Ultrastructural Localization of Concanavalin A Surface Receptors on Brush-Border Enterocytes in Normal Children and during Coeliac Disease

Pediatr. Res. 16: 441-445 (1982) Ultrastructural Localization of Concanavalin A Surface Receptors on Brush-Border Enterocytes in Normal Children and during Coeliac Disease DANIEL VASMANT, GERARD FELDMANN,"'' AND JEAN-LOUP FONTAINE Laboratoire d'Histologie, Embryologic, Cyrogknerique, Faculte de Medecine Xavier-Bichat, 75018 Paris and Service de Pediatric, H6pital Trousseau, 75012 Paris, France Summary Concanavalin A surface receptors were detected on brush border enterocytes from child intestinal biopsies by the concanavalin Aperoxidase method. O n electron microscopy, these receptors appeared as electron-dense deposits, located on a band running along the h e r m o s t part of the brush border membrane g l ~ c o c a l ~ xIn. the five control subjects tested, brush border length was 1-06 0.18 pm, and band thickness, 19 f 4 nm. Deposits were regular in and Out the entire membrane, reflecting g l ~ c o s ~ l a t i oofn the MWmal brush border. In seven ~ a t i e n t swith villous atrophy induced by gluten, the brush border was damaged and its length was 1.04 k 0.39 Pm. The thickness of the electron-dense deposit band was 19 6 nm; deposit shape was irregular and the band ran discontinuously along the membrane. The degree of this abnormality seemed to correspond to the degree of brush border damage. In six treated coeliac patients with normalized mucosa, the brush border was structurally normal but it was significantly longer (1.25 f 0.32 pm) than in the controls ( p < 0.01). The electron-dense band was significantly thinner (16 + 7 nm) than in the controls (p < 0.01). The distribution of the electron-dense deposits was sporadic in some parts of the band and regular in others. These results suggest abnormal glycosylation of the brush border membrane in coeliac disease, and might be due to the presence of abnormal glycoconjugates. It remains to be established if these changes are induced by gluten toxicity or are the consequences of nonspecific intestinal disorders. * * Speculation Clycoconjugates are the most important component of the brush border glycocalyx. During gluten-induced villous atrophy, the most mature enterocytes exhibit an abnormal glycocalyx. W e therefore investigated coeliac and noncoeliac patients by the concanavalin A-peroxidase method in order to examine brush border enterocyte glycoconjugates by electron microscopy. During villous atrophy in children with coeliac disease, both the number of damaged enterocytes and the cell turnover rate increase (27). The mechanisms of these changes are still controversial. No peptidase deficiency has ever been proved (19), but most authors agree that a local immunological reaction is involved (1 1, 22, 24). However, the way in which gluten starts this reaction is not known. Recently, Douglas (3) showed that gluten adherence increased on isolated brush border (BB) from coeliac intestinal biopsies. Weiser and Douglas (26) observed a decline in glycosyltransferase activity of the same BB. On the basis of these findings both authors suggested the presence of an abnormal glycoprotein on the BB membrane in coeliac disease (25). Earlier conventional ultrastructural studies had already revealed the presence of abnormal microvilli during gluten-induced villous atrophy (1,20,21, 23), but the abnormalities disappeared when the villi were nor- malized by a completely gluten-free diet (21, 23). Nevertheless, to our knowledge, no ultrastructural cytochemical study of the BB membrane has been yet made for the purpose of investigating g~ycocon,ugates~ with this view, we applied the concanavalin A-peroxidase method to biopsies from children displaying no clinical or histologic evidence of small bowel involvement, and from coeliac patients at various stages of their illness. Concanavalin A (Con A) is a lectin from Canavalia Enseformjs, which binds to a-~-glucopyranosyl,a-~-fructopyranosyl,and a-D-manopyranosyl (6). Covalent linkage between Con A and horseradish peroxion electron dase was easily obtained (2). Peroxidase was microscopy by the Graham and Karnovsky technique (7). Electron-dense deposits showed the distribution of the above three carbohydrates involved in the formation of BB membrane glycoproteins and glycolipids (9, 12). MATERIALS AND METHODS I Patients. Three groups patients were studied. prised five patients who acted as controls. All biopsies were from patients with Or malabsOr~gives tion and were On light patients' age and the for their Group I1 consisted of seven patients with active coeliac disease and villous atrophy (VA), whose diet included gluten. The degree of VA was assessed by light microscopy and was found to be subtotal in four cases and partial in three times. Coeliac disease (CD) was diagnosed on the basis of the E.S.P.G.A.N. criteria (14). When the disease was discovered after the age of 2 years, the gluten challenge was not always performed, but VA disappeared after 1 year on a gluten free diet, thus confirming the diagnosis of C D (4). Table 2 summarizes the characteristics of group 11patients. Group 111 included six CD patients on a completely gluten-free diet displaying normal mucosa on light microscopy. In this group, some biopsies were analyzed after an initial period of gluten-free (from 6 months to 3 years 9 months) and others, after a second period of such diet (from 1 year to 1 year 7 months), following a positive gluten challenge. C D diagnosis was based on the same criteria as those applied to group 11. Table 3 summarizes the features of group 111 patients. Me:hods. Intestinal biopsies were collected by Crosby capsule. Radioscopic checking of the capsule's position ensured that it remained at Treitz's angle. Each biopsy was divided into two parts. One part was fwed for light microscopy in Bouin's fluid and embedded in paraffin; 5 pm-thick sections were cut and stained with hematein-phloxine-safran and PAS. The other part of the biopsy was processed for ultrastructural cytochemical study as previously described (8). The intestinal tissue was futed at +4OC for 2 hr in a solution of 2.5% glutaraldehyde, buffered with 0.1 M phosphate buffer, pH 7.3, and washed in 0.1 M phosphate buffer, with a razor pH 7.3. The fragment was cut into 1 mm"1ocks 44 1 443 COELIAC DISEASE Table 4. Electron microscopic findings in groups I, II and I I I Length of microvilli (pm) (mean rf: S.D.) Total number of values' Group I (control patients) I1 (patients with villous atrophy) 111 (patients with normalized mucosa) * 0.18 250 1.06 350 + 0.392 1.25 + 0.323 1.04 300 Thickness of electron-dense deposits (nm) (mean f S.D.) 19 +4 19 rt 6' 16 +. 73 Appearance of electron-dense deposits Regularly in shape, evenly spaced out over the entire membrane Irregular in shape unevenly spaced out over the membrane Irregular and discontinuous in some parts of the deposit band, regular and continuous in others ' Five measurements were made on one enterocyte; 10 enterocytes were studied for each patient. Fifty (...truncated)