Antigenic "Hot- Spots" on the TSH Receptor Hinge Region

Abstract

The TSH receptor (TSHR) hinge region was previously considered an inert scaffold connecting the leucine-rich ectodomain to the transmembrane region of the receptor. However, mutation studies have established the hinge region to be an extended hormone-binding site in addition to containing a region which is cleaved thus dividing the receptor into $\underset{|}{\overset{\text{'}}{\alpha }}$ (A) and β (B) subunits. Furthermore, we have shown in-vitro that monoclonal antibodies directed to the cleaved part of the hinge region (often termed "neutral" antibodies) can induce thyroid cell apoptosis in the absence of cyclic AMP signaling. The demonstration of neutral antibodies in patients with Graves' disease suggests their potential involvement in disease pathology thus making the hinge a potentially important antigenic target. Here we examine the evolution of the antibody immune response to the entire TSHR hinge region (aa280-410) after intense immunization with full-length TSHR cDNA in a mouse (BALB/c) model in order to examine the immunogenicity of this critical receptor structure. We found that TSHR hinge region antibodies were detected in 95% of the immunized mice. The antibody responses were largely restricted to residues 352-410 covering three major epitopes and not merely confined to the cleaved portion. These data indicated the presence of novel antigenic "hotspots" within the carboxyl terminus of the hinge region and demonstrate that the hinge region of the TSHR contains an immunogenic pocket that is involved in the highly heterogeneous immune response to the TSHR. The presence of such TSHR antibodies suggests that they may play an active role in the immune repertoire marshaled against the TSHR and may influence the Graves' disease phenotype.

Keywords: Graves' disease (GD); TSH receptor; ectdomain; hinge antibodies; neutral antibodies.

Figure 1

Immunization scheme of the mouse model. Female Balb/c mice of 6–8 weeks were used in this study. Test animals (n = 20) were injected with expression plasmids pEGFP-hTSHR and control animals (n = 10) with pEGFP-N1 empty vector, respectively with 50 μl of plasmid DNA in the bicep femoris muscle and electroporated as described in materials and method. This immunization scheme was done every 3 weeks for a total of 4 times. The mice were bled at 0, 5, 10, 13, 19, and 24 weeks.

Figure 2

Receptor structure and hinge region amino acid sequence. (Left) Model of the full-length human TSH receptor showing the major concave leucine-rich domain (gray), the large hinge region (orange) connecting to the helical transmembrane domain (yellow) with its short carboxyl tail and extra- and intra- cellular loops (25). The intracellular loops of receptor harbor sites for downstream effectors such Gαs, Gαq, Gβγ, Gα12, and recruitment of β-arrestins thus forming the array of signaling units of the TSHR. (Right) A magnified image with amino acid sequence of the hinge region (aa 280–410) of the receptor with the “hot-spot” peptides marked in color. Shown here are the 3 “hot-spots” of the hinge region described in this study: peptide 23 (residues 352–371), peptide 24 (residues 367–386), and peptide 25 (residues 382–401) with there overlapping regions indicated in blue here. Note the 50 amino acid cleaved region (amino acids 316–366- indicated as dotted line in the above model) of the hinge has an overlap with peptide 23.

Figure 3

Thyroid function tests of immunized animals. (A) Serum TSH levels: T4 and TSH were detected using multiplex bead assay (Millipore) as described in materials and method. The average TSH level was 1.46 ± 1.26 ng/dL in the control animals vs. 2.22 ± 6.56 ng/dL in the immunized group. When the immunized group was subdivided into low and high TSH groups, 25% of mice had high TSH levels. (B) Serum T4 levels: Average T4 level was 4.29 ± 1.72 μg/dL in the control animals vs. 5.20 ± 2.32 μg/dL in the immunized group. 20% of the immunized animals had elevated T4 with corresponding suppressed TSH. There was significant difference (p < 0.0001) between the high TSH and the control group only.

Figure 4

Total TSH receptor antibodies in immunized mice. (A) Percentage of TSHR antibody: Flow cytometry was carried out at week 10 with serum samples from immunized and control animals using CHO-TSHR expressing cells to estimate TSH receptor binding antibodies. By this analysis, 95% of the immunized mice had conformational TSHR binding antibodies in their serum. The dotted line is the average response seen in the 10 control animals for TSHR antibodies (≤0.2%). (B) TSHR inhibiting antibody. To examine TSH blocking antibodies in the serum of immunized animals we used the TSHRGlo cells and performed the assay as described in M & M. K1-70 blocking monoclonal antibody was used as a positive control which gave a 39.7% inhibition to 60 μU of bTSH in the assay, as indicated by the top dotted line. The lower cutoff (bottom dotted line) is 2 standard deviations above the average of the control mice, which was 20.4%. By these criteria we observed 5 out of the 20 hTSHR immunized animals had blocking antibodies at week 10 and 80% of these were mice with high TSH.

Figure 5

Detecting linear antibodies by indirect peptide ELISA. (A) Antibodies to non-hinge region peptides: 17 overlapping 20-mer peptides representing the non-hinge region of the TSHR ectodomain spanning residues 1–281 of the ectodomain was used in the indirect peptide ELISA to detect linear binding antibodies in the serum of immunized mice. The cut-off as indicated by the dotted line is the highest value of all peptides reacted with control animal serum plus 2 standard deviation above the mean. Peptide 1 (residues 22–41) of the receptor showed overwhelming predominance of antibody response (85%) compared to other peptides 6 and 13. Antigenicity predicted by more than two open source software is marked below each peptide by stars. (B) Antibodies to hinge region peptides: To examine hinge specific antibodies in the immunized animals we used 9 peptides covering the entire hinge region (residues 277–415). The majority of immunized animals showed linear binding antibodies to peptides 23, 24, and 25 that were above cut-off (dotted line) determined by the control serum samples plus 2 standard deviation. Overall, 40% of the immunized mice showed antibody response to these novel hinge regions “hot-spots.” Antigenicity predicted by more than two open source software is marked below each peptide by stars.

Figure 6

Time course of response to hinge region “hot-spots” in immunize mice. Mice immunized with full-length hTSHR showed increasing antibody response to the hinge region “hot-spots” throughout the time course of the immunization described in Figure 1. As indicated here an antibody rise started as early as < 5 weeks and plateaued off at week 13 maintaining 85% positivity until the termination of the experiment. Dotted lines indicate immunization time points. Insert: Shows the percentage of hinge region “hot-spot” antibody positive mice by thyroid function at week 10. 56% of the antibody positive mice were in the normal TSH range and 22% were in each of the low and high TSH ranges.

Figure 7

Thyroid histology of immunized animals: Images of H&E stained slides were obtained using the Hamamatsu NanoZoomer digital scanner at 20x magnification and analyzed and depicted here by NDP.view2 software at 15x magnification. (A) Normal control mouse thyroid (B) Enlarged thyroid follicles (C) Dense micro infiltrates. These changes marked by arrow were all seen in some high hinge positive mice (2 out of 5 hTSHR immunized mice). The changes could not be conclusively associated with these antibodies because of the heterogeneity of the response observed in these immunized animals.

Figure 8

Adipogenesis in the retro orbital of immunized animals: Images of hematoxylin and eosin (H&E) stained slides were obtained using the Hamamatsu NanoZoomer digital scanner at 20x magnification and analyzed and depicted here by NDP.view2 software. Representative images of (A) Adipose tissue (ROI marked in red) around the optic nerve (ROI marked in yellow) of a control mouse. (B) Adipose tissue around the optic nerve of the TSHR immunized mice (C) To quantitate the changes in adipocytes in the control and immunized animals we used Image J on images at 10x magnification. The area of adipose was normalized by the area of the optical nerve on the same image for each orbit of animal. The immunized mice adipocyte to optic nerve area ratio was 2.88 ± 3.46 and that for the control mice was 2.59 ± 1.66. No significant difference in adipogenesis was seen in these immunized animals.