Blockage of T-cell costimulation inhibits T-cell action in celiac disease

Blockage of T-cell costimulation inhibits T-cell action in celiac disease

GASTROENTEROLOGY 1998;115:564–572 Blockage of T-Cell Costimulation Inhibits T-Cell Action in Celiac Disease LUIGI MAIURI,*,‡ SALVATORE AURICCHIO,* SA...

508KB Sizes 0 Downloads 24 Views

GASTROENTEROLOGY 1998;115:564–572

Blockage of T-Cell Costimulation Inhibits T-Cell Action in Celiac Disease LUIGI MAIURI,*,‡ SALVATORE AURICCHIO,* SALVATORE COLETTA,* GIULIO DE MARCO,* ANTONIO PICARELLI,§ MARCO DI TOLA,§ SONIA QUARATINO,\ and MARCO LONDEI\ *Department of Pediatrics, University ‘‘Federico II,’’ Naples, Italy; ‡Children’s Hospital Pausilipon, Naples, Italy; §Cattedra di Gastroenterologia, II Clinica Medica, University ‘‘La Sapienza,’’ Rome, Italy; and \Immunology Division, Kennedy Institute of Rheumatology, London, England

Background & Aims: Celiac disease is an exemplary model of T cell–mediated pathology. Therefore, therapeutic approaches that target T cells may successfully control this disease. CTLA-4 immunoglobulin (CTLA4Ig) can inhibit T-cell activation by blocking the engagement of CD28. We took advantage of this tool to define the pathogenic role of gliadin-specific T cells in the induction of celiac disease. Methods: Duodenal biopsy specimens from 7 treated celiac patients were challenged in vitro with gliadin and CTLA-4Ig or CD40-Ig. After 24 hours, the biopsy specimens were analyzed for the presence of characteristic modifications induced by gliadin challenge. Results: CTLA-4Ig down-regulated the expression of CD25, intercellular adhesion molecule 1, interleukin 2, and interferon gamma (stained lamina propria mononuclear cells/mm2; P F 0.05) induced by gliadin challenge, caused apoptosis of gliadin-specific T cells (apoptotic T cells/mm2; P F 0.05), and inhibited the production of antiendomysial antibody (P F 0.01). However, it did not control intraepithelial T-cell migration (P 5 NS) and Fas expression by enterocytes. Conversely, CD40-Ig only controlled production of antiendomysial antibody. Conclusions: In an organ culture model, CTLA-4Ig controls many but not all of the immunologic features of celiac disease.

ecent studies have outlined the molecular mechanisms involved in T-cell activation. It is now accepted that this event requires at least two separate but complementary signals.1,2 The first is the recognition of antigen, which is specific for each single T clone.3 The second, nonclonotypic, is practically common to all T lymphocytes.4,5 Although the concept of the two signals was postulated several years ago,1 only in the last few years has the precise nature of the molecules controlling the second signal been defined. Thus, it has been shown that the engagement of the CD28 molecule, expressed on the surface of T cells, is crucial for complete T-cell activation.5,6 At least two complementary molecules, CD80 (B7.1) and/or CD86 (B7.2), expressed on the APC,

R

can engage CD28.7 Any perturbation of this interaction will automatically alter the process of T-cell activation and may lead to a condition of T-cell nonresponsiveness or anergy.8 Another molecule expressed on the surface of T cells, called CTLA-4, competes with CD28 in binding CD80 and CD86.9 Both CD28 and CTLA-4 bind CD80 and CD86, although CTLA-4 does it with a much higher affinity.10 The higher affinity of CTLA-4 for CD80 or CD86 has been used, with the CTLA4 immunoglobulin (CTLA-4Ig) fusion protein, to mask CD80-86 expressed by APC in order to block the activation of CD28. These studies have suggested that CTLA-4Ig could disrupt the engagement of CD28, altering the required second signal and leading to a condition of T-cell tolerance.5,11,12 These in vitro studies were soon followed by in vivo investigations in which it was possible to show that CTLA-4Ig could allow acceptance of a graft and prolong its survival.13–15 In other studies, CTLA-4Ig treatment could prevent and even control autoimmune diseases.16,17 These studies showed that such a treatment could have a significant clinical impact in several immunity-mediated pathologies. Celiac disease is a pathological condition that is caused by the ingestion of gliadin,18 which initiates a cascade of events that is probably driven by gliadin-specific T cells.19 Of note, other antigens in addition to gliadin, in particular autoantigens such as tissue transglutaminase,20–22 play a role in the pathogenesis of this disease. Furthermore, it is not yet known whether intraepithelial leukocytes (IELs) or lamina propria lymphocytes (LPLs) cause the epithelial damage, because both IELs23 and LPLs24 have been reported to induce epithelial damage Abbreviations used in this paper: EMA, antiendomysial antibody; ICAM-1, intercellular adhesion molecule 1; IEC, intraepithelial compartment; IFN-g, interferon gamma; IL-2, interleukin 2; LPL, lamina propria lymphocyte; LPMNC, lamina propria mononuclear cell; MAb, monoclonal antibody; SEC, subepithelial compartment; TUNEL, terminal deoxynucleotidyl transferase–mediated deoxyuridine triphosphate digoxygenin nick-end labeling. r 1998 by the American Gastroenterological Association 0016-5085/98/$3.00

September 1998

CTLA–4Ig T CELLS AND EPITHELIAL CHANGES IN CELIAC DISEASE

and shedding. This pathology, normally treated with a gliadin-free diet, is currently underdiagnosed and has a very high incidence in the Northern Hemisphere. Because in vitro gliadin challenge of celiac patients’ biopsy specimens mimics many of the immunologic features observed in overt celiac disease,25 as well as the production of antiendomysial antibodies (EMAs),21 these organ culture studies provide an ideal bench test to elucidate whether disruption of the second signal by CTLA-4Ig could control celiac disease. This study would also clarify the possible role of pathogenic T cells in celiac disease. This report shows that the soluble fusion protein CTLA-4Ig efficiently influences the T cell–driven modifications observed in this organ culture model on gliadin challenge. The CTLA-4Ig treatment, however, does not control other specific features of gliadin challenge of small intestinal biopsy specimens from treated celiacs, such as lymphocyte migration and up-regulation of Fas by enterocytes. Thus, our report provides evidence of the efficacy of the CTLA-4Ig treatment in controlling human pathogenic T cells, but it also opens a novel view on the pathogenic mechanisms leading to full-blown celiac disease.

Materials and Methods Patients Seven patients with celiac disease admitted to the gastrointestinal unit for diagnostic purposes were diagnosed according to European Society for Pediatric Gastroenterology and Nutrition criteria. All of them showed duodenal villous atrophy and the presence of serum EMAs while on a glutencontaining diet. After at least 24 months of gluten-free diet, in the absence of clinical symptoms and serum EMAs, the patients (called ‘‘treated celiac disease’’ below) underwent a second duodenal biopsy.

Intestinal Samples Intestinal specimens from treated patients with celiac disease were obtained at the duodenal-jejunal flexure by peroral biopsy. Informed consent was obtained from all patients. All specimens were collected in ice-chilled tissue culture medium and cultured within 20 minutes. Each biopsy specimen was washed in 0.15 mol/L sodium chloride and examined with a dissecting microscope. One specimen from each patient was oriented and embedded in optimal cutting temperature compound (Tissue Tek; Miles Laboratories, Elkhart, IN), snapfrozen in isopentane cooled in liquid nitrogen, and then stored at 270°C until cryosectioning. Three-micrometer-thick sections were stained with hematoxylin and used for diagnosis; IELs were enumerated by counting CD3-positive cells per 100 enterocytes.

565

Preparation of the Culture Medium and Mucosal Tissue Culture Peptic-tryptic (PT) digest of gliadin from bread wheat was obtained as previously reported.26 Dr. Peter Lane (Basel, Switzerland) kindly donated CTLA-4Ig; CD40-Ig was obtained from Immunex (Seattle, WA). Mucosal samples from each patient were placed on a stainless steel mesh positioned over the central well of an organ culture dish with the villous surface of the biopsy specimens uppermost. The culture took place as previously reported.25 PT digest of gliadin from bread wheat was added at a concentration of 1 mg/mL.25 Just before the in vitro culture, CTLA-4Ig and CD40-Ig were added and carefully mixed to reach a final concentration of 5 µg/mL of culture medium. The intestinal samples were cultured for 6 and 24 hours in the presence of medium alone, PT gliadin digest, or PT gliadin plus CTLA-4Ig. In 4 patients (cases 1–4), CD40-Ig was similarly added to the culture medium together with gliadin digest (1 mg/mL) at the same final concentration. In 3 patients (cases 1–3), cultures with CTLA-4Ig alone at 5 µg/mL were also performed. In vitro cultures with gliadin digest together with CD40-Ig, as well as with CTLA-4Ig without gliadin, were used as control experiments. Because different samples from the same subject were cultured with and without PT digest and PT digest added with CTLA-4Ig, each subject constituted an internal control. At the end of the incubation, the specimens were harvested, snap-frozen in isopentane cooled in liquid nitrogen, and finally prepared for cryosectioning as described earlier. Culture supernatants were collected after culture and stored at 270°C.

Staining Technique Three-micrometer-thick cryostat sections of each sample were air-dried at room temperature and then fixed in acetone for 10 minutes. The experiments were carried out as previously reported.25 Serial sections were individually tested with monoclonal antibodies (MAbs) to different markers (intercellular adhesion molecule 1 [ICAM-1; 1:400; Dako, Copenhagen, Denmark], CD25 [1:30; Dako], CD3 [1:200; Dako], interleukin 2 [IL-2; 5 µg/mL 13A6 rat Ig], and interferon gamma [IFN-g; 5 µg/mL 166.5 mouse Ig]), and immunostaining was performed by following the alkaline phosphatase/anti–alkaline phosphatase method with New Fucsin (Merck Sharp & Dhome, Inc., West Point, PA) and naphtol AS-bi-phosphate (Sigma Chemical Co., St. Louis, MO).25 MAbs M3 and M38 Fas (CD95) specific (Immunex, Seattle WA, 1:30)27 were also used in 4 cases (subjects 1–4). The reaction product was developed by peroxidase staining technique with 3-amino-9-ethyl-carbazole (Sigma) and hydrogen peroxide. Staining of epithelial cells by M3 or M38 MAbs was arbitrarily graded from 0 to 12. Two-color staining technique was also performed as described previously.25

Morphometric Analysis The total number of mononuclear cells positive for the different markers was blindly calculated within a standard area

566

MAIURI ET AL.

of 1 mm2 of lamina propria with a calibrated ocular graticule aligned parallel to the muscularis mucosae. The number of CD3-positive cells in the different mucosal compartments was calculated as described previously.25 The total area of lamina propria was longitudinally divided into two parts of equal height, and the number of CD3-positive cells was calculated in the subepithelial compartment (SEC) and in the deep layer. In the SEC, little areas of lamina propria corresponding to the villus axis were independently evaluated and pooled to reach the standard reference area. The total number of CD3-positive cells in the intraepithelial compartment (IEC) was calculated as the percentage of enterocytes; CD3-positive cells were counted for at least 500 enterocytes in each sample.

EMA Detection in Culture Supernatants Detection of EMAs was sought for in undiluted culture supernatants by immunofluorescence (Eurospital Pharma, Trieste, Italy). Forty microliters was applied on each section of monkey esophagus for 30 minutes according to an experimental procedure previously reported.21 Two distinct observers blindly evaluated the results.

In Situ Detection of DNA Fragmentation of Lamina Propria Mononuclear Cells In situ detection of DNA fragmentation of lamina propria mononuclear cells (LPMNCs) was performed as described previously.28 Briefly, fresh intestinal specimens from 4 treated celiac patients were studied after challenge with gliadin digest, gliadin digest supplemented with CTLA-Ig, and CTLA-Ig alone. The specimens were immersed in optimal cutting temperature compound and rapidly frozen in liquid nitrogen, and 5-µm sections were cut in a cryostat. Sections were air-dried and fixed in acetone. To detect cells with DNA fragmentation, we used a modification of the terminal deoxylnucleotidyl transferase (TdT)-mediated deoxyuridine triphosphate digoxygenin nick-end labeling (TUNEL) method of Surh and Sprent.28,29 Tissue sections incubated with unlabeled nucleotides or with TdT alone were used as controls of specificity. The experiments were repeated at least three times each in the presence of positive and negative controls.28 The number of stained cells was calculated within a 1-mm2 standard area of lamina propria. The characterization of LPMNCs undergoing DNA fragmentation was performed by double-staining techniques with CD3 or CD68 MAbs as previously reported.25

Statistical Analysis Student’s two-tailed t test was used to compare specimens exposed to gliadin with those exposed to gliadin and CTLA-4Ig and with those exposed to medium alone for the same time of incubation. Nonparametric tests (Wilcoxon two-tailed) were also applied, and the results have been found concordant with those obtained with parametric tests. Fisher’s

GASTROENTEROLOGY Vol. 115, No. 3

exact test was applied to compare the expression of FAS after CTLA-4Ig treatment with that observed after challenge with gliadin alone. Statistical evaluation for EMAs was performed according to methods described in a previous report.21

Results CTLA-4Ig Is Able to Restrain Gliadin-Induced Features of Activation of Lamina Propria T Cells in Treated Celiac Disease Intestine After 6 hours of challenge with gliadin, CTLA4Ig, in all but 1 case (subject 5), caused a significant decrease in IL-2–positive (Figure 1A and B) and IFN-g– positive (Figure 1C and D) cells in the lamina propria (Table 1; P , 0.05 vs. cultures with gliadin alone). After 24 hours of in vitro culture, CTLA-4Ig restrained the immunologic modifications induced by gliadin challenge in all but 1 case (subject 5). CD25positive cells were significantly reduced in the lamina propria (P , 0.05 vs. cultures with gliadin alone; Figure 2) as well as ICAM-1–positive cells (P , 0.05 vs. cultures with gliadin alone; Figures 2 and 3). Because the soluble form of CTLA-4 that we used in our experiments is a fusion protein composed of the CTLA-4 extracellular domain and the Fc portion of the human IgG1 constant region, we used as an internal control another similarly constructed fusion protein. We chose CD40-Ig because of the effect of CD40 in B-cell activation30 and because it modulates the expression of CD80/86 on APC,31,32 whereas its natural ligand CD40L is expressed on activated T cells.33 When intestinal samples from 4 patients (cases 1–4) were cultured in the presence of gliadin digest and CD40-Ig, no differences were detected at any time point compared with gliadin alone (P 5 NS; Figure 2). Intestinal samples from 3 patients (cases 1–3) cultured in the presence of CTLA-4Ig alone did not show significant alterations compared with medium alone (Figure 2). CTLA-4Ig Is Able to Suppress the Production of EMA After Gliadin Challenge of Treated Celiac Disease Intestine We have recently shown that in vitro gliadin challenge of celiac biopsy specimens leads to the production of EMAs.21 Because CTLA-4Ig significantly controlled T cell–mediated modifications observed in our organ culture model on gliadin challenge, we checked whether the production of EMAs was also regulated. In all the five tested samples (cases 1–4 and 6), no EMAs

September 1998

CTLA–4Ig T CELLS AND EPITHELIAL CHANGES IN CELIAC DISEASE

567

Figure 1. Effect of CTLA-4Ig on the expression of (A and B) IL-2 and (C and D) IFN-g by LPMNCs induced by 6 hours of in vitro gliadin challenge in treated celiac disease biopsy specimens. After gliadin challenge, many LPMNCs express (A) IL-2 and (C) IFN-g; note also soluble IL-2 in the extracellular matrix (A). After CTLA-4Ig treatment of gliadin-stimulated biopsy specimens, only a few LPMNCs express (B) IL-2 and (D) IFN-g (alkaline phosphatase staining technique; original magnification 1503).

were detected in culture supernatants after culture with gliadin plus CTLA-4Ig (P , 0.01 vs. cultures with gliadin alone). In all cases, they were detected after gliadin challenge (Table 2). CD40-Ig was also effective in controlling the production of EMAs induced by gliadin challenge in 2 of 2 tested samples (Table 2). Table 1. Effect of CTLA-4Ig Treatment on the Expression of IL-2 and IFN-g by LPMNCs Induced by 6 Hours of Gliadin Challenge in Cultured Treated Celiac Disease Intestine IL-2 a Gliadin Median 23 Mean 22.6 Range 11.4–25.1 SD 9.5 aNo. bP

IFN-g a

Gliadin 1 CTLA4 b

Gliadin

Gliadin 1 CTLA4 b

7.5 9.5 3–24.7 8.1

20 19.8 12–27.4 5.7

11 10.5 5–15.9 4

of stained LPMNCs/mm2 of lamina propria. , 0.05 vs. cultures with gliadin alone.

CTLA-4Ig Induces T-Cell Apoptosis in Biopsy Specimens of Treated Celiac Disease Patients Challenged With Gliadin The previous set of data indicated that CTLA-4Ig was effective in controlling most of the T cell–mediated effects observed in our organ culture system. To establish how CTLA-4Ig restrained T-cell function, we determined the number of (CD3-positive) T cells undergoing apoptosis using the approach described by Surh and Sprent.29 The results of this analysis are reported in Table 3. The number of LPMNCs showing DNA fragmentation in treated celiac intestine was higher after challenge with gliadin plus CTLA-4Ig (Figure 4A) than after challenge with gliadin alone (after culture with gliadin plus CTLA-4Ig: median, 66/mm2 of lamina propria; range, 55.4–80/mm2 of lamina propria; mean, 66.8/mm2 of lamina propria; SD, 11.2/mm2 of lamina propria; after culture with gliadin alone: median, 17.7/mm2 of lamina propria; range, 5.8–24.8/mm2 of lamina propria; mean,

568

MAIURI ET AL.

GASTROENTEROLOGY Vol. 115, No. 3

Figure 2. Effect of CTLA-4Ig on the in vitro immune response to gliadin in treated celiac disease intestine. ICAM-1– and CD25–positive cells were numbered per square millimeter of lamina propria; CD3-positive cells in the SEC were numbered per square millimeter of lamina propria.18 The number of CD3-positive cells in the IEC was calculated as the percentage of enterocytes. j, Medium (n 5 7); 6, gliadin (n 5 7); h, gliadin 1 CTLA4 (n 5 7); M, gliadin 1 CD40 (n 5 4); Z, CTL4 (n 5 3). ••P , 0.001 vs. cultures with medium alone; *P , 0.05 vs. cultures with gliadin alone.

16.5/mm2 of lamina propria; SD, 9.4/mm2 of lamina propria; P , 0.05). CTLA-4Ig alone did not induce apoptosis (median, 14/mm2 of lamina propria; range, 7–18/mm2 of lamina propria; mean, 13/mm2 of lamina propria; SD, 5.5/mm2 of lamina propria; P 5 NS). LPMNCs that showed DNA fragmentation were defined as CD3-positive cells, as shown in Figure 4B. The percentage of total CD3-positive cells showing DNA fragmentation within 1 mm2 of lamina propria was nearly five times higher after culture with gliadin plus CTLA-4Ig (median, 9.95%; range, 8.3%–12.1%; mean, 10.1%; SD, 1.7%) than after culture with gliadin alone (median, 2.65%; range, 0.8%–3.7%; mean, 1.8%; SD, 0.8%; P , 0.05; Table 3). Interestingly, the percentage of lamina propria T cells affected by CTLA-4Ig (,10%) is the same affected by gliadin challenge in treated celiac disease intestine.25

Table 2. Production of EMAs in Culture Supernatants After 24 Hours of In Vitro Culture With Medium Alone or With Gliadin in Treated Celiac Intestine: Effect of CTLA-4Ig and CD40-Ig

Medium Gliadin aP bP

24 h of challenge

24 h of challenge with CTLA-4Ig

24 h of challenge with CD40-Ig

0/5 5/5 a

0/3 0/5 b

0/2 0/2

, 0.01 vs. cultures with medium alone. , 0.01 vs. cultures with gliadin alone.

CTLA-4Ig Does Not Control Gliadin-Induced Intraepithelial Infiltration of T Cells in Biopsy Specimens of Treated Celiacs We have previously described that gliadin challenge of treated celiac disease biopsy specimens induces subepithelial migration as well as intraepithelial infiltration of CD3-positive cells.25 Migrating IE CD3-positive cells do not express IL-2–receptor, ICAM-1, or DR molecules.25 CTLA-4Ig was unable to control the migration of CD3-positive cells in the SEC (P 5 NS vs. cultures with gliadin alone; Figure 2), or into the IEC (P 5 NS vs. cultures with gliadin alone; Figure 2). CTLA-4Ig Does Not Repress Enterocyte FAS Expression Induced by Gliadin Challenge in Treated Celiac Disease Biopsy Specimens We have recently observed that Fas is overexpressed on enterocytes of celiacs and that this expression could be up-regulated by gliadin challenge in treated Table 3. CTLA-4Ig in Conjunction With Gliadin Induces Apoptosis of Pathogenic T Cells

Gliadin Gliadin 1 CTLA-4Ig CTLA-4Ig aMean

% of CD3-positive cells undergoing apoptosis

Apoptotic cells (/mm2 of LP )

1.8 10.1 a 2.7

16.5 66.8 a 13

values; P , 0.05, gliadin 1 CTLA-4Ig vs. gliadin.

September 1998

CTLA–4Ig T CELLS AND EPITHELIAL CHANGES IN CELIAC DISEASE

569

Figure 3. Effect of CTLA-4Ig on the expression of ICAM-1 on 24 hours of in vitro gliadin challenge in intestinal explants from treated celiacs. ICAM-1 expression after challenge with (A) medium alone, (B) gliadin, (C) gliadin plus CTLA-4Ig. Note the strong expression of ICAM-1 in almost all LPMNCs after culture with gliadin (B), whereas lower expression was found after culture with medium alone (A) or with gliadin plus CTLA-4Ig (C) (alkaline phosphatase staining technique: the strong staining of the brush border is due to endogenous alkaline phosphatase; original magnification 1503).

celiacs’ biopsy specimens (Maiuri L., et al., submitted). We therefore tested whether CTLA-4Ig could control this enterocyte-specific, gliadin-driven modification. Interestingly, CTLA-4Ig treatment was unable to control the expression of Fas by the enterocytes induced by 24 hours of gliadin challenge. In all 4 tested patients, the epithelial expression of Fas was intense (21) after gliadin challenge as well as after challenge with gliadin plus CTLA-4Ig with respect to the pattern observed before in vitro

manipulation or after culture with medium alone (P 5 0.99 vs. cultures with gliadin alone; Table 4).

Discussion The development of novel therapies able to control immunity-mediated pathologies is the objective of many studies, because no satisfactory therapy for autoimmune diseases, allergies, and complications from transplan-

Figure 4. Detection of CD3-positive T cells undergoing apopotosis. DNA fragmentation, detected by the TUNEL technique, observed in biopsy specimens of treated celiac patients challenged in vitro with gliadin and CTLA-4Ig. Many LPMNCs are TUNEL positive; no staining at the nuclear level can be detected in enterocytes (A). The apoptotic lamina propria cells (dark blue; thin arrows) are CD3 positive (orange), although not all CD3-positive cells (orange, arrow) show sign of apoptosis (B) (original magnification 2803; A, peroxidase staining technique; B, double-color immunostaining, peroxidase [orange], and alkaline phosphatase [blue] staining techniques; hematoxylin counterstaining).

570

MAIURI ET AL.

GASTROENTEROLOGY Vol. 115, No. 3

Table 4. Effect of CTLA-4Ig Treatment on the Expression of FAS by Enterocytes Induced by 24 Hours of Gliadin Challenge in Cultured Treated Celiac Disease Intestine Increased expression of FAS by enterocytes (n) a Gliadin Gliadin 1 CTLA4 b

4/4 4/4

a With

respect to the pattern observed before in vitro manipulation and after challenge with the sole medium. bFisher’s exact test: P 5 0.99 vs. cultures with gliadin alone.

tation is available. Although different pathologies arise from these disorders, common molecular mechanisms are shared by these conditions. Central to all of them is the activation of pathogenic T cells.34,35 Recently, the basic players involved in the process of T-cell activation have been defined. It is now accepted that T cells require at least two signals to become fully activated.1,2 A recent study showed that the optimal approach for disrupting CD28 engagement, thus eliminating the second signal, is the use of soluble forms of CTLA-4.6 The disruption of the CD28 pathway has been successfully tested, as a therapeutic approach, in several animal models of immunity-mediated pathologies.2,13,14,16,17 For these reasons, we decided to study the effect of CTLA-4Ig in an organ culture model of celiac disase.25,28 In this disease, a single antigen, gliadin, drives an immune-pathological cascade considered to be controlled by T lymphocytes,19 which leads to tissue damage. In this model, it is possible to time the challenge of the pathogenic antigen and simultaneously to provide CTLA4Ig as a way to perturb the engagement of the CD28 expressed on T lymphocytes. Furthermore, celiac disease has other features that are of interest. Indeed, although gliadin, an exogenous antigen, is the triggering antigen implicated in celiac disease, an autoimmune component (EMA) is detected in all patients with celiac disease.21,22,36,37 Our report indicates that CTLA-4Ig is effective in controlling many of the immunologic features previously described by us as specific to gliadin challenge.21,25 Thus, no induction of CD25 on T cells, or sustained ICAM-1 expression, is observed. Moreover, the CTLA-4Ig treatment efficiently blocks the production of cytokines, such as IL-2 and IFN-g, most dominantly produced by T cells. This treatment is also able to control the production of EMA induced by gliadin, a trait that we have recently reported to be specific to the gliadin challenge.21 Whatever antigen(s) drives EMA production, CTLA-4Ig treatment prevents it. This indicates that T cells at the site of the lesion are essential for the production of these autoantibodies. Because it has been hypothesized that EMA may have a major pathogenic role

in celiac disease,20–22 these results stress the therapeutic relevance of CTLA-4Ig in celiac disease. Thus, to some extent, not surprisingly, we describe that CTLA-4Ig regulated the function of the likely pathogenic T cells. Other molecules, such as CD40/CD40L, are involved in the cross-talk between T cells and APC.30,33 We therefore used the fusion protein CD40-Ig to define the relevance of this receptor/coreceptor system in our model as well as to provide the ideal internal control for CTLA-4Ig. This fusion protein could only control the production of EMA but did not influence other T cell–driven events. This is of interest, because it has been shown recently that CD40L expressed on T lymphocytes plays an important role in T-cell activation38,39; therefore, CD40-Ig could have influenced T-cell function. Moreover, it has been shown that masking of CD40L can control ongoing autoimmune processes40 as well as graft rejection.15 Thus, at least in our organ culture model, CD40-Ig does not affect human pathogenic T cells, although it controls, as predicted, B lymphocytes30,33,41 by blocking the production of EMAs. CTLA-4Ig is quite successful in restraining gliadinspecific T cells by, at least in our system, inducing apoptosis of the antigen-specific T cells. Indeed, we observe a strict correlation between the number of T cells that become activated (,10%) after gliadin challenge25 and the percentage of T cells undergoing apoptosis (10.1%) after CTLA-4Ig treatment of gliadin-stimulated celiac disease biopsy specimens. Our study also provides an explanation of the mechanism of action of a potentially therapeutic agent such as CTLA-4Ig; apoptosis of antigenspecific T cells is observed when CTLA-4Ig is given in association with gliadin. This is not surprising, because engagement of CD28 augments the expression of Bcl-XL, a molecule known to have antiapoptotic effects,42 and CTLA-4Ig, by masking CD28, would clearly interfere with this essential T-cell survival process. However, it is possible that T cells at different stages of maturation (naive or memory) could differentially respond to CTLA4Ig treatment by inducing either anergy or apoptosis. Thus, CTLA-4Ig proved to be a possibly important therapeutic tool in controlling pathogenic T cells of the small intestine, which are probably involved in celiac disease. CTLA-Ig is, however, unable to control other unique modifications induced by gliadin challenge in the celiac disease organ culture model. This treatment does not abolish the migration of T lymphocytes, most of them CD28 positive, in the SEC as well as in the intraepithelial area of the mucosa, where T cells are almost all CD8 positive and CD28 negative.43 An increase of IELs, in particular CD8-positive CD28-negative T cells (data not

September 1998

CTLA–4Ig T CELLS AND EPITHELIAL CHANGES IN CELIAC DISEASE

shown), is a well-known property of the small intestinal mucosa of untreated patients with celiac disease.18 The inability of CTLA-4Ig to control this migration, in particular that in the SEC, is puzzling because these cells are CD28 positive and therefore allegedly sensitive to CTLA-4Ig action. Thus, the data reported in the present article are in agreement with those in our previous study,25 where we hinted that, in celiac disease, not all the pathogenic modifications were controlled by T cells. Also enigmatic is the incapacity of CTLA-4Ig to preclude the overexpression of Fas by enterocytes on gliadin challenge. Apoptosis of enterocytes is known to be a feature of untreated celiac disease intestine,44 and we have recently observed that the expression of Fas is increased on enterocytes of untreated celiac patients (Maiuri L, et al., manuscript submitted), and in treated celiac patients on gliadin challenge. Moreover, we have shown that apoptosis of enterocytes in celiac disease is inhibited by masking Fas with inhibitory anti-Fas MAbs (Maiuri L, et al., manuscript submitted). In this light, because CTLA-4Ig has no effect on T-cell migration and Fas expression by enterocytes, it may be argued that T cells do not drive celiac disease. Our data seem to confer a limited role to gluten-specific or eventually tissue transglutaminase– specific T cells, because CTLA-4Ig controls EMA production in the modulation of important steps of celiac disease. In conclusion, our data provide clear evidence that CTLA-4Ig, as studied in a human organ system culture, can actually interfere with the evolution of human pathologies by controlling pathogenic T cells. They also show that mucosal T lymphocytes may be targeted with CTLA-4Ig to control disease progression. Thus, they indicate the therapeutic potential of this treatment in celiac disease and probably in other immunemediated intestinal pathologies. Nevertheless, they provide the unexpected evidence that T cells might have a less important role in celiac disease and that a more complex pathogenic path is involved in this disease.

8. 9.

10.

11.

12.

13.

14.

15.

16. 17.

18.

19.

20.

21.

References 1. Bretscher P, Cohn M. A theory of self–nonself discrimination. Science 1970;169:1042–1049. 2. Lafferty KJ, Gill RG. The maintenance of self-tolerance. Immunol Cell Biol 1993;71:209-214. 3. Bluestone JA. New perspectives of CD28-B7–mediated T cell costimulation. Immunity 1995;2:555–559. 4. Liu Y, Linsley PS. Costimulation of T-cell growth. Curr Opin Immunol 1992;4:265–270. 5. June CH, Bluestone JA, Nadler LM, Thompson CB. The B7 and CD28 receptor families. Immunol Today 1994;15:321–331. 6. Linsley PS, Wallace PM, Johnson J, Gibson MG, Greene JL, Ledbetter JA, Singh C, Tepper MA. Immunosuppression in vivo by a soluble form of the CTLA-4 T cell activation molecule. Science 1992;257:792–795. 7. Azuma M, Ito D, Yagita H, Okumura K, Phillips JH, Lanier LL,

22.

23.

24.

25.

571

Somoza C. B70 antigen is a second ligand for CTLA-4 and CD28. Nature 1993;366:76–79. Schwartz RH. A cell culture model for T lymphocyte clonal anergy. Science 1990;248:1349–1356. Harper K, Balzano C, Rouvier E, Mattei MG, Luciani MF, Golstein P. CTLA-4 and CD28 activated lymphocyte molecules are closely related in both mouse and human as to sequence, message expression, gene structure, and chromosomal location. J Immunol 1991;147:1037–1044. Linsley PS, Brady W, Urnes M, Grosmaire LS, Damle NK, Ledbetter JA. CTLA-4 is a second receptor for the B cell activation antigen B7. J Exp Med 1991;174:561–569. Jenkins MK, Taylor PS, Norton SD, Urdahl KB. CD28 delivers a costimulatory signal involved in antigen-specific IL-2 production by human T cells. J Immunol 1991;147:2461–2466. Harding FA, McArthur JG, Gross JA, Raulet DH, Allison JP. CD28-mediated signalling co-stimulates murine T cells and prevents induction of anergy in T-cell clones. Nature 1992;356:607– 609. Lenschow DJ, Zeng Y, Thistlethwaite JR, Montag A, Brady W, Gibson MG, Linsley PS, Bluestone JA. Long-term survival of xenogeneic pancreatic islet grafts induced by CTLA4lg. Science 1992;257:789–792. Lin H, Bolling SF, Linsley PS, Wei RQ, Gordon D, Thompson CB, Turka LA. Long-term acceptance of major histocompatibility complex mismatched cardiac allografts induced by CTLA4Ig plus donor-specific transfusion. J Exp Med 1993;178:1801–1806. Kirk AD, Harlan DM, Armstrong NN, Davis TA, Dong Y, Gray GS, Hong X, Thomas D, Fechner JHJ, Knechtle SJ. CTLA4-Ig and anti-CD40 ligand prevent renal allograft rejection in primates. Proc Natl Acad Sci USA 1997;94:8789–8794. Finck BK, Linsley PS, Wofsy D. Treatment of murine lupus with CTLA4Ig. Science 1994;265:1225–1227. Perrin PJ, Scott D, Quigley L, Albert PS, Feder O, Gray GS, Abe R, June CH, Racke MK. Role of B7:CD28/CTLA-4 in the induction of chronic relapsing experimental allergic encephalomyelitis. J Immunol 1995;154:1481–1490. Marsh MN. Gluten, major histocompatibility complex, and the small intestine. A molecular and immunobiologic approach to the spectrum of gluten sensitivity (‘celiac sprue’). Gastroenterology 1992;102:330–354. Lundin KE, Scott H, Hansen T, Paulsen G, Halstensen TS, Fausa O, Thorsby E, Sollid LM. Gliadin-specific, HLA-DQ(alpha 1*0501,beta 1*0201) restricted T cells isolated from the small intestinal mucosa of celiac disease patients. J Exp Med 1993;178: 187–196. Maki M, Hallstrom O, Marttinen A. Reaction of human noncollagenous polypeptides with coeliac disease autoantibodies. Lancet 1991;338:724–725. Picarelli A, Maiuri L, Frate A, Greco M, Auricchio S, Londei M. Production of antiendomysial antibodies upon in vitro challenge of coeliac biopsies. Lancet 1996;348:1065–1067. Dieterich W, Ehnis T, Bauer M, Donner P, Volta U, Riecken EO, Schuppan D. Identification of tissue transglutaminase as the autoantigen of celiac disease. Nat Med 1997;3:797–801. Lin T, Brunner T, Tietz B, Madsen J, Bonfoco E, Reaves M, Huflejt M, Green DR. Fas ligand–mediated killing by intestinal intraepithelial lymphocytes. Participation in intestinal graft-versus-host disease. J Clin Invest 1998;101:570–577. Pender SL, Tickle SP, Docherty AJ, Howie D, Wathen NC, MacDonald TT. A major role for matrix metalloproteinases in T cell injury in the gut. J Immunol 1997;158:1582–1590. Maiuri L, Picarelli A, Boirivant M, Coletta S, Mazzilli MC, De Vincenzi M, Londei M, Auricchio S. Definition of the initial immunologic modifications upon in vitro gliadin challenge in the small intestine of celiac patients. Gastroenterology 1996;110: 1368–1378.

572

MAIURI ET AL.

26. De Ritis G, Occorsio P, Auricchio S, Gramenzi F, Morisi G, Silano V. Toxicity of wheat flour proteins and protein-derived peptides for in vitro developing intestine from rat fetus. Pediatr Res 1979;13: 1255–1261. 27. Alderson MR, Tough TW, Braddy S, Davis Smith T, Roux E, Schooley K, Miller RE, Lynch DH. Regulation of apoptosis and T cell activation by Fas-specific mAb. Int Immunol 1994;6:1799– 1806 Issn: 0953-8178. 28. Maiuri L, Raia V, De Marco G, Coletta S, de Ritis G, Londei M, Auricchio S. DNA fragmentation is a feature of cystic fibrosis epithelial cells: a disease with inappropriate apoptosis? FEBS Lett 1997;408:225–231. 29. Surh CD, Sprent J. T-cell apoptosis detected in situ during positive and negative selection in the thymus. Nature 1994;372: 100–103. 30. Clark EA, Ledbetter JA. How B and T cells talk to each other. Nature 1994;367:425–428. 31. Grewal IS, Foellmer GH, Grewl KD, Xu J, Hardardottir F, Baron JL, Janeway CA, Flavell RA. Requirement for CD40 ligand in costimulation induction, T cell activation, and experimental allergic encephalomyelitis. Science 1996;273:1864–1867. 32. Yang Y, Wilson JM. CD40 ligand-dependent T cell activation: requirement of B7-CD28 signalling through CD40. Science 1996; 273:1862–1864. 33. Callard RE, Armitage RJ, Fanslow WC, Spriggs MK. CD40 ligand and its role in X-linked hyper-IgM syndrome. Immunol Today 1993;14:559–564. 34. Londei M, Bottazzo GF, Feldmann M. Human T-cell clones from autoimmune thyroid glands: specific recognition of autologous thyroid cells. Science 1985;228:85–89. 35. Quaratino S, Feldmann M, Dayan CM, Acuto O, Londei M. Human self-reactive T cell clones expressing identical T cell receptor beta chains differ in their ability to recognize a cryptic self-epitope. J Exp Med 1996;183:349–358. 36. Maki M, Holm K, Lipsanen V, Hallstrom O, Viander M, Collin P, Savilahti E, Koskimies S. Serological markers and HLA genes among healthy first-degree relatives of patients with coeliac disease. Lancet 1991;338:1350–1353.

GASTROENTEROLOGY Vol. 115, No. 3

37. Ferreira M, Davies SL, Butler M, Scott D, Clark M, Kumar P. Endomysial antibody: is it the best screening test for coeliac disease? Gut 1992;33:1633–1637. 38. Grewal IS, Xu J, Flavell RA. Impairment of antigen-specific T-cell priming in mice lacking CD40 ligand. Nature 1995;378:617–620. 39. van Essen D, Kikutani H, Gray D. CD40 ligand-transduced co-stimulation of T cells in the development of helper function. Nature 1995;378:620–623. 40. Durie FH, Fava RA, Foy TM, Aruffo A, Ledbetter JA, Noelle RJ. Prevention of collagen-induced arthritis with an antibody to gp39, the ligand for CD40. Science 1993;261:1328–1330. 41. Grammer AC, Bergman MC, Miura Y, Fujita K, Davis LS, Lipsky PE. The CD40 ligand expressed by human B cells costimulates B cell responses. J Immunol 1995;154:4996–5010. 42. Boise LH, Minn AJ, Noel PJ, June CH, Accavitti MA, Lindsten T, Thompson CB. CD28 costimulation can promote T cell survival by enhancing the expression of Bcl-XL. Immunity 1995;3:87–98. 43. Lundqvist C, Baranov V, Hammarstrom S, Athlin L, Hammarstrom ML. Intra-epithelial lymphocytes. Evidence for regional specialization and extrathymic T cell maturation in the human gut epithelium. Int Immunol 1995;7:1473–1487. 44. Moss SF, Attia L, Scholes JV, Walters JR, Holt PR. Increased small intestinal apotosis in coeliac disease. Gut 1996;39:811–817.

Received November 26, 1997. Accepted June 2, 1998. Address requests for reprints to: Marco Londei, M.D., Kennedy Institute of Rheumatology, Immunology, 1 Aspenlea Road, Hammersmith, London, W6 8LH, England. e-mail: [email protected]; fax: (44) 181-383-4499. Supported by the Arthritis Research Campaign (United Kingdom) and the Associazione Italiana Celiachia. Presented in part at the annual meeting of the American Gastroenterological Association and the American Association for the Study of Liver Diseases, San Francisco, California, May 19–22, 1996, and the Annual Meeting of the European Society for Pediatric Gastroenterology and Nutrition, Munich, Germany, 1996.