Small-bowel mucosal biopsies were taken from 20 treated patients suffering from celiac disease (median age 54 years, range 23–73 years, females 80%), five untreated celiac patients (median age 48 years, range 43–71 years, females 100%) and six non-celiac control subjects suffering from dyspepsia (median age 53 years, range 24–70 years, females 50%). In all celiac patients the diagnosis was initially based on the European Society of Pediatric Gastroenterology and Nutrition criteria , meaning that they all had small-bowel mucosal villous atrophy with crypt hyperplasia in the duodenum while consuming gluten. All treated celiac patients involved in the current study had been on a strict gluten-free diet for at least one year (median duration of GFD three years, range 1–20 years), and all, as well as the non-celiac controls, showed normal small-bowel mucosal architecture. In contrast, all the untreated patients evinced subtotal villous atrophy with crypt hyperplasia in the small-bowel mucosa. The study protocol was accepted by the Ethical Committee of Tampere University Hospital and written informed consent was obtained from all patients and controls.
Small-bowel mucosal biopsies and organ culture system
Altogether seven small-bowel mucosal biopsy samples were obtained from each patient during upper gastrointestinal endoscopy. Two samples were immediately snap-frozen in liquid nitrogen with optimal cutting temperature compound (OCT, Tissue-Tek, Sakura Finetek Europe, Holland) and stored at -20°C until used. Further, two biopsies were immediately fixed in paraffin for investigation of the baseline small-bowel mucosal morphology. The remaining three biopsies were cultured for 24 or 48 hours at 37°C, either in the presence or absence of a peptic-tryptic digest of gliadin (1 mg/ml) prepared by a standard protocol described elsewhere [15, 21].
The organ culture method was implemented as originally described by Browning and Trier . Briefly, the biopsy samples were cultured in RPMI-1640 medium (Invitrogen-Gibco, Paisley, Scotland, UK) containing 15 % heat-inactivated fetal bovine serum (Invitrogen-Gibco), 100 μg/ml streptomycin (Invitrogen-Gibco), 100 U/ml penicillin (Invitrogen-Gibco), 4 mM L-glutamine (Invitrogen-Gibco), 50 μg/ml insulin (Sigma-Aldrich Co, St. Louis, Missouri, USA) and 10 mM HEPES buffer (Invitrogen-Gibco). Duodenal specimens were placed villi upwards on a sterile stainless-steel grid positioned over the medium in a central well of the organ culture dish (Falcon, Becton Dickinson and Co, USA). After 24 or 48 hours' incubation, culture supernatants were collected and stored at -70°C until analysed. Free fluid was removed from the samples, whereafter they were snap-frozen with OCT and stored at -20°C until processed for stainings.
EmA was detected in patients' serum and undiluted organ culture supernatants using an indirect immunofluorescence assay where human umbilical cord was used as antigen . A serum dilution of 1:≥5 was considered positive. Antibody titers for organ culture supernatants were graded according to the intensity of the staining as follows: negative (neg), weak positive (+) and strong positive (++, +++ or ++++). Samples were analyzed blindly without knowledge of the patients' clinical background.
TG2-antibodies were measured by enzyme-linked immunosorbent assay (ELISA, Celikey®, Phadia, Freiburg, Germany), according to manufacturer's instructions, both in serum samples (diluted 1:100) and in undiluted culture supernatants. In serum samples a unit value (U) ≥ 5U was considered positive. Since there is no cut-off value for TG2-antibody in culture supernatants, the crude antibody values are given only for comparison to EmA.
Small-bowel mucosal TG2-specific IgA deposits
The small-bowel mucosal TG2-targeted IgA deposits were investigated before and after 24 hours of organ culture. In earlier studies it has been shown that these mucosal IgA deposits are specifically targeted against TG2 in the small-bowel mucosa [3, 6]. In the studies in question, autoantibody specificity for TG2 was demonstrated by the fact that IgA eluted from duodenal mucosa bound intensively to purified TG2 in ELISA and Western blot . Similarly, a human recombinant TG2 was capable of depositing celiac-specific IgA in small-bowel sections from celiac disease patients . In addition, after removal of TG2 from the sections by a specific acid, both TG2 and IgA deposits disappeared from the mucosa .
To study the mucosal IgA deposits the 5-μm-thick unfixed cryostat sections were stained using a two-color immunofluorescence method as previously described . The monoclonal primary antibody against TG2 (Clone CUB 7402, NeoMarkers, Fremont, USA, 1:200) was used followed by the rhodamine-conjugated antimouse immunoglobulin antibody (Dako, A/S, Glostrup, Denmark 1:120) and the fluorescein isothiocyanate-conjugated rabbit antibody against human IgA (Dako, 1:40). In untreated celiac disease a clear extracellular subepithelial IgA deposition can be found below the basement membrane along the villous and crypt epithelium and around mucosal vessels; this is in contrast to non-celiac normal small-bowel samples, where IgA is detected only inside plasma and epithelial cells [4, 6, 22].
Determination of ECH and the number of CD25+ lymphocytes
ECH was measured under a light microscope (Olympus BX60, 40× magnification) after 24 hours' organ culture with or without PT-gliadin challenge using the analySIS 3.0 program, (Soft Imaging System GmbH, Munster, Germany). Altogether 30 enterocytes from three different villi of each specimen were analyzed and a mean ECH value was calculated for each biopsy sample .
CD25-positive T cells were detected in the lamina propria of small-bowel mucosa from biopsy samples cultured for 24 hours with or without PT-gliadin. The 5-μm-thick cryostat sections were fixed in acetone and incubated with goat normal serum (Vector Laboratories Inc., Burlingame, USA), whereafter they were incubated with mouse monoclonal antibody, human anti-CD25 (Dako, 1:25) for one hour and alexa-conjugated goat anti-mouse IgG (Invitrogen, 1:1000) for 30 minutes. Washes with PBS were performed between each antibody. The density of small-bowel mucosal CD25-positive T-cells in the lamina propria was calculated and presented as number of cells in a total area of one mm2 of lamina propria [12, 13].
Celiac disease is strongly associated with the HLA gene region, since over 95% of celiac disease patients have either HLA DQ2 or HLA DQ8 haplotype molecules [23, 24]. HLA-DQ typing was performed in each patient using DELFIA® Celiac Disease Hybridization Assay (PerkinElmer Life and Analytic Sciences, Wallac Oy, Turku, Finland).
Statistical analysis was performed using 2-tailed Wilcoxon Signed Ranks Test or Mann-Whitney U Test, as appropriate. P values lower than 0.05 were considered statistically significant.