Autoreactive epitopes within the human α-enolase and their recognition by sera from patients with endometriosis

of Autoimmunity

Journal

(1995)

8, 93 l-945

Autoreactive Epitopes within the Human a-Enolase and their Recognition by Sera from Patients with Endometriosis

Michael Walter,*f Heike Berg,* Freimut A. Leidenberger,# Karl-Werner Schweppe,s and Wolfgang Northemann* *Department of Molecular Biology, ELLA.3 Entwicklungslabor, D-791 14 Freiburg Germany, tDepartment of Biology, University of Freiburg, D-791 04 Freiburg, Germany, $Institute of Hormone and Fertility Research, University of Hamburg, D-22529 Hamburg, Germany, and SDepartment of Gynecology and Obsterrics, Central Hospital Ammerland, D-26655 Westerstede, Germany (Received

18 May

1995

and

accepted

12 July

1995)

Patients with endometriosis significantly develop autoantibodies directed against endometrial proteins, which may be involved in the aetiology of this gynaecological disease. Based on standard Western blot analysis, a 48 kDa protein was localized in the soluble protein extract of endometrial adenocarcinoma cells using sera from patients with clinically staged endometriosis and identified as the glycolytic enzyme a-enolase. The corresponding cDNA coding for the human ‘a-enolase was isolated from a human endometrial cDNA library and cloned into the vector pH6EX3, allowing the efficient expression of recombinant human a-enolase with an N-terminal histidine-hexapeptide as affinity ligand in Escherichia coli. The purified recombinant human a-enolase was evaluated as a specific antigenic tool for the diagnostic measurement of antiendometrial antibodies in sera from patients with endometriosis. With selected endometriosis sera, two linear autoreactive epitopes were localized within the recombinant human a-enolase using epitope mapping techniques, and they were characterized. 0 1995 Academic Press Limited

Introduction

Endometriosis is one of the most commonly encountered diseases in gynaecology affecting 7-l 2% of the general female population of reproductive age, and clinically Correspondence should ELLIS Entwicklungslabor,

be addressed to: Dr Michael Obere Hardtstrasse 18, D-791

Walter, Department 14 Freiburg, Germany.

of Molecular

Biology,

931 0896-841

l/95/060931

+ 15 $12.0010

0

1995

Academic

Press

Limited

932

M. Walter

et al.

characterized by pelvic pain, menstrual disorders, inflammatory and immunologic changes, and infertility in up to 3045% of the patients [for review see l-61. The most accepted theory suggests that the pathogenesis of endometriosis involves the uncontrolled implantation and proliferation of endometriotic epithelium cells outside their physiological location in the uterus, preferentially in the peritoneal cavity and with high frequency on the ovaries. However, the aetiology of endometriosis remains still largely unexplained. Several recent studies have indicated that alterations in both humoral and cell-mediated immune functions are the major cause of endometriosis, indicating the activation of the immune system. The reports have underlined the association between endometriosis and defects either in the cell-mediated immune system, including a decreased response of peritoneal lymphocytes such as natural killer cells towards autologous endometrial antigens [7-91 or in maturation and function of peritoneal macrophages [lo], both allowing the implantation and development of ectopic endometrial cells in the peritoneal cavity, respectively. Furthermore, it is now generally discussed that the artificially located endomenial cells as well as released proteins may be biochemically different from those of the uterine endometrium [l 11, and therefore recognized by the immune system as ‘foreign’ and provoke humoral generation of disease-specific autoantiboidies in analogy to other autoimmune diseases [12-l 81. The theory of an autoimmune response is strengthened by detection of a variety of other autoantibodies such as those directed against cardiolipin [19] and nuclear proteins [16, 201. The current diagnosis of endometriosis normally carried out by direct observation and biopsy after laparoscopy is not failproof because even women with no evidence of typical clinical symptoms at laparoscopy could be affected by microscopic and invisible, non-pigmented lesions of endometriosis which are routinely missed. Therefore, a reliable, sensitive and specific assay based on non-invasive serological’diagnosis is desired, in particular for evaluating and monitoring women for endometriosis with the aim of assisting and/or replacing diagnostic laparoscopy. Considering .the disease-specific autoantibodies generated against cellular components of endometrial cells as a potential serological marker of endometriosis, several attempts to identify the corresponding antigens within the endometrial tissue or derived cell lines have been recently described [13, 17, 211. However, the nature of these possible antigens is still not known. The objective of this study was to search for an endometrial protein which was highly significantly recognized by sera from patients with endometriosis, and therefore could serve as an antigenic target in an immunoassay system for the diagnostic measurement of endometriosis-specific autoantibodies. A 48 kD antigen was characterized in human endometrial adenocarcinoma cells by Western blot analysis, purified and identified as the a-enolase by protein microsequencing technique. In this paper we describe the cDNA cloning and expression of recombinant human cl-enolase in Escherichia coli mediated by high-efficiency expression vectors with respect to two objectives. First, we aimed to evaluate the chromatographically purified recombinant human a-enolase as a specific and suitable antigen for the diagnosis of endometriosis. Second, we were interested in localizing and characterizing linear antigenic epitope(s) within human a-enolase

Endometriosis-specific

autoantigenic

epitopes

within

the human

which reacted with the disease-specific autoantibodies endometriosis. Materials

a-enolase

933

in sera from patients with

and methods

Sera ji-om patients with endometriosis and blood donors The pathological sera (group 1, n=6) studied here were selected from a large pool of sera obtained from women with endometriosis which were clinically diagnosed at laparoscopy. The severity of disease was staged at III and IV according to the revised classification of endometriosis by the American Fertility Society [22]. Sera of blood donors (group 2, YZ=5) randomly chosen from a large pool of women in the corresponding reproductive age were used as negative controls. Cell culture The human endometrial adenocarcinoma cell line HEC-1B (ATCC HTB 113) was cultured as monolayers in Eagle’s MEM medium (Gibco) complemented with 10% fetal calf serum. Cultures of HEC-1B cells with 70-80% confluency were harvested. The protein concentration of the cell lysates was determined by the BCA-method (Sigma). SDS gel electrophoresis and Western blot analysis About 8 l.tg protein of total cell lysate were separated by 12% SDS-PAGE under reducing conditions and transferred electrophoretically onto nitrocellulose filter (Schleicher & Schuell) using a tram-blot, semi-dry electrophoretic transfer cell (Bio-Rad). Unoccupied protein binding sites on the filter were blocked with 5% nonfat dried milk for 60 min at room temperature. The filters were incubated with lo-fold dilutions of sera from patients and blood donors for 16 h at 4°C. The bound antibodies were visualized with anti-human IgG immunoglobulins @omega) conjugated with alkaline phosphatase after incubation for 45 min at room temperature according to the manufacturer’s protocol (Promega). PuriJication of a-enolase from HEC-1B

cells

HEC-1B cells with a protein mass of about 25 mg were harvested, resuspended in 50 ml hypotonic buffer (10 mM TrisHCl, pH 7.8, 10 mM KCl, 1.5 mM MgCl,, 0.5 mM DTT, 0.5 mM PMSF and 2 @ml of each following proteinase inhibitors: aprotinin, leupeptin, pepstatin and bestatin), and homogenized with 10 strokes of a glass-teflon homogeniser (tight-fitting). The cell homogenate was incubated at 0°C for 10 min and cleared of cell debris by two centrifugation steps at 1000 x g for 10 min and at 180,000 x g for 30 min. The final supernatant representing cytoplasmic proteins was adjusted to buffer A (20 mM Tris/HCl, pH 7.9, 1.5 mM MgCl,, 0.5 mM DTT, 0.5 mM PMSF) and then directly applied onto a 1 ml anion exchange column MonoQ HR5/5 (Pharmacia). The loaded column was developed with a salt gradient (O-l .O M NaCl) by FPL chromatography. The flow through and

934

M. Walter

et al.

eluted fractions were immunologically analysed with a selected serum of an endometriosis patient. The fractions of interest were adjusted to buffer B (50 mM sodium acetate, pH 6.0, 1.5 mM MgCI,, 0.5 mM DTT, 0.5 mM PMSF) and subsequently processed onto a 1 ml cation exchange column MonoS HR5/5 (Pharmacia). The bound proteins were successively eluted with buffer B by increasing the concentration of NaCl (O-l.0 M) and immunologically analysed as described above. The flow through containing essentially pure a-enolase was dialysed against buffer C (20 mM Na-Hepes, pH 7.9, 1.5 mM MgCl,, 1 mM DTT, 1 mM PMSF) and stored at - 80°C. Purification of antibodies directed against a-enolase About 1 .O mg natural or recombinant a-enolase was separated by 12% SDS-PAGE and transferred electrophoretically onto nitrocellulose filter. The immobilized a-enolase was stained with Ponceau S (Sigma), cut out as a small strip from the filter, and incubated with a IO-fold dilution of a selected patient’s serum according to the Western blot technique as described above. The filter strip was washed 3 times with excess of PBS containing 0.05% Tween-20. The a-enolase antibodies were eluted with PBS containing 3 M potassium thiocyanate and 0.1% BSA for 5 min, dialysed against 0.1 x PBS, and concentrated with Centriprep-30 (Amicon). Protein sequence analysis of a-enolase About 10 p,g of a-enolase purified from the HEC-1B cells were employed for a protein microsequencing technique which was carried out by Prof. F. Lottspeich, TopLab GmbH, Mi.inchen. Proteolytic fragments were generated by digestion of a-enolase with the endoproteinase LysC, gel purified, and N-terminally analysed by Edman degration. The resulting two amino acid sequences were analysed by homology search with the Swiss Prot database (PharmaciaHitachi). Generation of cDNA library derived from human HEC-1B

cells

Total RNA was extracted from solubilized endometrial HEC-1B cells with watersaturated phenol according to a standard protocol @‘x-omega). The poly(A)+RNA was isolated by chromatography using oligo(dT) cellulose (Sigma). For the generation of a human endometrial cDNA library primed with oligo(dT), 5.0 pg of poly(A)+RNA were used as template for the synthesis of double-stranded cDNA prior to the unidirectional cloning in the lambda-ZAP11 vector (Stratagene) according to the manufactuer’s protocol (Stratagene). The obtained endometrial cDNA library represented more than lo6 independent clones. PCR-ampljication

of a-enolase cDNA

Recombinant DNA was isolated from about lo6 plaques of the endometrial cDNA library according to a standard protocol with the lambda-phage DNA isolation kit (Quiagen). A total of 0.1 pg recombinant cDNA was amplified by the PCR with 5 Units Taq DNA polymerase (Pharmacia) in a 100 @reaction mix containing

Endometriosis-specific

autoantigenic

epitopes

within

the human

a-enolase

935

10 mM Tris/HCl, pH 8.0, 50 mM KCl, 1.5 mM MgCl,, 100 PM dNTPs, and 5 pmole of each primer. The amplification reaction was carried out for 34 cycles with the following cycling times: denaturation, 2 min at 9YC, annealing, 1 min at 56 min, and primer extension, 2 min at 72°C. Both the 5’-PCRprimer (5’CAGITGGATC CATGTCTATT CTCAAGATCC ATGCCA-3’) and the 3’-PCR primer (5’-ATCGAATTCT TACTTGGCCA AGGGGRTTTCT GAAGTTCCTG-3’) were designed to add on the terminal restriction sites BamHI and Eco RI for cloning purposes and corresponded to the homologous cDNA sequence of the published human a-enolase, respectively [23]. The a-enolase cDNA synthesized was subcloned in the Bluescript SK(-f) (Stratagene) to construct pEA48B and proven by sequence analysis according to standard protocol using T7 DNA polymerase (Pharmacia). Cloning and expression of a-enolase cDNA in E. coli The a-enolase cDNA was inserted into the BamHI and EcoRI sites of the prokaryotic expression vector pH6EX3 [24] which directed the expression of a recombinant fusion protein with the human a-enolase at its C-terminus and a histidine-hexapeptide at its N-terminus under the control of the strong tacpromoter [25]. The obtained clone pEA48H was used to transform the E. coli strain LE392 [26] in order to induce the expression of the corresponding recombinant a-enolase fusion protein with 1 mM IPTG for 6 h as described previously [24]. Purification of recombinant a-enolase The purification of the recombinant a-enolase fused to the N-terminal histidinehexapeptide as an affinity ligand was performed by single-step aflinity chromatography using a metal chelating matrix as detailed previously [27]. The induced bacterial cells were sedimented, resuspended in PBS, lysed with lysozyme, and separated into the soluble and insoluble cell fractions by centrifugation. The soluble cell fraction was applied onto a chelating Sepharose FF column (Pharmacia) charged with nickel ions. The loaded column was processed with a pH step gradient ofpH 6.0, 5.5, 5.0, and 4.0. Construction of a-enolase epitope cDNA library About 20 pg of the 1.3 kb a-enolase cDNA insert was isolated, multimerized, and fragmented by controlled sonication with 8 times 20 sec. pulses of about 100 watts at 0°C as described previously [28]. The sonicated cDNA was end-repaired with T4 DNA polymerase (Pharrnacia) and ligated into the SmaI site of pGEX-ST [29] coding for a fusion protein with the 28 kDa glutatione-S-transferase (GST) at their N-termini. After transformation of the E. coli strain LE392, the obtained colonies representing the a-enolase epitope cDNA library were collected and stored as glycerol stocks at - 80°C. Expression and screening if the a-enolase epitope cDNA libray The a-enolase epitope cDNA library was titered, plated with a density of about 300 colonies/85 mm petri dish, and incubated until the colonies were 0.5-l mm in

936

M. Walter

et al.

c*b

kDa

1 2 3 4 5 6 7 8 9101112

Figure 1. Western blot analysis of endometrial cell extact with sera from patients with endometriosis and from blood donors. Aliquots of HEC-1B cell lysates, each representing about 8 pg protein, were applied to 12% SDS-PAGE. The separated proteins were transferred onto a nitrocellulose filter and analysed by Western blotting using sera from patients with endometriosis (lanes 1-6) and from blood donors (lanes 7-l 1). A control filter was incubated with only conjugated anti-human immunoglobulins (lane 12).

diameter. The colonies were lifted with nitrocellulose filters (Schleicher & Schuell) and placed onto prewarmed agar plates containing 5 mM IPTG to induce the expression of the fusion proteins for 5 h at 37°C. The bacteria were lysed in 2.5% SDS for 20 min at 55°C and electrophoretically blotted onto the nitrocellulose membrane. The immunoreaction was carried out according to the Western blotting technique as described above using the purified anti-enolase antibodies. All clones positively screened for specific epitopes were confirmed by expression of the recombinant GST fusion proteins prior to Western blot analysis under denaturing and reducing conditions. The sera used in this study were diluted lo-fold. The identified cDNA inserts were characterized by DNA sequencing and aligned with the full-length cDNA of the human a-enolase. Results Identification

of human

a-enolase

as an endometrial

antigen

In a search for a disease-related antigen, the cell extract of the human endometrial adenocarcinoma cell line HEC-IB was examined by SDS-PAGE under reducing conditions followed by Western blot analysis with selected sera from patients with clinically staged endometriosis (Figure 1, lanes l-6) and from healthy female blood donors (Figure 1, lanes 7-l 1). A highly significant immunoreaction of the patient’s sera with a 48 kDa protein, primarily designated 48 kDa endometrial antigen (EA48), was observed. The EA48 was not recognized by sera from blood donors. For further identification and characterization of the EA48, the soluble cell extract from HEC-1B cells was fractionated by two subsequent chromatographic steps

Endometriosis-specific

autoantigenic

epitopes

within

the

human

a-enolase

937

1018 517/506 398 322l 220/221 1!54/134

M

1234

Figure 2. PCR amplification and cloning of u-enolase cDNA. The total input of 0.1 ug DNA derived from about 10” plaques of a human endometrial cDNA library was amplified 1.3 kb cl-enolase cDNA was inserted into Bluescript SK(f-) and cloned in E. co/i. Aliquots the 50-ul PCR reaction and from plasmid preparations representing 300 ul bacterial separated in 1% agarose gel and visualized by staining with ethidium bromide. PCR product cDNA (lane l), Bluescript SK(f-) (lane 2) and pEA48B linearized with EcoRI (lane digested with BamHI and EcoRl (lane 4). Lane M, size marker.

recombinant by PCR. The of 10 111from culture were of cl-enolase 3), pEA48B

using an anion and cation exchange matrix. The EA48 purified to almost homogeneity was not retarded by the anion exchange or by the cation exchange column, indicating its unpolar nature at a given pH. The N-terminal sequence analysis of proteolytic peptides derived from the purified EA48 resulted in two amino acid sequences, KTIAPALVSK and KFAGRNFRNPLAK. After homology studies with an available database (Swiss Prot), the EA48 was identified as the human 48 kDa glycolytic enzyme u-enolase (2-phosphopyruvate-hydratase, EC 4.2.1.1.). The obtained peptide sequences revealed sequence homologies of 100% with the corresponding amino acid sequence of human a-enolase at positions 71-80 and 422434, respectively (data not shown). Isolation

and characterization

of the human

a-enolase

cDNA

Based on the published cDNA sequence [23], oligonucleotide primers were designed for isolation of the cDNA coding for the 434 amino acids of full-length human a-enolase including the translational stop signal TAA by PCR amplification of a cDNA library derived from mRNA of cultured HEC-1B cells (Figure 2). The synthesized cDNA of 1305 bp in length was subcloned and confirmed by DNA sequence analysis. The a-enolase cDNA displayed only one base transition (T-C) at position 755, causing an exchange of the amino acid phenylalanine to serine at position 252, which however did not alter the autoantigenicity of the expressed human a-enolase (data not shown).

938

M. Walter

et al.

Ml23456769 Figure 3. Expression and purification of recombinant a-enolase by metal chelating aflinity chromatography. The E. colt’ strain LE392 transformed with pEA48H was cultured, induced with 1 mM IPTG for 6 h, and fractionated in order to purify the recombinant a-enolase fused to N-terminal histidinehexapeptide (arrow). Aliquots with equal amounts of protein representing about 50-ul bacterial culture were taken from several steps and separated by 12% SDS-PAGE. The recombinant a-enolase in the soluble cell fraction was purified by affinity chromtography using a chelating Sepharose loaded with Ni’+ ions and eluted with a pH step gradient. E. coli lysate without (lane 1) and after induction with IPTG (lane 2), insoluble cell fraction (lane 3), soluble cell fraction (lane 4), flow through (lane 5), fraction eluted at pH 6.0 (lane 6), pH 5.5 (lane 7), pH 5.0 (lane S), and pH 4.0 (lane 9). Lane M, molecular mass marker.

Expression

of recombinant

a-enolse

The a-enolase cDNA was inserted into the high-efficiency prokaryotic expression vector pH6EX3 [24]. The obtained clone, pEA48H, directed the synthesis of recombinant human a-enolase with a histidine-hexapeptide at its N-terminus, under the control of the strong tat-promoter. After transformation of the E. cola’ strain LE392 and induction with 1 mM IPTG for 6 h at 37’C, the expression of recombinant a-enolase was observed with the expected apparent molecular mass of about 50 kDa, mainly in the soluble cell fraction (Figure 3, lanes 2, 3, 4). Under optimal conditions, the recombinant a-enolase constituted up to 20% of total cellular proteins. Purification

and Western

blot analysis

of recombinant

a-enolase

with endometriosis

sera

The recombinant human a-enolase was expressed as a histidine-hexapeptide fusion protein which allowed a simplified purification protocol of the recombinant protein directly from the bacterial cell lysate by affinity chromatography based on the specific complex formation between the N-terminal histidine-hexapeptide and the immobilized nickel ions [24]. About l-2 mg recombinant human a-enolase with almost homogeneous purity was isolated from a 100 ml bacteria culture (Figure 3, lanes 5-9). Western blot analysis with the selected sera from patients with endometriosis and from blood donors, as indicated in Figure 1, confirmed the immunological identity of both the EA48 derived from the human endometrial carcinoma cells HEC-1B and the recombinant a-enolase (Figure 4). For the evaluation of the human

Endometriosis-specific

autoantigenic

epitopes

kDa $.. . 139.9 -- ::: 66.0 - ‘L; ILL :’ i!;

;<

i

,a,.!

,

within

the human

a-enolase

939

l,

47.0

33.3

z&6-,

20.7-

-

1 *. : 1’ i .r ’ ! ’ I..:; ; ’ 1<‘$I’ I . !:’

1234

5 6 7 8 9101112

Figure 4. Western blot analysis of the purified recombinant human u-enolase with sera from patients with endometriosis and from blood donors. About 0.75 ug of recombinant a-enolasepurified by aflinity chromatography was separated by 12% SDS-PAGE, transferred onto nitrocellulose filter and analysed by Western blotting using sera from patients with endometriosis (lanes l-6) and from blood donors (lanes 7-l 1). A control filter was incubated with conjugated anti-human immunoglobuhns (lane 12).

a-enolase as an antigenic tool for the serological diagnosis of endometriosis, in a preliminary study we measured the autoantibodies in 18 (51%) of 35 sera derived from patients with clinically staged endometriosis and in 2 (6%) of 30 sera from healthy female blood donors by Western blot analysis using the recombinant a-enolase (data not shown). Mapping

of linear autoantagenic epitopes within the human a-enolase

Approximately 2000 colonies of the a-enolase cDNA epitope library generated in the expression vector pGEX-ST [29] coding for recombinant fusion proteins with an N-terminal GST were screened for linear autoreactive epitopes of the human a-enolase with the purified anti-a-enolase antibodies obtained from a selected serum of a patient with endometriosis. In total, 29 clones were identified and characterized by Western blot analysis and DNA sequencing (data not shown). No cross-reaction of the patient’s serum was observed with the fusion moiety GST. After alignment of the cDNA sequences with the full-length a-enolase cDNA, two segments were localized between the nucleotide positions 157 and 262, and 619 and 7 14 coding for 35 and 32 amino acids bearing linear autoantigenic epitopes, respectively (Figure 5). These sequences were responsible for the amino acids between position 53 and‘87, and 207 and 238 of the human a-enolase (Figure 6). In order to study the frequency of the autoantibodies ‘d,irected to both the antigenic epitopes A and B, three sera from patients with endometriosis were selected from the larger pool with respect to strong reactions with the a-enolase. As shown in Figure 7, all three patients had developed autoantibodies which recognized the autoreactive epitope A (Figure 7B, C, D, lane 2), whereas only one of them additionally reacted with the autoreactive epitope B (Figure 7C, lane 3).

940

M. Walter

et al.

I 440 amino acids

Figure 5. Mapping of two linear autoreactive epitopes of human a-enolase by immunoscreening. The nucleotide sequence of 29 clones positively identified within the a-enolase epitope cDNA library with a serum from a patient with endometriosis were analysed by DNA sequencing and lined up with the full-length a-enolase cDNA. The two stretches of cDNA between positions 157 and 262, as well as 619 and 714 coding for 35 and 32 amino acids shown as shadowed boxes contained the sequences for two linear autoantigenic epitopes of the human a-enolase.

These data, confirmed by studying 12 patients’ sera (data not shown), indicated that the autoreactive epitope A may represent a predominant antigenic epitope of the a-enolase, and therefore may be useful for the diagnosis of endometriosis.

Discussion

The pathophysiology of endometriosis and endometriosis-associated reproductive failure has often been linked with abnormalities in immunological mechanisms such as an autoimmune response. Weed and Arquembourg [30] first suggested that autoimmune reaction may be one of the major causes of endometriosis. More recently, several reports, using a variety of immunoassay systems, seemed to strengthen the correlation between autoimmunity and endometriosis by demonstrating that patients with endometriosis have significantly high levels of antibodies directed against endometrial tissue and ectopic endometrial cells in the peritoneal fluid and serum [12-18, 311. However, an actual association between endometriosis and autoimmunity is still not proven and needs further investigation. The aetiology and nature of the disease-specific autoantibodies are still not completely understood. They may help to trigger macrophages for clearing the peritoneal cavity of ectopic tissue, and secondarily target the intrauterine endometrium causing infertility, decreased oocyte fertilization and/or implantation, decreased pregnancy rates in the in vitro fertilization process, and increased pregnancy loss [32]. Healthy and fertile women without endometriosis exhibit no circulating or peritoneal antiendometrial antibodies. It is now believed that antiendometrial antibodies

Endometriosis-specific

autoantigenic

epitopes

within

the human

a-enolase

An;TCTATTCTCMGATCCATGCCAGGGAGATCTTTOACTCTCGCGGGMTCCCACTGTT HSILXIEAREIPDSRGNPT”

941 60 20

GAGGTTGATCTCTTCACCTCAAAAGGTCTCTTCAGAGCTGCTGTGCCCAGTGGTGCTTCA BVDLFTSXGLFRAAVPSGAS

autoreactive

120 40

region

A (aa #53-87)

GATAAGACTCGCTATATGGGGAAG D X T R Y M

G

X

180 60

GGTGTCTCAAAGGCTGTTGAGCACATCAATAAAACTATTGCGCCTGCCCTGGTTAGCAAG GVSXAVEEINXTIAPALVSX

240 80

AAACTGAACGTCACAGAAC XLNVTE

300 100

GAGAAGATTGACAAACTGATGATCGAGATGGATGGAACA

GAAAATAAATCTAAGTTTGGTGCGAACGCCATTCTGGGGGTGTCCCTTGCCGTCTGCneA ENXSXFGANAILGVSLAVCX

360 120

GCTGGTGCCGTTGAGMGGGGGTCCCCCTGTACCGCCACATCGCTGACTTGGCTGGCAAC AGAVEXGVPLYREIADLAGB

420 140

TCTGMGTCATCCTGCCAGTCCCGGCGTTCMTGTCATCMTGGCGGTTCTCATGCTOGC SEVILPVPAFNVINGGEEAG

480 160

AACAAGCTGGCCATGCAGGAGTTCATGATCCTCCCAGTCGGTGCAGCAAACTTCAGGGM NXLAMQEFMILPVGAANPRE

540 180

GCCATGCGCATTGGAGCAGAGGTTTACCACAACCTGAADAATGTCATCAAGGAGAAATAT AMRIGAEVYENLXNVIXEXY

600 200

autoreactive

region

B (aa #207-238)

GGGAMGATGCCACCAA~GTGGGGGATGAAGGCGGGTTTGCTCCCAACATCCTGGAGAATl GXDATNVGDEGGFAPNILEN

660 220

AAAGAAGGCCTGGAGCTGCn:AAGACTGCTATTGGGAAAGCTGGCTACACTGAT XEGLXLLKTAIGKAGYTD

720 240

GTCATCGGCATGGACGTAGCGGCCTCCGAGTTCTCCAGGTCTGGGAAGTATGACt3!GGAC VIGMDVAASEFSRSGXYDLD

780 260

Tn:AAGTCTCCCOATOACCCCAGCAGGTACATCTCGCCTOACCAGCTGaCTOACCTGTA= FXSPDDPSRYISPDQLADLY

840 280

AAGTCCTTCATCAAGGACTACCCAGTGGTGTCTATCGAAGATCCCTTTGACCAGGATGAC XSFIXDYPVVSIEDPFDQDD

900 300

TGGGGAGCTTGGC!AGAAGTTCACAGCCAGTGCAGGAATCCAGGTAGTGGGGGATGATC,X! WGAWQXFTASAGIQVVGDDL

960 320

ACAGTGACCAACCCAAAGAGGATCGCCMGGCCGTGAACGAGAAGTCCTGCAACTGCCTC TVTNPKRIAXAVNEXSCNCL

1020 340

CTGCTCAAAGTCAACCAGATTGGCTCCGTGACCGAGTCTCTTCAGGCGTGCAAOCTGGCC LLXVNQIGSVTESLQACXLA

1080 360

CAGGCCAATGGTTGGGGCGTCATGGTGTCTCATCGTTCGGGGGAGACTOAAGATACCTTC QANGWGVMVSHRSGETEDTF

1140 380

ATCGCTGACCTGGTTGTGGGGCTGTGCACTGGGCAGATCAAGACTGGTGC.CCCTTGCCGA IADLVVGLCTGQIXTGAPCR

1200 400

TCTGAGCGCTTGGCCZAGTACAACCAGCTCCTCAGAATTGAAGRGWIGCTGGGCAOCAAG SERLAXYNQLLRIEEELGSX

1260 420

GCT AAG TTT GCC GGC AGG MC AXFAGRNFRNPLAK*

1305 434

Figure 6. Nucleotide autoreactive epitopes. boxed.

TTC AGA MC

CCC TTG GCC AAG TM

and amino acid sequence The investigated sequences

of human coding for

a-enolase cDNA the autoreactive

with the investigated domains A and B are

are potential serological markers for endometriosis, and therefore considered useful for diagnosis of the development of endometriosis as well as monitoring the effectiveness of various therapies.

942

M. Walter

et al.

A kDa

94.0 67.0 43.0 -

.-c;x.

B

C

D

kDa 60.0 -

49.5 -

325- -G 27.5 -

-

30.0 15.5 -

w

.e-

.

123

123

123

Figure 7. Western blot analysis of the recombinant autoreactive epitopes with sera from patients with endometriosis. Aliquots representing about 50 ul bacterial culture of E. coli LE392 transformed with pGEX-5T (lane 1) and plasmid constructs expressing the autoreactive epitope A (lane 2) and B as GST fusion proteins (lane 3) and induced with 1 mM IPTG for 5 h were separated by 12% SDS-PAGE (A) and analysed by Western blotting using three selected sera from patients with endometriosis (B-D).

The aim of the present study was to identify and characterize an endometriosisspecific antigen of the endometrium; various protein extracts derived from human tissues and cell cultures such as endometrium, ectopic endometrial tissue, and the endometrial adenocarcinoma cell lines HEC-1B (ATCC HTB 113), AN3-CA (ATCC HTB 11 l), and RL95-2 (ATCC CRL 1671) were analysed by traditional Western blot technique using sera from patients with clinically staged endometriosis and from healthy female blood donors. Interestingly, only the protein extract from HEC-1B cells demonstrated the most intense immunoreaction with the patients’ sera, and therefore was selected for further experiments (data not shown). Among the heterogeneous pattern of proteins with apparent molecular masses in the range between 30 and 130 kDa, a unique 48 kDa protein was observed which was significantly recognized by the sera from endometriosis patients (Figure l), and thus may represent a disease-related marker antigen. Similar data were published by Rajkumar et al. [17] and Gorai et al. [21] who identified a 34 kDa and a set of 26, 34, and 42 kDa proteins, respectively. The 48 kDa antigen was further purified and characterized by protein sequence analysis, and identified as the ubiquitous glycolytic enzyme a-enolase (EC 4.2.1.1.). The enolase exists intracellularly as a dimer composed of two equal subunits of 48 kDa. At least three highly homologous isozymes, a, p and y, are known, whereas the a isoform is found in most of the tissues [33]. Testing the other isoforms, we found no immunological reaction to the sera from patients with endometriosis (data not shown). Despite the high degree of homology among the three isoforms of enolase, it was suggested that the endometriosisspecific antibodies must recognize unique epitope(s) on this antigen. The role and function of the a-enolase as an antigen in autoimmune mechanisms which accompanied the aetiology and pathogenesis of endometriosis and the

Endometriosis-specific

autoantigenic

epitopes

within

the human

a-enolase

943

disease-associated infertility still remains unclear and needs further investigation. The possible importance of a-enolase in autoimmune events was supported by results of Moodie et al. [34], who also found anti-enolase antibodies in about 40% of sera from patients with vasculitis and lupus nephritis positively tested for circulating anti-neutrophil cytoplasmic antibodies (ANCA). However, any correlation between the pathogenesis of endometriosis and vasculitis mediated by a-enolase seems still completely hypothetical. With a primary aim of developing a non-surgical diagnosis of endometriosis by immunological measurement of anti-enolase antibodies in sera from endometriosis patients, we followed standard strategies. First, the full-length cDNA coding for the human a-enolase was isolated by PCR amplification of recombinant DNA derived from a human endometrial cDNA library. Second, the a-enolase cDNA was cloned in the high-efficiency expression vector pH6EX3 [24] which directed the synthesis of a recombinant fusion protein with a histidine-hexapeptide at its N-terminus and the full-length human a-enolase at its C-terminus under the control of the strong tat promoter [25]. The recombinant plasmid was used to transform the E. coli strain LE392 [26] prior to induction with IPTG. In contrast to the high-level expression of many other human autoantigens, the recombinant human a-enolase was not completely sequestered in insoluble inclusion bodies, but remained partially in the soluble cell fraction (Figure 4). Third, the recombinant a-enolase was purified to almost homogeneity from the total cell lysate by single-step affinity chromatography based on the specific interaction between the histidine-hexapeptide and matrix-immobilized nickel ions [24]. No interference of the N-terminally fused histidine-hexapeptide on autoantigenicity was observed. The identity of the recombinant a-enolase and the primarily isolated 48 kDa antigen was confirmed by immunological competition assays using patients’ sera (data not shown). Fourth, the significance of the diagnostic value of the purified recombinant human a-enolase was preliminarily evaluated in comparable Western blot analysis, whereas 5 1% of the clinically confirmed patients with endometriosis and 6% of healthy female blood donors in reproductive ages could be positively assayed for endometriosis-specific anti-enolase antibodies. A survey of larger serum collection such as from patients with endometriosis, from patients with other and/or related gynaecological disorders, and from healthy females staged negatively by laparoscopy is underway. With established mapping techniques [28], we were able to localize two regions of the human a-enolase of 35 and 32 amino acids in length carrying linear autoreactive epitopes which were significantly recognized by sera from endometriosis patients (Figures 5 and 6). Western blot analysis of both autoantigenic epitopes A and B expressed as recombinant fusion proteins with N-terminally fused GST with selected pathological sera demonstrated that the majority of the endometriosis patients had developed autoantibodies which reacted only with the autoreactive epitope A (Figure 7). Therefore, the epitope A was considered to be a predominant autoantigenic epitope. Future efforts based on a larger collection of sera from patients with endometriosis will preferentially focus on the fine mapping of this autoreactive epitope by corresponding blocking and binding experiments with synthetic peptides.

944

M. Waker et al. Acknowledgements

The authors would like to thank Drs N. J. Cook, H. Haubruck, B. Liedvogel, L. Mauch, T. Simon, G. Stahnke (ELIAS Entwickhmgslabor) and N. Hunt (Institute of Hormone and Fertility Research, University of Hamburg) for their helpful discussion and advice, as well as for critical reading of the manuscript. References 1. Homm, R. J. and S. Mathur. 1988. Autoimmune factors in endometriosis: Cause or Effect? Sent. Reprod. Endocrinol. 6: 279-284 2. Roennberg, L. 1990. Endometriosis and infertility. Ann. Med. 22: 91-96 3. Gleicher, N. 1992. Endometriosis: a new approach is needed. Hum. Reprod. 7: 821-824 4. Thomas, E. J. 1993. Endometriosis: Still an enigma. Brit. 3. Obstet. Gynecol. 100: 615-617 5. Olive, D. L. and L. B. Schwartz. 1993. Endometriosis. New Eng. 3. Med. 328: 1759-1769 6. Mangtani, P. and M. Booth. 1993. Epidemiology of endometriosis. 3. Epidemol. Commun. Health 41: 84-88

7. Steele, R. W., W. P. Dmowski, and D. J. Marmer. 1984. Immunologic aspects of human endometriosis. Am. 3. Reprod. 6: 33-36 8. Oosterlynck, D. J., F. J. Comille, M. Waer, M. Vandeputte, and P. R. Koninchx. 1991. Women with endometriosis show a defect in natural killer activity resulting in a decrease cytotoxicity to autologous endometrium. FertiZ. SteriZ. 56: 45-51 9. Wilson, T. J., P. J. Hertzog, D. Angus, L. Munnery, E. C. Wood, and I. Kola. 1994. Decreased natural killer cell activity in endometriosis patients: Relationship to disease pathogenesis. Fercil. Stertl. 62: 1086-1088 10. Halme, J., S. Becker, and S. Haskill. 1987. Altered maturation and function of peritoneal macrophages: Possible role in pathogenesis of endometriosis. Am. 3. Obstet. Gynecol. 156: 783-790 11. Sharpe, K. L., R. L. Zimmer, W. T. Griffin, and L. L. Penney. 1993. Polypeptides synthesized and released by human endometriosis differ from those of the uterine endometrium in cell and tissue explant culture. FernI. SteriZ. 60: 839-851 12. Wild, R. A., C. A. Shivers, and D. Medders. 1992. Detection of antiendometrial antibodies in patients with endometriosis: methodological issues. Fertil. StenZ 58: 518-521 13. Mathur, S., H. J. Chihal, R. J. Homm, D. E. Garza, P. F. Rust, and H. 0. Williamson. 1988. Endometrial antigens involved in the autoimmunity of endometriosis. Fe&Z. SteriZ. 50: 860-863 14. Badaway, S. Z. A., V. Cuenca, H. Freliech, and C. Stefanu. 1990. Endometrial antibodies in serum and peritoneal fluid of infertile patients with and without endometriosis. Fail. Steri’l. 53: 930-932 15. Kennedy, S. H., P. M. Starkey, I. L. Sargent, B. R. Hicks, and D. H. Barlow. 1990. Antiendometrial antibodies in endometriosis measured by an enzyme-linked immunosorbent assaybefore and after treatment with danazol and nafarelin. Obstet. Gynecol. 75: 914-91s 16. Taylor, I’. V., M. D. Maloney, J. M. Campbell, S. M. Skerrow, M. C. Nip, R. Parmar, and G. Tate. 199 1. Autoreactivity in women with endometriosis. Brit. 3. Obstet. Gynecol. 98: 680-684 17. Rajkumar, K., V. Malliah, and C. W. Simpson. 1992. Identifying the presence of antibodies against endometrial antigens. A preliminary study. 3. Reprod. Med. 37: 552-556 18. Femandez-Shaw, S., B. R. Hicks, P. L. Yudkin, S. Kennedy, D. H. Barlow, and P. M. Starkey. 1993. Anti-endometrial and anti-endothelial auto-antibodies in women with endometriosis. Hum. Reprod. 8: 3 lo-31 5

Endometriosis-specific

autoantigenic

epitopes

within

the human

a-enolase

945

19. Kennedy, S. H., B. Nun, S. A. Cederholm-Williams, and D. H. Barlow. 1989. Cardiolipin antibody level in endometriosis and systemic erythematosus. Fertil. St&. 52: 1061-1062 20. Gleicher, N., A. El-Roeiy, E. Co&no, and J. Friberg. 1989. Reproduction failure because of autoantibodies: Unexplained infertility and pregnancy wastage. Am. J. Obstet. Gynecol. 160: 1376-1385 21. Gorai, I., M. Ishikawa, R. Onose, F. Hirahara, and H. Minaguchi. 1993. Antiendometrial autoantibodies are generated in patients with endometriosis. Am. J. Reprod. Zmnzunol. 29: 116-123 22. ASF 1985. Revised American Fertility Society classification of endometriosis: 1985. Fertil. Stenl. 43: 351 23. Giallongo, A., S. Feo, R. Moore, C. M. Croce, and L. C. Showe. 1986. Molecular cloning and nucleotide sequence of a full-length cDNA for human alpha enolase. Proc. Natl. Acad. Sci. USA 83: 6741-6745 24. Berthold, H., M. Scanarini, C. C. Abney, B. Frorath, and W. Northernann. 1992. Purification of recombinant antigenic epitopes of the human 68-kDa (Ul) ribonucleoprotein antigen using the expression system pH6EX3 followed by metal chelating affinity chromatography. Prot. Expr. Puurif. 3: 50-56 25. Amann, E., J. Brosius, and M. Ptashne. 1983. Vectors bearing a hybrid trylac promoter useful for regulated expression of cloned genes in Escherichia coli. Gene 25: 167-178 26. Borck, K., J. D. Beggs, W. J. Brammar, A. S. Hopkins, and N. J. Murray. 1976. The construction in vitro of transducing derivates of phage lambda. Mol. Gen. Genet. 146: 199-205 27. Berg, H., M. Walter, L. Mauch, J. Seissler, and W. Northemann. 1993. Recombinant human proinsulin: Expression, purification and reaction with insulin autoantibodies in sera horn patients with insulin-dependent diabetes mellitus. J. Immunol. Methods 164: 221-231 28. Frorath, B., C. C. Abney, M. Scanarini, H. Berthold, N. Hunt, and W. Northernann. 1992. Mapping of a linear autoantigenic epitope within the human thyroid peroxidase using recombinant DNA techniques. 3. Biochem. (Tokyo) 111: 633-637 29. Berthold, H., B. Frorath, M. Scanarini, C. C. Abney, B. Ernst, and W. Northernann. 1992. pGEX-ST: An alternative system for expression and purification of recombinant proteins. Biotechnol. Letters 14: 245-250 30. Wood, J. C. and P. C. Arquembourg. 1980. Endometriosis: Can it produce an autoimmune response resulting in infertility? C&z. Obstet. Gynecol. 23: 885-893 3 1. Garza, D., S. Mathur, M. M. Dowd, L. F. Smith, and H. 0. Williamson. 1991. Antigenic differences between the endometrium of women with and without endometriosis. 3. Reprod. Medicine 36: 177-182 32. Damewood, M. D., J. S. Helsa, W. D. Schlaff, M. Hubbart, Jr D. Gearhart, and J. A. Rock. 1990. Effect of serum from patients with minimal to mild endometriosis on mouse embryo development in vitro. Fertil. SteriZ. 54: 917-920 33. McAleese, S. M., B. Dunbar, J. E. Fothergill, L. J. Hinks, and I. N. Day. 1988. Complete ammo acid sequence of the neurone-specific and isoform of inolase (NSE) from human brain and comparison with the non-neuronal a form (NNE). Eur. 3. Biochem. 178: 413-417 34. Moodie, F. D., B. Leaker, G. Cambridge, N. F. Totty, and A. W. Segal. 1993. Alpha-enolase: A novel cytosolic autoantigen in ANCA positive vasculitis. Kidney Znt. 43: 675-681