REVIEWS

Autoimmune Addison disease:

pathophysiology and genetic complexity
Anna L. Mitchell and Simon H. S. Pearce
Abstract | Autoimmune Addison disease is a rare autoimmune disorder with symptoms that typically develop
over months or years. Following the development of serum autoantibodies to the key steroidogenic enzyme,
21-hydroxylase, patients have a period of compensated or preclinical disease, characterized by elevations in
adrenocortocotropic hormone and renin, before overt, symptomatic adrenal failure develops. We propose that
local failure of steroidogenesis, causing breakdown of tolerance to adrenal antigens, might be a key factor
in disease progression. The etiology of autoimmune Addison disease has a strong genetic component in
man, and several dog breeds are also susceptible. Allelic variants of genes encoding molecules of both the
adaptive and innate immune systems have now been implicated, with a focus on the immunological synapse
and downstream participants in T lymphocyte antigen-receptor signaling. With the exception of MHC alleles,
which contribute to susceptibility in both human and canine Addison disease, no major or highly penetrant
disease alleles have been found to date. Future research into autoimmune Addison disease, making use of
genome-wide association studies and next-generation sequencing technology, will address the gaps in our
understanding of the etiology of this disease.
Mitchell, A. L. & Pearce, S. H. S. Nat. Rev. Endocrinol. 8, 306–316 (2012); published online 31 January 2012; doi:10.1038/nrendo.2011.245

Introduction
Autoimmune Addison disease (AAD) is a rare endo-
crine condition with a prevalence in white European
populations of 110–140 cases per million,1–3 making it
30-fold less prevalent than type 1 diabetes (T1DM) and
200-fold rarer than autoimmune thyroid diseases. Like
other autoimmune diseases, AAD is more frequent in
women (female:male ratio 1.5–3.5:1) and often pres-
ents in indivi­duals between the ages of 30 and 50 years,
although it can affect individuals at any age.1
The combination of fatigue, hypotension, weight loss
and skin hyperpigmentation associated with pathological
changes in the adrenal glands was first recognized by Dr
Thomas Addison, a British physician, in 1855.4 Addison’s
original description referred predominantly to patients
who had tuberculous infiltration of the adrenal glands.5
However, in the developing world, autoimmune destruc-
tion of the adrenal glands is now the most common cause
of Addison disease,1,6–8 although many other causes of
primary adrenal failure are known (Box 1). Regardless
of etiology, Addison disease was invariably fatal until the
first synthetic cortisol precursors were developed, which
were introduced as a treatment for this disorder in the
Institute of Genetic
Medicine, Newcastle 1940s. Such treatment revolutionized the management of
University, International AAD, transforming it into a chronic condition.
Centre for Life, Central
Parkway, Newcastle
Although AAD has long been recognized to be an
upon Tyne NE1 3BZ, UK inherited disorder, investigating the genetic basis of rare,
(A. L. Mitchell, complex diseases is challenging and a number of factors
S. H. S. Pearce).
potentially complicate the process. The genotype of a
Correspondence to:
A. L Mitchell
anna.mitchell@ Competing interests
ncl.ac.uk The authors declare no competing interests.
patient with AAD does not always predict their pheno-
type, and vice versa.9 In addition, mutations associated
with AAD show incomplete penetrance, which means
that an individual who inherits a known disease suscepti-
bility allele might not manifest the disease. Furthermore,
the occurrence of phenocopies means that environ­
mental factors could induce an AAD phenotype in an
indivi­dual without known disease susceptibility alleles
that is identical to that of a patient who does carry such
alleles. Finally, variants in any one of several genes could
result in the same phenotype (genetic heterogeneity), and
multiple variants are expected to be necessary to produce
the phenotype (polygenic inheritance).9
This Review article explores the pathogenesis of AAD,
in particular the potential mechanisms involved in the
breakdown of immune tolerance to adrenal antigens.
We also outline our current understanding of the genetic
architecture of this condition derived from studies in
humans and in breeds of dogs that can develop Addison
disease with a similar presentation to humans. Armed
with a more detailed understanding of pathophysiology
and the underlying genetic susceptibility to this rare con-
dition, it is possible that disease-modifying treatments
could be feasible in the near future.
Pathogenesis of AAD
In AAD, steroidogenic enzymes in the adrenal cortex,
normally recognized as self-antigens, become the targets
of a misdirected immunological attack. Circulating anti-
bodies directed against steroid 21-hydroxylase can be
detected in around 85% of patients who present with

306  |  MAY 2012  |  VOLUME 8  www.nature.com/nrendo

© 2012 Macmillan Publishers Limited. All rights reserved

idiopathic primary adrenal failure, defining them as
having AAD.10 The proportion of individuals with idio-
pathic Addison disease in whom steroid 21-hydroxylase
autoantibodies cannot be detected but who have disease
with an autoimmune basis (such as immune responses
to other adrenal components) remains unknown.
Steroid 21-hydroxylase autoantibodies predominantly
have an IgG1 isotype and target the carboxy terminal
of this enzyme.11 Although these antibodies inhibit the
enzymatic activity of steroid 21-hydroxylase in vitro
by preventing its interaction with cytochrome P450
oxido­reductase,12,13 this finding has not been replicated
in vivo.14 In particular, the intracellular location of steroid
21-hydroxylase, which is located on the smooth endo-
plasmic reticulum of intact cells, precludes its direct
interaction with circulating autoantibodies. Despite
their presumed lack of functional activity, the presence
of circulating steroid 21-hydroxylase autoantibodies is
a reliable predictor of the development of AAD, since
they indicate an ongoing autoimmune process within
the adrenal glands. In addition to steroid 21-hydroxylase,
other autoantigens have also been identified in patients
with AAD, including steroid 17‑α-hydroxylase15 and the
cholesterol side-chain cleavage enzyme.16
Natural history
A series of phases have been proposed to occur in the
natural history of steroid 21-hydroxylase autoantibody-
positive AAD (Figure 1).17 During the earliest (potential)
AAD phase, autoantibodies are present, but parameters
of adrenal function are normal and no clinical features
of disease are present. The subclinical phase follows, in
which adrenal function gradually becomes impaired but
clinical symptoms of AAD are not present. Typically,
during this phase, an initial rise in plasma renin and/or
adrenocorticotropic hormone (ACTH) levels is evident
before basal, and then stimulated, circulating cortisol con-
centrations become subnormal (<550 nmol/l after tetra­
cosactide stimulation). Finally, clinical AAD develops,
in which the affected individual becomes sympto­matic,
often with fatigue and pigmentation at first.
In a clinical study, 31% of initially asymptomatic
individuals with steroid 21-hydroxylase auto­antibodies
developed AAD, although progression from the first
identification of antibodies to overt adrenal failure took
between 3 months and 11 years in different patients.18
Moreover, not all individuals with these autoantibodies
develop progressive disease; many remain in the poten-
tial phase, never developing AAD, and some even revert
to being antibody-negative.17, 18 Patients with high titers
of steroid 21-hydroxylase autoantibodies tend to prog-
ress more rapidly than those with low titers. A case study
that described a woman with a 9‑year history of hyper-
pigmentation, raised circulating ACTH levels and high
steroid 21-hydroxylase auto­antibody levels but a normal
cortisol response to adminis­tration of synthetic ACTH,
highlights that indivi­dual responses to the presence
of an ongoing autoimmune response are highly vari-
able.19 Similarly, case reports of one patient with remis-
sion of apparently established AAD20 and another with
NATURE REVIEWS | ENDOCRINOLOGY VOLUME 8  |  MAY 2012  |  307
© 2012 Macmillan Publishers Limited. All rights reserved
REVIEWS
Key points
■■ Autoimmune Addison disease (AAD) is among the rarest of the autoimmune
endocrinopathies
■■ An aberrant immune response directed, at least in part, at the steroidogenic
enzymes of the adrenal cortex (in particular 21-hydroxylase) underlies most
cases of AAD
■■ The progression of AAD can be slow, with an insidious decline into adrenal
failure over months and years
■■ Circumstantial evidence suggests that the glucocorticoid-rich adrenal milieu
might give natural protection to adrenal antigens from the immune system
■■ AAD has a strong genetic component, but owing to its rarity the recurrence rate
in family members is of the order of 2%
■■ Several disease-susceptibility alleles have been identified, including three loci
associated with organ-specific autoimmunity (MHC, CTLA4 and PTPN22), and
others that encode proteins involved in innate immune responses
Box 1 | The etiologies of primary adrenal failure
Impaired steroidogenesis
■■ Defects in cholesterol biosynthesis
■■ Smith–Lemli–Opitz syndrome
■■ A-β-lipoproteinemia
■■ Defects in steroid biosynthesis
■■ Steroidogenic acute regulatory protein mutations
■■ Mitochondrial mutations
■■ Mutations in genes encoding steroidogenic enzymes,
resulting in congenital adrenal hyperplasia
■■ Scavenger receptor BI mutations129
Adrenal dysgenesis or hypoplasia
■■ DAX1 mutations
■■ SF1 mutations
■■ ACTHR mutations
■■ GPX1130
■■ NNT130
Adrenal destruction
■■ Autoimmune responses
■■ Metastatic malignancy
■■ Amyloidosis
■■ Hemorrhage
■■ Infections
■■ Adrenoleukodystrophy
■■ Sarcoidosis
Abbreviations: ACTHR, adrenocorticotropic hormone receptor
gene; DAX1, dosage-sensitive sex reversal-adrenal hypoplasia
gene 1; GPX1, glutathione peroxidase 1; NNT, nicotinamide
nucleotide transhydrogenase; SF1, steroidogenic factor 1.
subclinical AAD and steroid 21-hydroxylase anti­bodies
that became undetect­able following glucocorticoid
therapy 21,22 suggest that even in patients with impaired
steroidogenesis, adrenal autoimmunity might not always
be permanent.
Links with other autoimmune disorders
Patients might have isolated AAD, or the disease might
occur as part of an autoimmune polyendocrinopathy
syndrome (APS). APS type 1 (APS1, also known as
autoimmune polyendocrinopathy, candidiasis and ecto-
dermal dystrophy syndrome) is a monogenic, autosomal-
recessive syndrome resulting from mutations in the AIRE
gene on chromosome 21q22.23–25 APS1 is rare in most
populations, with a prevalence of only 3 cases per million
REVIEWS

Stage 0 Stage 1 Stage 2 Stage 3

■ Local steroidogenesis ■ Focal lymphocytic ■ Area of lymphocytic ■ Systemic response

suppresses adrenal infiltration occurs
antigen presentation
within the adrenal ■ Local steroidogenesis
cortex is impaired
■ Adrenal antigens
are presented
■ Inflammatory cells
are recruited
infiltration expands targeted against
adrenal antigens
■ A cycle of antigen occurs
presentation, inflammatory
cell recruitment, cytokine ■ On-going autoimmune
release and adrenocyte attack gradually
damage begins destroys the adrenal
cortex bilaterally
■ Local steroidogenesis is
further impaired, ■ Adrenal failure results
perpetuating the cycle

Antigen
presentation

Inflammatory
infiltrate

Local glucocorticoid
concentrations

Phase of disease Potential Subclinical disease Clinical disease

Figure 1 | A hypothetical model of the pathogenesis of AAD based on the breakdown of glucocorticoid-induced immune
privilege. Gradual immune-mediated destruction of the adrenal gland occurs in patients with AAD. Stage 0 corresponds to
the ‘potential’ phase in the natural history of this disorder. Once tolerance to adrenal antigens is lost (stage 1), increased
antigen presentation leads to increasing inflammatory infiltrates in the gland. Adrenocytes are damaged by this immune
response and, in stage 2, local glucocorticoid production is impaired. Stages 1 and 2 correspond to the ‘subclinical’ phase
of AAD. Eventually, in stage 3, the adrenal cortex is destroyed and steroidogenesis ceases, which corresponds to the phase
of ‘overt’ clinical AAD. Abbreviation: AAD, autoimmune Addison disease.

individuals in the UK and affects males and females
equally. 26 However, this syndrome does occur more
frequently in certain groups, including Iranian Jewish
(1:9,000),27 Sardinian (1:14,000),28 Finnish (1:25,000)29
and Norwegian populations (1:90,000). 30 Affected
individuals usually present in childhood with chronic
mucocutaneous candidiasis, adrenocortical failure and
autoimmune hypoparathyroidism. They might also have
hypoplasia of the dental enamel or nail dystrophy,31–33
and can go on to develop other autoimmune disorders,
such as T1DM and pernicious anemia, at any time of
life.31 The molecular basis of APS1 has been reviewed in
detail in this journal.34
Autoimmune polyendocrinopathy syndrome type 2
(APS2) is defined as AAD in conjunction with auto­
immune thyroid disease and/or T1DM,31 and is present
in approximately 50% of individuals with primary adrenal
failure.35 Other autoimmune conditions might also arise
in individuals with APS2,31 most commonly pernicious
anemia and vitiligo. Like isolated AAD, APS2 is more
common in women than men, and the first mani­festations
often appear in the fourth decade of life.36,37
Animal models of AAD
Animal studies might provide useful genetic information
about AAD owing to careful, controlled breeding and
environments.9 Allergic adrenalitis has been experimen-
tally induced in a range of animals and results in lympho-
cytic infiltration of the adrenal glands and autoantibody
production,38 as it does in humans. In contrast to human
AAD, however, the disease in laboratory animals does
not progress to impaired steroidogenesis; these models
are, therefore, of limited relevance to the human disease.
Dogs, however, can spontaneously develop Addison
disease and the canine form of the disease is analogous
to human AAD in a number of ways. Canine Addison
disease is predominantly a disease of middle-aged, female
dogs, and it has a clinical presentation very similar to
that in humans.39,40 In affected dogs, common present-
ing symptoms include lassitude, vomiting and diarrhea,
muscle weakness and poor appetite. Blood tests com-
monly show abnormal electrolyte levels, which indicate
mineralocorticoid as well as glucocorticoid insufficiency.
As in humans, an ACTH stimulation test is used to
confirm the diagnosis. Canine Addison disease can affect
any breed of dog, and has an incidence ranging from 1.5%
to 9% in susceptible breeds (Portuguese Water Dogs,
Nova Scotia Duck-tolling Retrievers, Standard Poodles,
Leonbergers and Bearded Collies).41–44 The notably high
incidence of the disease in certain breeds is likely to reflect
founder effects owing to substantial inbreeding of pedi-
gree dogs and the frequent use of popular sires (Figure 2).
These factors strongly support a genetic eti­ology for
canine Addison disease43 and a number of loci have
been identified that confer susceptibility to this disease,
mostly through linkage studies. In Portuguese Water
Dogs, loci on chromosomes 12 and 37 have been linked
with canine Addison disease and these were identified
as orthologues of human MHC and the cytotoxic T lym-
phocyte antigen‑4 (CTLA4) gene regions, respectively.45

308  |  MAY 2012  |  VOLUME 8  www.nature.com/nrendo

© 2012 Macmillan Publishers Limited. All rights reserved

This association between canine Addison disease and the
MHC locus has been replicated in other breeds, including
the Nova Scotia Duck-tolling Retriever.46
The identification of genetic susceptibility to canine
Addison disease at loci that have also been implicated
in the human form of this autoimmune disorder sug-
gests that they share an underlying autoimmune etiology.
Although this contention is supported by the predilection
of both human and canine Addison disease for female
individuals, the existence of a common etiology has not
been definitively established from an immunological
perspective. Although antibodies to adrenal cell anti-
gens have been found by indirect immuno­fluorescence
in a small number of dogs, along with immune cell infil-
trates in the adrenal glands, no specific auto­antibodies
(against steroid 21-hydroxylase, for example) have been
identified.47 Canine Addison disease might, therefore, be
heterogeneous, with an auto­immune eti­ology in certain
breeds, but with an underlying develop­mental abnor­
mality of steroidogenesis in others. Information from
studies in humans with AAD has been useful for research
in dogs, but not vice versa. Whether further investiga-
tion of susceptible breeds, perhaps using dense linkage
marker maps in larger cohorts, will yield any insights into
human AAD remains to be seen.
Genetic basis of human non-APS1 AAD
Epidemiology
The genetic aspect of AAD susceptibility was recog-
nized 40 years ago and since then multiple reports of
concordant monozygotic twins have been published.48–51
Familial clustering of cases of AAD also suggests a
genetic basis for the condition.52,53 However, unlike the
monogenic form of AAD that occurs in APS1, isolated
AAD and AAD as a component of APS2 are not inherited
in a Mendelian fashion. Instead, like most auto­immune
endocrinopathies,54 they have a complex genetic etiology
requiring the interaction of several genetic variants and
environmental factors to cause disease.
Pooled data from studies encompassing more than
600 unrelated (non-APS1) AAD probands from the
UK and Norway have revealed only 14 families (2.3%)
that contain affected AAD sibling pairs (S. H. S. Pearce
& E. S. Husebye, unpublished work). These epidemio­
logical data enable, for the first time, a sibling recurrence
risk ratio (the risk of disease in a sibling of an individual
with the disease, divided by the population prevalence of
the disease) of between 160 and 210 to be calculated for
AAD, in contrast to T1DM and Graves disease, which
have sibling recurrence risk ratios of approximately 15
and 10, respectively.55,56
In addition, 50% of individuals with AAD will develop
other autoimmune conditions,8 and many report auto-
immunity in their families.57 In 22 AAD pedigrees con-
taining multiple affected individuals from the UK and
Norway, 24 of the 48 individuals with AAD (50%) had
one or more additional autoimmune conditions, most
commonly autoimmune thyroid disease (occurring
in 29% of the AAD cohort) and T1DM (occurring in
15% of the AAD cohort). These pedigrees included 80
NATURE REVIEWS | ENDOCRINOLOGY VOLUME 8  |  MAY 2012  |  309
© 2012 Macmillan Publishers Limited. All rights reserved
REVIEWS
Male dog without Addison disease Female dog without Addison disease
Male dog with Addison disease Female dog with Addison disease
Figure 2 | An example of a pedigree of Portuguese Water Dogs with heritable
Addison disease. All dogs in this pedigree trace back to common ancestors (not
depicted). A clear female preponderance to AAD is seen (15 female dogs are
affected compared with 8 males) and the relatively high disease prevalence
reflects significant in-breeding between pedigree members. Reproduced from
Oberbauer A. M. et al. BMC Vet. Res. 2, 15 (2006) which is published under a
Creative Commons Attribution License by BioMed Central.
Addison-disease-free relatives of patients with AAD,
15% of whom also reported autoimmune conditions,
including autoimmune thyroid disease, T1DM, perni-
cious anemia and vitiligo. This clustering of various
autoimmune disorders in individuals with AAD and
their families suggests that there are common suscepti-
bility loci for these disorders. Indeed, the strength of this
familial clustering predicts that some highly penetrant
AAD susceptibility alleles might exist.
Molecular genetic studies
Large-scale genetic analysis in humans with AAD has
not been done because of the rarity of the disease. The
investigations carried out so far have been candidate-
gene association studies with a case–control design con-
ducted in small cohorts of patients (Table 1). Candidate
genes are selected on the basis of underlying biological
plausibility (that is, because they have been implicated in
a related disease, such as T1DM or autoimmune thyroid
disease, or because they are associated with rare, mono-
genic variants of the disease, such as APS1). Despite
extensive research and the strong genetic contribution to
AAD, few susceptibility loci have been identified to date
using this candidate-gene approach. No single known
locus except for MHC makes a major contribution to an
individual’s genetic susceptibility to AAD. The major-
ity of the genetic component of AAD etiology, therefore,
remains undefined.
The AAD susceptibility loci that have been des­
cribed tend to influence immunological pathways.
REVIEWS

Table 1 | Selected case–control, candidate-gene studies in patients with non-APS1 Addison disease
Gene or marker Populations Number of patients Odds ratio P value
MHC class II loci
HLA-DR3–DQ2 Norway37 94 3.6 <10–7
HLA-DR3–DQ2 Finland, Russia, Estonia 131
69 (14 Finnish) 14.5 <0.0001
HLA-DR4–DQ8*0404 Norway37 94 NR <10–5
HLA-DR4–DQ8*0401 <10–4
HLA-DRB1*0301 Norway132 414 22.13 6 × 10–20
HLA-DRB1*0404
Other loci
MICA*5.1 Italy65 28 6.52 0.0015
USA62 46 22.5 <10–5
CYP21Adel Finland61 12 25.0 NR
CYP21 ± L10, R102, A494 Finland64 12 8.9 NR
CLEC16A (rs12917716) Norway71 332 0.72 0.0004
UK71 210 1.06 0.71
Combined71 542 0.81 0.006
CYP27B1 –1260C>A Germany118 124 1.18 0.0062 (genotype);
0.3354 (allele)
UK119 104 1.71 0.003
VDR Fok1 (exon 2) Germany 133
95 NR 0.0351 (genotype);
0.3390 (allele)
FCRL3 (rs7528684) UK104 200 1.61 0.0001
PTPN22 rs2476601 UK 90
104 1.69 0.031
Germany89 121 1.03 0.9878
Norway91 302 1.39 0.016
UK92 251 1.63 0.008
Poland92 87 1.84 0.010
CTLA4 A>G (exon 1) UK81 90 1.64 0.008
CTLA4 JO30G>A UK83 40 1.9 0.02
Norway83 94 1.4 0.04
Combined83 134 1.5 0.03
NLRP1 (rs12150220) Norway107 333 1.25 0.007
Poland108 101 1.5 0.015
PDL1 (rs1411262) UK99 315 1.33 0.032
Norway99 342 1.34 0.026
Combined99 657 1.32 3.03 × 10–3
Abbreviation: NR, not reported.

Auto­immunity is thought to arise as a result of disrup-
tions to both adaptive and innate immune responses. In
patients with AAD, the majority of susceptibility loci
identified exert their effects through the adaptive arm of
the immune system; however, some AAD susceptibility
loci are in genes of the innate immune system (Figure 3).
Adaptive immunity
The MHC locus is a genomic region on chromosome
6p21 that encodes the human leukocyte antigens
(HLA) and many other genes with and without known
immuno­logical functions. HLA proteins are expressed
on the surface of antigen-presenting cells and display
peptides (both self and nonself ) to T cells for patho-
gen surveillance by the immune system. Autoimmune
diseases arise when a persistent immune response is
aberrantly triggered by self peptides; in autoimmune
endocrinopathies, this response leads to destruction of
hormone-producing cells.
Variants within the MHC class II genes are strongly
associated with several autoimmune conditions,
possibly because some HLA-DR or HLA-DQ poly­
morphisms encode protein variants that enable self-
peptides to enter the antigen-binding pocket (Figure 4).
Although an association between AAD and HLA-DR3
alleles58 has been confirmed by a number of studies,59–61
initially, only one study reported an association with
HLA-DR4.58 Subsequent studies then confirmed asso-
ciations between AAD and the DRB1*04–DQA1*0301–
DQB1*0302 haplo­type;37,62 heterozygote carriers of the
DR3-DQ2 and DR4-DQ8 haplotypes were particularly
susceptible to this disease. In one study of Norwegian
patients, the DRB1*0404–DQ8 haplotype was par-
ticularly strongly associated with AAD, whereas the
DRB1*0401–DQ8 haplotype was protective against this
disease.37 The latter finding was replicated in a study
from the USA. 63 The association of more than one
HLA allele with AAD supports the hypothesis that the
disease can be triggered by autoimmune responses to
a range of peptide epitopes of steroid 21-hydroxylase
and potentially of other enzymes, rather than just one
specific epitope.

310  |  MAY 2012  |  VOLUME 8  www.nature.com/nrendo

© 2012 Macmillan Publishers Limited. All rights reserved

1 2 3 4 5 6 7 8 9 10
PTPN22 CTLA4 MICA PD-L1
FCRL3 MHC
Figure 3 | Schematic representation of candidate-gene loci that confer susceptibility to autoimmune Addison disease. The
cytogenetic bands of each chromosome are shown and the approximate position of the genes known to confer
susceptibility to autoimmune Addison disease are indicated.
Other genes within the MHC region are also associ-
ated with AAD; however, as they demonstrate strong
linkage disequilibrium with HLA alleles, these genes
probably do not confer an independent risk of AAD but
instead act as markers for the HLA alleles carried on the
adjacent chromosomal segment. The CYP21A2 (cyto-
chrome P450, family 21, subfamily A, poly­p eptide 2)
gene, which encodes the steroid 21-hydroxylase enzyme,
located in the MHC class III region, is one such example.
Polymorphisms in this gene have been studied, but their
association with AAD has been widely attributed to
linkage disequilibrium with MHC class II alleles. 61,64
One gene that does seem to be independently associ-
ated with AAD is the MHC class I polypeptide-related
sequence A (MICA) gene;65 the strength of the associa-
tion with the MICA*5.1 polymorphism is greater than
might be expected from linkage disequilibrium with the
HLA-DR3–HLA-DR4 haplotype alone.62,66
The MHC class II transactivator (CIITA) gene on
chromosome 16p13 encodes a protein that is a crucial
regulator of MHC class II expression. Its product CIITA
has a clear role in immunity and deleterious mutations
in this protein result in a severe monogenic primary
immunodeficiency disease, known as bare lymphocyte
syndrome. 67 Polymorphisms in the CIITA gene have
also been linked to the development of systemic lupus
erythematosus68 and rheumatoid arthritis,69 as well as
to AAD in Italian and Norwegian populations.70,71 The
AAD-associated CIITA variants lie in the promoter and
intron 3 regions, but the mechanism by which these poly-
morphisms influence susceptibility to AAD remains to
be fully elucidated.
Engagement of co-stimulatory molecules is required
to complete the T lymphocyte activation signal and
enable an immune response to be mounted, following
the encounter of an antigenic peptide in the MHC-
binding groove by the T cell receptor complex (Figure 4).
The CTLA4 gene on chromosome 2q33 encodes a co-­
stimulatory molecule that has a role in downregulating
NATURE REVIEWS | ENDOCRINOLOGY VOLUME 8  |  MAY 2012  |  311
© 2012 Macmillan Publishers Limited. All rights reserved
REVIEWS
11 12 13 14 15 16 17 18 19 20 21 22 X Y
VDR CLEC16A NLRP1
CYP27B1 CIITA
T-cell responses.72 Genomic variants at the CTLA4 locus
have been implicated in the etiology of numerous auto-
immune conditions, including autoimmune thyroid
diseases,73–76 T1DM77,78 and rheumatoid arthritis.79,80
CTLA4 polymorphisms have also been associated with
AAD in a number of independent cohorts: an A or G
single nucleotide polymorphism (SNP) in exon 1, 81
an AT repeat in the 3' untranslated region of exon 3 82
and G or A alleles of the JO30 SNP downstream of this
gene.83 Carriers of the downstream AAD susceptibil-
ity alleles (3' UTR and intergenic region) are thought
to have decreased levels of a soluble, secreted isoform
of CTLA‑4 that is able to engage CD80 and CD86
mol­e cules on antigen-­presenting cells. These poly­
morphisms, therefore, might act as a positive regulator
of the adaptive immune response by enabling the activat-
ing T cell surface molecule, CD28, to access more of its
ligand. Plausible support for this mechanism comes from
non-obese, diabetic (NOD) mouse models.84,85 However
several studies have found paradoxically increased levels
of soluble CTLA‑4 in sera from humans with auto­
immune conditions, suggesting that the complexity of
cell-surface CTLA‑4–CD28 interaction and signaling is
incompletely understood.86,87
Like CTLA4, the PTPN22 gene on chromosome
1p13 encodes a negative regulator of T cell signaling,
tyrosine-protein phosphatase non-receptor type 22.
One particular PTPN22 variant (the T allele of SNP
rs2476601, associated with 1858C>T and Arg620Trp
substitutions at the DNA and amino acid sequence
level, respectively) has been implicated in susceptibility
to rheumatoid arthritis,88 T1DM89 and Graves disease.90
This variant has also been implicated in several genetic
association studies of AAD.90–92 Although LYP has an
established regulatory role in T-cell activation and sig-
naling, the mechanisms underlying the association of
the Arg620Trp variant with autoimmunity were initially
unclear. Some studies demonstrated increased B and
T-cell responses in individuals who are homozygous
REVIEWS
Soluble CTLA-4
CD28
CTLA-4 CD80 and
CD86
PD-1
CD4+ Antigen-
lymphocyte PD-L1 presenting
dendritic cell
MHC class II
Lck presenting
Fyn TCRα peptide
LYP
TCRβ
Figure 4 | Molecular interactions between antigen-presenting cells and CD4+
lymphocytes that are involved in autoimmunity. The dendritic cell presents a
peptide antigen, bound to an MHC class II molecule, to the CD4+ lymphocyte. The
peptide‑MHC complex is recognized by the T cell receptor complex. A co-
stimulatory signal, delivered via CD80 or CD86, is necessary for lymphocyte
activation. CD28 acts as a positive co-stimulator and CTLA‑4 is a negative co-
stimulator in this context. Soluble CTLA‑4 regulates lymphocytic activation by
competing with CD28 for binding to CD80 or CD86. The interaction between PD‑L1
and PD‑CD1 interaction inhibits T cell activation, as does the action of LYP to
dephosphorylate Fyn and Lck. Abbreviations: CTLA‑4, cytotoxic T lymphocyte
antigen‑4; Fyn, tyrosine-protein kinases; Lck, tyrosine-protein kinases; LYP,
lymphoid tyrosine-protein phosphatase non-receptor type 22; PD‑1, programmed
cell death protein 1; PD‑L1, programmed cell death 1 ligand 1; TCR, T-cell receptor.
for the variant, but other studies found the opposite
effect.93–95 A 2011 study showed that Arg620Trp is a
loss-of-function variant in both humans and mice,
resulting in unstable LYP that is more susceptible to
proteosome-mediated degradation. Levels of LYP are
reduced by the Arg620Trp variant and, therefore, its
inhibitory effects on T-cell signaling and activation are
diminished, which creates an environment that favours
auto­immunity.96 CD274 (which encodes programmed
cell death 1 ligand 1) on chromosome 9p24 is another
example of a gene that encodes a co-stimulatory mol-
ecule. Allelic variants of CD274 have been associated
with autoimmune conditions, including AAD.97–99
B lymphocytes, apart from being responsible for the
generation of antibodies and the humoral immune
response, also act as antigen-presenting cells, and
express cell-surface markers that regulate T cell signal-
ing. Fc receptor-like family members are B cell surface
receptors of the immunoglobulin receptor superfamily
encoded in the FcRL gene locus on chromosome 1q21,
and these genes have been implicated in suscepti­bility
to several autoimmune diseases. 100–103 For example,
FcRL3 is expressed predominantly on B lymphocytes,
but also on natural killer cells and regulatory T cells,
and a promoter polymorphism in FCRL3 has been
widely associ­ated with rheumatoid arthritis and other
systemic autoimmune disorders, particularly in Asian
populations. 101,102 A study of the FCRL3 locus in a
cohort of patients with AAD from the UK found that
the strongest association occurred with the T allele of
the SNP rs7528684 in the FCRL3 promoter (termed
FCRL3_3*T). This allele is predicted to have reduced
312  |  MAY 2012  |  VOLUME 8  www.nature.com/nrendo
© 2012 Macmillan Publishers Limited. All rights reserved
promoter activity.104 Interestingly, a haplotype consist-
ing of seven FCRL3 alleles, including FCRL3_3*T, was
associated with protection from multiple autoimmune
conditions in a Japanese cohort, but seemed to be a sus-
ceptibility haplotype for AAD in white Europeans. This
contradiction could reflect variations between popula-
tions, or differences in the pathogenesis of AAD—as
an organ-specific autoimmune disorder—versus that
of other autoimmune conditions. However, this appar-
ent discrepancy highlights the complexity of genetic
analysis in autoimmunity.
Innate immunity
Loci that confer susceptibility to AAD through vari-
ants in innate immune pathways are now emerging,
including NLRP1, CLEC16A, VDR and CYP27B1. The
innate immune system is concerned with the imme-
diate defense of the host from infection by patho-
gens in a generic, nonspecific manner and comprises
a number of cell types and mechanisms. Its primary
functions include pathogen recognition, immune-cell
recruitment through the production of chemokines
and cytokines, activation of the complement cascade
and activation of the adaptive immune system through
antigen presentation. Cytoplasmic pattern-recognition
receptors, such as the NOD-like receptors (NLRs), are
an innate mechanism for sensing microbial products.
These proteins contribute to the formation of inflam-
masomes, which activate proinflammatory cytokines
and trigger the adaptive immune response. Interest in
the role of NLRs in autoimmunity increased when an
association between the NOD2 (formerly CARD15)
locus and susceptibility to Crohn disease was identi-
fied.105 NLR defects might, therefore, result in a reduced
capacity to differentiate self antigens from nonself
antigens. Subsequently, polymorphisms in various
NLR genes have also been associated with other organ-­
specific autoimmune diseases, including vitiligo106 and
T1DM.107 In Norwegian patients, Leu155His coding
variant of NLRP1 (formerly known as NALP1) demon-
strated a statistically significant association with AAD107
and this finding was replicated in a Polish cohort. 108
NLRP1 is thought to have a key role in inflammasome
assembly and activates proinflammatory cytokines,
including IL‑1β, but the effect of this polymorphism
on NLRP1 function has yet to be demonstrated and
requires further study.
The CLEC16A (C-type lectin domain family 16,
member A) gene, which lies in close proximity to the
CIITA gene, encodes a protein of unknown function.
However, it is expressed on a number of professional
antigen-presenting cells, such as dendritic and natural
killer cells.109 Moreover, this protein possesses a C‑type
lectin-binding domain, indicating that it could be a cell-
surface receptor. Polymorphisms in CLEC16A have been
implicated in a number of autoimmune conditions,110–113
including T1DM.109,114 In a Norwegian study, an intronic
SNP (rs12917716) in CLEC16A was associated with
AAD, although this finding was not replicated in a UK
cohort of patients with AAD.71

Vitamin D has a suppressive effect on immunity
and autoimmunity; supplementation with vitamin D
prevents the onset of autoimmune diabetes in the
­d iabetes-prone NOD mouse 115 and vitamin D defi-
ciency has been associated with human autoimmune
disease including T1DM.116 Polymorphisms in both the
vitamin D receptor (VDR) gene on chromosome 12q12
and the CYP27B1 gene on chromosome 12q13 (which
encodes a cytochrome P450 1α-hydroxylase enzyme
that catalyses the conversion of 25-hydroxyvitamin D3
to the active 1,25-dihydroxyvitamin D 3) have been
associated with AAD. A small study in Germany found
evidence of association between VDR genotypes and
AAD; however, no individual VDR allele was signifi-
cantly associated with this disease.117 An independent,
larger cohort study is required to confirm VDR as a sus-
ceptibility locus in AAD. By contrast, convincing evi-
dence indicates that a promoter polymorphism in the
CYP27B1 gene (–1260C>A) is associated with T1DM,
Graves disease, Hashimoto thyroiditis and AAD. This
association was initially identified in a German AAD
cohort 118 and subsequently confirmed in a UK study.119
Hypothetical basis for AAD progression
As outlined above, there are a series of phases in AAD
development, from steroid 21-hydroxylase antibody
positivity, through compensated, pre-clinical adrenal
insufficiency (with raised plasma renin and ACTH)
to overt, symptomatic adrenal failure. The triggers
or factors that influence progression from one phase
to the next in AAD are not yet clear. It has been sug-
gested that the zona fasciculata is protected from
lympho­c ytic infiltration for a longer period than the
zona glomerulosa by high concentrations of locally
produced cortisol.120 We envisage a more generalized
role for the high intra-adrenal levels of glucocorticoid,
in that continuous local steroidogenesis might consti-
tute a milieu that exerts a natural immunosuppressive
effect, protecting the healthy adrenal gland from auto­
reactivity. In support of this hypothesis, cortisol levels
in the healthy adrenal cortex are in the 20–50 μmol/l
range, 121 and these concentrations are sufficient to
inhibit antigen-presenting cell function, 122 and to
induce a state of suppressed T lympho­c yte function
in vitro. 123–127 Moreover, focal areas of lymphocyte
infiltration have been found at autopsy in the adrenal
cortex of 7% of individuals who are under the age of
50 years, and their prevalence increases with age. 128 As
this proportion is far larger than that of people who
develop AAD, we can assume these infiltrates are tran-
sient or of no pathogenic importance in the majority
of individuals. However, it is possible that a key step
in the patho­genesis of AAD could be a local reduction
in gluco­corticoid concentrations at the site of a focal
lympho­c ytic infiltrate, which reduces immune toler-
ance to adrenal antigens and leads to lymphocyte acti-
vation (Figure 1). Steroidogenesis might then become
further impaired by augmentation of the immune cell
infiltrate, resulting in a vicious cycle of increasing
adreno­cyte damage and decreasing steroid production.
NATURE REVIEWS | ENDOCRINOLOGY VOLUME 8  |  MAY 2012  |  313
© 2012 Macmillan Publishers Limited. All rights reserved
REVIEWS
In this way, we can speculate that progressive and sys-
temic autoreactivity to adrenal antigens might be ini-
tiated and eventually lead to complete adrenal failure.
A similar mechanism might contribute to the high
rate of adrenal metastases observed in patients with
cancer, in that high concentrations of glucocorticoids
in the adrenal cortex might attenuate the presentation
of tumor antigens and, therefore, perturb immune sur-
veillance. If adreno­cortical steroidogenesis does indeed
protect the healthy adrenal gland from the develop-
ment of autoimmunity, this mechanism might help to
explain why AAD is a rare condition compared to other
autoimmune endocrino­p athies. A laboratory animal
model that more closely resembles AAD would greatly
aid our understanding of disease progression.
Conclusions
AAD is a condition that results from the interaction
of currently unknown environmental factors with
incompletely defined genetic factors in a susceptible
indivi­dual. The genetic etiology of AAD has been a
subject of research for more than three decades. The
number of susceptibility loci known to contribute to
this disease is gradually increasing, but those discov-
ered so far make only a modest contribution to disease
susceptibility. Further research is needed for a major
breakthrough in our understanding of the etiology of
AAD. Most genetic studies of AAD have been relatively
small and based on candidate-gene selection. Although
this hypothesis-driven approach has yielded some
successes, the gene or genes that confer most of the
susceptibility to AAD might never be identified using
this method.
Studies of larger cohorts of patients with AAD,
perhaps conducted through international collaboration
between research institutes, are required. In addition to
the candidate-gene approach described above, genome-
wide association studies could offer unique insights
into the genetic etiology of AAD, providing that enough
patients with AAD could be identified to give the study
sufficient statistical power. Whole-genome sequenc-
ing of the DNA of patients with AAD might also yield
useful insights into the genetic architecture of AAD,
again as long as the study was adequately powered,
although this method could be prohibitively expen-
sive. With an increased understanding of the etiology of
AAD and the biological pathways involved, new thera-
peutic targets and diagnostic tools could be developed
that will directly benefit individuals with AAD.
Review criteria
The PubMed database was searched using the terms
“Addison disease”, “autoimmunity”, “adrenal failure”,
“adrenal insufficiency” and “immunity”. The search was
limited to English-language papers only, and all years
of publication were included. Research papers in the
authors’ personal collections, not available online, were
also used.
REVIEWS
1. Kong, M. F. & Jeffcoate, W. Eighty-six cases of
Addison’s disease. Clin. Endocrinol. (Oxf.) 41,
757–761 (1994).
2. Laureti, S., Vecchi, L., Santeusanio, F. &
Falorni, A. Is the prevalence of Addison’s
disease underestimated? J. Clin. Endocrinol.
Metab. 84, 1762 (1999).
3. Løvås, K. & Husebye, E. S. High prevalence and
increasing incidence of Addison’s disease in
western Norway. Clin. Endocrinol. (Oxf.) 56,
787–791 (2002).
4. Addison, T. On The Constitutional And Local
Effects Of Disease Of The Supra-Renal Capsules
(1855).
5. Ten, S., New, M. & Maclaren, N. Clinical review
130: Addison’s disease 2001. J. Clin.
Endocrinol. Metab. 86, 2909–2922 (2001).
6. Nerup, J. Addison’s disease—a review of some
clinical, pathological and immunological
features. Dan Med. Bull. 21, 201–217 (1974).
7. Söderbergh, A. et al. Adrenal autoantibodies
and organ-specific autoimmunity in patients
with Addison’s disease. Clin. Endocrinol. (Oxf.)
45, 453–460 (1996).
8. Zelissen, P. M., Bast, E. J. & Croughs, R. J.
Associated autoimmunity in Addison’s disease.
J. Autoimmun. 8, 121–130 (1995).
9. Lander, E. S. & Schork, N. J. Genetic dissection
of complex traits. Science 265, 2037–2048
(1994).
10. Winqvist, O., Karlsson, F. A. & Kampe, O.
21-Hydroxylase, a major autoantigen in
idiopathic Addison’s disease. Lancet 339,
1559–1562 (1992).
11. Nikoshkov, A. et al. A conformation-dependent
epitope in Addison’s disease and other
endocrinological autoimmune diseases maps to
a carboxyl-terminal functional domain of human
steroid 21-hydroxylase. J. Immunol. 162,
2422–2426 (1999).
12. Nikfarjam, L. et al. Mechanism of inhibition of
cytochrome P450 C21 enzyme activity by
autoantibodies from patients with Addison’s
disease. Eur. J. Endocrinol. 152, 95–101
(2005).
13. Furmaniak, J. et al. Autoimmune Addison’s
disease—evidence for a role of steroid
21-hydroxylase autoantibodies in adrenal
insufficiency. J. Clin. Endocrinol. Metab. 79,
1517–1521 (1994).
14. Boscaro, M. et al. Hormonal responses during
various phases of autoimmune adrenal failure:
no evidence for 21-hydroxylase enzyme activity
inhibition in vivo. J. Clin. Endocrinol. Metab. 81,
2801–2804 (1996).
15. Krohn, K., Uibo, R., Aavik, E., Peterson, P. &
Savilahti, K. Identification by molecular cloning
of an autoantigen associated with Addison’s
disease as steroid 17 alpha-hydroxylase. Lancet
339, 770–773 (1992).
16. Winqvist, O., Gustafsson, J., Rorsman, F.,
Karlsson, F. A. & Kampe, O. Two different
cytochrome P450 enzymes are the adrenal
antigens in autoimmune polyendocrine
syndrome type I and Addison’s disease. J. Clin.
Invest. 92, 2377–2385 (1993).
17. Betterle, C. et al. The natural history of adrenal
function in autoimmune patients with adrenal
autoantibodies. J. Endocrinol. 117, 467–475
(1988).
18. Coco, G. et al. Estimated risk for developing
autoimmune Addison’s disease in patients with
adrenal cortex autoantibodies. J. Clin.
Endocrinol. Metab. 91, 1637–1645 (2006).
19. Torrejón, S., Webb, S. M., Rodriguez-Espinosa, J.,
Martinez de Osaba, M. J. & Corcoy, R. Long-
lasting subclinical Addison’s disease. Exp. Clin.
Endocrinol. Diabetes 115, 530–532 (2007).
314  |  MAY 2012  |  VOLUME 8  www.nature.com/nrendo
20. Smans, L. C. & Zelissen, P. M. Partial recovery
of adrenal function in a patient with
autoimmune Addison’s disease. J. Endocrinol.
Invest. 31, 672–674 (2008).
21. De Bellis, A. et al. Remission of subclinical
adrenocortical failure in subjects with adrenal
autoantibodies. J. Clin. Endocrinol. Metab. 76,
1002–1007 (1993).
22. De Bellis, A. A. et al. Time course of
21-hydroxylase antibodies and long-term
remission of subclinical autoimmune
adrenalitis after corticosteroid therapy:
case report. J. Clin. Endocrinol. Metab. 86,
675–678 (2001).
23. Finnish-German APECED Consortium. An
autoimmune disease, APECED, caused by
mutations in a novel gene featuring two PHD-
type zinc-finger domains. Nat. Genet. 17,
399–403 (1997).
24. Nagamine, K. et al. Positional cloning of the
APECED gene. Nat. Genet. 17, 393–398
(1997).
25. Pearce, S. H. et al. A common and recurrent
13‑bp deletion in the autoimmune regulator
gene in British kindreds with autoimmune
polyendocrinopathy type 1. Am. J. Hum. Genet.
63, 1675–1684 (1998).
26. Pearce, S. H. & Cheetham, T. D. Autoimmune
polyendocrinopathy syndrome type 1: treat with
kid gloves. Clin. Endocrinol. (Oxf.) 54, 433–435
(2001).
27. Zlotogora, J. & Shapiro, M. S. Polyglandular
autoimmune syndrome type I among Iranian
Jews. J. Med. Genet. 29, 824–826 (1992).
28. Rosatelli, M. C. et al. A common mutation in
Sardinian autoimmune polyendocrinopathy‑
candidiasis‑ectodermal dystrophy patients.
Hum. Genet. 103, 428–434 (1998).
29. Perheentupa, J. Autoimmune
polyendocrinopathy‑candidiasis‑ectodermal
dystrophy. J. Clin. Endocrinol. Metab. 91,
2843–2850 (2006).
30. Wolff, A. S. et al. Autoimmune polyendocrine
syndrome type 1 in Norway: phenotypic
variation, autoantibodies, and novel mutations
in the autoimmune regulator gene. J. Clin.
Endocrinol. Metab. 92, 595–603 (2007).
31. Neufeld, M., Maclaren, N. K. & Blizzard, R. M.
Two types of autoimmune Addison’s disease
associated with different polyglandular
autoimmune (PGA) syndromes. Medicine
(Baltimore) 60, 355–362 (1981).
32. Ahonen, P., Myllarniemi, S., Sipila, I. &
Perheentupa, J. Clinical variation of
autoimmune polyendocrinopathy‑
candidiasis‑ectodermal dystrophy (APECED) in a
series of 68 patients. N. Engl. J. Med. 322,
1829–1836 (1990).
33. Betterle, C., Greggio, N. A. & Volpato, M. Clinical
review 93: Autoimmune polyglandular syndrome
type 1. J. Clin. Endocrinol. Metab. 83,
1049–1055 (1998).
34. Akirav, E. M., Ruddle, N. H. & Herold, K. C. The
role of AIRE in human autoimmune disease.
Nat. Rev. Endocrinol. 7, 25–33 (2011).
35. Betterle, C., Volpato, M., Greggio, A. N. &
Presotto, F. Type 2 polyglandular autoimmune
disease (Schmidt’s syndrome). J. Pediatr.
Endocrinol. Metab. 9 (Suppl. 1), 113–123 (1996).
36. Dittmar, M. & Kahaly, G. J. Polyglandular
autoimmune syndromes: immunogenetics and
long-term follow-up. J. Clin. Endocrinol. Metab.
88, 2983–2992 (2003).
37. Myhre, A. G. et al. Autoimmune adrenocortical
failure in Norway autoantibodies and human
leukocyte antigen class II associations related
to clinical features. J. Clin. Endocrinol. Metab.
87, 618–623 (2002).
© 2012 Macmillan Publishers Limited. All rights reserved
38. Drexhage, H. A. Autoimmune endocrinopathies
(ed. Volpe, R.) 309–336 (Humana Press,
1999).
39. Greco, D. S. Hypoadrenocorticism in small
animals. Clin. Tech. Small Anim. Pract. 22, 32–35
(2007).
40. Peterson, M. E., Kintzer, P. P. & Kass, P. H.
Pretreatment clinical and laboratory findings in
dogs with hypoadrenocorticism: 225 cases
(1979–1993). J. Am. Vet. Med. Assoc. 208,
85–91 (1996).
41. Famula, T. R., Belanger, J. M. & Oberbauer, A. M.
Heritability and complex segregation analysis of
hypoadrenocorticism in the standard poodle.
J. Small Anim. Pract 44, 8–12 (2003).
42. Oberbauer, A. M. et al. Inheritance of
hypoadrenocorticism in bearded collies. Am. J.
Vet. Res. 63, 643–647 (2002).
43. Pedersen, N. C. A review of immunologic
diseases of the dog. Vet. Immunol.
Immunopathol. 69, 251–342 (1999).
44. Smallwood, L. J. & Barsanti, J. A.
Hypoadrenocorticism in a family of leonbergers.
J. Am. Anim. Hosp. Assoc. 31, 301–305 (1995).
45. Chase, K., Sargan, D., Miller, K., Ostrander, E. A.
& Lark, K. G. Understanding the genetics of
autoimmune disease: two loci that regulate late
onset Addison’s disease in Portuguese Water
Dogs. Int. J. Immunogenet. 33, 179–184 (2006).
46. Hughes, A. M., Jokinen, P., Bannasch, D. L.,
Lohi, H. & Oberbauer, A. M. Association of a dog
leukocyte antigen class II haplotype with
hypoadrenocorticism in Nova Scotia Duck Tolling
Retrievers. Tissue Antigens 75, 684–690 (2010).
47. Bowen, D., Schaer, M. & Riley, W. Autoimmune
polyglandular syndrome in a dog: a case report.
J. Am. Anim. Hosp. Assoc. 22, 649–654 (1985).
48. Heggarty, H. Addison’s disease in identical
twins. Br. Med. J. 1, 559 (1968).
49. Simmonds, J. P. & Lister, J. Auto-immune
Addison’s disease in identical twins. Postgrad.
Med. J. 54, 552–554 (1978).
50. Smith, M. E., Gough, J. & Galpin, O. P. Addison’s
disease in identical twins. Br. Med. J. 2, 1316
(1963).
51. Russell, G. A., Coulter, J. B., Isherwood, D. M.,
Diver, M. J. & Smith, D. S. Autoimmune
Addison’s disease and thyrotoxic thyroiditis
presenting as encephalopathy in twins. Arch. Dis.
Child. 66, 350–352 (1991).
52. Fairchild, R. S., Schimke, R. N. & Abdou, N. I.
Immunoregulation abnormalities in familial
Addison’s disease. J. Clin. Endocrinol. Metab. 51,
1074–1077 (1980).
53. Hewitt, P. H. Addison’s disease occurring in
sisters. Br. Med. J. 2, 1530–1531 (1957).
54. Vyse, T. J. & Todd, J. A. Genetic analysis of
autoimmune disease. Cell 85, 311–318 (1996).
55. Risch, N. Assessing the role of HLA-linked and
unlinked determinants of disease. Am. J. Hum.
Genet. 40, 1–14 (1987).
56. Vaidya, B., Kendall-Taylor, P. & Pearce, S. H. The
genetics of autoimmune thyroid disease. J. Clin.
Endocrinol. Metab. 87, 5385–5397 (2002).
57. Hemminki, K., Li, X., Sundquist, J. &
Sundquist, K. Familial association between
type 1 diabetes and other autoimmune and
related diseases. Diabetologia 52, 1820–1828
(2009).
58. Maclaren, N. K. & Riley, W. J. Inherited
susceptibility to autoimmune Addison’s disease
is linked to human leukocyte antigens-DR3 and/
or DR4, except when associated with type I
autoimmune polyglandular syndrome. J. Clin.
Endocrinol. Metab. 62, 455–459 (1986).
59. Boehm, B. O. et al. The HLA-DQ beta non‑Asp‑57
allele: a predictor of future insulin-dependent
diabetes mellitus in patients with autoimmune

Addison’s disease. Tissue Antigens 37, 130–132
(1991).
60. Huang, W. et al. Although DR3-DQB1*0201 may
be associated with multiple component
diseases of the autoimmune polyglandular
syndromes, the human leukocyte antigen DR4-
DQB1*0302 haplotype is implicated only in
beta-cell autoimmunity. J. Clin. Endocrinol. Metab.
81, 2559–2563 (1996).
61. Partanen, J., Peterson, P., Westman, P.,
Aranko, S. & Krohn, K. Major histocompatibility
complex class II and III in Addison’s disease
MHC alleles do not predict autoantibody
specificity and 21-hydroxylase gene
polymorphism has no independent role in
disease susceptibility. Hum. Immunol. 41,
135–140 (1994).
62. Park, Y. S. et al. Additional association of intra-
MHC genes, MICA and D6S273, with Addison’s
disease. Tissue Antigens 60, 155–163 (2002).
63. Yu, L. et al. DRB1*04 and DQ alleles:
expression of 21-hydroxylase autoantibodies
and risk of progression to Addison’s disease.
J. Clin. Endocrinol. Metab. 84, 328–335 (1999).
64. Peterson, P. et al. Steroid 21-hydroxylase gene
polymorphism in Addison’s disease patients.
Tissue Antigens 46, 63–67 (1995).
65. Gambelunghe, G. et al. Microsatellite
polymorphism of the MHC class I chain-related
(MIC‑A and MIC‑B) genes marks the risk for
autoimmune Addison’s disease. J. Clin.
Endocrinol. Metab. 84, 3701–3707 (1999).
66. Triolo, T. M. et al. Homozygosity of the
polymorphism MICA5.1 identifies extreme risk
of progression to overt adrenal insufficiency
among 21-hydroxylase antibody-positive patients
with type 1 diabetes. J. Clin. Endocrinol. Metab.
94, 4517–4523 (2009).
67. Dziembowska, M. et al. Three novel mutations of
the CIITA gene in MHC class II‑deficient patients
with a severe immunodeficiency.
Immunogenetics 53, 821–829 (2002).
68. Koizumi, K. et al. Single nucleotide
polymorphisms in the gene encoding the major
histocompatibility complex class II transactivator
(CIITA) in systemic lupus erythematosus. Ann.
Rheum. Dis. 64, 947–950 (2005).
69. Eyre, S. et al. Investigation of the MHC2TA gene,
associated with rheumatoid arthritis in a
Swedish population, in a UK rheumatoid arthritis
cohort. Arthritis Rheum. 54, 3417–3422
(2006).
70. Ghaderi, M. et al. MHC2TA single nucleotide
polymorphism and genetic risk for autoimmune
adrenal insufficiency. J. Clin. Endocrinol. Metab.
91, 4107–4111 (2006).
71. Skinningsrud, B. et al. Polymorphisms in
CLEC16A and CIITA at 16p13 are associated
with primary adrenal insufficiency. J. Clin.
Endocrinol. Metab. 93, 3310–3317 (2008).
72. Brunet, J. F. et al. A new member of the
immunoglobulin superfamily-‑CTLA‑4. Nature
328, 267–270 (1987).
73. Akamizu, T. et al. Association of autoimmune
thyroid disease with microsatellite markers for
the thyrotropin receptor gene and CTLA‑4 in
Japanese patients. Thyroid 10, 851–858
(2000).
74. Kotsa, K., Watson, P. F. & Weetman, A. P.
A CTLA‑4 gene polymorphism is associated with
both Graves disease and autoimmune
hypothyroidism. Clin. Endocrinol. (Oxf.) 46,
551–554 (1997).
75. Sale, M. M. et al. Association of autoimmune
thyroid disease with a microsatellite marker for
the thyrotropin receptor gene and CTLA‑4 in a
Japanese population. Proc. Assoc. Am. Physicians
109, 453–461 (1997).
NATURE REVIEWS | ENDOCRINOLOGY VOLUME 8  |  MAY 2012  |  315
76. Yanagawa, T., Hidaka, Y., Guimaraes, V.,
Soliman, M. & DeGroot, L. J. CTLA‑4 gene
polymorphism associated with Graves’ disease
in a Caucasian population. J. Clin. Endocrinol.
Metab. 80, 41–45 (1995).
77. Marron, M. P. et al. Insulin-dependent diabetes
mellitus (IDDM) is associated with CTLA4
polymorphisms in multiple ethnic groups. Hum.
Mol. Genet. 6, 1275–1282 (1997).
78. Nistico, L. et al. The CTLA‑4 gene region of
chromosome 2q33 is linked to, and associated
with, type 1 diabetes. Belgian Diabetes Registry.
Hum. Mol. Genet. 5, 1075–1080 (1996).
79. Seidl, C. et al. CTLA4 codon 17 dimorphism in
patients with rheumatoid arthritis. Tissue
Antigens 51, 62–66 (1998).
80. Vaidya, B. et al. An association between the
CTLA4 exon 1 polymorphism and early
rheumatoid arthritis with autoimmune
endocrinopathies. Rheumatology (Oxford) 41,
180–183 (2002).
81. Vaidya, B. et al. Association analysis of the
cytotoxic T lymphocyte antigen‑4 (CTLA‑4) and
autoimmune regulator‑1 (AIRE‑1) genes in
sporadic autoimmune Addison’s disease. J. Clin.
Endocrinol. Metab. 85, 688–691 (2000).
82. Kemp, E. H. et al. A cytotoxic T lymphocyte
antigen‑4 (CTLA‑4) gene polymorphism is
associated with autoimmune Addison’s disease
in English patients. Clin. Endocrinol. (Oxf.) 49,
609–613 (1998).
83. Blomhoff, A. et al. Polymorphisms in the
cytotoxic T lymphocyte antigen‑4 gene region
confer susceptibility to Addison’s disease. J. Clin.
Endocrinol. Metab. 89, 3474–3476 (2004).
84. Gerold, K. D. et al. The soluble ctla‑4 splice
variant protects from type 1 diabetes and
potentiates regulatory T‑cell function. Diabetes
60, 1955–1963 (2011).
85. Ueda, H. et al. Association of the T‑cell regulatory
gene CTLA4 with susceptibility to autoimmune
disease. Nature 423, 506–511 (2003).
86. Daroszewski, J. et al. Soluble CTLA‑4 receptor an
immunological marker of Graves’ disease and
severity of ophthalmopathy is associated with
CTLA‑4 Jo31 and CT60 gene polymorphisms.
Eur. J. Endocrinol. 161, 787–793 (2009).
87. Oaks, M. K. & Hallett, K. M. Cutting edge:
a soluble form of CTLA‑4 in patients with
autoimmune thyroid disease. J. Immunol. 164,
5015–5018 (2000).
88. Begovich, A. B. et al. A missense single-
nucleotide polymorphism in a gene encoding a
protein tyrosine phosphatase (PTPN22) is
associated with rheumatoid arthritis. Am. J.
Hum. Genet. 75, 330–337 (2004).
89. Kahles, H. et al. Sex-specific association of
PTPN22 1858T with type 1 diabetes but not with
Hashimoto’s thyroiditis or Addison’s disease in
the German population. Eur. J. Endocrinol. 153,
895–899 (2005).
90. Velaga, M. R. et al. The codon 620 tryptophan
allele of the lymphoid tyrosine phosphatase
(LYP) gene is a major determinant of Graves’
disease. J. Clin. Endocrinol. Metab. 89,
5862–5865 (2004).
91. Skinningsrud, B. et al. Mutation screening of
PTPN22: association of the 1858T-allele with
Addison’s disease. Eur. J. Hum. Genet. 16,
977–982 (2008).
92. Roycroft, M. et al. The tryptophan 620 allele of
the lymphoid tyrosine phosphatase (PTPN22)
gene predisposes to autoimmune Addison’s
disease. Clin. Endocrinol. (Oxf.) 70, 358–362
(2009).
93. Vang, T. et al. Autoimmune-associated lymphoid
tyrosine phosphatase is a gain‑of‑function
variant. Nat. Genet. 37, 1317–1319 (2005).
© 2012 Macmillan Publishers Limited. All rights reserved
REVIEWS
94. Rieck, M. et al. Genetic variation in PTPN22
corresponds to altered function of T and
B lymphocytes. J. Immunol. 179, 4704–4710
(2007).
95. Arechiga, A. F. et al. Cutting edge: the PTPN22
allelic variant associated with autoimmunity
impairs B cell signaling. J. Immunol. 182,
3343–3347 (2009).
96. Zhang, J. et al.The autoimmune disease-
associated PTPN22 variant promotes calpain-
mediated Lyp/Pep degradation associated with
lymphocyte and dendritic cell
hyperresponsiveness. Nat. Genet. 43, 902–907
(2011).
97. Brown, J. A. et al. Blockade of programmed
death‑1 ligands on dendritic cells enhances
T cell activation and cytokine production.
J. Immunol. 170, 1257–1266 (2003).
98. Hayashi, M., Kouki, T., Takasu, N., Sunagawa, S.
& Komiya, I. Association of an A/C single
nucleotide polymorphism in programmed cell
death-ligand 1 gene with Graves’ disease in
Japanese patients. Eur. J. Endocrinol. 158,
817–822 (2008).
99. Mitchell, A. L. et al. Programmed death ligand 1
(PD‑L1) gene variants contribute to autoimmune
Addison’s disease and Graves’ disease
susceptibility. J. Clin. Endocrinol. Metab. 94,
5139–5145 (2009).
100. Capon, F. et al. Fine mapping of the PSORS4
psoriasis susceptibility region on
chromosome 1q21. J. Invest. Dermatol. 116,
728–730 (2001).
101. Kochi, Y. et al. A functional variant in FCRL3,
encoding Fc receptor-like 3, is associated with
rheumatoid arthritis and several
autoimmunities. Nat. Genet. 37, 478–485
(2005).
102. Kyogoku, C. et al. Fcgamma receptor gene
polymorphisms in Japanese patients with
systemic lupus erythematosus: contribution of
FCGR2B to genetic susceptibility. Arthritis
Rheum. 46, 1242–1254 (2002).
103. Ehrhardt, G. R. et al. Fc receptor-like proteins
(FCRL): immunomodulators of B cell function.
Adv. Exp. Med. Biol. 596, 155–162 (2007).
104. Owen, C. J. et al. Analysis of the Fc
receptor‑like‑3 (FCRL3) locus in Caucasians with
autoimmune disorders suggests a complex
pattern of disease association. J. Clin.
Endocrinol. Metab. 92, 1106–1111 (2007).
105. Cummings, J. R. et al. The genetics of NOD-like
receptors in Crohn’s disease. Tissue Antigens
76, 48–56 (2010).
106. Jin, Y., Birlea, S. A., Fain, P. R. & Spritz, R. A.
Genetic variations in NALP1 are associated with
generalized vitiligo in a Romanian population.
J. Invest. Dermatol. 127, 2558–2562 (2007).
107. Magitta, N. F. et al. A coding polymorphism in
NALP1 confers risk for autoimmune Addison’s
disease and type 1 diabetes. Genes Immun. 10,
120–124 (2009).
108. Zurawek, M. et al. A coding variant in NLRP1 is
associated with autoimmune Addison’s disease.
Hum. Immunol. 71, 530–534 (2010).
109. Hakonarson, H. et al. A genome-wide association
study identifies KIAA0350 as a type 1 diabetes
gene. Nature 448, 591–594 (2007).
110. Zoledziewska, M. et al. Variation within the
CLEC16A gene shows consistent disease
association with both multiple sclerosis and
type 1 diabetes in Sardinia. Genes Immun. 10,
15–17 (2009).
111. Dubois, P. C. et al. Multiple common variants for
celiac disease influencing immune gene
expression. Nat. Genet. 42, 295–302 (2010).
112. Marquez, A. et al. Specific association of a
CLEC16A/KIAA0350 polymorphism with
REVIEWS
NOD2/CARD15– Crohn’s disease patients. Eur. J.
Hum. Genet. 17, 1304–1308 (2009).
113. Skinningsrud, B. et al. A CLEC16A variant
confers risk for juvenile idiopathic arthritis and
anti-CCP negative rheumatoid arthritis. Ann.
Rheum. Dis. 69, 1471–1474 (2009).
114. Todd, J. A. et al. Robust associations of four new
chromosome regions from genome-wide
analyses of type 1 diabetes. Nat. Genet. 39,
857–864 (2007).
115. Mathieu, C., Waer, M., Laureys, J., Rutgeerts, O.
& Bouillon, R. Prevention of autoimmune
diabetes in NOD mice by 1,25
dihydroxyvitamin D3. Diabetologia 37, 552–558
(1994).
116. Zella, J. B. & DeLuca, H. F. Vitamin D and
autoimmune diabetes. J. Cell Biochem. 88,
216–222 (2003).
117. Pani, M. A. et al. A polymorphism within the
vitamin D‑binding protein gene is associated
with Graves’ disease but not with Hashimoto’s
thyroiditis. J. Clin. Endocrinol. Metab. 87,
2564–2567 (2002).
118. Lopez, E. R. et al. A promoter polymorphism of
the CYP27B1 gene is associated with Addison’s
disease, Hashimoto’s thyroiditis, Graves’
disease and type 1 diabetes mellitus in
Germans. Eur. J. Endocrinol. 151, 193–197
(2004).
119. Jennings, C. E., Owen, C. J., Wilson, V. &
Pearce, S. H. A haplotype of the CYP27B1
promoter is associated with autoimmune
Addison’s disease but not with Graves’ disease
in a UK population. J. Mol. Endocrinol. 34,
859–863 (2005).
120. Betterle, C., Dal Pra, C., Mantero, F. &
Zanchetta, R. Autoimmune adrenal insufficiency
and autoimmune polyendocrine syndromes:
autoantibodies, autoantigens, and their
316  |  MAY 2012  |  VOLUME 8  www.nature.com/nrendo
applicability in diagnosis and disease prediction.
Endocr. Rev. 23, 327–364 (2002).
121. Dickerman, Z., Grant, D. R., Faiman, C. &
Winter, J. S. Intraadrenal steroid concentrations
in man: zonal differences and developmental
changes. J. Clin. Endocrinol. Metab. 59,
1031–1036 (1984).
122. Rozkova, D., Horvath, R., Bartunkova, J. &
Spisek, R. Glucocorticoids severely impair
differentiation and antigen presenting function
of dendritic cells despite upregulation of Toll-like
receptors. Clin. Immunol. 120, 260–271 (2006).
123. Franchimont, D. et al. Inhibition of Th1 immune
response by glucocorticoids: dexamethasone
selectively inhibits IL‑12‑induced Stat4
phosphorylation in T lymphocytes. J. Immunol.
164, 1768–1774 (2000).
124. Abe, M. & Thomson, A. W. Dexamethasone
preferentially suppresses plasmacytoid dendritic
cell differentiation and enhances their apoptotic
death. Clin. Immunol. 118, 300–306 (2006).
125. Cidlowski, J. A. et al. The biochemistry and
molecular biology of glucocorticoid-induced
apoptosis in the immune system. Recent Prog.
Horm. Res. 51, 457–490; discussion 490–491
(1996).
126. Moser, M. et al. Glucocorticoids down-regulate
dendritic cell function in vitro and in vivo. Eur. J.
Immunol. 25, 2818–2824 (1995).
127. Zubiaga, A. M., Munoz, E. & Huber, B. T. IL‑4 and
IL‑2 selectively rescue Th cell subsets from
glucocorticoid-induced apoptosis. J. Immunol.
149, 107–112 (1992).
128. Hayashi, Y., Hiyoshi, T., Takemura, T.,
Kurashima, C. & Hirokawa, K. Focal lymphocytic
infiltration in the adrenal cortex of the elderly:
immunohistological analysis of infiltrating
lymphocytes. Clin. Exp. Immunol. 77, 101–105
(1989).
© 2012 Macmillan Publishers Limited. All rights reserved
129. Vergeer, M. et al. Genetic variant of the
scavenger receptor BI in humans. N. Engl. J. Med.
364, 136–145 (2011).
130. Metherell, L. et al. Mutations in GPX1 and NNT,
endcoding antioxidant defense genes, cause
familial glucocorticoid deficiency [Abstract].
Endocr. Rev. 32 (meeting abstracts), OR22–25
(2011).
131. Gombos, Z. et al. Analysis of extended human
leukocyte antigen haplotype association with
Addison’s disease in three populations. Eur. J.
Endocrinol. 157, 757–761 (2007).
132. Skinningsrud, B. et al. Multiple loci in the HLA
complex are associated with Addison’s disease.
J. Clin. Endocrinol. Metab. 96, E1703–E1708
(2011).
133. Pani, M. A., Seissler, J., Usadel, K. H. &
Badenhoop, K. Vitamin D receptor genotype is
associated with Addison’s disease. Eur. J.
Endocrinol. 147, 635–640 (2002).
Acknowledgments
A. L. Mitchell is in receipt of a clinical research
training fellowship from the Medical Research
Council, UK. S. H. S Pearce’s research work is
supported by European Union Framework 7 grant
201167 to the Euradrenal Consortium and grants
G07017632 & G0900001 from the Medical
Research Council, UK. A. L. Mitchell and
S. H. S. Pearce thank their collaborator, Professor
Eystein Husebye, Dr Martina Ericksen and the group
at Haukeland University Hospital, Bergen, Norway, for
contributing family data on Addison disease gathered
from the Norwegian registry of organ-specific
autoimmune diseases for this Review.
Author contributions
Both authors contributed equally to all aspects of the
article.