Adrenal-Cortex Autoantibodies and Steroid-Producing Cells

Adrenal-Cortex Autoantibodies and Steroid-Producing Cells Autoantibodies in Patients with Addison’s Disease: Comparison of Immunofluorescence and Immunoprecipitation Assays

C. BETTERLE, M. VOLPATO, B. PEDINI, S. CHEN, B. REES SMITH, AND Downloaded from https://academic.oup.com/jcem/article/84/2/618/2864396 by guest on 25 September 2021 J. FURMANIAK Department of Clinical Immunology and Allergy, Istituto di Semeiotica Medica (C.B., M.V., B.P.), University of Padova, Padova, Italy; FIRS Laboratories (S.C., B.R-S., J.F.), RSR Ltd., Cardiff CF4 5DU, United Kingdom; and Department of Medicine University of Wales College of Medicine (B.R-S., J.F.), Cardiff GF4 4XN, United Kingdom

ABSTRACT had adrenal-cortex autoantibodies or autoantibodies to 21-hydroxy- Autoimmune Addison’s disease and premature ovarian failure are lase. A high association between these two markers in patients with characterized by the presence of organ-specific autoantibodies. The different forms of autoimmune Addison’s disease and in those with main adrenal and gonadal autoantigens have been identified and short- or long-standing disease was found. Steroid-producing cells cloned, and the relationship between the autoantibodies detected by autoantibodies were found in 26% of the patients with autoimmune immunofluorescence techniques and those detected by the new assays Addison’s disease, and autoantibodies to 17␣-hydroxylase and/or using recombinant autoantigens needed to be investigated. We stud- P450 side chain cleavage enzyme in 36% of the patients. Steroid- ied 165 patients with Addison’s disease: 143 patients had different producing cells autoantibodies were found in 11/13 (85%) of patients forms of autoimmune Addison’s disease (13 with idiopathic premature with idiopathic premature ovarian failure associated with autoim- ovarian failure) and 22 had nonautoimmune Addison’s disease. Ad- mune Addison’s disease, and autoantibodies to 17␣-hydroxylase renal-cortex autoantibodies and steroid-producing cell autoantibod- and/or P450 side chain cleavage were found 12/13 (92%) of patients; ies were measured by the immunofluorescence techniques. Autoan- the only case negative for all these three markers suffered from tibodies to steroid 21-hydroxylase, 17␣-hydroxylase, and P450 side Turner’s syndrome. chain cleavage enzyme were measured by immunoprecipitation assay Provided that a high standard of immunofluorescence technique is using 35S-labeled recombinant proteins. maintained, measurement of adrenal cortex autoantibodies or ste- Adrenal-cortex autoantibodies and autoantibodies to 21-hydroxy- roid-producing cells autoantibodies by either immunofluorescence or lase were found in 81% of the patients with autoimmune Addison’s immunoprecipitation assay is essentially equivalent. (J Clin Endo- disease. None of the patients with nonautoimmune Addison’s disease crinol Metab 84: 618–622, 1999)

OR many years, the indirect immunofluorescence tech- These IPAs are highly specific and sensitive, and a good F nique (IIT) on cryostat sections of human and animal consistent agreement between results of 21-OH Abs mea- adrenal-cortex glands has represented the most reliable surement using these assays and ACAs by IIT was found method for the detection of adrenal-cortex autoantibodies (6–8). However, one report from a different laboratory sug- (ACAs) in Addison’s disease (AD), where they are present in gested no agreement between ACAs and 21-OH Abs results more than 90% of patients with recent-onset autoimmune mainly in patients with long-standing autoimmune AD (10). AD and are absent in those with nonautoimmune disease (1). The IIT on cryostat sections of testis and/or ovary has until ACAs are also markers of potential AD, particularly in chil- recent years been the only method available for the detection dren (2, 3). It has been shown that 21-hydroxylase (21-OH) of steroid-producing cells autoantibodies (StCAs) (1). In pa- is the major adrenal autoantigen in autoimmune AD irre- tients with autoimmune AD, StCAs generally correlate with spective of whether the disease presents as isolated, in the the presence of primary gonadal failure (hypergonadotropic context of autoimmune polyglandular syndrome (APS) type hypogonadism) (1, 11). In the absence of primary gonadal 1 or 2 (4–8,) or as potential (2, 3). failure, StCAs can herald the future development of the dis- Two different immunoprecipitation assays (IPAs) for the ease (12, 13). detection of 21-OH autoantibodies (21-OH Abs) have been 35S-Labeled recombinant antigens have also been used for 35 developed, one based on S-labeled 21-OH produced in an the detection of 17␣-hydroxylase autoantibodies (17␣-OH in vitro transcription-translation system (6, 9), the other based 125 Abs) and P450 side chain cleavage enzyme autoantibodies on I-labeled recombinant 21-OH produced in yeast (8). (P450scc Abs) (7). In this preliminary study a close association between StCAs detected by IIT and 17␣-OH and P450scc Received July 15, 1998. Revision received October 28, 1998. Accepted Abs assayed by IPA was demonstrated (7). In contrast, others November 3, 1998. studies indicated that there were some discrepancies be- Address all correspondence and requests for reprints to: Corrado ␣ Betterle, Istituto di Semeiotica Medica, University of Padova, via Os- tween StCAs and 17 -OH and P450scc Abs measurement pedale Civile 105, Padova, Italy 35128. E-mail: [email protected]. mainly in patients with APS type 1 (4, 14–17).

618 AUTOANTIBODIES IN ADDISON’S DISEASE 619

The aim of this study was to investigate the prevalence of Indirect IITs ␣ ACAs and StCAs by IIT and 21-OH Abs, 17 -OH Abs, and ACAs were tested by the classical IIT using thin cryosections of P450scc Abs by IPA in a large cohort of patients with different normal human adrenal tissue (2, 3). Positive sera were retested by forms of autoimmune and nonautoimmune AD with or with- doubling dilution to the end point by the same method. StCAs were out idiopathic hypergonadotropic hypogonadism and the tested by the indirect complement fixation test using thin cryosections associations between these markers. of normal human ovarian tissue (13). IPAs Materials and Methods 35S-labeled 21–0H, 35S-labeled 17␣-OH, and 35S-labeled P450scc were Patients produced in an in vitro transcription/translation system (Promega Corp., Southampton, UK), and the labeled proteins used in IPAs were We studied 165 Italian patients with primary clinical AD: 143 were ␣ affected by autoimmune AD and 22 were affected by nonautoimmune as described previously (6, 7) to test 21-OH Abs, 17 -OH Abs, P450scc AD. In the autoimmune AD group, 21 patients had APS type 1 (mean Abs reactivity of each serum. Downloaded from https://academic.oup.com/jcem/article/84/2/618/2864396 by guest on 25 September 2021 age 12.1 yr, mean duration of the disease 8.1 yr, range 0–30 yr), 55 patients had APS type 2 (mean age 31 yr, mean duration of the disease Statistical analysis 7.4 yr, range 0–46 yr), 67 patients had isolated AD (mean age 23.6, mean The statistical significance of the associations between autoantibodies duration of the disease 4.8 yr; range 0–30 yr). Thirteen females were ␹ affected by idiopathic premature ovarian failure (POF) (12 with auto- detected by different methods was determined by -square test. Cor- immune POF and 1 with Turner’s syndrome). With regard to the du- relation between ACA titers and 21-OH Ab levels was calculated after ration of the disease, data were collected from 125 patients with auto- logarithmic transformation of ACA titers and 21-OH. immune AD, from 62 patients with a disease duration of Յ2 yr (short- standing disease), and 63 patients with a disease duration Ͼ2 yr (long- RESULTS standing disease). In the group of patients with nonautoimmune AD, 10 ACAs and 21-OH Abs patients had disease due to tuberculosis, 9 had adrenoleukodystrophy, 2 had adrenal insufficiency due to primary adrenal neoplasia, and 1 had In patients with autoimmune AD, ACAs were found in congenital AD due to enzyme defect. All sera were coded and tested 116/143 (81%) patients and 21-OH Abs in 116/143 (81%) blindly for ACAs, StCAs, 21-OH-Abs, 17␣-OH Abs, and P450scc Abs. The correlation between ACA titers and 21-OH Ab levels was calculated patients. All 22 patients with nonautoimmune AD were neg- in 85 patients after logarithmic transformation of 21-OH Ab levels and ative for ACAs or 21-OH Abs (Fig. 1a). Prevalence of ACAs ACA titers. and 21-OH Abs varied in the three different groups of pa-

FIG. 1. Prevalence of ACAs, 21-OH Abs, in patients with autoimmune and nonautoimmune Addison’s disease. 620 BETTERLE ET AL. JCE&M• 1999 Vol 84 • No 2

itive for ACAs but negative for 21-OH Abs and 5 were low positive for 21-OH Abs but negative for ACAs). ACAs and 21-OH Abs were significantly associated in patients with autoimmune AD (P Ͻ 0.00001), APS type 1 (P ϭ 0.002) and type 2 (P Ͻ 0.001), isolated autoimmune AD (P Ͻ 0.001), and short- or long-standing disease (P Ͻ 0.0001). The comparison between ACA titers and 21-OH levels in 85 patients is shown in Fig. 2 (r ϭ 0.85).

StCAs, 17␣-OH Abs, and/or P450scc Abs StCAs were found in 37/143 (26%) patients with uns- ␣ elected autoimmune AD and 17 -OH and/or P450scc Abs Downloaded from https://academic.oup.com/jcem/article/84/2/618/2864396 by guest on 25 September 2021 were found in 51/143 (36%) of patients. All were also positive for ACAs and 21-OH Abs. Of 22 subjects with nonautoimmune AD, only one patient in the adrenoleukodystrophy group was positive for P450scc Abs at low levels; all were ␣ FIG. 2. Comparison between ACA titers and 21-OH Abs levels in 85 negative for StCAs and 17 -OH Abs (Fig. 3a). The prevalence patients with autoimmune Addison’s disease. of StCAs and 17␣-OH and P450scc Abs greatly varied in the different patient groups with autoimmune AD as shown in tients with autoimmune AD (Fig. 1b). In patients with short- Fig. 3b. StCAs were found in 11/13 (85%) of patients with standing disease, the prevalence of ACAs was 90% (56/62) idiopathic POF associated with autoimmune AD, and and that of 21-OH Abs 92% (57/62); in patients with long- 17␣-OH and/or P450scc Abs in 12/13 (92%) of the patients. standing disease, the prevalence of ACAs was 79% (50/63) StCAs were also present in 26/130 (20%) patients without and that of 21-OH Abs 78% (49/63) (Fig. 1c). The relationship POF, whereas 17␣-OH and/or P450scc Abs were present in between the two tests is shown in Fig. 1d. Sera from 155/165 39/130 (30%) of this non-POF group (Fig. 3c). All but one of (94%) patients were concordant in the two assays and sera the 26 StCA-positive patients were also positive for 17␣-OH from 10 patients showed discrepant results (5 were low pos- and/or P450scc Abs. The association between these assays is

FIG. 3. Prevalence of StCAs, 17␣-OH, and P450scc Abs in patients with autoimmune and nonautoimmune Addison’s disease. AUTOANTIBODIES IN ADDISON’S DISEASE 621 shown in Figure 3d: 148/165 (90%) sera were concordant in different forms (APS type 1 and 2 and isolated), in different the two assays, 16 sera showed discrepant results (15 were stages (potential, subclinical, and clinical), and with different low positive for 17␣-OH Abs and/or P450scc Abs but neg- durations of the disease. ative for StCAs and 1 was positive for StCAs and negative for The reasons for the discrepancy between our observations 17␣-OH Abs and/or P450scc). StCAs were significantly as- and results reported by Falorni et al. (10) may be related to sociated with 17␣-OH and/or P450scc Abs in patients with the demanding technical aspects of the ACA determination APS type 1 (P ϭ 0.001) and 2 (P Ͻ 0.001), isolated autoim- by IIT rather than to the selection of patients or to the du- mune AD (P Ͻ 0.001), and POF (P Ͻ 0.0001). ration of autoimmune AD. Furthermore, the current study Table 1 shows the immunological combinations in the 13 shows that measurement of ACAs by IIT or 21-OH Abs by patients with idiopathic POF associated with AD. The patient IPA is essentially equivalent and that they remain the best affected by Turner’s syndrome was the only case negative for markers for the identification of the patients with autoim-

all the three autoantibodies. mune AD (clinical, subclinical, or potential). Downloaded from https://academic.oup.com/jcem/article/84/2/618/2864396 by guest on 25 September 2021 Idiopathic POF can be associated with autoimmune AD, Discussion but its prevalence is quite different in the various forms of the The laboratory diagnosis of autoimmune diseases is en- disease. It is often associated with APS type 1, whereas it is tering a new era as an increasing number of autoantigens less frequent in APS type 2 and quite rare in patients with recognized by circulating autoantibodies in organ-specific isolated autoimmune AD (19). StCAs are well-established autoimmune diseases have been identified and cloned (18). markers of POF associated with autoimmune AD but are In this context, 21-OH, 17␣-OH, and P450scc have been rec- quite rare in other forms of POF (1, 11, 20). ognized as the main autoantigens in adrenal and ovarian In the present study we showed that in patients with autoimmunity (1). Previous studies from our laboratory have autoimmune AD the frequency of StCAs greatly varied, but they are highly associated with measurements of autoanti- shown good agreement between ACAs detected by IIT and ␣ 21-OH Abs detected by IPA (6–8) not confirmed by others bodies to 17 -OH Abs and/or P450scc Abs, which have been mainly in patients with long-standing disease (10). identified as the main gonadal autoantigens (7). In patients ϭ with idiopathic POF in the context of autoimmune AD, The present report on a large number of sera (n 165) from ␣ patients with different forms of AD demonstrates unequiv- StCAs, and 17 -OH and/or P450scc Abs must be considered ocally that ACAs and 21-OH Abs measurements are closely good markers of autoimmune POF. The only patient with autoimmune AD and idiopathic POF negative for StCAs, associated, with 94% of the sera showing concordant results ␣ in the two assays. Only 6% of the sera gave discrepant results 17 -OH Abs, and P450scc Abs disclosed a nonautoimmune but were, in general, at a low titer. A strong association POF. However, in patients with autoimmune AD without POF there were 16 discrepant results between the two tests. between ACAs and 21-OH Abs was recently confirmed in a ␣ follow-up study on 58 patients with organ-specific autoim- In particular, sera from 15 patients were positive for 17 -OH mune diseases without overt hypoadrenalism (2, 3). Further- Abs and/or P450scc Abs but negative for StCAs (Fig. 3d). This may reflect a greater sensitivity of the IPA based on more, during the follow-up, all ACAs and 21-OH Abs- 35 positive children developed overt adrenal failure (3); in the S-labeled recombinant autoantigens produced in the in case of the adults, all of the 12 patients who progressed to vitro transcription/translation system. clinical AD during the follow-up were positive for both Moreover, in the case of autoimmune AD without POF, ACAs and 21-OH Abs (2). StCAs have been demonstrated to be good markers of po- The current study is in good agreement with our previous tential autoimmune POF only in females (12, 13). Patients observations (2, 3, 6–8) and confirms in a larger number of with adrenal autoimmunity without POF who are positive ␣ patients that ACA or 21-OH Ab measurements are valuable for 17 -OH Abs and/or P450scc Abs in the absence of StCAs markers for identifying patients with autoimmune AD in may also be at risk for developing POF. A further follow-up study of such patients should be helpful to assess the role of 17␣-OH Abs and/or P450scc Abs alone in the natural history TABLE 1. Immunological pattern in 13 patients with POF and autoimmune Addison’s disease (AD). of the POF (13). The autoimmune AD and POF appear to be diseases me- Patients ACAs/21-OH Abs StCAs 17␣-OH Abs P450scc Abs diated by cytotoxic T lymphocytes. ACAs and 21OH Abs and 1. APS 1 ϩ/ϩϩϩ ϩStCAs, 17␣-OH Abs, and P450scc Abs are likely to be mere 2. APS 1 ϩ/ϩϩϩ ϩmarkers of the autoimmune process. The study of the 3. APS 1 ϩ/ϩϩϩ ϩepitopes recognized by ACAs and 21-OH Abs was unable to 4. APS 1 ϩ/ϩϩϩ ϩ 5. APS 1a ϩ/ϩϪϪ differentiate patients with different forms of autoimmune 6. APS 2 ϩ/ϩϩϩ ϩAD or potential AD (21). Only the identification of the au- 7. APS 2 ϩ/ϩϩϪ ϩtoepitopes recognized by autoreactive T lymphocytes infil- 8. APS 2 ϩ/ϩϩϪ ϩtrating the adrenal cortex and the ovary could improve our ϩ ϩϩϪ ϩ 9. APS 2 / knowledge of the pathogenesis of this disorder. 10. APS 2 ϩ/ϩϪϪ ϩ 11. APS 2 ϩ/ϩϩϩ Ϫ Overall, our studies indicate that ACAs and 21-OH Abs are 12. APS 2 ϩ/ϩϩϩ ϩcharacteristic of the four different forms of autoimmune ADs 13. Isolated AD ϩ/ϩϩϩ ϩ(APS type 1, APS type 2, isolated, and potential) and StCAs ϩ, Positive; Ϫ, negative. and autoantibodies to 17␣-OH and/or P450scc appear to be a Turner’s syndrome. good markers of POF associated with adrenal autoimmunity. 622 BETTERLE ET AL. JCE&M• 1999 Vol 84 • No 2

ACA and StCA measurements are demanding technically 10. Falorni A, Laureti S, Nikoshkov A, et al. 1997 21-hydroxylase autoantibodies but provided that this is recognized, the measurement of in adult patients with endocrine autoimmune diseases are highly specific for Addison’s disease. Clin Exp Immunol. 107:341–346. adrenal autoantibodies by either immunofluorescence or im- 11. Irvine WJ, Chan MMW, Scarth L. 1968 Immunological aspects of premature munoprecipitation assays is essentially equivalent. ovarian failure associated with idiopathic Addison’s disease. Lancet II:886–887. 12. Ahonen P, Miettinen A, Perheentupa J. 1987 Adrenal and steroidal cell an- References tibodies in patients with autoimmune polyglandular disease type I and risk of 1. Betterle C, Volpato M. 1998 Review article: adrenal and ovarian autoimmu- adrenocortical and ovarian failure. J Clin Endocrinol Metab. 64:494–500. nity. Eur J Endocrinol. 138:16–25. 13. Betterle C, Rossi A, Dalla Pria S, et al. 1993 Premature ovarian failure: 2. Betterle C, Volpato M, Rees Smith B, et al. 1997 I. Adrenal cortex and steroid autoimmunity and natural history. Clin Endocrinol (Oxf). 39:35–43. 21-hydroxylase autoantibodies in adult patients with organ-specific autoim- 14. Krohn K, Uibo R, Aavik E, Peterson P, Savilahti K. 1992 Identification by mune diseases: markers of low progression to clinical Addison’s disease. J Clin molecular cloning of an autoantigen associated with Addison’s disease as Endocrinol Metab. 82:935–938. steroid 17 ␣-hydroxylase. Lancet. 339:770–773. 3. Betterle C, Volpato M, Rees Smith B, et al. 1997 II. Adrenal cortex and steroid 15. Uibo R, Aavik E, Peterson P, et al. 1994 Autoantibodies to cytochrome P450

21-hydroxylase autoantibodies in children with organ-specific autoimmune enzymes p450scc, P450c17, and P450c21 in autoimmune polyglandular disease Downloaded from https://academic.oup.com/jcem/article/84/2/618/2864396 by guest on 25 September 2021 diseases: markers of high progression to clinical Addison’s disease J Clin types I and II and in isolated Addison’s disease. J Clin Endocrinol Metab. Endocrinol Metab. 82:939–942. 78:323–328. 4. Winqvist O, Karlsson FA, Ka¨mpe O. 1992 21-hydroxylase, a major autoan- 16. Winqvist O, Gebre-Medhin G, Gustafsson J, et al. 1995 Identification of the tigen in idiopathic Addison’s disease. Lancet. 339:1559–1562. main gonadal antigens in the patients with adrenal insufficiency and associ- 5. Bednarek J, Furmaniak J, Wedlock N, et al. 1992 Steroid 21-hydroxylase is a ated ovarian failure. J Clin Endocrinol Metab. 80:1717–1723. major autoantigen involved in adult onset autoimmune Addison’s disease. 17. Peterson P, Uibo R, Pera¨nen J, Krohn K. 1997 Immunoprecipitation of ste- FEBS Lett. 309:51–55. roidogenic enzyme autoantigens with autoimmune polyglandular syndrome Colls J, Betterle C, Volpato M, Rees Smith B, Furmaniak J. 6. 1995 A new type I (APS I) sera; further evidence for humoral immunity to P450 c17 and immunoprecipitation assay for autoantobodies to steroid 21-hydroxylase in P450 21. Clin Exp Immunol. 107:335–340. Addison’s disease. Clin Chem. 41:375–380. 18. Song YH, Li Y, Maclaren N. 1996 The nature of autoantigens targeted in 7. Chen S, Sawicka S, Betterle C, et al. 1996 Autoantibodies to steroidogenic enzymes in autoimmune polyglandular syndrome. Addison’s disease and autoimmune endocrine diseases. Immunol Today. 17:232–238. premature ovarian failure. J Clin Endocrinol Metab. 81:1871–1876. 19. Anonymous. 1992 Enzymes as autoantigens (Editorial). Lancet. 339:779–780. 8. Tanaka H, Powell M, Chen S, et al. 1997 Autoimmune adrenal diseases—a 20. Sotsiou F, Bottazzo GF, Doniach D. 1980 Immunofluorescence studies on new sensitive assay for measurement of steroid 21-hydroxylase autoantibod- autoantibodies to steroid-producing cells, and to germline cells in endocrine ies. J Clin Endocrinol Metab. 82:1440–1446. disease and infertility. Clin Exp Immunol. 39:97–111. 9. Falorni A, Nikoshkov A, Laureti S, et al. 1995 High diagnostic accuracy for 21. Volpato M, Prentice L, Chen S, Betterle C, Rees Smith B, Furmaniak J. 1998 idiopathic Addison’s disease with a sensitive radiobinding assay for autoan- A study of the epitopes on steroid 21-hydroxylase recognized by autoanti- tibodies against recombinant human 21-hydroxylase. J Clin Endocrinol Metab. bodies in patients with or without Addison’s disease. Clin Exp Immunol. 80:2752–2755. 111:422–428.