The JI
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     
 

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kerkar, N.
Right arrow Articles by Vergani, D.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kerkar, N.
Right arrow Articles by Vergani, D.
The Journal of Immunology, 2003, 170: 1481-1489.
Copyright © 2003 by The American Association of Immunologists

Cytochrome P4502D6193–212: A New Immunodominant Epitope and Target of Virus/Self Cross-Reactivity in Liver Kidney Microsomal Autoantibody Type 1-Positive Liver Disease 1

Nanda Kerkar*,{dagger}, Kaushik Choudhuri*, Yun Ma*, Ayman Mahmoud*, Dimitrios P. Bogdanos*, Luigi Muratori{ddagger}, Francesco Bianchi{ddagger}, Roger Williams*, Giorgina Mieli-Vergani{dagger} and Diego Vergani2,*,{dagger}

* Institute of Hepatology, University College, London, United Kingdom; {dagger} Institute of Liver Studies, King’s College Hospital, Denmark Hill, London, United Kingdom; and {ddagger} Department of Internal Medicine, Cardioangiology, Hepatology, University of Bologna, Bologna, Italy


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Cytochrome P4502D6 (CYP2D6), target of liver kidney microsomal autoantibody type 1 (LKM1), characterizes autoimmune hepatitis type 2 (AIH2) but is also found in patients with chronic hepatitis C virus (HCV) infection. To provide a complete linear epitope B cell map of CYP2D6, we tested peptides spanning the entire sequence of CYP2D6. In addition to confirming previously described antigenic sites, we identified four new epitopes (193–212, 238–257, 268–287, and 478–497). CYP2D6193–212 is immunodominant and was the target of 12 of 13 (93%) patients with AIH2 and 5 of 10 (50%) HCV/LKM1-positive patients. Because LKM1 is present in both AIH2 and a viral infection, we tested whether Abs to CYP2D6193–212 arise through cross-reactive immunity between virus and self. We identified a hexameric sequence "RLLDLA" sharing 5 of 6 aa with "RLLDLS" of HCV2985–2990 and all 6 aa with CMV130–135. Of 17 CYP2D6193–212-reactive sera, 11 (7 AIH and 4 HCV) reacted by ELISA with the HCV homologue, 8 (5 AIH and 3 HCV) with the CMV homologue, and 8 (5 AIH and 3 HCV) showed double reactivity. Autoantibody binding to CYP2D6193–212 was inhibited by preincubation with HCV2977–2996 or CMV121–140. Recombinant HCV-nonstructural protein 5 and CMV-UL98 proteins also inhibited Ab binding to CYP2D6193–212. Affinity-purified CYP2D6193–212-specific Ab inhibited the metabolic activity of CYP2D6. The demonstrated similarity and cross-reactivity between CYP2D6193–212 and two unrelated viruses suggests that multiple exposure to viruses mimicking self may represent an important pathway to the development of autoimmunity.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
Liver kidney microsomal Ab type 1 (LKM1)3 is the diagnostic marker of autoimmune hepatitis type 2 (AIH2), AIH1 being characterized by the presence of antinuclear Ab and/or smooth muscle Ab (1). LKM1 is also present in up to 10% of patients with chronic hepatitis C virus (HCV) infection, where it appears to confer severity to the disease process, as shown by Giostra et al. (2), and to predispose to adverse reactions during treatment with IFN (3). LKM1, first described by Rizzetto et al. in 1973 (4), is known to recognize a 50-kDa protein in rat microsomes located primarily in the smooth endoplasmic reticulum (5) and a 48-kDa protein in human liver microsomes (6). This protein in human liver was later identified as cytochrome P4502D6 (CYP2D6) (7, 8, 9). The catalytic function of this cytochrome is inhibited in vitro by incubation with LKM1-positive serum (7).

The presence of B cell epitopes on CYP2D6 has been investigated by several groups (Fig. 1A), and the region spanning 254–271 has been reported to be highly antigenic in LKM1-positive AIH (10, 11, 12). Within this region, Manns et al. (11) reported that CYP2D6263–270 is recognized by 62% of AIH2 patients, and Yamamoto et al. (12) reported that CYP2D6257–269 is the most frequently recognized antigenic site in LKM1 AIH (85% of sera tested). They also described three other epitopes spanning 321–351, 373–389, and 410–429 regions of CYP2D6, recognized respectively by 53, 7, and 13% of the AIH2 cases (12). In contrast to patients with AIH2, reactivity to the CYP2D6254–271 region was not detected in chronic HCV patients seropositive for LKM1 (13). Two epitopes (200–214 and 321–339) recognized by LKM1 in chronic HCV infection were described by Parez et al. (14). More recently, Klein et al. (15) reported that the CYP2D6 sequence 196–218 is recognized by 68% of patients with AIH2, but by only 18% of LKM1/HCV-positive patients. Common to all of the above reports is the selection of putative antigenic sites for testing, an approach that could ignore other important epitopes. The aim of the present study was to provide a complete B cell map of CYP2D6 linear epitopes using a representative number of LKM1-positive sera from patients with AIH2 or chronic HCV infection. Epitopes recognized by LKM1-positive serum would then be analyzed for similarities with HCV and other pathogens, possible triggers of LKM1 production.



View larger version (22K):
[in this window]
[in a new window]
  		FIGURE 1. A, Epitope map of CYP2D6 illustrating documented epitopes: CYP2D6254–271 by Gueguen et al. (8 ); CYP2D6263–270 by Manns et al. (9 ); CYP2D6257–269, CYP2D6321–351, CYP2D6373–389, and CYP2D6410–429 by Yamamoto et al. (12 ); CYP2D6200–214 and CYP2D6321–339 by Parez et al. (14 ); CYP2D6196–218 by Klein et al. (15 ); CYP2D6193–212, CYP2D6238–257, CYP2D6268–287, and CYP2D6478–497 as described in this report. B, Detailed epitope map of CYP2D6193–218. Bi, CYP2D6200–214 described by Parez et al. (14 ) and recognized by 62.5% of LKM1-positive HCV patients, but not by AIH2 patients. Bii, CYP2D6196–218 described by Klein et al. (15 ) and recognized by 68% of AIH2 and 18% of LKM1/HCV patients. Biii, CYP2D6193–212, the new epitope described in this report recognized by 92% of AIH2 and 50% of LKM1/HCV patients.

 

   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Subjects

Twenty-three LKM1-positive patients were studied. Thirteen had classical LKM1-positive AIH (AIH2) diagnosed according to the recently revised criteria of the International Autoimmune Hepatitis Group (16). All had raised transaminases at presentation and were negative for HCV by PCR as well as other serological markers for hepatitis A-E and IgM to EBV and CMV. All had characteristic interface hepatitis on liver biopsy and responded to immunosuppressive treatment (Prednisolone with and without Azathioprine). Eleven patients were female with a median age of 10.6 years and a range from 3.1 to 18 years. LKM1 was detected by standard immunofluorescence, and the LKM1 titers at the time of testing ranged from 1/40 to 1/5120 with a median of 1/640. Four children had other autoimmune conditions: autoimmune thyroiditis in one, insulin-dependent diabetes mellitus in the second, sclerosing cholangitis in the third, and inflammatory bowel disease in the fourth. Three children were tested at diagnosis, four during remission on immunosuppressive therapy, and six during relapse.

Ten sera were from patients with liver disease attributed to chronic HCV infection. Six were female with a median age of 37 years and a range of 19 to 55 years. All were HCV RNA-positive (Amplicor; Hoffmann-LaRoche, Basel, Switzerland). LKM1 titers at the time of testing were 1/160–1/640 with a median of 1/640. One was receiving IFN-{alpha}, two had stopped IFN-{alpha} 5 mo and 4 years before this study, and one was on steroid treatment after having developed a transaminase flare 4 mo into IFN-{alpha} treatment. These patients have been previously described within a larger series of LKM1/HCV-positive patients treated with IFN-{alpha} (17). Six patients were not treated. Serum from seven healthy controls, four male and three female with a median age of 13 years (range 10–30 years), were also studied.

Peptides

Thirty-four 20-mer synthetic peptides spanning the entire 517-aa sequence of the protein and overlapping by 5 aa were constructed by 9-fluorenylmethyloxycarbonyl chemistry (Mimotopes, Clayton, Victoria, Australia) (Table I). Biotin was linked to the peptides through a spacer with alternating serine (S) and glycine (G) residues to avoid steric hindrance and reduce hydrophobicity (18). Peptides conformed to the following format: biotin-SGSG-peptide-amide, where SGSG is the spacer. A peptide encoding a randomly generated sequence of amino acids, biotin-SGSG-HEDYVNQSLRPTPLEISVRA-amide, was used as the negative control peptide. All peptides were >90% pure after reverse-phase HPLC. The identity of each peptide was confirmed by mass spectrometry. The reactivity of LKM1-positive serum against each of the 34 peptides constructed was tested by ELISA.


View this table:
[in this window]
[in a new window]
  		Table I. Peptide keya

 
Recombinant proteins

CMV alkaline exonuclease, the UL98 gene product, was a kind gift from Dr. D. J. Tenney from Bristol-Myers Squibb (Pharmaceutical Research Institute, Wallingford, CT). Recombinant human CMV-UL98 was produced using a baculovirus-insect cell expression system and was purified by high salt cell lysis, Q Sepharose chromatography, and phosphocellulose chromatography. The final material was dialysed against 20 mM Tris-HCl (pH 7.5), 25% glycerol, 0.1 mM EDTA, and 1 mM DTT to an NaCl concentration of 5 mM. The protein resolved as a single ~65-kDa band on a Coomassie-stained SDS-PAGE. It had a specific activity of 28 mg of activated calf thymus DNA digest/mg protein/30 min with a protein concentration of 83 µg/ml. HCV-nonstructural protein 5B (NS5B) (genotype 3a) was a kind gift from Dr. Seong of Yonsei Engineering Complex (Yonsei University, Seoul, Korea). Recombinant protein migrated as a single band at ~60 kDa on SDS-PAGE and was enzymatically active as assessed by a gel-based biochemical RdRp assay. The protein was dialyzed into 25 mM sodium phosphate (pH 6.8), 150 mM NaCl, 0.5 mM EDTA, 5 mM DTT, 0.5% Tween 20, and 55% glycerol at a final concentration of 60 µg/ml. Recombinant metabolically active CYP2D6, produced in a baculovirus-insect cell expression system and extracted as a microsomal preparation (10 mg/ml), was purchased from Cambridge BioScience (Cambridge, U.K.).

ELISA

A total of 200 ml/well of 2% BSA/PBS was added for 1 h at 20°C to a 96-well polysterene plate precoated with 5 mg/ml streptavidin (Mimotopes) to prevent nonspecific binding. After addition to each well of 100 ml of biotinylated peptide diluted to 1/1000 in PBS containing 0.1% sodium azide and 0.1% BSA, the plate was mildly agitated in a shaker (Dynex Technologies, Guernsey, U.K.) for 1 h at 20°C. One hundred microliters of patient serum diluted to 1/200 in 2% BSA/PBS containing 0.1% sodium azide was added to each well and incubated under mild agitation at 20°C for another hour. Thereafter, 100 ml of HRP-conjugated goat anti-human IgG (Sigma-Aldrich, Poole, Dorset, U.K.) diluted to 1/1000 in 2% BSA/PBS was added and incubated for 1 h without shaking at 20°C. The optimum concentrations of reagents at various steps of the immunoassay were determined in preliminary experiments by checkerboard titration. Initially, using a wide range of dilutions, the working conditions of the assay were identified and, subsequently, using smaller dilutions around the optimum, the final concentration of the test samples and reagents was determined (19). After each of the above steps, the plate was washed five times with PBS containing 0.1% Tween 20. Freshly prepared o-phenylene-diamine in citrate phosphate buffer was used as substrate, and the reaction was stopped with 4N sulfuric acid. Absorbance values were read at 490 nm on a spectrophotometer. In each plate, two wells were used as blanks, in which serum and peptide were omitted, and two additional wells were used for a positive and a negative control. The positive control consisted of an LKM1-positive serum (titer 1/5120), which in preliminary experiments was shown to react with CYP2D6253–272, containing a previously described immunodominant epitope (10). A randomly generated scrambled control peptide incubated with serum from a healthy individual was used as negative control. Each serum tested against experimental peptides was also tested against the control peptide. The final absorbance value was calculated by subtracting control peptide from the experimental peptide absorbance. Subtracted values exceeding 0.11 were considered positive, this value representing mean + 3 SD of 126 readings using serum from seven healthy subjects against 14 randomly selected CYP2D6 peptides. A synthetic peptide was considered to encompass an epitope when the subtracted absorbance was >0.5.

CMV IgG was tested by ELISA using CMV IgG kit (Captia; Trinity Biotech, Bray, Ireland).

Serum from the 10 LKM1-positive patients with chronic HCV infection was tested for rheumatoid factor IgM Abs using a commercially available kit (Rheumatoid Factor kit; Sigma-Aldrich).

Database search

Protein databases SWISS-PROT and PIR were searched using PROTEININFO to search for homologies between the newly identified epitopes and human microorganisms. Homologous peptides were constructed by 9-fluorenylmethyloxycarbonyl chemistry as detailed above (Mimotopes) (Table I).

Inhibition studies

To investigate whether the reactivity to CYP2D6193–212 and homologous viral peptides was due to cross-reactivity, competition ELISAs were performed using CYP2D6193–212 in solid phase and using CYP2D6193–212, HCV2977–2996, CMV121–140, and the irrelevant peptide as competitors in liquid phase. Test serum was diluted 1/400 with peptide solutions to give final peptide concentrations of 0.1 mg/ml, 0.5 mg/ml, and 1 mg/ml, respectively. The solutions were incubated at 37°C for 2 h and then tested for anti-CYP2D6193–212 reactivity by ELISA.

Alanine substitution studies

To further investigate the contribution of the HCV2977–2996 region 2985RLLDL (which is identical with CYP2D6204–208 and CMV130–134) to Ab reactivity to the full 20-aa sequence, five peptides were constructed in the format biotin-SGSG-peptide-amide, each with five contiguous alanine substitutions across the length of HCV2977–2996 (Table I). Alanine-substituted peptides were tested for reactivity to patient serum by ELISA. LKM1+ patient serum was tested at a dilution of 1/400, and HRP-conjugated goat anti-human IgG Ab was used at a dilution of 1/2000. All other experimental conditions were as described in ELISA.

Reactivity of LKM1-positive serum to recombinant proteins

ELISA microtiter plates (Nunc International, Roskilde, Denmark) were coated overnight at 4°C with recombinant CYP2D6, HCV-NS5, and CMV-UL98 proteins at a concentration of 2 µg/ml in sodium bicarbonate buffer (pH 9.6). After washing, plates were blocked with 100 µl of 5% BSA in PBS for 1 h at 37°C. One hundred microliters of serum from a patient with LKM1-positive AIH at a dilution of 1/200 was added after washing. Further steps were as described in ELISA.

Inhibition of reactivity to CYP2D6193–212 by recombinant proteins

Inhibition of reactivity to the immunodominant epitope by whole protein was investigated by performing a competition ELISA using CYP2D6193–212 in solid phase and using recombinant HCV-NS5 and CMV-UL98 proteins as competitor in liquid phase. Test serum (at a final dilution of 1/200) and recombinant proteins (at a final concentration of 0.3 µg/ml) were incubated for 2 h at 37°C and subsequently were tested for reactivity to CYP2D6193–212 as described in ELISA.

Affinity purification of LKM1-positive serum

An LKM1-positive serum sample (reactive to CYP2D6193–212, HCV2977–2996, and CMV130–134) was affinity purified over a Tetralink tetrameric avidin column (Promega, Southampton, U.K.) complexed with biotinylated CYP2D6193–212 per the manufacturer’s instructions. Briefly, 1 ml of biotinylated CYP2D6193–212 at a concentration of 0.25 mg/ml was passed over a tetrameric avidin column (equilibrated in PBS). CYP2D6193–212-specific Ab was captured by passing 5 ml of a "triple-reactive" LKM1-positive serum (diluted 1/4 in PBS) over the prepared column a total of five times. The column was washed with PBS to remove unbound material and captured Ab was eluted into 2 M Tris-HCl (pH 8.0) using 50 mM glycine (pH 3.0). The eluted Ab was concentrated to give a final volume of 1 ml (to match the original serum volume).

Inhibition of CYP2D6 metabolic activity by LKM1-positive serum

A commercially available kit was used to assess inhibition of CYP2D6 metabolic activity by LKM1-positive serum and affinity-purified Ab (CYP2D6/3-[2-(N,N-diethyl-N-methylamino)ethyl]-7-methoxy-4-methylcoumarin (AMMC) inhibitor screening kit; Gentest). The assay uses recombinant microsomal CYP2D6 (1.5 pM) in conjunction with AMMC as substrate (1.5 mM) and cofactor, producing a fluorescent metabolite (3-[2-(N,N-diethylamino)ethyl]-7-hydroxy-4-methylcoumarin hydrochloride) that is detected using a fluorescence plate reader (CytoFluor Series 4000; Applied Biosystems, Foster City, CA; excitation at 390 nm, emission at 460 nm). Serial dilutions of 1/25 to 1/18,225 of a known CYP2D6 inhibitor (quinidine, initial concentration 0.5 µm) and LKM1-positive serum from patients with AIH and HCV infection, as well as the affinity-purified Ab, were added to 96-well black microtiter plates (Labsystems Fluoro 96; Thermo Labsystems, Vantaa, Finland) containing CYP2D6 followed by the addition of substrate. Control reactions included CYP2D6 with no substrate, CYP2D6 with substrate (no inhibitor), and blank wells, which were otherwise subjected to the same experimental conditions. The reaction was allowed to progress for 30 min at 37°C, after which it was terminated with 80% acetonitrile, 100 mM Tris. Plates were read using the parameters described above. Percentage inhibition was calculated as follows: [(1 - (mean fluorescence with test compound - mean fluorescence of blank well/mean fluorescence without inhibitor - mean fluorescence of blank well)] x 100.

Structural homology modeling of CYP2D6

Swiss-Model and the Swiss-Pdb Viewer (http://www.expasy.org/swissmod) were used for investigating and analyzing the derived CYP2D6 protein structure (20). CYP2D6193–212 was modeled locally on the homologous region in Bacillus megaterium cytochrome P450 BM-3 (21, 22).


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Seven sera from healthy individuals were tested against 14 randomly selected peptides to establish a cutoff for Ab reactivity. The mean (SD) subtracted (from the random control peptide) OD was 0.0137(0.0366), giving a cutoff value of 0.11 (mean + 3 SD).

Sequences incorporating previously described epitopes (12) were recognized at the following frequencies: CYP2D6253–272 by 11 of 13 (85%) AIH2 and 1 of 10 (10%) LKM1/HCV patients, CYP2D6418–437 by 3 of 13 (23%) AIH2 and 1 of 10 (10%) LKM1/HCV patients, and CYP2D6343–362 by 1 of 10 (10%) LKM1/HCV patients. Individual serum from AIH2 patients recognized the sequences CYP2D6328–347 and CYP2D6373–392 (Fig. 2, A and B, open arrows). Four new epitopes were identified (Fig. 2, A and B, filled arrows). These comprised CYP2D6193–212, which was recognized by the serum from 12 of 13 (93%) AIH2 and 5 of 10 (50%) LKM1-positive patients with chronic HCV infection, CYP2D6268–287 by two AIH2 sera, and CYP2D6238–257 and CYP2D6478–497 by individual sera from LKM1/HCV patients. Within the immunodominant epitope CYP2D6193–212, the motif "RLLDLA" was found to have 5 of 6 aa identical and one similar to the hexameric sequence "RLLDLS" of HCV2985–2990 (located within HCV-NS5, the RNA-dependent DNA polymerase), A and S being conserved residues (23). The same motif also shared a 6-aa identity with CMV130–135 (located within the CMV alkaline exonuclease) (Fig. 3). All 17 sera (12 AIH2 and 5 LKM1/HCV) reacting with the newly identified epitope CYP2D6193–212 were tested by ELISA against the homologous HCV and CMV biotinylated viral peptides. Seven AIH2 and four LKM1/HCV sera reacted with both CYP2D6193–212 and HCV2977–2996, whereas five AIH2 and three LKM1/HCV sera reacted with both CYP2D6193–212 and CMV121–140. Triple reactivity to CYP2D6193–212/HCV2977–2996/CMV121–140 was observed in five AIH2 and three LKM1/HCV patients (Table II). Ab binding to CYP2D6193–212 was inhibited by addition of increasing amounts of competitor peptide. In the presence of 1 mg/ml of competitor, absorbance was reduced by 70–90%, whereas it remained virtually unaltered by addition of the scrambled control peptide (Fig. 4).



View larger version (26K):
[in this window]
[in a new window]
  		FIGURE 2. Reactivities of sera from 13 patients with AIH2 (A) and 10 LKM1-positive patients with chronic HCV infection (B) against 34 overlapping 20-mer peptides spanning residues 1–517 on CYP2D6. Absorbance values on the y-axis are plotted against the 34 individual peptides on the x-axis. The filled arrows mark the four newly identified epitopes: CYP2D6 193–212 (peptide 14) is recognized by 12 AIH2 (patients 1–9, 11, 12, and 13) and 5 LKM1/HCV cases (patients 14, 15, 17, 18, and 20); CYP2D6238–257 (peptide 17), recognized by LKM1/HCV patient number 22; (Figure legend continues)

 


View larger version (15K):
[in this window]
[in a new window]
  		FIGURE 3. Amino acid sequence homology between the HCV2977–2996, CMV121–140, and CYP2D6193–212. Amino acids are in standard single letter code. Identical residues between CYP2D6193–212 and the homologous viral peptides are within the box.

 

View this table:
[in this window]
[in a new window]
  		Table II. Recognition of homologous viral peptides CMV121–140 and HCV2977–2996 by LKM1-positive sera all reactive with the immunodominant epitopea

 


View larger version (15K):
[in this window]
[in a new window]
  		FIGURE 4. Inhibition of Ab binding to CYP2D6193–212 after incubation with increasing amounts of CYP2D6193–212 (•), HCV2977–2996 ({blacktriangleup}), CMV121–140 ({blacktriangledown}), and control peptide ({blacksquare}). Percentage binding is on the y-axis. Decrease in binding of >75% is seen after preincubation with 1 mg/ml of self and homologous viral peptides, but not with the control peptide.

 
Alanine substitution of the HCV2977–2996 region 2985RLLDL completely abolished reactivity by LKM1 Abs, as did substitution of the contiguous downstream 5-aa segment. All other alanine-substituted peptides produced comparable serum reactivities to wild-type HCV2977–2996 and to the homologous CYP2D6 immunodominant peptide CYP2D6193–212 (Fig. 5).



View larger version (16K):
[in this window]
[in a new window]
  		FIGURE 5. Reactivity of LKM1-positive serum to peptide HCV2977–2996and its alanine-substituted homologues. A, HCV2977–2996. B–F, Substitution of HCV2977–2996 with five contiguous alanine residues across the full length of the native peptide. Amino acids are represented in standard single letter code. Alanine-substituted regions are indicated in bold. The region of identity between CYP2D6193–212 and viral peptides is boxed. G, CYP2D6193–212. The previously established cutoff for a positive reaction is indicated by a line in the x-axis. Reactivity to HCV2977–2996 is abolished by alanine substitution of the region of homology and the contiguous downstream five residues.

 
Of eight patients reactive with CMV121–140 whose serum was available for testing, four (one AIH2 and three LKM1/HCV) were positive for anti-CMV IgG. All 10 sera from LKM1/HCV patients were negative for rheumatoid factor IgM Abs.

Local homology modeling of CYP2D6 using the cytochrome P450 BM-3 crystal coordinates as a template predicts CYP2D6193–212 to be located within a helical domain in a solvent-accessible surface region of the protein, compatible with Ab recognition on whole protein. Therefore, we tested the ability of LKM1-positive sera to recognize whole protein to confirm the physiological significance of the observed peptide reactivities. Two LKM1-positive sera, one from a patient with AIH and another from a patient with HCV infection, were tested for reactivity to recombinant CYP2D6, HCV-NS5, and CMV-UL98. A cutoff for recognition was established as the mean absorbance at 490 nm + 3 SD of 43 serum samples from healthy individuals (cutoff OD490 = 0.07) (24). Both LKM1-positive sera strongly recognized the recombinant proteins (Fig. 6). We then investigated the ability of these recombinant proteins to inhibit reactivity to the immunodominant CYP2D6 peptide, CYP2D6193–212, by a liquid phase ELISA inhibition assay. Preincubation of LKM1-positive serum with HCV-NS5 and CMV-UL98 resulted in 60.5 and 61.2% inhibition of reactivity against CYP2D6193–212 (Fig. 7), confirming that LKM1 Abs that recognize CYP2D6193–212 are cross-reactive with the HCV and CMV peptide homologues.



View larger version (11K):
[in this window]
[in a new window]
  		FIGURE 6. Reactivity of LKM1-positive serum to recombinant CYP2D6 (rCYP2D6), HCV-NS5 protein (rHCV-NS5), and CMV-UL98 protein (rCMV-UL98) by ELISA. A cutoff for positive reaction was determined as OD490 = 0.07 using serum from healthy individuals and is represented as a horizontal line.

 


View larger version (7K):
[in this window]
[in a new window]
  		FIGURE 7. Inhibition of Ab reactivity of LKM1-positive serum to CYP2D6193–212 by preincubation with 0.3 µg/ml of rHCV-NS5 and rCMV-UL98 or in the absence of competitor protein. Residual Ab reactivity to CYP2D6193–212 was detected by ELISA. Error bars represent + SD of six separate experiments.

 
To further test recognition of conformationally intact CYP2D6 by LKM1 serum and the LKM1 Ab species reactive to CYP2D6193–212, we assessed the ability of LKM1-positive serum and affinity-purified LKM1 Abs specific for the CYP2D6193–212 epitope to inhibit CYP2D6 enzymatic activity using a fluorometric assay. Both LKM-positive serum from a patient with AIH and from another with HCV infection inhibited CYP2D6 enzymatic activity, producing ~40% inhibition at a dilution of 1/225. Affinity-purified Ab specific for the cross-reactive immunodominant CYP2D6 epitope, CYP2D6193–212, was also able to inhibit metabolic activity of CYP2D6, producing ~20% inhibition at a dilution of 1/225. This suggests that CYP2D6193–212-specific Abs contribute ~50% of the inhibitory activity of LKM1 serum, demonstrating the preservation and antigenicity of this epitope within native CYP2D6 (Fig. 8). In contrast, serum from a healthy individual failed to inhibit enzymatic activity.



View larger version (23K):
[in this window]
[in a new window]
  		FIGURE 8. Inhibition of the metabolic activity of CYP2D6 by LKM1-positive serum. AMMC was used as substrate and formation of the fluorescent metabolite 3-[2-(N,N-diethylamino)ethyl]-7-hydroxy-4-methylcoumarin hydrochloride detected using excitation and emission parameters of 390 nm and 460 nm, respectively. Quinidine was used as a standard positive control (starting concentration = 0.5 µM) and serum from a healthy individual used as a negative control (Normal). In addition to sera from patients with LKM1-positive AIH (LKM1AIH) and LKM1-positive HCV infection (LKM1HCV), affinity-purified Ab (against CYP2D6193–212) from an LKM1-positive patient was tested for inhibitory activity (apLKM1). Data are represented as percentage decrease in metabolite formation relative to no test sample/inhibitor. Fifty percent inhibition is indicated as a dashed horizontal line. Error bars represent ± SD of duplicate samples; where no error bars are shown both samples gave identical readings.

 

   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
In this study of complete B cell epitope mapping, we found a new immunodominant epitope on cytochrome P4502D6, CYP2D6193–212, recognized by LKM1-positive serum both from patients with AIH2 and chronic HCV infection, which is likely to be the target of virus/self cross-reactive immunity. The fact that reactivity to this epitope has gone unnoticed to date is probably due to the selective epitope construction used in the past to map antigenic sites on the cytochrome. Searching for LKM1 epitopes, Gueguen et al. (10) produced 10 CYP2D6 constructs through the expression of CYP2D6 cDNA or its rat orthologous fragments in a prokaryotic system. Through the analysis of the results of Ab binding to these constructs, they inferred that a region of 33 aa (239–273) was the major antigenic site in AIH, being recognized by all LKM1 AIH sera tested. Reactivity in this region was dissected further by the use of three synthetic peptides (241–260, 254–271, and 264–281), and the sequence 254–271 was found to be the focus of recognition. Although describing CYP2D6254–271 as a major epitope on CYP2D6, the authors acknowledged that their strategy might have overlooked other epitopes. Based on Gueguen et al.’s (10) and Yamamoto et al.’s additional data (12) with new CYP2D6 constructs and using information derived from algorithmic prediction of antigenic sites (25), Yamamoto et al. (12) constructed 12 peptides spanning the region 241–429. Epitope 257–269, 5 aa shorter than the epitope described by Gueguen et al. (10), was recognized by 14 of 15 sera from patients with AIH2, confirming that this region is a major antigenic site. They also found that the sequences 321–351, 373–389, and 410–429 were recognized by eight, one, and two of the 15 sera, respectively. Akin to the previous study, however, a selective approach was used, with the possibility of failing to detect other reactive epitopes. In contrast, in the present study all potential epitopes were investigated systematically using peptides spanning the entire length of the CYP2D6 protein. With this approach, we were able to confirm both existence and recognition frequency of the previously described epitopes. In agreement with Yamamoto et al. (12), we found that CYP2D6253–272, encompassing the hitherto considered main epitope (10), was recognized by 85% of our patients with LKM1 AIH and by only 1 of 10 LKM1-positive HCV-infected patients.

Of the four new epitopes identified in the present study (CYP2D6193–212, CYP2D6238–257, CYP2D6268–287, and CYP2D6478–497), CYP2D6193–212 is immunodominant. Hints were already present in the literature, indicating the antigenic importance of this CYP2D6 region. After the previous finding of Yamamoto et al. (13) that LKM1 sera from HCV-infected patients react preferentially with a region comprised within amino acids 208–349 of CYP2D6, Parez et al. (14) constructed a series of peptides selected on the basis of their high probability of being antigenic sites, as predicted by computer analysis. They found that peptide 200–214 contained an epitope recognized by LKM1 in HCV infection, but not in AIH. On the basis of this finding and in the apparent unawareness of our initial reports describing the new epitopes (26, 27), Klein et al. (15) constructed a peptide 4 aa longer than that of Parez et al. (14) at both the N and C ends (CYP2D6196–218) and obtained surprising results. In contrast with Parez et al. (14), they found that only 18% of LKM1-positive HCV-infected patients, but 68% of LKM1-positive AIH patients, reacted with the new peptide.

The methodological strategy used in the present paper has enabled us to clarify the antigenic characteristics of the region 193–218 on CYP2D6 and to explain, at least in part, these contrasting results (Fig. 1B). We demonstrate that the "AIH epitope" resides within the 193–212 sequence, with arginine, arginine, and phenylalanine at positions 193–195 conferring virtually universal recognition. Sequences after 212 do not appear to have antigenic properties, as suggested by the fact that our own peptide 208–227 is ignored by both AIH and HCV LKM1-positive sera. This also indicates that the "HCV epitope" is contained in the preceding peptide spanning amino acids 193–212, recognized by 50% of our HCV-positive patients. Though it may be argued that the 3-aa difference at the N terminus could explain this higher reactivity frequency among our HCV-positive patients when compared with an 18% frequency described by Klein et al. (15), this contention is not supported by the data of Parez et al. (14), who found a frequency similar to ours using peptide 200–214, not containing amino acids 193–199. Because the three studies have used assays of similar sensitivity, methodological differences are unlikely to explain these discrepancies. It is possible that the autoantibody concentration was lower in the patients studied by Klein et al. (15), who defined their LKM1 sera as exceeding a 1/80 dilution without providing the exact titer. Our data show that the epitope recognized in HCV infection is not present beyond amino acid 208, because CYP2D6208–227 was not recognized by any LKM1/HCV-positive sera (28).

Interestingly, the dominant CYP2D6193–212 epitope found in the present study shares similarities with HCV and CMV. Modeling studies locate the immunodominant CYP2D6193–212 peptide to the surface of CYP2D6, making it accessible to Ab recognition in the context of the physiologically folded protein (21, 22). Similarly, the homologous HCV2977–2996 peptide, part of the HCV RNA-dependent DNA polymerase (NS5), lies within a solvent-accessible loop structure of the "thumb" domain of this protein, allowing it to serve as an epitope on the native folded protein (23). No crystal coordinates are available for the CMV alkaline exonuclease, which contains the region CMV121–140 homologous to CYP2D6193–212, but preliminary secondary structure analysis suggests it to be solvent accessible.

We have experimentally demonstrated these homologous self and viral sequences to be targets of cross-reactive virus/self Abs, suggesting the involvement of a molecular mimicry mechanism (29), where the physiological immune response to a pathogen targets sequence-sharing self Ags. This mechanism was invoked by Manns et al. (11) to explain the production of LKM1 in HCV infection, after having noted that CYP2D6263–271 and the HCV polyprotein HCV2772–2780 share sequence similarities. This hypothesis, however, was not tested in studies of cross-inhibition. We demonstrate that serum from LKM1-positive patients with AIH or HCV infection react concurrently with CYP2D6193–212, HCV2977–2996, and CMV121–140. Multiple reactivity is due to an Ab capable of recognizing both virus and self, as demonstrated by inhibition studies. Furthermore, we confirm by alanine substitution of the homologous HCV epitope HCV2977–2996 that Ab reactivity to this sequence maps to the region with homology to CYP2D6193–212 and CMV121–140. Finally, we show by ELISA and a CYP2D6 metabolic assay that CYP2D6193–212-specific Abs recognize and inhibit the activity of the native enzyme and that the observed peptide cross-reactivities among CYP2D6, HCV-NS5, and CMV-UL98 also operate on the whole Ags when tested in a competition assay.

The cross-reactive Abs detected are of the IgG isotype, implicating T cell "help" in their generation. Because T and B cell epitopes frequently overlap, it is possible that the cross-reactive humoral epitopes identified may also serve as T cell epitopes (29). Thus, the emergence of LKM1 may be the ultimate manifestation of molecular mimicry at the level of the T cell, giving rise to a T-dependent and class-switched cross-reactive humoral response (30).

In conclusion, we found a new immunodominant epitope of LKM on CYP2D6, reactivity to which is likely to arise as a consequence of its similarity with viral Ags. Multiple exposure to viral sequences shared with self may lead to the break of immunological tolerance through a multihit mechanism in predisposed individuals.



View larger version (42K):
[in this window]
[in a new window]
  		FIGURE 2B. continued.

CYP2D6268–287 (peptide 19), recognized by two AIH2 cases (patients 1 and 9); and CYP2D6478–497 (peptide 33), recognized by LKM1/HCV patient number 21. The open arrows mark previously described epitopes: CYP2D6253–272 (peptide 18), is recognized by 11 AIH2 cases (patients 1–7, 9, 10, 12, and 13) and LKM1/HCV patient number 21; CYP2D6328–347 (peptide 23), recognized by AIH2 patient number 2; CYP2D6343–362 (peptide 24), recognized by LKM1/HCV patient number 16; CYP2D6373–392 (peptide 26), recognized by AIH2 patient number 3; CYP2D6418–437 (peptide 29), recognized by three AIH2 cases (patients 3, 4, and 7) and LKM1/HCV patient number 14.

 

   Acknowledgments
 
We thank Dr. Paul Cheeseman (King’s College Hospital, London, U.K.) for his help with the artwork, Dr. Sylvia Nagl (Department of Biochemistry and Molecular Biology, University College, London, U.K.) for performing structural modeling of CYP2D6, Dr. Stephen Hodges for his help with the CYP2D6 enzymatic inhibition assay, and Dr. Munther J. Hussain for performing the IgG anti-CMV assay. The healthy children were recruited through the Children’s Liver Disease Foundation Gillingham and Medway Branch (Kent, U.K.). We also thank Dr. Tenney (Bristol-Myers Squibb) and Dr. Seong (Yonsei University) for their generous gift of recombinant proteins.


   Footnotes
 
1 G.M.-V. is supported by the Children’s Liver Disease Foundation (Birmingham, U.K.) and the Children Nationwide Medical Research Trust (London, U.K.). At the time the study was undertaken, N.K. held a scholarship from the Royal College of Physicians (London, U.K.), funded by Children’s Nationwide Medical Research Fund. Y.M. is a Dorothy Hodgkin Fellow of the Royal Society (London, U.K.). D.P.B. is supported by the Children’s Liver Disease Foundation. Back

2 Address correspondence and reprint requests to Dr. Diego Vergani, Institute of Liver Studies, Alex P. Mowat Laboratory, King’s College Hospital, Denmark Hill, London SE5 9RS, U.K. E-mail address: d.vergani{at}kcl.ac.uk Back

3 Abbreviations used in this paper: LKM1, liver kidney microsomal Ab type 1; AIH2, autoimmune hepatitis type 2; HCV, hepatitis C virus; CYP2D6, cytochrome P4502D6; NS5B, nonstructural protein 5B; AMMC, 3-[2-(N,N-diethylN-methylamino)ethyl]-7-methoxy-4-methylcoumarin. Back

Received for publication May 4, 2000. Accepted for publication November 8, 2002.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Homberg, J. C., N. Abuaf, O. Bernard, S. Islam, F. Alvarez, S. H. Khalil, R. Poupon, F. Darnis, V. G. Levy, P. Grippon, et al 1987. Chronic active hepatitis associated with antiliver/kidney microsome antibody type 1: a second type of "autoimmune" hepatitis. Hepatology 7:1333.[Medline]
  2. Giostra, F., A. Manzin, M. Lenzi, R. Francesconi, L. Solforosi, P. Manotti, L. Muratori, D. Zauli, M. Clementi, F. B. Bianchi. 1996. Low hepatitis C viraemia levels in patients with anti-liver kidney microsomal antibody type 1 positive chronic hepatitis. J. Hepatol. 25:433.[Medline]
  3. Vergani, D.. 1999. NOSA in HCV infection: markers or makers of disease?. Gut 45:328.[Free Full Text]
  4. Rizzetto, M., G. Swana, D. Doniach. 1973. Microsomal antibodies in active chronic hepatitis and other disorders. Clin. Exp. Immunol. 15:331.[Medline]
  5. Alvarez, F., O. Bernard, J. C. Homberg, G. Kreibich. 1985. Anti-liver-kidney microsome antibody recognizes a 50,000 molecular weight protein of the endoplasmic reticulum. J. Exp. Med. 161:1231.[Abstract/Free Full Text]
  6. Waxman, D. J., D. P. Lapenson, M. Krishnan, O. Bernard, G. Kreibich, F. Alvarez. 1988. Antibodies to liver/kidney microsome1 in chronic active hepatitis recognize specific forms of hepatic cytochrome P-450. Gastroenterology 95:1326.[Medline]
  7. Zanger, U. M., H. P. Hauri, J. Loeper, J. C. Homberg, U. A. Meyer. 1988. Antibodies against human cytochrome P-450db1 in autoimmune hepatitis type II. Proc. Natl. Acad. Sci. USA 85:8256.[Abstract/Free Full Text]
  8. Gueguen, M., A. M. Yamamoto, O. Bernard, F. Alvarez. 1989. Anti-liver-kidney microsome antibody type 1 recognizes human cytochrome P450 db1. Biochem. Biophys. Res. Commun. 159:542.[Medline]
  9. Manns, M. P., E. F. Johnson, K. J. Griffin, E. M. Tan, K. F. Sullivan. 1989. Major antigen of liver kidney microsomal autoantibodies in idiopathic autoimmune hepatitis is cytochrome P450db1. J. Clin. Invest. 83:1066.
  10. Gueguen, M., O. Boniface, O. Bernard, F. Clerc, T. Cartwright, F. Alvarez. 1991. Identification of the main epitope on human cytochrome P450 IID6 recognized by anti-liver kidney microsome antibody. J. Autoimmun. 4:607.[Medline]
  11. Manns, M. P., K. J. Griffin, K. F. Sullivan, E. F. Johnson. 1991. LKM-1 autoantibodies recognize a short linear sequence in P450IID6, a cytochrome P-450 monooxygenase. J. Clin. Invest. 88:1370.
  12. Yamamoto, A. M., D. Cresteil, O. Boniface, F. F. Clerc, F. Alvarez. 1993. Identification and analysis of cytochrome P450IID6 antigenic sites recognized by anti-liver-kidney microsome type-1 antibodies (LKM1). Eur. J. Immunol. 23:1105.[Medline]
  13. Yamamoto, A. M., D. Cresteil, J. C. Homberg, F. Alvarez. 1993. Characterization of anti-liver-kidney microsome antibody (anti-LKM1) from hepatitis C virus-positive and -negative sera. Gastroenterology 104:1762.[Medline]
  14. Parez, N., D. Herzog, E. Jacqz-Aigrain, J. C. Homberg, F. Alvarez. 1996. Study of the B cell response to cytochrome P450IID6 in sera from chronic hepatitis C patients. Clin. Exp. Immunol. 106:336.[Medline]
  15. Klein, R., U. M. Zanger, T. Berg, U. Hopf, P. A. Berg. 1999. Overlapping but distinct specificities of anti-liver-kidney microsome antibodies in autoimmune hepatitis type II and hepatitis C revealed by recombinant native CYP2D6 and novel peptide epitopes. Clin. Exp. Immunol. 118:290.[Medline]
  16. Alvarez, F., P. A. Berg, F. B. Bianchi, L. Bianchi, A. K. Burroughs, E. L. Cancado, R. W. Chapman, W. G. Cooksley, A. J. Czaja, V. J. Desmet, et al 1999. International Autoimmune Hepatitis Group report: review of criteria for diagnosis of autoimmune hepatitis. J. Hepatol. 31:929.[Medline]
  17. Muratori, L., M. Lenzi, M. Cataleta, F. Giostra, F. Cassani, G. Ballardini, D. Zauli, F. B. Bianchi. 1994. Interferon therapy in liver/kidney microsomal antibody type 1-positive patients with chronic hepatitis C. J. Hepatol. 21:199.[Medline]
  18. Rodda, S. J., G. Tribbick. 1996. Antibody-defined epitope mapping using the multipin method of peptide synthesis. Methods 9:473.[Medline]
  19. Voller, A., D. Bidwell, A. Bartlett. 1980. Enzyme-linked immunosorbent assay. N. R. Rose, and H. Friedman, eds. Manual of Clinical Immunology 2nd Ed.368. American Society of Microbiology, Washington, DC.
  20. Guex, N., M. C. Peitsch. 1997. SWISS-MODEL and the Swiss-Pdb Viewer: an environment for comparative protein modelling. Electrophoresis 18:2714.[Medline]
  21. Ravichandran, K. G., S. S. Boddupalli, C. A. Hasermann, J. A. Peterson, J. Deisenhofer. 1993. Crystal structure of hemoprotein domain of P450BM-3, a prototype for microsomal P450’s. Science 261:731.[Abstract/Free Full Text]
  22. de Groot, M. J., M. J. Ackland, V. A. Horne, A. A. Alex, B. C. Jones. 1999. Novel approach to predicting P450-mediated drug metabolism: development of a combined protein and pharmacophore model for CYP2D6. J. Med. Chem. 42:1515.[Medline]
  23. Lesburg, C. A., M. B. Cable, E. Ferrari, Z. Hong, A. F. Mannarino, P. C. Weber. 1999. Crystal structure of RNA-dependant RNA polymerase from hepatitis C virus reveals a fully encircled active site. Nat. Struct. Biol. 6:937.[Medline]
  24. Kerkar, N., Y. Ma, M. Hussain, L. Muratori, C. Targett, R. William, F. B. Bianchi, G. Mieli-Vergani, D. Vergani. 1999. A novel assay for detecting antibodies to cytochrome P4502D6, the molecular target of liver kidney mocrosomal antibody type 1. J. Immunol. Methods 223:227.[Medline]
  25. Deleage, G., F. F. Clerc, B. Roux, D. C. Gautheron. 1988. ANTHEPROT: a package for protein sequence analysis using a microcomputer. Comput. Appl. Biosci. 4:351.[Abstract/Free Full Text]
  26. Kerkar, N., Y. Ma, A. Mahmoud, L. Muratori, M. Lenzi, F. B. Bianchi, R. Williams, G. Mieli-Vergani, D. Vergani. 1998. Identification of new epitopes on cytochrome P4502D6 in autoimmune hepatitis and chronic hepatitis C infection. Hepatology 28:1.[Medline]
  27. Kerkar, N., A. Mahmoud, L. Muratori, M. Lenzi, F. B. Bianchi, R. Williams, G. Mieli-Vergani, D. Vergani. 1999. New viral triggers of autoreactivity to cytochrome P4502D6, the target of liver kidney microsomal antibody type1. J. Hepatol. 30:(Suppl.):58.
  28. Oldstone, M. B.. 1987. Molecular mimicry and autoimmune disease. Cell 50:819.[Medline]
  29. Tung, K. S.. 1994. Mechanism of self-tolerance and events leading to autoimmune disease and autoantibody response. Clin. Immunol. Immunopathol. 73:275.[Medline]
  30. Wuscherpfennig, K. W., J. L. Strominger. 1995. Molecular mimicry in T-cell mediated autoimmunity: viral peptides activate human T cell clones specific for myelin basic protein. Cell 80:695.[Medline]



This article has been cited by other articles:


Home page
 JEMHome page
M. Holdener, E. Hintermann, M. Bayer, A. Rhode, E. Rodrigo, G. Hintereder, E. F. Johnson, F. J. Gonzalez, J. Pfeilschifter, M. P. Manns, et al.
Breaking tolerance to the natural human liver autoantigen cytochrome P450 2D6 by virus infection
J. Exp. Med., June 9, 2008; 205(6): 1409 - 1422.
[Abstract] [Full Text] [PDF]

Home page
 NEJMHome page
E. L. Krawitt
Autoimmune Hepatitis
N. Engl. J. Med., January 5, 2006; 354(1): 54 - 66.
[Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
M. Vidali, M. Hidestrand, E. Eliasson, E. Mottaran, E. Reale, R. Rolla, G. Occhino, E. Albano, and M. Ingelman-Sundberg
Use of Molecular Simulation for Mapping Conformational CYP2E1 Epitopes
J. Biol. Chem., December 3, 2004; 279(49): 50949 - 50955.
[Abstract] [Full Text] [PDF]

Home page
 The Annals of PharmacotherapyHome page
M. Pompili, M. Basso, A. Grieco, F. M Vecchio, G. Gasbarrini, and G. L. Rapaccini
Recurrent Acute Hepatitis Associated with Use of Cetirizine
Ann. Pharmacother., November 1, 2004; 38(11): 1844 - 1847.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Kerkar, N.
Right arrow Articles by Vergani, D.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Kerkar, N.
Right arrow Articles by Vergani, D.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS