| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Studies |
Division of Clinical Sciences (R.A.A., E.H.K., E.A.W., P.F.W., A.P.W.), Northern General Hospital, University of Sheffield, Sheffield, S5 7AU, United Kingdom; and Third Department of Internal Medicine (T.E., T.O.), University of Yamanashi Medical School, Tamaho, Yamanashi 409-38, Japan
Address correspondence and requests for reprints to: Dr. R. A. Ajjan, Division of Clinical Sciences Northern General Hospital, University of Sheffield, Sheffield, S5 7AU, United Kingdom. E-mail: Ramzi{at}Ajjan.Freeserve.co.uk
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
A significant difference in the frequency of hNIS antibody-positive sera was found when patients with either Graves’ disease (GD) or autoimmune hypothyroidism (AH) were compared with normal controls (P = 0.01 and P = 0.03, respectively). Of 49 GD and 29 AH sera tested, 11 (22%) and 7 (24%), respectively, were found to contain hNIS antibodies. Differences were also significant when the antibody-binding indices of the control group of sera were compared with those of both the GD and the AH patient sera (P < 0.0001 and P = 0.001, respectively). In contrast, sera from 10 patients with Addison’s disease and 10 patients with vitiligo (without associated ATD) were all negative for antibody reactivity to the symporter. No differences were detected when the antibody binding indices of either the Addison’s disease or the vitiligo sera were compared with those of the normal sera group (P = 0.9 and P = 0.6, respectively).
Eight of the 11 (73%) GD and 3 of the 7 (43%) AH sera, which were positive for hNIS antibodies in the immunoprecipitation assay, were also found to inhibit iodide uptake in hNIS-transfected CHO-K1 cells, suggesting the existence of antibodies in some serum samples that bind to the symporter without modulating its function. Overall, a significant correlation was found between the iodide uptake inhibition and the binding assays for hNIS antibody detection (r = 0.49, P < 0.0001).
In summary, we have developed a specific and quantitative assay for the detection of hNIS binding antibodies in sera of patients with ATD. This system offers the advantage of studying antibody reactivity against conformational epitopes and will be useful in understanding the role of NIS autoreactivity in the pathogenesis of ATD.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
We have developed a bioassay that demonstrated that 30% of GD sera can modulate hNIS activity, an effect that was antibody-mediated (9). Although the bioassay is suitable for the investigation of conformational epitopes, it cannot detect antibodies that bind the hNIS without modulating its function. Taken together, the production of the hNIS in its native form (or at least a form close to it), suitable for application in a binding assay, seems necessary for the proper evaluation of autoantibodies against the symporter.
A common problem encountered in the study of autoantibody reactivity is the lack of pure uncontaminated antigen, often resulting in a background of nonspecific antibody reactivity, in turn making interpretation of the results difficult. Recently, autoantigens have been produced in an in vitro transcription and translation (TnT) system that offers the advantage of investigating antibody reactivity against a single protein (10, 11, 12, 13, 14, 15), rendering assays highly specific and sensitive.
In the present work, we describe the in vitro production of [35S]-labeled hNIS and the immunoprecipitation of the translated antigen by antibodies in sera from patients with ATD. We also demonstrate the presence of antibodies that bind to but do not block the activity of the symporter, thus extending the profile of autoreactivity to this novel autoantigen.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Sera from 49 GD (8 men, 41 women; mean age, 49 yr; age range, 23–80 yr) and 29 autoimmune hypothyroidism (AH) (4 men, 25 women; mean age, 51 yr; age range, 28–81 yr) patients were used in this study. Diagnosis of GD was based on the presence of hyperthyroidism, supported by one or more of the following features: a diffuse goiter, the presence of thyroglobulin or thyroid peroxidase antibodies, and evidence for thyroid-associated ophthalmopathy. AH was diagnosed by the presence of hypothyroidism and positive thyroglobulin or thyroid peroxidase antibodies. Of the 49 GD patients studied, 25 were newly diagnosed untreated GD patients and 12 had associated ophthalmopathy (grade II-IV), of whom 4 had also pretibial myxedema. Sera were also obtained from 10 patients with isolated vitiligo and 10 with Addison’s disease (AD), without any evidence of associated ATD. Sera from 20 normal individuals (9 men, 11 women; mean age, 31 yr; age range, 23–47 yr) were used as controls. Patient and control IgGs were prepared as described elsewhere (9).
The study was approved by the Ethics Committee of the Northern General Hospital, Sheffield, and all subjects gave informed consent. All sera were kept frozen at -20 C before analysis.
Antisera
Rabbit antiserum against a hNIS fragment (amino acids 466–522)
has been described previously (16) and was used as a positive control.
Antityrosinase rabbit antiserum 
PEP7, which was used as a negative
control, was a gift of Prof. V. Hearing (NIH, Bethesda, MD; Ref.
17).
hNIS complementary DNA (cDNA) constructs
The full-length hNIS cDNA and a truncated version encoding amino acids (aa) 1–643 and 1–612 of hNIS, respectively, both cloned into the eukaryotic expression vector pcDNA3, were a gift from Dr. S. M. Jhiang (The Ohio State University, Columbus, OH; Ref. 18). The plasmids were prepared from Escherichia coli JM109 (Promega Corp., Southampton, UK) using a Wizard Maxipreps DNA Purification System (Promega Corp.), according to the manufacturer’s protocol.
Iodide uptake and iodide uptake inhibition studies
A stable CHO-K1 cell line containing full-length hNIS and designated CHO-NIS12 was established using steps identical to those used in the isolation of CHO-NIS9, which contains truncated (aa 1–612) hNIS (9). The cell line was subsequently tested for iodide uptake activity, as detailed elsewhere (9). Briefly, hNIS-transfected CHO-K1 cells were cultured in 6-well plates, and iodide uptake was analyzed when these cells reached 100% confluence. Untransfected CHO-K1 cells were used as a control. 125I (17 Ci/mg; NEN Life Science Products, Hounslow, UK), 1.5–2 kBq, was incubated with the cells in 1 mL serum-free Ham’s F-12 medium (Life Technologies, Inc., Paisley, UK) for 30 min. Cells were then washed quickly with phosphate-buffered saline (PBS), solubilized with 1 mL of 1 mol/L NaOH, and radioactivity was counted in a Wallac, Inc. 1282 Compugamma counter (Wallac UK, Milton Keynes, UK). For untransfected cells, iodide uptake in cpm was assigned a value of 1. For CHO-NIS9 and CHO-NIS12, iodide uptake was expressed as: iodide uptake in cpm in transfected cells/iodide uptake in cpm in untransfected cells. The mean of duplicate cultures ± SEM was calculated and used to express the results.
The effect of perchlorate on the iodide uptake activity of both CHO-NIS9 and CHO-NIS12 was assessed by incubating the cells with 1 mmol/L perchlorate before the addition of 125I. Again, untransfected cells were used as a control. For perchlorate-treated untransfected cells, iodide uptake in cpm was assigned a value of 1. For perchlorate-treated transfected cells, iodide uptake was expressed as: iodide uptake in cpm in perchlorate-treated transfected cells/iodide uptake in cpm in perchlorate-treated untransfected cells. The mean of duplicate cultures ± SEM was calculated and used to express the results.
To determine the effects of control and ATD patient sera on symporter activity, either 100 µL serum or 2 mg/mL IgG in 1 mL serum-free medium were incubated with hNIS-transfected CHO-K1 cells at 37 C for 60 min. An additional 1 mL medium containing 3–4 kBq of 125I was added, and the cells were incubated for an additional 30 min, followed by washing, solubilization, and counting as above. Each serum/IgG was tested twice, and results expressed as the mean percentage of inhibition of iodide uptake by hNIS-transfected CHO-K1 cells. The mean ± SEM intra- and interassay coefficients of variation for the bioassay, obtained by testing 10 serum samples on three occasions, were 5.3 ± 0.9% and 6.7 ± 1.4%, respectively.
Rubidium uptake and rubidium uptake inhibition studies
CHO-NIS9 cells were cultured in 6-well plates, and 86Rb uptake was analyzed when these cells reached 100% confluence. Untransfected CHO-K1 cells were used as a control. 86Rb (>1Ci/g; NEN Life Science Products), 2 kBq, was incubated with the cells in 500 µL serum-free medium for 30 min. Cells were then washed quickly with PBS and solubilized with 200 µL of 1 mol/L NaOH. Scintillation fluid was added to 100 µL solubilized cells, and radioactivity was counted in a Compugamma counter.
To determine the effects of control and ATD patient sera on the activity of the Na+/K+ ATPase, hNIS-transfected CHO-K1 cells were incubated with either 50 µL serum or 2 mg/mL IgG in 500 µL serum-free medium. After 60 min of incubation at 37 C, 500 µL medium containing 4 kBq of 86Rb were added to the cells, and incubation continued for an additional 30 min. Cells were then washed, solubilized, and counted as above. Each serum/IgG was tested twice, and results were expressed as the mean percentage of inhibition of rubidium uptake by hNIS-transfected CHO-K1 cells.
In vitro-coupled TnT of hNIS
The full-length hNIS protein was produced in vitro from its cDNA using a TnT T7 Coupled Reticulocyte Lysate System (Promega Corp.). Briefly, 0.1 µg plasmid DNA was incubated for 120 min at 30 C with 25 µL rabbit reticulocyte lysate, 1 µL T7 TnT RNA polymerase, 1 µL amino acids minus methionine, 40 U RNasin (Promega Corp.), 2 µL TnT reaction buffer, and 4 µL translation-grade [35S]methionine (1000 Ci/mmol; 10 mCi/mL; Amersham Pharmacia Biotech, Aylesbury, UK) made up to 50 µL with nuclease-free water. The reaction was stored at -20 C until needed. Percentage incorporation of [35S]methionine was determined by trichloroacetic acid precipitation, according to the manufacturer’s protocol. Glycosylation of [35S]hNIS was initiated by the addition of 5 µL Canine Microsomal Membranes (Promega Corp.) to the reaction, as described above, but containing 1 µL TnT reaction buffer and 1 µg plasmid DNA. Deglycosylation of [35S]hNIS was achieved using N-Glycosidase F (Roche Molecular Biochemicals, Lewes, UK), as detailed by the supplier’s instructions. In brief, an aliquot of the glycosylated [35S]hNIS was incubated at 37 C for 12 h with 2 U enzyme in buffer containing 100 mmol/L sodium phosphate buffer (pH 6.5), 10 mmol/L EDTA, 0.1% SDS, and 0.5% Triton X-100.
Electrophoretic analysis and autoradiography
SDS-PAGE (19) of in vitro-translated products was performed in a 10% polyacrylamide resolving gel containing 325 mmol/L Tris-HCl (pH 8.8) and 0.1% SDS and a 4% polyacrylamide stacking gel containing 125 mmol/L Tris-HCl (pH 6.8) and 0.1% SDS. The gel running buffer contained 25 mmol/L Tris-HCl (pH 8.3), 192 mmol/L glycine, and 0.1% SDS.
Aliquots of in vitro-translated [35S]hNIS (both unglycosylated and glycosylated) and deglycosylated [35S]hNIS were added to SDS-sample buffer containing 63 mmol/L Tris-HCl (pH 6.8), 10% glycerol, 2% SDS, 5% 2-mercaptoethanol, and 0.001% bromophenol blue and heated to 100 C for 2 min before loading onto a gel. To visualize protein markers, the gel was stained with 0.05% Coomassie blue in 10% glacial acetic acid/25% isopropanol and destained with 10% glacial acetic acid/25% isopropanol each for 30 min at room temperature. The gel was then soaked in Amplify scintillant (Amersham Pharmacia Biotech) for 30 min at room temperature before drying at 80 C for 2 h onto 3MM chromatography paper (Whatman International Ltd., Maidstone, UK) under vacuum. Autoradiography was carried out at -70 C using x-ray film (Genetic Research Instrumentation Ltd., Dunmow, UK). Protein molecular weight standards (Sigma, Dorset, UK) consisted of myosin (205 kDa), ß-galactosidase (116 kDa), phosphorylase b (97 kDa), serum albumin (66 kDa), ovalbumin (45 kDa), and carbonic anhydrase (29 kDa).
Immunoprecipitation of [35S]-labeled hNIS
For each assay, an aliquot of the in vitro translation reaction mixture (equivalent to 20,000 cpm of trichloroacetic acid precipitable material) was suspended in 50 µL immunoprecipitation buffer containing 20 mmol/L Tris-HCl (pH 8.0), 150 mmol/L NaCl, 1% Triton X-100, and 10 µg/mL aprotinin (Bayer Corp., Newbury, UK). The serum to be analyzed was then added to a final dilution of 1:10, unless otherwise specified. After incubation overnight with shaking at 4 C, 50 µL protein G Sepharose 4 Fast Flow slurry (Pharmacia Biotech, Uppsala, Sweden), prepared according to the manufacturer’s protocol, was added and incubated for 1 h at 4 C. The protein G Sepharose-antibody complexes were then collected by centrifugation and washed six times for 15 min in immunoprecipitation buffer at 4 C. Immunoprecipitated radioactivity in cpm was evaluated in a LKB 1217 Rackbeta liquid scintillation analyzer (Wallac, UK).
Antibody binding was expressed as an immunoprecipitation index (hNISAb index) defined as: the cpm immunoprecipitated by each tested sample/the mean cpm immunoprecipitated by 20 normal controls. Each serum sample was assayed at least three times, and a mean hNISAb index was calculated. The mean ± SEM intra- and interassay coefficients of variation for the binding assay, obtained by testing five samples on three occasions, were 12.1 ± 1.6% and 14.8 ± 1.7%, respectively.
For analysis by SDS-PAGE, the protein G Sepharose-antibody complexes were resuspended in 100 µL SDS-sample buffer, boiled, and centrifuged and the supernatant was recovered for electrophoresis.
For dilution experiments, hNIS antibody-positive sera were analyzed in the immunoprecipitation assay at dilutions of 1:10, 1:50, 1:100, and 1:200. Specific anti-hNIS rabbit antiserum was tested in the assay at dilutions of 1:10, 1:50, 1:100, 1:200, 1:500, 1:1000, and 1:2000.
Absorption studies
For preparation of extracts, both CHO-K1 or CHO-NIS12 cells were
scraped on ice from the surface of tissue culture dishes. After washing
in PBS, 1 mL immunoprecipitation buffer containing 10 µg/mL aprotinin
(Bayer Corp.), 100 µmol/L
N-tosyl-phenylalanylchloromethyl ketone (Novobiochem, Nottingham,
UK), 100 µmol/L N-tosyl-lysyl chloromethyl ketone
(Sigma, Poole, UK), and 10 µmol/L pepstatin A
(Novobiochem) were added, and cells were lysed by two freeze thaw
cycles. The resulting lysate was sonicated on ice for six bursts of 6
sec each. Unlysed cells were then removed by centrifugation at
7000 x g. The total amount of protein in cell extracts
was determined by the method of Bradford (20).
Serum samples from a pool of 20 normal controls and from three ATD patients positive for hNIS antibodies were preincubated in immunoprecipitation buffer with CHO-K1 or CHO-NIS12 cell extracts, both containing equal amounts of total protein. After a 12-h incubation at 4 C, in vitro-translated [35S]hNIS was added and immunoprecipitation was carried out as described previously. For each serum, the results are expressed as: the mean cpm immunoprecipitated in three experiments as a percentage of the mean cpm immunoprecipitated in three experiments without preincubation with extract.
Immunoprecipitation assay for Pmel17 autoantibodies
Pmel17 antibody reactivity was determined by immunoprecipitation of [35S]-labeled Pmel17, as described previously (21). Pmel17-specific rabbit antiserum and Pmel17 cDNA are detailed elsewhere (21).
Statistical analysis
Differences in immunoprecipitation indices of the different groups were analyzed using Mann-Whitney tests. Differences in the frequency of iodide uptake inhibitory activity of GD and AH sera compared with controls were further tested using 2 x 2 contingency tables. Correlation was analyzed using Pearson’s correlation test.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To establish a stable cell line, CHO-K1 cells were transfected
with full-length hNIS and a total of 24 colonies were picked for
further analysis. Six colonies displayed 125I
uptake activity ranging from a 10- to 60-fold increase above
untransfected cells. One clone designated CHO-NIS12 was selected for
additional studies, with an iodide uptake activity that was 5-fold
higher than our previous stable cell line CHO-NIS9, which expressed a
truncated (aa 1–612) form of the symporter (Fig. 1
). It is not known whether this
increased activity is due to differences in the level of expression of
the symporter or whether it relates to a specific functional role of
the last 31 amino acids of the protein. The iodide uptake activity of
both CHO-NIS9 and CHO-NIS12 was completely abolished by incubating the
cells with 1 mmol/L perchlorate (Fig. 1).
|
; Ref. 9). On this basis, 16 GD (33%)
and 6 (21%) AH patient sera were positive for inhibition
activity (Table 1). Because a truncated
form of the symporter was used in this bioassay, some inhibitory
positive samples may have been missed if a critical epitope is situated
at the carboxy-terminal of the molecule. Therefore, 18 (11 GD and 7 AH)
sera were also tested using the new cell line CHO-NIS12, which
expressed the full-length hNIS. No additional positive samples could be
detected, suggesting that the CHO-NIS9 cells were sufficient for the
investigation of antibodies that modulate the hNIS function.
Furthermore, a highly significant correlation was detected between the
two cell lines by testing 10 different sera (r = 0.92,
P < 0.0001), indicating the validity of our previous
results using the truncated form of the hNIS.
|
|
). Rubidium uptake and rubidium uptake inhibition studies
Sera (three GD patients, three AH patients, and six controls),
which had been tested for iodide uptake inhibitory activity using
CHO-NIS9, were analyzed for their effect on the uptake of
86Rb in the same cell line. The AH and GD serum
samples chosen contained hNIS inhibitory activity. A mean percentage
(± SEM) inhibition of 86Rb uptake of
4.3 (±2.4) was determined for the group of control sera. Patient sera
with an inhibition activity of more than 22% (the mean percentage of
inhibition of rubidium uptake of six control sera + 3 SD)
were regarded as positive. On this basis, none of the patient sera
examined was positive for inhibition of
Na+/K+ ATPase activity
(Fig. 2A).
Six IgG samples from two normals, two AH patients, and two GD patients
were tested for their effect on rubidium uptake by CHO-NIS9 cells. The
AH and GD samples were selected because the sera contained iodide
uptake inhibitory activity. A mean percentage (±SEM)
inhibition of 86Rb uptake of 2.8 (±1.7) was
determined for the control IgGs. Any patient IgG with an inhibition
activity of more than 9.7% (the mean percentage of inhibition of
rubidium uptake of two control sera + 3 SD) was regarded as
positive. With respect to this, all four patient IgG samples tested
were negative for rubidium uptake inhibition activity (Fig. 2B).
Production of hNIS in a TnT system
In vitro-translated
[35S]-labeled hNIS was analyzed by SDS-PAGE and
autoradiography, revealing a protein product of 60 kDa (Fig. 3), which is similar in size to that
described for unglycosylated hNIS in human thyrocytes (16). On addition
of Canine Microsomal Membranes to the translation reaction, a band of
66 kDa was detected, which we assumed to represent the glycosylated
form of [35S]hNIS (Fig. 3). This increase in
molecular weight of hNIS could be specifically reversed by adding the
deglycosylating enzyme N-Glycosidase F to the reaction containing
glycosylated protein (Fig. 3).
|
The immunoreactivity of the in vitro-translated
recombinant [35S]hNIS was tested using rabbit
antisera at a 1:10 dilution; [35S]hNIS was
immunoprecipitated by anti-hNIS antiserum but not by antityrosinase
antiserum PEP7 (Fig. 4).
|
), whereas none of the AD or vitiligo samples was
positive (Fig. 5). Further analysis of GD
subgroups showed that 6 of the 25 (24%) untreated sera and 3 of 16
(19%) of those with extrathyroidal complications were positive.
|
2 = 6.2,
P = 0.01 and 2 = 4.5,
P = 0.03, respectively).
SDS-PAGE was used to check that the radioactivity immunoprecipitated by
each of the positive sera was due to [35S]hNIS
(Fig. 4). Positive sera immunoprecipitated a band of the correct size
when compared with that precipitated by hNIS-specific rabbit antiserum
used as a positive control.
Dilution experiments showed that the reactivity of hNIS antibody-positive sera was clearly evident with dilutions of up to 50-fold. The reactivity of the positive control anti-hNIS rabbit antiserum was detected with dilutions of up to 1000-fold.
Immunoprecipitation of [ 35S]-labeled glycosylated hNIS
We analyzed antibody binding to in vitro-translated [35S]-labeled glycosylated hNIS. Using 20 normal control sera, a reference range of hNISAb indices was established, as described previously. Fourteen GD and 10 AH sera were tested, including 12 (10 GD and 2 AH) samples that were positive in the bioassay. One of the GD sera that failed to bind unglycosylated [35S]hNIS showed reactivity against glycosylated [35S]-labeled symporter. In contrast, one AH serum that reacted with unglycosylated [35S]hNIS, failed to show binding the to glycosylated [35S]-labeled symporter.
Correlation between the bioassay and the binding assay for the detection of hNIS antibodies
The results of the bioassay and the binding assay for GD and
AH sera are summarized in Table 1. A significant correlation
(r = 0.49, P < 0.0001) was found
between the bioassay and the binding assay for hNIS antibody detection
(Fig. 6).
|
Cell extracts from CHO-NIS12 cells were used to preabsorb antibody
reactivity against the symporter, with extracts from untransfected
CHO-K1 cells being used as a control. Fig. 7 shows the results of absorption studies
on a pool of 20 normal controls and three ATD sera. Antibody reactivity
of the three hNIS antibody-positive sera was reduced by 20–35% when
preabsorbed with CHO-NIS12 extracts. Preincubation with CHO-K1 extracts
decreased antibody binding of the patient sera by 0–10%, indicating
some nonspecific absorption of antibody reactivity.
|
Sera were examined for antibodies to the melanocyte-specific protein Pmel17 to test the possibility that antibody binding to [35S]hNIS occurred due to nonspecific interactions with radiolabeled ligands. Ten GD, seven AH, eight control sera, and a Pmel17-specific rabbit antiserum (21) were analyzed for Pmel17 antibodies by immunoprecipitation (21). The patient sera selected were positive for hNIS reactivity in the immunoprecipitation assay. Antibodies to the melanocyte-specific protein Pmel17 were not detected in any of the patient sera. The mean immunoprecipitation index ± SEM of the Pmel17-specific rabbit antiserum, control, GD, and AH groups was: 3.63 ± 0.73, 1.00 ± 0.08, 1.01 ± 0.06, and 1.10 ± 0.10, respectively. No significant differences were found when comparing the immunoprecipitation indices of control sera with those of either GD or AH sera (P = 0.8 and P = 0.3, respectively).
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The molecular weight of the [35S]-labeled full-length hNIS was similar to that described for the full-length unglycosylated hNIS in thyroid cells (16) at 60 kDa. However, glycosylation of the symporter in the TnT system resulted in a 66-kDa molecule, which is smaller than the 77 kDa documented for the mature glycosylated hNIS in thyroid cells (16). This discrepancy in relative molecular weight may be due to less extensive glycosylation of the translated polypeptide or may be due to altered migration of the protein in SDS-PAGE with differing experimental conditions. The translated [35S]hNIS was immunoreactive and could be immunoprecipitated by specific anti-hNIS rabbit antiserum, but not by nonspecific antityrosinase rabbit antiserum, further confirming the identity of the translated and labeled material.
Overall, 22% (11 of 49) of GD and 24% (7 of 29) of AH sera reacted with the in vitro-translated unglycosylated [35S]hNIS, whereas AD, vitiligo (without associated ATD), or control sera showed no reactivity, indicating a high disease-associated specificity. Glycosylation of the [35S]hNIS had only a minor effect on increasing the sensitivity of the assay. Similar findings have been documented for the thyroid antigens thyroglobulin and thyroid peroxidase, as deglycosylation of these molecules had no effect on autoantibody binding (23). Interestingly, one sample that reacted with the unglycosylated [35S]hNIS, failed to show binding to the glycosylated symporter, suggesting that glycosylation can disrupt antibody binding sites in some cases.
Of the 11 GD sera positive in the hNIS binding assay, 8 (73%) were found to inhibit iodide uptake in hNIS-transfected CHO-K1 cells. This suggests the presence of hNIS antibodies in three samples that bind the symporter without modulating its function. With respect to the AH patient samples, three of the seven (43%) sera positive in the hNIS binding assay also inhibited iodide uptake in the bioassay, again suggesting that four of the sera contained hNIS antibodies that bind to the symporter but fail to affect its activity. However, because the truncated form of the hNIS (aa 1–612) was used in the bioassay, antibodies reacting with epitopes situated at the C-terminal end of the molecule may have missed. Further experiments on a CHO-K1 cell line expressing the full-length hNIS (CHO-NIS12) did not identify additional inhibitory positive samples among those sera that reacted with hNIS in the binding assay. In addition, sera were found to exert similar inhibitory activity on both CHO-NIS9 and CHO-NIS12 cell lines. The failure to find antibodies that inhibit the function of the full-length but not the truncated hNIS is not surprising, as the last 31 amino acids are likely to be intracellular, making this portion of the symporter inaccessible to antibodies in intact cells.
Although a significant correlation was detected between the bioassay and the binding assay, 11 of 22 (50%) sera that inhibited iodide uptake in hNIS-transfected cells did not bind the symporter. This could be due to a number of reasons. First, it is conceivable that the inhibitory activity of some sera is not antibody-mediated, although the results with purified IgGs described here and in our previous study (9) argue against this possibility. Second, some of the inhibition of iodide uptake may result from antibodies that do not act directly on the symporter. Experiments to analyze the effects of hNIS antibody-positive sera on the activity of the Na+/K+ ATPase, which is required indirectly for iodide uptake, indicated that the functioning of the sodium ion gradient was not affected by serum samples containing hNIS inhibitory activity. Alternatively, it may be that the in vitro-translated protein fails to bind these antibodies due to incorrect folding of the molecule, the absence of an abnormal pattern of glycosylation, or may simply reflect differential detection sensitivity in the two assays. This is not unprecedented, as similar findings have been documented for thyroid-stimulating autoantibodies (TSAb). In one study, one third of GD sera positive in the TSH binding inhibiting immunoglobulin assay (TBII) failed to bind in vitro-translated TSH receptor (13). In another study, using a similar system, none of the TBII-positive sera bound TSH receptor (14). Moreover, it is clear that TSAb and TBII levels do not correlate and some potent TSAb have no detectable TBII activity (24).
In summary, the present study describes the detection of antibodies in ATD sera that bind the hNIS expressed in a TnT system. About one third of hNIS antibodies were found to bind the symporter without modulating its activity. On the other hand, some hNIS antibodies can only be detected using a bioassay. Overall, we found that 19 of 49 (39%) GD and 10 of 29 (35%) AH sera were positive for hNIS antibodies in at least one of the assays, indicating that the frequency of antibody reactivity against the symporter in GD and AH sera is very similar. The immunoprecipitation assay further offers the possibility of carrying out epitope mapping of the hNIS, which would allow the detailed analysis of this novel thyroid autoantigen, in turn helping to determine the role that NIS autoreactivity plays in the pathogenesis of ATD.
|
Acknowledgments |
|---|
PEP7 antiserum. |
Footnotes |
|---|
2 Supported by grants from the Overseas Research Award and the
University of Sheffield.
Received January 27, 1999.
Revised October 5, 1999.
Accepted December 23, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  O. Dohan, A. De la Vieja, V. Paroder, C. Riedel, M. Artani, M. Reed, C. S. Ginter, and N. Carrasco The Sodium/Iodide Symporter (NIS): Characterization, Regulation, and Medical Significance Endocr. Rev., February 1, 2003; 24(1): 48 - 77. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. P. Taylor, R. A. Metcalfe, P. F. Watson, A. P. Weetman, and R. C. Trembath Mutations of the PDS Gene, Encoding Pendrin, Are Associated with Protein Mislocalization and Loss of Iodide Efflux: Implications for Thyroid Dysfunction in Pendred Syndrome J. Clin. Endocrinol. Metab., April 1, 2002; 87(4): 1778 - 1784. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. P. Armengol, M. Juan, A. Lucas-Martin, M. T. Fernandez-Figueras, D. Jaraquemada, T. Gallart, and R. Pujol-Borrell Thyroid Autoimmune Disease : Demonstration of Thyroid Antigen-Specific B Cells and Recombination-Activating Gene Expression in Chemokine-Containing Active Intrathyroidal Germinal Centers Am. J. Pathol., September 1, 2001; 159(3): 861 - 873. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Seissler, S. Wagner, M. Schott, M. Lettmann, J. Feldkamp, W. A. Scherbaum, and N. G. Morgenthaler Low Frequency of Autoantibodies to the Human Na+/I- Symporter in Patients with Autoimmune Thyroid Disease J. Clin. Endocrinol. Metab., December 1, 2000; 85(12): 4630 - 4634. [Abstract] [Full Text]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
