| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Articles |
1
Department of Orthopedic Surgery, University of Heidelberg, D-96118 Heidelberg, Germany.
2
Diabetes Research Institute, D-40225 Duesseldorf,
Germany.
3
Labor Dr. Koch & Dr. Merk, D-88476 Ochsenhausen,
Germany.
aAddress correspondence to this author at: Stiftung Orthopädische Universitätsklinik Heidelberg, Schlierbacher Landstrasse 200a, D-69778 Heidelberg, Germany. Fax 49-6221-969288; e-mail Wiltrud.Richter{at}ok.uni-heidelberg.de.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Methods: Chimeric proteins composed of GAD65 and IA2 residues were constructed, analyzed for their immune reactivity with monoclonal antibodies and sera, and used in a diagnostic assay with 35S-labeled protein as antigen.
Results: Length and order of GAD65 and IA2 sequences were critical for conservation of the conformational epitopes in the fusion protein. Among four chimera tested, only IA2(606–979)/GAD65(1–585) retained wild-type-like folding of GAD65 and IA2 domains and yielded a stable protein after baculovirus expression. Reactivity of GAD65 antibody- and IA2 antibody-positive sera from patients newly diagnosed with insulin-dependent diabetes mellitus, from ICA-positive prediabetics, and from ICA-positive first-degree relatives demonstrated conservation of the relevant autoreactive epitopes. The assay based on the in vitro translated fusion antigen had a sensitivity and specificity identical to those for detection of GAD65- and IA2-abs based on the two separate GAD65 and IA2 proteins.
Conclusions: Autoantigens such as GAD65 and IA2 can be combined successfully in a fusion protein of similar immune reactivity. This allows simultaneous detection of GAD65- and IA2-abs in a one-step screening assay and cost-effective identification of positive individuals at risk of diabetes or at onset of disease.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
80% of individuals at diagnosis of IDDM and
frequently appear years or even decades before the clinical onset of
the disease (1). Ongoing diabetes research is focusing on
optimizing the prediction and prevention of insulin deficiency in IDDM.
This will depend on large-scale screening for autoantibodies in
relatives and, ultimately, in the general population. Identification of autoantibody-positive individuals previously relied almost exclusively on the histochemical detection of ICAs on frozen sections of human pancreas (2)(3). The usefulness of ICAs for routine screening of susceptible individuals or the general population is, however, limited by the technical difficulties, poor precision, and variable sensitivity of this test in different laboratories (4)(5)(6).
Two islet autoantibody specificities definitely represent major
subfractions of ICAs: antibodies to the 65-kDa isoform of glutamic acid
decarboxylase (GAD65-abs) (7)(8) the
-aminobutyric acid-synthesizing isoform of glutamate decarboxylase
in human islets (9), and antibodies to islet antigen-2
(IA2-abs) (10), a protein tyrosine phosphatase expressed in
islet cells that has been described as a membrane protein composed of
979 amino acids (11). Recent studies reported that screening
for the two subspecificities of ICA, GAD65-abs and IA2-abs, may be an
appropriate alternative for identification of positive subjects;
however, the cost of using two specificities, instead of one, must be
considered. To date, the distinction of GAD65-abs from IA2c-abs has
been of no further therapeutic consequence for diagnosis of
IDDM, and screening for autoantibody-positive individuals at
risk in the general population would profit from easy one-step
identification of both antibody markers.
We therefore aimed to design GAD65/IA2 fusion proteins suitable for simultaneous detection of GAD65- and IA2-abs. Because of one-step production, purification, and quantification, the ideal fusion protein could save time, money, and labor compared with separate testing of different antibody markers and could provide a reasonable target for a screening step, allowing cost-effective combined detection of GAD65- and IA2-abs.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
monoclonal antibodies
The human monoclonal ICAs (MICAs) 1–10, derived from ICA-positive
patients newly diagnosed with type 1 diabetes, have been described
elsewhere (8)(13)(14)(15)(16). MICA 5 was not included
in this study because it is an IgG3 antibody that
does not bind to protein A-Sepharose. The GAD65-specific mouse
monoclonal antibody GAD-6 (17) was derived from the
Developmental Studies Hybridoma Bank, and an IA2-specific mouse
monoclonal antibody, 76F4B, was kindly provided by E. Bonifacio, Milan,
Italy.
construction of chimeric
gad65/ia2c
molecules
For construction of the chimera
GAD65(1–585)/IA2(606–979),
a NotI restriction site was introduced by PCR at amino acid
585 of GAD65, eliminating the stop codon. A second PCR fragment
[IA2(606–979)] was generated with the
restriction sites NotI at the 5' end and EcoRI at
the 3' end. A pVL1393 vector (PharMingen) containing
full-length GAD65 cDNA was cut with StuI and
EcoRI and ligated with both PCR fragments in one step.
The second chimera, IA2c(606–979)/GAD65(1–585), was designed by generation of an IA2c PCR fragment (encoding amino acids 606–979) with restriction sites NcoI and XhoI at the 5' and 3' ends, respectively. A GAD65 PCR fragment (1–585) with primers inserting restriction sites for XhoI (5') and SacI (3') was amplified. Both PCR fragments were cloned into the NcoI and SacI sites of pSP64poly(A) vector.
The third chimera, GAD65(234–585)/IA2(606–979), was assembled first by generating a GAD65 PCR fragment (coding for amino acids 234–442) with the restriction sites BamHI at the 5' end and StuI at the 3' end. The pVL1393 vector, which encodes for the chimera GAD65(1–585)/IA2(606–979), and the described GAD65 PCR fragment were cut with BamHI and StuI and ligated to obtain the final construct. Because of the cloning strategy, two additional glycine residues were generated between the residues of GAD65 and IA2.
The last chimera, IA2(606–979)/GAD65(244–585), was constructed first by generating an IA2 PCR fragment (coding for amino acids 606–979) with additional XmaI and NarI restriction sites at the 5' and 3' ends, respectively. The IA2 PCR fragment and a pSVSport1 vector, which included full-length GAD65 cDNA, were cut with XmaI and NarI and fused together by ligation. A leucine and a glutamic acid residue were generated by this cloning procedure between the two protein fragments. All plasmid constructs were verified by DNA sequence analysis.
autoradiography of recombinant proteins
Aliquots of the in vitro translated
[35S]methionine-labeled chimeric proteins were
analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis
(SDS-PAGE) on 8% polyacrylamide gels. The gels were fixed with 400
mL/L methanol–100 mL/L glacial acetic acid before being dried
under reduced pressure and exposed to x-ray films at -80 °C
for 1 day or longer.
ria for detection of gad65-ab and
ia2c-ab
GAD65- and IA2c-abs in sera were assessed by RIA as
described previously (18). In brief, 2.5 µL of the sample
was incubated overnight with 35S-labeled GAD65,
IA2c, or IA2c/GAD65 fusion proteins and immunoprecipitated with protein
A-Sepharose (Pharmacia). Sepharose-bound immune complexes were washed
in a membrane-bottomed microtiter plate and transferred to
scintillation vials, and the cpm were determined by a beta
counter (Wallac, Pharmacia). MICA 3 was used as the calibrator for
GAD65-abs, and an
IA2-ab+/GAD65-ab-
reference serum of a patient newly diagnosed with IDDM was selected as
the calibrator for IA2-abs. When GAD65- and IA2-abs were determined in
the same test, 2.5 µL of both MICA 3 and the IA2 reference serum were
applied (combined calibrator). At the first Combined Autoantibody
Workshop in 1995 (19), the GAD65-ab RIA had a sensitivity of
70% [95% confidence interval (95% CI), 56–83%] and a specificity
of 100% (95% CI, 96–100%). At the IDS Proficiency Workshop 2000,
our RIA with
IA2c(606–979)/GAD65(1–585)
achieved a sensitivity of 92% (95% CI, 81–98%) and a specificity of
98% (95% CI, 89–100%) when we applied our in-house 4% assay cutoff
according to ROC analysis.
expression of fusion proteins in
Sf9-insect cells
In vivo expression of the two fusion protein variants
GAD65(1–585)/IA2(606–979)
and
IA2(606–979)/GAD65(1–585)
was performed using the baculovirus expression system
(20)(21).
immunoblotting
Sf9-insect cell lysate was prepared at various time
points after infection, and the supernatant of a 100 000g
centrifugation was separated by SDS-PAGE. Proteins were transferred to
nitrocellulose (Bio-Rad) (22) before immunostaining
with a polyclonal rabbit anti-GAD65-ab or anti-IA2-ab (1:1000 dilution)
and alkaline phosphatase-conjugated anti-rabbit IgG antibody (1:20 000
dilution; Dianova). Bound antibodies were visualized on nitrocellulose,
using an alkaline phosphatase substrate solution of Nitro Blue
Tetrazolium/5-bromo-4-chloro-3-indolyl phosphate [Sigma; 66 and 33
µL, respectively, in 10 mL of 60 mmol/L Tris-HCl (pH 7.5), 325 mmol/L
NaCl, 0.5 mL/L Tween 20].
roc analysis
Sensitivity/specificity pairs were calculated by varying the
decision threshold values over the entire range of results
(23)(24). Sensitivity (the true-positive
fraction) was calculated with samples derived from 101 patients at
clinical onset of IDDM. The specificity (the true-negative fraction)
was calculated from 246 healthy individuals. The experimental cutoff
for each assay was determined by calculating the decision threshold as
follows: for each given threshold, the values for sensitivity and
specificity were added, and the value at the maximal sum was used as
the experimental cutoff. Test results above binding of the undiluted
calibrators (100% binding) and the linear range of the calibration
curve were set to 100%.
statistical analysis
Correlation and the agreement between the results obtained with
the assay using a GAD65/IA2 fusion protein and the RIA for combined
detection of GAD65- and IA2-abs (18) based on separate
antigens were examined with the method of Bland and Altman
(25)(26). In short, this method uses a plot of
the difference between the assays compared against their mean. This
kind of plot allows investigation of any possible relationship between
the measurement error and the true value. The mean of the two
measurements is the best estimate for the measurement error. For all
statistical methods, the Statistical Package for Social Science (SPSS)
was used.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). All four chimeric GAD65/IA2c proteins were successfully
expressed by in vitro transcription and translation in a reticulocyte
lysate system. The estimated molecular masses of the fusion
proteins were 108 and 72 kDa. However, analysis of the proteins by
SDS-PAGE revealed molecular masses of 97 and 74 kDa, respectively
(data not shown). Few additional bands were indicative for low-level
protein degradation or prematurely stopped translation.
|
epitope conservation in the chimeric
gad65/ia2c proteins
Correct folding of the subunits in chimeric proteins is crucial
for preservation of epitope recognition by disease-associated
autoantibodies. The conformational integrity of the GAD65 domains was
checked by RIA using the diabetes-related human monoclonal GAD65
antibodies MICA 1–10. These antibodies define nine distinct
conformational epitopes in GAD65, seven of which reside between amino
acids 240 and 585 of GAD65
(14)(16)(27). To monitor the
conformation of IA2c, one mouse monoclonal antibody (76F4B), which
recognizes an epitope of the cytoplasmic part of IA2 that is located
near the membrane, was available. Therefore, sera from two
IA2c-ab+, GAD65-ab-
patients were included, one of which was directed exclusively to the
COOH terminus of IA2c (amino acids 771–979), whereas the other
recognized several epitopes. Compared with wild-type GAD65 (Fig. 1A
),
the reactivities of MICA 1–10 were almost completely lost in the two
chimeras
GAD65(234–585)/IA2c(606–979)
(Fig. 1C
) and
IA2c(606–979)/GAD65(244–585)
(Fig. 1D
), although, according to the GAD65 residues present in the
chimeras, all MICAs except MICA 8 and MICA 9 were expected to react
(14). Recognition of the IA2c domain was conserved in both
constructs, but decreased reactivity was observed compared with the
IA2c molecule alone. In the fusion protein
GAD65(1–585)/IA2c(606–979)
(Fig. 1E
), all MICA epitopes except MICA 2 and MICA 7 were conserved.
Only for MICA 3 and MICA 10, however, was binding as high as with the
wild-type GAD65 protein. Again, IA2c-ab reactivity was reduced
approximately twofold compared with the IA2c molecule (Fig. 1B
). Within
chimera
IA2c(606–979)/GAD65(1–585),
all nine tested MICA epitopes and the IA2c reactivity were conserved
with no significant loss of binding compared with wild-type GAD65 and
IA2c (compare panels A, B, and F in Fig. 1
). Thus, the length of
the GAD65 fragment was crucial for its correct folding in the fusion
protein. The wild-type conformation of GAD65 was preserved only in the
IA2c(606–979)/GAD65(1–585)
chimera, where the COOH terminus of GAD65 was freely accessible.
In contrast, IA2c tolerated well a connection of its C-terminal end to
GAD65.
order of gad65 and
ia2c in fusion proteins is relevant for
in vivo expression
Recombinant baculoviruses were produced that coded for the two
fusion protein variants
GAD65(1–585)/IA2c(606–979)
and
IA2c(606–979)/GAD65(1–585).
Expression studies over a time interval of 31–53 h after infection of
Sf9-insect cells revealed that only the IA2c/GAD65 chimera
with GAD65 at the COOH terminus was expressed as a stable protein (Fig. 2
). This indicated that the orientation of the two protein
fragments GAD65 and IA2c within the fusion protein is crucial for
stable expression of correctly folded GAD65 and IA2c domains in
Sf9-insect cells.
|
preserved reactivity of sera from patients and individuals at risk
The immune reactivity of GAD65-ab+ and/or
IA2c-ab+ IDDM sera and of sera from ICA-positive
prediabetics with in vitro expressed
IA2c(606–979)/GAD65(1–585)
was analyzed by RIA. All 14 of the monospecific
GAD65-ab+/IA2c-ab- sera,
all 15 GAD65-ab-/IA2c-ab+
sera, and all 15 bispecific
GAD65-ab+/IA2c-ab+ sera
bound to chimera
IA2c(606–979)/GAD65(1–585).
In addition, all 15 sera from ICA-positive prediabetics recognized the
IA2c/GAD65 fusion protein (Fig. 3
). This demonstrated that the relevant IDDM- and
prediabetes-associated immune reactivity was preserved in this fusion
protein.
|
equal performance of
ia2c(606–979)/gad65(1–585) and the separate antigens gad65 and
ia2c in ria
To compare the performance of the GAD65/IA2c fusion protein with
the separate antigens in detection of GAD65- and IA2c-abs, we analyzed
sera from 101 newly diagnosed IDDM patients and from 245 control
subjects by RIA. The threshold for discrimination of positive from
negative sera was then selected by different means. In ROC plot
analysis, the optimal cutoff was at 4% antibody binding for both
assays. When the 99th percentile or maximum specificity was selected,
the cutoff changed accordingly (Table 1
). Irrespective of the mode of cutoff selection, the fusion
protein gave equal or slightly superior values for sensitivity and
specificity compared with a test in which the separate antigens were
used (Table 1
). To quantitatively approach the correlation and
agreement between the antibody assays, we used a method described
previously by Bland and Altman (25)(26). The
data shown for the 101 patients’ sera (Fig. 4
) revealed that the two test methods were not only highly
related (r = 0.99; R2
= 0.98; P <0.00001) but also showed good agreement,
especially near the cutoff (4–30% autoantibody binding). The mean
(0.37) indicated that the assay using the GAD65/IA2c fusion protein
measured on average slightly higher test scores than the combined
detection of GAD65-abs and IA2c-abs based on separate antigens. Thus,
the performance of the RIA based on the
IA2c(606–979)/GAD65(1–585)
fusion protein was almost indistinguishable from that of combined or
separate testing of GAD65- and IA2c-abs.
|
|
diagnostic power in patients, prediabetics, and individuals at risk
Sera from 15 prediabetic individuals and 8 ICA-positive sera
selected from 62 first-degree relatives were subjected to analysis of
GAD65-abs and IA2c-abs by tests based on the separate antigens and the
fusion protein. Independent of the mode of cutoff selection, the fusion
protein assay detected more IDDM sera (Fig. 5A
) and ICA-positive sera from prediabetics and from first-degree
relatives (Fig. 5, B and C
). Separate testing missed four sera from
patients with IDDM and one serum derived from a prediabetic individual
56 months before diagnosis of IDDM. Sera with low concentrations of
GAD65- and IA2c-abs just below the assay cutoff of the separate tests
may account for this superior performance of the fusion protein assay.
|
The fusion protein assay thus improved detection of patients’ sera compared with assays based on the separate antigens. Hence it can be recommended for further studies on risk assessment of IDDM.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Among four distinct GAD65/IA2c fusion proteins that included various
lengths and opposite orientations of the GAD65 and IA2c domains, only
chimera
IA2c(606–979)/GAD65(1–585)
showed complete epitope conservation in the GAD65 and IA2c domains
according to binding studies with 10 monoclonal antibodies and sera
specific for GAD65 or IA2c, respectively (Fig. 1
).
IA2c(606–979)/GAD65(1–585)
fusion protein, but none of the other tested chimera, performed almost
indistinguishably from the separate GAD65 and IA2c proteins in RIAs in
which sera from patients with IDDM and from ICA-positive prediabetics,
ICA-positive first-degree relatives, and control subject were analyzed
(Figs. 3
and 5
).
IA2c(606–979)/GAD65(1–585)
protein also retained the wild-type conformation of the GAD65 and IA2c
domains when expressed by recombinant baculoviruses in insect cells.
The fusion protein described here may be especially useful in
increasing the efficiency of screening for autoantibody-positive
subjects in risk groups or in the general population with the purpose
of preparing prevention studies.
Instead of testing all sera twice in two distinct assays, we suggest a new two-step strategy. The first step, based on the fusion protein assay, would identify the 1–3% expected positive samples; thus, only those samples will have to undergo thorough risk analysis in the second step, which involves looking for individual antibody markers (28), for epitope recognition (15), or for subclass distribution (29), which could further refine risk assessment. More thorough, labor-intensive, and expensive testing would thus be confined to only 1–3% of the total samples.
In addition to GAD65-abs and IA2c-abs, antibody specificities such as
islet cell, insulin, and proinsulin autoantibodies have been applied
for the prediction of IDDM. We did not include insulin or proinsulin
protein domains in our fusion proteins because studies on first-degree
relatives (30)(31) and in a population of
schoolchildren showed that their predictive value is below that of
GAD65-abs and IA2c-abs and did not enhance prediction in a family study
(31). Reduction of islet antibody reactivities to only the
major immunogenic parts of the GAD65 and the IA2 molecules was
evaluated in this study. It was successful only for the intracellular
part of IA2, which attracted most of the IA2-ab reactivity
(32)(33). Although most of the autoreactive
epitopes of GAD65 are located within the middle and the C-terminal
regions of GAD65 (amino acids 245–585)
(2)(7)(34), truncated GAD65 lost almost all of its
autoreactive epitopes when it was integrated in a chimeric protein with
IA2c (Fig. 1, C and D
). Thus, the length of GAD65 was important for its
correct folding, and a free GAD65 COOH terminus was required for stable
expression and full preservation of epitope recognition.
The close correlation between combined GAD65-ab and IA2c-ab detection based on the separate antigens and on our fusion protein IA2c(606–979)/GAD65(1–585) suggests this molecule as an ideal tool for economical diagnostic testing and population screening for prediction of IDDM. To confirm the usefulness of this combi-assay, its performance now needs to be compared with that of separate assays on a large unselected group of subjects.
In conclusion, fusion proteins combined from several autoantigens can reduce the effort for analysis of several autoantibody specificities to one test, allowing combined analysis of the desired markers. Thus, diagnosis, risk assessment, or prediction of autoimmune diseases can profit from this efficient and cost-effective strategy, especially in cases involving large sample collections with a low prevalence of positive sera and in autoimmune disorders in which the presence of target-specific antibodies, rather than their exact specificity, is of therapeutic consequence.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  M. Ankelo, A. Westerlund, K. Blomberg, M. Knip, J. Ilonen, and A. E. Hinkkanen Time-Resolved Immunofluorometric Dual-Label Assay for Simultaneous Detection of Autoantibodies to GAD65 and IA-2 in Children with Type 1 Diabetes Clin. Chem., March 1, 2007; 53(3): 472 - 479. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Lumbreras-Lacarra, J. M. Ramos-Rincon, and I. Hernandez-Aguado Methodology in Diagnostic Laboratory Test Research in Clinical Chemistry and Clinical Chemistry and Laboratory Medicine Clin. Chem., March 1, 2004; 50(3): 530 - 536. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Westerlund-Karlsson, K. Suonpaa, M. Ankelo, J. Ilonen, M. Knip, and A. E. Hinkkanen Detection of Autoantibodies to Protein Tyrosine Phosphatase-like Protein IA-2 with a Novel Time-resolved Fluorimetric Assay Clin. Chem., June 1, 2003; 49(6): 916 - 923. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. B. Sacks and A. Lernmark Molecular Manipulation of Autoantibody Testing in Type 1 Diabetes: Two for One Clin. Chem., May 1, 2001; 47(5): 803 - 804. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
