Graves' disease (GD), a prototypic example of human autoimmunity dominated by pathogenic autoantibodies, represents a complex multitissue syndrome that remains incompletely understood (1). Generation of activating antibodies directed against the TSH receptor (TSHR) accounts for thyroid glandular overactivity. Like many forms of autoimmunity, GD is associated with detectable autoantibodies directed at other antigens. It is uncertain whether the loss of immune tolerance to antigens besides TSHR plays any pathogenic role in GD. This question is particularly relevant to the manifestation of GD known as thyroid-associated ophthalmopathy (TAO), a process in which connective tissue in the orbit undergoes remodeling (2). Early clinical observations raised suspicion that antigens shared by orbit and thyroid might underlie the peculiar anatomical distribution of this disease. An obvious suspect is TSHR because it has been detected at relatively low levels in multiple depots of fatty connective tissue (3) where it possesses biological activity. But whether TSHR or the antibodies directed against it are directly or indirectly involved in the genesis of TAO has yet to be established. Support for its involvement derives from circumstantial evidence. For example, there exists a correlation between levels of anti-TSHR antibodies and the severity/activity of TAO (4). Moreover, these antibodies may have value in predicting disease outcome (5). Such associations suggest the potential for connectivity between TSHR and the disease but fall short of proving it. Another candidate antigen emerged decades ago with the detection of thyroglobulin in the diseased orbit (6) and the recent reports that have substantiated those findings.
Residual uncertainty concerning the identity of factors that underlie immune reactivity within the orbit in TAO has prompted a continued search for other antibody/antigen candidates. One such protein that has been linked to immune function is IGF-IR (7). It was nearly 20 years ago that Weightman et al (8) identified high-affinity IGF-I binding sites on the surface of orbital fibroblasts. Antibodies extracted from the sera of patients with GD (GD-IgGs) from 52% of donors with the disease, irrespective of whether or not they manifested clinical TAO, were found to displace 125I-IGF-I from these sites. In contrast, none of those from healthy donors exhibited displacement activity. Although these studies failed to identify the sites to which IGF-I was binding, the authors presumed them to be IGF-I receptor (IGF-IR). Subsequently, Pritchard et al (9) found that GD-IgG could activate the Akt/mTOR/p70s6k signaling pathway and induce expression of chemokines in TAO orbital fibroblasts, responses that were absent in fibroblasts from healthy tissue. Evidence that the activity of GD-IgG was mediated through IGF-IR came from studies demonstrating that an IGF-IR-blocking monoclonal antibody and a transfected dominant negative IGF-IR construct could abrogate these actions (10). Moreover, the IGF-IR-specific activating ligand Des 1–3 could mimic the actions of GD-IgG and IGF-I. These studies identified the binding sites to be IGF-IR using 125I-IGF displacement. Subsequent studies by Smith and Hoa (11) found that GD-IgG could also induce the generation of hyaluronan in TAO orbital fibroblasts. IGF-I could mimic these effects, but recombinant human TSH failed to do so. Subsequently, van Zeijl et al (12) also concluded that whereas recombinant human TSH was ineffective in up-regulating hyaluronan levels, GD-IgG enhanced its accumulation in cultures of differentiated TAO fibroblasts. Their studies failed to implicate TSHR-driven cAMP generation but suggested that an alternative mechanism, such as one related to the IGF-IR pathway, might be involved.
Two topically related papers appear in this issue of the JCEM, each presenting experimental findings that point us in different directions with regard to potential involvement of IGF-IR in GD. Minich et al (13) developed a novel assay for anti-IGF-IR antibodies, which they contend possesses diagnostic usefulness and is based on their generating an IGF-IR-luciferase fusion protein. In their study, human embryonic kidney cells (HEK 293 cells) were stably transfected with this construct. These cells were then subjected to immunoprecipitation with serum samples from healthy subjects and donors with TAO. Sera were examined from 108 consecutive patients with TAO, some sampled at multiple time points, and 92 healthy controls. Control sera were used to establish the lower limits for assay positivity. In applying these same criteria to sera from patients with TAO, the authors determined that a similar small fraction of samples from healthy controls and patients with GD (11 and 10%, respectively) tested positive for anti-IGF-IR antibody. The report included a number of characterizing and validating studies. Among them, the serum content of IGF-I failed to influence the assay results, and purified IgG behaved identically to serum samples from which it derived. The functional characteristics of anti-IGF-IR antibodies were assessed by treating HepG2 and MCF-7 cells with patient sera in the absence or presence of IGF-I. No stimulatory effects on receptor autophosphorylation could be ascribed to samples testing positive for anti-IGF-IR antibodies. Instead, these samples abrogated IGF-IR activation by IGF-I. Moreover, these IgGs reduced MCF-7 cell viability in the presence of IGF-I. Based on their findings, the authors conclude that anti-IGF-IR antibodies are uninvolved in the pathogenesis of TAO.
A second paper appearing in this issue of the JCEM from Varewijck et al (14) also reports studies attempting to detect anti-IGF-IR antibodies in patients with GD. In this case, the authors examined the relationship between levels of TSH-binding inhibitory Ig (TBII) and IGF-IR stimulatory activity as they relate to age in a cohort of 70 patients with GD. Their kinase receptor activation assay also utilized the HEK 293 line stably transfected with human IGF-IR for assessing phosphorylation of receptor tyrosine residues. Cell lysates were subjected to an anti-IGF-IR capture antibody and an anti-phospho-tyrosine detection antibody. In patients with TBII values at least 1 SD above the mean, a positive correlation existed between IGF-IR-stimulating activity and age. For those with lower levels of TBII, no such relationship could be identified. This escalating IGF-IR stimulating activity could be abolished by depleting sera of IgGs. The aggregate findings of this study include detecting IGF-IR stimulating antibody activities in a subset of patients with high TBII. Their observations differ from those of earlier studies that found IGF-IR stimulating activity in a majority of sera from patients with GD.
What accounts for the widely divergent results obtained in these 2 reports? Moreover, to what factor(s) might the conflicting evidence concerning involvement of IGF-IR in TAO and GD be attributed? Among them, synthesis of IGF-I, IGF-II, and IGF-I binding proteins (IGFBPs) was not monitored. Many, if not all, cell types express some or all of these factors, and yet their levels were not assessed in any study. Could the apparent failure to control for these variables have confounded experimental outcomes? In general, an underestimation of the complexities associated with the IGF-I/IGF-IR pathway appears to plague many studies. One need look no further than an examination of the biology surrounding IGFBPs, of which 6 have been identified (15). Not only do IGFBPs exert biological actions in an unligated state, but they also modulate the actions of IGF-I on IGF-IR and can influence post-IGF-IR signaling (16). The nuclear targeting of IGFBP3 (17) and its interaction with nuclear transcription factors such as retinoid X receptor (18) exemplify the diverse consequences of these proteins in eukaryotic cells. Because most of the parameters measured in the current studies rely on fold differences provoked in cultures receiving GD-IgG or IGF-I compared to those remaining untreated, differences in endogenous IGF-I, IGF-II, or IGFBPs generated during culture incubations could obscure potentially meaningful responses. Thus, the interpretation of the phosphorylation studies using HepG2 offered by Minich et al (13) appears to ignore the capacity of these cells to produce several IGFBPs that can attenuate activation of IGF-IR. In addition, both HepG2 and MCF-7 cells synthesize IGF-II, which could maximally stimulate receptor activation and cell proliferation. In that case, responses to the exogenous test factors might have been underestimated.
An important divergence of both Minich et al (13) and Varewijck et al (14) from earlier studies derives from the fact that neither of these reports relies on human fibroblasts as cellular targets in the assays used to detect anti-IGF-IR antibodies. Both Weightman et al (8) and Pritchard et al (9, 10) used human fibroblasts. The molecular neighborhood in which IGF-IR functions in primary human fibroblasts may differ from that of immortalized cell lines and tumor cells used in both current studies. The reports from Pritchard et al (9, 10) and Smith and Hoa (11) suggest that important intrinsic differences might set apart TAO orbital fibroblasts from other cell types. The potential impact of these apparent differences as determinants of cellular response, which have yet to be identified, should not be ignored.
The report by Minich et al (13) describes attempts to identify antibodies against IGF-IR using an assay based on immunoprecipitation. Important limitations of their techniques, correctly acknowledged by the authors, include the possibility that relatively low affinity antibodies may have gone undetected in their assay. This stems from the requirement for antibodies to possess high avidity to enable them to bring antigens out of solution. Important examples of low-affinity pathogenic antibodies have been described in the literature. Moreover, the likely medley of antibodies present in sera from patients with GD might block or modify anti-IGF-IR antibody interactions with the receptor in some cell types while promoting it in others. Another potential complication in interpreting the findings of Minich et al (13) relates to their setting the lower limits for assay positivity at a level that might be justified arithmetically but appears to be biologically arbitrary.
Resolving the conundrum of IGF-IR involvement in the pathogenesis of TAO will require open minds, acknowledging limitations of all studies, and more rigorous experimentation. We already know that the involvement of autoantibodies in TAO must be complicated. Substantial evidence currently exists that neither TSHR nor IGF-IR functions alone, but each is involved in crosstalk with a variety of cell surface and signaling molecules. Some investigators suspect interactions between TSHR and IGF-IR that involve the formation of physical and functional complexes associating the 2 proteins (19). Those studies indicate that blocking IGF-IR with a monoclonal antibody can attenuate signaling initiated at TSHR, observations that have been confirmed recently by another laboratory group (20). The aforementioned variables, including concentrations of IGF-I, IGF-II, and their molecular partners such as IGFBPs must be controlled for if studies performed under different conditions are to be compared. Ultimately, the observations made in vitro must be systematically applied to animal models that recapitulate critical aspects of the clinical disease. As the biological complexities of GD become better understood, concepts surrounding its pathogenesis must be appropriately revised. Strong opinion exists about whether antigens other than TSHR should be considered in the context of TAO (21). In the end, the value of identifying key molecules involved in the disease process will be measured by the expansion of our therapeutic armamentarium for difficult TAO. To this end, a multicenter clinical trial is being organized currently to assess the efficacy of IGF-IR blockade in severe, active TAO. Besides the obvious potential for clinical benefit to patients that would result from its success, the study may yield important insight into what role, if any, IGF-IR and its activation by autoantibodies might play in this disease.
Acknowledgments
The author is indebted to Ms. Linda Polonsky for her editorial expertise. Assistance from Ms. Dierdre Jeske and Mr. Mark Sandusky is gratefully acknowledged.
This work was funded in part by National Institutes of Health Grants EY008976, EY011708, and DK063121 and by continued support from the Bell Charitable Foundation.
Disclosure Summary: T.J.S. is a consultant for River View/Narrow River Management LP. He is an inventor on US Patents 7998681, 8178304, 6936426, and 8153121.
For articles see pages 752 and 769
- GD
- Graves' disease
- IGFBP
- IGF-I binding protein
- IGF-IR
- IGF-I receptor
- TAO
- thyroid-associated ophthalmopathy
- TBII
- TSH-binding inhibitory Ig
- TSHR
- TSH receptor.
References
- 1. Brent GA. Clinical practice. Graves' disease. N Engl J Med. 2008;358:2594–2605 [DOI] [PubMed] [Google Scholar]
- 2. Bahn RS. Graves' ophthalmopathy. N Engl J Med. 2010;362:726–738 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Feliciello A, Porcellini A, Ciullo I, Bonavolontà G, Avvedimento EV, Fenzi G. Expression of thyrotropin-receptor mRNA in healthy and Graves' disease retro-orbital tissue. Lancet. 1993;342:337–338 [DOI] [PubMed] [Google Scholar]
- 4. Gerding MN, van der Meer JW, Broenink M, Bakkar O, Wiersinga W, Prummel MF. Association of thyrotrophin receptor antibodies with the clinical features of Graves' ophthalmopathy. Clin Endocrinol (Oxf). 2000;52:267–271 [DOI] [PubMed] [Google Scholar]
- 5. Eckstein AK, Plicht M, Lax H, et al. Thyrotropin receptor autoantibodies are independent risk factors for Graves' ophthalmopathy and help to predict severity and outcome of the disease. J Clin Endocrinol Metab. 2006;91:3464–3470 [DOI] [PubMed] [Google Scholar]
- 6. Tao TW, Cheng PJ, Pham H, Leu SL, Kriss JP. Monoclonal antithyroglobulin antibodies derived from immunizations of mice with human eye muscle and thyroid membranes. J Clin Endocrinol Metab. 1986;63:577–582 [DOI] [PubMed] [Google Scholar]
- 7. Smith TJ. Insulin like growth factor-1 regulation of immune function: a potential therapeutic target in autoimmune diseases? Pharmacol Rev. 2010;62:199–236 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Weightman DR, Perros P, Sherif IH, Kendall-Taylor P. Autoantibodies to IGF-1 binding sites in thyroid associated ophthalmopathy. Autoimmunity. 1993;16:251–257 [DOI] [PubMed] [Google Scholar]
- 9. Pritchard J, Horst N, Cruikshank W, Smith TJ. Igs from patients with Graves' disease induce the expression of T cell chemoattractants in their fibroblasts. J Immunol. 2002;168:942–950 [DOI] [PubMed] [Google Scholar]
- 10. Pritchard J, Han R, Horst N, Cruikshank WW, Smith TJ. Immunoglobulin activation of T cell chemoattractant expression in fibroblasts from patients with Graves' disease is mediated through the insulin-like growth factor I receptor pathway. J Immunol. 2003;170:6348–6354 [DOI] [PubMed] [Google Scholar]
- 11. Smith TJ, Hoa N. Immunoglobulins from patients with Graves' disease induce hyaluronan synthesis in their orbital fibroblasts through the self-antigen, insulin-like growth factor-I receptor. J Clin Endocrinol Metab. 2004;89:5076–5080 [DOI] [PubMed] [Google Scholar]
- 12. van Zeijl CJ, Fliers E, van Koppen CJ, et al. Thyrotropin receptor-stimulating Graves' disease immunoglobulins induce hyaluronan synthesis by differentiated orbital fibroblasts from patients with Graves' ophthalmopathy not only via cyclic adenosine monophosphate signaling pathways. Thyroid. 2011;21:169–176 [DOI] [PubMed] [Google Scholar]
- 13. Minich WB, Dehina N, Welsink T, et al. Autoantibodies to the IGF1-receptor in Graves' orbitopathy. J Clin Endocrinol Metab. 2013;98:752–760 [DOI] [PubMed] [Google Scholar]
- 14. Varewijck AJ, Boelen A, Lamberts SW, et al. Circulating IgGs may modulate IGF-I receptor stimulating activity in a subset of patients with Graves' ophthalmopathy. J Clin Endocrinol Metab. 2013;98:769–776 [DOI] [PubMed] [Google Scholar]
- 15. Firth SM, Baxter RC. Cellular actions of the insulin-like growth factor binding proteins. Endocr Rev. 2002;23:824–854 [DOI] [PubMed] [Google Scholar]
- 16. Shen X, Xi G, Maile LA, Wai C, Rosen CJ, Clemmons DR. Insulin-like growth factor (IGF) binding protein 2 functions coordinately with receptor protein tyrosine phosphate β and the IGF-I receptor to regulate IGF-I-stimulating signaling. Mol Cell Biol. 2012;32:4116–4130 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Santer F, Bacher N, Moser B, et al. Nuclear insulin-like growth factor binding protein-3 induces apoptosis and is targeted to ubiquitin/proteasome-dependent proteolysis. Cancer Res. 2006;66:3024–3033 [DOI] [PubMed] [Google Scholar]
- 18. Schedlich LJ, Graham LD, O'Han MK, et al. Molecular basis of the interaction between IGFBP-3 and retinoid X receptor: role in modulation of RAR-signaling. Arch Biochem Biophys. 2007;465:359–369 [DOI] [PubMed] [Google Scholar]
- 19. Tsui S, Naik V, Hoa N, et al. Evidence for an association between thyroid-stimulating hormone and insulin-like growth factor receptors: a tale of two antigens implicated in Graves' disease. J Immunol. 2008;181:4397–4405 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Kumar S, Iyer S, Bauer H, Coenen M, Bahn RS. A stimulatory thyrotropin receptor antibody enhances hyaluronic acid synthesis in Graves' orbital fibroblasts: inhibition by an IGF-I receptor blocking antibody. J Clin Endocrinol Metab. 2012;97:1681–1687 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Wiersinga WM. Autoimmunity in Graves' ophthalmopathy: the result of an unfortunate marriage between TSH receptors and IGF-1 receptors? J Clin Endocrinol Metab. 2011;96:2386–2394 [DOI] [PubMed] [Google Scholar]