| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Articles |
1
Department of Dermato-Venereology, Semmelweis University of Medicine, H-1085 Budapest, Mária u. 41, Hungary.
2
Institute for Biochemistry II, Medical Faculty,
University of Cologne, Joseph-Stelzmann-Strasse 52, D-50931 Cologne,
Germany.
a Address correspondence to this author at: Institute for Biochemistry II, Medical Faculty, University of Cologne, Joseph-Stelzmann-Strasse 52, D-50931 Cologne, Germany. Fax 49-221-478-3109; e-mail Miklos.Sardy{at}Uni-Koeln.DE
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsDiscussionReferences |
|---|
Methods: Human TGc was expressed in the human embryonic kidney cell line 293-EBNA as a C-terminal fusion protein with the eight-amino acid Strep-tag II allowing one-step purification via streptavidin affinity chromatography. We carried out ELISA assays for IgA antibodies against TGc using calcium-activated human and guinea pig TGc. The sera were also tested on monkey esophagus sections by indirect immunofluorescence for IgA EMA. We examined 71 serum samples from patients with GSE (38 with celiac disease, 33 with dermatitis herpetiformis), including 16 on therapy, and 53 controls.
Results: The human TGc could be expressed and purified as an active enzyme giving a single band on a Coomassie-stained gel. The mean intra- and interassay CVs for the human TGc ELISA were 3.2% and 9.2%, respectively. The area under the ROC curve was 0.999. The specificity and sensitivity were 98.1% (95% confidence interval, 95.7–100%) and 98.2% (95.9–100%), respectively.
Conclusions: The human TGc ELISA was somewhat superior to the guinea pig TGc ELISA, and was as specific and sensitive as the EMA test. The human TGc-based ELISA is the method of choice for easy and noninvasive screening and diagnosis of GSE.
|
Introduction |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
The diagnosis of GSE is based on characteristic histological changes (villous atrophy, intraepithelial lymphocytosis, and crypt hyperplasia) seen in jejunal biopsies, followed by the regeneration of the mucosa after a gluten-free diet and relapse during subsequent gluten challenge (1)(2). However, serological tests may be helpful in the diagnosis of GSE because they offer a less invasive and less expensive alternative. These detect IgA antibodies directed against endomysial antigen, reticulin, or gliadin. IgG antibody-based tests are not sensitive for GSE in patients with normal serum IgA concentrations [Sárdy et al., unpublished observations, and Ref. (3)]. However, IgG-based tests are performed for the diagnosis in patients with selective IgA deficiency, which is known to be associated with GSE (3)(4). The IgA-class endomysial antibody (EMA) test is considered the serological method of choice because of its higher sensitivity and specificity when compared with the IgA-class anti-reticulin antibody and the IgA-class anti-gliadin antibody tests (5)(6)(7). EMAs are found in 60–70% of untreated patients with DH (8)(9) and in almost all untreated patients with CD (9). However, because the EMA test usually is performed on expensive esophagus sections from endangered primates, is labor-intensive and time-consuming, and is subjective in borderline cases, a demand for alternative test methods exists. The monkey esophagus sections can be replaced by umbilical cord sections (10) or immortalized human umbilical vein endothelial cells (11), eliminating one of the major disadvantages of the classical EMA test without loss of performance of the test.
In 1997, Dieterich et al. (12) identified tissue transglutaminase (TGc; EC 2.3.2.13) as the predominant, or perhaps the sole, endomysial autoantigen of CD. Later studies showed that TGc may also be the autoantigen of EMA-positive patients who have DH (13). An ELISA test for CD has been produced based on the commercially available guinea pig TGc (12). Although the amino acid sequence identity between guinea pig and human TGc is 82.8% (14)(15), after optimization using calcium activation, this test gave high sensitivity and specificity (>90%) (16)(17). However, the question has been raised whether the performance of the test can be further improved by using the human TGc as antigen because patient sera not recognized by the guinea pig TGc ELISA may have antibodies directed against epitopes of human TGc not conserved in the guinea pig enzyme.
To answer this question, we expressed the human TGc recombinantly and set up an ELISA based on the purified protein for detecting IgA anti-TGc antibodies. The results of this assay were compared with those from the ELISA with the guinea pig TGc and the EMA test on monkey esophagus.
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
recombinant expression of human TGc
The episomal eukaryotic expression vector pCEP-Pu/BM40SP, produced
from pCEP4 (Invitrogen) (20), was modified to introduce a
sequence encoding the Strep II tag (Institut für Bioanalytik) and
a stop codon into the multiple cloning site. The primers
5'-GGCCGCATGGAGCCATCCACAATTCGAAAAGTA-3' and
5'-GGCCTACTTTTCGAATTGTGGATGGCTCCATGC-3' were annealed together and
introduced into the NotI site, thus constructing a vector
(pCEP-Pu/BM40SP/C-Strep) that produces a C-terminal Strep II fusion
protein suitable for streptavidin affinity purification by a
StrepTactin® (Institut für Bioanalytik)
affinity column as described previously (21). We received
the full-length human TGc cDNA (GenBank accession number M55153, cloned
in pSP73) from Dr. Daniel Aeschlimann, Division of Orthopaedic Surgery,
University of Wisconsin, Madison, Wisconsin. This cDNA was amplified by
PCR using the 5' primer
5'-ATTAAGCTTGCCGCCACCATGGCCGAGGAGCTGGTC-3', and the 3' primer
5'-TAAGCGGCCGCGGGGCCAATGATGACATTC-3'. The 5' primer introduced a
new HindIII restriction site and a Kozak’s translation
initiation sequence; the 3' primer inserted a new NotI
restriction site and removed the stop codon. The
HindIII/NotI restriction enzyme-digested PCR
product was purified and inserted at the same restriction sites of the
pCEP-Pu/BM40SP/C-Strep to obtain the final expression vector
pCEP-Pu/TGc/C-Strep. The correct insertion and sequence of the full
construct were verified by cycle sequencing with the ABI Prism Big Dye
Terminator Cycle Sequencing Ready Reaction Kit, and the products were
resolved on an ABI Prism 377 Automated Sequencer (Perkin-Elmer/Applied
Biosystems).
Human embryonic kidney cells (293-EBNA; Invitrogen) were transfected with pCEP-Pu/TGc/C-Strep and harvested in cell culture in Dulbecco’s MEM NUT MIX F-12 (Life Technologies) medium containing 100 mL/L fetal bovine serum (Life Technologies), 10 g/L L-glutamine (Life Technologies), 200 kilounits/L penicillin (Life Technologies), and 200 mg/L streptomycin (Life Technologies). Cells were selected with 0.5 mg/L puromycin (Sigma). After the medium was removed and the cells washed with cold (4 °C) 0.25 mol/L sucrose, the cells were lysed mechanically in cold 0.25 mol/L sucrose. The lysate was cleared of particulate material by centrifugation at 27 200g for 30 min at 4 °C, followed by ultracentrifugation of the supernatant at 210 000g for 60 min at 4 °C. The supernatant was filtered with cheesecloth and 1 mmol/L phenylmethylsulfonyl fluoride (Fluka) was added as proteinase inhibitor; 12 mL of the supernatant was then passed over a StrepTactin affinity column of 3 mL volume equilibrated with sterile filtered 50 mmol/L Tris-HCl, pH 7.5, containing 1 mmol/L EDTA at 4 °C and at a flow rate of 0.4 mL/cm2 per minute. After extensive washing with equilibration buffer containing 1 mmol/L phenylmethylsulfonyl fluoride at a flow rate of 0.9 mL/cm2 per minute, the protein was eluted with equilibration buffer containing 1 mmol/L phenylmethylsulfonyl fluoride and 2.5 mmol/L desthiobiotin (Sigma) at a flow rate of 0.4 mL/cm2 per minute. Fractions (2 mL) were collected. The purification was controlled by Coomassie-stained SDS-PAGE and immunoblotting with monoclonal antibodies against TGc as described above. The protein concentration was estimated by SDS-PAGE and measured by the bicinchoninic acid protein assay reagent (Pierce) according to the protocol provided by the supplier, with bovine serum albumin as the calibrator.
TGc ACTIVITY ASSAY
TGc activity was measured by incorporation of
[1,4-1
H]putrescine (Amersham) for 30 min at
37 °C as described previously (22), with the only
difference being that the buffer contained 22.5 mmol/L dithiothreitol
to reduce any oxidized sulfhydryl groups important for catalytic
activity.
mass spectrometry
Mass spectrometry was performed by matrix-assisted laser
desorption using a Bruker Reflex III instrument equipped with a high
mass detector for linear detection. Sinapinic acid was used as the
matrix, and external calibration was carried out using singly, doubly,
and triply charged molecular ions of protein A.
sera and patients
The patients had been examined at the Gastroenterological
Departments of Internal Medicine or Paediatrics and the Department of
Dermato-Venereology of the Semmelweis University. The CD diagnosis was
confirmed by jejunal biopsy, whereas DH was confirmed by skin biopsy.
Serum samples were taken from 71 patients with GSE (33 with DH and 38
with CD), 26 with non-CD gastrointestinal diseases (such as Crohn
disease, food hypersensitivity, food intolerance, intestinal infection,
reflux esophagitis, non-CD diarrhea, and alimentary dystrophy), and 27
with other diagnoses, such as autoimmune diseases (systemic lupus
erythematosus and diabetes mellitus type I), different skin disorders
(pemphigus foliaceus, ichthyosis, and urticaria), cholelithiasis,
hepatosplenomegaly, retarded growth of other than gastrointestinal
orgin, as well as healthy controls. The mean ages and sexes of the
patient groups are presented in Table 1
. To obtain data on the sensitivity of the TGc ELISA, we
included in the current study sera from 16 treated patients (patients
on gluten-free diet). All serum samples were stored at -78 °C until
assayed.
|
ema test
Serum IgA antibodies were measured by an indirect
immunofluorescence method (3). All serum samples were
diluted 1:5 in phosphate-buffered saline (pH 7.4). Cryostat tissue
slides (10 µm) of the aboral part of the monkey
(Cercopithecidae family) esophagus were used as antigen.
Bound IgA was detected by 
-chain-specific, fluorescein
isothiocyanate-conjugated rabbit anti-human IgA antibodies (1:40 in
phosphate-buffered saline; Dako). All sera used in this study were
indisputably negative or positive for IgA EMA.
elisa
The ELISA method was similar to the calcium-activated test
described previously (16)(17). Briefly, 96-well
microtiter plates (Nunc MaxiSorp) were coated with 1 µg of guinea pig
TGc (Sigma) or human TGc in 100 µL of 50 mmol/L Tris-HCl, pH 7.5,
containing 5 mmol/L CaCl2 per well at 4 °C
overnight (at least 9 h). No blocking was used. After each step,
the wells were washed with 50 mmol/L Tris-HCl containing 10 mmol/L EDTA
and 1 mL/L Tween 20 (TET). Sera were diluted to various concentrations
with TET and incubated on the plates for 1.5 h at room
temperature. Bound IgA was detected by peroxidase-conjugated antibody
against human IgA (Dako), diluted 1:4000 in TET and incubated for
1 h at room temperature. The color was developed by the addition
of 100 µL of 60 mg/L 3,3',5,5'-tetramethylbenzidine substrate in 100
mmol/L sodium acetate, pH 6.0, containing 0.15 mL/L
H2O2 for 5 min at room
temperature. The reaction was stopped by adding 100 µL of 200 mL/L
H2SO4. The absorbance was
read in an ELISA reader at 450 nm.
The amount of protein and the serum concentrations used in the test were optimized. All serum samples were examined in triplicates, and triplicates of a negative and a positive reference serum as well as a buffer blank were included in each assay. The antibody concentrations were expressed in arbitrary units (AU), i.e., as percentages of the positive reference serum.
To obtain data on the effects of calcium activation, an experiment in which wells were coated with human TGc without CaCl2 in the coating buffer was also performed.
statistics
Absorbances (and thus titers given in AU values) did not show
gaussian distribution; thus for statistical description of titers from
the different patient groups, we present medians with their 95%
confidence intervals (95% CIs) (23), and for comparison,
the Mann–Whitney nonparametric, unpaired, two-tailed test was used
(24). For describing correlation of titers, the Spearman
correlation coefficient with its 95% CI and correlation analysis for
unpaired data of nongaussian distribution were used
(23)(24). For comparison of titers in the
calcium-activated and unactivated human TGc ELISA, the Wilcoxon
two-tailed signed-rank test was performed (24). For
description and comparison of the two ELISA systems, the ROC curves and
the areas under the ROC curves (AUCs) with their 95% CIs are presented
(25)(26)(27). For calculating confidence intervals of the AUCs,
a bootstrap technique, the bias-corrected and accelerated
(BCa) confidence interval method
(26)(27), was applied in addition to the most
frequently used method (25) because it is more appropriate
for describing confidence intervals of AUCs that are very close to the
maximum (1.0).
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
A) when visualized by Coomassie-stained SDS-PAGE. Immunoblot
analysis showed that the band reacted with monoclonal antibodies
against TGc (Fig. 1B
). The column bound almost all of the tagged
protein with no immunoreactivity appearing in the flow through (Fig. 1B
). The yield from the lysate of a confluent cell monolayer in a cell
culture dish of 13 cm diameter was ~200 µg. The molecular mass
calculated from the sequence of the human TGc is 77.3 kDa, and the
calculated molecular mass of the fusion protein (TGc having a
C-terminal tag of 10 amino acids) is 78.4 kDa. Mass spectrometry of the
fusion protein gave a molecular mass of 78.3 kDa. In cell lysates, the
activity of the expressed human TGc was 4.7-fold higher than the
background activity of transglutaminases present in untransfected
293-EBNA cells. The freshly purified human TGc showed similar or higher
activity than the guinea pig TGc from Sigma, but it lost activity on
storage. It is not known whether the human TGc has the same catalytic
activity as the guinea pig enzyme.
|
performance of the human TGc ELISA
The optimal coating concentration of human TGc was 1 µg/well.
Using highly positive sera from four patients for calibration, we
obtained a log-linear curve between dilutions of 1:250 and 1:32 000.
Four negative sera showed some signal at lower dilutions (>1:500).
Some positive sera showed a signal plateau at dilutions of 1:250 or
less. The ratio between the mean absorbance values of positive and
negative results at the dilution of 1:125 was 1:6, whereas at higher
dilutions, it was >1:10. Hence, in the assay a serum dilution of 1:250
was used. One positive and one negative reference serum sample was
included in each assay to control the test performance. The positive
serum was used as the "standard", and the absorbance results were
given as AU, calculated as a percentage of the standard serum. The mean
intra- and interassay CVs for the positive standard serum were 1.3%
and 14%, respectively. The mean intra- and interassay CVs (using serum
titers given in arbitrary units) for the other sera tested in the human
TGc ELISA were 3.2% (n = 124) and 9.2% (n = 15),
respectively. The median antibody concentration was 61.4 AU (95% CI,
45.1–78.5 AU; n = 55) for patients with untreated GSE (CD or DH)
and 12 AU (95% CI, 10.8–13 AU; n = 53) for controls; the
difference was significant (P <0.0001). The median antibody
concentration was 48.1 AU (95% CI, 20.8–85.6 AU; n = 16) for
treated patients, 12.1 AU (95% CI, 9.8–14.7 AU; n = 26) for
controls with gastrointestinal diseases, and 12 AU (95% CI, 10.7–13.0
AU; n = 27) for healthy individuals and controls with other
diagnoses. The area under the ROC curve was 0.999 (95% CI,
0.996–1.001; 95% CI with BCa method,
0.990–1.0; Fig. 2
).
|
A cutoff value of 18 AU was chosen, and sera with antibody
concentrations 
18 AU were labeled as human TGc ELISA positive. This
cutoff value gave a specificity and a sensitivity of 98.1% (95% CI,
95.7–100%) and 98.2% (95% CI, 95.9–100%), respectively (treated
patients were excluded). The coincidence of the human TGc assay with
the clinical diagnosis (excluding treated patients) was 106 of 108
(98.1%), giving one false-positive and one false-negative result (Fig. 2
).
performance of the guinea pig TGc ELISA
The optimal coating concentration of guinea pig TGc was 1
µg/well, and the optimal serum dilution was 1:250, as with the human
TGc ELISA. Each assay was performed parallel to the human TGc assay at
the same time, and the same serum samples and serum dilutions were
used. The mean intra- and interassay CVs of the positive standard serum
were 2.2% and 9.0%, respectively. The intra- and interassay CVs
(using serum titers given in arbitrary units) for the other sera tested
in the guinea pig ELISA were 2.8% (n = 124) and 13% (n =
15), respectively.
The median antibody concentration was 51.8 AU (95% CI, 34.2–63 AU;
n = 55) for the patients with untreated GSE (CD or DH) and 8 AU
(95% CI, 7.3–8.9 AU; n = 53) for controls; the difference was
significant (P <0.0001). The median antibody concentration
was 18 AU (95% CI, 9.2–69.9 AU; n = 16) for treated patients,
7.5 AU (95% CI, 6.6–9 AU; n = 26) for controls with
gastrointestinal diseases, and 8.5 AU (95% CI, 7.2–10.3 AU; n =
27) for healthy individuals and controls with other diagnoses. The area
under the ROC curve was 0.980 (95% CI, 0.958–1.002; 95% CI with
BCa method, 0.943–0.993; Fig. 3
).
|
A cutoff value of 14 AU was chosen, and sera with antibody
concentrations 14 AU were labeled as guinea pig TGc ELISA positive.
This cutoff value gave (excluding treated patients) a specificity and a
sensitivity of 96.2% (95% CI, 92.8–99.6%) and 92.7% (95% CI,
88.1–97.3%), respectively. The coincidence of the guinea pig TGc
assay with the clinical diagnosis (excluding treated patients) was 102
of 108 (94.4%), giving two false-positive and four false-negative
results (Fig. 3
).
effects of calcium activation
Thirty-two serum samples were tested in the ELISA for IgA
antibodies against human TGc with and without calcium activation. The
overall antibody titers did not show a significant difference
(P = 0.27). However, sera with anti-TGc titers <30 AU
in the calcium-activated assay were significantly lower in the assay
without calcium activation (n = 18; P = 0.009),
whereas higher titers were not significantly different (n = 14;
P = 0.35).
comparison of ema test with TGc ELISA
Excluding treated patients, with the exception of one
false-positive result, all of the patients with EMA-positive sera had
GSE (55 of 56, 98.2%). Twelve of 16 (75%) treated patients with GSE
were positive for EMAs. Comparing only the untreated EMA-positive
cases, the results of the human and guinea pig TGc ELISAs coincided
with the EMA test in 54 of 56 (96.4%) and 51 of 56 (91.1%) cases,
respectively. The serum that gave a false-positive result in the EMA
test was negative in both the human and the guinea pig TGc ELISAs. The
one serum that gave a false-negative result in the human TGc ELISA was
also negative in the guinea pig TGc ELISA. The 12 patients on
incomplete gluten-free diets with EMA positivity had also positive
anti-TGc IgA titers with both ELISA systems.
All patients negative for EMAs were either treated patients having GSE or patients not having GSE. Comparing only the untreated EMA-negative cases, the results of the human and guinea pig TGc ELISAs coincided with the EMA test in 51 of 52 (98.1%) and 50 of 52 (96.2%) cases, respectively. The one serum that gave a false-positive result in the human TGc ELISA was also positive in the guinea pig TGc ELISA; in addition, another serum that gave a false-positive result was detected by the guinea pig assay. Both false-positive sera were from patients with Crohn disease. The four EMA-negative patients with treated GSE were also negative by guinea pig TGc ELISA, but one of them was positive by human TGc ELISA.
The overall coincidence of the EMA test with the human and guinea pig ELISAs was 120 of 124 (96.8%) and 117 of 124 (94.4%), respectively.
comparison of human TGc ELISA WITH GUINEA PIGTGc ELISA
The results of the two ELISAs coincided in 119 of 124 (96%) of
all tested sera. In four discordant cases, the human assay was more
sensitive than the guinea pig assay, giving positivity in the human
assay. One of the sera was from an EMA-negative CD patient on a
gluten-free diet. In the fifth discordant case, the guinea pig ELISA
gave a false-positive result for a patient with Crohn disease. The
antibody titer, however, was also high (17.5 AU) in the human TGc
ELISA, almost reaching the cutoff value (18 AU).
The false results of the human TGc ELISA coincided with those of the guinea pig TGc ELISA. Both assays failed to recognize the serum of one EMA-positive CD patient, and both detected a patient having Crohn disease as positive. Both tests gave a correct, negative result in the case of a false EMA-positive patient.
The titers obtained with the two assays correlated well
(rs = 0.9377; 95% CI, 0.9121–0.9559;
P <0.0001); the correlation was theoretically exponential,
but in practice, it was linear, with an exponent of 1.05 (Fig. 4
). The difference between the AUCs was 0.019 (95% CI, -0.002
to 0.040; 95% CI with BCa method, 0.005–0.056).
|
|
Discussion |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
The molecular mass of the guinea pig TGc differs only slightly (0.1 kDa) from that of the human TGc when measured by mass spectrometry, but the guinea pig TGc migrates appreciably faster on SDS-PAGE than the human TGc. The TGc from human fibroblasts migrates with the same speed as our fusion protein (Sárdy et al., unpublished observation), although the difference between them is 1.2 kDa. These observations imply that the difference between the structures of the human and the guinea pig TGc is more profound than suggested by their high amino acid identity.
The guinea pig liver TGc preparation used for testing contains other protein contaminants that are not immunoreactive with monoclonal antibodies against TGc. However, because it had been used successfully by other authors in its original form (12)(16)(17), we did not purify it further. It cannot be excluded that the immunopositivity seen in microtiter wells might in some cases be attributable to reactivity against contaminants.
Dieterich et al. (16) and Sulkanen et al. (17) optimized the original method (12), by using calcium activation. Because the sensitivity of the calcium-activated human TGc ELISA was higher than that without calcium activation, we also use calcium in the assay with human TGc.
The setting of the cutoff values for the ELISAs was based on the ROC analysis of the tests. A cutoff value that provided perfect separation of individuals having or not having GSE could not be found, although the coincidence with the diagnosis by biopsy was very high in both tests. The guinea pig TGc ELISA was not able to detect two untreated patients with DH and one untreated patient with CD who could be detected by the human TGc ELISA. The 95% CIs of the sensitivities of the two ELISAs overlap; therefore, the sensitivity difference must be confirmed by further studies, but the results affirm the assumption that in a few cases autoantibodies are directed against epitopes of human TGc not conserved in guinea pig TGc. One serum from a CD patient with clear EMA positivity was not immunoreactive in either ELISA, and the titer values were so far below the cutoff that the results were probably not random. It is conceivable that TGc is not the only autoantigen in GSE and that in some rare cases, although EMA positivity occurs, no antibodies against TGc are present. This is also supported by the observation that the immunoabsorption of IgA-class autoantibodies against TGc by guinea pig TGc cannot completely abrogate antiendomysium activity (28), although these experiments should be repeated using calcium-activated human TGc because not all patients’ antibodies may cross-react with the guinea pig antigen.
The EMA test gave a false-positive result in an 8-year-old girl who had transient diarrhea in February 1998. Repeated EMA tests showed IgA binding in the intercellular spaces of smooth muscle cells. The jejunal histology was negative for GSE, and the diarrhea has not recurred. The fact that a serum was false positive by the EMA test but correctly diagnosed by both ELISAs also underlines the possibility of EMA positivity attributable to antigens other than TGc.
It is interesting that two EMA-negative patients with Crohn disease had TGc antibody titers above the cutoff in the guinea pig TGc ELISA; one of them also had a TGc antibody titer above the cutoff in the human TGc ELISA. CD and Crohn disease have been described in the same patient (29), but this association is very rare. In our two cases, associated CD cannot be ruled out, but because both titers were near the borderline (21.6 and 17.5 AU in the human TGc ELISA, 15.9 and 17.8 AU in the guinea pig ELISA), the increased titers might result from low-level IgA autoantibody production against TGc in Crohn disease rather than from that in active CD. This speculation is supported by the fact that in both ELISA systems, the median titers of patients with Crohn disease are greater than those of healthy individuals and patients with other gastrointestinal or non-gastrointestinal diseases. However, the differences and the number of patient sera tested in the present study are too small to allow us to judge the significance of this finding.
Patients on complete or incomplete gluten-free diet had a wide spectrum of antibody titers, and the results of the ELISAs were in good agreement with those of EMA tests. The ELISA with the human antigen turned out to be slightly more sensitive in this regard than the EMA test or the guinea pig TGc ELISA, recognizing one CD patient who was negative in the EMA test and guinea pig TGc ELISA as positive.
Compared with the other established systems, the human TGc ELISA was as specific and sensitive as the EMA test, and somewhat superior to the guinea pig TGc ELISA. The results show the high diagnostic value of all tested systems in this study, but in particular that of the human TGc ELISA, which has almost perfect sensitivity and specificity and does not have the disadvantages of EMA test. Thus, we conclude that the human TGc-based ELISA should be the method of choice for easy and noninvasive screening and diagnosis of GSE.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsDiscussionReferences |
|---|
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  H. E. Prince Evaluation of the INOVA Diagnostics Enzyme-Linked Immunosorbent Assay Kits for Measuring Serum Immunoglobulin G (IgG) and IgA to Deamidated Gliadin Peptides Clin. Vaccine Immunol., January 1, 2006; 13(1): 150 - 151. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Alaedini and P. H.R. Green Narrative Review: Celiac Disease: Understanding a Complex Autoimmune Disorder Ann Intern Med, February 15, 2005; 142(4): 289 - 298. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Van Meensel, M. Hiele, I. Hoffman, S. Vermeire, P. Rutgeerts, K. Geboes, and X. Bossuyt Diagnostic Accuracy of Ten Second-Generation (Human) Tissue Transglutaminase Antibody Assays in Celiac Disease Clin. Chem., November 1, 2004; 50(11): 2125 - 2135. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J.-J. Baudon, C. Johanet, Y. B. Absalon, G. Morgant, S. Cabrol, and J.-F. Mougenot Diagnosing Celiac Disease: A Comparison of Human Tissue Transglutaminase Antibodies With Antigliadin and Antiendomysium Antibodies Arch Pediatr Adolesc Med, June 1, 2004; 158(6): 584 - 588. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Peretti, F. Bienvenu, C. Bouvet, N. Fabien, F. Tixier, C. Thivolet, E. Levy, P. G. Chatelain, A. Lachaux, and M. Nicolino The Temporal Relationship Between the Onset of Type 1 Diabetes and Celiac Disease: A Study Based on Immunoglobulin A Antitransglutaminase Screening Pediatrics, May 1, 2004; 113(5): e418 - e422. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I Marai, Y Shoenfeld, N Bizzaro, D Villalta, A Doria, E Tonutti, and R Tozzoli IgA and IgG tissue transglutaminase antibodies in systemic lupus erythematosus Lupus, April 1, 2004; 13(4): 241 - 244. [Abstract] [PDF]
|
|
|
|
 |
|
 P D Thomas, A Forbes, J Green, P Howdle, R Long, R Playford, M Sheridan, R Stevens, R Valori, J Walters, et al. Guidelines for the investigation of chronic diarrhoea, 2nd edition Gut, July 1, 2003; 52(90005): v1 - 15. [Full Text]
|
|
|
|
 |
|
 E Tonutti, D Visentini, N Bizzaro, M Caradonna, L Cerni, D Villalta, and R Tozzoli The role of antitissue transglutaminase assay for the diagnosis and monitoring of coeliac disease: a French-Italian multicentre study J. Clin. Pathol., May 1, 2003; 56(5): 389 - 393. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. L. Dockerty, L. Williamson, and B. P. Wordsworth Endomysial antibodies in psoriatic arthritis patients Rheumatology, October 1, 2002; 41(10): 1195 - 1196. [Full Text] [PDF]
|
|
|
|
|
|
R C W Wong, R J Wilson, R H Steele, G Radford-Smith, and S Adelstein A comparison of 13 guinea pig and human anti-tissue transglutaminase antibody ELISA kits J. Clin. Pathol., July 1, 2002; 55(7): 488 - 494. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Martini, G. Mengozzi, G. Aimo, L. Giorda, R. Pagni, and C. S. Guidetti Comparative Evaluation of Serologic Tests for Celiac Disease Diagnosis and Follow-Up Clin. Chem., June 1, 2002; 48(6): 960 - 963. [Full Text] [PDF]
|
|
|
|
 |
|
 M. Sardy, S. Karpati, B. Merkl, M. Paulsson, and N. Smyth Epidermal Transglutaminase (TGase 3) Is the Autoantigen of Dermatitis Herpetiformis J. Exp. Med., March 18, 2002; 195(6): 747 - 757. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Martini, G. Mengozzi, G. Aimo, R. Pagni, and C. Sategna-Guidetti Diagnostic Accuracies for Celiac Disease of Four Tissue Transglutaminase Autoantibody Tests Using Human Antigen Clin. Chem., September 1, 2001; 47(9): 1722 - 1725. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
, treated CD or DH patients; 
, controls and untreated CD or DH
patients. The chosen arbitrary cutoff for positivity (dashed
line) is drawn at 18 AU. The ROC curve for the human TGc ELISA
is shown as an inset.
