained with 32 of 35
mAbs (Table II). Two of the mAbs (2H-3H11
and 2H-6H8) that were reactive with BCOADC-E2 by ELISA gave a
homogeneous nuclear staining pattern (Table II). The most interesting
data, however, were observed when liver sections of PBC and control
patients were examined using alkaline phosphatase immunohistochemical
staining. First, we note that C355.1, our known positive mAb control,
stained the apical region of BEC in liver sections from PBC in an
intense linear pattern, in contrast with liver sections from PSC
patients. Of the 35 mAb studied, 23 reacted to tissue sections. Two
distinct staining patterns were observed (Table III). The first pattern was the typical
mitochondrial staining identical to that observed on HEp-2 cells; this
was noted in Table II. Immunohistochemical staining of mitochondria in
liver sections was less sensitive than immunofluorescence of HEp-2
cells, with only 23 of 35 mAbs being AMA positive compared with 32 of
35 by HEp-2 cells (Table II). This may have been due to an alteration
in antigenicity caused by fixation in the paraffin-embedded sections.
The second pattern was an intense linear staining of the apical region
of BEC in liver tissue from patients with PBC. Of the 12 mAbs that
reacted to PDC-E2, 1 (2H-4C8) showed intense reactivity at the apical
surface of BEC (Fig. 4). Of the 15 mAbs
that reacted to OGDC-E2, 3 (2H-2D3, 2H-5G2, and 2H-6G10) showed intense
linear staining of the apical surface of BEC. Of the 5 mAbs that
reacted to BCOADC-E2, 2 (2H-5A12 and 1H-2G12) showed a similarly linear
pattern of reactivity at the apical surface of BEC. Of the remaining
mAbs (3 of 35), 1 (3H-2H4), which recognizes PDC-E2, BCOADC-E2, and
OGDC-E2, produced a striking mixture of both intense linear and
punctate staining at the apical surface of BEC from PBC patients. As
additional controls, the mAbs were also studied using liver sections
taken from patients with PSC (Table III). None of the mAbs that gave a
linear apical staining pattern on BEC of PBC tissue produced such a
pattern or intensity in PSC tissue.
|
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In a previous study, we constructed a trihybrid clone consisting of the lipoyl domains of PDC-E2, BCOADC-E2, and OGDC-E2 to be utilized as a potential diagnostic reagent and reported that nearly all PBC patient sera reacted with this trihybrid Ag (18). In the current study, mice immunized with the trihybrid molecule produced mAbs that were categorized based on immunoblot and ELISA analysis. The characterization of Ag specificities revealed that almost all (32 of 35) reacted with only one mitochondrial Ag when tested by ELISA.
To the best of our knowledge, only one anti-OGDC-E2 mAb has been previously reported (29), and to date there have been no anti-BCOADC-E2 mAbs described in the literature. Included in our present study are two anti-BCOADC-E2 mAbs, 1H-2G12 and 3H-5H1, which react with the E3 component of the BCOADC complex dihydrolipoamide dehydrogenase (data not shown). E3 is a component shared among the 2-oxoacid dehydrogenase complexes and is reported to be present in much smaller amounts than the corresponding E2s (30). The reactivity against E3 was removed by preabsorption of the Abs with BCOADC-E2, suggesting that some level of cross-reactivity exists between the E2 and E3 components of the BCOADC complex despite the low homology (38%) of the amino acid sequences between equivalent lipoyl domains. Epitope mapping of 1H-2G12 and 3H-5H1 showed that the reactivity was localized to subfragments of amino acids 1–115 and 84–227 of BCOADC-E2, implying that 1H-2G12 and 3H-5H1 recognize either the shorter segment, amino acids 84–115, of BCOADC-E2 or an as yet undetermined conformational domain of the molecule. Although autoreactivity against BCOADC-E3 in sera from patients with PBC have been shown to be of low prevalence and titer (31), further study may be necessary to determine the role and pathologic significance of the E3 component in PBC.
The staining of BECs of PBC livers with not only anti-PDC-E2 but
also anti-BCOADC-E2 and anti-OGDC-E2 mAbs is of great interest.
Previous reports on the staining of bile duct cells with
anti-PDC-E2 mAbs showed that one mAb, C355.1, produced intense
linear staining in the apical region of bile duct cells exclusively in
patients with PBC, suggesting the presence of an increased amount of
PDC-E2 or a cross-reactive molecule within this area (23, 32, 33). In a
recent study it was proposed that protein X, or the E3 binding protein,
is the cross-reactive protein, but control data are not presented and
the data are based on analogy only (34). In this study, seven of the
mAbs produced intense linear staining in the apical region of bile duct
cells exclusively in patients with PBC. Three of those mAbs (2H-5G2,
2H-2D3, and 2H-6G10) react with OGDC, one (2H-4C8) reacts with PDC, two
(2H-5A12, 1H-2G12) react with BCOADC, and one (3H-2H4) reacts with all
three mitochondrial proteins as determined by ELISA and immunoblot. All
seven of these mAbs produced strong, specific reactivities to their
respective Ags as determined by ELISA, as well as strong AMA patterns
in HEp-2 cells (Table II
). It is important to note that 23 of the 35
mAbs against PDC-E2, BCOADC-E2, and OGDC-E2 do give mitochondrial
staining in the bile ducts of both PBC and PSC patients. However, there
appear to be either additional forms of these enzymes or a
cross-reactive epitope(s) present in the apical region of patients with
PBC that is not present in the mitochondria. Thus, only 7 of 35 mAbs
produced against the mitochondrial enzymes recognize an epitope(s) that
is either conformationally distinct from, or cross-reacts with, their
native enzyme. While only a portion of the native lipoyl domains were
used to produce these Abs, all 35 reacted to some form of the native
protein, whether by ELISA or immunoblot. In addition, only one mAb to
BCOADC-E2 did not react to the enzymes in the mitochondria of HEp-2
cells and was not used for further immunohistochemical analysis. With
regard to PDC-E2, there is no evidence for increased production of this
enzyme at the transcriptional level detected in bile duct cells by in
situ mRNA hybridization using an antisense probe against PDC-E2 (15).
This does not rule out an increase in PDC-E2 due to translational or
catabolic defects.
Epitope mapping studies using shorter fragments of recombinant Ags indicated that shared epitopes recognized by mAbs are conformational rather than linear, as previously reported for AMA and murine anti-PDC-E2 Abs (11, 12, 13, 35). Although the levels of linear sequence homology between the lipoic acid binding domains of PDC-E2, BCOADC-E2, and OGDC-E2 is quite low, there may be similarity at the structural level that accounts for the similarity in the intense apical staining patterns seen in PBC-BEC when using mAbs to three different mitochondrial enzymes. These results are consistent with the hypothesis that an immune response to one mitochondrial Ag may induce a cross-reactive response to the other Ags containing similar structural components, i.e., the lipoic acid binding region. It should be stressed that at present this suggestion is based on the immunization of experimental animals and that this cross-reactivity may not be evident in humans.
In a previous study, we demonstrated the presence of mitochondrial Ags and AMA in bile from patients with PBC (36) and found a positive correlation between AMA in PBC sera and corresponding Abs and Ags in bile, suggesting the possibility of immune complex formation between the two. Such mitochondrial Ags, especially coupled with IgA AMA, may be trapped or accumulated in bile duct cells during their normal transport to the bile duct lumen via the polyimmunoglobulin receptor found only on bile duct cells. Therefore, the specific binding of the seven mAbs to the apical region, in addition to mitochondria of BEC, may reflect the presence of these enzymes complexed to IgA-AMA, which would not be present in PSC BEC. To confirm this hypothesis, the study of IgA AMA transcytosis and mitochondrial Ags will be necessary. The discovery of additional antigenic determinants located in a specific region of the target tissue of PBC that cross-reacts with the autoantigens raises several new avenues of investigation into the etiology of this disease.
|
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. M. Eric Gershwin, Division of Rheumatology, Allergy and Clinical Immunology, University of California, One Shields Avenue, TB 192, School of Medicine, Davis, CA 95616-8660. E-mail address:
3 Abbreviations used in this paper: PBC, primary biliary cirrhosis; AMA, antimitochondrial Ab; PDC-E2, pyruvate dehydrogenase complex E2; BCOADC-E2, branched chain 2-oxoacid dehydrogenase complex E2; OGDC-E2, 2-oxoglutarate dehydrogenase complex E2; BEC, biliary epithelial cell; GST, glutathione S-transferase; BHM, beef heart mitochondrial fraction; PSC, primary sclerosing cholangitis; TBS, Tris-buffered saline.
Received for publication April 23, 1998. Accepted for publication July 6, 1998.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-ketoacid dehydrogenase complexes. Proc. Natl. Acad. Sci. USA 85:1422.This article has been cited by other articles:
|
|
 |
|
  J. Mazumdar, E. H. Wilson, K. Masek, C. A. Hunter, and B. Striepen Apicoplast fatty acid synthesis is essential for organelle biogenesis and parasite survival in Toxoplasma gondii PNAS, August 29, 2006; 103(35): 13192 - 13197. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Irie, Y. Wu, L. S. Wicker, D. Rainbow, M. A. Nalesnik, R. Hirsch, L. B. Peterson, P. S.C. Leung, C. Cheng, I. R. Mackay, et al. NOD.c3c4 congenic mice develop autoimmune biliary disease that serologically and pathogenetically models human primary biliary cirrhosis J. Exp. Med., May 15, 2006; 203(5): 1209 - 1219. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Amano, P. S. C. Leung, R. Rieger, C. Quan, X. Wang, J. Marik, Y. F. Suen, M. J. Kurth, M. H. Nantz, A. A. Ansari, et al. Chemical Xenobiotics and Mitochondrial Autoantigens in Primary Biliary Cirrhosis: Identification of Antibodies against a Common Environmental, Cosmetic, and Food Additive, 2-Octynoic Acid J. Immunol., May 1, 2005; 174(9): 5874 - 5883. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Habelhah, A. Laine, H. Erdjument-Bromage, P. Tempst, M. E. Gershwin, D. D. L. Bowtell, and Z. Ronai Regulation of 2-Oxoglutarate ({alpha}-Ketoglutarate) Dehydrogenase Stability by the RING Finger Ubiquitin Ligase Siah J. Biol. Chem., December 17, 2004; 279(51): 53782 - 53788. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Amano, P. S. C. Leung, Q. Xu, J. Marik, C. Quan, M. J. Kurth, M. H. Nantz, A. A. Ansari, K. S. Lam, M. Zeniya, et al. Xenobiotic-Induced Loss of Tolerance in Rabbits to the Mitochondrial Autoantigen of Primary Biliary Cirrhosis Is Reversible J. Immunol., May 15, 2004; 172(10): 6444 - 6452. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. A. Long, C. Quan, J. Van de Water, M. H. Nantz, M. J. Kurth, D. Barsky, M. E. Colvin, K. S. Lam, R. L. Coppel, A. Ansari, et al. Immunoreactivity of Organic Mimeotopes of the E2 Component of Pyruvate Dehydrogenase: Connecting Xenobiotics with Primary Biliary Cirrhosis J. Immunol., September 1, 2001; 167(5): 2956 - 2963. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
