| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Technical Briefs |
1 Department of Reproductive and Development Science and Service of Dentistry, I.R.C.C.S. "Burlo Garofolo", and
2 Department of Biology, University of Trieste, Trieste, Italy;
aaddress correspondence to this author at: Clinica Pediatrica, Istituto per l’Infanzia "Burlo Garofolo", Via dell’Istria 65/1, IT-34100 Trieste, Italy; fax 39-040-3785-210, e-mail not{at}burlo.trieste.it
Undetected celiac disease (CD) can cause serious complications and excessive mortality (1)(2)(3). Noninvasive tests could be useful in physicians’ offices to identify patients for intestinal biopsies (4). Serum IgA anti-endomysium (AEA) and anti-human transglutaminase (anti-htTG) antibody assays are the most widely used laboratory tests for CD (5). Autoantibody determinations have been simplified by the use of rapid testing (6)(7), but when saliva specimens were evaluated for AEA testing, they were found to be not suitable (8).
To evaluate the potential utility of saliva as a sample for anti-htTG testing, we compared salivary IgA anti-htTG (measured by ELISA with a dot-blot method and an AEA test) with serum IgA anti-htTG antibodies (measured by ELISA and an AEA test) in patients with CD and in controls. Salivary IgA anti-htTG activity was analyzed for the IgA secretory chain by use of a monoclonal antibody. The recognition patterns of the saliva samples were compared with those for the serum samples and with the pattern for a gut-derived IgA monoclonal antibody to htTG from one CD patient, using two transglutaminase deletion mutants, one recognized and one not by the patients’ sera (9).
We studied 47 untreated CD patients (29 females and 18 males; median age, 19 years; range, 2–52 years) diagnosed in 2002–2003 according to ESPGHAN criteria (10) and 47 patients with celiac disease (26 females and 21 males; median age, 19 years; range, 4–50 years) on a gluten-free diet (GFD) for at least 12 months. We recruited 51 healthy controls with no history of gastrointestinal or autoimmune diseases (25 females and 26 males; median age, 14 years; range, 6–50 years) and identified samples from 49 individuals suffering from various gastrointestinal diseases (18 with Crohn disease, 10 with ulcerative colitis, 10 with milk allergy, 6 with acute diarrhea, 5 with gastroesophageal reflux). Of these 49 individuals, 11 had undergone intestinal biopsies showing nonceliac mucosa.
Serum IgA anti-htTG antibodies were measured by ELISA (11) in plates coated with tTG (1 µg/well). Serum samples diluted 1:100 were incubated for 1 h, and then 1 h more with anti-human IgA. Absorbance was read at 405 nm. The results were expressed as percentages of the positive control serum. The cutoff for IgA was set at <16%, >2 SD above the mean of 400 healthy individuals. Serum IgA AEA concentrations were measured by indirect immunofluorescence on human umbilical cord sections as described previously (11).
Whole unstimulated saliva (1–2 mL) was collected from each participant and centrifuged; the supernatant was stored in aliquots at -80 °C and thawed once before testing. Saliva samples were taken from four untreated patients and four healthy controls and used as positive and negative specimens for setting up the ELISA. Plates were coated with tTG (0.5 or 1 µg/well). The buffer solution (BS) contained 1, 10, 30, or 50 mL/L Tween 20, and saliva specimens were undiluted or diluted 1:1 and 1:4. Plates were washed and incubated for 1 h with phosphatase-conjugated anti-human IgA diluted 1:4000 (cat. no. A-3062; Sigma). Results are expressed as percentages of the positive control serum. The assay conditions chosen were those with the greatest differences between patients and controls.
The secretory form of salivary IgA anti-htTG was assayed by ELISA with a monoclonal anti-human IgA secretory chain antibody (cat. no. MCA1129; Serotec) diluted 1:500. To measure salivary AEA, samples were diluted 1:1 before assay with BS containing 50 mL/L Tween 20.
A dot-blot test was used to detect salivary anti-htTG antibodies (6). Samples were diluted 1:1 in BS containing 50 mL/L Tween 20, and then incubated for 5 min with tTG-activated squares of Immobilon-P membrane. The membrane squares were then washed and incubated for 5 min with anti-human IgA, and the immunocomplexes were revealed by substrate solution. Tests were performed by four operators blinded to the clinical and laboratory findings.
A phage display library of single-chain antibody fragments (scFv) was constructed (12). B lymphocytes from the intestinal biopsy of an untreated CD patient were used as a cDNA source for antibody cloning. The VH and VL immunoglobulin variable regions were amplified from the cDNA by use of V gene oligonucleotides that recognize human IgA V genes. Phages expressing scFv to tTG were selected by affinity chromatography.
The tTG-derived fragments were cloned and expressed as reported previously (9). In the present study we used one tTG-related fragment recognized by sera from CD patients (fragment 1; amino acids 1–376; 45 kDa) and one not recognized (fragment 2; amino acids 269–687; 60 kDa). Plates were coated with fragments (1 µg/well) for testing with the ELISA. Serum and saliva samples were diluted and tested as described above. The human monoclonal antibody was diluted as described previously (9). Results are expressed as absorbance.
Statistical comparison were performed with the Mann–Whitney test. The inter- and intraassay imprecisions (CV) of the ELISA were estimated from three replicates of four samples containing low (5–9%), intermediate (22%), and high (59%) antibody concentrations on 3 consecutive days.
The results for plates coated with 0.5 µg/well of tTG and incubated with saliva samples diluted 1:1 with BS plus 50 mL/L Tween 20 were those that best separated patients’ samples from controls (Table 1
). We observed no differences in the ELISA results when undiluted or 1:4-diluted saliva samples were used (data not shown). The intra- and interassay CV of this ELISA were 8.4% and 7.5%, respectively, for the samples with low antibody concentrations, 6.5% and 4.3% for samples with intermediate antibody concentrations, and 4.7% and 3.0% for the samples with high antibody concentrations. At a cutoff of 19.85%, ROC curve analysis gave a sensitivity of 87% (41 of 47; 95% confidence interval, 74–95%) and specificity of 91% (47 of 51 healthy controls and 44 of 49 sick controls; 95% confidence interval, 83–95.7%). Saliva samples were positive in 23 of 47 CD patients on GFD (Fig. 1A
). All six saliva specimens selected with different values of anti-htTG (range, 25–65%) displayed the IgA secretory chain, whereas all serum samples tested negative.
|
|
The serum samples were positive for both anti-htTG and AEA in all 47 untreated CD patients and 5 of 47 treated patients, but were negative in controls. No salivary AEA was detected in patients or controls. We used AEA to test the reactivity of two saliva samples with ELISA values for anti-htTG of 76% and 56% and two serum samples with values of 78% and 59%. The sera tested positive up to a 1:100 dilution, whereas the saliva specimens tested negative at a 1:1 dilution.
We used the dot-blot test during clinical examination and investigated a subset of saliva samples. This method was positive in 17 of 18 untreated CD patients, in 1 of 20 healthy controls, and 1 of 6 sick controls (Fig. 1A
).
The tTG fragments were tested with 9 randomly selected saliva samples that tested positive for IgA anti-htTG antibody (median, 29.7%; range, 20–35%), with 20 saliva samples negative for IgA anti-htTG antibody (median, 4.5%; range, 0.7–10%), with sera from 4 of the 9 CD patients and 3 of the 20 corresponding controls, and with the gut-derived monoclonal anti-tTG. The patients’ saliva samples recognized both tTG fragments with significantly higher values than controls (P = 0.0001). The four patient serum samples and gut-derived monoclonal anti-tTG from the CD patient recognized only tTG fragment 1, whereas the three control sera recognized neither (Fig. 1B
)
Although the saliva of untreated CD patients contains IgA secretory chain anti-htTG, the saliva-based ELISA was disappointing; had we used saliva specimens alone, six individuals with CD would have been missed. Furthermore, the availability of simple and highly sensitive assays that use a drop of blood (sensitivity, 100%; specificity, 98–100%) (6)(7) makes use of the saliva-based dot-blot test unnecessary.
The failure of saliva specimens with high concentrations of anti-htTG antibodies in the ELISA to identify tissue transglutaminase on umbilical cryosections may indicate that salivary IgA forms respond to transglutaminase epitopes not involved in the immune response of the gut, the site of synthesis of circulating anti-htTG antibodies (12). We have demonstrated that both sera from untreated CD patients and gut-derived tTG monoclonal antibodies recognize only selected tTG fragments (9). In the present study, saliva samples (but not sera) from untreated CD patients recognized both fragments. The gut-derived monoclonal antibodies to tTG from one individual with positive salivary results reacted specifically with fragment 1, confirming a dichotomy between the aerodigestive tract and the gut in the gluten-dependent autoimmune response. This supports earlier findings that saliva does not reflect immune events occurring in the intestine (13).
Oral duct-associated lymphoid tissue may play an independent role in the CD-related autoimmune response by offering a route for gliadin-derived peptides to activate B cells (14). Instead of migrating to the gut, these primed immune cells tend to home in on the oral mucosal site where they were initially stimulated (15). This could explain the increased number of T and B lymphocytes in the oral mucosa of CD patients testing negative by AEA, even in the presence of strict GFD or during supramucosal gliadin challenge (16)(17). The high number of treated patients positive only for salivary autoantibodies may be explained by oral duct-associated lymphoid tissue activation in response to amounts of gluten too small to induce gut activation. Our highly compliant treated patients probably have a detectable concentration of anti-htTG antibodies only in the gut, and the salivary response is a localized event.
In conclusion, because of its low sensitivity, saliva is not a useful source of CD-related autoantibodies for diagnostic purposes. However, salivary IgA anti-htTG antibodies may offer promise in studies on differences between the aerodigestive tract and the gut in the gluten-dependent mucosal immune response. In particular, the phage display library technique (12) allows study of the differences between autoantibodies to tTG originating from the intestinal and oral mucosa of CD patients with respect to antibody V chain family usage and antigen-driven somatic hypermutation.
Acknowledgments
This study was supported by research grants from the Italian Ministry of Health (RF174/02), from MIUR (cofin 2001068143_004), and from Telethon Foundation (E1141).
References
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  L. Pastore, G. Campisi, D. Compilato, and L. L. Muzio Orally Based Diagnosis of Celiac Disease: Current Perspectives Journal of Dental Research, December 1, 2008; 87(12): 1100 - 1107. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
) and controls (
) and in serum samples of patients (
) and controls (•). 
, reactivity of intestine-derived monoclonal antibody from a CD patient to tTG. The solid and dashed lines represent the median values for the control and patient samples, respectively. Pos, positive; Neg, negative.