Journal of Autoimmunity (2000) 15, 441–449doi:10.1006/jaut.2000.0452, available online at http://www.idealibrary.com on

Occurrence of IgA and IgG Autoantibodies to Calreticulinin Coeliac Disease and Various Autoimmune Diseases

D. Sanchez1, L. Tuckova1, P. Sebo1, M. Michalak2, A. Whelan3, I. Sterzl4,L. Jelınkova1, E. Havrdova5, M. Imramovska5, Z. Benes6, S. Krupickova1 andH. Tlaskalova-Hogenova1

1Department of Immunology, Institute ofMicrobiology, Czech Academy ofSciences, Prague, Czech Republic2Department of Biochemistry, Universityof Alberta, Edmonton, Canada3Trinity College, St James’s Hospital,Dublin, Ireland4Institute of Endocrinology, Prague,Czech Republic51st Medical Faculty, Charles University,Prague, Czech Republic6Thomayer’s Hospital, Prague, CzechRepublic

Received 16 June 2000Accepted 20 September 2000

Key words: coeliac disease,calreticulin, autoantibodies,autoimmune diseases

Calreticulin (CRT), a high-affintiy calcium binding protein and chaperone, wasrecently identified as one of the targets of autoantibodies in coeliac disease. Weevaluated the level of IgA and IgG antibodies to CRT in sera from patientswith coeliac disease and various autoimmune diseases. The level of antibodiesto gliadin (shown previously to cross-react with CTR), isolated enterocytesand tissue transglutaminase were determined for comparison. The mean levelof IgA antibodies to CRT was significantly higher (P<0.001) in sera fromcoeliac patients with active disease (139.9±11.2 AU/±SE) than in healthycontrols (20.9±1.7 AU). In sera of patients with systemic lupus erythematosus(SLE), insulin dependent diabetes mellitus (IDDM), multiple sclerosis (MS)and autoimmune thyroiditis (AT) or inflammatory bowel disease (IBD) themean level (25.8±3.7 to 38.1±5.6 AU) did not exceed the cut-off value. A lowlevel of these antibodies, however, was detected in some sera of patients withMS and IBD. The level of IgG anti-CRT antibodies was increased in coeliacpatients (mean 125.4±8.0 AU, P<0.001) when compared to that in healthycontrols (33.9±2.3 AU). The IgG anti-CRT antibodies were also detected inabout 30% of SLE patients sera (54.1±3.6 AU, P<0.001), but the mean levelreached only half that detected in coeliac patients. © 2000 Academic Press

Correspondence to: Dr L Tuckova, Department of Immunology,Institute of Microbiology, Czech Academy of Sciences, Vıdenska1083, 142 20 Prague 4, Czech Republic. Fax: 02-4721143. E-mail:tuckova@biomed.cas.cz

Introduction

Calreticulin (CRT) is a soluble Ca2+-binding, regulat-ory protein and chaperone (glu1, man5-binding) whichis present in a wide variety of cells [1–5]. This proteinis capable of interacting with various molecules suchas the �-subunit of integrins, steroid/nuclear hormonereceptors, vitamin D receptor, rubella virus RNA,perforin in cytolytic granules, Ro/SS-A polypeptideand (because of the structural homology with C1qreceptor) also with C1q, as well as modulating theirfunctions [4–10]. CRT is localized mainly in endo-plasmic reticulum, but is also found in other cellcompartments including cell membranes, and alsoextracellularly, as a consequence of infection, stressand cell death [3, 11–13]. When released to the extra-cellular environment, it could become clinicallyrelevant. CRT was determined to be one of the autoan-tigens in systemic lupus erythematosus (SLE) [14, 15],

4410896–8411/00/080441+09 $35.00/0

Sjogren’s syndrome, rheumatoid arthritis, completecongenital heart block [16–18], and in some infectiousdiseases (Schistosoma mansoni, Onchocerca volvus andPlasmodium falciparum) [19–20]. However, variousmethods and CRT of different origin were used foranti-CRT antibody detection in these diseases.

In a pilot study we found elevated levels of IgAautoantibodies against calreticulin in patients withcoeliac disease [21, 22]. Coeliac disease, a gluten-sensitive enteropathy, is induced in genetically sus-ceptible individuals by wheat gluten or relatedprolamines. This disease shares many features charac-teristic of autoimmune diseases: strong genetic associ-ation with MHC class II (HLA-DQ2 and HLA-DQ8)antigens, higher occurrence in females than in males,frequent association with other autoimmune diseases[such as insulin dependent diabetes mellitus (IDDM),thyroiditis, rheumatoid arthritis] and presence ofcirculating autoantibodies [23–25]. The sera fromcoeliac patients contain high levels of IgA and IgGantibodies (Abs) to gliadin and IgA autoanti-bodies against endomysial (EMA), reticulin (ARA),jejunal and enterocyte antigens [25–28]. The endo-mysial antigen has recently been identified as tissue

© 2000 Academic Press

442 D. Sanchez et al.

transglutaminase (tTG) [29]. We have reported thereactivity of anti-gliadin Abs with structures on ratand human enterocytes [22, 30, 31]. Cross-reactiveenterocyte proteins were isolated by affinity chroma-tography (using anti-gliadin Abs), and the first entero-cyte antigen was identified as calreticulin [21, 22].Furthermore, similar structures recognized by isolatedpatients IgA anti-gliadin Abs on gliadin, enterocytesand calreticulin were identified by inhibitory enzyme-linked immunoassay (ELISA, using 40 syntheticdodecapetides of �-gliadin) and the binding specificityof one of the inhibitory peptides (EQVPLVQQQQF)was determined [32].

The aim of our study was to perform a reproduci-bile ELISA assay for CRT autoantibodies, using puri-fied recombinant human CRT. We used this method toestimate the level of IgA and IgG antibodies to CRT insera from patients with coeliac disease and otherautoimmune diseases, such as systemic lupus ery-thematosus (SLE), IDDM, multiple sclerosis (MS), andautoimmune thyroiditis and with inflammatory boweldisease (IBD—ulcerative colitis and Crohn’s disease).To complete the study, we determined the level of Absto gliadin, tissue transglutaminase and enterocyteantigens, markers of coeliac disease in the same sets ofsera. The specificity of the binding of patients’ serawith CRT was verified by Western blotting.

Material and Methods

Patients and healthy controls

The sera tested were from 75 adults (mean age 42,range 20–60 years) and 11 children (mean 11, range3–13 years) with active coeliac disease and from 20child patients on a gluten-free diet (GFD), all diag-nosed on the basis of ESPGAN criteria [33]. Sera weretaken at the time of first small bowel biopsy androutinely tested for EMA. EMA-positive sera of thesepatients with positive biopsy findings were used inthe study. Sera from 20 child patients on GFD wereobtained after a 10–12-month period on GFD and werefound to be EMA-negative. Sera from 19 patients withulcerative colitis (five treated with hydrocortisoneand/or prednisone) and 15 sera from Crohn’s diseasepatients (eight treated with hydrocortisone and/orprednisone) were combined in an inflammatory boweldisease group (34 patients in all, age range 19–75years, mean age 36.7 years). Sera obtained frompatients with autoimmune diseases included 44patients with IDDM (mean age 23 years, range 18–68years), 46 patients with MS (mean age 34, range 17–45years, all treated with IFN-�), 75 patients with sys-temic lupus erythematosus (mean age 46, age range18–61 years, six treated with immunosuppressivedrugs, the others with corticosteroids—at the time ofblood sampling 10 of them were without drugs) and29 patients with autoimmune thyroiditis (mean age 37,age range 21–56 years, all positive for thyroid peroxi-dase Abs and treated with thyroxin, in euthyroid

stage). Sera from 59 healthy adults without clinicalsigns of disease (mean age 24, range 21–38 years) and16 healthy children (mean age 12, range 5–17 years)were used as control. Sera of patients with combineddisorders, autoimmune disease and coeliac diseasewere excluded from the tested groups.

Preparation of recombinant human calreticulin

Calreticulin was expressed as glutathione-S-transferase (GST) fusion protein from pGEX-CRT 3Xplasmid in Escherichia coli BNN103. One-litter cultureswere grown in LB medium containing ampicillin(150 �g/ml) at 37°C to A600=0.4 and the expression ofGST-CRT fusion protein was induced by addingisopropyl-1-thio-�-D-galactopyranoside (IPTG) to afinal concentration of 1 mM. The culture was shakenat 37°C for 4 h and then rapidly chilled and centri-fuged for 20 min (2000×g, 4°C). The pellet wasresuspended in PBS containing 5 mM EDTA and1 mM PMSF (PBS/EDTA/PMSF) and the bacteriawere sonicated using a probe-tip sonicator. The fusionprotein was purified by affinity chromatographyon glutathione-Sepharose 4B (Pharmacia Biotech,Sweden) column. After equilibration of the columnwith 50 mM Tris buffer pH 8.0 containing 10 mMreduced glutathione (Sigma, Germany) and subse-quent washing with PBS/EDTA/PMSF, the bacterialsonicate was loaded overnight at 4°C. The columnwas then washed with PBS/EDTA and bound pro-teins were eluted by Tris-glutathione buffer. PurifiedGST-CRT was digested by the endopeptidase FactorXa (Sigma-Aldrich, USA) and recombinant CRT wasisolated by chromatography as previously described[34–36].

Isolation of intestinal epithelial cells

Enterocytes were isolated from the jejunum of 2–3-month-old Wistar rats. Intestinal epithelial cells wereisolated as described by Raul et al. [37] and Kolınskaet al. [38]. Briefly, the cells were detached from the gutin sequential populations by gently shaking theeverted jejunum in phosphate buffered saline (PBS)with addition of 1.5 mM Na2EDTA and 0.5 mM DL-dithiothreitol at 37°C. Ca2+ and Mg2+ were omitted.The suspension of isolated cells was gently centri-fuged at 4°C, 900×g for 5 min, washed three timesin PBS and diluted to a final concentration of5×105 cells/ml. After fixation with 0.25% glutaralde-hyde, plates with enterocytes were stored in PBScontaining 1% bovine serum albumin (BSA) and 0.1%NaN3.

Estimation of antibodies to gliadin, tissuetransglutaminase, endomysium, enterocytes andcalreticulin

ELISA for determination of IgA and IgG anti-calreticulin and IgA anti-gliadin Abs and anti-enterocyte Abs was performed according to Tuckova

Anti-calreticulin Abs in autoimmune diseases 443

et al. [31]. IgA antibodies to tTG were detected by amodified method described earlier [29]. Briefly, thefollowing antigens: human recombinant GST-CRT,purified recombinant CRT, guinea-pig liver tTG(Sigma-Aldrich) and crude gliadin (Sigma) were usedin a final concentration of 5 �l/ml in PBS. The 96-wellpolystyrene plates (Gama, Ceske Budejovice, CzechRepublic and Nunc-MaxiSorp plates, Denmark—inthe case of tTG), were coated overnight at 4°C. Ratenterocytes were used at a concentration of5×105 cells/ml, sedimented cells were fixed with0.25% glutaraldehyde. Blocking solution—1% BSA inPBS (Sevac, Czech Republic) was also used as anegative control. Each dilution of standard sera orsamples was used in triplicate and incubated for 2 h.After incubation, the plates were washed with PBSand PBS-0.1% Tween (PBS-T) and then peroxidase-labelled swine anti-human IgA or IgG (Px-SWAHu/IgA or Px-SWAHu/IgG, Sevac) Ab, diluted 1:750 in10% normal pig serum in PBS were added to thewells. After 1-h incubation at room temperaturethe plates were washed with PBS and PBS-T and theenzyme reaction was developed by adding a solutioncontaining 1,2-o-phehylenediamine (0.7 mg/ml) in0.1 M phosphate buffer (pH 6.0) containing 0.06%H2O2, or in 0.1 M sodium citrate (pH 4.2) in the case ofanti-tTG and anti-CRT assay. The reaction wasstopped by 2 M H2SO4 and optical density was readon a spectrophotometer at 492 nm (Titertek MultiscanMCC/340, Flow Lab, Irvine, UK). The internal labora-tory standard (reference serum) was prepared frompooled positive coeliac patients’ sera and used in allELISA tests.

An immunofluorescent method using humanumbilical cord tissue sections was used for determi-nation of anti-endomysial IgA antibodies [39].

SDS-PAGE and ECL Western blot analysis

SDS-PAGE was performed as described earlier [40].Cell lysate was diluted in sample buffer, boiled for5 min at 100°C, loaded into 10% polyacrylamide geland subjected to SDS-PAGE under non-reducing con-ditions. To prevent denaturation CRT samples werenot boiled before use. Separated proteins were trans-ferred to nitrocellulose membranes (Hybond-C pure,Amersham International, UK) in 96 mM glycine,12.5 mM Tris buffer containing 20% methanol [41]using the Transblot system (Biorad, Richmond, CA,USA) overnight at 40 mA. The nitrocellulose was cutinto strips and blocked with 4% non-fat powderedmilk in PBS-T for 2 h at room temperature and thenincubated with patients’ or control sera diluted 1:100in the blocking solution. After washing with PBS andPBS-T, the secondary antibody—pig anti-human IgA-�-chain (The Binding Site, UK) diluted 1:2000—in theblocking solution was added. After 1-h incubation andrepeated washing, ECL reagents (Amersham, UK) andautoradiography (film X-Omat RA, Kodak, USA) wereused for detection.

Statistical methods

The results are presented as non-parametric data [42];antibody levels are not directly proportional to theantigen-binding capacity of the sample. The Mann–Whitney test was used for comparison of antibodylevels between groups. Correlation analysis was usedto evaluate the relation between serum level of IgAAbs to human recombinant calreticulin, gliadin, tissuetransglutaminase and enterocyte antigens.

The optical density of reference serum was taken as100% (100 arbitrary units, AU). The level of Abs wascalculated as a percentage of the reference serumsample, expressed in AU±standard error (SE). Allpatients’ sera were tested in triplicates in at least twoindependent experiments. The results of estimation ofAb level in 59 healthy adult donor sera were used tocalculate cut off values, defined as the mean valueplus 2 SD.

Results

Figure 1. SDS-PAGE analysis of E. coli lysate enriched byexpressed GST-calreticulin (lane 1), purified recombinantGST-calreticulin (lane 2), and calreticulin (lane 3). Proteinswere separated on 10% polyacrylamide gel and stained withServa Blue R.

Screening of IgA autoantibodies to calreticulin

The occurrence of antibodies to CRT was tested in serafrom seven groups of adult patients with autoimmunediseases: SLE, IDDM, MS, autoimmune thyroiditis(AT), diseases affecting gut mucosa, i.e. coeliac diseaseand IBD and healthy controls. Isolated recombinanthuman GST-CRT and CRT were prepared as a fusionprotein in E. coli (Figure 1); both forms (previouslyshown to yield comparative results) were used as anantigen in ELISA.

444 D. Sanchez et al.

Individual values of IgA anti-CRT Ab in sera fromhealthy controls (59 adults) and from each group ofpatients are shown in Figure 2. The level of anti-CRTAbs reached the highest value in the group of coeliacpatients (75 adults, all biopsy proven and EMA-positive). The mean value of these antibodies in serafrom patients with active coeliac disease (139.9±11.2 AU/±SE) was significantly higher (P<0.001) thanin healthy controls (20.9±1.7 AU). In all other groupstested the mean value of IgA antibodies did not exceedthe cut off value (46.0 AU, mean+2 SD; ranging from25.8±3.7 to 38.1±5.6 AU). A low level of IgA anti-CRT Ab was, however, detected in sera from somepatients mainly with IBD and MS (P<0.01 and 0.05respectively) as documented in Figure 2 and Table 1.

For comparison, IgA Abs against tTG, gliadin andisolated enterocytes, serological markers of coeliacdisease, were estimated in the same sets of seraobtained from adult patients and healthy controls.Figure 3 shows significantly elevated mean levelsof Abs to tTG, gliadin and isolated enterocytes

(111.5±9.8, 86.4±5.2, 129.7±5.7) in sera from coeliacpatients, when compared to those in healthy controls(10.7±1.1, 10.4±1.0, 35.9±2.1) (P<0.001). The presenceof a low level of IgA Abs to gliadin was detected insome sera from patients with MS (P<0.05) and IgAAbs against tTG in patients with SLE (P<0.05). TheAbs to isolated enterocytes were enhanced in somesera of patients suffering from AT, IBD (P<0.001) andSLE (P<0.01) (Figure 3, Table 1).

The effect of treatment of coeliac disease by GFD onthe level of IgA anti-CRT antibodies was analysed bycomparing the Ab level in children with active coeliacdisease with those on a diet (11 newly diagnosed and20 treated children) and with healthy controls (16children). The GFD resulted in a substantial disease ofIgA anti-CRT Abs as well as Abs to other antigenstested (Table 2).

Screening of IgG antibodies to calreticulin

The IgG anti-CRT antibodies were tested in sera fromall groups of adult patients with various autoimmunediseases. A significantly higher level of IgG anti-CRT Abs was detected in coeliac patients (mean125.4±8.0 AU) (P<0.001) when compared with adulthealthy controls (mean 33.9±2.3 AU). Levels exceed-ing the cut off value were detected in about 30% ofsera from patients with SLE (mean value 54.1±3.6,ranging from 1.7 to 188.9 AU). Low levels, onlyslightly above the cut off value, were found in sera ofpatients with other diseases (Figure 4).

The sensitivity and specificity of IgA and IgG anti-CRT antibody estimation was calculated for twogroups of patients i.e. those with coeliac disease(n=75) and SLE (n=75) with statistically significantincreased mean levels of Abs, and for healthy controlsas documented in Table 3.

0

400A

U (

IgA

)

100

200

300

C CD SLE IDDM MS AT IBD

Figure 2. Scattergram of IgA anti-calreticulin Abs levels insera from adult patients with active coeliac disease (CD,n=75), systemic lupus erythematosus (SLE, n=75), insulindependent diabetes mellitus (IDDM, n=44), multiplesclerosis (MS, n=46), autoimmune thyroiditis (AT, n=29),inflammatory bowel disease (IBD, n=34) and adult healthycontrols (C, n=59). The IgA level is expressed in arbitraryunits (AU). Solid lines represent mean, values above thecut-off (46.5 AU) are considered positive.

Table 1. Comparison of anti-CRT*, anti-tTG**, anti-gliadin and anti-enterocyte Abs levels in sera of adultpatients with coeliac disease, and other diseases to healthy controls*** (Mann–Whitney test)

Antibodies Autoimmune diseases

Against Class CD(n=75)

SLE(n=75)

IDDM(n=44)

MS(n=46)

AT(n=29)

IBD(n=34)

Calreticulin IgA P<0.001 NS NS P<0.05 NS P<0.01Calreticulin IgG P<0.001 P<0.001 P<0.01 P<0.001 P<0.05 NStTG IgA P<0.001 P<0.05 NS NS NS NSGliadin IgA P<0.001 NS NS P<0.05 NS NSEnterocytes IgA P<0.001 P<0.01 NS NS P<0.001 P<0.001

*CRT (calreticulin), **tTG (tissue transglutaminase), ***59 adults.n=number of patients.

Western-blot analysis of patient’s sera reactivitywith human CRT

The specificity of binding of coeliac patient’s sera withCRT was confirmed by Western-blot analysis. Thepurification procedure of recombinant GST-CRT from

Anti-calreticulin Abs in autoimmune diseases 445

bacterial lysate by affinity chromatography and itsdigestion by factor Xa, resulting in recombinanthuman CRT, was shown on SDS-PAGE (Figure 1). Asdocumented in Figure 5 (lanes 1–4) the antibodies insera form two different coeliac patients (with theanti-CRT Ab level of about 300 AU) recognized

specifically both GST-CRT and CRT, while binding ofAbs with free GST and reactivity of sera from healthydonors were not observed (lane 5).

0

300A

U (

IgA

)

100

200

C CD SLE IDDM MS AT IBD

A

0

200

AU

(Ig

A)

100

C CD SLE IDDM MS AT IBD

B

0

300

AU

(Ig

A)

100

200

C CD SLE IDDM MS AT IBD

C

cut-off

Figure 3. Distribution of IgA Abs to tissue transglutaminase(A), gliadin (B) and rat enterocytes (C) in sera from adultpatients with coeliac disease (CD, n=75), systemic lupuserythematosus (SLE, n=75), insulin dependent diabetesmellitus (IDDM, n=44), multiple sclerosis (MS, n=46),autoimmune thyroiditis (AT, n=29), inflammatory boweldisease (IBD, n=34) and adult healthy controls (C, n=59).Values are expressed in AU. Solid lines represent the mean,and values above cut-off (which equals 25.3 AU for Abs totTG, to gliadin 24.3 AU and to enterocytes 67.5 AU) areconsidered positive.

Discussion

In the present study we have demonstrated that thelevel of both IgA and IgG anti-CRT Abs in coeliacpatients significantly exceed the values found inpatients with other diseases and in healthy controls. Inthe group of adult coeliac patients (with active form ofthe disease) we have detected three serum sampleswith increased level of anti-tTG Abs, but with a lowlevel of anti-gliadin Abs (below or equal to the cut-offvalue). In contrast, an increased level of anti-gliadinAbs but a low level of anti-tTG Abs was detected intwo patients’ sera. Interestingly, in all these serumsamples the level of anti-CRT Abs was significantlyincreased. Therefore we suggest that the estimation ofIgA Abs to CRT in coeliac patients could becomeuseful for diagnostic purposes, especially in cases ofthe discrepancy in the level of IgA anti-gliadin andanti-tTG Abs described above. Furthermore, casereports and also our recent pilot study concerning theanalysis of sera from coeliac patients who did notrespond to GFD (refractory sprue), showed anincreased level of IgA anti-CRT and anti-enterocyteAbs at the same time when the other serologicalmarkers of coeliac disease are missing [43]. Thisfinding indicated the possibility of using anti-CRTAbs detection as a diagnostic tool for this form ofcoeliac disease.

A low level of IgA anti-CRT Abs, only slightly abovethe cut-off value, was detected in some patients withIBD and MS. Interestingly, the low IgA response toCRT was detected in patients with digestive disordersand in patients with autoimmune diseases associatedwith coeliac disease [23–25]. In the study we havetested 232 sera of patients with various disorders, fourpatients (two with IDDM, one with AT and one withSLE) were diagnosed on the basis of this study also ascoeliac patients (due to the positivity of serologicalmarks including EMA). This means that the frequencyof silent coeliac cases was 2–4 times higher in thegroups of patients with studied autoimmune diseasesthan in normal population.

The elevated level of IgA and IgG anti-CRT and IgAanti-gliadin Abs in sera of patients with MS and IgAanti-CTR Abs in patients with IBD could be a conse-quence of increased intestinal permeability observedin patients suffering from these diseases [44, 45].

High prevalence of antibodies to CRT of the IgAclass was recently reported also in sera of patientswith primary biliary cirrhosis, alcoholic liver diseaseand in infection caused by Yersinia enterocolitica usingWestern-blot technique [46].

An increased level of IgG anti-CRT Abs was deter-mined in about 30% of SLE patients. Our results are inkeeping with previously reported data [9, 15] concern-ing IgG autoantibodies against CRT in sera with SLEpatients. Furthermore, these authors detected elevated

446 D. Sanchez et al.

Table 2. The levels of IgA antibodies in sera from children with active coeliac disease, those on a gluten freediet (GFD) and healthy controls

Samples from/numberIgA antibodies to:

Calreticulin Gliadin tTG Enterocytes

Coeliac patients/11 66.9±6.0 74.4±12.9 113.7±21.6 68.4±8.6Patients on GFD/20 14.2±3.7 7.9±1.8 7.5±1.2 36.1±2.6Healthy controls/16 22.5±3.5 9.0±1.7 12.5±2.1 27.4±2.8

Data are expressed as mean±SE.

0

400

AU

(Ig

G)

100

200

300

C CD SLE IDDM MS AT IBD

Figure 4. Scattergram of IgG Abs to calreticulin in sera fromadult patients with coeliac disease (CD, n=75), systemiclupus erythematosus (SLE, n=75), insulin dependent dia-betes mellitus (IDDM, n=44), multiple sclerosis (MS, n=46),autoimmune thyroiditis (AT, n=29), inflammatory boweldisease (IBD, n=34) and healthy controls (C, n=59). Valuesare expressed in AU (solid lines represent mean, cut off is69.4 AU).

Table 3. Specificity and sensitivity of IgA and IgG detection ofanti-calreticulin auto-Abs in adult coeliac and SLE patients

Disease/patient no. Sensitivity* Specificity**IgA IgG IgA IgG

Coeliac disease/75 92 92 — —SLE/75 11 25.3 — —Healthy controls/59 — — 85.5 79.7

*Percentage of positive (above cut-off) patients with active disease.**Percentage of negative (below cut-off) healthy controls.

Figure 5. Western blot analysis of reactivity of coeliacpatients’ and healthy control sera with calreticulin. SDS-PAGE separated proteins, GST-calreticulin (lanes 1, 2) beforeand after cleaving by factor Xa—(mixture of CRT andGST—lanes 3, 4, 5) were transferred to nitrocellulose, incu-bated with two different coeliac patients’ sera (lanes 1, 3 and2, 4) or control serum (lane 5) and the specific binding of IgAantibodies (in coeliac patients’ sera) was detected using ECLreagents and autoradiography.

level of CRT in sera, associated with immune com-plexes including C1q and also bound to the surfaceof neutrophils. The target sequences recognized byanti-CRT Abs of SLE patients were localized predomi-nantly to the N-region and the binding of C1q tothe N- and P-domain. This interaction with CRTmay interfere with the ability of C1q to associatewith immune complexes, possibly preventing theactivation of the complement cascade at the site ofinflammation [14, 15, 47–48].

We have recently shown that anti-gliadin antibodiesisolated from coeliac sera cross-react with human

CRT. The sequence similarities of gliadin antigenicepitopes and short stretches on CRT, recognized byAbs of coeliac patients, have been localized to thecalcium binding, proline rich, P-domain of humanCRT [22, 32]. The high level of anti-CRT Abs in coeliacpatients’ sera could be explained by the cross-reactionof anti-gliadin Abs with CRT. The immune response toCRT could be spread to CRT cryptic antigenic deter-minants. One could hypothesize that the mechanismof antigenic spreading and the response to new anti-genic determinants of CRT may play a role, especiallyin the severe form of coeliac disease, the refractorysprue, and could cause the elevation of anti-CRT Abseven after treatment with the gluten-free diet. Hencetesting the level of anti-CRT Abs could be also usefulto follow-up the course of the disease.

The up-regulation of CRT on the membrane ofendothelial or epithelial cells under stress conditions(heat shock, inflammatory stimuli, viral infection,heavy metals) and its release induced, e.g., by UVBirradiation, necrosis and increased rate of apoptosiscould lead to activation of immunocompetent cells.

Anti-calreticulin Abs in autoimmune diseases 447

These mechanisms and the significance of CRT homol-ogy with some of the parasitic proteins and C1q undernormal and pathophysiological conditions are cur-rently discussed [for review, see 49, 50, 51]. Whateverthe reason might be, once in the extracellular environ-ment, CRT becomes a target for both innate andcell-mediated immune responses [52–54]. Thus,understanding the mechanism of surface exposure orrelease of CRT from cells or its homologue from themicroorganism into the circulation represents a basicrequirement for elucidation of the immune responsesto CRT (biologically important, multifunctional mol-ecule) and its role in the development of autoimmunediseases.

Coeliac disease, as a disease with many autoim-mune features, is characterized by the presence ofautoreactive T lymphocytes and Abs that recognizevarious autoantigens. The structural and functionalanalysis of these autoantigens represents one of thenecessary steps in understanding the immunopatho-genic mechanisms operating in this unique disease.

AcknowledgementsWe thank Drs J. Vencovsky and J. Kafkova forpatients’ sera collection, A. Dedicova, for statisticalanalysis, P. Kasparova and L. Kopecka for their skilfultechnical assistance and valuable help with and prep-aration of the manuscript. This work was supportedby grants A 7020808/1998 of the Academy of Sciences,NI/5051-3, N/5264-3, N/4205-3 of the Ministry ofHealth, 306/98/0433, 310/98/0432, 310/00/1373 ofGrant Agency and 96149 of Ministry of Education,Czech Republic.

References1. Krause K., Michalak M. 1997. Calreticulin. Cell: 88:

439–4432. Spiro R.G., Zhu Q., Bhoyroo V., Sollingen H.D. 1996.

Definition of the lectin-like properties of the molecularchaperon calreticulin, and demonstration of itscopurification with endomannosidase from rat liverGolgi. J. Biol. Chem. 271: 11588–11594

3. White T.K., Zhu Q., Tanzer M.L. 1995. Cell surfacecalreticulin is a putative mannoside lectin whichtriggers mouse melanoma cell spreading. J. Biol. Chem.270: 15926–15930

4. Wheeler D.G., Horsford J., Michalak M., White J.H.,Hendy G.N. 1995. Calcium inhibits vitamin D3 signaltransduction. Nucleic Acid Research 23: 3268–3274

5. Mery L., Measaeli N., Michalak M., Opas M., LewD.P., Krause K.H. 1996. Overexpression of calreticulinincreases intracellular Ca2+ storage and decreasesstore-operated Ca2+ influx. J. Biol. Chem. 271:9332–9339

6. Coppolino M.G., Leung-Hagesteijn Ch., Dedhar S.,Wilkins J. 1995. Inducible interaction of integrin �2 �1with calreticulin. J. Biol. Chem. 270: 23132–23138

7. Atreya Ch.D., Singh N.K., Nakhasi H.L. 1995. TheRubella virus RNA binding activity of humancalreticulin is localized to the N-terminal domain.J. Virol. 69: 3848–3851

8. Andrin Ch., Pinkoski M.J., Burns K., Atkinson E.A.,Krahenbuhl O., Hudig D., Fraser S.A., Winkler U.,Tschopp J., Opas M., Bleackley R.Ch., Michalak M.1998. Interaction between a Ca2+-binding proteincalreticulin and perforin, a component of the cytotoxicT-cell granules. Biochemistry 37: 10386–10394

9. Kishore U., Sontheimer R.D., Sastry K.N., Zappi E.G.1997. The systemic lupus erythematosus (SLE) diseaseautoantigen-calreticulin can inhibit C1q associationwith immune complexes. Clin. Exp. Immunol. 108:181–190

10. Sim R.B., Moestrup S.K., Stuart G.R., Lynch N.J., Lu J.,Schwaeble W.J., Malhorta R. 1998. Interaction of C1qand the collectins with the potential receptorscalreticulin (cC1qR/collectin receptor) and megalin.Immunobiol. 199: 208–224

11. Eggleton P., Reid K.B.M., Kishore U., Sontheimer R.D.1997. Clinical relevance of calreticulin in systemiclupus erythematosus. Lupus 6: 564–571

12. Zhu Q., Zelinka P., White T., Tanzer M.L. 1997.Calreticulin-integrin bidirectional signaling complex.Biochem. Biophys. Res. Commun. 232: 354–358

13. van den Berg R.H., Farber-Krol M.C., Sim R.B., DahaM.R. 1998. The first subcomponent of complement,C1q, triggers the production of IL-8, IL-6, andmonocyte chemoattractant peptide-1 by humanumbilical vein endothelial cells. J. Immunol. 161:6924–6930

14. Chen S.T., Nguyeng T.Q., Yang Y.S., Capra J.D.,Sontheimer R.D. 1996. Calreticulin binds hYRNA and52-kDa polypeptide component of the Ro/SS-Aribonucleoprotein autoantigen. J. Immunol. 156:4484–4491

15. Kovac H., Campbell I.D., Strong P., Johnson S., WardF.J., Reid K., Eggleton P. 1998. Evidence that C1q bindsspecifically to CH2-like immunoglobulin � motifpresent in the autoantigen calreticulin and interfereswith complement activation. Biochemistry 37:17865–17874

16. Rokeach L.A., Haselby J.A., Meilof J.F., Smeenk R.J.T.,Unnasch T.R., Greene B.M., Hoch S.O. 1991.Characterisation of the autoantigen calreticulin.J. Immunol. 147: 3031–3039

17. Boem J., Orth T., an Nguyen P., Soeling H.D. 1994.Systemic lupus erythematosus is associated withincreased autoantibody titers against calreticulin andGrp 94, but calreticulin is not the Ro/SS-A antigen.Eur. J. Clin. Invest. 24: 248–257

18. Routsias J.G., Tzioufas A.G., Sakarellos-Daitsiotis M.,Sakarellos C., Moutsopoulos H.M. 1993. Calreticulinsystemic peptide analogues: anti-peptide antibodies inautoimmune rheumatic diseases. Clin. Exp. Immunol.91: 437–441

19. Orth T., Dornner T., Meyer zum Buschenfelde K.H.,Mayet W.J. 1996. Complete congenital heart block isassociated with increase of autoantibody titers againstcalreticulin. Eur. J. Clin. Invest. 26: 205–215

20. Khalife J., Trottein F., Schacht A.M., Godin C., PierceR.J., Capron A. 1993. Cloning of the gene encoding aShistosoma mansoni homologue to human Ro-SS-Aautoantigen. Mol. Biochem. Parasitol. 57: 193–202

21. Karska K., Tuckova L., Steiner L.,Tlaskalova-Hogenova H., Michalak M. 1995.Calreticulin the potential autoantigen in celiac disease.Biochem. Biophys. Res. Commun. 209: 597–605

448 D. Sanchez et al.

22. Tuckova L., Karska K., Walters J.R.F., Michalak M.,Rossmann P., Krupickova S., Verdu E., Saalman R.,Hanson L.A., Tlaskalova-Hogenova H. 1997.Anti-gliadin antibodies in patients with celiac diseasecross-react with enterocytes and human calreticulin.Clin. Immunol. Immunopathol. 85: 289–296

23. Trier J.S. 1991. Celiac sprue. N. Engl. J. Med. 24:1709–1719

24. Strober W., Fuss I.J. 1999. Gluten-sensitive enteropathyand other immunologically mediated enteropathies. InMucosal Immunology. P.L. Ogra, J. Mestecky, M.E.Lamm, W. Strober, J. Bienenstock, J.R. McGhee Jr, eds.Academic Press, San Diego/London/Boston,pp. 1101–1128

25. Velluzzi F., Caradonna A., Boy M.F., Pinna M.A.,Cabula R., Lai M.A., Piras E., Corda G., Mossa P.,Atzeni F., Loviselli A., Usai P., Marotti S. 1998. Thyroidand celiac disease: clinical, serological, andechographic study. Am. J. Gastroenterol. 93:976–979

26. Troncone R., Ferguson A. 1991. Anti-gliadin antibodies.J. Pediatr. Gastroenterol. Nutr. 12: 150–158

27. Valeski J.E., Kumar U., Beutner E.H., Lerner R.A.,Chorzelski T.P. 1990. Immunology of celiac disease: tissueand species specificity of endomysial and reticulinantibodies. Int. Arch. Allergy Appl. Immunol. 93: 1–7

28. Karpati S., Burgin-Wolff A., Krieg T., Meurer M., StolzW., Braun-Falco O. 1990. Binding to human jejunum ofserum IgA antibody from children with coeliacdisease. Lancet 36: 1335–1338

29. Dieterich W., Ehnis T., Bauer M., Donner P., Volta U.,Riecken E.O., Schuppan D. 1997. Identification oftissue transglutaminase as the autoantigen of celiacdisease. Nature Med. 3: 797–801

30. Tlaskalova-Hogenova H., Stepankova R., Fric P.,Kolınska J., Vancikova Z., Jılek M., Kocna P., Lipska L.,Slaby J., Dvorak M. 1990. Lymphocyte activation bygliadin in celiac disease and in experimentallyinduced enteropathy of germ-free rats. In Advances inMucosal Immunology. T. MacDonald, S.J. Challacombe,P.R. Bland, C.P. Stokes, V. Heatley, A. McMowat, eds.Kluwer Academic Publishers,Dordrecht/Boston/London, pp. 781–786

31. Tuckova L., Tlaskalova-Hogenova H., Farre M.A.,Karska K., Rossmann P., Kolınska J., Kocna P. 1995.Molecular mimicry as a possible cause of autoimmunereactions in celiac disease? Antibodies to gliadincross-react with epitopes on enterocytes. Clin. Immunol.Immunopathol. 74: 170–176

32. Krupickova S., Tuckova L., Flegelova Z., Michalak M.,Walters J.F.R., Whelan A., Harries J., Vencovsky J.,Tlaskalova-Hogenova H. 1999. Identification ofcommon epitopes on gliadin, enterocytes, andcalreticulin recognised by antigliadin antibodies ofpatients with coeliac disease. Gut 44: 168–173

33. Walker-Smith J.A., Guadalini S., Schmitz J., ShmerlingD.H., Visakorpi J.K. 1990. Revised criteria for diagnosisof coeliac disease. Report of working group ofEuropean Society of Pediatric Gastroenterology andNutrition. Arch. Dis. Child. 65: 909–911

34. Baksh S., Michalak M. 1991. Expression of calreticulinin Escherichia coli and identification of its Ca2+ bindingdomains. J. Biol. Chem. 266: 21458–21465

35. Pinsky D.J., Kuwabara K., Schmidt A.M., Lawson Ch.,Benedict C., Broekman J., Marcus A.J., Malinski T.,

Ryan J., Stern D.M. 1996. Calreticulinin in thrombosis.In Calreticulin. M. Michalak, ed. Springer-Verlag,Heidelberg, pp. 155–168

36. Wingfield P.T. 1997. Purification of recombinantproteins. In Current Protocols in Protein Science. J.E.Coligan, B.M. Dunn, H.L. Ploegh, eds. Wiley and Sons,Inc, USA Supplement 9, pp. 6.0.1.–6.6.19

37. Raul F., Simon P., Kedinger M., Haffen K. 1977.Intestinal enzyme activities in isolated villus andcrypts cells during postnatal development of the rat.Cell Tissue Res. 176: 167–178

38. Kolınska J., Kraml J., Zakostelecka M., Kadlecova L.1990. Actinomycin D suppressedhydrocortisone-induced changes in distribution andsialylation of brush-border enzymes in jejunal villi andcrypts. Biochem. Int. 21: 581–591

39. Ladinser B., Rossipal E., Pittschieler K. 1994.Endomysium antibodies in coeliac disease: animproved method. Gut 35: 776–778

40. Laemmli U.K., Favre M. 1979. Maturation of the headof bacteriophage T4. I DNA packing events. J. Mol.Biol. 80: 575–599

41. Towbin H., Staehelin T., Gordon J. 1979. Electro-phoretic transfer of proteins from polyacrylamide gelsto nitrocellulose sheets: procedure and someapplications. Proc. Natl Acad. Sci. USA 76: 4350–4354

42. Arranz E., Ferguson A. 1993. Intestinal antibodypattern of celiac disease: occurrence in patients withnormal jejunal biopsy histology. Gastroenterology 104:1263–1272

43. Rolny P., Sigurjonsdottir H.A., Remotti H., NilssonL.A., Ascher H., Tlaskalova-Hogenova H., Tuckova L.1999. Role of immunosuppressive therapy in refractorysprue-like disease. Am. J. Gastroenterol. 94: 219–225

44. Yacyshyn B., Meddings J., Sadowski D.,Bowen-Yacyshyn M.B. 1996. Multiple sclerosis patientshave peripheral blood CD45RO+ B cells and increasedintestinal permeability. Dig. Dis. Sci. 41: 2493–2498

45. Hollander D., Vadheim C., Brettholz E., PettersonG.M., Delahunty T., Rotter J.I. 1986. Increasedintestinal permeability in patients Crohn’s disease andtheir relatives. Ann. Intern. Med. 105: 883–885

46. Kreisel W., Siegel A., Bahler A., Spamer C., Schilty E.,Kist M., Seilnacht G., Klein R., Berg P.A., Heilmann C.1999. High prevalence of antibodies to calreticulin ofthe IgA class in primary biliary cirrhosis: a possiblerole of gut-derived bacterial antigens in its aetiology?Scand. J. Gastroenterol. 34: 623–628

47. Lieu T.S., Sonthaimer R.D. 1997. A subpopulation ofWIL-2 cell calreticulin molecules is associated withRo/SS-A ribonucleoprotein particles. Lupus 6: 40–47

48. Sontheimer R.D., Lieu T.S., Capra J.D. 1993.Calreticulin. The diverse functional repertoire of a newhuman autoantigen. The Immunologist 1: 155–160

49. Eggleton P., Llewellyn D.H. 1999. Pathophysiologicalroles of calreticulin in autoimmune disease. Scand.J. Immunol. 49: 466–473

50. Conway E.M., Liu L., Nowakowski B.,Steiner-Mosonoyi M., Ribeiro S.P., Michalak M. 1995.Heat shock-sensitive expression of calreticulin. Biol.Chem. 170: 17011–17016

51. Zhu Q., Belitz H.D., Ashkenazi A., Idar D. 1997.Calreticulin-integrin bidirectional signaling complex.Biochem. Biophys. Res. Commun. 232: 354–358

Anti-calreticulin Abs in autoimmune diseases 449

52. Basu S., Srivastava P.K. 1999. Calreticulin, apeptide-binding chaperone of the endoplasmicreticulum, elicits tumor- and peptide-specificimmunity. J. Exp. Med. 189: 797–802

53. Coppolino M.G., Woodside M.J., Demaurex N.,Grinstein S., St-Arnaud R., Dedhar S. 1997. Calreticulin

is essential for integrin-mediated calcium signallingand cell adhesion. Nature 386: 843–847

54. Sontheimer R.D., Nguyeng T.Q., Chen S.T., Lieu T.S.,Capra J.D. 1995. The unveiling of calreticulin—A clinicrelevant tour of modern cell biology. J. Invest. Med. 43:362–370